Report No. ACS14253 Republic of Haiti Towards Greater Fiscal Sustainability and Equity: A Discussion of Public Finance June 29, 2015 GMFDR Latin America and Caribbean Document of the World Bank . . . Standard Disclaimer: This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. . Copyright Statement: The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. ii Acronyms and Abbreviations AFD Agence Française de Développement (French agency for development) AIDS Acquired Immunodeficiency Syndrome ARI Acute Respiratory Infection ASC Agent de Santé Communautaire (community health worker) BM Banque Mondiale BMPAD Bureau de Monétisation des Programmes d'Aide au Développement (Monetization Office of Development Assistance Programs) BRH Banque de la République d’Haïti (Haiti Central Bank) BSEIPH Bureau du Secrétaire d’Etat à l’Intégration des Personnes Handicapées (Office of the Secretary of State for Integration of Persons with Disabilities) CAL Centre de Santé avec Lit (health centre with bed) CAS Caisse d'Assistance Sociale (Social Assistance Fund) CDB Caribbean Development Bank CHE Catastrophic Health Expenditure CIDA Canadian International Development Agency CNMP Commission Nationale des Marchés Publics (National Procurement Commission) CONATEL Conseil National de Télécommunication (National Telecommunication Regulator) CSAFP Conseil Supérieur de l’Administration et de la Fonction Publique CSCCA Cour Supérieure des Comptes et du Contentieux Administratif (Supreme Court of Accounts and Contentious Administrative Proceedings) CSL Centre de Santé sans Lit (health centre without bed) DAB Directorate of Administration and Budget DAO Dossier d’Appel d’Offres (Bidding Document) DASIP Direction de l'Analyse et du Suivi des Investissements Publics (Public Investment Analysis and Monitoring Unit) DDP Document Définitif du Projet (Final Project Document) DEA Data Envelopment Analysis DEC Direction d’Evaluation et de Contrôle (Evaluation and Control Directorate) DGB Direction Générale du Budget (Budget Directorate) DTDCP Direction du Trésor, Dette, et Comptabilité Publique (Treasury Directorate) DHS Demographic and Health Survey DIP Direction de l’Investissement Public (Directorate of Public Investment) DPC Direction de la Pension Civile (Directorate for Public Pensions) DPES Direction de la Programmation Economique et Sociale (Economic and Social Programming Directorate) DSE Direction du Suivi et de l’Evaluation (Directorate of Monitoring & Evaluation) ECVMAS Enquête sur les Conditions de Vie des Ménages après le Séisme (survey on the living conditions of households after the earthquake – household survey) EMMUS Haiti Mortality, Morbidity, and Service Utilization Survey (DHS in English) EPPLS Entreprise Publique de Production de Logement Social (Public Enterprise for the Promotion of Social Housing) EPSSS Evaluation de la Prestation des Services de Soins de Santé (Service Provision Assessment) EPT Education Pour Tous (Education For All) EU European Union FAES Fonds d'Assistance Economique et Social (Economic and Social Assistance Fund) FER Fonds d’Entretien Routier (Road Maintenance Fund) FIOP Fiche d’Identification et d’Opération des Projets (Identification and Project Operation File) iii FNE Fonds National de l’Education (National Education Fund) FY Fiscal Year GDP Gross Domestic Product GL General Ledger GoH Government of Haiti GPRSP Growth and Poverty Reduction Strategy Paper GSP Groupe Santé Plus (Health group – private company) HCR Hôpital Communautaire de Référence (Community Referral Hospital) HIMO Travaux à Haute Intensité de Main d’Oeuvre (Labor-intensive public works) HIS Health Information System HIV Human Immunodeficiency Virus HRMIS Human Resources Management Information System HTG Haitian Gourdes IADB Inter-American Development Bank IBERS Institut du Bien-Etre Social et de Recherches (Institute of Social Welfare and Research) IBRD International Bank for Reconstruction and Development IDA International Development Association IDB Inter-American Development Bank IGF Inspection Générale des Finances (Inspectorate General of Finances) IHE Institut Haïtien de l’Enfance (Haitian Institute of Childhood) IHSI Institut Haïtien de Statistique et d’Informatique (Haitian Statistics Agency) IMF International Monetary Fund IMR Infant Mortality Rate INTOSAI International Organization of Supreme Audit Institutions KF Kore Fanmi (Family Support – Community Social Worker Initiative) LdF Loi de Finances MARNDR Ministère de l’Agriculture, des Ressources Naturelles et du Développement Rural (Ministry of Agriculture, Natural Resources and Rural Development) MAST Ministère des Affaires Sociales et du Travail (Ministry of Social Affairs and Labor) MCFDF Ministère de la Condition Féminine et des Droits des Femmes (Ministry for Women and Women’s Rights) MDG Millennium Development Goals MEF Ministère de l'Economie et des Finances (Ministry of Economy and Finance) MENFP Ministère de l'Education Nationale et de la Formation Professionnelle (Ministry of National Education and Vocational Training) MICT Ministère de l’Intérieure et des Collectivités Territoriales (Ministry of Interior and Local Government) MIS Management Information Systems MMR Maternal Mortality Ratio MNCH Maternal, Neonatal, and Child Health MoP Ministry of Planning MPCE Ministère du Plan et de la Coopération Extérieure (Ministry of Planning and External Cooperation) MSPP Ministère de la Santé Publique et de la Population (Ministry of Public Health and Population) MTPTC Ministère des Travaux Publics, Transports et Communications (Ministry of Public Works, Transport, and Communication) NGO Non Governmental Organization NHA National Health Account OFATMA Office d'Assurance Accidents du Travail, Maladie et Maternité (Office of Insurance for Work Accidents, Illness and Maternity) OMRH Office des Ressources Humaines (Office of Human Ressources) ONA Office National d’Assurance Vieillesse (National Office for Old Age Insurance ) iv ONART Office National de l’Artisanat (National Handicrafts Office) ONM Office National de la Migration (National Migration Office) PAARP Plan d’Action pour l’Accélération de la Réduction de Pauvreté (Plan for Accelerating the Reduction of Extreme Poverty) PaP Port-au-Prince PARDH Plan d’Action pour le Relèvement et le Développement d’Haïti (Action Plan for Haitian National Recovery and Development) PDS Plan Directeur de Santé (Health Master Plan) PEFA Public Expenditure and Financial Accountability PFM Public Financial Management PIMS Public Investment Management System PIP Programme d’Investissement Public (Public Investment Program) PMS Paquet Minimum de Santé (Minimum Package of Services) PNCS Programme National de Cantines Scolaires (National School Canteen Program) PNS Politique Nationale de Santé (National Health Policy) PSDH Plan Stratégique de Développement d’Haïti (Strategic Plan for the Development of Haiti) PSUGO Programme de Scolarisation Universelle Gratuite et Obligatoire (Free and Compulsory Universal Schooling Program) PTI Plan Triennal d’Investissement (Triennial Investment Plan) RBF Results-Based Financing SAI Supreme Audit Institution SDI Schéma Directeur Informatique (IT Master Plan) SIGMP Système Informatisé de Gestion des Marchés Publics (IT System for Procurement Management) SP Social Protection SSN Social Safety Net SYSCOMPTE Système de gestion des Comptes Courants (Accounting System for Specific Budget Items) SYSDEP Système Informatisé de gestion des Dépenses Publiques (Public Expenditure Management System) SYSPIP Système Informatisé de gestion du Programme d’Investissement Public (Public Investment Management System) TB Tuberculosis THE Total Health Expenditure U5MR Under-five Mortality Rate UAS Unité d’Arrondissement de Santé (District Health Unit) UCE Unité de Contrôle d’Exécution UEP Unité d’Etude et de Programmation (Programming and Analysis Unit) ULCC Unité de Lutte Contre la Corruption (Anti-Corruption Unit) UN United Nations UNCITRAL United Nations Commission on International Trade Law UNDP United Nations Development Program UNICEF United Nations Children’s Fund UPM Unité de Passation de Marché (Procurement Unit) USAID United States Agency for International Development WB World Bank WFP World Food Program WHO World Health Organization WHS World Health Survey v Acknowledgement W e would like to thank the members of the Haiti Country Team, as well as all the partners and stakeholders in Haiti who have contributed to the preparation of this document in a strong collaborative process. We are very grateful for the generosity exhibited in providing us with substantive inputs, knowledge and advice. The Team was led by Raju Jan Singh (Program Leader, LCC8C) and the table below identifies the full list of team members who have contributed their time, effort and expertise, and their affiliations. We wish to thank for their helpful suggestions and insights our peer reviewers: Margaret Grosh (Practice Manager, GSPDR), Sebastian James (Senior Investment Policy Officer, GTCDR), Roland Kpodar (Senior Economist, IMF), David Cal McWilliam (Senior Economist, GMFDR), Patrick Ramanantoanina (Senior Operations Officer, GEDDR), and Anand Rajaram (Practice Leader, GGODR). Mary Barton-Dock (Special Envoy for Haiti, LCC8C) and Miria Pigato (Practice Manager, GMFDR) advised the Team throughout the stages of the process. The Team is also extremely grateful for the cooperation of the Haitian authorities and for the invaluable contributions of senior officials and governmental agencies. We acknowledge in particular the following institutions: the Ministry of Economy and Finance, the Ministry of Education, the Ministry of Planning and External Cooperation, the Ministry of Public Health and Population, the Ministry of Social Affairs, and the Prime Minister’s Office, as well as the Central Bank of Haiti (BRH). The Team wishes to thank all the participants of the numerous workshops where preliminary results have been presented and discussed, in particular Mme Laleau, Head of the Economic Studies (Ministry of Economy and Finance) and Berny Duvalsaint, Secretary of the PER Steering Committee. The Team gratefully acknowledges also the support from donors present in Haiti: AFD, Brazil, Canada, European Union, IADB, IMF, Mexico, USAID, UNDP, and UNICEF. A strong emphasis has been put throughout the preparation of this report on training and capacity building. Special attention has been put on ensuring as much as possible that the various analytical tools and technics used in this document be transferred to our Haitian counterparts, and many of the quantitative work was carried out jointly with Haitian teams. In this regard, the Team would like to thank particularly all the Haitian participants who attended our various workshops on the Tariff Reform Impact Simulation Tool (TRIST), the subsidy simulation toolkit (SUBSIM), and BOOST, as well as Ibrahim El ghandour (GGODR), Olivier Jammes (Consultant, GEDDR), and Paulo Verme (Senior Economist, GPVDR) for running these trainings. The Team is also grateful to a number of colleagues who provided insights, comments and support in the course of the report preparation. They include Lucy Basset (Social Protection Specialist, GSPDR), Malaika Becoulet (Consultant, GTIDR), Ludmilla Buteau (Consultant, GHNDR), Pierre Bonneau (Program Leader, LCC8C), Daniel Boyce (Practice Manager, GGODR), Michelle Keane (Lead County Officer, LCC8C), Donald Mphande (Lead Financial Management Specialist, GCFDR), Deo Ndikumana (Senior Country vi Officer, LCC8C), Prosper Nindorera (Senior Procurement Specialist, GGODR), Maki Noda (Consultant, GSPDR), Elizabeth Ruppert Bulmer (Lead Economist, GCJDR), Paolo Verme (Senior Economist, GPVDR), and Kanae Watanabe (Country Officer, LCC8C). Data in Haiti is particularly challenging. This study built on the recently-completed Poverty Assessment, carried out jointly by the World Bank and the ONPES. The expansion in donor assistance and the availability of concessional financing over the past decade has compounded, however, the fragmentation of fiscal data. In this regard, the Team would like to thank the Gates Foundation financing the BOOST Initiative in Haiti, as well as Massimo Mastruzzi (Senior Economist, GGODR) and Leif Jensen (Senior Public Sector Specialist, GGODR) for assisting the Team in this process. Our work on fuel price subsidies would not have been possible without the financial support of the Energy Sector Management Assistance Program (ESMAP), and the help of Sameer Shukla (Senior Energy Specialist, GEEES) and Rohit Khanna (Practice Manager, GEEES). HAITI PER TEAM GLOBAL PRACTICE/CROSS-CUTTING AREA TEAM MEMBERS Communication Christelle Chapoy, Berdine Edmond Education Melissa Adelman, Juan Baron, Eva Junyen, Axelle Latortue Energy & Extractives Frederic Verdol Governance Emeline Bredy, Mamadou Deme, Ibrahim El ghandour, Eduardo Estrada, Rubens Lacerd, Andy MacDonald, Fabienne Mroczka, Renas Sidahmed, Gerard Verger Haiti CMU Gabrielle Dujour, Nellie Sew Kwan Kan, David Lighton, Raju Jan Singh, Paula White Health, Nutrition & Population Eleonora Cavagnero; Marion Cros; Sunil Rajkumar Macro Economics & Fiscal Management Kassia Antoine, Calvin Djiofack, Evans Jadotte, Naoko Kojo, Jan Loeprick, Julie Lohi, Sandra Milord, Emilio Sacerdoti, Erik von Uexkull, Konstantin Wacker Poverty Facundo Cuevas, Federica Marzo, Aude-Sophie Rodella, Thiago Scot Social Protection Carine Clert, Victoria Strokova, Frieda Vandeninden Social, Urban, Rural and Resilience Michel Matera, Rafael Van der Borght vii Contents ACRONYMS AND ABBREVIATIONS ........................................................................................................................ III ACKNOWLEDGEMENT ........................................................................................................................................... VI EXECUTIVE SUMMARY ........................................................................................................................................XIV COUNTRY PROFILE .............................................................................................................................. 1 Haiti Has Opportunities …. ..............................................................................................................................1 … Hampered by Political Instability and Natural Disasters .............................................................................1 … As Well as Weak Structural Policies and Institutions ...................................................................................3 LIVING WITH TIGHTER BUDGET CONSTRAINTS ................................................................................... 8 Macroeconomic Stability Has Been Maintained … ..........................................................................................8 … With Greater Revenue Mobilization and a Surge in Aid ..............................................................................9 … Allowing Higher Public Spending ...............................................................................................................11 … Despite Widening Deficits, Debt is Expected to Remain Sustainable … ....................................................12 … But Financing Constraints Are Becoming Tighter ......................................................................................15 … While Natural Disasters Would Call for Larger Fiscal Buffers ....................................................................16 PROTECTING PRIORITY SPENDING .................................................................................................... 17 Haiti Has a Vision Calling for Higher Human and Physical Capital … .............................................................17 … Reflected in the Rise in Public Investment … .............................................................................................21 … And Greater Priority Given to Social Sectors … .........................................................................................24 … But Haiti’s Inclusiveness Remains Limited .................................................................................................27 TOWARDS GREATER AND MORE EQUITABLE REVENUE MOBILIZATION ............................................ 29 Despite Some Increase, Revenue Mobilization Remains Low and Regressive… ...........................................29 … With Exemptions and Ill-Designed Tax Brackets Eroding Direct Taxes… ...................................................32 … And Indirect Taxes Hampered by Weak Revenue Administration and Unclear Exemptions ....................38 … But Could Be Bolstered By a Simple VAT ...................................................................................................40 … And Taxes on International Trade Could Be Streamlined ..........................................................................41 GETTING A GREATER GROWTH DIVIDEND FROM PUBLIC INVESTMENT ............................................ 46 High Public Investment Doesn’t Translate into Faster Growth in Haiti… ......................................................46 … Because of an Overly Elaborate and Rarely Respected Regulatory Framework .......................................50 … Little Strategic Guidance to Select Projects ...............................................................................................52 … A Budgeting Process That is Not Followed ................................................................................................56 … Weak Procurement… .................................................................................................................................57 … Fragmented Project Execution and Administration… ................................................................................61 … Limited Physical and Financial Monitoring… .............................................................................................65 … And a Lack of Ex-post Evaluation (Project Auditing and Evaluation) .........................................................68 INCREASING SUSTAINABILITY AND IMPROVING FURTHER HEALTH OUTCOMES ............................... 70 Although Low, Health Indicators Have Improved … ......................................................................................70 … But Disparities Remain … ...........................................................................................................................73 … With Relatively High Health Spending Financed by Donors ......................................................................75 … Access to Health Is Limited Especially For the Poor … ...............................................................................78 … And Public Spending Is Not Geared to Better Services. .............................................................................80 viii KEEPING CHILDREN IN SCHOOL AND IMPROVING EDUCATION OUTCOMES ..................................... 92 Despite Progress, Education Outcomes are Low … .......................................................................................92 … Driven by a Greater Education Supply from the Non-Public Sector … ......................................................96 … While Public Spending Has Been Recently Rising … ..................................................................................99 … Its Composition Is Not Conducive to Better Services … ...........................................................................101 … With Cost Remaining a Main Obstacle to Access… ..................................................................................102 … And Disparities across Regions and Poor Services ...................................................................................103 TOWARDS A MORE EFFECTIVE AND BETTER TARGETED SOCIAL PROTECTION SYSTEM ................... 105 Growing Investments in Social Safety Nets .................................................................................................105 … Needed By A Vulnerable Population … ....................................................................................................109 … But Too Fragmented … ............................................................................................................................113 … Too Small … ..............................................................................................................................................116 … Not Well Targeted … ................................................................................................................................120 … And of Limited Impact .............................................................................................................................124 PREVENTING A RETURN OF FUEL SUBSIDIES ................................................................................... 127 Fuel Price Subsidies Implied a Heavy Burden on the Budget… ...................................................................127 … Are Regressive… .......................................................................................................................................130 … And Could Return. ...................................................................................................................................137 CONCLUDING OBSERVATIONS ........................................................................................................................... 139 REFERENCES....................................................................................................................................................... 142 LIST OF ANNEXES Annex 1 : PIM Process Flowcharts ............................................................................................................................156 Annex 2 : Correlates with Enrollment .......................................................................................................................161 Annex 3 : Correlates with Overage Status ................................................................................................................162 Annex 4 : EDE PEP Programs: A Glossary ..................................................................................................................163 Annex 5 : Price Structure for Petroleum Products (2010/2011) ...............................................................................165 LIST OF BOXES Box 2.1 : Electricité d’Haiti ..........................................................................................................................................13 Box 2.2 : Petrocaribe and Haiti ...................................................................................................................................14 Box 3.1 : Benchmarking the Drivers of Shared Prosperity: An Application to Haiti....................................................20 Box 3.2 : Fiscal Data Challenges ..................................................................................................................................25 Box 4.1 : International Experience with Tax Incentives ..............................................................................................33 Box 4.2 : Removing Nuisance Taxes ............................................................................................................................35 Box 4.3 : The CATT Methodology for Benchmarking Customs Performance .............................................................39 Box 4.4 : The World Bank’s Tariff Reform Impact Simulation Tool (TRIST) ................................................................45 Box 5.1 : Common Features of PIM in Donor-Dependent Countries..........................................................................49 Box 5.2 : Haiti Legislative Framework .........................................................................................................................51 Box 5.3 : The Share of Recurrent Expenditures in Investment Projects .....................................................................55 Box 5.4 : INTOSAI Conditions Supporting Corruption .................................................................................................63 Box 5.5 : Project Completion ......................................................................................................................................69 ix Box 6.1 : Health Care Provision in Haiti ......................................................................................................................81 Box 6.2 : Definitions of Operating Budget Terms .......................................................................................................82 Box 6.3 : Definition of Technical Efficiency .................................................................................................................89 Box 6.4 : More Health for Every Dollar: Results-Based Financing ..............................................................................90 Box 6.5 : The Right Incentives Lead to Measurable Results in Rwanda ......................................................................91 Box 7.1 : Education Provision in Haiti .........................................................................................................................98 Box 8.1 : Methodology Note on Social Protection Spending ....................................................................................117 Box 9.1 : The Transport Sector in Haiti .....................................................................................................................135 LIST OF FIGURES Figure 1.1 : Annual GDP Growth vs. Change in Government, 1971-2013 (percent) ....................................................2 Figure 1.2 : Annual GDP Growth vs. Occurrence of Natural Disaster, 1971-2013 (percent) ........................................2 Figure 1.3 : Obstacles to Growth ..................................................................................................................................4 Figure 1.4 : Change in Governance Indicators, 2004-13 ...............................................................................................6 Figure 1.5 : Logistic Performance Index, 2014 ..............................................................................................................7 Figure 1.6 : Port Tariffs Estimated Cost Per TEU*, 2009 (in US Dollar). ........................................................................7 Figure 1.7 : Electric Power Consumption, 2011 ............................................................................................................7 Figure 2.1 : Inflation, 2004-14.......................................................................................................................................8 Figure 2.2 : Gross International Reserves, 2004-14 (Millions of US Dollars) ................................................................8 Figure 2.3 : Fiscal Revenues, 2004-13 (In Percentage of GDP) ....................................................................................9 Figure 2.4 : General Government – Total Revenues, 2000-12 (In Percentage of GDP) ..............................................10 Figure 2.5 : Public Expenditures by Source of Financing, 2001-13 (In Percentage of GDP) .......................................11 Figure 2.6 : General Government Expenditure - International Comparison 2004-13 ................................................12 Figure 2.7 : Central Government Fiscal Balance, ........................................................................................................12 Figure 2.8 : Current Account Balance, 2004-14 (Percentage of GDP) ........................................................................12 Figure 2.9 : Haiti – Stock of Debt from Petrocaribe, 2009-14 (In Percentage of GDP) ...............................................14 Figure 2.10: External Debt (PV), 2014-35 (Percentage of Exports G&S + Remittances) ..............................................14 Figure 2.11: International Aid, 2008-25 (In Percentage of GDP) .................................................................................15 Figure 2.12: Government Deposits, 2009-14 (In Percentage of GDP) .........................................................................15 Figure 2.13: Petrocaribe Financing, 2008-17 (Millions of US Dollars) .........................................................................15 Figure 3.1 : Extreme Poverty Simulations ...................................................................................................................18 Figure 3.2 : Economic Magnitude of Estimated Parameters ......................................................................................19 Figure 3.3 : Life Expectancy at Birth, 2010 (Years) .....................................................................................................20 Figure 3.4 : Cabinet Changes, 2003 (10 Years Average) .............................................................................................20 Figure 3.5 : Income Effects of Closing the Gap (In Percentage of the Gap) ................................................................21 Figure 3.6 : Current Expenditure, 2005-14 (In Percentage of GDP) ............................................................................23 Figure 3.7 : Operating and Capital Expenses, 2005-14 (In Percentage of GDP)..........................................................24 Figure 3.8 : Changes in Sectoral Composition –..........................................................................................................26 Figure 3.9 : Project Activities Financed by Petrocaribe Funds, 2008-13, (USD Millions) ............................................26 Figure 3.10: Project Activities Financed by Petrocaribe Fund, 2008-13, (Percentage of Total) ..................................26 Figure 3.11: Donors Financing, 2010-12 (Percentage of Total) ...................................................................................27 Figure 3.12: International Comparison - Social Spending, 2013 or latest (In Percentage of GDP) ..............................28 x Figure 4.1 : Fiscal Revenue, 2009-13 (In Percent of GDP) ..........................................................................................30 Figure 4.2 : Tax-to-GDP Ratio......................................................................................................................................30 Figure 4.3 : Ratio Direct To Indirect Taxation, 2009 or 2011 ......................................................................................31 Figure 4.4 : Corporate (left) and Personal (right) Income Tax Rates in Regional Comparison (Percent)....................32 Figure 4.5 : Income Tax Rate Thresholds In Regional Comparison (Per Capita GDP) .................................................36 Figure 4.6 : Hypothetical Income Tax Revenue Collection Based on 2012 Household Survey Results ......................37 Figure 4.7 : Hypothetical Effect of Changes in Income Tax Brackets on Revenue Collection by Income Group ........38 Figure 4.8 : Results of Haiti Customs Assessment: Distance from ..............................................................................39 Figure 4.9 : Tax Revenue Collection around the Introduction of VAT ........................................................................41 Figure 4.10: Applied Protection from Tariffs and Inspection Fees for Four Reform Scenarios ...................................45 Figure 5.1 : Capital Investment (a) and (b) .................................................................................................................47 Figure 5.2 : Public Management Efficiency Index .......................................................................................................48 Figure 5.3 : Public Investment Management Index – PIM (0=Lowest, 4=Highest) .....................................................48 Figure 5.4 : Four Stages of the Project Life Cycle ........................................................................................................50 Figure 5.5 : Required Process for Adding New Projects to the PIP .............................................................................54 Figure 5.6 : Illustrates the Budget Formulation Process for the Preparation of the Investment Budget. ..................57 Figure 5.7 : Contracting Methods, 2012-13 ................................................................................................................58 Figure 5.8 : Complexity in Procurement Processes ....................................................................................................60 Figure 5.9 : PIP Project Execution Processes ..............................................................................................................62 Figure 5.10: Treasury funded Public Investment ........................................................................................................65 Figure 5.11: Cash Flows National Treasury Funded Projects 2012-13 ........................................................................67 Figure 5.12: Monitoring and Evaluation on Projects, in PIP, 2011-13 (in Number of Projects) ..................................69 Figure 6.1 : Maternal Mortality, 2013 (Per 100,000 live birth) ...................................................................................71 Figure 6.2 : Under-5 Child Mortality, 2013 (Per 1,000 live births) ..............................................................................71 Figure 6.3 : Health Service Utilization (In Percent of Population) ..............................................................................71 Figure 6.4 : Changes in Under-5 Child Mortality Rates, 1990-2013 (per 1,000 live births) ........................................72 Figure 6.5 : Maternal Mortality Ratio, 1990-2013 (Per 100,000 live births)...............................................................72 Figure 6.6 : Changes in Life Expectancy at Birth, ........................................................................................................72 Figure 6.7 : Health Service Utilization, 1994-2012 (In Percent of Population) ..........................................................72 Figure 6.8 : Change in Infant Mortality, 2006-12 (Percentage Points) .......................................................................73 Figure 6.9 : Change in Under-5 Child Mortality, 2006-12 (Percentage Points) ...........................................................73 Figure 6.10: Change in Stunting Rate, 2006-12 (Percentage Points) ...........................................................................73 Figure 6.11: Change in Diarrhea Prevalence Rate, 2006-12 (Percentage Points) ........................................................73 Figure 6.12: Child Health Outcome Indicators, 2012 (Per 1,000 Live Births) ..............................................................75 Figure 6.13: Health Expenditure Per Capita, 2012 (Current US$) ................................................................................76 Figure 6.14: Health Expenditure Per Capita PPP, 1995-2012 (Constant 2005 US$) ....................................................76 Figure 6.15: Public Health Expenditure, 2012 (Percent of GDP) .................................................................................76 Figure 6.16: Public Health Spending, 2000-12 (Percent of Government Expenditure) ...............................................76 Figure 6.17: Health Expenditure, 2005-10, (Percentage of Total) ...............................................................................77 Figure 6.18: Health Expenditure, 2011-12 (Percentage of Total) ................................................................................77 Figure 6.19: Health Expenditure by Source, 2004-2013 ..............................................................................................77 Figure 6.20: Barriers to Health Care Access for Women Aged 15-49, 2012 or 2013 (Percentage of Total) ................78 xi Figure 6.21: Enrollment of Populations in Risk-Pooling...............................................................................................78 Figure 6.22: Incidence of Catastrophic Health Expenditures at ..................................................................................80 Figure 6.23: Salaries, Excl. Domestic Investment Budget ............................................................................................82 Figure 6.24: Investment Budget from Public Funds, 2012-13 (Percentage of Total) ..................................................83 Figure 6.25: Utilization of the Different Types of Structures by Area, 2013 ...............................................................83 Figure 6.26: Density of Medical Personnel, 2013 (Per 10,000 Inhabitants) ...............................................................84 Figure 6.27: Density of Inpatient Beds, 2013...............................................................................................................84 Figure 6.28: Density of Medical Personnel by Department – Private for-Profit Sector Excluded ...............................85 Figure 6.29: Number of Beds, exc. Private For-Profit (Per 10,000 Inhabitants) ..........................................................85 Figure 6.30: Density of (a) Medical (b) Paramedical Personnel, and (c) Community Agents by Location, ..................86 Figure 6.31: Technical Efficiency Score, 45 First-Level ................................................................................................87 Figure 6.32: Consultations Per Doctor and Nurse, exc. Private For-Profit Sector, 2013 (Daily Numbers) ..................88 Figure 6.33: Consultations per Medical Staff ..............................................................................................................88 Figure 6.34: International Comparisons-Share of Administrative Personnel In Total Number of Primary Health Facilities (Percentage Total) ..................................................................................................................89 Figure 7.1 : Share of 3-25 Year Olds Currently ............................................................................................................92 Figure 7.2 : Haiti’s Literacy Rate is between LIC and LAC Averages ............................................................................93 Figure 7.3 : Figures (a) and (b) - Gross and Net Enrollment Ratios, 2001-12 .............................................................94 Figure 7.4 : Youth Enrolled in School by Age and .......................................................................................................95 Figure 7.5 : Primary Schools, 1930-2011 ....................................................................................................................97 Figure 7.6 : Donor Financing for Education Sector, 2010-13 ......................................................................................98 Figure 7.7 : MENFP Annual Budget Funded by Domestic Resources, 2005-13 ........................................................100 Figure 7.8 : Student Beneficiaries by Department, 2013 ..........................................................................................100 Figure 7.9 : Public Expenditures on Education .........................................................................................................101 Figure 7.10: Education Budget, 2014 (domestic resources only, percentage of total) .............................................101 Figure 7.11: Education Expenditures by Type, 2012 .................................................................................................103 Figure 8.1 : EDE-PEP SSN Financing, By Source, 2013 (Percentage of Total) ............................................................108 Figure 8.2 : Programs Under EDE PEP .......................................................................................................................108 Figure 8.3 : Chronic Poverty, Multidimensional Deprivation and Transient Poverty, 2012 (Percent of Total) ........110 Figure 8.4 : Poverty Headcount by Age Group, 2012 (In Percent) ............................................................................112 Figure 8.5 : Key Risks, the Life Cycle, and Social Protection in Haiti .........................................................................113 Figure 8.6 : Public Expenditure on Social Protection, 2013 (Percentage of GDP) ....................................................116 Figure 8.7 : Social Safety Net Spending – International Comparison, 2013 or latest (Percent of GDP) ...................119 Figure 8.8 : Distribution of Social Protection Benefits (Benefit Incidence), 2012.....................................................120 Figure 8.9 : Access to Social Security by Income Group, 2012 (Percentage of Total) ...............................................121 Figure 8.10: Coverage of Social Assistance Programs and Distribution of Beneficiaries ...........................................122 Figure 8.11: Coverage of Social Assistance Programs by Age Group, 2012, ..............................................................123 Figure 8.12: EDE PEP Coverage (FAES-Executed) by Region and Program Type, 2013 .............................................124 Figure 8.13: Numbers of Meals Distributed, Kantin Mobile, 2013 (Percentage of Total Population) .......................124 Figure 8.14: Benefit Amounts and Contribution to Consumption of Beneficiaries ...................................................125 Figure 9.1 : Evolution of Domestic Price and International Prices of Oil Products, 2004-14 ....................................128 Figure 9.2 : Oil Subsidies Costs, 2010-13 ..................................................................................................................128 Figure 9.3 : Total Post-Tax Costs of Petroleum Subsidies, 2011 (Percentage of GDP) .............................................129 xii Figure 9.4 : Total Post-Tax Costs of Petroleum Subsidies, 2011 (In percent of General Revenues) .........................129 Figure 9.5 : Oil Subsidies and Public Spending Allocation, 2010-12 .........................................................................129 Figure 9.6 : Expenditure on Petroleum Products Per Decile, 2012 (Percentage of Budget) ....................................130 Figure 9.7 : Fuel Subsidies Received By Group, 2012 (Percentage of Total) ............................................................131 Figure 9.8 : Composition of Expenditure by Groups, by Decile, 2012 ......................................................................132 Figure 9.9 : Expenditures by Groups, 2012 ...............................................................................................................132 Figure 9.10: Welfare Loss by Direct Effect (Percentage) ...........................................................................................133 Figure 9.11: Intensity In Oil Products By Sector ........................................................................................................134 Figure 9.12: Effect of Prices from the IO Table ..........................................................................................................134 Figure 9.13: Contribution to Inflation ........................................................................................................................134 Figure 9.14: Welfare Loss by Decile (Percent) ..........................................................................................................136 Figure 9.15: Main Source of Lighting, 2012 (percent Households) ..........................................................................137 Figure 9.16: Households Using Kerosene As Main Source of Lighting, 2012 (Percent) .............................................137 Figure 9.17: Total Loss of Welfare by Decile ..............................................................................................................137 Figure 9.18: Welfare Losses by Urban/Rural Areas ...................................................................................................137 LIST OF TABLES Table 1.1 : Frequency and Impact of Natural Disasters, 1971-2014 .............................................................................3 Table 3.1 : Fiscal Indicators (In Percent of GDP) ........................................................................................................22 Table 4.1 : Customs Duties and Inspection Fees (2013) .............................................................................................43 Table 4.2 : TRIST Simulations on Four Scenarios of Tariff Reform ..............................................................................44 Table 5.1 : Set of Projects Reviewed by Ministerial Affiliations..................................................................................50 Table 5.2 : Procurement Thresholds for State institutions (in millions of Haitian Gourdes) .....................................59 Table 5.3 : Summary of Applications Relating to the Expenditures and Investment Chain .......................................66 Table 6.1 : Maternal and Children Health Coverage by Income Group, 2005-6 and 2012 .........................................74 Table 6.2 : Participation Incidence by Income Groups, 2013 (in Percent of Beneficiaries) ........................................79 Table 6.3 : Technical Efficiency In Haiti And Other Low-Income Countries ................................................................87 Table 7.1 : Late Starts, Repetition, and Drop out Contribute to Low System Efficiency ............................................95 Table 7.2 : National Exam Passing Rates, 2012-13 (In Percent)..................................................................................96 Table 7.3 : Characteristics of Primary Schools by Department, 2010-11 (In Percent) ..............................................104 Table 8.1 : Social Insurance: Institutions and Programs in Haiti, December 2014 ...................................................114 Table 8.2 : EDE PEP Fragmentation: Institutions and Programs by October 2014 ...................................................115 Table 8.3 : EDE PEP Executed Spending (Actual) in 2013 .........................................................................................118 Table 8.4 : Key Donors’ Contributions (Estimates) - Social Protection. ....................................................................119 Table 8.5 : Alignment of EDE PEP Programs with Risks and Vulnerabilities Across the Life Cycle ...........................126 Table 9.1 : Distribution of Oil Subsidies by Fuel Product, 2009-13 (Billion Gourdes) ..............................................129 Table 9.2 : Average Yearly Expenditure on Fuels (HTG)............................................................................................131 Table 9.3 : Fuel Subsidies Received By Income Decile (Percentage of Total) ...........................................................131 Table 9.4 : Subsidies (Million Gourdes) Given Different Scenarios for Oil Prices .....................................................138 xiii Executive Summary Objectives 1. Haiti has a vision to become an emerging economy by 2030. Haiti has comparative advantages, including its proximity and access to major markets; a young labor force and a dynamic diaspora; and substantial geographic, historical, and cultural assets. Areas of economic opportunity for Haiti include agribusiness, light manufacturing and tourism. Recognizing these opportunities, the Government of Haiti issued in May 2012 a Strategic Development Plan (PSDH), aiming at building a new modern, diversified, resilient, competitive and inclusive economy, respectful of its environment and in which people’s basic needs are met. Achieving this objective would require ambitious double digit growth rates, a significant break from the past, possibly based on an expansion of agriculture, construction, manufacturing, and tourism. 2. Haiti’s growth performance in the last four decades has been overall disappointing, however, and poverty remains endemic. A history of vested interests, political instability, and natural disasters, as well as poor governance, inadequate infrastructure and limited skills, have prevented the country to realize up to now its aspirations, trapping it in a low equilibrium. GDP per capita fell by 0.7 percent per year on average between 1971 and 2013. As a result, Haiti is the poorest country in the Latin America and Caribbean (LAC) region and among the poorest in the world. The overall poverty headcount amounts to about 59 percent and extreme poverty to 24 percent in 2012, indicating that almost 6.3 million Haitians cannot meet their basic needs and 2.5 million cannot even cover their food needs. Furthermore, with a Gini coefficient at 0.6, Haiti has the highest income inequality in the region and one of the highest in the world. 3. Haiti experienced a return of donor assistance and greater access to concessional financing over the past decade. Haiti has long been characterized by its very low fiscal revenue mobilization, seriously constraining its ability to carry out needed developmental spending (infrastructure, health, education). A lot of basic services are provided in Haiti by non-State actors. Following greater political stability and particularly the 2010 earthquake, the budget has benefitted from exceptional donor assistance with external grants increasing from 2 percent of GDP in 2004 to 8.1 percent in 2013, peaking at 12.1 percent of GDP in 2010. Meanwhile, in addition to benefiting from the HIPC and the MDRI initiatives, Haiti received additional debt cancellation in the aftermath of the earthquake, reducing its total external debt to 8.9 percent of GDP in 2011 and providing borrowing space that the country used for concessional financing from Venezuela. 4. These resources allowed Haiti to finance a strong expansion in capital spending. Greater donor assistance and the availability of concessional financing have allowed an expansion in capital spending, a substantial shift in the country’s priorities as reflected by the growing share of public investment: it represented in 2014 more than half of total public spending compared to a third in 2005. In line with xiv higher own fiscal revenue, current expenditure has also been on the rise, increasing from just below 10 percent of GDP in 2005 to about 13 percent GDP in 2014, driven by a higher wage bill: wages in relation to GDP rose significantly from 2005 to 2014 (from 3.5 to about 6 percent of GDP). 5. Priority spending in Haiti has also increased. The fragmented fiscal data available would suggest that these additional resources have been channeled to the reconstruction and social sectors, consistent with the decline in poverty and improvements in human development indicators (such as in education) observed over the same period. Overall, the share of resources allocated to social sectors seems to have expanded from an average of about 16 percent of total public spending for the period 2007 to 2010 to about 28 percent of total public spending after the earthquake, for the period 2010 to 2012. In addition, 22 percent of donor assistance on average is estimated to have gone towards social sectors over the period of 2010 to 2012. 6. The composition of public spending may, however, not be conducive to better services. Part of the growing wage bill reflects, however, increases in support, administrative or security staff, and not in staff directly delivering social services. Furthermore, operating costs more specifically (e.g. equipment, utilities) have not kept up with rising public investment, remaining broadly unchanged over the past decade. Without the needed resources to operate the newly built investments and equipment, it is questionable whether the new facilities could be maintained or deliver the expected growth dividends. 7. Furthermore, despite the recent increase in social public spending, Haiti’s economic inclusiveness is still limited. Social spending remains limited and the delivery of basic services highly inequitable. Public spending in health, education, and social protection amounts to 5 percent of GDP, below comparator countries, limiting the government’s ability to offer equal opportunities to its citizens. At the same time, many large spending items such as fuel subsidies clearly favor the rich. In the absence of government, basic services such as health and education are mainly provided by non-government actors. Seventy to eighty percent of primary school students attend non-public schools, placing a substantial financial burden on households and delivering achievements closely linked with household income. Outcomes are equally unfavorable to the poor in the health sector and only a small share of the poor can benefit from basic safety nets. 8. Haiti needs now to adjust to tighter budget constraints. An expansion of aid and concessional borrowing has allowed Haiti to increase public spending and catch up with regional comparators. It has, however, also increased its reliance on foreign assistance as source of financing. Donor assistance and concessional financing represent about 70 percent of the financing for public capital spending. Social sectors including health, education and social protection rely on donor assistance for 45 percent of their financing in 2012. This heavy reliance on donor and concessional financing makes these spending items particularly vulnerable to the decline in aid and in international oil prices (which affect the availability of concessional resources from Venezuela). These tighter constraints could put into question some of the recent progress achieved in human development, making the country’s balancing act between developmental needs and fiscal sustainability even more challenging. In this regard, this report examines more particularly how to: (i) mobilize greater fiscal revenue; xv (ii) enhance the growth dividend of public investment; (iii) improve the efficiency of spending in critical sectors such as health, education, and social protection; and (iv) protect the budget from a return of fuel price subsidies. 9. Against this backdrop, this study aims at contributing to an evidence base to inform decision- making. This report builds on newly available data, including the 2012 household survey, the 2012 Demographic and Health Survey (DHS), and the 2011/12 School Census, to provide a better understanding of the equity and sustainability issues in the delivery of some basic services such as health and education, and engage policy makers and other actors on gaps and priorities. The surge in donor assistance and the availability of concessional financing have exacerbated the fragmentation of fiscal data in Haiti. As part of the study, a special effort has been made to partially consolidate fiscal data and gather a more comprehensive picture of the use of public resources (BOOST Initiative). Finally, the document builds on the recently completed Public Investment Management Diagnostic and Assessment (PIMDA) that has led to an action plan to improve the management of public investment in Haiti. Findings Greater and More Equitable Revenue Mobilization 10. While own fiscal revenue has increased over the past decade, Haiti has still one of the lowest revenue mobilization rates in the region. Haiti’s own fiscal revenues have improved going from less than 10 percent of GDP in 2004 to 12.6 percent of GDP in 2014. Despite this rise in revenue, Haiti has still one of the lowest rates of revenue mobilization in the region. Furthermore, Haiti’s tax system tends to be regressive, relying heavily on indirect taxes. The ratio of direct to indirect taxes stood at about 30 percent in 2011, a level inferior to that of most countries in LAC and to the average of Low-Income Countries, and is largely explained by the fact that a sizable share of Haiti’s revenues comes from international trade. Tax systems relying relatively more on direct taxes tend to be more progressive because in such systems the burden of taxes weighs differently on economic agents of varying income levels. Moreover, while direct taxes tend to be harder to administer, they are frequently seen as an important indicator and tool of successful state-building as they tend to create a more direct citizen-state interaction and voice. 11. With income tax rates largely comparable to the regional average, a large share of revenue is, however, lost because of exemptions. Corporate and personal income tax rates are largely comparable to the regional average and cannot explain Haiti’s low tax burden. For instance, Haiti’s corporate income tax rate of 30 percent is slightly above the regional average. Exemptions, however, are estimated to have amounted in 2011 to the equivalent of more than half of all income tax collected. Most exemptions derive from the 2002 investment code that grants 15-year-tax exemptions for companies in free zones as well as 5-10 year exemptions for specific investment projects that are considered desirable from a development perspective. International experience tends to show, however, that lower taxation does not compensate for an unfavorable investment climate and foster little additional investment. Anecdotal evidence suggests a number of taxes that generate little revenue but impose high compliance costs on businesses (“nuisance” taxes) and reducing them would be a better alternative to attract investors. xvi 12. The top personal income tax rate seems to apply only to very few taxpayers. While top income tax rates are also in line with other countries in the region (30 percent), income brackets are defined in such a way that the top rates only apply to a small share of taxpayers. There are currently four tax bands for personal income tax: 10 percent on incomes above HTG 20,000 and below 100,000; 15 percent up to HTG 250,000; 25 percent up to HTG 750,000; and 30 percent for everything above that. The maximum rate is thus levied at incomes above more than twenty times the national average per capita GDP, more than twice the ratio for the next highest comparator country (Guatemala at about eight times). 13. Substantial fiscal revenue shortfalls also occur at the borders. The largest share of fiscal revenue in Haiti (60 percent) is collected at the border through customs duties, inspection fees, and sales and excise taxes on imported goods. While no complete data is available on TCA and excise tax exemptions, detailed transaction level data on taxes collected at the border makes it possible to calculate the value of exemptions on imported products. Estimating the foregone revenue by comparing the actual revenue collected to the amount an importer should have normally paid had the official rate been applied suggests that Haiti gave up in 2013 the equivalent of 14 percent of total sales tax revenue collected at the border, plus another 19 percent in ad valorem excise taxes. For tariffs, the revenue shortfall rises to 50 percent and the classification of exemptions in customs data is insufficient to thoroughly track their justification. 14. Finally, while Haiti is one of the most open economies, inspection fees increase the country’s effective protection. While Haiti is one of the most open economies with very low tariffs, it imposes various fees that add up to the cost of importing goods. A 5 percent inspection fee is applied to almost all imported products, increasing tariff revenues by about 75 percent in 2013. Inspection fees in the magnitude applied in Haiti mask the true level of protection in the country. Being applied across the board with an identical rate for almost all products, they are unlikely to be in line with the country’s strategic trade objectives that would normally suggest placing a higher tariff on consumer goods, and lower tariffs on key inputs and capital goods. Meanwhile, Haiti is a member of CARICOM, but applies lower tariffs. Alignment of Haiti’s tariffs to the current CARICOM Common External Tariff could generate significant additional revenue, but also substantially raise protection and costs for firms and consumers. One way forward could be to replace the inspection fees with a combination of higher regular tariffs and a removal of exemptions with no loss in total revenue and no increase in protection. Getting a Greater Growth Dividend from Public Investment 15. Despite rising public investment, economic growth has not accelerated so far in Haiti. Haiti still performs poorly for selected infrastructure indicators, such as access to electricity, roads or ports. In this context, high public investment should be expected to contribute in reducing bottlenecks to faster growth. This paradox is, however, nothing new in Haiti. Several reasons have been put forward, ranging from deficiencies in the country’s national accounting system to chronic lack of maintenance or simply the unproductive nature of the investment itself. Poor past donor coordination and the high volatility of external aid have also been argued to have affected the impact of investments in Haiti. 16. This disappointing outcome stems in part from very weak public investment management. Haiti’s public investment management exhibits a number of distinctive features and practices common to countries that are aid-dependent, including weak appraisal capacity and reliance on donors to design good xvii projects. These hamper the effectiveness of public investments. Sectoral strategies to guide the prioritization of projects are lacking. This leads to a Public Investment Program composed of projects that are neither fully assessed nor prioritized. Furthermore, there is no effective ex-ante control on disbursements based on the physical progress of projects against plans. More importantly, domestically- funded capital expenditures are not properly accounted for, tracked and reported, creating an environment conducive to a lack of transparency and accountability, as well as to mismanagement of scarce public resources. Finally, even though the existing legal framework is acceptable for the management of public investments, its requirements are rarely respected, with numerous processes and procedures that, when not redundant, are excessively elaborate. Increasing Sustainability and Improving Further Health Outcomes 17. Health indicators have improved over the last two decades. Child health indicators have improved for instance with the under-five mortality rate declining from 145 deaths per 1,000 live births in 1990 to 73 deaths per 1,000 births in 2013. Haitians can now expect to live 8 more years than back in 1990, two additional years gained compared to the progress made by the average LAC country. The gap in health outcomes between rich and poor households has also narrowed. Significant gains have been made in infant mortality for the poorest quintile, as well as the second and third quintiles over the last decade. Reduction in stunting and in the prevalence of diarrhea has been more pronounced for poor income groups. 18. Health indicators in Haiti remain, nevertheless, poor and disparities persist. The under-five mortality rate is four times higher than the LAC regional average. The 2012 DHS generally shows large inequalities in health outcomes and results that are worse for the poorest quintiles. Infant mortality was thus found to be 62 (per 1,000 live births) for the highest quintile of income, versus 104 for the lowest quintile of income in 2012. Compared to the highest quintile, the number of children suffering from stunting was four times higher in the lowest quintile in 2012 (DHS, 2012), a gap that has not changed since 2006. 19. While the level of total health expenditure per capita is relatively high in Haiti, public spending in health is very low. Health expenditure per capita in Haiti has increased in line with the average of low- income countries and at USD 76.6 per capita stood in 2012 above what the average low-income country spends. According to the WHO, USD 50 per capita would ensure the provision of a basic care package to the population. Haiti was allocating, however, only 5 percent of government spending to health in 2013- 14, well behind the 15 percent recommended by the Abuja Declaration, and government spending accounted for only about 1.5 percent of GDP, a low rate compared to other low- and middle-income countries, spending on average between 2 and 6 percent of GDP. This high level of per capita expenditure was thus mainly financed by external funding, which surged after the earthquake but has been declining since 2012. 20. Despite the recent increase in donor assistance, lack of finance still prevents the poor from seeking treatment. In 2012, total health expenditure was largely financed by donors in Haiti, more so than in other low-income countries, questioning the sustainability of the recently-achieved progress. The surge in donor assistance seems to have been accompanied by a substantial decline in the contribution of xviii households to their health expenditure. This development should have improved access to health services for the poorest. Yet, the most recent household survey reveals that 65 percent of households in the lowest consumption quintiles did not seek medical attention due to their lack of money against 39 percent for the highest consumption quintiles. Moreover, lack of money for treatment is most frequently cited as a barrier to health care in Haiti by women aged 15-49 (76 percent). 21. The composition of public spending may not be conducive to better service delivery. The health sector is a good example, showing that overall operational spending has not kept up with the recent rise in public investment. While operating expenditure has changed little, the investment budget from public funds (Treasury and Petrocaribe) in this sector has doubled between 2006 and 2013. Moreover, the wage bill may be crowding out critical operating expenses (such as medicine and equipment) with the share of the operating budget allocated to personnel costs representing about 90 percent over the period 2006 to 2012. Furthermore, public investment may not be targeted where it is most needed: in 2012, for instance, nearly a third of the commitments of national investment funds were allocated to the construction or rehabilitation of hospitals and only 22 percent of the commitments were allocated to investment projects for dispensaries and health centers, although these facilities account for 87 percent of health care delivery in Haiti. 22. The productivity of the medical personnel of first-level primary facilities is low. Only 13 percent of health facilities are estimated to produce health services efficiently: they have neither excessive operating costs, nor excessive personnel. In particular, the number of consultations is low with one patient per hour, much lower than in comparable countries. A second occupation could explain low productivity with evidence revealing often that half of the medical personnel derive an income from sales activities or from a private practice. Excessive administrative personnel may also undermine efficiency. Administrative personnel represents 40 percent of total institutional personnel although in general the need for such personnel in primary level facilities is low. These facilities are small and have little or no equipment (for dispensaries), requiring a low level of maintenance and management, and therefore few administrative personnel. Keeping Children in School and Improving Education Outcomes 23. Today, the majority of children are in school in Haiti. The majority of preschool age children, and 90 percent of children of official primary school age (6-11), are in school. Educational attainment has also been rising steadily across cohorts of young adults. In 1994, fewer than 30 percent of 15-19 year olds had reached lower secondary school, a figure which had risen to over 50 percent for women and over 40 percent for men by 2012. 24. Despite this progress, education indicators remain low with late primary school entry, high dropout, and limited learning. Compared to its LAC neighbors, Haiti has the highest share of adults with no education, and the highest share of 15-19 year olds who have not completed primary school. The adult literacy rate is about 77 percent, midway between the LIC and the LAC averages. Furthermore, many students, particularly in poor communities, learn little. Indicators of the quality of education, including teacher knowledge and learning materials available at the school level, suggest that many children, particularly poor ones, are receiving a low-quality primary education. Assessments administered in early xix grades in selected schools have found that fundamental skills are acquired very slowly or not at all, particularly in schools in poor communities. For example, the average third grader reads only 23 words per minute, well below the estimated speed of 35-60 words per minute required for comprehension of a basic text. 25. Growth in education supply has primarily come from the non-public sector, making Haiti one of the countries with the largest non-public sectors at the primary and secondary levels. According to the 2010-11 School Census (the most recent published), 94 percent of preschool, 78 percent of primary, and 73 percent of secondary students attend non-public schools. These shares are among the highest in the world. Tuition waivers for students in non-public primary schools have, however, been put in place to help children access school. Donors are financing a complete cycle of primary school for cohorts of children in disadvantaged communities by providing tuition waivers to non-public schools, while the authorities are financing a similar program financed by a new tax on international phone calls and money transfers. Declining donor financing and a recent decision by the Ministry of Education to stop funding tuition waivers for new cohorts of first graders in non-public schools could therefore threaten the gains in access made in recent years. 26. As in health, however, the composition of public spending is not conducive to better services. The operating budget of the Ministry of Education is skewed towards non-“front line” staff. In 2013, for instance, almost 70 percent of the operating budget is allocated to salaries and personnel costs. Although, this share is in line with other LAC countries (where 60-80 percent of the total education budget is absorbed by the wage bill), the public sector in these countries pays the majority of all primary and secondary teachers, while in Haiti the State only directly employs those in the small public sector. As a result, about 42 percent of salary expenditures go to staff who are not working as teachers, directors, or inspectors. Furthermore, half of non-salary operating expenditures go towards textbooks, national exams, and the national school canteen program. 27. Despite weak progression and learning, households spend a substantial amount on sending children to school. Among all households with children age 6-14 in school, 93 percent report positive education expenditures. These expenditures are substantial: on average, these households reported spending 10 percent of total annual household consumption on education in the 2011-12 school year. Consequently, costs remain a major obstacle to education for poor households: when asked why school- aged children are not currently in school, 83 percent of households cite costs as the primary reason. Towards a More Effective and Better Targeted Social Protection System 28. The chronic and transient poor should be an important focus of an effective social protection system in Haiti. Although poverty is widespread in Haiti, rural areas and specific geographic departments require special attention from the social protection system. Nearly half the households are considered chronically poor because they live below the moderate poverty line and lack at least three of the seven basic dimensions defining non-monetary well-being. Such a large share of chronic poverty, unusual in Latin America, highlights the structural challenges of poverty reduction and limitations of the contributions of social protection interventions without combined interventions around basic services and income generation. Haiti’s social protection system should also be able to include the transient poor with xx one million people living slightly above the poverty line but vulnerable to drop below if a shock (weather or health-related) occurred. Natural disasters, in particular, have great disruptive potential partly because they have such a serious impact on agriculture, which represents the main source of livelihood for most of the poor, especially in rural areas. 29. During the last three years (2012-14), there have been encouraging signs of increased attention for social protection as a framework to address both short-term and long-term vulnerability and poverty reduction. In 2012, the national strategy for fighting hunger and malnutrition, ‘Aba Grangou’, was launched, with the aim of coordinating donor and government activities in the nutrition sector. A notable step toward the formation of a social protection strategy was taken in 2014 with the launch of the three-year action plan “Thinking and Fighting for a Haiti without Poverty: Action Plan for Accelerating the Reduction of Extreme Poverty” (PAARP). This document defines a set of flagship programs constituting a social assistance strategy named “EDE PEP” to protect the vulnerable population living in extreme poverty throughout their life cycle, in the short and medium term, to ensure long-term investment in human capital and to provide opportunities to overcome the condition of extreme poverty. The PAARP also identifies a set of elements required to underpin the implementation of a social protection system, such as a national targeting system, a unique beneficiary registry, and an integrated service delivery model at communal level through a network of multi-sectorial agents, as well as local coordination on social protection. 30. Despite these efforts, social protection spending has still a long way to go to become more aligned with challenges at hand. More investment is particularly needed for children under five as this group suffers from the highest poverty rates and as a lack of investment in young children has irreversible negative consequences. As shown in the education and health findings, there is a need for a set of social protection and promotion interventions that would enable the poorest households (especially those in rural areas with young children) to overcome hurdles in building, accumulating, and preserving their human capital in the face of repeated shocks. 31. Furthermore, the various programs building Haiti’s social protection system are excessively fragmented, small, and not targeted well enough to make a difference. Within government’s public spending on social protection, both the contributory (social insurance) and non-contributory pillars (social safety nets) are spread across multiple institutions, challenging the country’s weak administrative capacity. Public spending in social protection is low, amounting to about 5 percent of GDP with subsidies for electricity and petroleum products representing a considerable share (2 percent of GDP). In addition, social assistance benefits are not well targeted: the results of the most recent household survey suggest that as much as half of social assistance benefits accrues to the non-poor. As a consequence, social protection programs have limited impact on poverty and inequality. Finally, most EDE PEP programs are financed principally by concessional borrowing, which poses challenges in terms of their sustainability. Preventing a Return of Fuel Price Subsidies 32. The size of oil subsidies in Haiti had increased dramatically over the last five years, imposing a large burden on public finances. Haiti raised retail fuel prices in October 2014, a notoriously difficult reform, especially in a fragile situation. This move combined with a decline in international oil prices has xxi allowed to eliminate the fuel price subsidies that were burdening the budget. Like other net oil-importing developing countries, Haiti has faced economic difficulties with high oil prices. Haiti has been for a long time cautious passing through the increase in international oil prices to consumers. In 2010, it abandoned a mechanism of automatic pass-through, and adopted a price control system through adjustments in taxes. As the gap between international and domestic prices widened, the cost of keeping retail petroleum prices fixed was placing an increasingly significant burden on public finances (nearly 2 percent of GDP in 2013). 33. Subsidized fuels are mainly consumed by the wealthiest households. Poor households in Haiti consume very little fuel, both in absolute terms and as a share of their total budget. As a result, the poor receive only 1.6 percent of total subsidies accruing directly to households. Since direct usage of petroleum products is so limited among the poor, wealthier households benefit disproportionally from subsidies granted to these products: over 93 percent of fuel subsidies going directly to households benefit the richest 20 percent of the population. Even though there are no data on expenditures on kerosene, there is some indication, nevertheless, that poor households in rural areas would be the most affected by increases in the prices of this product: among the rural poor, over 80 percent of households use kerosene as main source of fuel. Fuel subsidies are particularly regressive when compared to subsidies on other categories of consumer expenditures, such as food, education and health: because these items represent a larger share in their budget, the poor would receive between 30 percent and 40 percent of the benefits from a universal reduction in education, health, and food prices. 34. Increases in fuel prices have large effects on electricity and transportation, however, and poor households are especially vulnerable to increases in food prices. Directly, households are affected through their own consumption of gasoline or kerosene. Indirectly, they are also affected since petroleum products are used as intermediary products in many sectors and their higher price will feed into the price of the final good produced by these sectors. The analysis of the indirect effects suggests that an increase in fuel prices would have the largest direct impact on transportation and electricity, in which oil products account for more than 30 percent of total inputs. Even though the production of food is not intensive in fuel, rises in food prices nevertheless impose a sizeable loss of welfare on the poorest households, due to the large share of food in their budgets. 35. The budget remains vulnerable to a rebound in international oil prices. Fuel subsidies have been eliminated in October 2014. International oil prices are, however, volatile and without the introduction of an automatic price adjustment mechanism for petroleum products, fuel price subsidies could return to haunt the budget, but this time under tighter financing constraints. The subsidies, which amounted to around 2 percent of GDP in 2013, could climb to over 3 percent of GDP with international prices at USD 100/barrel. Going Forward 36. Against this background, some options going forward present themselves. While the report wanted to provide mainly a diagnostic and a better understanding of some critical sectors such as health and education, some broad options going forward can nevertheless be suggested as basis for discussion: xxii  Data: Overall tighter financing constraints make it even more urgent to collect timely data to track public spending appropriately and make it more efficient in different sectors. Differences in budget classifications and the absence of an effective Single Treasury Account lead to a fragmentation of fiscal data and prevents a comprehensive monitoring of public spending from budget appropriation to payment. Fiscal reporting needs thus to be strengthened to allow an appropriate tracking of public spending and set in place the minimum conditions for transparency and accountability in the use of public resources. Completing the roll-out of the Single Treasury Account and unifying the budget classifications among the various sources of financing would be important steps in this direction.  Donors: With declining donor assistance, greater donor coordination is needed to make the most of these resources. A master plan for critical sectors translating the priorities of these sectors into areas of intervention could federate the actions of the donor community, allowing it to coordinate its technical and financial contributions, and avoid potential duplication and inefficiencies. Such a plan would have indicators and deadlines, triennial operational plans, and an annual action plans to structure the monitoring of its implementation.  Budget composition: The composition of public spending has room for improvement to enable better service delivery. The growing wage bill, although in line with economies at similar levels of development, has been the driving force of total current spending, but most of it is accounted for by staff not directly involved in service delivery. Sharp increases in capital outlays have not contributed to the acceleration of growth, due in part to an inadequate resource allocation for operating and maintenance expenses (e.g., equipment, utilities), thereby limiting growth dividends of investment.  Fiscal revenue: Balancing fiscal sustainability and development needs will remain a daunting challenge for Haiti as long as the country does not mobilize substantially more own fiscal resources. In this regard, a review would be needed of existing exemption regimes for all tax instruments with priority to customs duties and corporate income tax where revenue losses are the greatest. An important step would be the introduction of a clear classification system in customs data to track the justification of exemptions. Where tax exemptions or other investment incentives are used, a clear framework needs to be put in place to measure and track their costs. At the same time, it is important to track and monitor the benefits these incentives were expected to deliver (such as employment creation or technology transfer).  Greater growth dividend: Haiti’s development needs are daunting and will need to be addressed. In this regard, the overall public investment management needs to be improved through a strengthening of project evaluation, selection, programming, execution, and control. As a first step, the priority should be put on implementing basic elements of formal project selection and appraisal, execution and controls such as the need for investment projects to (i) advance only if they are mature enough (i.e. developed in a concept or in a justification); (ii) be monitored (including physically); (iii) use resources adequately; (iv) are controlled ex-post as required. xxiii  Improve the performance of health facilities: For a long time the focus has been put on increasing the resources or the inputs. With shrinking budgets, the time may have come to shift the logic upside down and pay more attention to the results or the services actually delivered by the health facilities through a more general use of results-based financing (RBF). RBF could provide monetary and non-monetary incentives to health workers in order to increase their productivity, discouraging them from exercising another occupation. Health facilities may also have very low productivity because they have too few patients. To improve the use of their resources, as well as health outcomes, such facilities may want to reinforce primary health level services and operate mobile clinics to increase access to health services, particularly for the poor.  More accountable schools to increase learning in the classroom: The majority of schools at all levels in Haiti are non-public, and operate with little oversight or accountability. With limited public resources, the sector will remain predominantly non-public for the foreseeable future and efforts should focus on making the schools more accountable for the quality of service they deliver. In this regard, efforts could focus on the comprehensive identification of schools or students benefiting from the different programs offered by the Ministry of Education, including the donor-financed Tuition Waiver Program (TWP) and PSUGO and linking this support with school achievements. Greater oversight could also be called for through a mandatory teaching license based on demonstrated competencies and a mandatory school identity card, leading to certification. In addition, targeted in-service training programs for teachers could be put in place to improve instruction in the classroom and learning.  Better targeting in social protection: Haiti’s social protection suffers from fragmentation of programs and interventions. The tighter resource constraints related to a reduced access to concessional borrowing and donor assistance provide an opportunity to design a less fragmented and better targeted system of social safety nets. In the meantime, continuing progress in the implementation of the social registry of potentially eligible beneficiaries is required.  Protecting the budget: With fiscal resources remaining limited, unbounded calls on the budget should be contained. In this regard, the re-emergence of fuel price subsidies remain a risk. Higher international oil prices could come back again to haunt the budget. There is thus a need to return to an automatic pass-through with this mechanism given to an independent agency to reduce political influence in the setting of fuel prices. xxiv MAP xxv COUNTRY PROFILE This chapter reviews briefly Haiti’s economic potential and discusses the main obstacles the country is facing in achieving faster economic growth. Haiti has comparative advantages, including its proximity and access to major markets, a young labor force and a dynamic diaspora, and substantial geographic, historical, and cultural assets. A history of political instability, violence, and natural disasters, as well as poor governance, inadequate infrastructure, and limited skills, has prevented the country from realizing up to now its aspirations. Haiti Has Opportunities …. 1.1 Haiti’s geography, resources, and history provide it with opportunities . Haiti occupies the western, smaller portion of the island of Hispaniola, which it shares with the Dominican Republic. Both by area and population, Haiti is the third largest Caribbean nation (after Cuba and the Dominican Republic), with 27,750 square kilometers and an estimated population of 10.4 million people. The country has comparative advantages, including its proximity and access to major markets; a young labor force and a dynamic diaspora; and substantial geographic, historical, and cultural assets.1 Areas of economic opportunity for Haiti lie in agribusiness, light manufacturing and tourism. According to the World Economic Forum, Haiti’s economic fundamentals could allow the country to become a vibrant economy and grow by 6-8 percent a year if adequate policies were in place.2 … Hampered by Political Instability and Natural Disasters 1.2 Overall Haiti’s growth performance over the last four decades has been disappointing and poverty remains endemic. A history of political instability, violence, and natural disasters has prevented the country from realizing its aspirations, trapping the country in a low equilibrium and keeping it as one of the poorest and most unequal equal countries in the world. GDP per capita fell by 0.7 percent per year on average between 1971 and 2013. As a result, the overall poverty headcount amounts to about 59 percent and extreme poverty to 24 percent in 2012, indicating that almost 6.3 million Haitians cannot meet their basic needs and 2.5 million cannot even cover their food needs. 1.3 Political violence has occurred regularly throughout Haiti’s history, leading to government instability (Figure 1.1). At Independence in 1804, Haiti was at the forefront of history, being the first nation 1 Haiti has traditionally enjoyed substantial preferential market access to the United States and Canada, but has not succeeded in fully exploiting this advantage. For example, the country has not been able to fill its quota for exports to the United States under the Hemispheric Opportunities through Partnership Encouragement II (HOPE II) initiative or under the Haiti Economic Lift Program (HELP) Act. These laws allow Haiti to assemble textiles, whatever the origin of the imported fabrics, and to export them to the United States duty free and tax free but are due to end in 2020, unless renewed. 2 World Economic Forum (2011). -1- to abolish slavery. Since then, however, with the exception of the 30-year period of autocratic rule under Francois Duvalier (Papa Doc) and his son Jean-Claude Duvalier (Baby Doc) (1957-1986), Haiti has known a succession of short-lived governments. Lacking sufficiently long periods of stability, the country has struggled to develop the institutional mechanisms and policy fundamentals essential to development progress. 1.4 The past decade has been comparatively stable, nevertheless. Over the past ten years, with security provided by a large United Nations Stabilization Mission in Haiti (MINUSTAH), two presidents have been chosen by election and the current President, Michel Martelly, was the first to accede to the Presidency by election from a party in opposition. However, stability remains fragile, with frequent changes in Government and repeated delays in electoral calendars. Such delays led to the lapsing of Parliament in January 2015, leaving the executive to broker a deal with the opposition and agree on a transitional government that would organize elections. 1.5 Furthermore, the Haitian population is one of the most exposed in the world to natural disasters—hurricanes, floods and earthquakes.3 Between 1971 and 2013, Haiti’s economy has been subjected to numerous shocks with natural disasters occurring almost every year and adverse effects on economic growth (Figure 1.2). The country has a higher number of disasters per km2 than the average of the Caribbean countries (Table 1.1). In 2008, tropical storms and hurricanes caused losses estimated at 15 percent of GDP. The earthquake on January 12, 2010 killed 220,000 people, displaced 1.5 million people, and implied losses of the equivalent of 120 percent of GDP. Figure 1.1 : Annual GDP Growth vs. Change in Figure 1.2 : Annual GDP Growth vs. Occurrence of Government, 1971-2013 (percent) Natural Disaster, 1971-2013 (percent) 10 10 5 5 0 0 -5 -5 -10 -10 -15 -15 1999 2005 2011 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 2001 2003 2007 2009 2013 1999 2005 2011 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 2001 2003 2007 2009 2013 Occurrence GDP growth Ocurrence GDP growth Sources: UN, World Bank and International Media Sources: UN, World Bank and EM-DAT: The OFDA/CRED International Disaster Database 3 From 1993 to 2012, Haiti has experienced two droughts, one earthquake, 31 floods and 26 tropical storms/hurricanes. -2- Table 1.1 : Frequency and Impact of Natural Disasters, 1971-2014 Country/ Number of Disasters Disasters/ Disasters/ Deaths/ Total Group Natural / Year Land Surface Population Population Damage/ Disasters (`000 sq. km) (millions) (millions) GDP Haiti 137 3.1 5.0 13 23,427 1.776 exc. Earth- 136 3.1 4.9 13 1855 0.22 quake (2010) Dominican Rep. 60 1.4 1.2 6 311 0.05 Jamaica 34 0.8 3.1 13 102 - Nicaragua 66 1.5 0.5 11 2363 0.33 Honduras 70 1.6 0.6 9 3298 0.40 El Salvador 51 1.2 2.5 8 687 0.34 Guatemala 82 1.9 0.8 5 1754 0.12 Costa Rica 58 1.3 1.1 12 72 0.04 Panama 46 1.0 0.6 12 80 0.01 Other Caribbean 129 2.9 0.3 30 86 0.19 States* Source: EM-DAT: The OFDA/CRED International Disaster Database. *Antigua and Barbuda, The Bahamas, Barbados, Belize, Guyana, Suriname, Dominica, Grenada, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines and Trinidad and Tobago. Simple average. … As Well as Weak Structural Policies and Institutions 1.6 Haiti’s weak structural policies and institutions hamper faster economic growth (Figure 1.3). Poor governance and political instability have been identified in previous assessments as the major impediments to sustainable development, along with weak public sector capacity and accountability, followed by low levels of education and badly deteriorated infrastructure.4 Doing Business Indicators 2015 rank Haiti at 180 out of 189 countries, while the 2014–15 Global Competitiveness Index (GCI) ranks it 137 out of 144 countries. According to these indicators and the investors’ survey of the GCI, the major constraints in Haiti’s business environment include poor infrastructure, limited access to finance, and cumbersome administrative procedures to start a business. 4 See for instance World Bank (2002). -3- Figure 1.3 : Obstacles to Growth Governance Is a Concern... … Infrastructure Inappropriate, ... Governance Indicators, 2013 Quality of Port Infrastructure, 2013 (percentile rank) (1=extremely underdeveloped to 7=well developed and efficient by international standards) Control of Corruption Low Income Countries Rule of Law Latin America and Caribbean Government Effectiveness Small Caribbean Countries Voice and Accountability Haiti 0 50 100 Haiti LIC LAC 0 2 4 6 Source: World Bank Source: World Economic Forum ... Especially Access to Electricity, ... … Human Development Low… Access to Electricity, 2011 Human Development Index, 2013 (percentage of population) (overall score) Least Developed Low Income Countries Countries Caribbean Latin America and Caribbean Latin America and Caribbean Haiti Haiti 0 50 100 0.0 0.2 0.4 0.6 0.8 Source: World Bank Source: UNDP -4- Figure 1.3: Obstacles to growth (continued) … And the Business Climate Unfavorable … … Restricting Competition and Opportunities … Business Indicators, 2015 or Latest Intensity of Local Competition, 2015 (Doing Business) (1=low to 7=high, scores and position of Haiti) 49 Private credit (in % 31 of GDP, 2013) 19 Latin America and Caribbean Strengh of minority 5 inverstor protection 5 (0-10) 2 Low Income Countries 31 Starting a business 29 (days) 97 Haiti 140/144 0 50 100 150 LAC LIC Haiti 3.0 3.5 4.0 4.5 Source: World Bank Source: World Economic Forum ... And Encouraging the Population to Seek Its Future Abroad… ... Or in Low Paid Jobs. Stock of Migrants of Hatian Origin, 1960- Underemployment, 2012 2010 (percentage of employed population) (in thousands) 1200 Sud-Est 79.7 Nippes 79.5 1000 Grand'Anse 79.3 Nord-Ouest 78.2 Sud 77.7 800 Centre 77.1 Nord 76.0 600 Artibonite 75.5 Nord-Est 74.6 400 Ouest 56.4 200 National 70.0 Urban 57.3 Rural 80.3 0 1960 1970 1980 1990 2000 2010 0 20 40 60 80 100 Source: World Bank Source: ECVMAS -5- 1.7 Although governance indicators have Figure 1.4 : Change in Governance Indicators, 2004-13 improved, they remain low. The increase in (Percentile Rank) political stability after the low point of 2004 came with an improvement in a number of governance Control of Corruption indicators (Figure 1.4). The implementation of the Automated System for Customs Data (ASCUYDA) Rule of Law was stepped up, for instance, after the earthquake. Indicators for control of corruption, Government Effectiveness rule of law, government effectiveness, as well as voice and accountability have all registered Voice and improvements in Haiti over that period, although Accountability remaining low compared to the average of the 0 5 10 15 20 LAC region and to that of low-income economies. Source: World Bank 1.8 Haiti’s infrastructure also falls short. Island economies are extremely dependent on the quality, frequency and cost of the means of transport that link them to markets which represent both outlets for their products and supply sources for the needed imported goods. The efficiency and effectiveness of transport, whether by road, by sea, or by air, therefore contribute to the competitiveness of these countries. The quality of transport and logistics services in Haiti is poor, however, ranking 144th out of 160 countries on the World Bank’s logistics performance index (LPI) in 2013 and trailing behind its competitors (Figure 1.5).5 The road network in Haiti is in poor condition and many parts of the country have little connectivity.6 Haiti is also less integrated into the global shipping line network than many developing countries.7 Furthermore, the costs of loading and unloading a standard container at Port-au-Prince are by far the highest of the Caribbean ports (Figure 1.6).8 5 The LPI is based on six core dimensions of trade-related services, including customs, infrastructure, international shipments, logistics competence, tracking and tracing, and timeliness. 6 The sharp deterioration of Haiti’s road network, the inadequate packaging of fruit and vegetables and the use of vehicles unsuited to the transport of agricultural goods are responsible for considerable commercial losses. In the North West, for instance, transport costs for a stem of bananas from the farm to the primary market are estimated to represent 25 percent of the sale price or 45 percent of the profit margin (World Bank, 2013). 7 According to UNCTAD’s liner shipping connectivity index (LSCI), Haiti ranks lower than the main players in the Caribbean: Haiti’s score of 5 is much lower than scores by Jamaica (21), Bahamas (27), and the Dominican Republic (23) that host transshipments ports. The liner shipping connectivity index (LSCI) is made up of five components: 1) the number of companies that provide services from/to a country’s ports; 2) the size of the largest ship providing services from/to a country’s port (measured in Twenty foot Equivalent Units--TEU); 3) the number of services that connect the country’s port(s) to other countries’ ports; 4) the total number of ships operating from/to the country’s port; and 5) the total container carrying capacity of those ships. 8 According to a 2009 study mandated by the Haitian Chamber of Commerce, costs in Port–au-Prince are five times higher than in the ports of the Dominican Republic (World Bank, 2013). -6- Figure 1.5 : Logistic Performance Index, 2014 Figure 1.6 : Port Tariffs Estimated Cost Per TEU*, 2009 (1=lowest, 5=highest) (in US Dollar). 3.5 Port Everglades 144 of 166 3.0 Miami 2.5 Port of Spain, T&T 2.0 Port of Point Lisas, T&T 1.5 1.0 Dominican Ports Authority 0.5 Puerto Rico Ports Authority Haiti 0.0 Terminal Varreux Private Operators (Public Berths) 0 200 400 600 Stevedoring Wharfage Ship Charges** Source: TranSystems Bahamas Dominican Rep. Jamaica Guyana Haiti * Based on a 700 TEU ship, discharging and loading a mixture of 20-ft and 40ft containers. ** Pilotage, Port Dues, Light Dues & Dockage Source: World Bank 1.9 In particular, the quality of public utilities Figure 1.7 : Electric Power Consumption, 2011 is poor. The provision of electricity is unstable and (kWh Per Capita). there are frequent power cuts and surges, which 900 can result in serious damage to industrial 800 700 equipment. Despite this poor quality of service, 600 the cost of electricity is among the highest in the 500 400 region. In 2011, industry was charged $0.32/KwH, 300 compared to $0.18 in Nicaragua, $0.17 in the 200 100 Dominican Republic and $0.06 in Bangladesh.9 0 Furthermore, the supply of electricity covers only a small proportion of the country, with rural areas being particularly neglected. Per capita Source: World Bank consumption of electricity in Haiti is substantially lower than in other Caribbean countries, for example, it is only two percent of the level in the Dominican Republic (Figure 1.7). Access to water and sanitation in Haiti remains also limited.10 1.10 Surveys also point to the lack of qualified labor.11 One major constraint faced by enterprises in Haiti is the difficulty in finding technicians that are well qualified, particularly in new technologies. This forces the country to position itself as a low cost producer for goods and services requiring limited skills. A deficient labor market may be at fault, with no institutional mechanism to enable the exchange of information between labor demand and supply. More generally, the level of education among the adult population remains low: 45.7 percent of the adult population (60.5 percent of households’ heads) have never attended school or have not completed primary education. Not mastering basic skills such as literacy and numeracy when starting work represents a major impediment for their insertion in the labor market and, more importantly, for their ability to absorb post-school training either on or off the job, and to adapt to changing job requirements. 9 IFC (2011) 10 64 percent of the population has access to an improved drinking water source (48 percent in rural areas) and 31 percent access to improved sanitation facilities (16 percent in rural areas). 11 See for instance ILO (2010) -7- LIVING WITH TIGHTER BUDGET CONSTRAINTS 12 This chapter discusses the overall trends in public expenditure, fiscal revenue, and borrowing. Haiti has managed over the past decade to increase its own fiscal revenue. The re-engagement of donors led to an increase in assistance that accelerated following the 2010 earthquake. Debt relief has also provided the country with space to access concessional borrowing. The availability of these resources has allowed public spending in Haiti to catch up with comparators. The sustainability of this high level of public spending is, however, now put into question by the decline in aid and lower oil international prices that are reducing the availability of concessional financing. Against this backdrop, the need to mobilize own fiscal revenue and improve the efficiency of public spending have become all the more urgent. Macroeconomic Stability Has Been Maintained … 2.1 Haiti’s macroeconomic environment has broadly improved over the past decade. Growth has regained some momentum following the earthquake, driven by services and construction (Table 2.1). Inflation was brought back to single digits and has remained contained at these levels since 2009 (Figure 2.1). International reserves have increased and have been maintained at an adequate level, covering about 5 months of imports (Figure 2.2). Between 2009 and 2011, Haiti’s stock of external debt drastically shrank following debt relief. In addition to benefiting from the HIPC and the MDRI initiatives, Haiti received additional debt cancellation in the aftermath of the 2010 earthquake. As a result, total external debt fell to 8.9 percent of GDP in 2011. Figure 2.1 : Inflation, 2004-14 Figure 2.2 : Gross International Reserves, 2004-14 (In Percentage) (Millions of US Dollars) 30 3000 25 2500 20 2000 15 1500 10 1000 5 500 0 0 Sources: IMF and Haiti's Statistical Institute (IHSI) Sources: Central Bank of Haiti (BRH) 12 Prepared by Kassia Belo da Silva Antoine, Ibrahim El ghandour, Michel Matera, Emilio Sacerdoti, and Rafael Van der Borght. -8- … With Greater Revenue Mobilization and a Surge in Aid 2.2 Haiti received an unprecedented amount of support and aid in response to the 2010 earthquake. With the re-engagement of donors after the political turmoil, external grants increased from 2 percent of GDP in 2004 to 8.1 percent in 2013. Following the earthquake, private donations are reported to have reached USD 3.1 billion (47 percent of GDP). 58 donors made pledges totaling USD 5.5 billion (83 percent of GDP) to help Haiti at the International Donor’s Conference in New York on 31 March, 2010. The Office of the Special Envoy for Haiti reports that bilateral and multilateral donors had pledged USD 9.28 billion (140 percent of GDP) in humanitarian and recovery funding for 2010 to 2012. Of these pledges, USD 5.63 billion (60.7 percent) has been disbursed. From the available figures, it appears that NGOs and private contractors have been the primary intermediate recipients of this assistance, extending their already significant presence in the provision of social services (Ramachandran and Walz, 2012). Although very little money has gone directly to the Government of Haiti, these resources have nevertheless contributed to the recent increase in fiscal revenue (Figure 2.3). Figure 2.3 : Fiscal Revenues, 2004-13 (In Percentage of GDP) 25 20 15 10 5 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Domestic revenues Customs Other revenues Grants Source: Ministry of Finance (MEF) and IMF 2.3 Domestic revenue mobilization has also improved. Although more modestly, Haiti’s own fiscal revenues have also improved during these years going from less than 10 percent of GDP in 2004 to 12.6 percent of GDP in 2014 (Table 2.1). Since 2011, however, progress in increasing revenue collection has slowed and even reversed with tax revenues in terms of GDP declining from their peak in 2012 of about 13 percent of GDP. As discussed in Chapter 4, some of the factors behind this decline relate to the inefficiency of the tax collection agencies, as well as revenue shortfalls stemming from the freeze of domestic fuel prices and exemptions granted on imports. -9- Table 2.1 : Macroeconomic Indicators 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 (annual percentage change) National accounts and prices Real GDP 1.8 2.3 3.3 0.8 3.1 -5.5 5.5 2.9 4.2 2.8 Consumer price index (average) 28.3 14.2 9.0 14.4 3.4 4.1 7.4 6.8 6.8 3.9 External sector Exports (goods, f.o.b.) 21.3 7.7 5.5 -6.0 12.4 2.2 36.3 2.2 12.3 3.4 Import (goods, f.o.b) 8.0 18.3 4.5 30.2 -3.6 38.3 7.3 -11.1 24.3 1.9 Money and credit Base money (currency in circulation and gourde deposits) 0.6 5.5 7.6 13.9 9.5 31.2 6.0 -3.7 15.1 1.7 Credit to private sector 21.1 5.5 10.3 25.8 14.7 -6.3 25.5 30.6 16.8 11.2 (in percentage of GDP) Fiscal accounts Revenues 13.1 13.5 15.8 15.1 17.8 23.9 21.9 23.4 20.8 19.7 Expenditures 13.8 15.2 15.6 18.2 22.4 21.7 25.5 28.2 28.0 26.1 Overall balance -0.6 -1.7 0.2 -3.1 -4.6 2.2 -3.6 -4.8 -7.1 -6.4 Balance exc. grants -4.1 -5.0 -5.1 -7.5 -11.3 -10.0 -12.7 -15.4 -15.2 -13.5 Public debt, Stock 34.2 28.2 25.5 30.0 19.4 15.9 11.5 15.8 19.5 24.1 Balance of payment Current account balance 0.7 -1.8 -1.5 -3.1 -1.9 -1.5 -4.3 -5.7 -6.3 -6.3 Exports (goods and services) 14.0 14.5 13.2 14.0 15.7 15.3 17.5 16.8 18.6 19.0 Imports (goods and services) 41.7 45.3 40.8 43.9 42.9 64.7 59.0 53.2 52.3 51.8 Workers remittance - - - 20.9 20.9 22.3 20.6 20.4 21.1 22.7 Foreing direct investment 0.6 3.4 1.3 0.5 0.8 2.7 1.6 2.0 1.9 1.1 External debt, Stock 34.2 28.2 25.5 30.0 19.4 12.9 8.9 13.8 17.7 21.4 Memorandum items: Nominal GDP (millions of gourdes) 168,035 197,183 220,110 250,590 267,880 266,952 302,854 328,061 364,526 388,809 Nominal GDP (millions of US dollars) 4,310 4,757 5,885 6,549 6,585 6,623 7,517 7,890 8,453 8,713 Exchange rate (average) 39.0 41.4 37.4 38.3 40.7 40.3 40.3 41.6 43.1 44.6 Gross reserves (millions of US dollars) 279 388 596 759 1,018 1,832 2,045 2,346 2,384 1,914 in months of imports 1.6 2.0 2.5 3.2 2.8 5.0 5.8 6.4 6.3 5.2 Sources: Minitry of Finance (MEF), Central Bank (BRH), Haiti's Statistical Institute (IHSI), IMF and BOOST database. *Note: Annual data covers Haiti's fiscal year (Oct. 1 to Sep. 30) 2.4 This expansion in overall revenue Figure 2.4 : General Government – Total Revenues, allowed Haiti to catch up with comparator 2000-12 (In Percentage of GDP) countries. Haiti’s overall fiscal revenues registered a gradual increase, reaching about 20 percent of GDP in 2014 (compared to 11.3 percent of GDP in 2004) and catching up with comparators (Figure 2.4). Haiti’s mobilized revenue is now closer to the regional average and above the average of countries at similar levels of development. The country is, however, much more reliant on donor assistance with about 40 percent of its fiscal revenue coming from donors in 2013, compared to about one percent on average for Small Caribbean States or in LAC (Table 2.2). Nicaragua and Honduras depend on donors for about 10 percent and 8 percent of their fiscal revenue, respectively. -10- Table 2.2 : Fiscal Revenues - International Comparison (In Percentage of GDP) Caribb.States El Salvador Guatemala Dominican Costa Rica Nicaragua Honduras Panama Jamaica Small Haiti Rep. LAC 2013 2013 2010 2013 2010 2012 2012 2012 2012 Total 20.8 27.5 13.6 18.4 17.4 18.9 11.6 13.9 18.7 26.4 23.1 Domestic 12.7 26.9 13.5 16.5 16 18.2 11.6 13.9 18.6 25.2 21.9 Grants 8.1 0.6 0.1 1.9 1.4 0.7 0 0 0.1 1.2 1 Source: MEF and IMF … Allo wing Higher Public Spending 2.5 Public expenditure has been on Figure 2.5 : Public Expenditures by Source of Financing, the rise since 2004. The availability of 2001-13 (In Percentage of GDP) higher overall fiscal revenues described in 30 the previous section has allowed Haiti to 25 expand public spending and catch up with comparator countries (Figures 2.5 and 20 2.6). In 2004, Haiti’s public expenditure 15 amounted to 11.4 percent of GDP, the 10 lowest level in the LAC region. Since then, public spending increased gradually up to 5 2009, when it reached about 22 percent 0 of GDP. This trend has accelerated after 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 the earthquake, when public expenditure Current - Treasury Capital - Treasury increased to 28 percent of GDP in 2012. Capital - Petrocaribe Capital - Grants Even though, public expenditure has Sources: Ministry of FInance (MEF) and IMF declined in the last two years, at 26 percent of GDP in 2014, it remains high by historical standards. As shown in Figure 2.5, the main factor in this change has, however, been the rise in external aid since 2004, and especially following the earthquake. -11- Figure 2.6 : General Government Expenditure - International Comparison 2004-13 (In Percentage of GDP) … Despite Widening Deficits, Debt is Expected to Remain Sustainable … 2.6 Haiti’s twin fiscal and current account deficits have widened. The fiscal deficit averaged 1.3 percent of GDP during the period 2005 to 2008, but widened to 7.1 percent and 6.4 percent of GDP in 2013 and 2014, respectively (Figure 2.7). This widening fiscal deficit mainly reflected transfers to the public electricity company (EDH), which amounted to about 1.5 percent of GDP in 2013, and the fuel retail price freeze that had a fiscal cost of almost 2 percent point of GDP (Box 2.1). The counterpart of the rising fiscal deficit was a rising external current account deficit (Figure 2.8). Both deficits are largely financed by concessional (Petrocaribe) flows from Venezuela. Figure 2.7 : Central Government Fiscal Balance, Figure 2.8 : Current Account Balance, 2004-14 2004-14 (Percentage of GDP) (Percentage of GDP) 4 1 0 2 -1 0 -2 -2 -3 -4 -4 -5 -6 -6 -8 -7 Sources: Ministry of Finance (MEF), IMF and World Bank staff calculations Sources: IMF and Central Bank of Haiti (BRH) 2.7 While increasing, Haiti’s debt is expected to remain at a sustainable level . As previously mentioned, Haiti benefitted from significant debt relief during 2009-11, with external debt declining from 30 percent of GDP in 2008 to 9 percent in 2011. The country has taken the opportunity of this greater -12- borrowing space to contract loans on concessional terms with Venezuela through the Petrocaribe agreement (Box 2.1 and Figure 2.9). The availability of these funds has allowed the financing of higher levels of public investments and some social programs. Despite a steady accumulation of debt towards Venezuela (amounting at end-2014 to about 18 percent of GDP and accounting for 88 percent of Haiti’s outstanding external public debt), the latest joint Bank-Fund Debt Sustainability Analysis concludes that the country’s risk of debt distress is moderate (Figure 2.10). Box 2.1 : Electricité d’Haiti Due to poor management, the performance of the national, vertically integrated electricity utility Electricité d’Haïti (EDH) has deteriorated over time. Worsening commercial performance has led to a lack of infrastructure maintenance and the quality of electricity service has rapidly deteriorated, including frequent service interruptions and large voltage fluctuations. The sector’s institutional framework is obsolete (e.g. not allowing EDH to tap into important renewable energy potential), sector policies are out of date, planning and monitoring of sector activities is inadequate, and vested interests have hampered reform efforts. The compound effect of the earthquake on power generation, transmission, and distribution further aggravated EDH’s weak performance by delaying key modernization activities (e.g. new billing system and the rehabilitation of the Peligre hydropower plant). In 2015, EDH’s losses stood at 57 percent of the electricity generated (of which commercial losses represented nearly 35 percent). Only one-fourth of Haiti’s population has access to electricity, and in rural areas, access is around 5 percent. EDH’s weak grid infrastructure, poor commercial performance, and inadequate controls over sub -contracted electricity generation by IPPs have led to a financial drain on Government resources. Due to its inability to meet electricity demand and in an attempt to expand electricity availability, EDH has subcontracted part of the production of electricity to IPPs. 13 Unable to cover its operating expenses, including fuel costs and power purchases in part because of low bill collection rates, EDH has relied on fiscal transfers from the Treasury averaging USD200 million annually in recent years (equivalent to 10 percent of the national budget and 1-2 percent of GDP).14 Past governance reforms to preserve the integrity of public resource management in the electricity sector have produced modest results. Achievements to-date include: (i) EDH Commercial Recovery action plans (2012-13; 2013-14; and 2014-15); (ii) a National Directions Paper (endorsed by Cabinet on September 15, 2013); and (iii) a Government Energy White Paper (in draft). Since 2014, an active dialogue on the management and reform of the energy sector has been initiated within the Government’s Energy Commission and with the strategic partners in the sector. 13 IPPs provided 60 percent of total electricity generation in 2011. 14 Electricity transfers are unbounded due to their dependence on international oil prices, electricity supply, unsanctioned theft and bill non-payment, inter alia. -13- Figure 2.9 : Haiti – Stock of Debt from Petrocaribe, Figure 2.10: External Debt (PV), 2014-35 2009-14 (In Percentage of GDP) (Percentage of Exports G&S + Remittances) 20 90 18 80 16 70 14 60 12 50 10 40 8 30 6 20 4 10 2 0 0 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2008 2009 2010 2011 2012 2013 2014 Source: IMF Sources: IMF and World Bank Box 2.2 : Petrocaribe and Haiti The Petrocaribe initiative, created in 2005, has 18 members. 15 The agreement aims to promote cooperation among state energy operators, in terms of technology, energy policy, joint exploration, refining, sales and investment in the energy sector. The most relevant aspect of the Petrocaribe agreement is the stable supply of oil from Venezuela to other members, at favorable financing conditions. Through the agreement, Venezuela commits to providing oil to the members based on quotas established bilaterally at international market prices. Haiti joined the initiative in March 2007 and started benefiting from the agreement in October 2007, with a quota of 14 thousand barrels per day (b/d). Petrocaribe Financing Conditions Price of Barrel (dollars) Share of Financing Repayment Period (years) Grace Period (years) (% of total) 150 70 23 2 100 60 23 2 80 50 23 2 50 40 23 2 40 30 23 2 30 25 15 2 Source: BMPAD The sale of these oil products in the domestic market creates considerable resources for the government. Since 2008, the Petrocaribe flows have amounted cumulatively to 25 percent of GDP. An autonomous agency of the Ministry of Finance, (the Bureau de Monétisation du Programme d'Aide au Développement , BMPAD), manages these resources on behalf of the Haitian government. BMPAD plays an intermediary role between Haitian fuel purchasers and the Venezuelan supplier. Petrocaribe resources have been used to finance investment and social projects, as well as support the electricity sector. Whenever the government decides that new projects will be financed using these resources, it publishes a resolution listing these projects and the amounts to be financed in the official gazette. Disbursements into the projects are reported in BMPAD’s webpage o n a regular basis. These projects suffer nevertheless from the same shortcomings as the rest of the public investment program, notably a lack of proper assessment, prioritization and monitoring. 15 Antigua and Barbuda, Bahamas, Belize, Cuba, Dominica, Grenada, Guatemala, Guyana, Haiti, Honduras, Jamaica, Nicaragua, Dominican Republic, St. Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Suriname and Venezuela. -14- … But Financing Constraints Are Becoming Tighter 2.8 The exceptional level of aid following the 2010 earthquake is rapidly declining, however, eroding fiscal buffers. Donor assistance has declined from its peak after the earthquake to 7 percent of GDP in 2014 (Figure 2.11). Additional fiscal revenue was not mobilized or public spending cut to offset fully this substantial decline in resources and government deposits needed to be drawn upon (Figure 2.12). In light of Haiti’s reliance on donor assistance to maintain the country’s fiscal position on a sustainable path, greater efforts are called for to mobilize more own fiscal revenue, improve the efficiency of existing public spending programs, and increase their growth dividends, as will be discussed in more detail in the next Chapters. Figure 2.11: International Aid, 2008-25 Figure 2.12: Government Deposits, 2009-14 (In Percentage of GDP) (In Percentage of GDP) 20 18 16 14 12 10 8 6 4 2013 2017 2021 2008 2009 2010 2011 2012 2014 2015 2016 2018 2019 2020 2022 2023 2024 2025 Source: IMF 2.9 The recent decline in international oil Figure 2.13: Petrocaribe Financing, 2008-17 prices is further tightening fiscal resources. (Millions of US Dollars) According to the Petrocaribe agreement, lower 450 international oil prices imply a higher payment 400 rate and less concessional financing. With 350 lower international oil prices, Haiti has 300 experienced a substantial decline in the 250 availability of concessional financing (Figure 200 2.13). International oil prices are expected to 150 100 remain low in the near future, making it even 50 more challenging for the country to balance 0 the need to ensure macroeconomic stability 2008 2009 2010 2011 2012 2013 2014 2015p 2016p 2017p while protecting social and investment crude at US$50 crude below US$40 crude at US$100 World Bank projections spending needed for social cohesion and actual Sources: Government of Haiti and World Bank staff growth. -15- … While Natural Disasters Would Call for Larger Fiscal Buffers 2.10 Natural disasters involve an extremely high opportunity cost. Work was carried out jointly with the Ministry of Economy and Finance (MEF) to develop several hypothesis regarding contingent liabilities for the government in the event of disasters. The results suggest that the annual average fiscal cost for the government can be estimated at around one percent of GDP. This estimated cost was based on modeling used by the CCRIF and is linked to damages caused by tropical hurricanes (0.8 percent of GDP) and by earthquakes (0.2 percent of GDP). In the absence of a financial strategy to manage this risk and appropriate fiscal buffers, natural disasters will require reallocating public funds and involve high opportunity costs. 2.11 Less frequent, higher-impact events can pose a real threat to the sustainability of public finances. While frequent, low-impact events constrain fiscal space, less frequent, higher-impact events can imply significant fiscal shocks and pose a real threat to the sustainability of public finances. While the government can mobilize a portion of its financing with respect to low-impact events through budget reallocations or deposits, these funds will not be sufficient to address major events that give rise to costs that far exceed the government’s financing capacity. The fiscal shock associated with a hurricane that occurs once every 50 years, for instance, is estimated at about 10 percent of GDP, representing a much greater risk to fiscal sustainability and macroeconomic stability. -16- PROTECTING PRIORITY SPENDING 16 This chapter reviews recent trends in public spending, both by economic classification and by ministries. Haiti needs to build its human and physical capital to achieve its aspirations of becoming an emerging economy. Despite serious data limitations, the overall picture would suggest that the availability of both donor assistance and concessional financing over the past decade has allowed the country to increase spending in these priority areas. While capital spending has accelerated, operating expenses have, however, not followed, questioning the viability of many public investments. Moreover, although the wage bill has been driving the increase in current spending, most of it is accounted for by staff not directly involved in service delivery. The heavy reliance on donor and concessional financing makes social sectors particularly vulnerable to a decline in these resources. Furthermore, despite the recent increase is social spending, the role of the State in basic services such as health and education remains limited and access is inequitable. Haiti Has a Vision Calling for Higher Human and Physical Capital … 3.1 Building on its opportunities, Haiti has a vision to become an emerging economy by 2030. The Government of Haiti issued in May 2012 a Strategic Development Plan (PSDH), building on the Action Plan for the Recovery and Development of Haiti (PARDH) prepared with its international development partners following the 2010 earthquake. The PSDH details the PARDH’s vision and strategic plan for the country’s long-term development, and the four major areas of work for the recovery and development of Haiti (territorial reform, economic reform, social reform and institutional reform). The Plan aims at building a new modern, diversified, resilient, competitive and inclusive economy, respectful of its environment and in which people’s basic needs are met. This objective would require ambitious double digit growth rates, a significant break from the past, based on an expansion of agriculture, construction, manufacturing, and tourism. This strategy remains, however, overly vague and provides little guidance on the appropriateness of the composition of public spending. 3.2 Faster economic growth alone will, however, not be enough to bring significant improvements in the living standards of most Haitians. Simulations show that if growth in Haiti up to 2030 were to follow its historical performance, poverty would hardly decline. Under the more sustained performance of one percent real per capita growth rate observed in Haiti over the 2005-2009, poverty reduction would still fall significantly short of reaching the goal of extreme poverty of 3 percent or less by 2030 (Figure 3.1). Reducing extreme poverty by almost half in 15 years, to about 14 percent, would require an overall GDP growth of 3.3 percent per year, the average growth registered in post-earthquake/reconstruction years (2011-2014). Assuming unchanged income distribution, per capita GDP would need to grow by about 7 16 Prepared by Kassia Antoine Bela da Silva, Ibrahim El ghandour, Eduardo Estrada, Konstantin Wacker, and Emilio Sacerdoti. -17- percent per year for extreme poverty to fall to 3 percent by 2030. This would require a two- to three-fold acceleration in Haiti’s growth rate with respect to its best performing years: a very ambitious outcome. 3.3 Policies to ensure more inclusiveness are needed. Increasing growth of the bottom 40 percent by around one percentage point more than the best observed average growth rate (2005-2009) would add considerable impetus to poverty reduction. In this scenario, per capita real GDP of the bottom 40 percent would grow twice as fast as the mean. Such a performance would cause poverty to decrease by 5 percentage points more than in the distribution neutral growth scenario, coming closer but still falling short from reaching the 3 percent target for extreme poverty by 2030. To achieve the 3 percent target, a combination of faster and more inclusive growth would be needed: a growth of about 4 percent per year with the income of the bottom 40 growing at twice that speed. But what would be the measures needed to deliver this faster growth in lower income groups? Figure 3.1 : Extreme Poverty Simulations Poverty Rate Projections, 2013-2030 Poverty Rate Projections, 2013-2030 (Percentage of Population) (Percentage of Population) -18- 3.4 Human capital and political Figure 3.2 : Economic Magnitude of Estimated Parameters stability matter more for lower income (In Percentage Points) groups. To provide a more specific framework to assess the appropriateness Cabinet Changes of the composition of public spending, we Inflation built on Araujo et al. (2014) and examined Inflation variation more closely the factors that could lead to faster growth for lower income groups. School enrollment First, the correlation between potential Mobile phone use drivers of income and the income growth of Transparency the poorest 20 and poorest 40 percent of Life expectancy households were estimated, using household data for a sample of 100 -0.03 0.00 0.03 0.06 0.09 0.12 countries (Box 3.1). This measure gives us bottom 20 bottom 40 mean Source: World Bank staff the expected impact of changing one of these drivers on the income growth of the poorest. Second, following Araujo et al. (2014), the standard deviation of each variable is taken to illustrate the magnitude of possible or realistic change one could expect in this variable. Multiplying the two gives a rough estimate of the expected effect of a realistic change in a variable for income developments. The analysis finds that health (proxied by life expectancy), as well as education (measured by school enrollment) are important for increasing income generation at the bottom of the income distribution (Figure 3.2). Reasonable changes in political stability (measured by the number of cabinet changes and within-regime instability) matter slightly more to the lower income groups. Infrastructure (measured by mobile phone usage, a common measure in the literature) and transparency, although not disproportionally favoring the lowest income group, are nevertheless important for overall income growth. 3.5 As an alternative, Haiti’s performance in key variables was benchmarked to a reference group. As a reference group, the average of an aspirational group of countries was taken (Belize, Dominican Republic, Ecuador, El Salvador, Jamaica, Paraguay, and Peru). In terms of income levels, these LAC countries broadly stand today where Haiti aspires to be after 2030, i.e. beyond a threshold of USD 2,310. Figure 3.3 illustrates Haiti’s position relative to its aspirational peers with respect to life expectancy. It shows that the country is in the lowest third percentile of the health distribution and lags behind the group average by approximately ten years. Figure 3.4 does the same for cabinet changes, highlighting that other peers suffered from comparable political instability as well (especially Peru, Ecuador, and Paraguay). -19- Box 3.1 : Benchmarking the Drivers of Shared Prosperity: An Application to Haiti 17 We estimate economic magnitudes for key correlates of income growth at the mean, the bottom 40 percent and the bottom 20 percent of the income distribution across a set of about 100 countries. While econometric exercises investigating determinants of GDP growth have been numerous, less is known about factors influencing household incomes at different segments of the income distribution. Empirical work by Dollar and Kraay (2002) and Dollar et al. (2013) broadly rejects the idea that other factors than mean income growth would influence incomes of the poor, thus suggesting that growth would be mostly distribution neutral. Other studies have attempted, by contrast, to outline key drivers of pro-poor growth (e.g. Bourguignon, 2003; Ravallion and Chen, 2007; Christiaensen et al. 2013). Our results add to this discussion and suggest that some factors indeed allow growth to be more beneficial for the bottom 40 percent or the bottom 20 percent of the income distribution. Our data set is mainly based on Dollar et al. (2013). The data covers household data from the World Bank’s POVCALNET and the Luxembourg Income Study (LIS) databases. Household income data is organized in “spells,” i.e. income changes between two survey years, calculated as average annual log differences.18 These are calculated for average income, income at the bottom 20 percent or the bottom 40 percent. We focus on those non-overlapping spells that are at least five years long, providing 299 spells for 117 countries with a median spell length of 6 years, which is also the preferred sample of Dollar et al. (2013). To this data, we add macroeconomic variables, especially measures of political stability or institutions that are of particular interest for Haiti. Based on a fixed effect estimation, the results confirm that the income at the bottom 20 percent and bottom 40 percent grow with mean income. The estimated effect is a little smaller than previously thought, although the estimated elasticity around 0.85 is statistically not significantly different from one. We also find that better health and education outcomes, as well as lower political instability, have an economically relevant effect for income growth that goes beyond the effect via mean income (i.e. it is stronger at the bottom of the distribution. Figure 3.3 : Life Expectancy at Birth, 2010 (Years) Figure 3.4 : Cabinet Changes, 2003 (10 Years Average) 80 80 1.2 120 70 69 70 1.0 100 60 60 97 50 50 50 0.8 82 82 80 40 42 40 71 30 31 33 30 0.6 60 20 22 20 19 0.4 40 10 10 32 0 3 0 0.2 20 6 9 9 0.0 0 2010 value p-rank (rhs) 2003 value p-rank (rhs) Source: World Bank Source: World Bank 17 Based on a background paper prepared for the SCD by Kassia Antoine, Raju Singh, and Konstantin M. Wacker. 18 POVCALNET data is either income or consumption, LIS data is disposable income. We still refer to “income” in our paper. -20- 3.6 Again, closing the gaps in human Figure 3.5 : Income Effects of Closing the Gap capital and achieving greater political (In Percentage of the Gap) stability seem to have the highest potential for lower income groups. Life expectancy at birth Multiplying the gap in the indicators Mobile phone use between Haiti and the reference Cabinet Changes aspirational group by the unconditional effects estimated above captures the Transparency potential income gains Haiti may School enrollment experience from closing the gap in these Inflation variation key variables. This idea, based on Araujo et al. (2014), not only highlight the areas Inflation where the country lags behind most but 0 5 10 15 implicitly weights this gap by the economic mean bottom 40 bottom 20 Source: World Bank relevance of closing it.19 Figure 3.5 shows the economic income effect from closing the gap relative to the reference aspirational group for different parts of the income distribution. Variables relating to health, education, political stability clearly stand out. Again, closing Haiti’s gap in infrastructure, although not disproportionally favoring the lowest income groups, is important for overall growth. Given Haiti’s recent performance in maintaining macroeconomic stability, the gap with the aspirational reference group in terms of low and stable inflation is small and any incremental narrowing would have only a limited effect on income growth. This does not mean, however, that preserving macroeconomic stability is not important. … Reflected in the Rise in Public Investment … 3.7 The strong expansion in capital spending has been the most striking development over the past decade. As mentioned in the previous Chapter, greater donor assistance and the availability of concessional financing resources have allowed an expansion in capital spending (Table 3.1). This increase also highlights a substantial shift in the country’s priorities as reflected by the growing share of public investment: it represented in 2014 more than half of total public spending (compared to a third in 2005). Donor assistance and concessional financing represent, nevertheless, about 70 percent of the financing for public capital spending, and makes this item particularly vulnerable to a decline in the availability of these resources. 19 While it is implicitly assumed that a one-unit improvement in closing a large gap demands less effort than an equivalent one- unit improvement in a small gap, this might not necessarily be the case. Our exercise hence does not explicitly take into account the costs or efforts of closing this gap. -21- Table 3.1 : Fiscal Indicators (In Percent of GDP) 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Revenues 13.1 13.5 15.8 15.1 17.8 23.9 21.9 23.4 20.8 19.7 Fiscal Revenues 9.7 10.2 10.5 10.7 11.2 11.8 12.8 12.8 12.7 12.6 Domestic revenues 6.3 6.5 7.2 7.2 7.4 7.3 8.1 8.6 8.0 8.7 Customs 2.5 3.1 3.1 3.2 3.3 4.3 4.5 4.2 3.9 3.4 Other current revenues 0.8 0.6 0.3 0.4 0.4 0.2 0.3 0.1 0.8 0.5 Grants 3.5 3.3 5.2 4.4 6.7 12.1 9.1 10.6 8.1 7.0 Expenditures 13.8 15.2 15.6 18.2 22.4 21.7 25.5 28.2 28.0 26.1 Current 9.6 9.8 8.6 10.7 11.6 11.2 11.6 11.9 11.8 12.6 Wages and salaries 3.4 3.3 3.7 4.7 5.0 5.5 4.9 5.1 5.5 5.8 Goods and services 2.6 3.1 1.4 3.4 2.9 2.6 2.5 3.5 3.1 3.4 Interest payments 0.7 0.5 1.1 0.7 0.8 0.6 0.4 0.4 0.5 0.5 Internal 0.4 0.1 0.3 0.4 0.4 0.2 0.1 0.1 0.1 0.2 External 0.3 0.4 0.8 0.3 0.4 0.4 0.4 0.3 0.4 0.3 Subsides and transfers 3.0 2.8 2.4 2.0 2.9 2.5 3.8 2.9 2.8 2.9 Capital 4.1 5.4 7.0 7.5 10.8 10.5 13.9 16.3 16.1 13.5 Domestically financed 1.3 1.0 1.6 2.2 4.1 5.5 5.8 5.9 9.0 7.5 o/w Treasury 1.3 1.0 1.6 2.2 1.5 4.4 3.4 3.0 4.6 2.7 o/w Petrocaribe 0.0 0.0 0.0 0.0 2.6 1.1 2.5 2.8 4.0 3.8 Foreign-financed 2.9 4.4 5.4 5.2 6.7 5.0 8.1 10.4 7.1 6.0 Overall balance -0.6 -1.7 0.2 -3.1 -4.6 2.2 -3.6 -4.8 -7.1 -6.4 Balance exc. grants -4.1 -5.0 -5.1 -7.5 -11.3 -10.0 -12.7 -15.4 -15.2 -13.5 Adjustment (unsettled payment obligation) 0.0 0.0 0.0 0.0 -0.2 1.4 0.0 0.0 0.9 -0.1 Financing 0.6 1.7 -0.2 3.1 4.4 -0.8 3.6 4.8 8.1 6.3 External financing 0.6 2.0 0.0 2.7 3.0 3.6 5.4 4.7 4.6 4.0 Loans 1.6 1.9 0.7 2.7 3.0 3.6 5.4 4.7 4.6 4.0 Disbursments 2.4 1.9 1.5 3.4 3.7 3.7 5.4 4.7 4.8 4.3 Amortizations -0.8 0.0 -0.8 -0.7 -0.7 -0.1 0.0 0.0 -0.2 -0.3 Arrears -1.0 0.2 -0.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Internal financing 0.0 -0.4 -0.6 0.0 0.8 -4.4 -1.8 0.1 3.5 2.3 Banking system 0.0 -0.3 -0.6 -0.1 0.2 -5.2 2.4 -1.1 3.8 2.7 BRH 0.0 -0.2 -0.4 0.0 0.2 -4.2 -1.0 -0.1 0.6 1.4 Commercial banks 0.0 -0.1 -0.1 -0.1 0.0 -1.0 -1.4 -0.9 3.2 1.3 Nonbank financing 0.0 -0.1 0.0 0.1 0.6 0.9 0.6 1.2 -0.3 -0.4 Arrears 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Debt reschedule 0.0 0.0 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 HIPC 0.0 0.0 0.4 0.3 0.5 0.0 0.0 0.0 0.0 0.0 Sources: Minitry of Finance (MEF), IMF and BOOST database. *Note: Annual data covers Haiti's fiscal year (Oct. 1 to Sep. 30) -22- 3.8 Wages have been driving the rise Figure 3.6 : Current Expenditure, 2005-14 in current expenditure. As discussed (In Percentage of GDP) previously, current expenditure has been 7 on the rise, increasing from just below 10 6 percent of GDP in 2005 to about 13 percent 5 GDP in 2014, in line with higher mobilization of own fiscal revenue. A higher 4 wage bill has been mostly driving this 3 development (Figure 3.6). Wages in 2 relation to GDP rose significantly from 2005 1 to 2014 (from 3.5 to about 6 percent of 0 GDP). This expansion reflects partly a 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 catching up. The share of wages and Wages and salaries Goods and services salaries in total current expenditure had Interest payments Subsides and transfers Source: Ministry of Finance (MEF) and IMF been declining from an average of below 45 percent in 2000–03 to an estimated average of about 35 percent in 2004–06. The decline in the wage bill had been particularly pronounced after 2000, reflecting the freeze on nominal wages in the public sector. At about 3 percent of GDP, the government wage bill during this period was below the levels of 6–7 percent of GDP observed in other low-income countries. Looking at the other current spending items, the cancellation of part of Haiti`s external debt through HIPC and MDRI in 2008 and further debt relief after the earthquake lowered interest payments. Outlays on transfers and subsidies, as well as government consumption of goods and services, were broadly declining before the earthquake, but have been on the rise since. Both spending items averaged about 2.5 percent of GDP over the period 2005 and 2010. This average increased to above 3 percent of GDP in the period following the earthquake (2011-14). 3.9 However, the high share of the wage bill in current spending limits the role of the budget as an effective policy tool. Even if the wage bill in terms of GDP in Haiti is in line with economies at similar levels of development, it takes up a significant portion of current spending, reflecting the country’s very low overall revenue and expenditure levels. Already back in 2005, total spending on wages and salaries was about 35 percent of public current outlays. In 2014, this ratio had increased to almost half of Haiti’s current budget (46 percent). The preponderance of wages and salaries in current expenditures translates into more rigidity, hampering the ability of fiscal policy to shift spending from one category to another if needed, as the expansion of the payroll increases mandatory spending as well as future expenditure on civil servants’ pensions. -23- 3.10 Furthermore, the composition of Figure 3.7 : Operating and Capital Expenses, 2005-14 public spending could also put into (In Percentage of GDP) question the viability of many 18 investments. As public investment has 16 increased, current spending has not 14 followed. Moreover, within this constant 12 envelope, operating costs more specifically 10 (e.g. equipment, utilities) have remained 8 broadly unchanged over the past decade 6 (Figure 3.7). Part of the growing wage bill 4 reflects increases in support, 2 administrative or security staff, and not in 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 staff directly delivering social services (see Operating expenses Capital expenses Chapters 6 and 7 for a more detailed Source: Ministry of Finance (MEF) and IMF discussion). Without the needed resources to operate the newly built investments and equipment, it is questionable whether these new facilities could be maintained or deliver the expected growth dividends. … And Greater Priority Given to Social Sectors … 3.11 Evidence suggests that greater priority has been given to social sectors. Assessing developments in public spending according to a functional classification is particularly challenging in Haiti (Box 3.2). Fragmentation of the fiscal data in terms of different budget classifications, periods, and stages of spending reported makes it extremely difficult to paint an overall picture of spending priorities over the past decade and any related shifts. Nevertheless, patching together the disparate available data for social spending suggests that the functional composition of public spending has grown closer to the country’s priorities as described in the first section of this Chapter. Overall, the share of resources allocated to social sectors seems to have expanded from an average of about 16 percent of total public spending for the period 2007 to 2010 to about 28 percent of total public spending after the earthquake, for the period 2010 to 2012. In addition, 22 percent of donor assistance on average is estimated to have gone towards social sectors over the period of 2010 to 2012 where data are available. -24- Box 3.2 : Fiscal Data Challenges Haiti does not have a consolidated financial statement. Fiscal data is widely scattered and has been a challenge for any work on public expenditure in Haiti over the recent past. As in many countries relying on aid, there is no consolidated financial statement combining current and capital spending. Information on current expenditure data is easily accessible and was extracted from Haiti’s expenditure management system, SYSDEP ( Système d’Informatisation des Dépenses). The data include all the administrative units of the central government. There is no information management system, however, recording public investment expenditures in Haiti. Data for domestically-funded investment expenditure was drawn from Budget Review Laws ( Lois de Règlement) and execution reports for the Public Investment Program (PIP) (Bilan d’Execution du PIP) provided by the Ministry of Economy and Finance and the Ministry of Planning and External Cooperation. The data includes projects financed from domestic sources for all the administrative units of the central government. For foreign-financed investments, though with some limitations, sectoral data exist on commitment and disbursements, collected by the Ministry of Planning and External Cooperation through the External Assistance Management Module (Module de Gestion de l’Aide Externe - MGAE) set up in 2009. Budget classifications are problematic. Haiti’s budget classification system does not contain an explicit module for the functional classification of expenditures. This means that the information system is unable to produce reports linking each budget line with a given function. Like in many countries the administration classification has to be used as a proxy. Moreover, a large portion of externally-financed expenditures is executed outside the budget system, with donors using their own implementation arrangements. As a result, the investment budget follows different classification depending on the sources of financing (Treasury i.e. own resources, Petrocaribe, and donors). Furthermore, on the domestically-funded part of the investment budget, there is no information on actual payments. While current spending can be followed from budget allocation to payment, only information up to the disbursements for specific domestically-funded projects is available. Recorded disbursements (décaissement) are transfers made to bank accounts held by individual line ministries for each project and hence do not necessarily imply final project expenditure execution. The information on the overall position of the government in the banking system tells us what level of spending has actually taken place and allows a discussion on broad trends according to the economic classification. Discussing trends along functional lines within this overall spending is much more challenging, however, because information on individual project account balances is not yet available. The ongoing rolling out the Single Treasury Account is expected to provide more detailed information on the payments made out of the project accounts and allow for a more detailed reporting of investment expenditure in the future. -25- 3.12 Treasury resources have been Figure 3.8 : Changes in Sectoral Composition – geared more towards social sectors. (In Percent of Total Treasury Spending) Looking at each source of financing in turn, 16 one observes that the share of the 12 country’s own fiscal resources managed by 8 the Treasury have been over the past decade increasingly allocated to the social 4 sector (including the ministries of health, 0 education and social affairs). This average -4 hides, however, wide differences (Figure -8 3.8). The period before the earthquake -12 (2007-2009) seems to have been -16 characterized by the need to strengthen 2007-2009 2009-2011 2011-2013 2007-2013 institutions and human capital with the Economic Sector Political Sector Social Sector Culture Sector social and political sectors seeing an Sources: Ministry of Finance (MEF) and IMF. increase in their share of spending (the latter includes the ministries of justice and interior). The earthquake triggered a clear shift in priorities towards the economic sector (including ministries in charge of reconstruction such as the ministries of planning and public works). More recently (2011-2013), with the need of reconstruction receding, the social and political sectors have regained attention. 3.13 Petrocaribe funds are reported to have been directed to infrastructure and, to a lesser extent, social programs. Data on Petrocaribe made available by BMPAD, even though it does not follow administrative or functional classification, gives an idea of the investment activities financed by these resources (Figures 3.9 and 3.10). From 2008 to 2013, Petrocaribe funds have been mostly allocated to infrastructure projects (58 percent), followed by social programs (9 percent), housing (7 percent) and agriculture (6 percent). The availability of Petrocaribe resources channeled to infrastructure may explain partly why Treasury financing shifted away from this sector. Figure 3.9 : Project Activities Financed by Petrocaribe Figure 3.10: Project Activities Financed by Funds, 2008-13, (USD Millions) Petrocaribe Fund, 2008-13, (Percentage of Total) 360 Housing 7% Social 9% Agriculture 300 6% 240 Others 7% 180 Territories 120 2% Infrastruct. Water & 60 58% Sanitation 4% 0 2008 2009 2010 2011 2012 2013 Economy Infrastructure Agriculture Others Energy 2% Territories Water & Sanitation Economy 5% Energy Housing Social Source: BMPAD Source: BMPAD -26- 3.14 Donor assistance seems also to have broadly favored infrastructure. Neither the Treasury nor the MEF monitor the expenditures financed by donors. The Ministry of Planning and External Cooperation aggregates information on the projects financed by donors through the Module de Gestion de l’Aide Externe (MGAE) set up in 2009 and started recording data from 2010 onwards. The MGAE system has its own limitations, however, as MGAE data does not always match data on official transfers in the Balance of Payment, or to the data reported by the IMF. Nevertheless, the data reported through the MGAE helps to reach a more detailed view of the composition of public expenditure. From 2010 to 2012, external aid has covered a broad range of sectors, but seems to have been mainly channeled to activities in infrastructure (infrastructure and transport), health and education (Figure 3.11). 20 3.15 The reliance of social sectors on Figure 3.11: Donors Financing, 2010-12 donor assistance for their financing (Percentage of Total) seems to have increased. While the Policies on Transport Health Others availability of donor assistance has health and Energie 6% 12% 1% population 4% allowed for financing of an expansion in 2% social sectors, it has also increased the Non specified reliance of these sectors on donor 8% Debt related 19% financing. Social sectors including health, Infra. & Social In-kind education and social protection are Services 9% 5% considered the most aid-dependent Agriculture sectors in the country. Overall, social Gov. & Civil 4% Society sectors relied on donor assistance for 16 7% percent of their financing in 2010.21 This Entreprises Education Water & Financial services 1% Sanitation ratio has expanded to 45 percent in 9% 7% Multi-sector 2% 6% Sources: Foreing Aid Management Database (MGAE) and World Bank Staff calculations 2012. This heavy reliance on donor financing makes social sectors particularly vulnerable to the decline in donor assistance and could put into question the sustainability of some programs (see Chapters 6, 7 and 8 for a more detailed discussion). … But Haiti’s Inclusiveness Remains Limited 3.16 Despite the recent increase in social public spending, Haiti’s economic system is not very inclusive. The Bank’s 1998 Poverty Report noted that "Haiti has never had a tradition of governance aimed at providing services to the population or creating an environment conducive to sustainable growth.” A number of Haitian and international observers broadly agree that the Haitian State remains ineffective and delivers little to its population. “Haitian institutions have never provided justice, education or healthcare to the majority of the population” (Lockhart and Forman, 2013). Instead, a small economic elite has supported a state that makes only negligible investments in human resources and basic infrastructure (World Bank, 1998). “Governance and state capacity to effectively formulate and implement sound policies, and to deliver core public services to the population, are weak. Furthermore, 20 Debt relief is another important component (14 percent). 21 Social sectors have been defined for this purpose as including the following MGAE classifications: education, infrastructure and social services, policies on health and population, and health. -27- the state is present largely in the major urban centers and has been unable to provide basic services or infrastructure to large portions of the population” (Buss, 2013). 3.17 Social spending remains limited Figure 3.12: International Comparison - Social Spending, 2013 and the delivery of basic services highly or latest (In Percentage of GDP) inequitable. Despite the recent increases 25 in public spending in social sectors, public spending in health, education, and social 20 protection amounts to 5 percent of GDP, 15 below comparator countries, limiting the 10 government’s ability to offer equal opportunities to its citizens (Figure 3.12). 5 At the same time, many large spending 0 items such as fuel subsidies clearly favor the rich (Chapter 9). In the absence of government, basic services such as health and education are mainly provided by Sources: World Bank and Ministry of FInance (MEF) non-government actors (Chapters 6 and 7). Seventy to eighty percent of primary school students attend non-public schools, placing a substantial financial burden on households and delivering achievements closely linked with household income. Outcomes are equally unfavorable to the poor in the health sector. -28- TOWARDS GREATER AND MORE EQUITABLE REVENUE MOBILIZATION 22 This Chapter reviews recent trends in fiscal revenue. Haiti has managed over the past decade to increase its fiscal revenue mobilization ratio. This ratio remains, nevertheless, relatively low, revenues originate mainly from the Metropolitan area or at the borders, and the overall tax system is regressive. Exemptions are eroding the tax base, and would call for a review and close monitoring. Personal income tax thresholds do not refl ect the country’s socio -economic structure with the highest marginal rate applied only to a small number of taxpayers. In addition, although Haiti has very low trade tariffs, it levies substantial inspection fees. Despite Some Increase, Revenue Mobilizati on Remains Low and Regressive… 4.1 While own fiscal revenue has increased over the past decade, Haiti has still one of the lowest revenue mobilization rates in the region. As mentioned in Chapter 2, fiscal revenue as share of GDP both from domestic sources and trade has increased over the recent past (Figure 4.1). This improvement has been driven primarily by an increase in domestic income and sales tax revenue collected in Port-au-Prince. In addition, a new tax instrument was introduced in 2012: the Fond National d’Education (FNE) which is financed through taxes on international telephone calls and money transfers with the purpose to support the Programme de Scolarisation Gratuite et Universelle (see Chapter 7). Despite this progress, Haiti has still the second lowest revenue-to-GDP ratio for all countries in the region (Figure 4.2). Increasing the tax- to-GDP ratio is a common part of moving towards higher levels of per capita GDP. Cross-country data nevertheless suggest that Haiti currently performs slightly below average when compared to countries with similar per capita income levels. 22 Prepared by Jan Loeprick and Erik von Uexkull. -29- Figure 4.1 : Fiscal Revenue, 2009-13 (In Percent of GDP) 14 12 Other (inc. FNE) 10 Customs duties (inc. inspection fees) 8 Other taxes Sales tax 6 Excise taxes 4 Domestic taxes in provinces 2 Income taxes (P-au-P only) 0 2009 2010 2011 2012 2013 Sources: Ministry of Finance (MEF), IMF, BOOST database and authors' estimate. 2009 2010 2011 2012 2013 Fiscal Revenue 11.2 11.8 12.8 12.8 12.7 Domestic taxes 7.4 7.3 8.1 8.6 8.0 Income taxes (P-au-P only) 2.3 2.2 2.5 3.0 2.6 Domestic taxes in provinces 0.5 0.4 0.4 0.5 0.5 Excise taxes 0.7 0.5 0.3 0.3 0.3 Sales tax 3.5 3.2 3.7 3.7 3.7 Other taxes (P-au-P only, inc. discrepancies) 0.5 0.9 1.2 1.0 1.0 Customs duties (inc. inspection fees) 3.3 4.3 4.5 4.2 3.9 Other (inc. FNE) 0.4 0.2 0.3 0.1 0.8 Sources: Ministry of Finance (MEF), IMF and BOOST database Figure 4.2 : Tax-to-GDP Ratio (a) Tax / GDP Regional Comparison (b) Tax-to-GDP relative to Per Capita GDP (Percent) (Percent and Log of PC Income) 30 35 25 30 20 25 15 Tax revenue / GDP 20 10 5 15 0 10 Haiti 5 7 8 9 10 11 12 natural log of pc GDP Sources: Haitian Revenue Authorities and World Bank, World Development Indicators. Sources: Haitian Revenue Authorities and World Bank, World Development Indicators. 4.2 Furthermore, the regional distribution of revenue collection is heavily skewed towards Port-au- Prince and border posts. The Tableau des Opérations Financières de l’Etat (TOFE) currently only provides -30- a rudimentary breakdown of the geographic origin of fiscal revenues in the sense that all revenues collected in the provinces are lumped together in one category with no distinction by instrument. This limited data indicates that in 2013 only approximately 4 percent of total revenue was collected in the provinces. While the largest share of revenue (60 percent) is collected at the border through customs duties, inspection fees, and sales and excise taxes on imported goods, the remaining 36 percent are collected in Port-au-Prince. 4.3 Haiti’s tax system generates not only limited resources for the government, but also tends to be regressive. Haiti’s fiscal revenues rely heavily on indirect taxes which affect consumers independently of their income level. The ratio of direct to indirect taxes stood at about 30 percent in 2011, a level inferior to that of most countries in LAC and to the average of Low-Income Countries, and is largely explained by the fact that a sizable share of Haiti’s revenues come from international trade (Figure 4.3). Tax systems relying relatively more on direct taxes tend to be more progressive because in such systems the burden of taxes weighs differently on economic agents of varying income levels. Moreover, while direct taxes tend to be harder to administer, they are frequently seen as an important indicator and tool of successful state-building as they tend to create a more direct citizen-state interaction and voice (Di John, 2010; OECD 2014). Figure 4.3 : Ratio Direct To Indirect Taxation, 2009 or 2011 (Percent) Panama Low income countries Jamaica Costa Rica Honduras Dominican Rep. Nicaragua Haiti 0 20 40 60 80 Sources: IMF and World Bank 4.4 Compared to other countries in the region, Haiti relies heavily on customs duties for revenue generation. Revenue from customs duties and a 5 percent inspection fee collected on imports amount to about 4 percent of GDP, more than twice the regional average of 1.8 percent of GDP, and also significantly above the ratios of neighboring countries with similar overall revenue-to-GDP ratios like Costa Rica and Dominican Republic (both at 1.1 percent of GDP). Streamlining trade policy and its application with a view to reduce distortive effects, and enhance border efficiency and transparency will be important to ensure future growth and productivity, and a number of trade policy reforms are currently being debated. However, given the importance of tariffs and fees collected at the border, such reforms must also be guided by a careful assessment of their potential effects on fiscal revenue, a topic which is further explored in this chapter. -31- … With Exemptions and Ill -Designed Tax Brackets Eroding Direct Taxes … 4.5 A high share of corporate income tax revenue is lost because of exemptions. Corporate and personal income tax rates are largely comparable to the regional average (Figure 4.4) and cannot explain Haiti’s low tax burden. For instance, Haiti’s corporate income tax rate of 30 percent is slightly above the regional average (the dotted line) and above the rates in most regional comparators with higher tax revenue-to-GDP ratios. Exemptions, however, are estimated to have amounted in 2011 to the equivalent of about 63 percent of all income tax collected.23 Most exemptions derive from the 2002 investment code that grants 15-year-tax exemptions for companies in free zones as well as 5-10 year exemptions for specific investment projects that are considered desirable from a development perspective. Figure 4.4 : Corporate (left) and Personal (right) Income Tax Rates in Regional Comparison (Percent) 40 35 30 25 20 15 10 5 0 Source: USAID ‘Collecting Taxes’ 2013 dataset. 4.6 International experience suggests that tax exemptions should be used with caution as a means to attract FDI, especially in countries with challenging investment climates. Tax incentives to attract investment to developing countries are often seen as needed to compensate investors for unfavorable investment conditions and limited provision of public goods. International experience suggests, however, that this ‘substitution effect’ often does not materialize in practice and that tax incentives can create windfalls for investments that would have taken place in any case (Box 4.1). According to the World Bank’s Doing Business Indicators, Haiti ranks particularly low on indicators such as starting a business (187th out of 189) and protecting investors (170th) that are of strong concerns for potential investors, but can be resolved through policy reforms that are less costly and potentially more effective in attracting investors. In addition, IMF (2013) identifies several aspects of the current tax code – rate and calculation method of the tax on dividends, non-deductibility of foreign taxes on international income, strong audit requirements, and high upfront-payments of taxes – that if reformed are likely to have a positive impact on investment. Under these circumstances, reforms to improve the business climate are likely to have a stronger impact in attracting FDI without the adverse of tax exemptions. 23 IMF (2013) -32- Box 4.1 : International Experience with Tax Incentives Every investment incentive policy has potential costs and benefits. Benefits may arise through higher revenue from possibly increased investment and social benefits such as jobs, positive externalities, and signaling effects from this increased investment. The costs are due to revenue losses from investments that would have been made even without the incentives and indirect costs such as economic distortions and administrative and leakage costs. It is difficult to quantify these elements, but trying to do so provides a useful conceptual tool for policymakers analyzing the general framework for incentives as well as targeted incentives for anchor investments, export-oriented and mobile investments, extractive industries, and so on. A key factor in the cost – benefit equation is the question of additionality (‘Would a given investment have occurred in any case had no incentive been granted?’). If the answer to this question is yes, the benefits that can be observed in relation with the investment cannot be associated with the incentive that was provided. On the other hand, the costs of the incentive are real regardless of whether the incentive was effective or not. While most international studies find a significant negative relationship between the marginal tax rate and investment inflows in developed economies, results for developing countries differ. The balance of evidence suggests that, for many developing countries, fiscal incentives do not effectively counterbalance unattractive investment conditions such as poor infrastructure, macroeconomic instability, and weak governance and markets. The Figure below shows that for countries with weak investment climates, a lower marginal effective tax rate (METR) has limited impact on FDI. For example, having an METR of 20 percent instead of 40 percent raises FDI by 1 percent of GDP for countries ranked in the bottom half in terms of investment climate —while the same difference in METR has an effect eight times greater for countries in the top half. Effectiveness of Tax Incentives for Investment Attraction To confirm this finding, the World Bank’s Investment Climate Advisory conducted econometric studies and surveys of investors in developing countries. The econometric studies overwhelmingly conclude that the investment climate is more important than tax breaks or other nontax incentives. The surveys found that factors related to the investment climate—such as ease of import and export, availability of local suppliers, regulatory framework, adequate infrastructure, and the country’s geographic location—rated higher than incentives as a primary motivation for investment. For Thailand, 81 percent of investments would have been made even without incentives. In Jordan, Mozambique, and Serbia 70 percent or more of investments would have been made anyway, so incentives were redundant. Such surveys also reveal, however, that tax incentives are more important for exporters who operate in tight international competition and tend to be internationally mobile. Source: James (2013) -33- 4.7 Where tax exemptions or other investment incentives are used, a clear framework needs to be in place to measure and track their costs. Work is in progress in Haiti to measure more thoroughly and track foregone revenue from tax exemptions. This is a welcome step. In the absence of such information, decision makers may not fully factor costs from taxes that were never collected (and thus do not appear in the budget) into their decisions. In addition to foregone revenue, tax exemptions can also have indirect negative effects, for instance through administrative costs, encouragement of rent seeking behavior, and creation of economic distortions (James, 2013). 4.8 At the same time, it is important to track and monitor the benefits of tax exemptions. In order for the policy maker to decide whether exempting an investor was a good decision or not, not only the amount of the foregone fiscal revenue should be identified, but also the expected benefits regularly monitored. This applies to benefits such as employment creation, technology transfer, or demonstration effects. In doing so, a critical step of analysis is to determine whether a given investment is actually a result of the tax exemption, or would have occurred in any case. 4.9 Furthermore, anecdotal evidence suggests that a number of taxes exist in Haiti that generate little revenue but impose high compliance cost on businesses (“nuisance” taxes). While such taxes generate very little revenue, they can have adverse effects on the tax system as a whole by encouraging informality and adding complexity to tax administration. Removing nuisance taxes can therefore be an effective complement to ongoing reform efforts to strengthen tax administration, while at the same time contributing to a more favorable business environment. Box 4.2 presents international experience for the identification and removal of nuisance taxes. -34- Box 4.2 : Removing Nuisance Taxes So called nuisance taxes are instruments that raise very little revenue, but can impose substantial compliance burdens on companies. Administering a large number of different instruments for the same taxpayer population is by definition inefficient, and often the revenue generated by such taxes barely covers their administrative cost. Anecdotal evidence suggests that a number of nuisance taxes currently exist in Haiti. In order to reduce compliance cost of firms and use limited tax administration capacity more efficiently, it is therefore important to complement tax administration reform efforts with a review of tax instruments to identify and remove such nuisance taxes. International experience suggests that despite negligible revenue collected from nuisance taxes, a reform tends to be a difficult undertaking as it can involve various entities and stakeholder with often conflicting interests. The overall reform objective should be the combination of as many taxes applied to the business sector as possible into a single tax administered and collected by the same agency. Such a reform typically requires: (i) thorough review of all current instruments, their tax base, legal basis, and revenue collected. Such an exhaustive inventory is needed for all subsequent decisions on the removal and streamlining of instruments; (ii) the definition of clear criteria to streamline these instruments. For example, all taxes which serve solely to generate revenue and are not directly related to a service could be assessed based on the average revenue generated by each instrument over the last 3 years. When intakes have been negligible and below a certain threshold, their elimination could be envisaged; and (iii) the development of transfer mechanisms to ensure sufficient resource flows to line ministries and local institutions that previously levied their own taxes. In this process, user charges that are directly related to a service, or fulfill a regulatory function would have to be distinguished from those that only serve to generate revenue. Such taxes related to direct services or a regulatory function would have to be evaluated against more sophisticated criteria (i.e. justification of regulatory function, necessity/quality of services provided). In a second phase, taxes would actually be consolidated to reduce significantly the number of instruments and thus the interaction with tax officials. The most drastic reform option for achieving this is based on the reversal of the burden of proof for the existence of a tax instrument. 24 This approach has been pioneered in Sweden and subsequently adopted in several countries (Hungary, Mexico, Kenya), and has been further developed by the World Bank’s Foreign Investment Advisory Service for reform licensing regulation. A reversal of the burden of proof would mean that line ministries, local authorities and other “owners” of tax instruments would need to prove that their taxes match pre-defined criteria for tax instruments. If they fail to justify their taxes, or do not notify its existence past a set deadline, these taxes would be eliminated by default. This mechanism reverses the incentives to share data and information required to identify nuisance taxes. All ‘surviving’ taxes would be listed in a centralized registry. The registry wou ld have positive legal security: no regulation not in the registry can be enforced against a business. 4.10 Personal income tax rates are in line with the regional average, but apply only to very few taxpayers. There are currently four tax bands for personal income tax: 10 percent on incomes above HTG 20,000 and below 100,000; 15 percent up to HTG 250,000; 25 percent up to HTG 750,000; and 30 percent for everything above that (Figure 4.5). These thresholds are high relative to per capita GDP compared to other countries in the region: The maximum rate is only levied at incomes above more than 2000 percent 24 The following discussion of the reversal of proof principle draws upon FIAS (2010) -35- of per capita GDP, more than twice the ratio for the next highest comparator country (Guatemala at 757 percent). In other words, only very high incomes are taxed at the maximum rate. Even the threshold for the second highest income tax band at 25 percent exceeds the regional average for applying the maximum rate (the dotted line), and the majority of regional comparators apply thresholds that even fall below the third highest tax band of 15 percent in Haiti. Figure 4.5 : Income Tax Rate Thresholds In Regional Comparison (Per Capita GDP) 2100 1800 1500 1200 900 600 300 0 Sources: USAID ‘Collecting Taxes’ 2013 dataset and calculation by authors. 4.11 Household data suggests that most fiscal revenue could be generated at the top end of the income distribution. Figure 4.6 shows average monetary income by 0.2 percentiles of the working age (15+) population. 62 percent of the working age population report no monetary income at all. Another 14.4 percent report monetary income below the minimum income tax threshold of HTG 20,000. 16.2 percent report monetary income that would put them in the 10 percent income tax bracket, and the equivalent shares are 5.2 percent, 1.6 percent, and 0.6 percent respectively for the 15, 25, and 30 percent brackets. If perfect income tax collection and no strategic response in taxpayer behavior were to be assumed, this would imply that 11 percent of income tax revenue would be generated by taxpayers in the 10 percent bracket, and 20, 24, and 46 percent respectively by taxpayers in the 15, 25 and 30 percent brackets.25 This calculation illustrates the importance of ensuring the actual collection of income tax revenue in particular at the top end of the income distribution. Perfect tax collection is a strong assumption that is unlikely to hold in practice as income taxes tend to be collected mainly on incomes in the formal sectors and public service. In addition, household survey data tends to be less accurate at the top end of the income distribution where outliers in terms of very high incomes are not easily captured. Figure 4.6 is thus merely an illustration of how the income tax brackets are aligned to the socio-economic structure of the population. A more thorough review of income tax brackets would rely on a combination of household survey data and actual income tax information to factor in collection efficiency, and gain a more accurate picture of top income earners. 25 The strategic response by businesses and individuals to tax policy thresholds has been investigated in a number of studies. See for instance, Kleven and Waseem (2013) and Bruhn and Loeprick (2014) looking at thresholds in Pakistan and Georgia. -36- Figure 4.6 : Hypothetical Income Tax Revenue Collection Based on 2012 Household Survey Results (Tax Rate in Percent and Income in Millions of Gourdes 4.12 There might be potential for expanding revenue mobilization from the top 10 percent of the income distribution by reviewing tax brackets. Figure 4.7 shows results of a simple illustrative calculation based on the 2012 household survey, as in the previous paragraph based on the assumption of full tax compliance and no strategic response in taxpayer behavior, and serving for illustrative purposes only. Two reform scenarios are applied: Example 1 simply lowers the threshold for the 30 percent maximum rate from HTG 750,000 to HTG 350,000. The effect is a relatively small increase in overall revenue collection (about 4 percent) which is borne entirely by tax payers in the top 1 and 96-99th percentile of income distribution that are moved from the 25 to the 30 percent bracket. A more extensive reform scenario is analyzed as example 2, which includes not only the same threshold for the 30 percent band as example 1, but also a general lowering of minimum thresholds with the exception of the HTG 20,000 threshold for the 10 percent rate. It also introduces a fourth bracket at 20 percent to close the relatively large gap between the 15 and 25 percent bands. The effect is much more substantial in terms of the additional revenue generation (+ 13 percent). In terms of its distributional impact, taxpayers below the 90th income percentile would still remain unaffected, but the impact is spread more evenly across taxpayers in the top 10 percentile. 4.13 Further analytical work could help review income tax brackets under equity and efficiency perspectives. While household data can give an indication of the alignment of tax brackets with the country’s socio-economic reality, further analysis of income tax collection could focus on reviewing tax collection data by income groups. Such analysis could aim to identify measures of collection efficiency, in particular for high income taxpayers with high revenue collection potential, and inform a potential review of income tax brackets in terms of equity (ensuring a ‘fair’ distribution of the tax burden) and efficiency (revenue generation relative to administrative cost). -37- Figure 4.7 : Hypothetical Effect of Changes in Income Tax Brackets on Revenue Collection by Income Group … And Indirect Taxes Hampered by Weak Revenue Administration and Unclear Exemptions 4.14 Customs administration seems to suffer from significant performance shortfalls. Since about 60 percent of all tax revenue is collected by the customs authorities at the border, the ability of the customs to collect these taxes efficiently and effectively is critical for overall revenue mobilization. A recent World Bank analysis using the Customs Assessment Trade Toolkit (CATT) concluded that customs procedures in Haiti were moving closer to international good practices (World Bank, 2014). Critical indicators of practical performance such as efficiency, degree of control, and trade facilitation were, nevertheless, underperforming (Figure 4.8 and Box 4.3). The CATT analysis highlighted several areas of recent or ongoing improvement, and an ambitious reform and modernization agenda is currently under implementation. Among the priorities for improvement in the CATT report are the better systemization of available information, enhanced coordination between sectors, the development of a manual of operational procedures, improvement of internal communication, and upgrading of available infrastructure. The customs administration relies on the globally used SYDONIA (ASYCUDA in English) software for tracking and managing customs data, but some of its more advanced functions are not fully exploited, and transactions are not always registered in real time. -38- Figure 4.8 : Results of Haiti Customs Assessment: Distance from International Best Practice process orientation 100 80 transparency 60 strategic thinking 40 20 0 facilitation control effectiveness efficiency Source: World Bank (2014) Box 4.3 : The CATT Methodology for Benchmarking Customs Performance The Customs Assessment Trade Toolkit (CATT) is an integrated tool for measuring customs performance across countries and over time. The tool provides relevant information on customs performance by collecting data and information from different countries against a set of 120 high level indicators related to customs’ operations and producing performance benchmarks. With these, the tool can be used to evaluate the relative strengths and weaknesses of any customs administration and compare them to good practices. The distinctive feature of the CATT is that all data is objectively verifiable and thus measurements are comparable across countries and across time periods. The CATT does not simply measure the formal use of good practices, as it might be possible that most customs administrations around the world operate within a legal framework that promotes the use of simplified procedures. The CATT also considers the performance of a customs office, such as how long in fact customs administrations take to control declarations. To address this question, the real time that goods take to be released is measured. To get this information, the CATT evaluation exercises reviews to determine how the customs administration operated in practice. For example, the release of goods is compared against reasonable international standards. The CATT team has picked good practices from the existing literature on customs good practices in a wide range of areas, including simplified procedures; selective controls; risk analysis; advanced data/entry level; electronic data submission; increased use of information technology; improved management of special customs procedures; enhanced transparency and partnership with private sector; infrastructure; process orientation; and public service orientation. The 120 measurements (called high level benchmark indicators) derived from these areas are grouped into seven dimensions (Process Orientation, Strategic Thinking, Control, Efficiency, Effectiveness, Facilitation, Transparency) that summarize the overall state of a customs office, creating a quantifiable view of the overall practice and performance of a customs office. Source and further information: www.customscatt.org 4.15 Substantial fiscal revenue shortfalls also occur at the border. While no complete data is available on TCA and excise tax exemptions, detailed transaction level data on taxes collected at the border makes it possible to calculate the value of exemptions on imported products. The foregone revenue can be estimated by comparing the actual revenue collected to the amount an importer should have normally -39- paid, had the official rate been applied. The results suggest that In 2013 Haiti gave up the equivalent to 14 percent of total sales tax revenue collected at the border, plus another 19 percent in ad valorem excise taxes.26 For tariffs, the reserve shortfall rises to 50 percent (Table 4.1). The bulk of foregone TCA revenue is from products that are either industrial inputs (21 percent of revenue lost) or capital equipment (23 percent of revenue lost). Exemptions on capital goods can be beneficial in terms of promoting investment, while exemptions on industrial inputs can be absorbed as windfall profits rather than passed on to consumers if competition is weak. As with business tax exemptions, carefully monitoring the costs versus the expected benefits of such exemptions is key for policy makers to understand their full impact. 4.16 The classification of exemptions in customs data is insufficient to thoroughly track their justification. Typically, customs procedure codes would give a clear indication of the legal basis for the exemption (i.e. import to free zone, exemption under a specific paragraph in the investment code) that makes it possible to use this data to track the costs of specific exemption rules. In the case of Haiti, most exempted imports are simply classified as “exempt of all taxes” or exempt of specific tax instruments. Exemptions attributed to foreign embassies, NGO and government entities account for only a small share of total exemptions. It is also noteworthy that close to 5 percent of exemptions are granted on imports that are classified as “fully taxable”. Improving the classification of exemptions to more thoroughly explain the justification for a given exemption would be an important first step for better monitoring of exemption costs. The same principles discussed above regarding the costs and benefits of TCA and excise exemptions apply to tariff exemptions. … But Could Be Bolstered By a Simple VAT 4.17 Important efficiency gains could be expected from a move towards a regular VAT instead of the turnover tax (TCA) currently applied. Because the deductibility for the TCA is limited, the instrument can be applied to consecutive sales, including inputs. When the input taxes are not refunded, this leads to tax accumulation and tax on tax (cascading). Depending on the production chain, the cumulative tax burden could thus become very high in cases where no deductibility is allowed, which renders affected goods and services less-competitive. A particular concern is that the non-deductibility of domestic inputs sourced by exporters undermines efforts to promote linkages and the goal of strengthening the overall competitiveness of the Haitian export sector. Technical work to determine the transition of the TCA towards a regular VAT (e.g. removing deductibility restrictions, building a sound refund administration, introducing a zero rate for exporters) and a schedule for its implementation is currently under way. 4.18 International experience suggests that introducing a VAT can lead to substantial increases in tax revenue for low and lower middle income countries if it is designed properly. Data from a sample of nine low and lower middle income countries that have introduced VATs since the 1990s is summarized in Figure 4.9. It shows that on average, tax revenue as a share of GDP increased from around 13 to 14.6 percent of GDP within three years after the introduction of the VAT. As VATs are sometimes introduced in the context of larger tax reforms, it is also important to note that the share of taxes on goods and services also increased significantly over the same time period, indicating that VAT revenue is indeed contributing to 26 Exemptions on non-ad valorem excise taxes cannot be calculated with available data, but these taxes account for a relatively small share of total revenue. -40- the positive change. However, there is substantial variation in the data, suggesting that these positive outcomes are not automatic with the introduction of a VAT (Figure 4.9). Figure 4.9 : Tax Revenue Collection around the Introduction of VAT (Percentage of GDP and of Total Revenues) 16 36 Introduction 35 15 of VAT 34 33 32 14 31 30 13 29 28 12 27 -3 -2 -1 0 1 2 3 Tax revenue, % of GDP (lhs) Revenue from taxes on goods and services, % of total (rhs) Source: Author’s calculation based on data from World Bank / World Development Indictors and USAID ‘Collecting Taxes’ 2013 dataset. 4.19 The implementation strategy for the reform process is critical for success. A number of failed VAT introductions can be explained by political challenges and poor planning, in particular regarding administrative capacity constraints. Ghana, for instance, repealed a VAT a few months after its initial introduction in 1995. In 1997, the VAT was successfully re-introduced, following more extensive preparation of the revenue administration and business community and relying on a significantly higher mandatory registration threshold to limit the number of taxpayers in the regime. More generally, the establishment of self-assessment procedures, effective enforcement tools, a functioning refund-process for exporters and sufficient capacity to perform audits are critical for the successful implementation of a VAT (Grandcolas, 2005). 4.20 Its design should be simple. General recommendations in designing VAT systems typically include a simple (single or, sometimes, dual) rate structure to facilitate administration and compliance, and a comparatively high mandatory registration thresholds to exempt the majority of small traders (Gnossen, 1992, International Tax Dialogue, 2005, Kloeden, 2011). However, based on the extensive country experience in adopting VAT, it has also been noted that not all these design features are “attainable or even desirable in the context of a particular country at a particular time” (Bird and Gendron, 2011). Some flexibility regarding the level of exemption thresholds, use of exemptions of multiple rates may thus be required in the political process to build support for a VAT. … And Taxes on International T rade Could Be Streamlined 4.21 While Haiti is one of the most open economies, trade taxes account for a substantial amount of fiscal revenue. Trade taxes from customs duties and inspection fees are the largest source of fiscal revenue, jointly accounting for one third of total government revenue (Figure 4.1). While Haiti is one of the most open economies to trade with very low tariffs, it imposes various fees that add up to the cost of -41- importing goods into the country. A 5 percent inspection fee is applied to almost all imported products increasing tariff revenues by about 75 percent in 2013. 4.22 Ongoing discussions on tariff changes are unlikely to lead to significant revenue effects. The 2013 budget was proposing a number of tariff adjustments. The proposed tariff changes would have affected one fifth of total tariff lines at 8 digit level. The revenue implications of these tariff adjustments were simulated using the World Bank’s Tariff Reform Impact Simulation Tool (TRIST) (Box 4.4). The results suggest that the revenue gains from this reform would be limited, equivalent to about 4 percent increase in tariff revenue (scenario 1 in Table 4.2 and Figure 4.10). Changes in the level of protection would also be limited for most sectors, but the highest increase in protection by 1.3 percentage points is for capital goods, which may have detrimental effects on innovation and productivity growth in the longer term (Eaton and Kortum, 2001). 4.23 Alignment of Haiti’s tariffs to the current CARICOM Common External Tariff could generate substantial additional revenue, but also substantially raise protection and costs for firms and consumers. Haiti is a member of CARICOM, but applies lower tariffs. As shown under scenario 2 in Table 4.2, an alignment with the regional external tariff would lead to a substantial increase in tariff revenue for Haiti (30.1-47.7 percent) while revenue from inspection fees and excise taxes would decline as imports decline. The effect on TCA revenue would be ambiguous due to the fact that TCA is levied on the tariff inclusive value of imports. Depending on the reactiveness of imports to the change in tariffs, the effect could thus be slightly positive or negative. The total change in revenue collected at the border is expected to increase between 7.6 and 14.8 percent. However, it shows that this would come at the cost of a substantial increase in protection, and hence domestic prices, in particular for products in the category food and beverages (both for household consumption and industrial inputs). This reform scenario does not include any change in the 5 percent inspection fee, suggesting that while Haiti would align its tariffs with regional neighbors under the CET, actual levels of protection including the inspection fee would be substantially higher. 4.24 Inspection fees could be replaced with a combination of higher regular tariffs and the removal of exemptions with no loss in total revenue. As explained in the previous paragraph, inspection fees in the magnitude applied by Haiti mask the true level of protection in the country. Being applied across the board with an identical rate for almost all products, they are unlikely to be in line with the country’s strategic trade objectives that would normally suggest placing a higher tariff on consumer goods and lower tariffs on key inputs and capital goods. Scenarios 3 and 4 therefore show different illustrative examples for replacing the inspection fees with tariffs. Scenario 3 is a proportionate 75 percent increase on all existing tariffs, while scenario 4 combines a 50 percent reduction in all tariff exemptions with a proportionate 20 percent increase on all existing tariffs. Both scenarios are designed to have a minimal effect on total revenue (Table 4.2), but Figure 4.10 shows that their impact on protection would vary significantly across sectors. Compared to the status quo inclusive of inspection fees, scenario 3 would significantly reduce protection for industrial inputs and capital goods at the expense of higher tariffs on consumer goods. Scenario 4 would provide for lower tariff protection on all sectors except for fuels and lubricants, which account for the major share of tariffs exemptions to be removed under this scenario. -42- Table 4.1 : Customs Duties and Inspection Fees (2013) Import value Customs duties Inspection fee hypothetical hypothetical hypothetical revenue weighted amount of revenue hypothetical amount of revenue without effectively tax rate foregone share of revenue without effectively weighted tax foregone share of collected exemptions applied tax without revenue foregone collected exemptions applied rate without revenue foregone (HTG mln.) (HTG mln.) (HTG mln.) rate exmptions (HTG mln.) revenue (HTG mln.) (HTG mln.) tax rate exmptions (HTG mln.) revenue 11 - Food and beverages (industrial use) 24,529 617 674 2.5% 2.7% 57 8.4% 1,225 1,226 5.0% 5.0% 2 0.2% FROMENT (BLE) DUR 3,783 113 132 3.0% 3.5% 20 14.9% 189 189 5.0% 5.0% 0 0.0% GRUAUX ET SEMOULES DE FROMENT(BLE) 385 0 13 0.0% 3.5% 13 99.2% 19 19 5.0% 5.0% 0 0.2% SUCRE CHIMIQMT PUR,A L'ETAT SOLIDE YC LE SUCRE INVERTI OU (INTERVERTI) 102 3 15 2.7% 15.0% 13 81.8% 5 5 5.0% 5.0% 0 0.0% 12 - Food and Beverages (household consumption) 22,181 2,355 2,647 10.6% 11.9% 292 11.0% 1,102 1,109 5.0% 5.0% 7 0.7% MORCEAUX ET ABATS DE COQS ET POULES, CONGELES 2,672 580 668 21.7% 25.0% 88 13.1% 133 134 5.0% 5.0% 1 0.4% PATE CONCENTREE DE TOMATE NON CONDITIONNEE POUR LA VENTE AU DETAIL 64 3 13 4.5% 20.0% 10 77.3% 3 3 5.0% 5.0% 0 0.0% PREPARATIONS POUR SAUCES...CONDIMENTS, ASSAISONNEMENTS. 157 25 31 16.1% 20.0% 6 19.7% 8 8 5.0% 5.0% 0 0.3% 2 - Industrial supplies not elsewhere specified 30,426 1,489 2,110 4.9% 6.9% 621 29.4% 1,460 1,521 4.8% 5.0% 61 4.0% FLACONS,BONBONNES ET ARTICLES SIMILAIRES EN MATIERES PLASTIQUES 748 14 187 1.9% 25.0% 173 92.3% 37 37 5.0% 5.0% 0 0.8% BOUTEILLES,BONBONNES,FLACONS...EN VERRE. 667 27 100 4.0% 15.0% 73 73.4% 33 33 5.0% 5.0% 0 0.1% AUTRES BOUCHONS 238 7 60 2.9% 25.0% 53 88.4% 12 12 5.0% 5.0% 0 0.2% 3 - Fuels and lubricants 35,515 879 7,121 2.5% 20.1% 6,242 87.7% 523 1,765 1.5% 5.0% 1,242 70.4% ESSENCE POUR MOTEURS (GAZOLINE) NAPLITE ET BENZITE. 12,245 837 7,077 6.8% 57.8% 6,241 88.2% 275 612 2.2% 5.0% 337 55.0% PREPARATION LUBRIFIANTE POUR MATIERES AUTRES QUE CEUX DU NO 340311. 26 0 1 1.3% 5.0% 1 74.9% 1 1 5.0% 5.0% 0 0.0% HUILES MINERALES SPECIALES ET AUTRES HUILES LUBRIFIANTES COMPOSEES. 809 40 40 5.0% 5.0% 0 0.4% 40 40 5.0% 5.0% 0 0.2% 4 - Capital goods (except transport equipment), and parts and accessories thereof 9,976 237 308 2.4% 3.1% 71 23.1% 425 494 4.3% 5.0% 69 14.0% PARTIES ET ACCESSOIRES DES MACHINES DU N. 84.71 115 1 6 0.7% 5.0% 5 85.8% 1 6 0.7% 5.0% 5 85.8% GROUPES ELECTROGENES D'UNE PUIS. >375 KVA 104 2 5 1.9% 5.0% 3 62.8% 4 5 3.9% 5.0% 1 22.7% PARTIES DES MACHINES DES NOS 85.01, 85.02 79 1 4 1.0% 5.0% 3 79.3% 1 4 1.0% 5.0% 3 79.2% 5 - Transport equipment and parts and accessories thereof 9,393 737 831 7.8% 8.8% 94 11.4% 432 470 4.6% 5.0% 38 8.1% VEH. TOUT TERRAIN , CYL. > 2500 CM3 , DIESEL , NEUFS 1,114 78 111 7.0% 10.0% 33 30.0% 44 56 4.0% 5.0% 12 21.0% ACCUMULATEURS AU PLOMB AUTRE QUE DU 850710 203 17 41 8.6% 20.0% 23 57.0% 4 10 2.1% 5.0% 6 57.0% VEH.TOUT TERRAIN , CYL. >= 2200 CM3 , MOINS DE 2 TONNES , DIESEL , NEUFS 788 32 39 4.0% 5.0% 8 19.8% 35 39 4.4% 5.0% 4 11.2% 6 - Consumer goods not elsewhere specified 12,235 1,199 1,279 9.8% 10.5% 80 6.3% 571 612 4.7% 5.0% 40 6.6% IMPRIMES AUTRE QUE DU NO 491191 103 9 21 8.3% 20.0% 12 58.5% 3 5 2.8% 5.0% 2 44.4% MEUBLES EN PLATIQUES AUTRE QUE CEUX DU NO 940370 11 61 5 12 8.6% 20.0% 7 57.1% 1 3 2.1% 5.0% 2 57.1% OUVRAGES EN MATIERES PLASTIQUES ET MATIERES DES N.3901 A 3914 NDCA 130 8 13 6.0% 10.0% 5 39.7% 6 6 4.8% 5.0% 0 3.1% 7 - Goods not elsewhere specified 72 0 0 0.2% 0.2% 0 0.4% 0 0 0.1% 0.1% 0 0.4% Grand Total 168,922 7,513 14,970 4.4% 8.9% 7,457 49.8% 5,737 7,197 3.4% 4.3% 1,460 20.3% Source: Authors’ calculation based on raw data provided by Haiti Customs. -43- Table 4.2 : TRIST Simulations on Four Scenarios of Tariff Reform Scenario 1 Scenario 2 Scenario 3 Scenario 4 Replace inspection fee, implementing tariff Remove inspection fee, status quo (bln. aligning tariffs with reduce all exemptions by reforms in 2014 budget law increase all customs tariffs HTG) CARICOM CET (% change) 50% and increase all customs (% change) by 75% (% change) tariffs by 20% (% change) low E high E low E high E low E high E low E high E Total imports 144.3 -0.1% -0.2% -1.4% -2.6% 0.2% 0.4% 0.0% 0.1% Total Tariff Collection 7.5 3.9% 3.5% 47.7% 30.1% 72.6% 70.2% 71.8% 64.3% Total Inspection Fee Collection 5.8 -0.1% -0.3% -1.8% -3.9% -100.0% -100.0% -100.0% -100.0% Total Excise Tax Collection 0.8 -0.2% -0.4% -0.3% -0.6% -0.3% -0.5% 1.0% 2.0% Total TCA Collection 10.4 0.1% -0.1% 1.3% -1.8% 4.5% 4.5% 3.0% 4.5% Total reveneu 24.5 1.2% 1.0% 14.8% 7.6% 0.7% 0.0% -0.1% -1.8% Source: Authors’ calculations, using 2013 Haiti customs data and the World Bank’s Tariff Reform Impact Simulation Tool (TRIST) with low elasticities 1.5 for substitution and 0.5 for demand, and high elasticities 5 for substitution and 1 for demand. See Box 4.3 for more information on TRIST. -44- Figure 4.10: Applied Protection from Tariffs and Inspection Fees for Four Reform Scenarios from left to right bars for each sector represent: status quo: current tariff and inspection fee collection (2013 data) 25 Scenario 1: implementing tariff reforms in 2014 budget law Scenario 2: aligning tariffs with CARICOM CET 20 Scenario 3: Remove inspection fee, increase all tariffs by 75% Scenario 4: Replace inspection fee, reduce all exemptions by 50% and increase all tariffs by 20% 15 10 5 0 Total Food and Food and Industrial Fuels and Capital Transport Consumer imports beverages beverages supplies lubricants goods equipment goods (industrial) (household) Tariff Inspection fee Source: Authors’ calculations, using 2013 Haiti customs data and the World Bank’s Tariff Reform Impact Simulation Tool (TRIST ) . Box 4.4 : The World Bank’s Tariff Reform Impact Simulation Tool (TRIST) TRIST was designed with the specific task of providing policy makers with detailed insights by product and tax instrument into the short-term effects of trade reform. By its comparative static nature, TRIST allows the comparison of two states - one in which the base values of policy instruments (such as tariffs) are unchanged and another in which these base values are exogenously changed. TRIST is an Excel based tool that does not require expert knowledge in modelling or specialized software and is designed to handle very disaggregated information. An integral part of TRIST is the trade model that underlies the quantification of the effects of trade reform. It is based on five core assumptions: First, the model is derived from standard consumer demand theory and utilizes elasticities to determine the magnitude of the demand response to the price changes that result from a tariff reform. Second, the calculations are based on the standard Armington (1969) assumption of imperfect substitution between imports from different trading partners since consumers distinguish products by the place of production. This intuitive assumption is standard in empirical international trade work and implies that a fall in the price of imports from country A relative to country B will only lead to a partial and not complete substitution of imports from country B with imports from country A. Third, the model does not allow for direct substitution between different products. In other words, each product is modeled as a separate market and in isolation from other markets. This is perhaps the strongest assumption used in the model. However, a relaxation would not only complicate computations but would also generate a need for a range of additional ad-hoc assumptions regarding the precise design of the additional substitution effect and its parameterization. Fourth, it is assumed that all changes in tariffs are fully passed on and that the world price remains unchanged. That is to say an infinite supply elasticity is assumed of imports so that changes in demand in the importing country are small enough relative to the world so that they have no effect on the world price of the product. Fifth, the trade model in TRIST is a partial equilibrium model that treats demand for each product in isolation from the rest of the economy. Hence, it does not take into account inter- and intra-sectoral linkages or the economy wide impacts of tariff changes. But this is not the primary objective of TRIST, which is designed so as to avoid the degree of aggregation of the data that would be necessary in order to implement economy wide computable equilibrium models and to remain simple and transparent in its assumptions, with the flexibility to adjust the key parameters. A detailed discussion of the trade model in TRIST can be found in Brenton et al. (2009). As most partial and general equilibrium trade models, TRIST uses elasticities as the parameters of the model that determine how trade flows react to a given change in prices. Elasticities are notoriously difficult to estimate and so detailed and robust estimates of the required elasticities for import demand and substitutions between imports from difference trading partners are not readily available in the literature. In the absence of empirically estimated country specific elasticities the model is calibrated with different scenarios of standard elasticities for substitution between trading partners (‘exporter substitution’) and the overall effect on import demand (‘demand elasticity’). Standard elasticities, shown in TRIST result tables as "low ɛ", are set at 1.5 for exporter substitution and 0.5 for demand substitution, whereas "high ɛ" are set at 5 for exporter substitution and 1 for demand substitution. Source: World Bank (2014) -45- GETTING A GREATER GROWTH DIVIDEND FROM PUBLIC INVESTMENT 27 This chapter examines the public investment management system in Haiti by looking at four key steps: project identification, planning, execution, and ex-post evaluation. The regulatory framework for public investment in Haiti is complex and not consistently applied. Sectoral strategies to anchor government broader priorities into sector programs and projects are missing. Tools, staff capacity and skills to adequately evaluate and select projects are also missing, leading to the funding of projects that are neither fully assessed nor prioritized. Internal control processes and mechanisms are inadequate which, combined with an absence of project technical monitoring and financial accounting, translate into high fiduciary risks. Ex-post audit of investment projects is ineffective. Overall, the system is weak and would call for strong action if the country’s growth outlook and social indicators are to be improved. High Public Investment Doesn’t Translate into Faster Growth in Haiti… 5.1 Despite rising public investment, economic growth has not accelerated so far in Haiti. The level of public investment expenditure in Haiti is high in relation to Latin American countries and Low Income Countries, but has not generated a commensurate acceleration of economic growth. Despite the high level of capital expenditure, Haiti’s growth rate has been lagging behind, questioning the efficiency of public investment (Figure 5.1). As discussed in Chapter 1, Haiti still performs poorly for selected infrastructure indicators, such as access to electricity, roads or ports. Against this backdrop, high public investment should have been expected to contribute in reducing bottlenecks to faster growth. 27 Prepared by Josue Akre, Emeline Bredy Mamadou Deme, Andy MacDonald, Fabienne Mroczka, and Gerard Verger. -46- Figure 5.1 : Capital Investment (a) and (b) (a) Capital Expenditures, 2003-2013 (Percentage of GDP) (b) GDP Growth, 2003-2013 (Real GDP 2002 = 100) 5.2 This paradox is, however, nothing new in Haiti. High levels of investments have been common throughout the 1965-2004 period, interrupted only during the embargo of the first half of the 1990s, when investment decreased by around 4 percentage points of GDP. After this decline, investment climbed back to levels above 20 percent of GDP, attaining at some point levels of more than 30 percent of GDP. And yet, those high levels of investment did not translate into economic growth. Several reasons have been put forward, ranging from deficiencies in the country’s national accounting system to chronic lack of maintenance or simply the unproductive nature of the investment itself. Poor past donor coordination and the high volatility of external aid have also been argued to have affected the impact of investments in Haiti. This chapter will focus on the quality of the public investment management system. 5.3 The heavy reliance on donors has weakened Haiti’s public investment management. As indicated in a recent IMF study, the efficiency of public investments in Haiti ranks among the lowest quartile in the sampled countries drawn from many regions of the world (Figure 5.2) and particularly low in the early stages of the investment cycle (Figure 5.3). Haiti’s public investment management exhibits a number of distinctive features and practices common to countries that are aid-dependent, including-weak appraisal capacity and reliance on donors to design good projects. These hamper the effectiveness of public investments (Box 5.1). While the Government’s Strategic Development Plan (PSDH) provides broad guidance on Government priorities, sectoral strategies to guide the prioritization of projects are lacking. This leads to a Public Investment Program (PIP) composed of projects that are neither fully assessed nor prioritized. Furthermore, there is no effective ex-ante control on disbursements based on projects’ physical progress against plans. More importantly, domestically funded capital expenditures are not properly accounted for, tracked and reported, creating an environment conducive to a lack of transparency and accountability, as well as to mismanagement of scarce public resources. Finally, even though the existing legal framework is acceptable for the management of public investments, its requirements are rarely respected, with numerous processes and procedures that, when not redundant, are excessively elaborate. -47- 1 2 3 4 0 Belize Congo Source: IMF Solomon Islands West Bank/Gaza Yemen Lao Sao Tome&Princ. Gambia Burundi Togo Senegal Gabon Chad Quartile 1 Sierra Leone Sudan Haiti Swaziland Appraisal Selection Evaluation Managing Trinidad&Tob. Source: IMF Guinea Nigeria 0.0 Barbados Djibouti 0.0 Tanzania Kyrgyz Rep. Egypt Appraisal Uganda Indonesia 0.5 Kenya Azerbaijan Quartile 2 Benin Cambodia Pakistan Mozambique Selection Albania Montenegro 1.0 Ethiopia Haiti Low Income Countries -48- Mauritania Jamaica Mongolia Kosovo 1.2 El Salvador Managing Namibia (0=Lowest, 4=Highest) 1.3 Malawi 1.5 Philippines Ghana Cote d'Ivoire Zambia Quartile 3 Turkey 1.7 Lesotho FYR Macedonia Evaluation 2.0 Ukraine Madagascar Figure 5.2 : Public Management Efficiency Index Serbia Bangladesh Belarus Burkina Faso Afghanistan Latina America and Caribbean Mali 2.5 Jordan Rwanda Moldova Figure 5.3 : Public Investment Management Index – PIM (0=Lowest, 4=Highest) Botswana Kazakhstan Quartile 4 Armenia Bolivia Peru Thailand Tunisia Colombia Brazil South Africa Box 5.1 : Common Features of PIM in Donor-Dependent Countries PIM systems in donor-dependent settings tend to exhibit the following distinctive features: Investment guidance, project development and preliminary screening: government strategy documents, such as a PRSP, tend to be directed towards the donors, rather than covering both external and domestic investment in an integrated and coherent manner. They are at a level of generality that limits the extent to which they can provide a basis for preliminary screening of projects, and are often not supported by effective sector strategies. Formal project appraisal: there is a reliance on donors to conduct appraisal, with a serious lack of appraisal capacity within government; and a lack of guidance on defining the project preparation process and on how to appraise domestically-financed projects and PPPs. Donor capacity-building on appraisal tends to be agency- specific, with little or no domestic training capacity. Independent review of appraisal: reflecting reliance on donors, there is a lack of capacity for independent review, either of donor projects or domestically-financed projects. Project selection and budgeting: the budget is divided into a recurrent and a development budget, with weak integration between the two and substantial off-budget aid. The use of PIPs remains quite common, but these can be poorly connected to fiscal policy and the budget. In practice, a PIP tends to be more a coordination tool than a tool to manage the project portfolio strategically or to help enforce review of individual project proposals before they can be considered for budget funding. Agreement by a donor to finance a project is tantamount to the project being included in the budget – subject to basic screening for consistency with a PRSP (which is not difficult given their generality) and any required counterpart financing being affordable. Project implementation: unpredictability of donor funding (especially budget support) interrupts project implementation due to lack of alternative financing. Weak project management capacity induces donors to set up multiple Project Implementation Units (PIUs) within implementing agencies that initially help to speed implementation and compliance with fiduciary standards, but which cut across and impact negatively on in- line capacities and accounting and reporting systems. Procurement is undertaken by PIUs or donors to donor standards rather than national procurement standards. Project adjustment: reliance on donors to trigger review of any projects that are off-track. No similar mechanism for domestically-financed projects. Facility operation: formal hand-over procedures on completion of donor projects, but inadequate asset registration systems; and inadequate funding for operations and maintenance, in part due to weak integration of recurrent costs of donor projects in fiscal policy and budgets. Basic completion review and evaluation: reliance on donors to undertake reviews and evaluations of their projects. Otherwise, little or no systematic basic post-project review, let alone evaluation, and little systematic use made of findings from donor evaluations to improve future project design and implementation. Source: Rajaram et al, (2010) 5.4 The chapter structure follows the project cycle from planning to completion (Figure 9.4). The analysis draws among other sources on (i) interviews with key PIM stakeholders, and (ii) a representative sample of 20 recent investment projects (Table 5.1 and Figure 5.4).28 The sample focused on the eleven ministries with the largest investment budgets (76 percent of the total number of projects in the 2012-13 investment budget, included projects from all four sources of investment funding - Treasury, Petrocaribe, bilateral and shared donor financing, and represented all Government sectors. 28 Total population of approximately 700-800 projects. -49- Table 5.1 : Set of Projects Reviewed by Ministerial Affiliations Figure 5.4 : Four Stages of the Project Life Cycle Project Cycle Iden fica on Planning Execu on Evalua on •Consistent with •Ongoing •Monitoring by •MPCE, DPES, PSDH projects execu ng DIP, DSE •Consistent with •Opera onal agency) •Sector Ministry Sectoral Plan •External •MEF Stratgey •Financial monitoring •CSCCA •Included in Plan (MPCE, MEF, Sector •Donors Triennial •FIOP Investment Plan ministries, •Projects in donors) •Project File prepara on (sector ministry •Quarterly •Feasibility Study submits to its Reports •Socio- UEP and MPCE •Annual balance economic for analysis) sheet of study physical •Project progress and Document financial (DDP) statement from •FIOP beginning of •Inclusion in project PIP with above Documenta on Source: Haitian authorities … Because of an Overly Elaborate and Rarely Respected Regulatory Framework 5.5 Haiti’s public financial management system has not significantly evolved since 2008. Reports from various sources (World Bank 2008,29 PEFA 2011,30 IMF 2013,31 and the Directorate of Monitoring and Evaluation of Haiti’s Ministry of Planning32) describe the same shortcomings across the spectrum of Public Financial Management (PFM) components, with few marginal improvements. Box 5.2 outlines the PFM legislative framework. 29 Haiti Public Expenditure Management and Financial Accountability, World Bank 2008. 30 PEFA, PEFA Evaluation Haïti, January 2012. 31 IMF Country Report No. 13/91, Haïti. Selected Issues, March 2013. 32 Diagnostic du système de suivi et d’évaluation programme et projets d’investissement public, janvier, MPCE/DES, 2014. -50- Box 5.2 : Haiti Legislative Framework Haiti, Constitution, 1987. Finance Law Decree, February 16, 2005 on preparation and execution of finance laws. July 2001 Regulation setting the State’s expenses budgetary nomenclature. November 2001 Regulation setting the State’s resources budgetary nomenclature. Arrêté, February 16, 2005 setting general regulation on public accounting. Regulation, October 2006 setting the State’s chart of accounts. Law N° CI. 06 2009 009 establishing general rules related to procurement and concession agreements for work and public goods. Decree, October 4, 1984 setting up the Public Investment Fund within the Ministry of Planning. Arrêté, September 17, 1985 defining application conditions of the Public Investment Fund. Arrêté September 5, 2009 establishing public procurement thresholds and the National Procurement Commission’s intervention thresholds. Arrêté October 26, 2009 determining application conditions of general rules on public procurement and concession agreements work of public service. Arrêté October 26, 2009 establishing procedures for the organization and functioning of the CNMP, the National Procurement Commission. Arrêté December 19, 2005 establishing special rules for the statute of Inspectors of IGF, Treasury Public Accountants and Budget Financial Controllers. Decree, March 17, 2006 creating the Ministry of Economy and Finance decentralized technical service called the General Inspectorate of Finance. Decree May 17, 2005 revising the Civil Service Statute. Decree November 23, 2005 establishing the organization and functioning of the SAI. Source : http://www.sdn.mefhaiti.gouv.ht/sdn_lois.php#A2005 5.6 Even though the existing legal framework is acceptable for the management of public investments, its requirements are overly elaborate for a capacity-challenged environment. The Haitian legal framework includes sound laws, decrees, arrêtés, and regulations covering all aspects of preparation, presentation, and control of budget execution related to financial transactions. The public investment management system more particularly is governed by a Decree and an Arrêté.33 The Decree of October, 4 1984 created the Public Investment Fund within Ministry of Planning and External Cooperation (MPCE) and the Arrêté of September 17, 1985 set out its execution. Both documents include information guiding the operation of the PIM system in Haiti. In addition, MPCE has recently completed a procedures manual governing the preparation and execution of investment projects in Haiti.34 Regulation establishes the requirements for the identification, preparation and execution of 33 Decree of October 4, 1984 creating the Public Investment Fund within the Ministry of Planning; and Arrêté of September 17, 1985, stating the execution of the October 4, 1984 decree. 34 Manuel des Procédures sur la Gestion de l’Investissement Public Tome I: Élaboration du Programme d’Investissement Public, Tome II: Exécution, Suivi et Évaluation du Programme d’Investissement Public, 2014 -51- investment projects. This framework requires a great number of processes and procedures that, when not redundant, are overly elaborate and prescriptive. 5.7 As a result, the formal requirements are rarely respected. Failure to adhere to procedures and controls, as designed in the regulatory framework, is not sanctioned. The requirements are hence largely nominal in practice, not consistently applied or enforced. Our sample indicates the lack of compliance with the existing regulatory framework throughout the project lifecycle. For example, the review showed that first level screening seldom takes place; thus calling into question whether projects meet minimum criteria of consistency with the strategic goal of government or meet budget classification tests for inclusion as a project rather than as a recurrent spending item. This is a critical step in a well-functioning public investment management system to ensure that limited resources are not wasted and contribute to the Government’s policy agenda. Similarly, from our sample of 20 projects, only 6 (30 percent) of these had all of the required project documentation to be included in the PIP (e.g. Project Document, supporting studies, financial plans, FIOP). A quarter of the projects did not even have FIOPs (Identification and Project Operation File), although the existing regulatory framework requires projects to prepare one to receive disbursements of the allocated credits. … Little Strategic Guidance to Select Projects 5.8 While the Strategic Development Plan provides broad guidance on Government priorities, many sectors and ministries remain without a sector strategy that translates this document into sector priorities. Broad strategic guidance for public investment is an important starting point to anchor decisions and to guide sector-level decision-makers. Strategic guidance ensures that investments are chosen on the basis of development policy priorities. As mentioned in Chapter 3, the Government of Haiti completed in May 2012 a Strategic Development Plan (PSDH), detailing the country’s vision to become an emerging economy by 2030. This vision is, however, too broad to serve as a basis for project selection. While some sectors have strategies, there is again not enough information in terms of timing and selection criteria that could assist ministries decide which project to include in the PIP. As a result, links between the Government’s investment projects, its national policies and its underlying strategies are tenuous. From our sample of 20 projects, only a quarter were supported by a sector strategic plan. 5.9 Against this backdrop, ministries face difficulties in setting priorities. The regulatory framework in Haiti requires ministries to prepare a list of potential projects that are then to be subjected to feasibility studies on the basis of key operational, financial, socio-economic and environmental factors to avoid selecting wasteful or unfeasible projects (Figure 5.5) provides a description of the required processes for registering a project in the PIP). In reality, however, the process is seldom followed. Since many ministries lack a formal strategy document that establishes policy priorities and associated projects, the MPCE selects frequently the projects that will appear in the PIP, sometimes on the basis of political considerations. In the sample of 20 projects, only 7 had been prioritized for PIP inclusion by the responsible ministry. 5.10 Consequently, projects included in the PIP are often only at the concept stage, and have not been fully developed. Basic documents such as the Project Document, feasibility studies, and -52- economic, financial and social assessments are often not prepared and submitted to the MPCE. Nonetheless the project is included in the PIP (to ensure its inclusion in the budgetary document) because of:  A lack of capacity and resources to conduct and analyze feasibility studies, and economic, financial and social assessments and prepare Project Documents.  A lack of clear guidance on the criteria to be employed as to when, and how pre-feasibility, feasibility and socio-economic studies should be undertaken. 5.11 As a result, the selected projects are often not fully mature, leading to delays in their implementation. The lack of rigor in project analysis, justification and selection leads to immature projects being included in the budget which may fail to be successfully implemented or even begun. Consequently, the PIP does not represent a strong government commitment to perform. This contributes to delays in the execution of projects. 5.12 Alterations to projects once approved are frequent. This top-down approach can preclude ministries from executing activities they consider priorities through the normal PIP channel and encourage them to find alternative resources to execute these activities, or (with more adverse consequences) to change the objectives of projects after they have been approved in PIP. Moreover, the MPCE can and does arbitrarily replace PIP projects, without transparency or stated justification, thus reinforcing the top-down character of actual prioritization. In the sample of 20 projects, 4 were not maintained in their original, approved PIP form. -53- Figure 5.5 : Required Process for Adding New Projects to the PIP Source: Haitian authorities 5.13 The PIP includes recurrent or operational expenses. Recurrent costs are often not considered in the operating budget when establishing the investment project. The lack of a multi-year expenditure plan doesn’t provide incentives for increasing operating costs as a result of completion of investment project. As a result, the PIP includes non-investment spending and over-estimates the amount of public investment taking place in the country, distorting measurement of its impact on economic growth following implementation. In our sample of 20 projects, almost a third was operating budget projects. These were generally related to the provision of subsidies to commodities, individuals, and industries. 5.14 The PIP, and by default the National Budget, often does not include all donor-funded projects. This results in an undervalued PIP and the potential for a lack of convergence between the -54- objectives of the externally-financed projects and the objectives of the government and sectors. This is counter to the spirit of the Paris Declaration, specifically on the use of in-country systems wherever feasible. This also means that the related project costs are not included in the PIP, are not reported in the Haitian national budget, are not fully monitored as external debt and are not controlled by government financial systems. Finally, the impact on the operating budget of completed projects funded by externally-funded projects is often neglected. Our discussions revealed, for example, that the operating costs of an externally-financed hospital had not been taken into consideration, impacting its entry in operation once completed (Box 5.3). Box 5.3 : The Share of Recurrent Expenditures in Investment Projects Investment programs and project are not reported by economic classification, making it difficult to identify the recurrent share of the investment budget. However, some investment programs are included in SYSDEP with their respective economic classification (in 2012-13 approximately 50 percent of PIP in monetary terms was included in SYSDEP). The Figure below depicts the four top spending categories for investment programs recorded in SYSDEP. It is evident from the Figure that wages and salaries account for the largest share of spending across the observed period. Top Spending Categories for Investment Programs in SYSDEP 2009-13 (percent, budgeted) 80 70 60 50 40 30 20 10 0 2009 2010 2011 2012 2013 Wages and salaries Services and other fees Purchase of goods Immobilization Source: Haitian authorities Moreover, looking at the nature of investment projects in the PIP, certain types of “recurrent” projects can be identified, such as projects with a lifecycle that extends beyond 4 years or projects with an “institutional” or “appui” label. The Table below shows the amounts budgeted for “recurrent” projects, compared to the total amount of treasury funded projects in that same year. Budgeted Amounts for "Recurrent" Projects Selected Budgeted Treasury % projects (prevision) funded projects 2007 879,019,890 6,205,872,899 14.2 2008 855,445,146 4,000,000,000 21.4 2009 728,362,595 4,360,191,490 16.7 2010 2,603,714,685 11,300,000,000 23 2011 2,432,000,000 12,501,500,000 19.5 2012 2,636,072,697 15,397,886,784 17.1 2013 3,000,148,456 18,011,105,415 16.7 Source: World Bank Estimates -55- … A Budgeting Process That is Not Followed 5.15 While there is a well-defined annual budget cycle for investments, it is seldom followed. The process for the preparation and the execution of the operating budget and the investment budget are established by decree, according to a set schedule. These two budgets are prepared in parallel through separate processes managed by the MEF and the MPCE, respectively. However, in reality the set schedule is seldom followed and important delays have been observed. In addition, the current budget cycle does not permit appropriate ministry participation, particularly when it comes to project selection. At times, selection (deletions, additions) is performed by the MPCE without reference to the sponsoring ministry. 5.16 Despite the Triennial Investment Plan, true multi-year planning is still lacking. While a three- year investment plan has been prepared for 2014-16 reflecting the first three years of the PSDH, this tool is not used as a multi-year planning tool. It is not anchored in available resources, and the procedures to make changes to the PTI and its monitoring are still unclear. In addition, the three-year plan is not sufficiently grounded on ongoing programs and projects, disconnecting the PTI and the PIP. Against this backdrop, some projects that have already been selected, or even under implementation, may be abandoned or deferred. The review found, for instance, some examples of abandoned or deferred projects in the health sector. 5.17 Strategic coordination among ministries for multi-sectoral projects is limited. The MPCE tends to treat such projects as a number of independent projects, with little, if any, ongoing project coordination. For example, delays in the construction of basic infrastructures, such as roads, has been experienced during the construction of a new health center, jeopardizing the effective use of the hospital once built. This contributes to higher risks for project completion and a potential adverse impact on total project outlays. Figure 5.6 illustrates the budget formulation process for the preparation of the investment budget. 5.18 Furthermore, project classification used in the budget and in the PIP is unreliable. The same PIP reference can be used for more than one project. It is not uncommon for some projects to be assigned different codes from one year to another, or for identical codes to be assigned to different programs or projects. This means that the reporting of the budget vs. actual expenditures cannot be readily determined; such reporting is the foundation of accountability for the use of public resources, including investment resources. -56- Figure 5.6 : Illustrates the Budget Formulation Process for the Preparation of the Investment Budget. Source: Haitian authorities … Weak Procurement … 5.19 Ministries lack sufficient capacity to perform their procurement responsibilities . The study noted a particular lack of resources and expertise of staff involved in the preparation of tender -57- documents, launching public tenders, and their evaluation – all of which have seriously slowed the procurement process. This is one of the factors that have fostered the creation of parallel structures outside of the official public procurement process. Except in the ministry of agriculture, where a procurement unit is operational, ministries do not currently have procurement units whose main function would be to facilitate and ensure more transparency in the procurement process. 5.20 A large share of contracts are awarded using Figure 5.7 : Contracting Methods, 2012-13 single source selection, resulting in limited (Number of Contracts Reviewed by CNMP) 80 competition. Even though there has been some 70 progress over the last year, about half of all contracts 60 50 are still selected through single sourcing (Figure 5.7). 40 30 Furthermore, while the percentage of competitive 20 contracting has increased, the quality of competitive 10 0 bidding is still low. This limited competition can 2012 2013 Direct contracting single source selection represent additional costs for the budget to carry out Limited competition Open competition Source: Haitian authorities public investments. 5.21 Procurement planning is insufficient and there is limited compliance with legal provisions. A review of our sample indicates that only five percent of the projects reviewed were included in a procurement plan. More generally, the legal and regulatory framework mandates the ministerial procurement units to publish and transmit their procurement plans for acquisitions above the threshold for competitive tenders to the National Procurement Commission.35 At the same time, Decree October 26, 2009, Article 220 requires departments and agencies to submit the quarterly reports on all procurements, regardless of the threshold. However, for 2012-13, only three ministries and five public institutions had submitted their procurement plans and none their quarterly reports. 5.22 The validation of transactions at each stage of the process introduces further delays. Figure 5.9 shows the various steps needed to validate, control, and execute procurement transactions as prescribed by law.36 In total, it should take no more than about four months to process a request for goods and services. However, the actual processing time regularly exceeds the legally required interval. The discussions revealed that at the earliest disbursements for new projects occur at the beginning of the second half of the fiscal year. Indeed, the procurement process is slowed down at each step by an administrative procedure that follows every hierarchical level in the organization, where every issue is submitted to the top for resolution before the authorization flows back down to its point of origin. 5.23 Stakeholders have incompatible responsibilities in the public procurement system. The study identified a number of instances where procurement stakeholders had multiple and conflicting roles. Examples include: 35 Article 66 of the Decree of 26 October 2009 and Article 5 para . 4 of the Act 36 Loi fixant les règles générales relatives aux marchés publics et aux conventions de concession d'ouvrage de service public, CL 06 2009-09. -58-  The National Procurement Commission is both a regulatory body that carries out ex-ante validation controls over procurements, as well as ex-post financial investigations and independent audits, while at the same time advising in case of disputes.37  The SAI, whose main role is to perform an ex-post audit of the government’s accounts and to verify stated assertions must also perform the ex-ante function that “give(s) its reasoned opinion on all draft contracts opinions, agreements and conventions to financial, commercial or industrial which the State is party”.38  At the ministerial level, the Ministerial Procurement Committees prepare Biding Documents and its members participate in the Committee of bid opening and evaluation of tenders.39 5.24 Public procurement thresholds are too high for the levels of budget expenditure. Under the current regime, a large proportion of purchases of goods and services are excluded from competitive tender by the National Procurement Commission or under the provisions of the Law on Public Procurement.40 Procurements under HTG 8 million (approximately USD180,000) are not covered by the law; such procurement can be implemented with a memorandum invoice by the ministry. Between 8 million HTG and the three limits of 20, 25, and 40 million HTG for procurements of works, goods and services, and professional services respectively, procurements are regulated by the law on Public Procurement. Above these three respective thresholds, procurements are regulated by law and controlled by the National Procurement Commission. Table 5.2 presents the details. Table 5.2 : Procurement Thresholds for State institutions (in millions of Haitian Gourdes) 41 Procurement Process Works Goods and service Professional services Memorandum Invoice .. Amount<8 .. Direct procurement 8 ≤ Amount < 40 8 ≤ Amount < 25 8 ≤ Amount < 20 Public tender Amount ≥ 40 Amount ≥ 25 Amount ≥ 20 5.25 There are cases of misuse of the emergency provisions for direct contracting. The Law (Article 7) and Decree relating to a declared State of Emergency allows the government to suspend all ex-ante controls on the release of funds, accounting for expenses to undertake projects, budget allocations, procurement of contracts and the SAI ex-ante visa requirements for procurements (Figure 5.8), which can lead to some undesirable results (increase use of direct contracting, absence of document review from the CNMP, and observed deficiencies in contracts).42 Our discussions revealed that, although this procedures ought to be exceptional, there is evidence of overuse with a number of cases using emergency procedures in areas where no emergency had occurred. 37 Loi fixant les règles générales relatives aux marchés publics et aux conventions de concession d'ouvrage de service public (CL Act,) Article 10, Juin 2009. 38 Decree, Establishing the organization and functioning of the Superior Court of Auditors and Administrative Disputes, Article 5, le 23 November 2005 39 Arrêté, Révisant les seuils de passation de marchés publics #.117, le 4 décembre 2006 40 Ibid. 41 Arrêté, Réglementation des Marchés Publics, # 12, le 14 Février, 2005. 42 Loi sur l'état d'urgence of 15 April 2010 and related Decree of September 9, 2008. -59- Figure 5.8 : Complexity in Procurement Processes Source: Haitian authorities -60- … Fragmented Project Execution and Administration … 5.26 Although projects financed from domestic resources are generally included in the PIP, they can follow multiple, different processes and procedures (Figure 5.9). Projects financed from domestic resources include those financed by the Treasury, the National Education Fund, the Roads Maintenance Fund, the Petrocaribe Fund, and the Debt Cancellation Fund. While all these funds manage the budget’s own resources, each of them has a different budget execution process. This affects how projects are selected, implemented, managed and funded (disbursements and accounting). These different processes and procedures hamper both a clear overall view of the use of state resources and the government’s public policy objectives. 5.27 Executing entities often operate separately from ministries. Executing entities have been put in place at the behest of donors to compensate for the inefficiency of internal government services. Although these entities have more flexibility in project planning and implementation, and potentially can deliver implementation results more swiftly, they entail several potential downsides. First, there is potentially weaker linkage with the government’s overall investment strategy when there are significant investments being managed outside the government system. Second, because these donor-supported entities typically provide higher levels of remuneration, they often skim off the most skilled ministerial staff, draining these scarce resources from departments. In addition, public expenditure information within government is less accessible and comprehensive, understating the levels and types of capital investments being undertaken in the country. But the main drawback of these separate executing agencies is that they reduce incentives for the government and the donors to carry out the institutional reforms necessary to make public services more efficient. -61- Figure 5.9 : PIP Project Execution Processes Source: Haitian authorities -62- 5.28 The disbursement process for the payment Box 5.4 : INTOSAI Conditions Supporting of the quarterly instalments is extremely Corruption cumbersome and highly centralized (Annex 1 and  Excessive concentration or centralization of authority Annex 2). All payment requests, regardless of their  Complex financial administration with multiple importance, are uniformly processed as ordinary layers of bureaucracy Ministry mail. For example, before reaching the  Shared or non-existent accountability within the processes functional step required, all requests are submitted  Weaknesses in organizational and administrative through the mail receiving services, the Assistant systems Director, the Director General (sometimes several  Inadequate internal controls and internal audit mechanisms times), and the Minister before the requested action  Inadequate external audit mechanisms can be sanctioned. At various stages, the request is  A lack of transparency in PFM and inadequate accompanied by letters of transmittal, as applicable, reporting  Inadequate salaries for civil servants signed by the Directors General or Ministers. INTOSAI has identified “complex financial administration with Source: INTOSAI multiple layers of bureaucracy” as one of the process characteristics that pose high fiduciary risks to the operations of any government (Box 5.4).43 5.29 Despite or perhaps because of its complexity, this process lacks effective internal controls and in some cases is not followed. For example, the team noted:  A lack of periodic (for example, quarterly) field visits to verify physical and financial progress.  Compliance checks of expenditure are made on the basis of photocopied documents, rather than originals, eliminating the effectiveness of the controls.  The mere fact that a request is routed through, and a visa is given by, a director, the deputy director, the head of department, and the project manager for processing does not strengthen the quality of controls. This is because, contrary to the prerequisite for the issuance of a visa, there is no verification required or conducted at many of these stages.  In some cases the disbursement process is not followed. In 2012, 20 percent of the disbursements were done directly from the MEF without following the traditional procedures requiring the review of the request by the MPCE. 5.30 Procedures at the line ministry are equally cumbersome. At a minimum, the origination of the request, its commitment, payment and liquidation of all investment project expenses take place at the level of the line ministry. Cumbersome procedures hamper these operations. Within ministries, a request for funds issued by the technical services may require as many as twelve steps before the check is delivered to the beneficiary. Again, multiple steps do not equate with improved internal financial controls. 5.31 Sector ministries do not have an effective tool for monitoring their investment budget availability or type of expenditures. The only real indicator of budgetary availability is the balance in 43 INTOSAI, Assessment of Conditions Encouraging Corruption, Vienna workshop, April 2003. -63- the current account opened for each project.44 Absent an effective tool to monitor cash balances of individual projects, ministries will have to make allocations among the different projects by using a separate set of books that are outside of the formal financial system. 5.32 Payments are made on the basis of pro forma invoices and therefore before the liquidation phase. Supporting documentation to justify receipt of the goods or service is not attached with the final bill, even retrospectively, to the request justifying the payment. It is only since the creation of the Public Accountants that the service rendered will be justified on an ex-ante basis. 5.33 The establishment of public accountants is expected to improve the monitoring and reporting of accounts. Since the Decree of February 16, 2005 establishing the public accountant function, the MEF has been putting in place a network of public accountants, in many cases in line ministries. The Decree reaffirms the principle of segregation of duties and good governance by stipulating the role of the public accountant and the “ordonnateur”. With the implementation of the Single Treasury Account, MEF is now in the process of centralizing this network of public accountants. Their responsibilities for capital expenditure are to control, support, pay and account for investment expenditures and maintain financial records. 5.34 The establishment of Public Accountants Units constitutes a significant advance. The Public Accountants Units, if properly implemented, should help in addressing major ex-ante control issues through:  Recognition of more disaggregated investments: the public accountant could disaggregate actual expenditure charged to subaccount 3900 - Expenditure of projects by project. In order not to overload the accounting function, these more detailed entries should be strictly limited to what is essential (i.e., by project and by expenditure type such as works, goods, services). This temporary practice should cease as soon as sector ministries have the capacity to keep their own detailed records of project funds.  Elimination of prepayment of services, with payment only when services have been actually rendered or when goods have been delivered. This is the worldwide practice in the public and private sectors. There is sound justification for this change, as it significantly improves financial spending controls. Under this change, which the Government is now in the process of testing, the public accountant is responsible for ensuring compliance with budget execution. Gathering the necessary documents to ensure proper use of public funds is part of the accountant’s obligations. In this context, public accountants that work at the payment stage require two proofs: the supplier’s final invoice and proof of delivery bearing the stamp of the recipient of goods and services. 5.35 The creation of financial comptrollers is also very promising. In addition to the network of public accountants, there is also a network of financial comptrollers reporting to the MEF’s Budget Directorate whose responsibility includes the control and verification of budget availability. The 30 44 The SYSDEP system is intended for use for management of the operating expenditures in the operating budget. -64- financial comptrollers have, however, not yet been allocated to all line ministries. In their absence, line ministry staff must obtain their payments visa from the MEF, lengthening the payment cycle. But the main drawback is the separation of financial control from the ministry. Effective control requires an intimate knowledge of the underlying business documents that are verified in the process. … Limited Physical and Financial Monitoring… 5.36 Accounting for investment projects is Figure 5.10: Treasury funded Public Investment limited. An entry by Treasury shows only the 2012-13, (In Billions of Haitian Gourdes) transfer of funds to the ministry’s single 16 investment account (starting in 2014) or the project’s account at the Central Bank. 12 Ministries do not have an accounting system 8 for monitoring expenditures by nature; and 15.4 even when they use the existing SYSCOMPTE 4 system (designed for another purpose), the 6.5 4.3 2.6 0.3 0.3 corresponding entries are not included in the 0 overall General Ledger of the State. This 2012 2013 Opening Balance practice greatly limits control over public Expenditures Available balances in project accounts (BRH) investment expenditure and poses a high Source: Haitian authorities fiduciary risk. Ministries’ opening balances are high while reported expenditures in any given year are low (Figure 5.10). Cash balances at the BRH are not increasing, however. The low levels of available cash in the project accounts at the BRH suggest that actual (but not reported) spending levels are much higher. 5.37 This practice does not allow for meaningful control and reporting on PIP activities. This practice falls short of generally accepted accounting principles and, perhaps more importantly does not allow for meaningful control and reporting on PIP activities, as there cannot be any preparation of consolidated financial reporting for investment projects. A number of basic principles of public accounting are not met: o Unity, which requires visibility of the budget and effective financial control. o Completeness, which requires the recognition of all charges. o Comprehensiveness, i.e. the provision of sufficient information to enable effective control over budget execution. o Fairness, which implies completeness, consistency and accuracy of the financial information provided. 5.38 There is no integrated computerized management system for public finances or independent accounting system for public investments. There are more than 25 free-standing applications that respond to specific needs, but none of them caters to the project budgetary, financial and accounting management needs. In addition, where there is a need for additional functionality not provided by these stand-alone systems, requirements are met by the use of Excel spread sheets. This is not simply -65- a “process” shortcoming: the lack of an integrated information system has significant implications for the oversight of public resources, because it effectively undermines any audit trail and there is no security control over changes made in Excel (Table 5.3). Table 5.3 : Summary of Applications Relating to the Expenditures and Investment Chain Name Description Principal Limitation Degree of Integration o Semi-automatic linkage to o Intended only for the SYSDEP; requires ELABUD Budget development preparation of the the intervention of operating budget supplier o Only freezes budget allocation after the visa o Registers of the expense by transaction in the Expenditure management, Public Accountant SYSDEP General Ledger Financial controls of requisitions o Only supports (GL) (Automatic expenditures directly posting to GL) paid by the Treasury o Excludes information on budget availability Treasury and current accounts o No interface with SYSCOMPTE o Development management system General Ledger application; not yet stabilized Financial monitoring of public o Application appears to SYSPIP n.a investments have been abandoned o General Ledger / Module of o No interface with TBMS (Integrated financial GL / TBMS None SYSCOMPTE and and administrative CHEK management System) CHEK Check printing and control None o No GL interface Monitoring System for Public o Under development. Investment and for Projects o Not interfaced with SYSGEP o Application appears to financed by external aid ELABUD or GL have been abandoned External aid management o No interface with program for donors and NGOs o At present, covers only MGAE the expenditure donors system 5.39 Treasury cash management does not support a smooth, uninterrupted implementation cycle for capital projects. There is no process in place that allows Treasury and the ministries to project their financing needs, and thus, prioritize their transfer requests to allow for a smooth project implementation. As a result, there are cash shortfalls that result in rationing, reducing the efficient implementation of all capital projects the financing of which is delayed. In 2011-13, national treasury funded projects received approximately 60 to 70 percent of their appropriated credits. In addition, about 80 percent of disbursements for national treasury funded projects were received during the last 4 months of the fiscal year in 2012, resulting in projects not having sufficient time to execute their activities as planned (Figure 5.11). A similar trend can be observed in 2013, with a spike in February due to post Hurricane Sandy emergency activities, which led to a national budget increase. -66- Figure 5.11: Cash Flows National Treasury Funded Projects 2012-13 (In Percent of Credits) 70 60 50 40 30 20 10 0 2012 2013 Source: Haitian authorities 5.40 Monitoring of projects is sparse. In principle, monitoring of projects appears to be in place. Quarterly project monitoring reports, as well as annual reports (bilan), are required to be submitted to the MPCE by line Ministry Planning Units (UEPs) and the MPCE should review this information. However, these reports are seldom produced and presented in a timely manner. For 2012, approximately 37 percent of annual reports were submitted to the MPCE (for 247 projects financed by treasury, only 92 annual reports were received). During that period, certain ministries such as MPCE, MCFDF, MHAVE and MAE did not submit the required annual reports, which did not prevent them from getting disbursements). The fiscal year 2013 did not fare better, with only 52 annual reports received by the MPCE for 252 projects financed by treasury (approximately 20 percent). Even when quarterly or annual monitoring reports are submitted, the quality is questionable with often no information available on verifiable implementation indicators. For 2012, a total of 542 activities were programmed for the 92 annual reports received. Information, however, was only available for 170 activities (or 31 percent of total activities) and no information was available for 301 activities (or 56 percent). 5.41 Both at the central (MPCE) and the line ministries levels, physical monitoring of projects is insufficient. From the discussions, it appeared that the MPCE had not performed any physical supervision on-site visit in the last five years. Because of lack of budget funds to perform the work, their technical teams rarely go into the field to inspect projects, trusting instead the progress reports of the departmental directorates. However, these reports are not treated as a priority and their production is too irregular for use as a basis for monitoring. In fact, some institutions have even been unable to provide any follow-up report of projects in the PIP. Against this backdrop, the Government cannot obtain progress reports, the exact situation of a project, or the challenges it faces in order to take timely corrective actions or simply report the status. Under these conditions, it is also very difficult to assess the quality of the reports produced by contractors. -67- … And a Lack of Ex-post Evaluation (Project Auditing and Evaluation) 5.42 Created in 2008, IGF is still a nascent internal audit entity. IGF has the responsibility, among other things, to periodically audit: (i) the system for public investments, in particular the management of investment funds, programs and investment projects, independent of the sources of financing; and (ii) the methods of acquisition, conservation, allocation and the accounting of fixed assets that are part of the assets of the state.45 At present, however, its 20 auditors are limited as to what they can be reasonably expected to achieve. They do not prepare a risk-based audit plan, and to date have not conducted any capital project-related audits, despite the significant amounts mobilized in some of the investments. Their focus has been on system-based audits, rather than transaction-based compliance audits that are required for the public investment budget and system. 5.43 The development of ministry-based internal audit units is unlikely in the near future. There is a long-term vision of achieving this goal, but it remains to be proposed, approved, and implemented. In any case, it represents a decade-long development period for full implementation; as such, it does not represent a viable medium-term option. Meanwhile, IGF should be used as a technical and institutional resource for line ministries and agencies’ internal audit needs and requirements. 5.44 The external audit function in Haiti is performed by the Supreme Audit Institution (SAI), the Superior Court of Auditors and Administrative Disputes. The organic law states that: “the CSCCA is an independent institution whose mission is to judge the actions of the Public Administration, Accounts Officers and Public Accountants and assist the Parliament and the Executive in controlling the execution of the laws and regulations related to Budget and Public Accounting.46 5.45 The SAI so far plays a limited role in investment budget execution. As part of the development of its opinion on the government’s financial reports, it examines the public investment system under the assumption that ministries exercise appropriate ex-ante control over the individual investment projects. As previously noted, it also exercises an ex-ante role in the validation of contracts.47 5.46 The requirement for a tripartite (SAI, MPCE and line ministry) audit team to conduct a closing audit of projects has never been enforced. The regulatory framework requires that a project be audited at the completion by a team composed of representatives of the SAI, the line Ministry and MPCE after a completion report is produced by the line ministry; however, there are no sanctions for non-compliance. The SAI has not yet performed financial closing audits at project completion nor did it request any project completion reports. The SAI receives no information of the history of the project’s execution, no copies of the quarterly audits that are to be conducted by the responsible ministry’s M&E unit, its technical division or by MPCE. 45 Décret: Inspection Générale des Finances (IGF), No 47 of 25 May 2006, Article 5. 46 Décret: Etablissant l'organisation et le fonctionnement de la Cour Supérieure des Comptes et du Contentieux Administratif, Article 2, 24 du 10 mars 2006. 47 There is considerable debate within Haiti and among donors over the appropriateness and the utility of this ex-ante role. However, it continues to date. -68- 5.47 There is no significant ex-post control over the Government’s investment budget (Figure 5.12). Given the absence of an internal audit function, continuing budget and staff restrictions on ministries that limit physical and financial quarterly inspections and the inability of MPCE to adequately monitor projects, it is unlikely to see in the future a significant increase in ex-post controls (Box 5.5). Figure 5.12: Monitoring and Evaluation on Projects, in PIP, 2011-13 (in Number of Projects) 900 800 700 600 500 400 300 200 100 0 # projects in PIP Physically Audited by IGF Audited by monitored by CSCCA MPCE 2012 2013 Source: Haitian authorities Box 5.5 : Project Completion The closing of projects is an important component of the investment project accountability loop. The Figure below shows the performance of line ministries during this phase. Project Completion/Evaluation in our Sample (Number of Projects) 18 16 14 12 10 8 6 4 2 0 Ministry closing Asset transferred Joint closing inspection audit Yes No Ongoing or Inapplicable Source: Haitian authorities Source: Haitian authorities From the 20 projects surveyed, eight had undergone a closing inspection by the line ministry, the rest had not (either because they were on-going, operating budget projects included in the investment budget or were not registered with the reporting ministry). The regulations48 required a joint closing audit by a three-person team, composed of representatives from the ministry, MPCE and the SAI. Some ministries reported having gone through a joint audit. However, it was not possible to obtain evidence of these audits. Several ministries seemed unaware of their responsibilities for ensuring that completed projects are audited and the assets transferred to the ministry’s accounts. 48 Decree of October 4, 1984 creating the Public Investment Fund within the Ministry of Planning; and Arrêté of September 17, 1985, stating the execution of the October 4, 1984 decree. -69- INCREASING SUSTAINABILITY AND IMPROVING FURTHER HEALTH OUTCOMES 49 This chapter discusses recent trends in health indicators in Haiti, the structure of health provision, and the role of public spending. Donor assistance to the health sector has substantially increased over the past decade with improvements in health indicators, and narrowing gaps between income groups and regions. The role of the State remains, however, very small and access to health services remain limited. In light of its heavy reliance on donor assistance and given the falling level of external resources, the health sector faces a particular challenge in ensuring a sustainable financing model for the future. Furthermore, the composition of public spending may not be conducive to better services with increases in operating expenditure not keeping up with the expansion in public investment. Health services are unevenly distributed geographically and the number of daily visits is low, possibly because of medical personnel holding second jobs. Going forward, the authorities are thinking about making greater use of results-based financing. Although Low, Health Indicators Have Improved … 6.1 Health indicators in Haiti are poor. Maternal mortality rate in Haiti is almost five times higher than the regional LAC average of 85 and the under-five mortality rate is also four times higher than the LAC regional average of 18 (Figures 6.1 and 6.2). These outcomes are, however, in line with the average of countries of similar levels of economic development and much lower than the average of fragile states. Compared with other countries in Latin America and Africa, utilization of health services remains also lower. For example, Haiti has the lowest rate of deliveries by skilled birth attendants (37 percent) in the region and compared to low-and middle-income African countries. The immunization coverage rate is also much lower in Haiti (45 percent) than in other Latin American and African countries (Figure 6.3). The diarrhea treatment rate is one of the services for which coverage in Haiti is comparable or better than in other Latin American and African countries. 49 Prepared by Eleonora Cavagnero, Marion Cros, and Andrew Sunil Rajkumar. -70- Figure 6.1 : Maternal Mortality, 2013 Figure 6.2 : Under-5 Child Mortality, 2013 (Per 100,000 live birth) (Per 1,000 live births) Fragile states Fragile states 600 100 500 90 Haiti LIC LIC 80 Haiti Caribbean small states Caribbean small states 400 70 60 300 50 40 200 30 LAC LAC 100 20 10 0 0 Source: World Development Indicators Source: World Development Indicators Figure 6.3 : Health Service Utilization (In Percent of Population) Burkina Faso 67 66 (2010) 48 81 59 Ghana (2008) 57 53 79 84 Benin (2012) 87 54 48 44 Kenya (2009) 43 72 68 Honduras 83 46 (2012) 60 85 Guyana 93 89 (2009) 59 56 37 Haiti (2012) 36 58 45 0 20 40 60 80 100 Skilled birth attendance Institutional delivery Diarrhea treatment Immunisation Sources: Demographic and Health Surveys and World Health Organization 6.2 Maternal and child health indicators have improved in Haiti, however, over the last two decades. Despite the devastating earthquake of 2010, maternal and child health seem to have improved over the last two decades: the under-five mortality rate (U5MR) declined from 145 deaths per 1,000 live births in 1990 to 73 deaths per 1,000 births in 2013 (Figure 6.4). In addition, the maternal mortality ratio (MMR) between 1990 and 2013, from 670 deaths per 100,000 live births in 1990 to 380 -71- deaths per 100,000 live births in 2013 (Figure 6.5).50 Haitians can now expect to live 8 more years than back in 1990, two additional years gained compared to the average LAC country (Figure 6.6). All health service utilization indicators improved (Figure 6.7): Between 1994-5 and 2012, deliveries by skilled birth attendants increased by 76 percent, institutional deliveries increased by 125 percent, the treatment of diarrhea increased by 87 percent, and immunization coverage by 50 percent (DHS, 1994- 95 and 2012). Figure 6.4 : Changes in Under-5 Child Mortality Rates, Figure 6.5 : Maternal Mortality Ratio, 1990-2013 1990-2013 (per 1,000 live births) (Per 100,000 live births) 0 700 -10 650 -20 Caribbean small states 600 -30 550 -40 LAC -50 500 -60 450 Fragile states -70 400 Haiti -80 350 -90 300 LIC -100 1990 1995 2000 2005 2013 Source: World Development Indicators Source: World Health Organization (WHO) Figure 6.6 : Changes in Life Expectancy at Birth, Figure 6.7 : Health Service Utilization, 1994-2012 1990-2012 (In Years) (In Percent of Population) Haiti 9 30 LIC Immunization 33 41 8 45 Fragile states Caribbean small states LAC 7 31 Diarrhea treatment 41 44 6 58 5 16 Institutional delivery 17 4 22 36 3 21 Skilled birth attendance 24 2 26 37 1 0 20 40 60 DHS 1994-95 DHS 2000 DHS 2005-6 DHS 2012 0 Source: World Development Indicators Sources: Demographic and Health Surveys and World Health Organization 6.3 The gap in health outcomes between rich and poor households has also narrowed. Significant gains have been made in infant mortality for the poorest quintile, as well as the second and third quintiles over the last decade (Figures 6.8 and 6.9). Meanwhile, surprisingly, infant mortality increased slightly for the fourth quintile, as well as for the richest quintile in 2012 (DHS, 2012). This could be the 50 MMR numbers, estimates from the World Health Organization, are however, not as reliable as those for IMR and U5MR, calculated from household surveys. -72- result of the earthquake which affected the Metropolitan area where households are relatively richer compared to the rest of the country. Reduction in stunting and in the prevalence of diarrhea has been more pronounced for poor income groups (Figures 6.10 and 6.11). Figure 6.8 : Change in Infant Mortality, 2006-12 Figure 6.9 : Change in Under-5 Child Mortality, (Percentage Points) 2006-12 (Percentage Points) 20 20 15 15 10 10 5 5 0 0 -5 -5 -10 -15 -10 -20 -15 -25 -20 -30 Q1 Q2 Q3 Q4 Q5 Total Q1 Q2 Q3 Q4 Q5 Total Sources: Demographics and Health surveys Sources: Demographics and Health surveys Figure 6.10: Change in Stunting Rate, 2006-12 Figure 6.11: Change in Diarrhea Prevalence Rate, (Percentage Points) 2006-12 (Percentage Points) 0 0 -2 -1 -2 -4 -3 -6 -4 -8 -5 -10 -6 -12 -7 -14 -8 Q1 Q2 Q3 Q4 Q5 Total Q1 Q2 Q3 Q4 Q5 Total Sources: Demographics and Health surveys Sources: Demographics and Health surveys … But Disparities Remain … 6.4 Income disparity in health outcomes remain. Despite some improvements, the 2012 DHS generally shows large inequalities in health outcomes and results that are worse for the poorest quintiles. Infant mortality was thus found to be 62 (per 1,000 live births) for the highest quintile of income, versus 104 for the lowest quintile of income in 2012. Compared to the highest quintile, the number of children suffering from stunting was four times higher in the lowest quintile in 2012 (DHS, 2012), a gap that had not changed from 2006. However, the prevalence rates for diarrhea and Acute -73- Respiratory Infections (ARIs) improved for the lowest quintiles and ended up matching those in the highest quintiles (DHS, 2005-6 and 2012). The prevalence rate of ARIs was thus 14 percent for children in the lowest quintile and 13 percent for children in the highest quintile in 2012, whereas in 2005-6, the children in the lowest quintile had a prevalence rate of ARIs that was twice as high as that of children in the highest quintile (DHS, 2005-6 and 2012). 6.5 The poorest benefit the least from health services, despite improvements over time. The children’s vaccination coverage and diarrhea treatment rates seem to be more consistent across wealth index quintiles in 2012 than in 2005-6 (DHS, 2005-6 and 2012), but inequalities persist for the other health services. 52 percent of children with ARIs received treatment in the richest quintile against only 23 percent of children with ARIs in the poorest quintile (DHS, 2012). In addition, coverage of institutional deliveries was eight times higher for the richest quintile (76 percent) than for the poorest quintile (9 percent) in 2012 (DHS, 2012), which highlights the fact that the poorest have very limited access to maternal health services. Table 6.1 : Maternal and Children Health Coverage by Income Group, 2005-6 and 2012 (In Percent of Population) Q1 Q2 Q3 Q4 Q5 Total DHS 2005-6 Immunization 34 40 45 37 56 41 ARI treatment 27 31 41 40 40 35 Diarrhea 34 38 47 54 54 44 treatment Skilled Birth — — — — — 54 Skilled Attendant — — — — — 26 Delivery Health 5 8 17 35 58 22 Facility Stunted 41 37 34 18 8 29 DHS 2012 Q1 Q2 Q3 Q4 Q5 Total Immunization 43 46 52 42 41 45 ARI 23 32 36 52 52 38 Diarrhea 57 52 59 61 62 58 Skilled Birth — — — — — 67 Skilled Attendant — — — — — 37 Delivery Health 9 20 38 51 76 36 Facility Source: Demographic Health Services, 2006 and 2012 -74- 6.6 Geographical disparities have not Figure 6.12: Child Health Outcome Indicators, 2012 disappeared either. The Metropolitan area (Per 1,000 Live Births) has the highest infant mortality rates, due Nord-Ouest to the earthquake, but the service coverage Nippes Grande-Anse is generally better compared to other Sud departments. The Metropolitan area has an Centre infant mortality rate of 81 deaths per 1,000 Artibonite births, or 1.26 times the national rate, and Nord-Est Nord an under-five mortality rate of 108 deaths Sud-Est per 1,000 births, or 1.17 times the national Ouest average (Figure 6.12). In terms of medical Metropole coverage indicators, The Metropolitan area National average has the highest institutional delivery rate 0 20 40 60 80 100 120 and the lowest underweight rate. under-5 mortality rate infant mortality rate Immunization coverage remains similar to Source: ICF International, 2012. The DHS Program STATcompiler - http://www.statcompiler.com the national average (34 percent), while most departments are doing better than Port-au-Prince in this area. The department of the Centre also has infant and under-five mortality indicators that are more alarming than the national average. In contrast, the department of the North-West presents the best child health indicators, with an infant mortality rate of 41 deaths per 1,000 births and an under-five mortality rate of 57 deaths per 1,000 births (DHS, 2012). … With Relatively High Health Spending Financed by Donors 6.7 Total health expenditure per capita is higher in Haiti than in other low-income countries, but still much lower than the average of countries in the region. Health expenditure per capita in Haiti has increased in line with the average of low-income countries and at USD76.6 stood in 2012 above what the average low-income country spends (Figures 6.13 and 6.14). According to the WHO, USD50 would ensure the provision of a basic care package to the population (Sachs, 2001).51 Haiti reached that target, spending 1.5 times that amount in 2011-13. This level of expenditure was, however, financed mainly by donors. 51 WHO estimated US$34 in 2001 prices which would correspond to US$50 in 2014 prices, calculated with a similar methodology than Cavagnero et al. (2008). -75- Figure 6.13: Health Expenditure Per Capita, 2012 Figure 6.14: Health Expenditure Per Capita PPP, (Current US$) 1995-2012 (Constant 2005 US$) 1000 120 900 100 800 700 80 600 60 500 400 40 300 200 20 100 0 0 1996 2007 1995 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2008 2009 2010 2011 2012 LAC Caribbean Fragile Haiti LIC small states states Haiti LIC Source: World Development Indicators Source: World Development Indicators 6.8 While the level of total health expenditure per capita is relatively high in Haiti, public spending in health is very low. In 2013-14, Haiti was allocating only 5 percent of government spending to health, well behind the 15 percent recommended by the Abuja Declaration, and government spending accounted for only about 1.5 percent of GDP, a low rate compared to other low- and middle- income countries (Figures 6.15 and 6.16). Senegal, Benin, Bolivia, and Ghana – countries broadly at the same development level and facing similar health challenges – allocate 10 percent of government expenditure to health, and in the case of Rwanda, this figure reaches 22 percent. Domestic expenditure on health represents between 2 and 6 percent of GDP in other low- and middle-income countries (WHO, 2014). While the share of government spending allocated to health in Haiti was above the average in LAC and LIC countries, it plummeted in the aftermath of the earthquake. This event marked a substantial change in the financing of the sector (Figures 6.17 and 6.18). Figure 6.15: Public Health Expenditure, 2012 Figure 6.16: Public Health Spending, 2000-12 (Percent of GDP) (Percent of Government Expenditure) 18 16 14 12 10 8 6 4 2 0 2003 2010 2000 2001 2002 2004 2005 2006 2007 2008 2009 2011 2012 Haiti LAC LIC Source: World Development Indicators -76- Figure 6.17: Health Expenditure, 2005-10, Figure 6.18: Health Expenditure, 2011-12 (Percentage of Total) (Percentage of Total) Public Public Out-of- domestically domestically pocket funded Out-of- funded 16% 26% pocket 8% 39% External resources External 35% resources 76% Source: World Development Indicators, World Health Organization Sources: National Health acounts, World Health Organization and MSPP. 6.9 Out-of-pocket expenditure has been recently rising substantially, as donor assistance has declined. Following the earthquake, donor assistance in the health sector has surged. Donors represented 28 percent of total health expenditure in 2010. This share rose to 90 percent in 2011, following the earthquake. Since then, donor contributions have progressively declined, covering about 26 percent of health spending in 2013 (Figure 6.19), questioning the sustainability of the recently- achieved progress. This situation puts more financial pressure on households with out-of-pocket expenditure rising substantially. Households only paid 5 percent of health spending in 2011. This share has expanded to almost 50 percent in 2013. This development is alarming in a country where about 60 percent of the population live below the poverty line. Figure 6.19: Health Expenditure by Source, 2004-2013 (% of Total Health Expenditure) -77- … Access to Health Is Limited Especially For the Poor … 6.10 Despite this recent increase in donor assistance, lack of finance still prevents the poor from seeking treatment. The surge in donor assistance seems to have been accompanied by a substantial decline in the contribution of households to their health expenditure. This development should have improved access to health services for the poorest. Yet, the most recent household survey reveals that 65 percent of households in the lowest consumption quintiles did not seek medical attention due to their lack of money against 39 percent for the highest consumption quintiles (World Bank and ONPES, 2014). Moreover, lack of money for treatment is most frequently cited as a barrier to health care in Haiti by women aged 15-49 (76 percent). The second problem faced by these women is distance (43 percent) (DHS, 2012). Distance and lack of money are also the most frequently cited barriers preventing women from seeking medical attention in other low-income countries (Figure 6.20). After Burundi, Haiti is the country where the lack of money is most frequently cited as a barrier. Figure 6.20: Barriers to Health Care Access for Women Aged 15-49, 2012 or 2013 (Percentage of Total) 90 77 76 80 72 70 62 60 52 52 47 50 44 44 43 40 41 36 37 40 34 29 27 30 25 23 24 19 21 21 18 20 11 9 8 10 2 0 Liberia Mali Benin Tanzania Burkina-Faso Burundi Haiti Obtain permission Lack of money distance Not willing to go alone Source: Haiti: MSPP, IHE and ICF International, 2013; Benin: Calverton, Maryland, USA: INSAE and ICF International, 2013; Mali: CPS, INSTAT, INFO-STAT and ICF International, 2014; Burkina Faso: INSD and ICF International, 2012; Liberia; Tanzania; NBS and ICF Macro. 2011; Burundi: ISTEEBU, MSPLS, and ICF International. 2012. Note: the total per country is not equal to 100 percent as respondents can give several answers. Figure 6.21: Enrollment of Populations in Risk-Pooling 6.11 In Haiti, risk-pooling mechanisms are Systems, 2012 or 2013 (Percentage of Total) weak. Only 4 percent of the population is 100 covered by a health insurance system in Haiti 92 (World Bank and ONPES, 2014). This rate is 80 very similar to that of other low-income countries such as Benin (Benin DHS, 2010) and 60 54 Mali (Mali DHS, 2012), where only 3 percent 40 of the population is covered by health insurance (Figure 6.21). Low-income 20 12 10 countries such as Rwanda, on the other hand, 4 3 3 0 show a very high enrollment rate, since 92 percent of Rwanda’s population is enrolled in a mutual health insurance program. Such a Sources: Lagomarcino (2012) and Demographic and Health Surveys -78- system allows poor households to be covered without the burden of out-of-pocket spending at the point of care (MOH, 2010), thus protecting the poor against the financial risks caused by health problems. Donors fund, however, almost half of the mutual health system in Rwanda (Lagomarcino, 2012). Given the large informal sector in Haiti and the fragmentation in donor funding, traditional risk- pooling mechanisms may be difficult to implement. A more workable option could be to exempt specific groups or health services at public facilities: an option that would require, however, public money. 6.12 As a result, public facilities are not necessarily used by the poor. Dispensaries serve mainly rural areas, which are poor areas in majority. Yet, among households going to the public dispensary 49 percent are non-poor (Table 6.2). Furthermore, hospitals serve a majority of noon-poor: 53 percent of households going to public hospitals are non-poor. This could be the result of higher payments required in public hospitals. By contrast, community health interventions seem to be pro-poor to a greater extent: 66 percent of the beneficiaries of these services are poor (moderate and extreme poor). Finally, among households buying medications from street vendors, the majority are poor (43 percent are even extremely poor), indicating that poor households may prefer self-treatment rather than seeking treatment at public facilities. Table 6.2 : Participation Incidence by Income Groups, 2013 (in Percent of Beneficiaries) Public Public CHW & Traditional Private Pharmacy Ambulant Other Dispensary Hospital Mobile Healer Providers Drug Clinics Seller Non-Poor 49 53 34 33 68 74 38 46 Moderate 19 13 34 15 9 5 19 23 Poor Extreme 32 34 32 53 23 21 43 32 Poor Total 100 100 100 100 100 100 100 100 Source: ECVMAS, 2013 6.13 Catastrophic health expenditures (CHE) in Haiti are more important among poor households. When people have to face out-of-pocket spending, the amount can be so high in relation to income that it results in a “financial catastrophe” for the household. These expenditures are not necessarily related to catastrophic events (such as accidents or natural disasters) but to the sizeable burden on household welfare, as they may cut down on necessities such as food and clothing, or are unable to pay for their children’s education (WHO, 2005). Estimating CHE by using a 25 percent threshold of non- food expenses suggests that 3.4 percent of households incur CHE (Figure 6.22).52 This rate is two to three times higher among poor households than among non-poor households: the CHE rate was 3.7 percent among moderately poor households, 5 percent among the extreme poor, and 1.7 percent among the non-poor. This rate was three times higher in rural areas (5.0 percent) than in urban areas 52 This threshold is in line with the consensus among the WB and the WHO (WB-WHO, 2013). These estimates were calculated using the 2012 ECVMAS. -79- (1.6 percent), suggesting that the poor and households in rural areas are more vulnerable to health shocks than the non-poor and households in urban areas. Figure 6.22: Incidence of Catastrophic Health Expenditures at Household Level, 2012 (Percentage) Men 3.4 Women 3.5 Urban 1.6 Rural 5.0 Lowest wealth quintile 4.1 2nd wealth quintile 6.1 3rd wealth quintile 3.2 4th wealth quintile 2.1 Highest wealth quintile 1.5 Extreme poor 5.0 Moderate poor 3.7 Non-poor 1.7 National average 3.4 0 1 2 3 4 5 6 7 Source: World Bank estimates based on ECVMAS, 2012 and ADePT. … And Public Spending Is Not Geared to Better Services. 6.14 The composition of public spending may not be conducive to better service delivery. As noted in Chapter 3, overall operational spending has not kept up with the recent rise in public investment. The health sector is a good example. The investment budget from public funds (Treasury and Petrocaribe) in this sector has doubled, going from 0.07 percent of GDP to 0.15 percent of GDP between 2006 and 2013. By contrast, the increase in operating expenditure has been more modest, amounting to 0.52 percent of GDP in 2006 and 0.7 percent in 2012. 6.15 Moreover, the wage bill may be crowding out critical operating expenses. From 2006 to 2012, the share of the operating budget allocated to personnel costs represented about 90 percent (Figure 6.23). In order to improve health outcomes, however, not only personnel is needed, but also and more importantly the right mix of personnel and other inputs such as medicine and equipment. In comparison, other low- and middle-income countries allocate a smaller share of their operating budget (excluding investment budget from public funds) to human resources (65 percent in Honduras, 58 percent in Ghana and respectively 58, 53, 43 and 22 percent in Tanzania, Uganda, Burkina Faso and Benin). Non-salary operating expenditures include spending such as vaccines and medicines, i.e. critical elements of the health system. Yet the share of expenditure allocated to consumer goods declined from 6 percent in 2006 to 4 percent in 2012 (Box 6.1).53 53 4 percent of donor commitments were assigned to projects supplying inputs and medicines, which remain very low. Programs to combat endemic diseases (11 percent of the national investment budget) and priority health programs (9 -80- Box 6.1 : Health Care Provision in Haiti Health in Haiti is managed by a three-tier system:  The primary health sector is managed by the District Health Unit (UAS). The UAS corresponds to the administrative division of the country and oversees the primary level: Community Referral Hospitals (HCRs), CALs, CSLs and dispensaries. Dispensaries, CSLs and CALs provide primary health services related to maternal and child health, family planning, medical emergency and the treatment of communicable diseases (TB, HIV/AIDS, Malaria), as well as prevention and health promotion services. In addition, each of these three types of institutions provide community health services through assembly stations, mobile clinics and home visits conducted by community agents, nurses/birth attendants, and matrones (traditional birth attendants). HCRs offer four types of health services in the following specialties: internal medicine, pediatrics, gynecology and obstetrics, and surgery.  The secondary level is managed by the departmental health directorate (DDS), which provides technical support to departmental hospitals, which are the referral hospitals at the departmental level. In principle, patients who cannot be treated by HCRs are referred to departmental hospitals. These offer more specialized services than HCRs, such as ophthalmology, urology, dermatology and orthopedics.  Finally, the tertiary level corresponds to the central level of the MSPP, which plays a regulatory role, is in charge of developing health policies and strategies as well as standards and procedures, and plays a monitoring and evaluation role. It directly oversees academic institutions or hospitals as well as the other two levels. The MSPP is also required to supervise the actions of the private for-profit sector even if in practice the MSPP has little information on the private sector. Governance of the Haitian Health System Source: World Bank, adapted from the 2012 Health Master Plan The first-level primary sector accounts for three-quarters of the provision of care in Haiti. There were 907 health facilities in 2013 in Haiti (IHE and ICF International, 2013), 359 dispensaries, 298 CSLs and 129 CALs, which represents 87 percent of the total provision of care. 11 percent of the 907 health facilities were HCRs (second- level primary sector), and the remaining 2 percent were departmental hospitals (8) and University Hospitals/Institutes (8), mainly in Port-au-Prince. The for-profit sector plays a substantial role in the provision of primary health care services. While the for- profit private sector manages 24 percent of the overall provision of care, it focuses on primary care services. The for-profit private sector provides 29 percent of CSLs and 31 percent of HCRs, while there is no for-profit departmental hospital and only 1 in 8 hospital/university institutes is a private for-profit one. The public sector includes the health facilities that are managed by the MSPP, the non-profit sector (structures managed by NGOs and the Church) and the mixed sector (managed by the MSPP and NGOs). Facilities run by the MSPP account for 38 percent of the provision of health care, those managed by the non-profit sector account for 18 percent, and mixed-sector facilities account for 20 percent (IHE and ICF International, 2013). percent) also include activities for the procurement of medicines and laboratory inputs, increasing the share of the commitments of national investment programs going to medicines or non-salary operating expenditures. -81- Box 6.2 : Definitions of Operating Budget Terms Personnel costs: sums spent on basic salaries and wages before deduction of tax withholdings at source, and of contributions to social insurance and civil pension funds, which are at the expense of employees. Compensation payments, premiums and bonuses are planned. Service expenditures and miscellaneous charges: these notably include basic services fees (communication, provision of water, electricity, and gas), promotion costs (printing costs), transportation and travel expenses, training costs, provision of services by third parties, intermediary fees and expenses, petty cash service expenditures to be broken down. Consumer goods and small equipment: supplies and small equipment including office equipment and surgical, medical, and pharmaceutical tools; chemicals and energy supplies, subsistence products, and textiles and apparels. Tangible assets: this is capital expenditure for payments made for the acquisition of movable fixed assets (furniture, machinery and equipment) or non-depreciable property (land) or depreciable property (buildings, edifices, etc.). Grants, assessments, contributions, benefits, compensation: a) grants (non-repayable unrequited payments that the State performs for current purposes or to achieve policy objectives or various objectives; b) assessments and contributions: expenses related to the share that the State has committed to pay for membership in various organizations or institutions; c) benefits: unrequited sums allocated by the State for social assistance benefits, incentives and aid, scholarships and awards, and grants of any kind; d) compensation: exceptional and unforeseen expenses. Other public spending: in this category are grouped ordinary expenses whose nature does not allow their classification in the above categories. Source: Government of the Republic of Haiti. Nomenclature of State budget expenditures and User guide. 2001 Figure 6.23: Salaries, Excl. Domestic Investment Budget (Percent of Government Budget) 100 90 80 70 60 50 40 30 20 10 0 Source: World Bank 6.16 Furthermore, public investment may not be targeted where it is most needed. In 2012, for instance, nearly a third of the commitments of national investment funds were allocated to the construction or rehabilitation of hospitals (Figure 6.24). Only 22 percent of the commitments were allocated to investment projects for dispensaries, health centers (Centre de santé sans lit, CSL and Centre de santé avec lit, CAL), although these facilities account for 87 percent of health care delivery in Haiti (Box 6.2). Focusing national investment programs on hospital construction may not be -82- sustainable in the long run. Indeed, the construction of new facilities will require additional needs in operating costs (e.g. personnel, maintenance) over the long term, and given the current fiscal space constraints of the health sector, these costs may not be covered. Figure 6.24: Investment Budget from Public Funds, 2012-13 (Percentage of Total) Priority health Equipment Programs 9% procurement against endemic at health illnesses facility level 11% 3% CSL Drugs and (Construction) medical supply 10% 4% CAL Governance (development) 19% 12% University hospitals and specialized HCR centers (development) (Contruction) 25% 7% Sources: BOOST and World Bank Staff calculations 6.17 Health services seem also to be unevenly distributed geographically. The results of the most recent household survey would suggest that metropolitan and urban populations are the main beneficiaries of public expenditure on health. According to this survey, 51 percent of the metropolitan population and 45 percent of the urban population consult a public provider (funded by the government or donors) against 43 percent of the rural population, which comprises a majority of the country’s poor (Figure 6.25). However, the metropolitan and urban populations (respectively 36 percent and 33 percent) tend to consult private providers more often than the rural population (23 percent). Moreover, households in rural areas have a greater tendency to self-medicate and consult traditional healers. Figure 6.25: Utilization of the Different Types of Structures by Area, 2013 (Percentage of Total) 100% other 90% Traditional birth attendant 80% 70% Ambultory drug seller 60% Mobile clinic 50% Pharmacy and lab 40% 30% Traditional healer 20% Community health workers 10% 0% Private providers metropolitan area urban rural Public providers Source: World Bank calculations based on ECVMAS, 2013. -83- 6.18 Both the density of medical staff as well as hospital beds is low relative to low-income countries and the regional average (Figures 6.26 and 6.27). There are 9.5 medical staff per 10,000 inhabitants in Haiti (counting medical staff in private-for profit facilities), in line with the average of Sub-Sahara Africa countries, but much lower than the regional average and also much lower than the average of countries at the same stage of economic development. Similarly, the density of hospital beds is low. Haiti has 7 beds per 10,000 inhabitants below the number observed on average in the region and in economy of a similar level of development. Figure 6.26: Density of Medical Personnel, Figure 6.27: Density of Inpatient Beds, 2013 2013 (Per 10,000 Inhabitants) (Per 10,000 Inhabitants) 100 30 90 80 24 25 70 21 60 20 71.5 50 40 15 30 20 10 14.9 7 10 20.4 9.1 6.9 0 2.6 5.1 2.5 5 Haiti Low Income Latin America Sub-Saharan Countries and the Africa 0 Caribbean Haiti Low Income Countries Latin America and the Medical Doctor Nurses Caribbean Source: WHO-Health Statistics, 2013 Source: WB estimates (Haiti) based on SPA dataset (2013) and WHO 2013 for other countries. 6.19 There are also disparities in the departmental distribution of medical personnel and beds. The departments of the South-East, Artibonite, Grand’Anse, North, North-West and Nippes have a density of doctors, nurses, and birth attendants, which is below the national average (Figures 6.28 (a) and (b)). The department of the West has the highest density of doctors, nurses and midwives per 10,000 inhabitants, followed by the North-East (9 per 10,000 inhabitants), the Centre (8.9 per 10,000 inhabitants), and the South (8.8 per 10,000 inhabitants). In addition, the departments of the North- East, South-East, Artibonite, and Grand’Anse, have a low number of hospital beds for a population of 10,000 inhabitants, with respectively 2.7, 3.3, 3.6 and 3.7 hospital beds per 10,000 inhabitants (Figure 6.29). In contrast, the departments of the South, West and Centre have a higher number of beds than the national average for public institutions (which is 5.6 beds per 10,000 inhabitants). -84- Figure 6.28: Density of Medical Personnel by Department – Private for-Profit Sector Excluded a. Density of medical personnel, 2013 b. Density of pharmacists and community health (per 10,000 inhabitants) agents, 2013 (Per 10,000 Inhabitants) 8 7.5 10 0.4 0.1 0.4 0.1 7 9 8 0.1 0.2 6 5.2 5.2 0.1 4.7 7 2.6 0.1 5 4.4 0.1 4.1 6 0.1 3.6 0.1 3.6 3.6 2.1 2.3 2.7 4 3.2 5 4.3 2.9 3.0 2.5 3 4 3.3 2.1 2.2 2.4 2.9 3 3.4 2.2 2.7 2 1.6 3.1 3.6 3.3 2.2 2 1.6 1.8 1 0.3 0.1 0.2 0.1 0.1 0.2 0.2 0.2 0.1 0.2 0.2 1 2.9 1.4 1.6 1.8 1.1 2.1 1.5 1.5 0.9 1.4 2.0 0 0 Pharmacist Community Health Worker Doctor Nurse Auxiliaire-Nurse Midwive Source: World Bank calculations based on the SPA dataset (IHE and ICF International, 2013). Source: World Bank calculations based on the SPA dataset (IHE and ICF International, 2013). Figure 6.29: Number of Beds, exc. Private For-Profit (Per 10,000 Inhabitants) 9 8.5 8 7.1 7 6.1 6 5.5 5.6 4.6 5 4.0 3.6 3.7 4 3.3 2.7 3 2 1 0 Source: World Bank calculations based on the SPA dataset (IHE and ICF International, 2013). 6.20 There are disparities in the distribution of medical and paramedical personnel between urban and rural areas. The densities of medical and paramedical personnel are 2.5 and 2 times lower in rural areas than in urban areas. However, the density of health agents is 1.3 times higher in rural areas than in urban areas. There are 11.4 doctors, nurses and midwives per 10,000 inhabitants in urban areas against 4.3 in rural areas (Figure 6.30). The density of paramedical personnel is also higher in urban areas than in rural areas. The density of health agents is 4.1 per 10,000 inhabitants in rural areas and 3.1 per 10,000 inhabitants in urban areas, which is explained by the fact that rural areas can be geographically more difficult to reach. Inaccessible zones are indeed more numerous in rural areas, requiring a higher number of health agents. -85- Figure 6.30: Density of (a) Medical (b) Paramedical Personnel, and (c) Community Agents by Location, Excluding the Private For-Profit Sector, 2013 (per 10,000 Inhabitants) Density of Medical Personnel 12 10 8 6 4 2 0 Urban Rural National average Medical doctor Nurse Auxiliaire-Nurse Midwife Source: World Bank calculations based on the SPA dataset (IHE and ICF International, 2013). Density of paramedical personnel 0.30 0.25 0.20 0.15 0.10 0.05 0.00 Urban Rural National average Pharmacist Dentist Source: World Bank calculations based on the SPA dataset (IHE and ICF International, 2013). Density of community agents 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Urban Rural National average Source: World Bank calculations based on the SPA dataset (IHE and ICF International, 2013). -86- 6.21 The productivity of the medical personnel of first-level primary facilities is low. A recent study was conducted on technical efficiency for a sample of 45 first-level primary facilities (dispensaries, CSLs and CALs). The sample consisted of 11 CALs, 21 CSLs, and 13 dispensaries in the departments of the West, the North-East, the North-West and the Centre for 2011 (Box 6.3).54 Institutions that have a technical efficiency score of one are efficient. The study shows that only 13 percent of health facilities (6 out of 45) produced health services efficiently: they had neither excessive operating costs, nor excessive personnel. The rest of the facilities (with a technical efficiency score of less than one had an excess of either personnel or operating costs for the number of services provided (Figure 6.31). The average score for Haiti was 0.48 low compared to similar studies on the technical efficiency of the primary health sector carried out in other low-income countries (Table 6.1). Furthermore, health facilities in the metropolitan area tended to have a higher technical efficiency score than facilities in rural and urban areas because they have more patients. The average technical efficiency score was 0.36 in urban areas, 0.42 in rural areas, and 0.82 in the metropolitan area (Cros and Zeng, 2014). Figure 6.31: Technical Efficiency Score, 45 First-Level Primary Facilities, (0=lowest, 1=highest) 30 27 25 20 Number of health facilities 15 10 9 6 5 3 0 <0.50 0.50 - 0.75 0.75-0.99 1 Source: Cros M., Zeng, W. (2014). Table 6.3 : Technical Efficiency In Haiti And Other Low-Income Countries Country Percent of Sample That Is Average Sample Author Non-Efficient (<1) Score Haiti 83 % 0.48 Random selection of 45 health Cros, M., Zeng, but 60 % have a score < 0.5 facilities in 4 departments W. 2014 Burkina Faso - 0.86 25 primary health facilities Marshall et al., 2011 Ethiopia 75 % 0.57 60 health posts in 7 rural districts Sebastian et al., out of 35 2007 Ghana 78 % 0.88 Random selection of 86 health Akzali. et al., facilities 2011 Guatemala 71 % but 53 % 0.78 34 health posts Hernandez, have a score > 0.9 2013 Source: Cros, M., and Zeng, W. (2014). 54 This study on technical efficiency was based on the number of visits (output), and total expenditure and total number of personnel (input). -87- 6.22 In particular, the number of consultations is low. The results indicate that the national average is six daily consultations per medical personnel.55 By comparison, the number of daily consultations per medical personnel in Haiti is 1.5 times lower than the average in Liberia and approximately 2.5 times lower than in Cambodia and Rwanda (Figures 6.32 and 6.33). If a doctor works 7 hours daily, or 420 minutes, this means that this doctor sees one patient per hour (420/6 = 60). It should be noted that nurses were included, given that in dispensaries, there are no doctors, and patients also consult nurses for prenatal care and family planning in CSLs and CALs. Figure 6.32: Consultations Per Doctor and Nurse, Figure 6.33: Consultations per Medical Staff exc. Private For-Profit Sector, 2013 (Daily Numbers) (Daily Numbers) 12 25 10 10 10 20 8 6 7 15 6 6 6 6 6 5 5 4 10 4 5 2 0 0 Haiti (2013) Liberia Cambodia Rwanda Morocco (2009) (2008) (2011) (2014) Sources: SPA, 2013 (Haiti); USAID, 2009 (Liberia); Royal Government of Cambodia, 2008 Source: World Bank calculations based on the SPA dataset (IHE and ICF International, 2013). (Cambodia); MSH, 2011 (Rwanda); Kingdom of Morocco, 2014 (Morocco). 6.23 A second occupation could explain low productivity. The survey on human resources conducted by the World Bank, USAID and the MSPP revealed that 30 percent of the medical personnel in the North-East, 49 percent in the Centre, and 49 percent in the North-West derived an income from sales activities or from a private practice. This was confirmed by focus groups with medical personnel and validated by health department directors at a workshop in Port-au-Prince in June 2013 (World Bank, USAID, MSPP, 2013). It is thus possible that in some cases the medical personnel actually work part-time instead of full-time at facility level, but are paid full time by an NGO or the MSPP, and engage in other activities such as having their own practice, selling medicines or phone cards. Such side activities increase their salary by at least 5 percent and in some cases double it. This might explain the low number of consultations and poor productivity of the medical personnel. 55 A study by the World Bank, USAID, and the MSPP, confirms these results: in 2012 the medical personnel and health agents were carrying out an average of four visits per day (institutional and community) in dispensaries and 6 visits per day in CSLs/CALs. -88- Box 6.3 : Definition of Technical Efficiency Applied to the health sector, technical efficiency consists in achieving a maximum level of consultations or hospitalizations in a health facility with a given level of inputs (Street, 2011). There are 2 cases: 1) input- oriented technical efficiency, which aims at determining the proportion by which inputs (personnel, other expenses) must be used to produce a given number of consultations; 2) results-oriented technical efficiency, which studies the additional number of consultations possible without having to change the health facility’s number of inputs (Coelli, 1996). The use of linear programming, known as data envelopment analysis (DEA), is a non-parametric method that determines the number of healthcare facilities included on an efficiency frontier. This method gives a technical efficiency score based on the number of inputs, such as personnel, current expenditures, and results, i.e. consultations. The technical efficiency score is comprised between 0 and 1. A score of 1 means that the health facility is on the efficient frontier and is efficient. A score below 1 demonstrates poor performance, especially if the score is close to 0. Initially applied in the industrial sector, this methodology is increasingly used in the health sector to measure the technical efficiency of hospitals or primary health care facilities. 6.24 Excessive administrative personnel may also undermine efficiency. The same study shows that administrative personnel represent 40 percent of total institutional personnel. In MSPP facilities, administrative personnel represent 45 percent of institutional personnel and 39 percent of institutional personnel in non-profit institutions (NGOs). There is no standard on the number of administrative personnel, but in general, the need for such personnel in primary level facilities is low. These facilities are small and have little or no equipment (for dispensaries), requiring a low level of maintenance and management, and therefore few administrative personnel. Furthermore, given the low daily productivity previously noted (6 visits per doctor and nurse), the high proportion of administrative personnel is not justified. This ratio seems also high in comparison to economies at similar development levels and facing similar health challenges (Figure 6.34) Figure 6.34: International Comparisons-Share of Administrative Personnel In Total Number of Primary Health Facilities (Percentage Total) 50 45 40 40 35 33 30 29 30 25 20 20 15 10 5 0 Haiti Afghanistan Rwanda Liberia Liberia (dispensaire) (CSL/CAL) Source: Haiti: Cros, M. and Zeng, W., 2014; Afghanistan: Transitional Islamic Government of Afghanistan, 2003; Rwanda: USAID, 2011; Liberia: USAID, 2009 6.25 Going forward, the government aims at implementing results-based financing (RBF) to improve productivity and technical efficiency of primary-level facilities. Currently in a pilot phase, the -89- MSPP is working with partners on the expansion of the RBF national strategy. The fact that health facilities have limited technical efficiency can certainly be explained by their very low productivity, having few patients. To improve the use of their resources, it is therefore in the interest of such facilities to operate mobile clinics to increase the number of consultations. RBF could provide monetary and non-monetary incentives to health workers in order to increase their productivity, discouraging them from exercising another profession (Boxes 6.4 and 6.5). Box 6.4 : More Health for Every Dollar: Results-Based Financing What is Results-based financing? Results-Based Financing (RBF) is an instrument that links financing to pre-determined results, with payment made only upon verification that the agreed-upon results have actually been delivered. RBF for health refers to any program that transfers money or goods to either patients when they take health-related actions (such as having their children immunized) or to healthcare providers, when they achieve performance targets (such as immunizing a certain percentage of children in a given area). What Makes Results-based Financing different? Traditionally funding for health has been directed toward inputs—salaries, construction, training, equipment. Improved health was assumed to follow, but this has not always happened. Despite billions of dollars over the last decade, many countries in Africa are still falling short, particularly in areas that require a functioning health system. Sub-Saharan Africa, for instance, has the highest rate of maternal deaths in the world with an average of about 900 deaths per 100,000 live births. Child deaths and malnutrition are also serious problems. The fundamental issue is the poor performance of the public health care system, including low levels of physical access in some places; poor quality of care; a lack of adequate incentive structures for health workers; weak management; and inadequate data of sufficient quality to monitor and evaluate progress. Individuals must demand services; health workers must be motivated to deliver adequate care; and the institutions they work for must be encouraged to make the systemic changes required to achieve health goals. RBF flips the whole equation on its head, starting with the result —more children immunized, for example—and letting health workers and managers on the ground decide how to achieve them. The Potential of RBF A number of developing country experiences strongly suggest that RBF can work. There are currently three countries (Rwanda, Burundi and Sierra Leone) with nationwide programs and 14 countries with ongoing pilots. These programs help improve health; strengthen health systems; spur innovation, creativity and country ownership; and encourage reforms that confer authority and flexibility to local service-delivery levels, fostering problem-solving where it is most needed. When poor patients or households have been offered financial or material rewards for adopting health- promoting practices, they respond and health indicators improve. Similarly, when health workers and facilities are given bonuses upon achieving targets, those targets tend to be met. Results-based financing has also been shown to help to increase patient demand for health services. In addition to improving health, results-based financing can also contribute to strengthening a country’s health information system. Because accurate monitoring and evaluation of RBF schemes require the development of robust health information and management systems, incorporating the RBF concept, even into donor funds aimed at specific diseases, reinforces efforts to improve the timeliness, credibility and accuracy of national reporting and monitoring, thus contributing to improving the overall capacity of a country’s health system. Source: L. Morgan (2012) -90- Box 6.5 : The Right Incentives Lead to Measurable Results in Rwanda In an effort to improve maternal and child health, Rwanda began paying for performance at the health facility level in 2006. At the time, health workers and facilities were in short supply (only 36 hospitals and 369 health centers in a country of nine million people, and only one doctor per 50,000 inhabitants). Many people lacked access to care, and the quality of care was often low. In 2001, three non-governmental organizations working in Rwanda attempted to address the problem by raising health workers’ salaries. Nothing changed. Then they tried linking bonuses directly to performance – for example, if the health worker or facility could show that ten more women had given birth in a facility rather than at home where women risk dying from complication, they would receive a bonus. Paying for performance worked. Following three successful RBF pilots, the government of Rwanda designed and implemented a nationwide RBF scheme, folding a rigorous impact evaluation into the roll out. Results released in 2009 revealed significant improvements in the deliveries and preventive care visits by young children. Source: F. Niyuhire (2010) -91- KEEPING CHILDREN IN SCHOOL AND IMPROVING EDUCATION OUTCOMES 56 This chapter discusses recent trends in education indicators in Haiti, the structure of education provision, and the role of public spending. With greater donor assistance and an increasing role of the State through tuition waivers, education access has improved, but education indicators remain low with late primary school entry, high dropout, and limited learning. Households still spend a substantial amount for sending their children to school. Geographic disparities in service delivery remain with the West Region having more schools per capita, and quality is poor. Monitoring tuition waivers needs to be improved, and linked to better learning conditions in schools and actual learning in the classroom. Results-based financing offer an opportunity to maintain access, increase accountability, and improve learning, but better Monitoring and Evaluation systems are needed for this. Despite Progress, Education Outcomes are Low … 7.1 Educational attainment has been rising steadily across cohorts of young adults. In 1994, fewer than 30 percent of 15-19 year olds had reached lower secondary school, a figure which had risen to over 50 percent for women and over 40 percent for men by 2012 (Measure DHS). These gains are due in part to increased access to schools, as the number of low-cost, non-public schools has expanded, and to the increasing role of the State in Figure 7.1 : Share of 3-25 Year Olds Currently financing primary education in recent years in School, 2012 (Percent of Total) through tuition waivers and school canteen programs. 100 90 7.2 Today, the majority of children are in 80 70 school. The majority of preschool age 60 children, and 90 percent of children of official 50 primary school age (6-11), are in school 40 (Figure 7.1). In 2001, the national net 30 enrollment ratio for primary school stood at 20 10 about 60 percent, and by 2012 had risen to 72 0 percent.57 Similarly, the overall secondary net 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Sources: World Bank staff estimates using ECVMAS 2012 enrollment ratio rose from 22 percent to 47 56 Prepared by Melissa Adelman. 57 The discrepancy between the primary net enrollment ratio and the overall 90 percent enrollment rate of 6-14 year olds is due to the large number of primary school age children in preschool. -92- percent. These increases reflect progress in both the share of children in school and the accordance between age and grade. 7.3 Despite this progress, educational attainment and learning are still relatively low. Compared to its LAC neighbors, Haiti has the highest share of adults with no education, and the highest share of 15-19 year olds who have not completed primary school. Literacy rates in all departments, including the West, are lower than the LAC average, and in several departments are close to the global average for Low-Income Countries (LIC) (Figure 7.2). Nationally, the adult literacy rate is about 77 percent, midway between the LIC and the LAC averages. Figure 7.2 : Haiti’s Literacy Rate is between LIC and LAC Averages Notes: Data on literacy for countries other than Haiti comes from different sources, and therefore methodological differences may affect the comparisons between Haiti and other countries. Sources: World Development Indicators; World Bank staff estimates using ECVMAS 2012 7.4 Haiti’s school enrollment still lags behind the region and most children are over age for their grade. While the majority of children are now in school in Haiti, within the Latin America and Caribbean region, only Nicaragua (88 percent), Guatemala (92 percent), and Honduras (94 percent) also have enrollment rates below 95 percent for this age group (SEDLAC 2014). School enrollment begins to drop off around age 15, but 73 percent of 18 year olds report still being in school.58 In addition, substantial distortions between age and grade remain, driving large differences between net and gross enrollment rations at every level, until participation drops off steeply at the tertiary level Figure 7.3 (a) (b). 58 Enrollment here is based on answers to the survey question asking if children are ‘currently in school’ rather than on administrative records. -93- Figure 7.3 : Figures (a) and (b) - Gross and Net Enrollment Ratios, 2001-12 (Percent of Children in Primary School) a) Primary Net Enrollment Ratios, 2001-12 (Percent) b) Net and Gross Enrollment Ratios, 2012 (Percent) 100 140 90 80 120 70 60 100 50 40 80 30 20 60 10 0 40 20 0 Primary Lower Upper Tertiary Secondary Secondary 2001 2012 NAR GAR Sources: UNESCO Institute for Statistics; World Bank EdStats; ECVMAS (2012) Sources: UNESCO Institute for Statistics; World Bank EdStats; ECVMAS (2012) 7.5 Children from poor, rural households, living in departments far from Port-au-Prince, are more likely to be over-age or out of school. To understand better the correlates of school enrollment and over-age status, probit regressions were estimated (results are presented in Annexes 2 and 3). Holding other characteristics equal, for a HTG 1,000 increase in annual household per capita consumption (worth about 4 percent of the national poverty line), the probability of school enrollment rises by 0.2 percentage points. Department of residence also has a significant effect on school enrollment: children living in Artibonite, West, and South-East (departments surrounding Port-au- Prince, and also with the lowest poverty rates) are all more likely to be in school and less likely to be over age than children living in other departments. Children from urban households are 2 percentage points more likely to be in school and 19 percentage points less likely to be over age than children living in rural areas. 7.6 Children living in households with less educated heads are also less likely to be in school and more likely to be over age. The results of the probit regressions also show that children living in households where the household head completed primary school are 4 percentage points more likely to be in school than children in households where the head had no formal schooling. They are also 23 percentage points less likely to be over age than children in households where the head had no formal schooling. These correlations between children’s school enrollment and household head education and household poverty (discussed above) demonstrate how poverty can persist from one generation to the next through education. 7.7 The primary level of the education system is highly inefficient: children start primary school two years late on average and fewer than 60 percent will reach the last grade of the cycle. While the official age for beginning primary school is 6, the average child enters first grade for the first time at 7.8 years old, after having spent two or more years in some form of preschool (Table 7.1). This distortion grows over time, as about 10 percent of children repeat and 2-6 percent drop out of each grade of primary. These repetition rates are higher than the averages in the rest of LAC and Sub- Saharan Africa. Using a simulated cohort approach, these rates imply that only about 58 percent of -94- children in first grade will arrive at sixth grade, and only 29 percent will reach the final year of upper secondary.59 For those who do eventually complete primary, the average ages by grade suggest that it takes 7 to 8 years on average to complete 6 years of primary school. Therefore, identifying and addressing the drivers behind late primary school starts, as well as the high repetition and dropout rates, is critical to increasing educational attainment. Table 7.1 : Late Starts, Repetition, and Drop out Contribute to Low System Efficiency Grade Average Prescribed Percent Expected Percent Expected To Age Age To Repeat Drop Out 1 8.1 6 12 % 2% 2 9.9 7 10 % 1% Primary 3 11.5 8 11 % 2% 4 12.8 9 9% 3% 5 13.8 10 7% 3% 6 15.3 11 11 % 5% 7 15.9 12 6% 3% secondary secondary Lower 8 16.8 13 3% 4% 9 17.8 14 10 % 5% 3 18.5 15 4% 4% 2 19.6 16 4% 7% Upper Retho 20.6 17 29 % 13 % Philo 20.8 18 9% 30 % Global comparisons Sub-Saharan Africa 8.6 % average primary repeaters Latin America & Caribbean 5.7 % Note: Percentages expected to repeat and to drop out after grade X are based on the share of individuals who were in grade X last year who are in grade X again this year (repeaters) and who are not in school this year (dropouts). Source: UNESCO Institute for Statistics; World Bank staff estimates using Demographic and Health Survey 2012. 7.8 After age 14, girls start dropping Figure 7.4 : Youth Enrolled in School by Age and out of school slightly faster than boys. As Gender, 2012 (Percentage of Total) Figure 7.4 shows, girls and boys are 100 enrolled in school at roughly equal rates 90 80 through age 14, and then girls begin to 70 leave school more quickly than boys. This 60 trend is similar across individuals in poor 50 and non-poor households, as well as urban 40 30 and rural areas. It may be due in part to 20 girls’ faster progression: in the final year of 10 primary school, the average girl is 14.9 0 years old compared to 15.6 years old for 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 boys. It may also be related to many other Female Male Sources: World Bank staff estimates using ECVMAS 2012 social and economic factors, including for example childbearing and migration. 59 The simulated cohort approach used to calculate the survival rates is based on the UNESCO Institute for Statistics method: survival rate from grade A to B = (students in grade B – repeaters in grade B)/students in grade A. -95- 7.9 Furthermore, many students, particularly in poor communities, learn little. Assessments administered in early grades in selected schools have found that fundamental skills are acquired very slowly or not at all, particularly in schools in poor communities. For example, assessments conducted in schools in the Artibonite and Nippes found that the average third grader could only read 23 words per minute, well below the estimated speed of 35-60 words per minute required for comprehension of a basic text.60 Students sitting for the national exams are a relatively select group, considering that many children do not reach beyond sixth grade, and that sitting for the first two exams requires paying a fee (250 gourdes in sixth grade; 350 in ninth). 2013 passing rates were about 75 percent at grades 6 and 9, falling to 29 percent in Rheto and 38 percent in Philo. These rates vary by department, as does the share of students actually sitting for the exams (Table 7.2). Given the weaknesses in basic skills suggested by small-scale studies, nationally representative learning assessments at earlier ages are needed to understand the challenges faced by the majority of Haitian students. Table 7.2 : National Exam Passing Rates, 2012-13 (In Percent) 2012-2013 Passing Rates Share of 6th Grade Students 6th 9th Rheto Philo Sitting For Exam* Ouest 77 77 30 37 70 Nord 82 79 22 35 58 Nord'Est 75 78 29 42 62 Nord'Ouest 78 71 21 28 61 Centre 70 82 32 46 56 Artibonite 79 71 26 35 60 Sud 77 65 23 33 52 Sud'Est 76 80 27 34 91 Grand'Anse 75 68 28 37 60 Nippes 72 70 28 47 58 National 77 76 29 38 64 *Note: Shares are calculated using 2010-11 School Census and 2010-11 exam results. Source: World Bank staff estimates using data from MENFP. … Driven by a Greater Education Supply from the Non-Public Sector … 7.10 Growth in education supply has primarily come from the non-public sector. While all children in Haiti are constitutionally entitled to a basic education, the State has historically not fulfilled this obligation. Formal education in Haiti is structured in four levels and, in practice, the majority of education services at all levels have been provided by the non-public sector (Box 7.1). Figure 7.5 shows that there are about 7 times more non-public schools than public schools providing education at the primary level. A similar ratio also exists at the secondary level. Non-public schools currently operate in a largely unregulated fashion, as the existing licensing system has never been fully enforced. However, schools participating in certain public financing programs, such as the donor-financed Tuition Waiver Program (TWP), must comply with specific criteria for funding. 7.11 Today, Haiti has one of the world’s largest non-public sectors at the primary and secondary levels. According to the 2010-11 School Census (the most recent published), 94 percent of preschool, 60 Research Triangle Institute 2010; USAID 2012. -96- 78 percent of primary, and 73 percent of secondary students attend non-public schools.61 These shares are among the highest in the world. Slightly fewer than half of non-public primary schools are religiously affiliated, with Protestant affiliated schools making up the majority of these. Another 35 percent are privately owned schools, and the remaining 15 percent or so were founded and are run by communities. Among non-public secondary schools, 60 percent are non-religious and privately owned, and about 30 percent are religiously affiliated, with Protestant schools again making up the majority. While data from the 2012 ECVMAS show that poor children are more likely to attend public primary school, the majority of all children attend non-public schools. Figure 7.5 : Primary Schools, 1930-2011 (Thousands of Schools) 61 Preliminary results from the 2013-14 School Census suggest that these shares have stayed fairly stable from 2010-11 to 2013-14. -97- Box 7.1 : Education Provision in Haiti Formal Education in Haiti is structured in four levels (Figure below):  Preschool is meant to serve children from age 2 to 5, and is considered to have four levels based on these ages: poupons, petits, moyens, and grands. However, this structure is not formally mandated by public policy.  Fundamental education consists of three cycles. The first two cycles, consisting of 6 grades for children ages 6-11, are considered primary education. After these first 6 grades, children may enter into vocational programs, or continue to the third fundamental cycle (lower secondary) which consists of 3 grades for children ages 12-14.  Secondary: after lower secondary children may continue to secondary (upper secondary) education, which consists of 3-4 grades, depending on the model followed by the school, or enter vocational programs.  Superior (tertiary) education includes a range of university, technical, and vocational programs. Formal Education System in Haiti Sources: MENFP (Ministère de l’Education et de la Formation Professionnelle ) 7.12 International partners remain Figure 7.6 : Donor Financing for Education Sector, 2010-13 an important source of funding. As in (In Billions of Gourdes) many sectors in Haiti, donor assistance 9 for education has fallen since the large 8 commitments made immediately after the 2010 earthquake (Figure 7.6). 7 However, at nearly 2 billion gourdes in 6 2013, it remains an important source of 5 funds for the education sector. 4 Between 80 and 90 percent of funds 3 have been designated for primary 2 education in recent years, and the major 1 donors include the IADB, CIDA, World 0 Bank, WFP, UNICEF, UNESCO, USAID, 2010 2011 2012 2013 Source: World Bank staff estimates using data from Ministry of Finance (MEF) AECID, and AFD. Two of the largest donor-financed programs are the TWP and the national school canteen program (Programme National de Cantines Scolaires – PNCS). TWP finances a complete cycle of primary school for cohorts of children in disadvantaged communities by providing tuition waivers to non-public schools, and PNCS funds school meals and supplementation. These programs, financed by multiple partners, supported over one million students in the 2013-14 school year. -98- 7.13 Declining aid flows put into question, however, further progress and could even threaten recent improvements. While primary enrollment rates have increased substantially in recent decades, enrollment is still not close to universal. The most disadvantaged children – including the poorest, those living without their parents, and those with disabilities – remain out of school. At the same time, declining donor financing and a recent decision by the Ministry of Education to stop funding tuition waivers for new cohorts of first graders in non-public schools threaten the gains in access made in recent years. … While Public Spending Has Been Recently Rising … 7.14 Public spending in education has been increasingly recently. Over the past decade, the budget of the Ministry of Education has increased in line with the national budget and accounted for about 14 percent of the total national budget (Figure 7.7). The budget of the Université d’Etat d’Haiti, a body independent of MENFP, has also increased in line with the national budget, accounting for 1- 1.5 percent of the budget annually in 2012. Since the inception of PSUGO (Programme de Scolarisation Gratuite et Universelle) in 2011, public spending on education has, however, increased markedly. In an effort to fulfill students’ entitlement to a basic education, the current administration introduced taxes on international phone calls and money transfers, to be placed in a National Education Fund (Fonds National d’Education - FNE).62 Modeled after the donor-supported Tuition Waiver Program (TWP), PSUGO provides tuition waivers for students in non-public primary schools and fee waivers for students in public primary schools through a per-student transfer made to schools. Other types of support are also provided to schools, but information on precisely what these are is not currently available. During the 2012-13 school year, over HTG 2 billion were spent on waivers, supporting approximately 1.4 million students across over 10,700 schools (GoH 2014). 7.15 Overall, the addition of the tuition waiver programs and school canteen program have increased funding for primary education, and likely improved the equity of spending. While the two tuition waiver programs, PSUGO and TWP, as well as PNCS, are part of the investment budget, they primarily consist of recurrent expenditures, and exclusively support primary schools. Each program is targeted in different ways to more disadvantaged Haitians. Although 40 percent of students benefitting from PSUGO are in the West, PSUGO and the Tuition Waiver Program both provide waivers of USD90 per student, thereby self-targeting low-cost schools that serve poor areas (Figure 7.8).63 In addition, schools were selected for participation in the TWP to reach purposefully rural schools. The PNCS has targeted schools located in areas of nutritional vulnerability. 62 A US$0.05 fee is charged on every international phone call coming in and going out of Haiti, a US$1.50 fee is charged on every international money transfer. Although the FNE has not yet been approved by the Senate, the funds have been channeled through the national budget (outside of the MENFP budget) to finance PSUGO. 63 The waiver amount of $90 was set around 2007, based on earlier studies of the minimum costs of running a school in Haiti. New research is needed to understand how these costs have changed over time. -99- Figure 7.7 : MENFP Annual Budget Funded by Domestic Resources, 2005-13 (Billions of Gourdes and Percentage of Total Budget (Rhs)) 8 16 7 14 6 12 Investments 5 10 4 8 Operations 3 6 Salaries 2 4 MENFP (% of total budget), rhs 1 2 0 0 Source: World Bank staff estimates using data from the Ministry of Economy and Finance (MEF). 7.16 Implementation challenges have plagued PSUGO, however, and beneficiaries are likely overestimated. In 2013, the government’s anti-corruption unit (Unité de Lutte Contre la Corruption - ULCC) published the results of an audit of about 30 percent of schools listed as PSUGO participants. The report finds that across all schools, 25 percent more students are reported to the program than actually exist, and in 11 percent of schools, there are major over-counts of at least 50 students (ULCC, 2013). The report also cites several cases of “ghost” schools: duplicated names and school information in the PSUGO records, schools that could not be physically located, and some cases of known fraud (schools invented and registered in the program for the private gain of specific individuals). In addition, the report raises several concerns about how the program is run, including dozens of cases of over or under payment of transfers to schools, lack of verification mechanisms, and incomplete record keeping. Figure 7.8 : Student Beneficiaries by Department, 2013 (In Percent of Total) 100% 90% Sud 80% Sud Est 70% Artibonite 60% Centre 50% Grand Anse 40% Nippes 30% Nord 20% Nord Est 10% Nord Ouest 0% Ouest Source: Government of Haiti 2014; World Bank staff estimates using data received from MEF. 7.17 Even after the introduction of PSUGO, total public spending on education in Haiti is relatively low in a global context. Education spending in Haiti is about 3.5 percent of GDP, below most LAC -100- neighbors, as well as the global average for Low Income Countries (Figure 7.9). However, this spending must be considered relative to the share of the sector supported by public finances. Of the roughly 3 million young people in primary and secondary school in Haiti, about 50 percent are subsidized by PSUGO and/or are attending public schools. While this marks a substantial increase in the State’s role, a large share of students remain without public support. Figure 7.9 : Public Expenditures on Education (In Percent of GDP) 8 7 6 5 4 3 2 1 0 Source: World Bank staff estimates using data from MEF; World Development Indicators … Its Composition Is Not Conducive to Better Services … 7.18 MENFP’s operating budget is Figure 7.10: Education Budget, 2014 (domestic resources skewed towards non-“front line” staff. only, percentage of total) In 2013, for instance, almost 70 percent of the operating budget is allocated to Services Consum. 13% goods salaries and personnel costs (Figure 1% 7.10). This salary expenditure is in line Tangible with other LAC countries, where 60-80 assets 1% percent of the total education budget is Intangible absorbed by the wage bill (World Bank, assets Salaries 0% 2014). However, in these countries, the 69% public sector pays the majority of all Subsidies primary and secondary teachers, while and quotas Others in Haiti the State only directly employs 7% 9% those in the small public sector. Fewer Source: Ministry of Finance (MEF) than 400 inspectors are currently employed to serve the approximately 17,000 primary and secondary schools in the country, and their salaries account for less than 1 percent of the wage bill. In addition to providing public education, MENFP is meant to play a large role in regulating and supporting schools, particularly non-public schools. As a result, about 42 percent of salary expenditures go to staff who are not working as teachers, directors, or inspectors. -101- 7.19 Half of non-salary operating expenditures go towards textbooks, national exams, and the national school canteen program. MENFP both provides free textbooks and subsidizes students’ purchase of textbooks in public primary schools. This spending accounts for up to 24 percent of non- salary expenditures. A 2011 audit by the General Inspectorate of the Ministry of Finance (Inspection Générale des Finances – IGF) shows that approximately 66 percent of the textbook program budget went to subsidizing books, while the remaining 34 percent went to providing free books (MEF, 2012). While over 2.5 million books were purchased, the audit found that only 19 percent of a sample of public schools had actually received any books. IGF concludes that the textbook program lacks a clear policy under which to operate (including on criteria for targeting beneficiaries, or requirements to return free books) and that management of the textbook program is insufficient, with little documentation available to show that books were ultimately given or sold at subsidized prices to students. The administration of national exams accounts for 14 percent of non-salary expenditures, and the operating expenses of PNCS account for another 9 percent.64 This leaves slightly more than 1 billion gourdes, or 14 percent of the operating budget, for all other expenditures. … With Cost Remaining a Main Obstacle to Access … 7.20 Total spending on education in Haiti is relatively high. In total, approximately HTG 30-40 billion are spent on education in Haiti annually, with the majority of funds coming from households. Estimates based on expenditure data from the 2012 household survey suggest that in total, households spend over HTG 20 billion per year on education at all levels in Haiti. In the past two years, the State has spent over HTG 10 billion per year, and international partners have disbursed HTG 2-3 billion. In total, this spending adds up to over 10 percent of GDP. 7.21 Despite weak progression and learning, households spend a substantial amount on sending children to school. Among all households with children age 6-14 in school, 93 percent report positive education expenditures. These expenditures are substantial: on average, these households reported spending 10 percent of total annual household consumption on education (including the education of all students in the household) in the 2011-12 school year, and this share is equal for poor and non- poor households. Because poor households have more school-age children and lower total consumption, they spend less than half as much per child compared to non-poor households: HTG 3,600 compared to about HTG 11,400 per 6-14 year old child per year. Households may therefore sort into schools, particularly non-public schools, based on costs. Research using the 2002-3 School Census finds that school fees are positively correlated with school infrastructure (latrines, electricity), smaller class sizes, and more teaching materials (Demombynes, Holland, and Leon 2010). Poor households in particular spend relatively more on uniforms and books, and less on transport, compared to non-poor households (Figure 7.11). Against this backdrop, costs remain a major obstacle to education for poor households: when asked why school-aged children are not currently in school, 83 percent of households cite costs as the primary reason (ECVMAS 2012) 64 During the last years, there have been major investments on nutrition, health and school feeding programs funded by partners. The two largest school feeding programs, funded and implemented by the World Food Program, the Canadian Government, and the World Bank, covered nearly 600,000 children in 2014 at a cost of an estimated USD 30 million per year. -102- Figure 7.11: Education Expenditures by Type, 2012 (In Percent) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Poor Not Poor Tuition Books Uniforms Transport Other Sources: World Bank staff estimates using ECVMAS 2012 7.22 For poor households, tuition spending over the past decade has increased less quickly than for non-poor households, potentially due in part to the growth in public financing. In 2001, the bottom 60 percent of households spent about HTG 450 per child for primary tuition; this had risen to about HTG 1,700 in 2012, for an annual growth rate of 12.8 percent (IHSI 2003; ECVMAS 2012). In comparison, tuition for the top 40 percent of households grew at nearly 18 percent per year, and overall consumer price inflation averaged over 13 percent per year. This slightly slower growth in spending may be due in part to the growth of Government and donor financing in the sector. … And Disparities across Regions and Poor Services 7.23 Geographic disparities in service delivery remain. The West department has more schools per capita and generally better-resourced schools than other departments. While only 14 percent of primary students in the West attend public schools, its large population and high enrollment rates mean that the West has the highest number of public primary students among all departments. In general, schools in the West, and in urban areas more broadly, appear to be better-resourced: primary schools in the West have lower average student-teacher ratios, and are more likely to have a library and electricity (Table 7.3). Across departments, student-teacher ratios are in between the averages for LAC and Sub-Saharan Africa, while most departments have fewer electrified schools than the African average. While there appear to be a sufficient number of teachers, low levels of infrastructure across the country indicate that students are often lacking environments conducive to learning. The physical proximity of schools also varies across the country. For example, the number of schools per capita varies substantially across communes, ranging from 1 school (primary or secondary) up to 20 schools per 1,000 residents age 0-18 (Data sources: IHSI 2012; 2010-11 School Census). Comparing public to non-public primary schools, student-teacher ratios are substantially higher in public schools, but measures of physical infrastructure and resources are similar on average. -103- Table 7.3 : Characteristics of Primary Schools by Department, 2010-11 (In Percent) Percent of Schools with Share of Public Student-Teacher Canteen Library Electricity Enrollment Ratio (Units) Ouest 14 24.4 34 24 54 Nord 29 31.7 33 16 29 Nord'Est 31 36.8 46 14 16 Nord'Ouest 20 28.3 32 11 11 Centre 19 35.2 45 5 13 Artibonite 26 28.4 34 16 30 Sud 27 28.1 39 8 18 Sud'Est 31 29.1 36 14 16 Grand'Anse 31 32.3 26 8 8 Nippes 21 28.8 26 11 14 Sub-Saharan Africa 89 40.7 28 Latin America & Caribbean 83 20.8 Sources: UNESCO Institute for Statistics; World Bank EdStats, 2010-11 Haiti School Census. 7.24 The quality of service delivery in education is poor. Indicators of the quality of education, including teacher knowledge and learning materials available at the school level, suggest that many children, particularly poor ones, are receiving a low-quality primary education. Instructional quality and the provision of learning materials are generally believed to be very limited (MENFP, 2013). For example, in French language and math assessments of primary school teachers in the Central Plateau, where the questions were drawn from teacher training institute exams, only 10 percent (French) and 22 percent (math) of teachers were able to answer at least half of the questions correctly (MENFP and World Vision, 2013). Low quality contributes to high repetition and dropout rates, and ultimately to low educational attainment, as children with weak basic skills are unable to or gain little from completing primary and continuing to secondary. 7.25 Efforts to increase accountability and improve quality, particularly in the large non-public sector, are growing, but much remains to be done. In August 2014, MENFP announced several policy measures aimed at increasing public accountability and improving quality in the education sector. These measures include mandatory registration of all teachers and schools with the MENFP to be instituted through automatic, provisional granting of registration cards followed by processes of checking qualifications. In addition, new evaluations prior to the sixth grade will be instituted, and schools performing particularly badly will be placed under Ministry supervision. While these are welcome efforts, implementation and financing plans have not yet been articulated, and how these new plans fit with ongoing efforts is not clear. For example, how the new establishment identity cards for schools relate to the ongoing effort to modernize the school licensing system, which has been led by DAEPP (Direction d’Appui aux Ecoles Privées et Partenariat), has not yet been articulated. In addition, MENFP has been planning to administer the Early Grade Reading and Math Assessments (EGRA and EGMA) to nationally representative samples of third grade students, and it is not clear how the newly announced assessments will relate to these plans -104- TOWARDS A MORE EFFECTIVE AND BETTER TARGETED SOCIAL PROTECTION SYSTEM 65 This chapter examines public expenditures on social protection, with a particular focus on social safety nets (SSNs). The review shows a positive increasing trend in SSNs spending, which corresponds to the g overnment’s recent efforts to promote this area and implement new interventions under the umbrella of the Ede-Pèp Program. The overall SSNs system remains, however, fragmented, challenging the country’s weak administrative capacity. Going forward, the persistent low coverage and poor adequacy of interventions should be addressed, as well as institutional responsibilities clarified. Recent initiatives – such as the establishment of a national registry of beneficiaries and a national targeting tool – are key elements to improve the efficiency and equity of SSNs. Growing Investments in Social Safety Nets 8.1 In many countries, attempting to estimate public expenditures in social protection is arduous. Lack of consensus on the scope of social protection, fragmentation of interventions across Ministries, and/or a weak or nonexistent sectoral strategy make a discussion of social protection challenging (Grosh, 2004). These data challenges are exacerbated in a fragile and aid-dependent country like Haiti. Yet Haiti’s high levels of poverty and vulnerability and a policy momentum for reform have prompted the need for such a review. Social protection programs and policies, and especially social safety nets (SSN) - when correctly developed and appropriately funded - can directly address extreme poverty and help households both to manage shocks, and invest in their children and their assets. In this context, a review of public spending in social protection can help assess whether government’s policy and programs are going in a direction that contributes to objectives of poverty reduction and equity. 8.2 The state has traditionally played a limited role in providing a safety net for the poor and vulnerable in Haiti. Observers (e.g Lamauthe-Brisson, 2013) have pointed out that historically, Haiti never had a social movement that challenged in any meaningful way the laissez-faire of the state on the specific issue of social protection. The Duvalier dictatorship (1957-1986) left no room for such movements. Public social security organizations (insurance and social assistance) were created in the late sixties but remained poorly developed. The post-dictatorship period saw the creation of new institutions such as the social investment fund FAES (Fonds d’Assistance Economique et Sociale) in 1990. However, years of political instability (up until the crisis of Aristide’s departure in 2004), combined with the country’s constant exposure to economic or natural disasters, contributed to a 65 Prepared by Carine Clert, Frieda Vandeninden, and Victoria Strokova -105- context of constant emergency and prevented the development of long-term plans in which the provision of basic services and minimum social protection would have been essential elements. 8.3 The concept of social protection as a public policy that can contribute to poverty reduction and social risk management emerged progressively during the second half of the 2000s. Marked by relative political and economic stability, this period saw the emergence of some structural reforms. The 2007 National Growth and Poverty Reduction Paper (DSNCRP) covering the period of 2008-10 – the first poverty reduction strategy since the democratic election of René Préval in 2004 – listed poverty reduction and the improvement of living conditions as national objectives. This marked a milestone toward official recognition of the need to “protect” the poor. Emerging trade unions also approved a roadmap that included strategic recommendations for social protection (gender equality, development of contributory and non-contributory mechanisms for workers and vulnerable groups). Similarly, some donor partners began to show interest in food and nutrition security, which led them to consider a potential safety net agenda that could address both the short-term need to respond to constant shocks and the longer-term goal of reducing chronic poverty. 8.4 An exploratory review of public spending on safety nets during that period revealed that despite progress, safety nets still remained marginal. Existing programs at that time were insufficient to meet the challenges of dire poverty, constant exposure to weather shocks, high food prices, and weak human development indicators (Borgarello, 2009). Estimates of SSN spending reached 0.7 percent of GDP (including donors’ contributions), which ranked Haiti in the bottom quintile of low- income countries with low SSN spending. Lack of coordination of social protection policies and fragmentation among governmental entities also reflected the absence of a national social protection strategy. These conditions sparked partners’ interest in policy change. 8.5 The January 2010 earthquake halted this nascent interest in social protection and marked the beginning of three full years that were understandably dominated by humanitarian and short- term responses. The earthquake affected only a portion of the country (mainly West and Southeast regions) but caused over 200,000 deaths, considerable economic and infrastructure damage estimated at 120 percent of GDP. The government’s Action Plan for National Recovery and Development of Haiti (PARDNH) included the establishment of a social protection system as a critical factor for recovery and growth in Haiti. This post-disaster plan focused both on short-term goals (for immediate reconstruction) and long-term objectives (until 2030). Since no safety net structure existed that could be rolled out quickly, interventions had to be designed from scratch. With the massive infusion of resources, the earthquake both exposed and exacerbated the deficiency in safety nets by multiplying the “project approach”, interventions, and actors. 8.6 During the last three years (2012-14), there have been encouraging signs of increased attention to the social sector agenda, and more recently to social protection as a framework to address both short-term and long-term vulnerability and poverty reduction needs. The institutional apparatus was progressively rebuilt following the earthquake, and efforts toward reform were resumed, including improvement of economic governance and mid-term investment planning. Education was made a top policy priority. In 2011, the Government launched a national program to achieve universal access to basic education called PSUGO (Programme de Scolarisation Universelle -106- Gratuite et Obligatoire), which provides tuition waivers for students in public and non-public schools (as discussed in Chapter 7). In the health sector, the Ministry launched a National Health Strategy, an important attempt to improve coordination of interventions, to strengthen the Ministry’s stewardship role, and to rally donors around key priorities. In 2012, the national strategy for fighting hunger and malnutrition, ‘Aba Grangou’, was also launched, with the aim of coordinating donor and government activities in the nutrition sector. 8.7 A notable step toward the formation of a social protection strategy was taken in 2014 with the launch of the three-year action plan “Thinking and Fighting for a Haiti without Poverty: Action Plan for Accelerating the Reduction of Extreme Poverty” (PAARP). This document provides concrete instruments for the implementation of the social protection system and defines a set of programs constituting a social assistance strategy named EDE PEP. The PAARP also identifies a set of elements required to underpin the implementation of a social protection system, such as a national targeting system, a unique beneficiary registry, and an integrated service delivery model at communal level through a network of multi-sectorial agents (along the Kore Fanmi approach),66 as well as local coordination on social protection (GOH 2014a and GoH 2014b). 8.8 The national social assistance strategy “EDE PEP” represents the framework for several flagship government programs. The goal of EDE PEP is to protect the vulnerable population living in extreme poverty throughout their life cycle, in the short and medium term, to ensure long-term investment in human capital and to provide opportunities to overcome the condition of extreme poverty. EDE PEP is based on four complementary pillars: a) social inclusion, b) development of human capital, c) economic inclusion, and d) development of a decent environment. The key initiatives and programs under EDE PEP are listed in Figure 8.1 below and detailed in Annex 4, which provides a glossary of programs. 8.9 Most EDE PEP programs are financed, however, principally by concessional borrowing, which poses challenges in terms of their sustainability. As illustrated Figure 8.1, the source of funding varies according to each program: PSUGO is financed by special taxes,67 Kore Etidyan, Kore Ti Gran Moun, Restaurants communautaires and PNCS by the national treasury, while the rest of EDE PEP programs are financed through concessional borrowing from Venezuela through the Petrocaribe agreement. As previously discussed, the recent decline in international oil prices is substantially reducing the availability of this source of financing , putting into question the sustainability of many of these social protection programs. 66 Kore Fanmi (family coaching in Creole) is a Government initiative to link vulnerable families to available social services and provide them with some basic services and behavior change counseling. Both the World Bank and UNICEF are supporting this program. 67 The PSUGO is financed through the National Education Fund (Fonds National d’Education - FNE), which is financed by taxes on international phone calls and international money transfers (collected mainly through two companies, CONATEL and BRH). See the chapter 6. -107- Figure 8.1 : EDE-PEP SSN Financing, By Source, 2013 (Percentage of Total) Public treasury Special taxe 34% 45% Petrocaribe 21% Sources: Staff estimates using FAES (2013), Primature (2014), Ministry of Education (MENFP), Ministry of Social Affairs and Labor (MAST). Figure 8.2 : Programs Under EDE PEP EDE PEP Social inclusion Human Capital Economic initiatives Environment Recurrent Periodic Health Education Kore Moun Panye Family Ranje Kay planning PSUGO Kore Peyizan Andikape Solidarité Katie Kore Ti Fight against School Kantin Mobill cholera Ti Credit HIMO Gran Moun feeding Community Community Bon health Literacy Restaurants Solidarité centers programs Ti Manman Carte Rose Kore Etidyan Cheri Notes: HIMO = labor intensive public works. PSUGO = Programme de Scolarisation Universelle Gratuite et Obligatoire. Source: GoH 2014a 8.10 Meanwhile, various donors are also showing signs of increasing interest in social protection. The United Nations Development Program (UNDP) has supported the human capital approach of the newly established conditional cash transfer Ti Manman Chéri. The World Bank and United Nations Children’s Fund (UNICEF) are assisting the FAES with new models of support to and guidance of poor households in rural areas. The United States Agency for International Development (USAID) and the International Labor Organization (ILO) have been working with the Ministry of Social Affairs and Labor -108- (MAST) and striving to identify capacity-building needs. Efforts to coordinate with the government have also been reinforced. For instance, the recent development of a national targeting system includes the collaboration of different partners under the leadership of the government.68 8.11 The publication of poverty data, the first in a decade, marks an important milestone and transformational opportunity for evidence-based social protection.69 The EDE PEP framework, and in general any social protection strategy in a given country, should ideally respond to vulnerability and poverty needs. New evidence is now available with the household living conditions survey of 2012-13: this presents an historic opportunity to compare demand with supply, and to assess whether the latter is up to the task. This is discussed in later sections of this chapter. … Needed By A Vulnerable Population … 8.12 The chronic poor should be an important focus of an effective social protection system in Haiti. Although poverty is widespread in Haiti, as discussed in Chapter 1, rural areas and specific geographic departments require special attention from the social protection system. Nearly half the households are considered chronically poor because they live below the moderate poverty line and lack at least three of the seven basic dimensions defining non-monetary well-being (Figure 8.3).70 Almost 70 percent of rural households are considered chronically poor, compared with 20 percent in urban areas, highlighting the particular difficulty of emerging from poverty in rural Haiti. Nationwide, only 14 percent of households are transiently poor, and 12 percent of households reside above the poverty line but are deprived of access to basic services, suggesting that they lack the assets to improve substantially their well-being and remain out of poverty. Such a large share of chronic poverty, unusual in Latin America, highlights the structural challenges of poverty reduction and limitations of the contributions of social protection interventions without combined interventions around basic services and income generation. 68 The committee in charge of the targeting tool, led by MAST, includes representatives from FAES, as well as principal donors such as USAID, UNICEF, UNDP, WFP, the World Bank, and international NGOs such as CARE and Action Contre la Faim. The committee aimed to develop a targeting tool that will respond to the specific needs of two major programs, Kore Fanmi and Kore Lavi (nutrition and food voucher program), as well as to serve the country more broadly by means of a national tool. 69 World Bank (2015). 70 World Bank and ONPES (2014) uses the following definitions for chronic and transient poor: Chronic poor are living below the poverty line and lack access to a basic set of services: education (the household head is literate; all school-age children are in school); health (the food security index); water (access to an improved source of drinking water, including treated water); sanitation (access to improved sanitation); energy (access to a sustainable energy source); and habitat (a home constructed of non-hazardous materials). Households deprived in at least three of these dimensions are considered multi- dimensionally poor or chronic poor (López-Calva, 2013). The transient poor lack monetary resources, but have access to basic services and are more likely to be able to move above the poverty line. In other words, they are among the monetary poor but are not multi-dimensionally deprived. -109- Figure 8.3 : Chronic Poverty, Multidimensional Deprivation and Transient Poverty, 2012 (Percent of Total) 67 49 45 29 20 21 12 14 14 10 11 8 Chronic Deprived Transient Non-poor National Rural Urban Sources: ECVMAS 2012 and World Bank staff calculations. 8.13 Haiti’s social protection system should also be able to include the transient poor. Haiti’s social protection system should also be capable of deploying short-term responses given the constant exposure to shocks, which disproportionately affect the poor and the extreme poor. With respect to the transient poor, one million people live slightly above the poverty line and could be pushed below by a shock. In terms of vulnerability to shocks, while a typical Haitian household faces multiple shocks annually, nearly 75 percent of households were economically impacted by at least one shock in 2012. This percentage rises to 95 percent for the extreme poor. Natural disasters, in particular, have great disruptive potential partly because they have such a serious impact on agriculture, which represents the main source of livelihood for most of the poor, especially in rural areas. Indeed, the evidence shows that the most common covariate shocks are weather- or climate-related, while the most important idiosyncratic shocks are health-related. This carries important implications for ensuring financial access to health services for the poorest as part of a broad-based social protection strategy.71 8.14 Poor households face tremendous constraints to building, accumulating, and preserving human capital although it should be potentially their biggest asset.72 Both ECVMAS and other surveys heighten a potential role for social protection in bridging access gaps to nutrition and social services that have a bearing on human capital. Financial constraints and other demand-side barriers to access and use of education and health services appear to be high in Haiti, resulting in poor investment in human capital. As mentioned already in Chapter 1, the main barriers to human capital accumulation identified using the various data sources available are summarized below:  The poor in Haiti suffer from malnutrition early in life and from subsequent food insecurity. According to the 2012 Demographic and Health Survey (DHS) one-fifth of children under five are chronically malnourished. The national food insecurity rate was 38 percent in 2011 and 41 percent for rural areas (CNSA 2011). According to ECVMAS 2012, poor households were much 71 Covariate shocks affect large shares of the population/community (such as natural disasters or epidemics), while idiosyncratic shocks affect individuals (such as sickness, death, or job loss). 72 Human capital is defined here broadly as a set of intangible assets, skills and knowledge that can create economic value and generate more valuable labor outcomes. -110- more likely to report repeated lack of food or hunger when going to bed than non-poor households. Importantly, households with young children (under the age of 5) are much more likely to experience repeated food shortage. This is a particular cause of concern, as proper nutrition in early life is crucial for brain development and subsequent life outcomes (Alderman and King, 2006).  The poor are much more likely to be illiterate (30 percent for the poor versus 15 percent for the non-poor) and to face lower levels of education achievement (7 years of education for the non-poor versus 4.4 for the poor). Poverty is also much higher for households headed by persons with no education (77.6 percent of households headed by a person with no education are poor).  Poverty proves to be an important barrier to school enrollment. This is further supported by the results derived from asking directly why children are not in school: costs are cited as the primary reason by 83 percent of households (ECVMAS 2012).  The poor also face higher barriers to health access. Financial barriers are the key obstacle to access to health care for the poorest. The 2012 DHS pointed out that 76 percent of women age 15-49 declared that lack of money deterred them from seeking health treatment.  Finally, poor households are also more likely to resort to strategies that are harmful to human capital accumulation. When there is a covariate economic shock in the community, a staggering 56 percent of households in extreme poverty change their nutritional behavior as opposed to 37 percent of the non-poor households. The two events which are most likely to lead to the removal of a child from school are changes in household composition (death or birth of a household member) or decrease in monetary help from outside of the household. 8.15 The different risks supported by the population require different responses in terms of social protection policies. Looking at poverty headcounts by age group and risks across the life cycle also offers important insights regarding where public interventions in social protection should go. The main findings are summarized below and key facts are detailed in Figure 8.4. The main policy implications – or demand in social protection - ensuing from each key fact on risks borne by the population throughout the life cycle, are summarized in Figure 8.5.  Poverty is highest among children and relatively lower among adults. Almost 70 percent of pre-school age children (below 6 years old) live in poor households. While still very high, poverty among adults is about 53.3 percent and almost half of all poor are adults aged 18-64 years old. School-age children represent about 30 percent of all poor, with the second highest poverty rate of 64.1 percent. Poverty rate among the elderly (above 65 years old) is relatively lower, but higher than that for adults. Poor elderly represent the smallest share of all poor (about 5 percent), in line with their small share in total. The major risks are not attending school or dropping out of school for several reasons, particularly related to money or early pregnancy.  A non-negligible number of children are involved in child labor and many continue to serve as “restavèks”, who work as domestic servants outside of their parents’ household. Restavèks -111- are difficult to identify in household survey data, yet some studies have shown the significance of this problem.73  Many adults (18-64 years old) in Haiti face a risk of unemployment or not earning sufficient income. This underscores the need to reinforce the “promotive” features of a social protection system, as poor households need to be connected to skill-enhancing and better income- earning opportunities. The principal risks facing adults are: unemployment, under- employment, low or variable income, informality, working but not earning enough income to cover basic needs (working poor), unstable livelihoods and lack of access to physical and financial capital.  Particular risks for women were highlighted. The unemployment rate for women is twice as high as for men. The disparity is even higher in rural areas, where women are almost three times more likely to be unemployed than men. In rural areas, female-headed households have lower access to agricultural inputs such as seeds, which can lead to lower productivity, thereby creating a gender gap.  Youth (15-24 years old) were also found to face additional challenges entering the labor market, with the lowest rates of participation and employment, and the highest rates of unemployment and informal employment.  Finally, the elderly (65 years and older) in Haiti are vulnerable to poverty and have to rely on support from their families. The main risk for the elderly is the lack of any pension (contributory or non-contributory scheme) and/or access to healthcare, and the dependency on family and charity for survival. Given demographic dynamics in Haiti, the elderly tend to be somewhat neglected by antipoverty programs. As indicated above, the elderly represent less than 5 percent of the poor (Figure 8.4); however, poverty is still prevalent among this group as more than half of those above 65 are poor. Figure 8.4 : Poverty Headcount by Age Group, 2012 (In Percent) 69.1 64.1 56 53.3 52.6 47.7 30.5 28 17 14.4 4.8 5 Child (<6yo) School age (6-17yo) Adult (18-64yo) Elderly (65+yo) Porverty headcount Share of poor Share of population Sources: World Bank Staff calculations based on ECVMAS 2012. 73 While ECVMAS 2012 includes only 91 observations with household members indicated as “Domestique = restavèk”, a 2009 study by the Pan American Development Foundation found, for instance, that there may be as many as 225,000 restavèks living in Haiti (Pierre et al. 2009). -112- Figure 8.5 : Key Risks, the Life Cycle, and Social Protection in Haiti Key facts on risks • Malnutrition: 22% suffer from chronic malnutrition (DHS 2012) • Mortality: Mortality rate is 92 deaths per 1000 live births, nearly 6 times higher than the Young regional average of 16 (WHO) children Key implications • Structural programs aimed at malnutrition addressing supply and demand barriers (ages 0-5) • A more comprehensive early childhood policy Key facts on risks • Non-enrollment or school drop-out rates: approx. 200,000 children (ages 6-14) are estimated to have dropped out of school (ECVMAS 2012) • Child labor: 20% of all children are involved in some form of work activity, 6% of children (ages 10-15) in the poorest quintile work and don’t attend school School age • Restavecs: up to 225,000 children are restavecs (PAHO 2009) children Key implications (ages 6-17) • Structural social promotion programs to cover the direct and indirect costs (opportunity costs of child labor) of education for poor children Key facts on risks • Unemployment: 28.3% and up to 38% in urban areas and in Port-au-Prince • Not earning adequate income: 80% of the rural poor live in households headed by a working person (61% in urban areas and 56% in Port-au-Prince) • Gender: Women are at a disadvantage in the labor market Key implications Adults • Urban areas: temporary income generation programs that also incorporate improvement of training and skills-building (ages 18-64) • Rural areas: holistic programs for the extreme poor combining consumption smoothing and food security, productive projects and access to financial capital • Address specific constraints for women (care for children, the elderly or disabled) Key facts on risks • Lack of stable income: more than half of people over 65 are poor • Lack of access to health care or other care Elderly Key implications • Targeted social pensions, solutions for access to health care or other care (ages 65+) … But Too Fragmented … 8.16 The various programs building Haiti’s social protection system are excessively fragme nted. Within Government’s public spending on social protection, both the contributory (social insurance) -113- and non-contributory pillars (social safety nets) are spread across multiple institutions. While such dispersion may not be unusual, it is problematic for a country with limited resources. Furthermore, this fragmentation risks overwhelming the country’s limited fiscal and administrative capacity, and reducing the impact of public expenditures. 8.17 Social insurance: Social insurance schemes are implemented by three different institutions, depending on the program type (health and pensions) and work activity (public/private) (Table 8.1). Contributory schemes for private sector workers are managed under the general authority of the Ministry of Social Affairs and Labor (MAST), but two separate institutions are responsible for the implementation of private social insurance schemes, financed by employers’ and employees’ contributions: (i) the Office of Work Accidents, Illness and Maternity Insurance (OFATMA), which is responsible for providing work injury and invalidity benefits to private workers; and (ii) the National Office for Old Age Insurance (ONA), which manages the private pension scheme. Regarding public sector workers, until recently health insurance and pension schemes were implemented by two different institutions: (i) the Directorate for Civil Pensions (DPC) in the Ministry of Finance (MEF), which manages pensions for civil servants and former military personnel and covered about 50,000 Haitian civil servants and their dependents in 2012 (300,000 people in total); and (ii) the private company for health insurance of public sector workers, the Groupe Santé Plus. Since 2014, health insurance for both public and private workers has been consolidated under a single umbrella, OFATMA. Table 8.1 : Social Insurance: Institutions and Programs in Haiti, December 2014 Ministry and Programs Target group Institution MAST – ONA Private pension Private sector workers MEF - DPC Public pension Civil servants MAST-OFATMA Public health insurance (since Oct. 2014/formerly Civil servants provided by Groupe Santé Plus) Disability insurance Private sector workers Work related insurance Private sector workers 8.18 Social Safety Nets (SSN): Government spending on social safety nets is fragmented across at least four Ministries and involves more than 20 programs, depending on the scope and list of programs included. Given the objectives of this paper, and as explained in the introduction, this paper considers the following categories of spending: (i) all social safety net programs under the EDE PEP flagship umbrella;74 and (ii) social subsidies for electricity and petrol. As illustrated in Table 8.2 below, nine Ministries share oversight over different EDE PEP programs and 11 public agencies and/or Ministries are responsible for their execution.75 In addition, the government provides universal subsidies for electricity and petrol, which are managed by the Ministry of Economics and Finance (MEF). Donor support is equally fragmented. 74 Definitions of EDE PEP programs are available in Annex 8.1. 75 Beyond EDE PEP, other small social safety net interventions and ad-hoc initiatives have been identified, such as those implemented by the social assistance fund CAS (training centers) and the Institute for Social Welfare and Research (IBERS) under the authority of the MAST. Other institutions are likely to finance SSN (e.g. BSEIPH and ONM) but data was not available at the time of this report. -114- Table 8.2 : EDE PEP Fragmentation: Institutions and Programs by October 2014 Oversight Ministry Executing Ministry or Public EDE PEP Main target Program (Ministère) de Tutelle) Agency Programs Groups Type MAST Caisse d'Assistance Sociale Kore Ti Gran Moun Elderly Cash transfer (CAS) Kore Moun Andikape Elderly/Disabled Cash transfer FAES Kore Ti Gran Moun Elderly Cash transfer FAES / BSEIPH Kore Moun Andikape Disabled Cash transfer MAST Resto Communautaire n.a. Food FAES Resto Communautaire : n.a. Food Resto Pèp (Closed) FAES Panye Solidarite - - FAES Kantin Mobil n.a. Food OFATMA Carte Rose n.a. Access to Health Services FAES Kredi Ti Biznis (Ti Kredi) Women-headed hse Access to credit Ministry of Human Rights FAES Ti Manman Cheri Children Conditional Cash and Fight Against Extreme Transfer Poverty Ministry of Education MENFP School meals PNCS Students Food (MENFP) MENFP PSUGO (School Tuition Students Fee waivers Waiver Program) Secrétariat d'Etat à Literacy Programs Illiterate adults Training l'Alphabétisation FAES Kore Etidyan Students Cash transfer (scholarship) Ministry of Health MSPP Family Planning Women from 15 to In kind (MSPP) 49 years old Fight against Cholera n.a. In kind Commun. health centers n.a. In kind Ministry of the Promotion Kore Peyizan In kind of Peasantry (MARNDR) FAES  Kore plantè (seeds and Poor Farmers tools) FAES  Kore pechè (fishing Poor Fishermen equipment) MARNDR  Distribisyon betay Poor Farmers (distribution of livestock) Ministry of Women’s FAES Kredi Fanm Lakay (Ti Women in Credit Conditions (MCFDF) Kredi) agriculture Ministry of Interior FAES avec appui Direction Bon solidarité (bon Poor (emergency Cash transfer (MICT) Protection Civile dijans) situtation) Prime Minister’s Office. UCLBP (Unité de Construction Ranje Kay Katie Poor Refurbishment of houses (Primature) de Logement et Bâtiments in poor neighborhoods Publics) Ministry of Public Works SMCRS (Service Métropolitain HIMO (Labor intensive Unclear targeting. Cash for work (MTPTC) de Collecte de Résidus Solides) public works) Self-targeting Source: Staff based on Review of EDE PEP Official Documents (GoH 2014) and Key Informants’ Interviews -115- … Too Small … 8.19 Overall public expenditure levels for social protection in Haiti reached about 5 percent of GDP in 2013, when (i) contributions to social insurance (1.2 percent of GDP); (ii) public spending on social safety nets (all EDE PEP or 1.4 percent of GDP); and (iii) subsidies for electricity and petrol (2 percent of GDP) are included.76 However, SSN spending (Figure 8.6) is strikingly low compared to that on universal subsidies that are regressive and fail to protect the poor and vulnerable against shocks (Box 8.1). Figure 8.6 : Public Expenditure on Social Protection, 2013 (Percentage of GDP) 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Social Insurance SSN Subsidies Sources: World Bank staff estimates using data from Ministry of Socal Affairs and Labor (MAST), Ministry of Finance (MEF), BOOST, FAES (2013) and Primature (2014). 76 Social insurance contributions do not represents the actual payments to insured beneficiaries (pensioners and beneficiaries of health insurance). Actual payments - for which data are not available - are financed out of contributions and therefore represent only a subset of overall contributions. -116- Box 8.1 : Methodology Note on Social Protection Spending Given the multiplicity of actors, definitions of social protection, and lack of systematic data reporting, data collection for a public expenditure review on social protection is an arduous task. A comprehensive overview of spending on social protection requires disaggregated data spending (i) from the various social Ministries (MAST, MENFP, MSPP) and their different agencies; (ii) from the social insurance agencies within MAST and MEF (ONA, OFATMA, DPC and Groupe santé plus); (iii) from MAST agencies responsible for social safety nets (CAS and IBERS); and (iv) from the FAES ( Fonds d’Assistance Economique et Sociale ). As a result, the choice was made to focus primarily on 2013 and to present time-series data only when available (EDE-PEP programs) and feasible. Similarly, only major donors’ contributions to social safety nets (programs and technical assistance) are listed. 1. Social Insurance: (Contributions): o ONA and OFATMA (private schemes): data provided directly by the MAST o Public health insurance: data provided directly by MEF o Public pensions: BOOST database, World Bank 2014 2. Social Safety Nets: o EDE-PEP programs: FAES (2014), “EDE PEP, Programme nationale d’assistance Sociale: 1 an de mise en œuvre - 2 ans d’existence ”, Rapport 2013-2014 and FAES (2013), “EDE PEP, Programme nationale d’assistance sociale: Rapport 2012-2013, 1 an de mise en œuvre”, Rapport 2012-2013 o PNCS : data provided by MENFP o PSUGO : Primature de la République d’Haiti (2014), ‘Bulletin d’information Programme de Scolarisation Universelle Gratuite et Obligatoire’, Bulletin vérité, N°1, February 2014 o MAST Community restaurants: MAST (2013), ‘Rapport de suivi et d’évaluation du projet: programme de protection et d’assistance sociales sous-programme de restaurant communautaires’, Port -au- Prince, Novembre 2013 o SSN from CAS relating exclusively to programs mapped to EDE-PEP: data provided by MAST’s Directorate of Studies: (Kore Ti Gran Moun; Restaurant Communautaire). 3. Petrol and electricity subsidies: Data provided by MEF 8.20 Social Insurance. Social insurance contributions (from employers and employees) represent about 1.2 percent of GDP in 2013. Old age pensions for private sector workers is the largest scheme, as contributions to pensions for public sector workers, public health insurance and disability benefits for private workers reach only 0.11 to 0.13 percent of GDP.77 Currently, about 400,000 workers are insured, yet only 3,000 retired receive benefits. Contributions to public pensions, including old age benefits for retired military, military staff and other civil servants have recently declined. The number of beneficiaries has remained stable since 2010, with about 11,000 civil servants receiving old age benefits. Contributions to health insurance (public scheme) have also been steady over the last five years. Health insurance for private workers only provides work-related disability benefits and maternity benefits but does not offer health care insurance. Since 2012, OFATMA has attempted to provide free health care to specific target groups through the distribution of insurance cards (‘Carte Rose’) on a pilot basis. However, no specific analysis of the experience is available to date.78 77 Contributions to public pension schemes correspond to government contributions only. 78 The Carte Roz pilot program aims at evaluating the costs of the access to the expected health package. In 2012-2013, about 400 health insurance cards were distributed and in fiscal year 2013-2014, the pilot extend to a group of university students but no additional data (including on the costs involved) is available. -117- 8.21 Social Safety Nets. Government’s social safety net spending has reached about 1.4 percent of GDP in 2013. Table 8.3 outlines spending data for EDE-PEP programs executed by MENFP, MAST and FAES for 2013.79 Table 8.3 : EDE PEP Executed Spending (Actual) in 2013 Program Title HTG In % In % of GoH Funding of GDP Expenditure Source PSUGO 2,229,097,353 0.61 2.94 Special tax Community Restaurants 550,000,000 0.15 0.73 Public treasury Panye Solidarité 490,673,688 0.13 0.65 Petrocaribe Kore Ti Gran Moun (MAST rogram 461,964,112 0.13 0.61 Public treasury for elderly and disabled) Kore Etidyan 443,322,200 0.12 0.58 Public treasury Ti Manman Cheri 265,507,530 0.07 0.35 Petrocaribe PNCS 235,000,000 0.06 0.31 Public treasury Bon Solidarité (bon dijans) 111,527,353 0.03 0.15 Petrocaribe Kore Peyisan 92,775,257 0.03 0.12 Petrocaribe Kantin Mobil 72,870,077 0.02 0.10 Petrocaribe Kore Moun Andikape 1,548,077 0.00 0.00 Petrocaribe Total 4,954,285,648 1.36 6.53 Source: FAES 2013, Primature 2014, MENFP, key informants’ interviews and WEO database. 8.22 This evolution represents an important increase compared with past spending and modifies the relative position of Haiti vis-à-vis other low income countries in terms of SSN spending (Figure 8.7). Borgarello (2009) estimated that SSN represented 0.7 percent of GDP in 2008, including donor funding and in these estimates, government spending represented only 22 percent of total SSN. This recent increase in public SSN is mainly a consequence of the introduction in 2011 of the PSUGO fee waiver program and of other EDE PEP programs starting 2013. In fact, when analyzing the composition of EDE PEP, it is useful to distinguish between PSUGO as a large tuition waiver program (representing 0.6 percent of GDP) and other EDE PEP programs (representing 0.8 percent of GDP).80 In this context, spending on social safety nets has recently grown in comparison to that of other low-income countries. 79 According to the preliminary budget, other EDE-PEP programs (literacy, HIMO, cholera, community health centers) were going to be implemented from 2013 or 2014 on (GoH 2014a and Key informants’ interviews). 80 PSUGO is presented separately as its funding is atypical. SSN executed by MAST include community restaurants (restaurants communautaires) and an aggregate figure for social protection for the elderly and vulnerable. The assumption was made that the latter refers to the social program for people with disabilities under EDE-PEP, Kore Ti Gran Moun. -118- Figure 8.7 : Social Safety Net Spending – International Comparison, 2013 or latest (Percent of GDP) 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.61 0.5 0.75 0.0 SSN spending Spending on EDE PEP PSUGO Sources: Wolrd Bank staff estimates using dta from FAES (2013), Primature (2014), Ministry of Education (MENFP), Ministry of Socail Affairs and LAbor (MAST) and IMF. 8.23 In addition, donors’ interventions are widely fragmented, which compounds the difficulty of estimating exact contributions (Table 8.4). In addition to the concessional borrowing that finances most EDE- PEP programs, other SSN interventions are financed by various grants and implemented by external partners. Some examples include: the food voucher Kore Lavi program financed by USAID (budget of about USD30 million, included in the social protection budget of USD80 million over 3 years) and implemented by CARE; various school canteen programs financed by multilateral organizations (UNASUR, BDC) and by bilateral donors (AFD and Canada) and operated by World Food Program (WFP), the World Bank and the donors’ Partnership Education for All, which is coordinated by UNESCO; and other bilateral projects, such as milk distribution financed by the Brazilian cooperation (budget of USD3 millions) and implemented by WFP.81 Partners also contribute to the financing of some government programs with a technical assistance component: this is the case of the community agent (social worker) initiative Kore Fanmi, implemented by FAES and financed by UNICEF (for $547,418 in 2012). Table 8.4 : Key Donors’ Contributions (Estimates) - Social Protection. Partners Budget in $US Period Budget per year ($US million) USAID 80 millions 4 years (from 2013) 20,00 UNASUR 1 million 1 year (2013) 1,00 Multilateral UNICEF 1.9 million 3 years (2012, 2012, 2014) 0,63 World Bank 24.5 2 years (2013-2014) 12,25 UNDP 1.7 million 3 years (2013-2015) 0,57 Caribbean Development Bk. NA NA AFD 10 million 2 years (2013 and 2014) 5,00 Bilateral Canada NA Brazilian cooperation 3 million 3 years (2011-2013) 1,00 Partial estimate of donors spending ($US million) 40,45 Partial estimate of donors spending (% of GDP) 0,46 Source: Key Informants 81 Donors and NGOs devote substantial resources to the school feeding programs. The World Bank and other donors are supporting the government in developing a school feeding policy that would clarify roles, responsibilities, accountability, objectives, financing, and would make the use of the resources more efficient (See Education Chapter). -119- 8.24 Subsidies. The level of government spending on subsidies has increased since 2007 and represents the largest share of the total GoH’s spending on social protection. Government spending on subsidies represented almost 3 percent of GDP in 2013. According to estimates from the Borgarello (2009), the surge in spending was initially driven by the increase in electricity spending (from 0.84 percent in 2007 to 2.1 percent in 2012), as fuel subsidies actually decreased during the period. A major concern, however, is the issue of regressivity of government subsidies. For instance, in the case of fuel subsidy: 95 percent of the subsidy accrues to the richest quintile (see Chapter 9 for a more detailed discussion). … Not Well Targeted … 8.25 ECVMAS data points to the urgent need to improve the targeting of social assistance benefits, since as much as half of social assistance benefits accrues to the non-poor.82 The share of beneficiaries in the first two quintiles is less than half, but the size of transfers tends to be bigger for higher quintiles, leading to a much more regressive distribution of benefits (Figure 8.8).83 This is especially the case for other transfers,84 but also holds for food aid and, to some extent, for scholarships. Figure 8.8 : Distribution of Social Protection Benefits (Benefit Incidence), 2012 (Percentage of Total) 100 90 80 70 60 50 40 30 20 10 0 All social Scholarship Other transfers Food aid assistance Q1 Q2 Q3 Q4 Q5 Source: World Bank calculations using ADePT, based on ECVMAS 2012. 8.26 A major finding of the most recent household survey is that access to social security (contributory programs) is out of reach for most Haitians, especially the poor, leading to a lack of protection in old age or in case of sickness or disability. Only salaried employees working in the formal sector have access to the limited social insurance schemes existing in Haiti. Among the active 82 This share is reflective only of the programs captured in ECVMAS. It is currently not possible to estimate what share of other assistance provided by the Government or NGOs goes to the poor. 83 Note that households are ranked on consumption net of social assistance transfers, so this is not simply as a result of household moving up the quintiles due to the transfers themselves. 84 The number of observations for other transfers is very small, so these estimates are less reliable. -120- population; i.e. those in the labor force, employees in wage employment constitute only one-fifth, which corresponds to less than 10 percent of the population. 8.27 Because of high levels of informality, only 11 percent of wage employees have access to social security, primarily concentrated in the upper quintiles of the population.85 Of wage workers, only a small share (11 percent) have access to social security, while the overwhelming majority do not (Figure 8.9). Access to social security is greatest among individuals in the richest quintile. Two-thirds of employees with social security belong to the top quintile, while only 5 percent are from the second poorest quintile. Given the prevalence of informality in rural areas, access is concentrated in the urban areas, particularly in Port-au-Prince. Figure 8.9 : Access to Social Security by Income Group, 2012 (Percentage of Total) Q1 Q2 25 0.3% 5.0% 20.9 Q3 7.8% 20 15 11.4 Q4 9.7 20.4% 10 4.6 4.8 Q5 5 66.5% 0.4 0 Q1 Q2 Q3 Q4 Q5 Total Source: World Bank calculations based on ECVMAS 2012 Source: World Bank calculations based on ECVMAS 2012 8.28 Access to health insurance through employment in a firm registered in OFATMA is also very low. A very low percentage (less than 4 percent) of the Haitian population has access to health insurance administered by OFATMA. Most of the households with OFATMA belong to the highest consumption quintile and live in the Metropolitan area. This is not surprising given that OFATMA is only available to employees of formal firms and their families, and receives social contributions from both employers and employees on a voluntarily basis (Cross et al.). As discussed earlier, there could be potential signs of future increases in the coverage of health insurance, notably with the pilot introduction of ‘Carte Rose’, a voluntary contributory health insurance which aims to extend coverage to the entire population, including informal workers. However, the pilot seems still limited. 8.29 Because they lack access to contributory programs, poor Haitians have limited protection against poverty in old age or in case of disability or sickness. Very few people are eligible for contributory pensions when they retire, and those who are tend to be much better off. ECVMAS 2012 data shows that only 2.6 percent of the elderly (65 years and older) receive pensions (old age, disability) and the majority of them are non-poor. Pension beneficiaries overwhelmingly reside in 85 For the purposes of this note, individuals employed as wage earners who contribute to social security or receive social security benefits, such as paid sick leave or maternity/paternity leave, are considered to have access to social security. The questions used for this analysis refer to individual employees. A set of questions referring to whether a firm is registered with ONA or OFATMA provides similar levels of potential access to social security schemes. -121- urban areas (92 percent) and almost half (43.2 percent) live in the PaP metropolitan area. These results are consistent with the fact that access to social security is very limited in rural areas. 8.30 Turning to SSN coverage, social assistance coverage, is alarmingly low and well below the identified needs.86 According to ECVMAS 2012 data, only about 8 percent of the Haitian population received non-contributory social assistance benefits such as scholarships, food aid or other transfers in 2012 (Figure 8.10a).87 Overall, social coverage seems to be progressive: about 12.5 percent of the population in the poorest quintile receive some social assistance benefits compared with about 5 percent in the two richest quintiles. A positive signal, given the high prevalence of rural poverty, is that overall coverage is slightly higher in the rural areas, primarily because of the larger share covered by food aid (8.8 percent compared to 5.3 percent in urban areas).88 More than 60 percent of food aid beneficiaries reside in rural areas (Figure 8.10b). However, the beneficiaries of scholarships or other transfers are slightly more likely to be in urban areas. Figure 8.10: Coverage of Social Assistance Programs and Distribution of Beneficiaries a. Population covered, 2012 b. Beneficiaries, urban vs rural, 2012 (Percent) (Percentage of total) 14 100% 12.5 11.7 12 11.3 80% 10 9.6 8.3 60% 8 7.1 6.8 6.0 40% 6 5.4 5.2 4.7 4.0 4 20% 1.7 2 0.9 0.9 0.9 0.5 0.3 0.3 0.5 0.4 0.4 0.3 0% 0.2 0 All SA Scholarship Food aid Other Poor Total Q1 Q2 Q3 Q4 Q5 transfers All social assistance Food aid Scholarship Other transfers Urban Rural Source: World Bank calculations based on ECVMAS 2012. Source: World Bank calculations based on ECVMAS 2012. 8.31 Social assistance coverage of various population groups is, nevertheless, not even: young children are underrepresented among social assistance beneficiaries, which is a concern given their vulnerability. Looking at the coverage by age group (Figure 8.11), children under 5 have the lowest coverage, with only 7.4 percent of all children under the age of 6 benefiting (indirectly) from social 86 This section uses two main methods to estimate how much of the social safety nets actually go to the poor. First, it draws on the ECVMAS 2012. Second, it uses administrative data from FAES to shed light on the extent of coverage of the EDE PEP programs, which are not reflected in the ECVMAS. 87 Social protection programs are very fragmented and often small in scale and coverage, so without surprise household survey data don’t capture many beneficiaries of such programs. This small number of observations should be kept in mind when interpreting the results. Preliminary findings from ECVMAS 2013 also confirm that overall coverage is low, on the order of 16 percent for social protection and about 13 percent for social assistance, not including assistance from NGOs and religious organizations, coverage of which is estimated at about 5.5 and 0.8 percent of the population, respectively. 88 In contrast, more than half of the population benefits from remittances, which arguably play a role of an informal safety net in Haiti. -122- assistance benefits. This is of particular concern given that this group suffers from the highest poverty rates, as shown earlier. While coverage of school-age children is also quite low, they are much more likely to benefit from programs targeted at schools, such as free school meals (PNCS) or PSUGO, which are not captured in the survey. EDE-PEP seems to lack sufficient focus on early childhood. The plan for reduction of extreme poverty includes a few programs in the early childhood window; however, most of the interventions (community restaurants, disaster response programs, and health interventions) are insufficient and not tailored to the needs of this age group. For example, health insurance is available only in urban areas and is contributory, so it is unlikely to reach the most vulnerable, while community pharmacies do not focus on preventative health and malnutrition, a critical priority for young children. There is also little available to prevent malnutrition or proactively improve children’s development potential early in life. Second, although child labor and restavèk are a considerable phenomenon, this risk is not addressed by the programs under the EDE PEP strategy. Third, half of people above 65 are poor and coverage of contributive pensions is very limited, yet the non- contributory cash transfer program only will be able to cover 30,000 people. Figure 8.11: Coverage of Social Assistance Programs by Age Group, 2012, (Percentage of Total Population) 9.7 Elderly (65+ yo) 8.5 0.0 1.3 7.9 Adult (18-64 yo) 7.0 0.7 0.1 8.0 School age (6-17 yo) 6.8 0.8 0.1 7.4 Child (0-5 yo) 7.4 0.0 0.0 0 2 4 6 8 10 All SA Food aid Scholarship Other transfers Source: World Bank calculations using ADePT, Note: Direct and indirect beneficiaries 8.32 The geographic distribution of government spending on its flagship, pro-poor safety net initiatives could be better aligned with the poverty incidence. Administrative data available from FAES on the extent of coverage for the last two years shows low coverage of some EDE PEP programs in regions with the highest poverty rates. The Centre region, with a poverty rate of 73 percent, has a similar coverage in terms of in-kind and in-cash transfer to the Ouest region (poverty rates of 37 percent). Similarly, Nord-Ouest is the region with the highest prevalence of poverty (81 percent), but in-kind coverage is not higher than average. Coverage of cash transfers (Ti Manman Chéri, Kore Etidyan, Kore Moun Andikape, Bon Solidarite, Bon Dijans) is lower than in-kind programs in every region (Figure 8.12). Furthermore, meal distribution is not aligned with regional incidence of poverty, but is mainly concentrated in the capital. In 2012, the program of meal distribution Kantin Mobile was especially active in Port-au-Prince (Figure 8.13). This is not surprising as this program focuses on poor urban areas, to the detriment of rural areas where poverty and food insecurity are relatively greater. -123- Figure 8.12: EDE PEP Coverage (FAES-Executed) by Region and Program Type, 2013 (Percentage of Total Population) 14 90 12 80 70 10 60 8 8.0 50 6 40 30 4 3.2 20 2 10 0 0 Cash In-kind Poor Source: data from FAES 2013 Figure 8.13: Numbers of Meals Distributed, Kantin Mobile, 2013 (Percentage of Total Population) 500 100 450 90 400 80 350 70 300 60 250 50 200 40 150 30 100 20 50 10 0 0 Meals distributed (thousands) Poor (% of pop), rhs Meals distributed (% of pop), rhs Source: data from FAES 2013 … And of Limited Impact 8.33 Unsurprisingly, ECVMAS indicates that social protection programs have limited impact on poverty and inequality because of their low coverage and small value. Without social protection transfers, including pensions, poverty headcount would be less than one percentage point higher compared to the current poverty rate. ECVMAS data also points out that the value of most social assistance benefits (cash or food) is very small and, hence, these benefits contribute relatively little to the consumption of beneficiaries. With the exception of other transfers that are larger in absolute terms, other social assistance benefits are much less generous Figure 8.14, chart a). Scholarships, albeit small in absolute value, also contribute a large share (almost 37 percent) to the poorest quintile’s consumption (Figure 8.14 chart b). Overall, however, social assistance benefits contribute only 10.6 percent to the poorest quintile’s consumption. Contribution to consumption tends to decline as -124- quintiles grow richer: for richer quintiles consumption is larger compared to the value of benefits, so benefits are relatively more important for the poorest. Figure 8.14: Benefit Amounts and Contribution to Consumption of Beneficiaries a. Average annual per capita transfer by benefit type, b. Contribution to consumption of beneficiaries, 2012 2012 (in Gourdes) (Percentage of total consumption per item) 35 8.3 6.2 Food aid 6.4 30 4.6 4.8 25 36.8 2.4 Scholarship 2.0 2.5 20 3.5 48.6 15 63.8 Other transfers 14.6 5.8 49.0 10 10.6 All social 7.9 5 6.6 assistance 4.4 6.4 0 Scholarship Food aid All SA Other Consumption 0 20 40 60 80 transfers Q1 Q2 Q3 Q4 Q5 Source: World Bank calculations using ADePT, based on ECVMAS 2012. Source: World Bank calculations using ADePT, based on ECVMAS 2012 8.34 Design may also be revisited in some cases. While full analysis of the adequacy of all EDE PEP programs is not possible given data limitations, there are concerns over adequacy of some programs which deserve consideration. The CCT Ti Manman Cheri brings into focus attention to human capital by incentivizing mothers to keep their children in school. However, due to the budget restrictions and GoH decisions, the transfer is awarded to households only for a year. This puts into question its effect in the medium term, despite promising short-term effects that are anecdotally reported at present. In this case, the issue of adequacy is not just with respect to the amount but the duration of the program. -125- Table 8.5 : Alignment of EDE PEP Programs with Risks and Vulnerabilities Across the Life Cycle Planned Projects Life Cycle Number of Risk Under Beneficiaries Stage ÉDE PÈP (2016) 1. Early Childhood Malnutrition unaddressed n/a Mortality unaddressed n/a Poor child development unaddressed n/a 2. School age childhood Low school enrolment and drop-out PSUGO 1,500,000 School feeding 1,200,000 Ti Manman Chri 100,000 Child labor/Restavek unaddressed n/a 3. Youth Unemployment unaddressed n/a Poor educational outcomes Kore Etidyan 30,000 4. Adulthood Unemployment, lack of access to credit Ti Kredi 6,500 Female unemployment unaddressed n/a Low income, insecure livelihoods Kore Peyizan 100,000 Unemployment, low income HIMO (public works) n/a Poor living conditions, poor sanitation Ranje Kay Kartier/ 25 districts Banm Lumie- Banm Lavi Illiteracy Literacy 150,000 5. Old Age Low income Kore Ti Gran Moun 30,000 6. All Cycles Disability Kore Moun Andikape 30,000 Malnutrition and food insecurity Resto Communautaire 150,000 Natural disaster/emergency Panye Solidarité 600,000 Kantin Mobil 1,000,000 Bon Dijans n/a Disease/lack of access to health care Campaign for n/a prevention of cholera Community health n/a centers Carte Roz 2,500,000 Inadequate living conditions (lack of Ranje Kay Kartier/ n/a access to sanitation, drinking water or Banm Lumie- Banm waste management) Lavi but insufficient Violence unaddressed n/a Source: World Bank calculations based on data from FAES -126- PREVENTING A RETURN OF FUEL SUBSIDIES 89 This chapter discusses the distributional and price effects of adjusting fuel prices in Haiti. The rise in retail fuel prices in October 2014, combined with a decline in international oil prices has allowed Haiti to eliminate the fuel price subsidies that were burdening the budget. Such subsidies favor the rich, more so than subsidies on food, health or education. International oil prices are, however, volatile and without the introduction of an automatic price adjustment mechanism for petroleum products, fuel price subsidies could return to haunt the budget, but this time under tighter financing constraints. Fuel price subsidies are, however, only one dimension of broader development challenges: more generally, the overall policy framework for urban transport and renewable energies could be strengthened. Fuel Price Subsidies Implied a Heavy Burden on the Budget… 9.1 Like other net oil-importing developing countries, Haiti has faced economic difficulties with high oil prices. Haiti has been for a long time cautious passing through the increase in international oil prices to consumers. In 2010, it abandoned a mechanism of automatic pass-through, and adopted a price control system through adjustments in taxes. Taxes that should have been applied to petroleum products (e.g. excises) were foregone and, when there were no more revenues to forgo, the government actively financed the difference in prices.90 Since 2010, the year of the devastating earthquake, until October 2014, the retail prices for petroleum products changed only once, in March 2011 (Figure 9.1).91 89 Prepared by Calvin Djiofack, Julie Saty Lohie and Thiago Scot. 90 Appendix 5 presents the structure of oil products. 91 In October 2014, the government increased the domestic price of oil for the first time since March 2011, as follows: about 7.5 percent for gasoline (to 215 gourdes per gallon), 9 percent for diesel (to 177 gourdes per gallon), and 9 percent for kerosene (to 171 gourdes per gallon). In February 2015 the increases were reversed to reflect lower international oil prices, with gasoline prices falling to 195 gourdes per gallon; diesel to 157 gourdes per gallon; and kerosene to 156 gourdes per gallon. -127- Figure 9.1 : Evolution of Domestic Price and International Prices of Oil Products, 2004-14 (In Gourdes) 250 200 150 Gazoline 91 Gazoline 95 100 Diesel Kerosene 50 International Price 0 Nov-05 Nov-10 Dec-07 Juin-10 Dec-12 Aug-04 Juin-05 Oct-08 Oct-13 Oct-14 Fev-07 Fev-12 Juil-07 Juil-12 Avr-06 Aug-09 Avr-11 Jan-05 Sept-06 Mai-08 Mar-09 Jan-10 Sept-11 Mai-13 Mar-14 Sources: IMF and Government of Haiti 9.2 The size of oil subsidies in Haiti had increased dramatically over the last five years, imposing a large burden on public finances. As the gap between international and domestic prices increased, the cost of keeping retail petroleum prices fixed was placing an increasingly significant burden on public finances (nearly 2 percent of GDP in 2013). Fuel subsidies in Haiti rose from around 3.4 percent of total revenues in 2009 to 11 percent in 2011, and to 15 percent in 2013 (Figure 9.2). In 2013, the Haitian government forewent revenue and spent 88.33 Gourdes, 30 Gourdes, and 52 Gourdes per gallon to subsidize gasoline, diesel, and kerosene, respectively (Table 9.1).92 In early 2013, eliminating subsidies would have meant an effective price increase for fuels of about 30 percent.93 Figure 9.2 : Oil Subsidies Costs, 2010-13 (Billions of Gourdes and Percentage of GDP) 92 These subsidies represent about 31 percent of the total price per gallon for gasoline, 16 percent for diesel and 23 percent for kerosene. 93 The simulations presented in this section were performed before the price changes of October 2014. All the baseline prices discussed in here, therefore, refer to those fixed in the period between April 2011 and September 2014. Average international oil prices prevailing at the time were about USD 90/barrel. -128- Table 9.1 : Distribution of Oil Subsidies by Fuel Product, 2009-13 (Billion Gourdes) Fiscal Year Products 2010 2011 2012 2013 Gazoline 95 0.105 0.604 1.004 2.066 Gazoline 91 0.269 1.565 2.301 2.529 Gasoil 0.541 1.93 1.217 1.898 Kérosène 0.162 0.341 0.187 0.399 Total 1.078 4.44 4.709 6.892 Source. MEF (2014) 9.3 Fuel price subsidies in Haiti were larger, relative to income, than most countries in the region and at a similar level of development (Figures 9.3 and 9.4). While Haiti spent 1.5 percent of its GDP on oil subsides in 2011, the rest of Caribbean countries spent on average just 0.9 percent of their GDP. The Dominican Republic, which shares the island with Haiti, spent less than 0.1 percent of GDP on fuel subsidies. The amount of public resources allocated to fuel price subsidies in Haiti was also huge compared to expenditures required to enhance growth and reduce poverty (Figure 9.5). Government spending on fuel price subsidies had exceeded public spending on health and education in recent years. Figure 9.3 : Total Post-Tax Costs of Petroleum Figure 9.4 : Total Post-Tax Costs of Petroleum Subsidies, 2011 (Percentage of GDP) Subsidies, 2011 (In percent of General Revenues) 1.6 12 1.4 10 1.2 1.0 8 0.8 6 0.6 4 0.4 0.2 2 0.0 0 Dominican Low Income Caribbean World Haiti Dominican Low Income Caribbean World Haiti Republic Group Republic Group Sources: IMF (2014) and Government of Haiti Sources: IMF (2014) and Government of Haiti Figure 9.5 : Oil Subsidies and Public Spending Allocation, 2010-12 (Billions of Gourdes) 9 8 7 6 5 4 3 2 1 0 MENJS MSPP MAS MJSAC MCFDF Oil subsides 2010 2011 2012 Source: Government of Haiti -129- 9.4 Furthermore, the price differential with the Dominican Republic may have encouraged smuggling. The large difference in oil subsidies between Haiti and the Dominican Republic had created a strong incentive to smuggle petroleum products to the higher-priced destination (the Dominican Republic), increasing the budgetary burden for Haiti. The price of gasoline in the Dominican Republic was 30 percent higher, the price of diesel 32 percent higher, and the price of kerosene 24 percent higher, than in Haiti (IMF, 2014). Anecdotal evidence suggests that informal trade was taking place. … Are Regressive… 9.5 Subsidized fuels are mainly consumed by the wealthiest households. Poor households in Haiti consume very little fuel, both in absolute terms and as a share of their total budget. On average, poor households spend only 112 Gourdes a year on fuels, or 0.11 percent of their total annual budget. 94 95 In comparison, the average annual expenditure on petroleum products in the richest decile is 18,900 Gourdes, or 5.68 percent of total consumption (Figure 9.6 and Table 9.2). Part of this difference is explained by take-up rates: less than 3 percent of poor households report having spent any amount on fuel during the last 12 months, whereas almost 17 percent of households in the richest decile did. Figure 9.6 : Expenditure on Petroleum Products Per Decile, 2012 (Percentage of Budget) 6 5.68 5 4 3 2 1.46 1 0.46 0.41 0.08 0.03 0.07 0.09 0.18 0.14 0 Decile 1 Decile 2 Decile 3 Decile 4 Decile 5 Decile 6 Decile 7 Decile 8 Decile 9 Decile 10 Sources: ECVMAS (2012), World Bank and Mnistry of FInance (MEF) calculations 94 All data discussed in this section, unless noted otherwise, is from the 2012 “Enquête sur les Conditions de Vie des Ménages après le Seisme” (ECVMAS 2012). The official poverty rates for Haiti are 58.5 percent for moderate poverty and 24.4 percent for extreme poverty. 95 ECVMAS 2012 does not have a specific question on expenditure on kerosene, which is widely used in Haiti, especially by poorer and rural households. Despite not being able to quantify the expenditure on that product, we do provide a more qualitative discussion on utilization later in this chapter. -130- Table 9.2 : Average Average Yearly Yearly Expenditure Expenditure on Fuels Fuels (HTG) on(HTG) Decile 1 40.05 Decile 2 22.29 Decile 3 61.70 Decile 4 95.46 Decile 5 221.88 Decile 6 197.19 Decile 7 753.99 Decile 8 746.97 Decile 9 3,097.85 Decile 10 18,877.00 Total 3,560.20 Source: ECVMAS (2012) and World Bank and MEF calculations 9.6 As a result, universal fuel subsidies are very regressive. The poor receive only 1.6 percent of total subsidies accruing directly to households (Table 9.3 and Figure 9.7). Since direct usage of petroleum products is so limited among the poor, wealthier households benefit disproportionally from subsidies granted to these products. Over 93 percent of fuel subsidies going directly to households benefit the richest 20 percent of the population. Poor households, on the other hand, receive only 1.6 percent of total subsidies, despite comprising 58 percent of the population. Extreme poor households benefit even less, receiving a meager 0.3 percent of subsidies. 9.7 Fuel subsidies are particularly regressive when compared to subsidies on other categories of consumer expenditures, such as food, education and health. Expenditures on food represent more than 60 percent of the total budget of households in the poorest decile, but the share of food decreases steadily as expenditure level rises (Figure 9.8). Households in the richest decile spend only 40 percent of their budget on food. The share of household expenditures on health is mostly stable across the consumption distribution, whereas the share of education increases slightly for the deciles 7 – 8 before declining for richer households. Thus the poor would receive between 30 percent and 40 percent of the benefits from a universal reduction in education, health, and food prices (Figure 9.9). Table 9.3 : %Fuel Subsidies Received By Income of total subsidies received by Figure 9.7 : Fuel Subsidies Received By Group, 2012 Decile (Percentage quintile of Total) (Percentage of Total) Decile 1 0.1% Decile 2 0.0% Poor Decile 3 0.1% 2% Decile 4 0.2% Decile 5 0.6% Decile 6 0.5% Decile 7 2.1% Decile 8 2.3% Decile 9 10.8% Decile 10 83.2% Total 100.0% Non-Poor 98% Sources: ECVMAS (2012), World Bank and Ministry of Finance (MEF) calculations -131- Figure 9.8 : Composition of Expenditure by Groups, by Decile, 2012 (Percentage of Total Budget) 40.1 Decile 10 5.4 5.7 1.9 53.0 Decile 9 7.2 1.5 2.0 54.6 Decile 8 8.5 0.4 1.2 56.7 Decile 7 8.7 0.5 1.2 58.2 Decile 6 7.4 0.1 1.5 61.6 Decile 5 7.7 0.2 1.2 61.9 Decile 4 7.6 0.1 1.9 63.1 Decile 3 7.5 0.1 1.9 64.6 Decile 2 7.7 0.0 1.6 64.9 Decile 1 7.0 0.1 1.6 0 10 20 30 40 50 60 70 Food Education Fuels Health Sources: ECVMAS (2012) and World Bank calculations Figure 9.9 : Expenditures by Groups, 2012 (Percentage of Total) 100 90 80 70 67.0 70.8 60 98.4 63.6 50 40 30 20 36.4 33.0 29.2 10 1.6 0 Fuels Education Health Food Poor Non-Poor Sources: ECVMAS (2012) and World Bank 9.8 An increase in fuel prices could also have indirect effects on households. Directly, households are affected through their own consumption of gasoline or kerosene. Indirectly, they are also affected since petroleum products are used as intermediary products in many sectors and their higher price will feed into the price of the final good produced by these sectors. To capture both these effects, the interdependence of sectors needs to be taken into account and a Social Accounting Matrix (SAM) multiplier analysis has been used for this purpose. A SAM is an input- output table, representing the transactions being carried out between sectors for a given year. It describes all the income received and expenditures made by households and various sectors, thus -132- capturing the linkages within the economy. Intuitively, our model aims at estimating the loss of consumers’ purchasing power in real terms given a fixed budget and higher fuel prices. 9.9 As discussed previously, the direct effect of a rise in fuels prices affects almost exclusively the richest two deciles.96 Since fuels are a very small part of poor households’ consumption basket, the direct effect of a price increase is negligible (Figure 9.10). A 30 percent increase in fuel prices was assumed to assess the income distribution effects of this reform. Whereas the direct effect of eliminating subsidies would reduce the real consumption of the average poor household by less than 0.05 percent, households in the richest decile would lose 1.65 percent. On average, a Haitian household can be expected to suffer a 0.62 percent drop in consumption due to the direct effect of higher fuel prices. Figure 9.10: Welfare Loss by Direct Effect (Percentage) 0.0 -0.2 -0.02 -0.01 -0.02 -0.02 -0.05 -0.04 -0.13 -0.12 -0.4 -0.6 -0.42 -0.8 -1.0 -1.2 -1.4 -1.6 -1.65 -1.8 Bottom Decile 2 Decile 3 Decile 4 Decile 5 Decile 6 Decile 7 Decile 8 Decile 9 Top Decile Decile Sources: ECVMAS (2012), World Bank and Ministry of FInance (MEF) calculations 9.10 Increases in fuel prices have large effects on electricity and transportation, however. The analysis of the indirect effects based the input-output matrix estimates the effect of increasing fuel prices on each sector of the economy. An increase in fuel prices would have the largest direct impact on transportation (Box 9.1) and electricity, in which more than 30 percent of total input comes from oil products (Figures 9.11 and 9.12). Nonetheless, sectors that do not use fuel as a direct input can also be affected through input-output chains: even though the production of food is not intensive in fuel, its price increases more than one percent after a 30 percent rise in fuel prices because of the pass-through from agricultural and transportation prices. The overall effect on the price level would nevertheless be contained: a 30 percent rise in fuel prices would lead to a one-time adjustment of about 3 percent in the overall price level (Figure 9.13). The main contributions to this increase would come from urban and interurban transportation (0.6 percentage points each), food (0.5 percentage points) and fuel (0.5 percentage points). 96 The simulations presented in this section were performed mostly using SUBSIM, The Subsidy Simulation stata package. For reference see Araar and Verme (2012). -133- Figure 9.11: Intensity In Oil Products By Sector Figure 9.12: Effect of Prices from the IO Table (Percentage of Total Oil Usage) (Percentage Increase) 40 35.0 13.3 33.5 14 35 12 11.3 30 25 10 20 8 15 6 10 5.0 4.0 4 2.6 5 2.1 1.9 1.7 1.1 0.9 0.8 1.5 2 1.1 0 0 Sources: World Bank and Ministry of Finance (MEF) calculations Sources: World Bank and Mnistry of Finance (MEF) calculations Figure 9.13: Contribution to Inflation (Percentage Points) 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Sources: World Bank and Ministry of Finance (MEF) calculations -134- Box 9.1 : The Transport Sector in Haiti Urban transportation in Haiti is dominated by private light vehicles and the “tap -taps”. As of 2005, the “tap-taps”, small buses or pickup trucks providing collective transportation, comprised almost 40 percent of the vehicle fleet in Haiti.97 Two-thirds of those used gasoline as their main fuel, but this number could be significantly lower nowadays, as differential taxation between gasoline and diesel has encouraged the use of the latter. Fuel costs are approximately half of total transportation costs, and changes are usually passed immediately to consumers, especially for price increases. Although transport prices are administered in Haiti, simulations using a Vehicle Operating Cost model (VOC) and assuming full pass-through suggest that a 10 percent increase in fuel prices would increase total transportation costs by 4 – 5 percent (4.7 percent for tap-taps).98 The behavior of transporters in relation to prices seems, however, to be asymmetric: 61 percent of transporters surveyed in 2014 declared that they would raise tariffs if fuel prices increased, whereas only 43 percent said that they would pass price decreases to consumers (World Bank, 2014). Price increases of transportation affect mainly the richest households, but would perpetuate the lack of access of the poor to services, and may affect competitiveness. Expenditures on collective transportation are, on average, less than 2 percent of total budget for poor households, partly because many do not use the transport system. In that sense, increases in transport prices reinforce the exclusion of poorer households from transportation services and may hinder their access to economic opportunities. Tap-Tap Cost Expenditure on transport, by quintile Structure (percentage of total expenditure) 6 Interest 26% 5 4 Fuel 49% 3 Depreciation 0% 2 Crew 13% 1 Maintenance 0 Labor Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 2% Maintenance Parts Tire Lubricants Long transport Daily transport Total 8% 1% 1% Source: World Bank (2014) Source: World Bank staff calculations based on the ECVMAS 2012 There is considerable potential to improve energy efficiency and better manage fuel prices over the long term. Developing an efficient fuel stocking strategy is an important step towards reducing price volatility once subsidies are phased out. Guaranteeing lower prices also involves reducing the costs of internal fuel transportation. On the efficiency front, several measures could help reduce fuel waste: encouraging the renewal of the vehicle fleet, improving the conditions of roads and, in a more general sense, reducing transport needs by better urban planning. In the short term, targeted subsidies can be applied to compensate both transporters and consumers. International experience offers examples of targeted policies that reduce the costs of transportation for specific vulnerable groups: the “vale-transporte” for formal workers in Brazil; the free cable cars in Medellin and Rio de Janeiro; and other subsidies for students and the unemployed. Improving access to financing and reducing insurance costs could lower the costs faced by transporters and, therefore, final prices for consumers. 97 Enquete Transport, IHSI (2005) 98 The Vehicle Operating Costs Model (VOC) is a stand alone program developed by the World Bank that estimates vehicle operating costs (See technical note on VOC Available at http://www.worldbank.org/transport/publicat/wbtp-234.pdf). -135- 9.11 Poor households are especially vulnerable to increases in food prices. The modest (1.05 percent) rise in food prices nevertheless imposes a sizeable loss of welfare on the poorest households, due to the large share of food in their budgets (Figure 9.14). The lowest deciles are hit by a 0.7 percent decrease in real consumption due to food price increases alone, close to half of the total losses they suffer in the simulation. As the level of income increases, higher food prices loose in importance as a source of welfare loss, and increases in electricity and transportation prices become more important,. Figure 9.14: Welfare Loss by Decile (Percent) Decile 1 Decile 2 Decile 3 Decile 4 Decile 5 Decile 6 Decile 7 Decile 8 Decile 9 Decile 10 0.0 -0.5 -1.0 -1.5 -2.0 -2.5 -3.0 -3.5 -4.0 -4.5 Fuel Food Transportation Long transportation Electricity Other indirect effects Sources: ECVMAS (2012), World Bank and Ministry of Finance (MEF) calculations 9.12 Furthermore, price increases of kerosene would mostly affect poorer, rural households. Even though there are no data on expenditures on kerosene, there is some indication that poor households in rural areas would be the most affected by increases in the prices of this product.99 Over 50 percent of households use kerosene as the main source of lighting, since electricity is available (with or without meters) for only about 30 percent of households (Figures 9.15 and 9.16). In rural areas, however, the use of kerosene is much more prevalent. Among the rural poor, over 80 percent of households use kerosene as main source of fuel. In urban settings, reliance on kerosene is much lower, with the exception of the extreme poor (bottom 25 percent). Thus, an increase in the price of kerosene would be mostly felt by the most vulnerable segments of the population. 99 Even though ECVMAS 2012 provides information about expenditure on biomass (charcoal, wood…), there is no survey question on kerosene expenditure. The survey does provide, however, information on the main source of lighting for the household. Since the question is exclusively for the main source used, the data discussed here can be read as a lower bound for the number of households that consume kerosene. ECVMAS 2012 does not provide details on different types of fuels consumed. Instead, the only question on fuel consumption refers to “Essence” , which can be interpreted as any fuels used in vehicles. Therefore, the scenario of a 30 percent increase in the fuel price refers to a consumption-weighted average of increasing gasoline by 44 percent and diesel by 18 percent. -136- Figure 9.15: Main Source of Lighting, 2012 Figure 9.16: Households Using Kerosene As Main (percent Households) Source of Lighting, 2012 (Percent) Sources: ECVMAS (2012), World Bank and MEF calculations 9.13 To sum up, the total effect of a fuel price increase falls mainly on non-poor, urban households. Assuming a 30 percent in increase in prices of petroleum products, the consumption of non-poor urban households would decline by more than 3 percent. Poor households face real losses between 1.5 – 2 percent, whereas households in the richest decile face decreases of over 4 percent in their real consumption (Figure 9.17). The average loss in consumption is higher for urban households (more than 3 percent) than for rural households (2.5 percent), because urban households spend more on fuel and transportation than rural households do (Figure 9.18). Figure 9.17: Total Loss of Welfare by Decile Figure 9.18: Welfare Losses by Urban/Rural Areas (Percent) (In Percent) 0.0 -0.5 -1.0 -1.5 -2.0 -2.5 -3.0 -3.5 Urban Rural Indirect effects Direct effects Sources: ECVMAS (2012), World Bank and Ministry of Finance (MEF) calculations Sources: ECVMAS (2012), World Bank and MEF calculations … And Could Return. 9.14 The budget remains vulnerable to a rebound in international oil prices.100 In October 2014, the authorities decided to increase petroleum prices, while international oil prices themselves were starting to decline, eliminating the need for oil subsidies. For a given price at the pump, simple simulations illustrate, however, how changes in international prices could quickly make fuel price subsidies return. The subsidies, which amounted to around 2 percent of GDP in 2013, would climb to over 3 percent of GDP with international prices at USD 100/barrel, but drop 100 ECVMAS 2012 does not provide details on different types of fuels consumed. Instead, the only question on fuel consumption refers to “Essence”, which can be interpreted as any fuels used in vehicles. Therefore, the scenario of a 30 percent increase in the fuel price refers to a consumption-weighted average of increasing gasoline by 44 percent and diesel by 18 percent. -137- to zero when prices hit USD 75/barrel. For oil prices lower than that threshold, current prices at the pump actually provides additional revenues to the government. Table 9.4 : Subsidies (Million Gourdes) Given Different Scenarios for Oil Prices Products / International Total (% Gas Gasoil Kerosene Total Oil prices (USD/barrel) GDP) 40 -4,898 -8,931 -1,848 -15,677 -4.3% 50 -3,000 -6,765 -1,426 -11,191 -3.1% 60 -1,101 -4,599 -1,005 -6,705 -1.8% 70 797 -2,433 -583 -2,219 -0.6% 80 2,696 -267 -161 2,267 0.6% 90 4,594 1,898 261 6,753 1.9% 100 6,493 4,064 682 11,239 3.1% Note: The simulation considers the relation between average international oil prices and import prices of oil products in 2012/2013, and performs a linear extrapolation for different scenarios. 9.15 There is thus a need to return to an automatic pass-through. Lower international oil prices have led people to the streets, demanding lower retail petroleum prices. The authorities in early 2015 have twice reduced retail prices. In the past, an automatic price adjustment mechanism had been in place, but was abandoned in March 2010. Under this arrangement, any movement in international oil prices that would imply a variation in retail fuel prices greater than 5 percent was fully passed through. As a result, retail fuel prices had fluctuated more or less in line with international oil prices. Consideration should be given to resume such a mechanism. 9.16 With the uncertainty surrounding the future course of international oil prices, reducing or even eliminating the influence of politics in the setting of fuel prices would also be critical. The implementation of the new automatic price adjustment mechanism should be given to an independent agency in order to ensure, as much as possible, that politics do not interfere with needed price adjustments. Such an agency could also be in charge of regulating the oil sector and could include representatives from the ministries involved, importers, distributors, and possibly civil society. 9.17 More generally, the overall policy framework for urban transport and renewable energies could be strengthened. Fuel subsidies are only one dimension of broader development challenges. Thinking about a subsidy mechanism to contain transport price increases is an opportunity to discuss more broadly about the need to structure better the urban transport sector with the aim of improving the security of vehicles and the skills of drivers. Similarly, discussions should be carried out on reducing the reliance of Haiti’s economy on oil through greater efficiency in the use of energy and the development of renewable sources of energy (such as solar or wind). -138- Concluding Observations 1. Human development indicators have improved in Haiti over the past decade. Haiti is the poorest country in the Latin America and Caribbean region, and among the poorest in the world. The country has long been characterized by its very low fiscal revenue mobilization, seriously constraining its ability to carry out needed developmental spending (infrastructure, health, education). Against this backdrop, greater donor assistance and access to concessional financing has allowed Haiti to increase priority spending over the past decade. The fragmented fiscal data available would suggest that these additional resources have been channeled to the reconstruction and social sectors, consistent with the decline in poverty and improvements in human development indicators (such as in education) observed over the same period. 2. Despite this progress, challenges remain. Haiti’s economic system is not very inclusive. Finance remains, one of the main obstacles for Haitians to access basic services in health or education. When accessed, the service provided is often not of the quality needed: problems of efficiencies plague the sectors with health professionals giving few visits and school teachers not using their time efficiently. Social safety nets programs and education subsidies have benefited from greater financing, but the programs remain fragmented and ill-targeted. In social protection, the low coverage is staggering given the high levels of chronic poverty, especially in rural areas. 3. Haiti needs to deal now with tighter budget constraints. The decline in aid and in concessional financing is reducing the country’s fiscal space and putting the sustainability of its higher level of public spending into question. This is all the more worrisome as social sectors are heavily depended on these resources for their financing. Within this context, the government faces an urgent need to use resources more efficiently and equitably in order to have a greater development impact while keeping the fiscal accounts on a sustainable path. This study aimed at contributing building an evidence base to inform decision making. 4. Against this background, some options going forward present themselves . While the report wanted to provide mainly a diagnostic and a better understanding of some critical sectors such as health and education, some broad options going forward can nevertheless be suggested as basis for discussion:  Data: Overall tighter financing constraints make it even more urgent to collect timely data to track public spending appropriately and make it more efficient in different sectors. Differences in budget classifications and the absence of an effective Single Treasury Account lead to a fragmentation of fiscal data and prevents a comprehensive monitoring of public spending from budget appropriation to payment. Fiscal reporting needs thus to be strengthened to allow an appropriate tracking of public spending and set in place the minimum conditions for transparency and accountability in the use of public resources. Completing the roll-out of the Single Treasury Account and unifying the budget classifications among the various sources of financing would be important steps in this direction. -139-  Donors: With declining donor assistance, greater donor coordination is needed to make the most of these resources. A master plan for critical sectors translating the priorities of these sectors into areas of intervention could federate the actions of the donor community, allowing it to coordinate its technical and financial contributions, and avoid potential duplication and inefficiencies. Such a plan would have indicators and deadlines, triennial operational plans, and an annual action plans to structure the monitoring of its implementation.  Budget composition: The composition of public spending has room for improvement to enable better service delivery. The growing wage bill, although in line with economies at similar levels of development, has been the driving force of total current spending, but most of it is accounted for by staff not directly involved in service delivery. Sharp increases in capital outlays have not contributed to the acceleration of growth, due in part to an inadequate resource allocation for operating and maintenance expenses (e.g., equipment, utilities), thereby limiting growth dividends of investment.  Fiscal revenue: Balancing fiscal sustainability and development needs will remain a daunting challenge for Haiti as long as the country does not mobilize substantially more own fiscal resources. In this regard, a review would be needed of existing exemption regimes for all tax instruments with priority to customs duties and corporate income tax where revenue losses are the greatest. An important step would be the introduction of a clear classification system in customs data to track the justification of exemptions. Where tax exemptions or other investment incentives are used, a clear framework needs to be put in place to measure and track their costs. At the same time, it is important to track and monitor the benefits these incentives were expected to deliver (such as employment creation or technology transfer).  Greater growth dividend: Haiti’s development needs are daunting and will need to be addressed. In this regard, the overall public investment management needs to be improved through a strengthening of project evaluation, selection, programming, execution, and control. As a first step, the priority should be put on implementing basic elements of formal project selection and appraisal, execution and controls such as the need for investment projects to (i) advance only if they are mature enough (i.e. developed in a concept or in a justification); (ii) be monitored (including physically); (iii) use resources adequately; (iv) are controlled ex-post as required.  Improve the performance of health facilities: For a long time the focus has been put on increasing the resources or the inputs. With shrinking budgets, the time may have come to shift the logic upside down and pay more attention to the results or the services actually delivered by the health facilities through a more general use of results-based financing (RBF). RBF could provide monetary and non-monetary incentives to health workers in order to increase their productivity, discouraging them from exercising another occupation. Health facilities may also have very low productivity because they have too few patients. To improve the use of their resources, as well as health outcomes, such facilities may want to reinforce primary health level services and operate mobile clinics to increase access to health services, particularly for the poor. -140-  More accountable schools to increase learning in the classroom: The majority of schools at all levels in Haiti are non-public, and operate with little oversight or accountability. With limited public resources, the sector will remain predominantly non-public for the foreseeable future and efforts should focus on making the schools more accountable for the quality of service they deliver. In this regard, efforts could focus on the comprehensive identification of schools or students benefiting from the different programs offered by the Ministry of Education, including the donor-financed Tuition Waiver Program (TWP) and PSUGO and linking this support with school achievements. Greater oversight could also be called for through a mandatory teaching license based on demonstrated competencies and a mandatory school identity card, leading to certification. In addition, targeted in-service training programs for teachers could be put in place to improve instruction in the classroom and learning.  Better targeting in social protection: Haiti’s social protection suffers from fragmentation of programs and interventions. The tighter resource constraints related to a reduced access to concessional borrowing and donor assistance provide an opportunity to design a less fragmented and better targeted system of social safety nets. In the meantime, continuing progress in the implementation of the social registry of potentially eligible beneficiaries is required.  Protecting the budget: With fiscal resources remaining limited, unbounded calls on the budget should be contained. In this regard, the re-emergence of fuel price subsidies remain a risk. Higher international oil prices could come back again to haunt the budget. There is thus a need to return to an automatic pass-through with this mechanism given to an independent agency to reduce political influence in the setting of fuel prices. -141- References Akazili J., M. Adjuik, S. Chatio, D. Kanyomse, A. Hodgson, M. Aikins, and J. Gyapong (2008), “What are the technical and allocative efficiencies of public health centers in Ghana?”, Ghana Medical Journal 42, no. 4, pp 149-155. Andrews, C., M. Das, J. Elder, M. Ovadiya and G. Zampaglione (2012), “Social Protection in Low Income Countries and Fragile Situations”, Social Protection and Labor Discussion Paper No. 1209, The World Bank, Washington, D.C. Araar, A. and P. Verme (2012), "Reforming subsidies: a tool-kit for policy simulations", Guidebook and SUBSIM 1.0 manual, Policy Research Working Papers No. 6148, The World Bank, Washington D.C. Armington, P. (1969), “A Theory of Demand for Products Distinguished by Place of Production”, International Monetary Fund Staff Papers XVI, pp 159-78. Backiny-Yetna, P., N. Garbiras and F. Marzo (2013), “Poverty in Haiti: Methodological Note on the aggregate of consumption using ECVMAS 2012”, mimeo. Baez, J. and I. Santos (2007), “Children’s Vulnerability to Weather Shocks: A Natural Disaster as a Natural Experiment”, Working Paper, The World Bank, Washington D.C. Barrientos, A. and D. Hulme (2008), “Social Protection for the Poor and Poorest in Developing Countries: Reflections on a Quiet Revolution”, Brooks World Poverty Institute Working Paper, No. 30, Manchester, UK. Beeston, K. (2010), “Starting from Scratch: Building a Brighter Future for Haiti’s Disabled Children”, June 14, 2010, The Guardian, www.theguardian.com. Bellos, A., K. Mulholland, K. L. O’Brien, S. A. Qazi, M. Gayer, and F. Checchi (2010), “The burden of acute respiratory infections in crisis-affected populations: a systematic review.” Conflict and Health 4, no. 3, doi:10.1186/1752-1505-4-3. Bird, R., and P.P. Gendron (2011), The VAT in Developing and Transitional Countries, vol.1, Cambridge University Press. Borgarello, A. (2009), “Haiti Social Safety Net Assessment”, World Bank Social Protection Non- lending technical Assistance Program, Mimeo. Bourguignon, F. (2003), “The growth elasticity of poverty reduction: Explaining heterogeneity across countries and time periods”, Inequality and Growth: Theory and Policy Implications, edited by Eicher, T. S. and S. J. Turnovsky, pp 3-26, Cambridge, Massachusetts. Brenton, P., C. Saborowski, C. Staritz and E. von Uexkull (2011), “Assessing the Adjustment Implications of Trade Policy Changes Using the Tariff Reform Impact Simulation Tool (TRIST)”, World Trade Review 10, no. 02, pp 249-276. Bruhn, M. and J. Loeprick (2014), “Small Business Tax Policy, Informality, and Tax Evasion - Evidence from Georgia”, World Bank Policy Research Working Paper 7010, The World Bank, Washington, D.C. -142- Buss, T. F (2013), “Foreign Aid and the Failure of State Building in Haiti under the Duvaliers, Aristide, Préval, and Martelly”, WIDER Working Paper No. 2013/104. Carpenter, S., R. Slater and R. Mallett, Richard (2012), “Social Protection and Basic Services in Fragile and Conflict - Affected Situations” Working Paper 8, Secure Livelihoods Research Consortium. Cayemittes, M., M. Fatuma Busangu, J. Bizimana, B. Barrère, B. Sévère, V Cayemittes and E. Charles (2013), Enquête Mortalité, Morbidité et Utilisation des Services, Haïti 2012, MSPP, IHE and ICF International, Calverton, Maryland, United States. Cellule de Planification et de Statistique (CPS/SSDSPF), Institut National de la Statistique (INSTAT/MPATP), INFO-STAT and ICF International (2014), Enquête Démographique et de Santé au Mali 2012-2013, CPS, INSTAT, INFO-STAT and ICF International, Rockville, Maryland, United States. Christiaensen, L., C. P. Punam, and A. Sanoh (2013), “Africa’s Growth, Poverty and Inequality Nexus - Fostering Shared Prosperity”, Mimeo, The World Bank, Washington, D.C https://editorialexpress.com/cgibin/conference/download.cgi?db_name=CSAE2014&pap er_id=381. Coa R., L.H. Ochoa and MEASURE DHS (2009) Encuesta Nacional de Demografía y Salud, Bolivia ENDSA 2008, Macro International Inc. Calverton, Maryland, United States. Coady, D., M. El-Said, R. Gillingham, K. Kpodar, P. Medas and D. Newhouse (2006), “The Magnitude and Distribution of Fuel Subsidies: Evidence from Bolivia, Ghana, Jordan, Mali, and Sri Lanka”, IMF Working Paper 06/247. Coelli, T.J. (1996), A guide to DEAP: version 2.1: a Data Envelopment Analysis (Computer) Program. Coordination Nationale De La Sécurité Alimentaire (CNSA) (2011), Enquête Nationale de la Sécurité Alimentaire (ENSA). Cros, M. and W. Zeng (2014), “Measuring Efficiency of Primary Health Care Facilities in Haiti using Data Envelopment Analysis (DEA)”, Accepted for a poster presentation at the 3rd Global Symposium on Health Systems Research in Cape Town, 30 Sept. - 3 Oct. 2014. Demombynes, G., H. Peter and L. Gianmarco (2010), “Students and the Market for Schools in Haiti.” Policy Research Working Paper 5503, The World Bank, Washington, D.C. Denizer, C., D. Kaufmann and A. Kraay (2011), “Good Countries or Good Projects? Macro and Micro Correlates of World Bank Project Performance”, Policy Research Working Paper 5646, The World Bank, Washington, D.C. Devereux, S, (2000) Social Safety Nets for Poverty Alleviation in Southern Africa, A Research report for DFID, ESCOR Report R7017. Di John, J. (2010), “Development as State-making - Taxation, Resource mobilization and state performance”, UKaid Crisis States Research Centre Working Paper no. 84. Dollar, D., and A. Kraay (2002), “Growth is Good for the Poor”, Journal of Economic Growth 7, no. 3, pp 195-225. -143- Dollar, D., T. Kleineberg, and A. Kraay. (2013), “Growth Still Is Good for the Poor”, World Bank Policy Research Working Paper 6568. Duryea, S., S. Galiani, H. Nopo, and C. Piras (2007), “The Educational Gender Gap in Latin America and the Caribbean”, Working Paper No. 502, The Inter-American Development Bank, Washington, D.C. Easterly, W. and M. Woolcock (2006), “Social Cohesion, Institutions, and Growth.” Working Papers 94, Center for Global Development, Washington, D.C. Eaton, J. and S. Kortum (2001), “Trade in Capital Goods”, European Economic Review 45, no. 7, pp 1195-1235. Echevin, D. (2011), “Vulnerability and Livelihoods Before and After the Haiti Earthquake”, Policy Research Working Papers, The World Bank, Washington, D.C. Echevin, D. (2011), “Characterizing Poverty and Vulnerability in Rural Haiti: A Multilevel Decomposition Approach”, MPRA Paper 35659, http://mpra.ub.uni-muenchen.de/35659/. Echevin, D. (2011), “Vulnerability to Asset Poverty in Sub-Saharan Africa”, MPRA No. 35660. http://mpra.ub.uni-muenchen.de/35660/. ECLAC (2013), Social Panorama of Latin America 2012, United Nations: Santiago, Chile. FAES (2013), “EDE-PEP, Programme nationale d’assistance Sociale: 1 an de mise en œuvre”, Rapport 2012-2013. FAES (2014), “EDE-PEP, Programme nationale d’assistance Sociale: 1 an de mise en œuvre – 2 ans d’existence ”, Rapport 2013-2014. FIAS (2010), “How to Reform Business Licenses”, World Bank Foreign Investment Advisory Service, Washington D.C. Filmer, D. (2004), “If You Build It, Will They Come? School Availability and School Enrollment in 21 Poor Countries”, Policy Research Working Paper 3340, The World Bank, Washington, D.C. Filmer, D. (2008), “Disability, Poverty, and Schooling in Developing Countries: Results from 14 Household surveys” The World Bank Economic Review 22, no. 1, pp 141-163. Ghana Statistical Service (GSS), Ghana Health Service (GHS), and ICF Macro (2009), Ghana Demographic and Health Survey 2008, Accra, Ghana. Government of the Islamic Republic of Afghanistan, Ministry of Public Health (2014), Aid Coordination and SWAP Manual. ______ (2010), Afghanistan National Health Accounts 2008-2009. ______ (2003), A Basic Package of Health Services for Afghanistan, Transitional Government. Government of Cambodia, Ministry of Health (2009), Cost and revenue comparisons for a sample of contracted and non-contracted health facilities, USAID, BASICS. -144- Government of the Republic of Ghana (2011), “Joint Review of Public Expenditures and Financial Management”, Supported by the World Bank, the Reconstruction Credit Institute (KfW) and DANIDA. Government of the Republic of Guyana, Ministry of Health (MOH), Bureau of Statistics (BOS), and ICF Macro (2010), Guyana Demographic and Health Survey 2009, MOH, BOS, and ICF Macro Georgetown, Guyana. Government of the Republic of Haiti (2014), “PSUGO – Bulletin Verité No. 1”, Primature de la République d’Haïti, Port au Prince, Haiti. ______ (2014a), “Penser et Lutter vers une Haïti sans pauvreté : plan d’action pour l’accélération de la réduction de la pauvreté extrême”, http://www.mpce.gouv.ht/sites/default/files/parpsynthese.pdf. ______ (2014b), “Luttons pour une Haïti sans pauvreté: plan d’action pour la réduction de la pauvreté extrême”. ______ (2013), “EDE PEP: Stratégie Nationale d’Assistance Sociale”, http://www.edepep.gouv.ht/4.03.2013.strategieEDEPEP_communication.pdf . ______ (2001), Nomenclature des dépenses budgétaires de l’État et guide d’utilisation. Government of the Republic of Haiti, Directorate-General for Budget (2014), Budget de fonctionnement du Ministère de la Santé Publique et de la Population de 2003-2004 à 2013- 2014. Government of the Republic of Haiti, Ministry of Planning and External Cooperation (MPCE) (2012), Plan Stratégique de développement d’Haïti. Pays Émergent en 2030. ______ (2013), Grandes Réalisations, 2012-2013. Government of the Republic of Haiti, Ministry of Public Health and Population (2014), Rapport des Comptes Nationaux de Santé 2011-12. ______ (2013), Rapport des Comptes Nationaux de Santé 2010-11. ______ (2013), Plan de Santé et de Développement d’Haïti 2012-2022. ______ (2013), Plan Directeur de la Santé 2012-2022. ______ (2012), Politique Nationale de la Santé. ______ (2011), Rapport de la Carte Sanitaire du Pays, Programming and Evaluation Unit. Government of the Republic of Haiti, Ministry of Economy and Finance (2012), État d’exécution des dépenses budgétaires par institution et par secteur, exécutées à partir des fonds du Trésor Public d’octobre 2011 à septembre 2012. Government of the Republic of Haiti and Inter-American Development Bank (2013), Rapport d’Evaluation du Système de Passation et Gestion des Marchés Publics d’Haïti. -145- Government of the Republic of Honduras, Secretaría de Salud, Instituto Nacional de Estadística (INE), ICF International (2013), Encuesta Nacional de Salud y Demografía 2011-2012, SS, INE and ICF International, Tegucigalpa, Honduras. Government of the Republic of the Ivory Coast, Ministère de la Santé et de l’Hygiène Publique. Comptes Nationaux de la Santé, 2007-8, 2010. Government of the Republic of Kenya, Ministry of Medical Services, Ministry of Public Health and Sanitation (2010), Kenya National Health Accounts 2009-10. Government of the Republic of Rwanda, Ministry of Health (2010), Rwanda Community-Based Health Insurance Policy, Ministry of Health, Kigali . Government of the Republic of Senegal, Agence Nationale de la Statistique et de la Démographie (ANSD), and ICF International (2012), Enquête Démographique et de Santé Continue (EDS- Continue 2012-2013), ANSD and ICF International, Calverton, Maryland, United States. Government of the United Republic of Tanzania, Ministry of Health and Social Welfare (2012), National Health Accounts 2010. ______ (2012), Public Expenditure Review, Prepared by Ministry of Finance and Development Partners PER Macro Group. Grandcolas, C. (2005), “The Occasional Failure in VAT Implementation: Lessons for the Pacific”, Asia-Pacific Tax Bulletin. Grosh, M., C. del Ninno, E. Tesliuc and A. Ouerghi (2008), For Protection and Promotion: The Design and Implementation of Effective Safety Nets, The World Bank, Washington, D.C. Grosh, M. (2004), “Social Protection Treatment in Public Expenditure Reviews”, Guidance note, The World Bank, Washington, D.C. Hanushek, E. and L. Woessmann (2009), “Do Better Schools Lead to More Growth? Cognitive Skills, Economic Outcomes, and Causation”, National Bureau of Economic Research (NBER), Working Paper No. 14633, NBER, Cambridge, MA. Harvey, P., R. Holmes, R. Slater and E. Martin (2007), “Social Protection in Fragile States”, Working Paper, Overseas Development Institute, London, UK. Health Systems 20/20 Project. Democratic Republic of the Congo (2011), National Health Accounts 2008-2009, Health Systems 20/20 project, Abt Associates Inc, Bethesda, MD. Hernandez, A. R. and M. S. Sebastian (2014), “Assessing the technical efficiency of health posts in rural Guatemala: a data envelopment analysis”, Global Health Action, no. 7. Hidalgo, S. and M. Theodate (2011), “Inter-Agency Real-Time Evaluation of the Humanitarian Response to the Earthquake in Haiti”, Inter-Agency Standing Committee (IASC), Report. http://www.alnap.org/pool/files/1395.pdf . Hollingsworth, B., P. J. Dawson, and N. Maniadakis (1999), “Efficiency measurement of health care: a review of nonparametric methods and applications.” Health Care Management Science, no. 2, pp 162-172. -146- Human Development Index (HDI). “Value.” New York: United Nations Development Programme, https://data.undp.org/dataset/Human-Development-Index-HDI-value/8ruz-shxu. Independent Evaluation Group (forthcoming), World Bank Group Assistance to Fragile and Conflict- Affected States. An IEG Evaluation, The World Bank, Washington, D.C. Institut Haïtien de Statistique et d`Informatique (IHSI) (2012), Population Totale, Population de 18 Ans et Plus Ménages et Densités Estimes en 2012, Ministry of Economy and Finance, Port au Prince, Haiti. ______ (2007), Projections de Population Totale, Urbaine, Rurale et Economiquement Active, Ministry of Economy and Finance, Port au Prince, Haiti. ______ (2005), "Enquête sur le secteur transport". ______ (2003), Enquête Sur les Conditions de Vie des Ménages (ECVH), Ministry of Economy and Finance, Port au Prince, Haiti. Institut Haïtien de l’Enfance (IHE) and ICF International (2014), Evaluation de la Prestation des Services de Soins de Santé (EPSSS) Haïti 2013, IHE and ICF International, Rockville, Maryland, United States. Institut National de la Statistique et de l’Analyse Economique (INSAE) and ICF International (2013), Enquête Démographique et de Santé du Bénin 2011-2012, INSAE and ICF International, Calverton, Maryland, United States. Institut National de la Statistique et de la Démographie (INSD) and ICF International (2012), Enquête Démographique et de Santé et à Indicateurs Multiples du Burkina Faso, 2010, INSD and ICF International, Calverton, Maryland, United States. Instituto Nacional de Estadísticas y Censos (INEC) and Ministerio de Salud (MINSA) (2002), Encuesta Nacional de Salud y Demografía 2001, Nicaragua and Macro International Inc., Calverton, Maryland, United States. Institut de Statistiques et d’Etudes Economiques du Burundi (ISTEEBU), Ministère de la Santé Publique et de la Lutte contre le Sida [Burundi] (MSPLS) and ICF International (2012), Enquête Démographique et de Santé Burundi 2010, ISTEEBU, MSPLS, and ICF International, Bujumbura, Burundi. International Finance Corporation (2011), “Economic Integrated Zone, Market Analysis”, Washington, D.C. International Labor Organisation (2010), “La Promotion du Travail Décent dans la Reconstruction et le Développement d'Haïti après le Tremblement de Terre de 2010”, Geneva, Switzerland. International Monetary Fund (IMF) (2014), World Economic Outlook Database, April 2014. ______ (2014), "Mécanisme de Tarification Automatique des Produits Pétroliers”. ______ (2013), Optimizing Fiscal Policy for High and Inclusive Growth in Haiti, IMF Country Report 13/91. -147- ______ (2011), “Investing in Public Investment: An Index of Public Investment Efficiency”, IMF Working Paper IMF, Washington, D.C. ______ (2008), “Haiti: Interim Poverty Reduction Strategy Paper”, IMF Country Report 08/115, IMF, Washington, D.C. ______ (2006), “Haiti: Interim Poverty Reduction Strategy Paper”, IMF Country Report 06/411, IMF, Washington, D.C. International Tax Dialogue (2005), “The Value Added Tax – Experience and Issues”, Background Paper Prepared for the International Tax Dialogue Conference on the VAT in Rome, March 15–16 2005. INTOSAI (2003), Assessment of Conditions Encouraging Corruption, Vienna workshop, Bibliography. .James, S. (2013), Tax and Non-Tax Incentives and Investments: Evidence and Policy Implications, World Bank Investment Climate Advisory Services. Josephson, E. and M. Bode (2013), Fiscal Space Analysis for Haiti, The World Bank, Washington D.C. Kauffmann, D. (2010), “Review of Cash-Transfer Coordination in Haiti Following the Earthquake of January 2010”, Cash Learning Partnership, http://www.cashlearning.org/downloads/GroupURDpercent20- percent20Cashpercent20coordinationpercent20-percent20Haiti.pdf. Kauffmann, D. and O. Collins (2012), “Comparative Study of Emergency Cash Coordination Mechanisms”, Cash Learning Partnership, http://reliefweb.int/sites/reliefweb.int/files/resources/120618percent20Groupepercent20URD_Cas hpercent20coordination_comparativepercent20study_Final.pdf. Kenya National Bureau of Statistics (KNBS) and ICF Macro (2010), Kenya Demographic and Health Survey 2008-09, Calverton, Maryland. Khwaja, M.S., R. Awasthi, and J. Loeprick (eds.) (2011), “Risk-Based Tax Audits - Approaches and Country Experiences”, The World Bank, Washington D.C. Kingdom of Morocco (2014), “Revue des dépenses publiques”, Report no. 86115-MA, The World Bank, Washington D.C. Kirigia, J. M., L. G. Sambo, A. Renner, W. Alemu, S. Seasa and Y. Bah (2011), “Technical efficiency of primary health units in Kailahun and Kenema districts of Sierra Leone”, International Archives of Medicine 4, no. 15. Kirigia, J. M, L. G. Sambo, and H. Scheel (2001), “Technical efficiency of public clinics in Kwazulu- Natal province of South Africa”, East African Medical Journal, no. 78, 1-3. Kleven, H.J. and M. Waseem (2013), “Using Notches to Uncover Optimization Frictions and Structural Elasticities: Theory and Evidence from Pakistan”, Quarterly Journal of Economics 128, no. 2, pp 669-723. Kloeden, D. (2011), “Revenue Administration Reforms in Anglophone Africa Since the Early 1990s”, IMF Working Paper 11/162. -148- Kpodar, K. (2006), “The Distributional Effects of Oil Price Changes on Household Expenditures: Evidence from Mali”, IMF Working Paper 06/91. Lagomarsino, G., A. Garabrant, A. Adyas, R. Muga, and N. Otoo (2012), “Moving towards universal health coverage: health insurance reforms in nine developing countries in Africa and Asia.” Lancet, no. 380, pp 933–43. Lamauthe-Brisson, N. (2013), “Social protection systems in Latin America and the Caribbean: Haiti.” Project Documents, Economic Commission for Latin America and the Caribbean and United Nations. Lombardo, A. (2012), “Mapping of Social Protection Programmes in Haiti”, Oxfam, UK and UNICEF. López-Calva, L.F., D. Battiston, G. Cruces, A.M. Lugo, M. Santos (2013), “Income and Beyond: Multidimensional Poverty in Six Latin American Countries”, Social Indicators Research, Springer, vol. 112, no. 2, pp 291-314. Lunde, H. (2008), Youth and Education in Haiti: Disincentives, Vulnerabilities, and Constraints, FAFO, Oslo, Norway. Mallet, R. and R. Slater (2012), “Growth and Livelihoods in Fragile and Conflict-Affected Situations” Working Paper 9, Secure Livelihoods Research Consortium. Marek, T., R. Eichler, and P. Schnabl (2006), “Allocation de ressources et acquisition de services de santé en Afrique : Qu’est-ce qui est efficace pour améliorer la santé des pauvres? ”, Série des documents de travail sur le développement humain dans la Région Afrique n°105. The World Bank, Washington, D.C. Marschall, P. and S. Flessa (2009), “Assessing the efficiency of rural health centers in Burkina Faso: an application of Data Envelopment Analysis”, Journal of Public Health, no. 17, pp 87-95. Ministry of Economy and Finance (MEF) (2014), "Evaluation des effets de la réforme des subventions aux carburants et ses implications", Presentation at the Second Workshop on Options for Fuel Price Reform, May 2014, Port-au-Prince, Haiti. Ministry of Economy and Finance (MEF), Inspectorate General of Finances (2012), Audit du Programme de Subvention et de Dotation de Manuels Scolaires 2009-2010, Rapport Final, MEF, Port au Prince, Haiti. Ministry of National Education and Vocational Training (MENFP) (2013), Programme D’Interventions Prioritaires en Education (PIPE) : 2013-2016, MENFP, Port au Prince, Haiti. Ministry of Planning and External Cooperation (MPCE), Direction de l’Investissement Public (2013), “Analyse de l’Exécution du Programme d’Investissement Public 2012/2013”. Ministry of Public Health and Population (MSPP) (2013), Rapport des Comptes Nationaux de Santé 2010-11. Ministry of Public Health and Population (MSPP), Haitian Child Institute, and ICF International (2013), “2012 Haiti: Mortality, Morbidity, and Service utilization Survey: Key Findings”, Final Report, MSPP, IHE, and ICF, Maryland, United States. -149- Mohapatra, S., B. Moreno-Dodson and D. Ratha (2012), “Migration, Taxation, and Inequality”, World Bank Economic Premise Note 80, The World Bank, Washington, D.C. Moreno-Serra, R., S. Thomson, and K. Xu (2013), “Measuring and comparing financial protection”, Health system performance comparison, an agenda for policy, information and research, edited by Irene P. and P. C. Smith, pp 223-255, Open University Press. Morgan, L. (2012), “Results-Based Financing for Health: Performance Incentives in Global Health — Potential and Pitfalls”, The World Bank, Washington D.C. Moser, C. (2006), “Asset-Based Approaches to Poverty Reduction in a Globalized Context”, Global Economy and Development Working Papers, The Brookings Institution, Washington, D.C. Moser, C. (1998), “The Asset Vulnerability Framework: Reassessing Urban Poverty Reduction Strategies”, World Development 26, no. 1, pp 1-19. National Bureau of Statistics (NBS) [of Tanzania] and ICF Macro (2011), Tanzania Demographic and Health Survey 2010, NBS and ICF Macro, Dar es Salaam, Tanzania. Niyuhire, F (2010), “More Health for Every Dollar: Results-Based Financing”, The World Bank, Washington D.C. OECD (2009), “Taxation, State Building and aid”, OECD Factsheet-Update. Pan American Health Organization (PAHO) (2010), Haiti Health Facilities: Monitoring Health Service Response Capacity, http://ais.paho.org/phip/viz/haiti_healthfacilities_v2.asp . Pierre, Y. F., G. R. Tucker, and J-F.Tardieu (2009), ʺLost Childhoods in Haiti: Quantifying Child Trafficking, Restaveks, and Victims of Violence, Pan American Development Foundation, Port-au-Prince. Primature de la République d’Haïti (2014), “Bulletin d’information Programme de Scolarisation Universelle Gratuite et Obligatoire”, Bulletin vérité no. 1. Rajaram, A., T. Minh Le, N. Biletska, and J. Brumby (2010), “A Diagnostic Framework for Assessing Public Investment Management”, WPS5397, The World Bank, Washington, D.C. Ravallion, M., and S. Chen (2007), “China's (Uneven) Progress against Poverty”, Journal of Development Economics, 82, no. 1, pp 1-42. Rawson, I. and G. Berggren (1973), “Family Structure, Child Location and Nutritional Diseases in Rural Haiti.” Journal of Tropical Pediatrics and Environmental Child Health, no. 19, pp 288- 298. Research Triangle Institute (2010), Haiti : Early Grade Reading Assessment - Rapport pour le Ministère de l’Education et la Banque Mondiale, RTI : Research Triangle Park, NC. Sachs, J. (2001), “Investing in Health for Economic Development: Report of the Commission on Macroeconomics and Health”, Commission on Macroeconomics and Health, Geneva. Saksena, P., K. Xu, and V. Durairaj (2010), “The drivers of catastrophic expenditures: outpatient services, hospitalization or medicines?” World Health Report, Background Paper no. 21. -150- San S., M., and H. Lemma (2010), “Efficiency of the health extension programme in Tigray, Ethiopia.” BMC International Health and Human Rights 10, no. 16. Schieber, G., C. Cashin, K. Saleh, and R. Lavado. (2012), “Health Financing in Ghana”, The World Bank, Washington, D.C., doi:10.1596/978-0-8213-9566-0. Scheiman, I., J. Langenbrunner, J. Kehler, C. Cashin, and J. Kutzin (2010), “Sources of funds and revenue collection: reforms and challenges”, Implementing health financing reform: lessons from countries in transition, edited by J. Kutzin, C. Cashin, and M. Jakab, p. 87-118, The World Health Organization, Geneva. Secure Livelihoods Research Consortium (2012), “Social Protection and Basic Services in Conflict- Affected Situations: What Do We Know?”, Briefing Paper 2, ODI, London. SEDLAC (2014), Socio-Economic Database for Latin America and the Caribbean – CEDLAS and the World Bank, Enrollment in LAC Database. Shepard, D., D. Hodgkin, and Y. E. Antony (2000), Analysis of hospital costs: a manual for managers, The World Health Organization, Geneva. Street, A., J. O’Reilly, P. Ward, and A. Mason (2011), “DRG-based hospital payment and efficiency: Theory, evidence, and challenges”, Diagnosis Related Groups in Europe. Moving towards transparency, efficiency and quality in hospitals, edited by R. Busse, A. Geissler, W. Quentin, and M. Wiley, pp 93-114, European Observatory on Health Systems and Policies Series, WHO Regional Office for Europe, Copenhagen. Suzuta, E. (2011), “Education in Haiti: An overview of Trends, Issues and Plans”, WISE Qatar Foundation, http://www.wiseqatar.org/sites/default/files/files/4851/eduoverview0921_draft_with_cover.pdf Taskforce on innovative funding international financing for health systems, http://www.internationalhealthpartnership.net/en/about-ihp/past-events/high-level- taskforce-for-innovative-international-financing-of-health-systems/. UN Special Envoy for Global Education (2013), Accelerating Progress to 2015 Haiti. UNESCO, UNESCO Institute for Statistics database, UNESCO, Paris, France, http://data.uis.unesco.org/Index.aspx?DataSetCode=EDULIT_DS&popupcustomise=true&lang=en Unité de Lutte Contre la Corruption (ULCC). 2013. Rapport de Vérification du Programme de Scolarisation Universelle Gratuite et Obligatoire (PSUGO). ULCC: Port au Prince, Haiti. ______ (2014), Education For All Global Monitoring Report 2013/14, UNESCO, Paris, France. UNICEF (2011), Children in Informal Alternative Care, Discussion Paper, UNICEF, New York. ______ (1999), Child Domestic Work, Innocenti Digest, International Child Development Center, Florence. USAID (2012), The Early Grade Reading Assessment in Haiti, USAID: Washington, D.C. ______ (2011), Health service delivery costing. Health center analysis, Management Sciences for Health (MSH). -151- ______ (2009), Costing the basic package of health services in Liberia at clinics and health centres in Liberia. Rebuilding Basic Health Services Project (RBHS), JSI Research and Training Institute, Inc., the Johns Hopkins Center for Communications Programs (JHUCCP), JHPIEGO, and Management Sciences for Health (MSH. Verner, D. (2008), “Labor Markets in Rural and Urban Haiti: Based on the First Household Survey for Haiti”, Policy Research Working Paper 4574, The World Bank, Washington, D.C. Wagenaar, A. C., M. J. Salois, and K. A. Komro (2009), “Effects of beverage alcohol price and tax levels on drinking: a meta-analysis of 1003 estimates from 112 studies” Addiction, no. 104, pp 179–190. World Bank (2014), Haiti Poverty Assessment: Poverty Profile and Trends, The World Bank, Washington D.C. ______ (2014), Résultats de l'évaluation douanière (CAT‐R) préparés pour le Service des douanes de Haiti, The World Bank, Washington, D.C. ______ (2013), “Strengthening Competitiveness and Promoting Diversification in Haiti - Diagnostic Trade Integration Study”, The World Bank, Washington, D.C. ______ (2013), Stop Conflict, Reduce Fragility and End Poverty: Doing Things Differently in Fragile and Conflict-affected Situations, World Bank Paper, The World Bank, Washington, D.C. ______ (2012), “Health Equity and Financial Protection Datasheets”, Latin America and Sub- Saharan Africa, www.worldbank.org/povertyandhealth . ______ (2012), “Interim Strategy Note for the Republic of Haiti for FY13-FY14”, Report No. 71885- HT, The World Bank, Washington, D.C. ______ (2011), Interim Strategy Note for the Republic of Haiti for CY 2012, Report No. 65112-HT, The World Bank, Washington, D.C. ______ (2010), Public Expenditure Tracking and Service Delivery Survey – Education and Health in Honduras, Latin America and the Caribbean Regional Office, The World Bank, Washington D.C. ______ (2010), Promoting Nutrition Security in Haiti: An Assessment of Pre- and Post-Earthquake Conditions and Recommendations for the Way Forward, The World Bank, Washington, D.C. ______ (2009), “Burkina Faso: The Budget as Centerpiece of PRSP Implementation”, Public Expenditure Review, Africa Region, The World Bank, Washington D.C. ______ (2008), Haiti: Public Expenditure Management and Financial Accountability Review, The World Bank, Washington, D.C. ______ (2006), Haiti: Social Resilience and State Fragility in Haiti. Report No. 36069, The World Bank, Washington, D.C. ______ (2004), Benin Enhancing the Effectiveness of Public Spending. A Review of Three Sectors, The World Bank, Washington D.C. -152- ______ (1998), Haiti The Challenges of Poverty Reduction, Report No. 17242-HA, Volume I, The World Bank, Washington, D.C. ______ , World Bank EdStats: Education Statistics (database), The World Bank: Washington, D.C., http://datatopics.worldbank.org/education/. World Bank and ONPES (2014), Haiti. Investing in people to fight poverty in Haiti: Reflections for evidence-based Policy-making (Poverty Assessment). World Bank Group: Washington D.C. World Bank, USAID, MSPP (2013), Études sur les ressources humaines et les dépenses de santé dans les départements du Nord-Est, Plateau Central et Nord-Ouest. World Economic Forum (2011), The Global Competitiveness Report 2011-2012, Geneva. World Health Organization (2014), Initiative Pour un Monde Sans Tabac. Taxation des produits du tabac, http://www.who.int/tobacco/economics/taxation/fr/ . ______ (2014), Initiative Pour un Monde Sans Tabac. Taxation du tabac: un financement novateur des systèmes de santé, http://www.who.int/tobacco/economics/innovative_financing/fr/ . ______ (2014), Haiti: Health Profile, Geneva. ______ (2014), “Trends In Maternal Mortality: 1990 to 2013”, Estimates by WHO, UNICEF, UNFPA, the World Bank and the United Nations Population Division. ______ (2013), “Enforcing bans on tobacco advertising, promotion and sponsorship,” WHO Report on the Global Tobacco Epidemic. ______ (2013), World Health Statistics, the World Health Organization, Geneva. ______ (2012), NHA Report 2005-2006, the World Health Organization, Geneva. ______ (2011), The Abuja Declaration: Ten Years On. ______ (2010a), “Monitoring the Building Blocks of Health Systems: A Handbook of Indicators and Their Measurement Strategies”, The World Health Report 2010, Geneva, http://www.who.int/healthinfo/systems/WHO_MBHSS_2010_full_web.pdf . ______ (2010b), “Health systems financing: the path to universal health coverage,” The World Health Report 2010, Geneva. ______ (2010), Rapport sur la Santé dans le Monde, The World Health Organization, Geneva. ______ (2008), Toolkit on Monitoring Health Systems Strengthening: Health Systems Financing, http://www.who.int/healthinfo/statistics/toolkit_hss/EN_PDF_Toolkit_HSS_Financing.pdf, the World Health Organization, Geneva. ______ (2005), Distribution of health payments and catastrophic expenditures methodology, by Ke Xu. Discussion paper, no 2. ______ (2004), “The Impact of Health Expenditure on Households and Options for Alternative Financing.” Report EM/RC51/4, Regional Committee for the Eastern Mediterranean, the -153- World Health Organization, Cairo, http://www.who.int/health_financing/documents/cov- emrc-healthexpenditureimpact ______ (2002), World Health Survey, A household questionnaire, World Health Organization, Evidence and Information for Policy, the World Health Organization, Geneva. ______ (2001), “Macroeconomics and Health: Investing in Health for Economic Development”, Report of the Commission on Macroeconomics and Health, the World Health Organization, Geneva. ______, Website, http://www.who.int/nha/country/hti/en/ World Health Organization and the World Bank (2013), Monitoring Progress towards Universal Health Coverage at Country and Global Levels: A Framework, Joint WHO and World Bank Group Discussion Paper 1. World Vision (2012), Le Système de Protection de l’Enfant en Haïti, World Vision, Port au Prince, Haiti. Xu, K., D. Evans, K. Kawabata, R. Zeramdini, J. Klavus and C. Murray (2003), “Household Catastrophic Health Expenditure: A Multi-Country Analysis.” Lancet 362, pp 111-117. Yves-Francois, P., G. Smucker and J. Tardieu (2009), “Lost Childhoods in Haiti: Quantifying Child Trafficking, Restaveks and Victims of Violence”, Final Report, Pan American Development Foundation. Zeng W., M. Cros, K. Dilley, and D. Shepard (2013), “Impact of performance -based financing on primary health care services in Haiti.” Health Policy and Planning. Zephyr, D., A. Cordova, H. Salgado and M. Seligson (2011), “Haiti in Distress: The Impact of the 2010 Earthquake on Citizen Lives and Perceptions”, Joint Study by USAID and Vanderbilt University, http://www.vanderbilt.edu/lapop/haiti/2010-Haiti-in-Distress-English.pdf . -154- Annexes -155- Annex 1 : PIM Process Flowcharts A. Investment Projects Processing at MPCE DSE -156- B. Investment Projects Processing at MEF DASIP -157- C. Investment Projects Processing at MEF Treasury -158- D. Investment Projects Processing at Line Ministry: From Request for Funds to Check Issue -159- E. Investment Projects Processing at Line Ministry: From Check Issue to Payment to Beneficiary -160- Annex 2 : Correlates with Enrollment School Participation is Unequal-Poverty, Location, Disability, and Other Characteristics Strongly Correlate with Enrollment Dependent Variable: Currently In School Variable Marginal effect Standard error Age 0.034*** (0.011) Gender (1 = male) -0.004 (0.008) Child of household head 0.026* (0.014) Disabled -0.499*** (0.109) Total number of children in household (age 0-18) 0.000 (0.004) Annual hhld consumption per capita (per 1000 gourdes) 0.002*** (0.001) Household head attended but did not complete primary -0.011 (0.013) Household head completed primary 0.038*** (0.013) Household head completed lower secondary or above 0.050*** (0.010) Urban area of residence 0.020* (0.012) Artibonite 0.047*** (0.012) Centre 0.009 (0.018) Grand'Anse 0.025* (0.014) Nippes 0.059*** (0.007) Nord 0.022 (0.014) Nord-Ouest 0.015 (0.021) Ouest 0.031** (0.015) Sud 0.020 (0.016) Sud-Est 0.047*** (0.012) Mean value of dependent variable 0.9065 Observations 4,939 Notes: The regression is estimated for children ages 6-14. Marginal effects evaluated at sample means. Omitted household head education level = no schooling. Omitted department = Nord-Est. Robust standard errors clustered at the household level. *Significant at 10percent. **Significant at 5percent. ***Significant at 1percent. -161- Annex 3 : Correlates with Overage Status Poverty, Location, Gender, and Other Characteristics Strongly Correlate With Overage Status Dependent Variable: At Least 2 Years Over Age For Current Grade Variable Marginal Effect Standard Error Age 0.191 (0.195) Gender (1 = male) 0.079*** (0.028) Child of household head -0.206*** (0.045) Disabled 0.231 (0.167) Total number of children in household (age 0-18) 0.052*** (0.013) Annual hhld consumption per capita (per 1000 gourdes) -0.0030*** (0.001) Household head attended but did not complete primary -0.084** (0.039) Household head completed primary -0.234*** (0.042) Household head completed lower secondary or above -0.285*** (0.045) Urban area of residence -0.192*** (0.036) Artibonite -0.175** (0.074) Centre -0.057 (0.078) Grand'Anse 0.034 (0.077) Nippes -0.105 (0.078) Nord -0.084 (0.074) Nord-Ouest -0.254*** (0.070) Ouest -0.115* (0.066) Sud -0.164** (0.081) Sud-Est -0.040 (0.082) Mean value of dependent variable 0.5165 Observations 2,380 Notes: The regression is estimated for children ages 10-14. Marginal effects evaluated at sample means. Omitted household head education level = no schooling. Omitted department = Nord-Est. Robust standard errors clustered at the household level. *Significant at 10percent. **Significant at 5percent. ***Significant at 1percent. -162- Annex 4 : EDE PEP Programs: A Glossary EDE PEP has been Haiti’s Government umbrella for social assistance programs since 2013. Its objective is to allow beneficiaries to escape extreme poverty and invest in developing human capital and economic inclusion. Under EDE PEP, 19 social assistance programs are in operation through the MAST, FAES, OFATMA, MSPP, MENFP, PNCS, UCLBP and the MARNDR.  Bon Solidarité (or Bon Dijan) is a one-time cash transfer granted to extreme poor who need immediate intervention. This is an emergency program that targets [assuming it’s still going on] households affected by natural disasters.101 Implementation: FAES.  Carte Rose (or Kat Solidarite) is a contributory health insurance program that provides access to a basic health care package. Implementation: OFATMA.  Community health centers. This program aims to implement community pharmacies providing basic health care (implementation starts in 2014-2015). Implementation: MSPP.  Community restaurant is an in-kind transfer in the form of free meals, provided by 300 restaurants, to people in disadvantaged neighborhoods in the metropolitan area of Port-au- Prince. Implementation: FAES (resto PEP) and MAST.  Family planning aims to facilitate the establishment of family planning health care centers (implementation starts in 2014-2015). Implementation: MSPP.  Fight against cholera aims to facilitate the implementation of cholera vaccination health care centers (implementation starts in 2014-2015). Implementation: MSPP.  HIMO is a program to develop infrastructures in rural areas (to be determined in 2014-2015). Implementation: MARNDR.  Kantin Mobil is an in-kind transfer that provides warm meals to extreme poor living in urban areas. Distribution is prioritized for areas that require urgent intervention. Implementation: FAES.  Kore Etidyan is a cash transfer targeted to vulnerable families, with a household member less than 30 years old attending university. Implementation: FAES.  Kore Moun Andikape is a monthly cash transfer program for individuals between the ages of 18 and 65 living in extreme poverty who have at least one disability that prevents them from working, and who receive no subsidy. They must live in households, not in disability assistance centers. Implementation: FAES and CAS.  Kore Peyizan aims to support vulnerable households involved in agricultural production by providing a set of agricultural inputs, fertilizers, etc. Implementation: FAES and MARNDR. 101 Bon Dijans and Panye Solidarite started in a period of emergency after Hurricane Sandy. However, the GoH kept them as non-emergency programs for the most vulnerable, hence the change of name from Bon Dijan to Bon Solidarite. -163-  Kore Ti Gran Moun is a cash transfer program that targets households in extreme poverty with an individual over 65 who is not receiving any pension. Implementation: FAES and MAST (through the CAS).102  Literacy [teaching] program (“alphabétisation”) targets individuals from 18 to 65 who are illiterate. Implementation: SEA/MENFP.  Panye Solidarité is an in-kind transfer program that provides a locally produced food kit to households living in extreme poverty, in rural areas. This is an emergency program that targets households affected by natural disasters. Implementation: FAES.  PSUGO is the program for universal school enrollment. It provides a per-child subsidy to public and private schools. Implementation: MENFP.  Ranje Kay Kartier/ Banm Lumie-Banm Lavi programs consist of improving urban infrastructures (according to the needs by area). Implementation: UCLBP.  School Meals [or Canteens] is an in-kind transfer that offers school meals to children during school days. Implementation: PNCS.  Ti Kredi offers micro-credits to female-headed households engaged in a commercial/productive activity in urban and rural areas. Implementation: FAES.  Ti Manman Cheri is a conditional cash transfer for a mother living in disadvantaged urban areas, granted on the condition that her children are enrolled in school. Implementation: FAES 102 FAES provides a non-contributory transfer of 400 HTG and MAST provides a pension of 1000 HTG after 20 years of contributions (may be voluntary for informal workers). Harmonization of the program is planned. -164- Annex 5 : Price Structure for Petroleum Products (2010/2011) 2010-2011 91 95 Diesel Kero PrIx CIF 3.02 3.05 3.01 3.06 Taux annuel moyen 40.66315 40.66315 40.60315 40.66315 Frais financiers 4.50% 4.50% 4.50% 4.50% Total valeur en douane 3.1559 3.187226 3.145475 3.197724 Frais de vérification 5.0.0% 5.o.o.P,6 5.0001i 5.00% Droits de douane 57.80% 57.8oP,6 0.00% 0.00% Frais portuaires 0..32% 0.31% 0.30% 0.29% Droits d'accise 5.40% 5.40% 4.00% 4.00% Accise variable 5.30% 5.25% 3.13% 0.34% Total DTR 73.81% 73.76% 12.43% 9.63% Prix Ex-douane 223.0541 225.198 143.8016 142..S473 Marge compagnie 6.79% 0.082329 0.053ID 0.054132 Marge distributeur 9.02% 10.44% 8.60% 7.61% Transport en province 1.21% 1.20% 1.22% 1.20% MCDT% 17.03% 18.67% 14.97% 13.02% MCDT 21.84831 24.19457 19.15234 16.93213 . DTR + MODT 90.84% 92.43% 21.40% 22.65% Valeur accise Variabfe ($) 0.167228 0.161228 0.098369 0.010821 Pc en $ 6.02 6.13 4.01 3.91 Pc en gdes 244.90 249.39 162.95 159.48 Redevance carburant, gdes 1 1 1 1 Prix final 245.90 250.39 163.95 160.48 Prix à la pompe (Pp) 195 200 162 161 Sources: Direction de l’Inspection Fiscal, Calcul de l’Etude Source: IMF -165-