Report No. 36344-HT Haiti Options and Opportunities for Inclusive Growth Country Economic Memorandum June 1, 2006 Poverty Reduction and Economic Management Unit Caribbean Country Management Unit Latin America and the Caribbean Region Document of the World Bank ABBREVIATIONSAND ACRONYMS APN AutoritC PortuaireNational (National Port Authority) BRH Banque de la RCpublique d'Haiti (National Bank of Haiti) CAE Country Assistance Evaluation CAMEP Centrale Autonome MCtropolitaine d'Eau Potable (Metropolitan Water Authority) CDD Community DrivenDevelopment CEM Country Economic Memorandum CNIMP CommissionNational Interimaire des MarchCs Publics CNRA National Commission on Administrative Reform CPI Consumer Price Index CSCCA Cour SupCrieure des Comptes et du Contentieux Administratif (Supreme Audit Institution) DR Dominican Republic EDH National Electricity Company (Electricite` d'Hai`ti) EGRO Economic Governance ReformOperation EU European Union FER Fonds d'Entretien Routier (Road Maintenance Fund) FL Lavalas Family (Fanmi Lavalas) FY Fiscal Year GDP Gross Domestic Product GNI Gross NationalIncome HDI Human DevelopmentIndex HERO Haiti Economic Recovery and Opportunity Act HIV Human Immunodeficiency Virus HLCS HouseholdLivingConditions Survey (Enqudte des Conditions de Vie) ICF Interim Cooperation Framework IDA International Development Association IDB Inter-AmericanDevelopment Bank IHSI National Institute of Statistics and Informatics (Institut Haitien de la Statistique et de 1'Informatique) IMF InternationalMonetary Fund LAC Latin America andCaribbean Region LICUS Low Income Countries Under Stress MDG Millennium Development Goals MEF Ministry of Economy and Finance MENJS Ministitre de 1'Education Nationale, de la Jeunesse et des Sports (Ministry of National Education, Youth and Sports) MINUSTAH UNMission for the Stabilization of Haiti NGO Non-GovernmentalOrganization OED Operations Evaluation Department PAHO PanAmerican Health Organization PNH Police Nationale d'Haiti (Haitian National Police) PRSP Poverty Reduction Strategy Paper SME Small and MediumEnterprises TELECO Telecommunications utility TI Transparency International ... 111 Acknowledgements This report was completedunder the leadership of Mauricio Carrizosa (Sector Manager, LCSPE) and Caroline Anstey (Country Director, LCC3C). The team, led by Antonella Bassani (LCSPR) and Katherine Bain (LCSPR), consisted of Carlos Mollinedo (LPBWB), GabrielDemombynes(LCSPP), LinnHammergren(LCSPS), Mary Morrison (LCSFP), Samuel Carlson (LCSHE), Gillette Hall (SDV), Luc Razafimandimby (LCSHS), Roberts Waddle (consultant), Manuel Orozco (consultant), Margarita Chgvez de Silva (LCC3C) and Constance Polite(LCSPR). Valuable contributions were provided by Guillermo Perry (LCRCE), MathurinGbetibouo (LCCHT), ChingboonLee (LCSHD), Charles Feinstein (LCSFP), Garry Charlier (LCSER), Julio Velasco (LCSPE), Manuel Schiffler (LCSFW), Juan Navas-Sabater (CITPO), Nicolas Peltier (LCSFT), Joelle Dehasse (LCC3C), Stephanie Anne Kuttner (SDV), Ahmadou Mustapha Ndiaye (LCSFM), Patricia Macgowan (LCSPT), Dorte Verner (LCSEO), Clemencia Torres (LCSFE), Franka Braun (LCSPR), Massimiliano Paolucci (LCC3C) and Maria Beatriz Orlando (LCSPP). Peer reviewers were Ian Bannon (SDV), Brendan Horton (AFTP3) and J. HumbertoLopez (LCRCE). iv IGrowth.Poverty.InequalityandEducation........................................................ . 118 I1. The Provision of Education inHaiti: Current Constraints and Challenges......121 111 A Strategy for ImprovingEducation ServiceProvision inHaiti...................... . 130 TABLES Table 1.1: Average Annual Real Growth Rates of GDP per Capita. 1961-2000 (in percent) ................................................................................................................................ 2 Table 1.2: Explaining Changes in Growth Between Decades (change in the average growth rate of real GDP per capita)................................................................................... 10 Table 1.3: Volatility of Output per Capita (1961-2000)................................................... 10 Table 1.4: Explaining Changes in Growth Between Decades in Haiti and Other LAC Countries (change in the average growth rate of real GDP per capita - 1990s compared to 1980s) ............................................................................................................................ 11 Table 1.5: Explaining Changes in Growth Between Decades in Haiti and Other SSA Countries (change in the average growth rate of real GDP per capita - 1990scomparedto 1980s) ................................................................................................................................ 12 Table 1.6: Growth Simulations, 2000-2010 (change in the average annual growth rate of real GDP per capita) .......................................................................................................... 13 Table 1.7: Haiti's Governments, 1986-2005.................................................................... 14 Table 1.8: SelectedMacro Economic Indicators (2000-2005)......................................... 19 Table 1.9: VAT in Selected Countries.............................................................................. 22 Table 1.10: Summary Indicators on Hiring & Firing Workers ........................................ 23 Table 2.1: HouseholdIncome by Source.......................................................................... 38 (Means Percentages Across Households).......................................................................... 38 Table 2.2: Amount of RemittancesReceived.................................................................... 51 Table 2.3: How Do Remittance Recipients Use the Money?........................................... 52 Table Al: HeadcountPoverty Estimates for Haiti........................................................... 56 Table A2: Poverty by Population Subgroups ................................................................... 57 Table A3: Returnsto Education inHaiti .......................................................................... 58 Table A4: Summary Household Characteristics for Various Years................................. 59 Table A4: (continued) Summary Household Characteristics for Various Years .............60 Table A5: US$1-a-Day Poverty Estimates for Various Years Based on Poverty Projection Exercise............................................................................................................ 61 Table A6: US$2-a-Day Poverty Estimates for Various Years Based on Poverty ProjectionExercise............................................................................................................ 62 vi Table A7: Gini Coefficient Inequality Estimates for Various Years Based on Projection Exercise.............................................................................................................................. 62 Table A8: US$1-a-Day Poverty and Inequality Estimates from a Projection Forward from 1986to 2001 ............................................................................................................. 63 Table 4.1: Haiti's Infrastructure Compared......................................................................96 Table 5.1: Trends inKey Education Indicators. 1994-2000.......................................... 121 Table 5.2: Top Five Reasons for Not Attending School. in Order of Importance. by Quintile ............................................................................................................................ 123 Table 5.3: Haiti's Current Public Spending (2005/06) and EFA Medium-TermTargets for Education ................................................................................................................... 131 Table 5.4: Projected Costs of Achieving EFAby 2011(US$M) .................................. 132 Table 5.5: EstimatedPublic Financing Capacity andFinancing Gap for Achieving EFA Under Scenario 2 ............................................................................................................. 133 FIGURES 2 Figure 1.2: Growth Accounting Decomposition ................................................................. Figure 1.1: Per Capita Real GDP Growth. 1991-2004........................................................ 5 Figure 1.3: TFP contribution to Growth inHaiti and LAC ................................................ 5 Figure 1.4: Investment to GDP Ratio andPer-capita GDP Growth (1995-2003) ..............6 Figure 1.5: GDP Per-Capita inHaiti. Dominican Republic and LAC................................ Figure 1.6: Growth Accounting Decomposition inHaiti and DominicanRepublic ..........8 8 Figure 1.7: Most Problematic factors for Doing Business According to Entrepreneurs 14 Figure 1.8: Haiti: Kaufman Indexesof InstitutionQuality.............................................. Operating inHaiti.............................................................................................................. 15 17 Figure 1.10: International Comparison of Cost of Electricity ......................................... Figure 1.9: International Comparison of Key Indicators on Infrastructure..................... 17 Figure 1.11: International Comparison of Key Indicators on Education......................... Figure 1.12: Emigration Rate of top 20 SkilledEmigrationCountries, 2000 ..................18 18 Figure 1.13: International Comparison of Key Indicators on Saving and Financing......20 (percentage of total educatedlabor) .................................................................................. 18 Figure 1.14: Top 20 Developing-Country Recipients of Remittances, 2003 (percentage Figure 1.15: Concentration of Loan Portfolio.................................................................. 20 of GDP).............................................................................................................................. 20 Figure 1.16: Credit by Activity (% of total loanportfolio) .............................................. 20 Figure 1.17: Bank Spreads................................................................................................ 21 Figure 1.18: Assembly Sector Wage Costs (inUS$ per hour) ......................................... 23 Figure 2.1: Headcount Poverty Rates inHaiti.................................................................. 28 Figure 2.2: Headcount Poverty Ratesfor Low Income Countries Worldwide Using US$1-a-Day Poverty Line ................................................................................................. 29 Figure 2.3: Headcount Poverty Rates and Concentration of Poor by Region UsingUS$l- a-day Poverty Line............................................................................................................. 30 vii Figure 2.4: The Consequences of Poverty for FoodConsumption: Percentagesof 30 Figure 2.5: Income DistributioninHaiti ........................................................................... Households Facing Food Shortages................................................................................... Figure 2.6: HouseholdSize andNumber of Children by Income Quintile inHaiti.........31 32 Figure 2.7: DependencyRatio by Income Quintile inHaiti............................................. 33 Figure 2.8: Average Years of Educationby Age and Gender.......................................... Figure 2.9: Average Years of Education by Income Quintiles (AdultsAged 25 to 65)..34 34 Figure 2.10: MeanYears of Education vs.Inequality inYears of Education (AdultsAges 35 Figure 2.11: Literacy Rates by Age and Gender .............................................................. 25 to 65) for Countries inLatin American and the Caribbean......................................... 35 Figure 2.12: School Enrollment Pyramids by Age and Level of Schooling .................... 36 Figure 2.13: Education Attainment Profileby Economic Status, Age 15-19, Haiti 1994- 95 and 2001........................................................................................................................ 37 Figure 2.14: Share of Dwellings Made of Low-Quality Materials by Income Quintiles 39 Figure 2.15: InfrastructureAccess by Income Quintiles.................................................. Figure 2.16: Percentageof Households with Access to Infrastructure by Region ...........40 39 Figure 2.17: Infant andChild Mortality inHaiti and the DominicanRepublic ...............40 Figure 2.18: Infant Mortality RatesinHaiti by Region, 2001.......................................... 41 Figure 2.19: Agricultural Technology and Credit Usedby Farming Households, By 44 Figure 2.20: Child Malnutrition inHaiti and the DominicanRepublic............................ Income Quintile ................................................................................................................. 45 Figure 2.21: Household Assets and Services Over Time 1995-2001............................... 47 Figure 2.22: US$1-a-Day HeadcountPoverty Estimates for 1986, 1995,2000, 2001. 48 Figure 2.23: Consumption andWorkers Remittances (1985-2003)................................. Point EstimatesandConfidence Intervals......................................................................... Figure 2.24: Percentageof Households Receiving Remittancesby Income Quintile ......50 51 Figure 5.2: Non-Public Primary Enrollment by Type of Provider ................................ Figure 5.1: Educational Attainment of Primary Breadwinner by Income Decile .........120 122 Figure 5.3: Enrollment by Income Quintile................................................................... 123 Figure 5.4: Stated Reasonfor Leaving School among 6-24 year olds .......................... 124 Figure 5.5: Household Spendingon Primary Education Per Childby Income Quintile125 Figure 5.6: Average Primary and Secondary Education SpendingAmong All 126 Figure 5.7: MENJS Licensing by Type of Non-Public School ..................................... Households, as a percent of averagehousehold revenue, by quintile ............................. 134 viii EXECUTIVESUMMARY 1. The Haitian population has demonstrated resilience and creativity in the face o f severe challenges. Marred by political instability, economic mismanagement and exogenous shocks, Haiti has suffered negative economic growth in three decades of the last forty years. Even when economic growth has taken place, it has not been sustained. Haiti's pattern of socio-economic development has also been characterized by marked inequalities in access to productive assets and public services, the result of exclusionary policies and ineffective public institutions. The resulting widespread poverty has meant that less of the gains from growth, when this has materialized, have been shared by the poor. Inturn, the inability o f poor Haitians to exploit growth-promoting opportunities for investment in physical and human capital has created a vicious circle of weak economic growth and persistent poverty and inequality. 2. Following the 2006 presidential and parliamentary elections, the Government of Haiti and the international community have before them new opportunities to promote inclusive growth. Within this context, the challenge i s to identify and implement actions to overcome the legacy of past decades and set Haiti on a path to high, sustained and broad based growth. Constraints to growth and poverty reduction are widespread and severe, and they are interrelated and mutually re-enforcing. As discussed in Chapter one, in addition to political instability, the key constraints are: (i)ineffective institutions and poor economic governance, (ii) inadequate infrastructure, and (iii) access to and poor quality of education. Chapter two highlights Haiti's widespread poverty and high inequality in income and in access to basic services. The constraints to growth discussed in Chapter one and the poverty and inequality characteristics reviewed in Chapter two point to the need to ensure that the benefits of growth are shared through polices which, on the one hand, provide equal access to services for the poorest regions and households and, on the other, build institutions that are responsive, effective and accountable to their citizens. 3. Chapter three provides a historical overview of the development of the Haitian state and its impact on institutional capacity and economic governance. It sets out the foundations for a state building agenda in Haiti which, based on experience in fragile states around the world, is known to be the priority in helping such countries to set the basis for broad-based growth. This chapter presents a set of recommendations aimed at building the institutional capacity of core public agencies and limiting the opportunities for elite capture and corruption. It then outlines a new role for the state in terms of service delivery and in engaging users of services and citizens more broadly. Chapters four and five provide a more detailed discussion of how service delivery in infrastructure and education can be improved by strengthening the capacity and accountability of service providers and ensuring that users, especially the poor, have access. In doing so, the report lays out a short and medium term agenda for improving basic state functions in an efficient, transparent and accountable way, as well as expanding basic infrastructure and improving access to and quality of primary education. The successful implementation of this agenda would require close partnership and sustained commitment of both Government and donors. Table 1 attached below summarizes the main i recommendations contained in the above chapters. These are consistent with the broad thrust and build on the interventions identified in the Interim Cooperation Framework preparedin2004 by the Government with the support of the donor community. I.Trends,DeterminantsandConstraintsofHaiti'sEconomicGrowth 4. Haiti's economy is in a long-term slump. Between 1961 and 2000, Haiti's real income per-capita fell by an average of 1 percent per year, resulting in a per capita real GDP reduction of 45 percent over the period. Positive growth rates have been observed only in the decade of the 1970s as a result of greater political stability, paradoxically achieved during a dictatorship. Indeed, Haiti's economic performance has been one of the worst in Latin America and the Caribbean (LAC) region and similar to that observed inthe worst performingsub-SaharanAfrican (SSA)countries. 5. As discussed in Chapter one, Haiti's growth experience in the 1970s and the second half of the 1990s shows that the country i s capable of experiencing rapid growth. However, these spurts in economic activity were not sustained due to either restrictive or ineffective economic policies, or political instability. Rapid economic expansion during the 1970s was achieved through the interaction of private sector dynamism in exports of agricultural crops and assembly manufactures, and public sector investment in infrastructure. This spurt ineconomic activity was not sustainedinthe early 1980s as the Government increasingly intervened by employing fiscal and trade policies that were restrictive of the private sector and biased against exports, created monopolistic public enterprises and spent public funds without increasing the country's productive or absorptive capacity. The 1990s showed two opposite trends: in the first half, the economy entered in a profound recession (due to high political instability, a donor- imposed trade embargo and natural disasters); after 1995, the economy recovered, boosted by a significant but short lived increase in external aid. Yet the per-capita real GDP growth was negative for that decade. Growth performance in the first half of 2000s remained poor as a result of continued political instability. 6. The relatively better growth performance in the 1970s is associated with growth in total factor productivity (TFP), while in the rest of the period of analysis (1960-1969 and 1981-2003) TFP contribution to growth was negative. Haiti's contribution of physical capital went from one of the lowest in the region in the 1960s to being at about the regional standards in the 1980s, and it declined afterwards. The contribution of human capital shows only a slight upward trend since the 1960s. The above trends are linked to declining productivity in the agricultural sector, the unproductive nature of investment and measurementdifficulties facing the country's national accounting system, and the limitedcontribution of human capital throughout the period. 7. This chapter also reviews the empirical evidence providedby Loayza et al. (2005) on the determinants of economic growth inHaiti. This shows that the bulk of the change in real per capita growth is not explained by the traditional growth determinants as currently measuredand that there i s a large negative effect on growth not captured by the panel regression. A comparison with other LAC and SSA countries points to the model's inability to capture the negative effects of political instability and violence on growth. Forecasting Haiti's economic growth under an ambitious reform scenario (with policy 11 .. determinants of growth moving to the top 25 percent of the LAC region and the world), Loayza et al. project a significant potential change of 6-8 percent, with both structural and stabilization policies playing a role in supporting an increase in growth. Within these, improvements in public infrastructure and education would render the largest contribution to economic growth. As discussed in this Chapter, these forecasts tend to underestimate the impact of improved political stability and reduction in violence on economic growth. 8. This chapter's analysis of the constraints to growth in Haiti in a comparative perspective depicts the country as one of the most difficult environments in the world in which to do business. It identifies inadequate infrastructure, political instability and inefficient bureaucracy, and education as the critical constraints to growth in Haiti. Political instability has weakened private sector confidence and the quality of public institutions. While sustainedpolitical stability will be a necessary condition for restoring growth in Haiti, an improvement in the quality of public institutions and economic governance will also be critically needed to improve the business environment. Infrastructure limitations are also significant, but would need to be addressed with a strong emphasis on rehabilitation and maintenance. Similarly, much needed improvements in human capital need to be accompanied by the lifting of other restrictions to growth, least a swift improvement ineducation will only result inincreased brain drain with a limited impact on growth. 11. Poverty and Inequality 9. The analysis of poverty conditions and trends in Haiti detailed in Chapter two highlights a number of key points. The country i s the poorest in LAC and among the poorest in the world, with more than half of the population (54 percent) living below the US$1-a-day and 78 percent living below the US$2-a-day line (2001 data). In terms of other human welfare indicators, Haiti ranks at or near the bottom in the region, and among the worst in the world. An overwhelming portion of the rural population lives in poverty (86 percent). There are also large pockets of urban poverty in slum areas in Port- au-Prince, while many intermediate cities and small municipalities have low poverty rates. Wide disparitiesexist regionally, with poverty being the lowest in the Ouest region (34 percent), which includes the capital of Port-au-Prince, and highest in the Nord-Est region (81 percent). Nonetheless, even in the Ouest region poverty i s extremely highby international standards. Indeed, the poverty rate for the Ouest region, the richest in Haiti, i s higher than that of any country in Latin America and the Caribbean. 10. Haiti also suffers from substantial inequality, with nearly half of the national income going to those in the richest 10 percent of the population. Little variation i s seen in the quality of living conditions across the poorest 80 percent (which encompass the group defined as poor in the US$2-a-day poverty line). Inequality in Haiti i s among the highest in the LAC region, which in turn has higher levels of inequality than any other area of the world. Due to the country's widespread poverty, lack of security, frequency of natural disasters and dearth of formal private or public insurance mechanisms, most Haitians face high vulnerability. While one cannot expect absolute poverty to fall in Haiti without positive high growth, the country's high inequality has meant that less of 111 the gains from growth, when this has materialized, have been shared by the poor. Inequalities in access to infrastructure and social services and in ownership of assets have made it harder for poor Haitians to take up the opportunities afforded by the limited economic growth that the country has experienced. In turn, high inequality and poverty have been a constraint on the country's growth as poor Haitians' inability to exploit growth-promoting opportunities for investment in physical and human capital has perpetuated the low growth-high poverty circle. 11. Although recent data offers a clear portrait of poverty inHaiti inrecent years, the picture of what has happened over time in the country i s much hazier. Growth figures calculated from national accounts data show a near continual decline in GDP per capita over the 1990s. At the same time, two pieces of evidence suggest that individual welfare may have improved along some lines during that decade. Indeed, Demographic and Health Survey (DHS) data show improvements in asset ownership, housing conditions, some child health measures, and child nutrition and school attainment between 1995 and 2000. Additionally, estimates based on household surveys suggest that poverty and inequality rates have not changed substantially over the last two decades. Chapter two suggests that part of the explanation can be found in the fact that, while GDP per capita declined, consumption levels were maintained by remittances. In particular, the Haitian population has become increasingly reliant on the money sent b y the Haitian diaspora living abroad. The acceleration in the growth of remittances since the mid 1990s likely accounts for the fact that poverty may not have changed substantially during that period and some welfare measures may have improved modestly, even while GDP per capita continued to decline. Indeed, many Haitians have avoided falling into complete destitution in the face of many shocks (both domestic and exogenous) by using remittances for food consumption, and education and health services. 111. StateBuilding,Securityand Institutions 12. Contemporary Haiti lacks a set of public institutions capable o f carrying out the basic state functions in an effective, responsive and accountable fashion. In recent years, institutions have lost trained personnel and other resources, have not modernized their procedures and provided few incentives for staff performance. Political instability and budget constraints are part o f the explanation but, as Chapter three describes, the institutional gap has its roots in Haiti's historical development. Since independence, the lack of effective and accountable institutions and of rules that are aimed at addressing and responding to public interest have placed the country in a perennial cycle of contested governments, political instability, popular uprisings and violent regime changes. 13. Chapter three argues that Haiti's overarching challenge i s to reform public institutions. This entails strengthening the role of the state in four key areas: (i) capacity to guarantee citizen security within a rule of law, (ii) capacity for economic governance (sound macro-economic management, public administration, and market formation and regulation), (iii)capacity to plan and oversee basic service provision and infrastructure development - whether the state provides services and infrastructure directly or works through NGOs and private firms, and (iv) transparency and accountability. These areas iv are key since they address the basic functions of any state and have the potential for reducing poverty and inequality while promoting broad-based growth. 14. Since state-building i s an incremental process, the Government will need both a medium term strategy and a short term program of priority actions. Ensuring public security and the rule of law is the area where state failure i s most critical in Haiti,both for its direct consequences and for its indirect impact on institution building. Indeed, ensuring a dramatic improvement in security will be critical for the success of any other intervention in the country. This chapter recommends that short term actions focus on: improving coordination between MINUSTAH and the Police Nationale Haitienne (PNH), accelerating the reform of the PNH and targeting slum areas with multi- dimensional development interventions. Over the medium term, it i s recommended that the Government sustain police reform with elements to improve the relationship between the PNH and citizens, and strengthen other elements o f the judicial and penal system, startingwith criminal courts inurban centers and gradually broadening coverage to other geographic areas andjurisdictions. 15. Inthe area of economic governance, the chapter highlightsrecent advances made inHaitito increase transparency and efficiency inthe use of public resources andexternal assistance, notably changes in the legal framework for budget formulation and execution, the setting up o f critical institutions and agencies, and efforts at disseminating basic information and engaging some groups to monitor the implementation of the economic governance reforms. Going forward, Chapter three recommends that the short term focus of Government efforts be on ensuring the full implementation of the new legal framework by building capacity of central government agencies for revenue, expenditure and human resource management and then gradually extending coverage to offices handling financial management and planning functions in sectoral ministries. Ensuring these core institutions function appropriately and in coordination i s key for encouraging both a steady flow of resources (including those of donors) and their efficient use. The chapter also points to the need for redistributive policies--through increased collection of progressive taxes to mobilize much needed public resources, improved efficiency and targeting of public spending, and the provision o f public transfers to the poorest linkedto incentives to buildhumancapital. 16. Finally, this chapter highlights the importance o f promoting a transparent, results- based approach to public service delivery and engaging stakeholders in a way that promotes voice and accountability around public action. This requires building ministerial capacity to plan and track basic service delivery, whether by public or private providers at the central or decentralized level. This should focus on equalizing access to productive assets for the poorest. Where private sector provision of services i s working, it should be supported and scaled up. To ensure accountability and responsiveness, in the short term the Government should share basic information on service delivery, initiate efforts to improve public expenditure tracking, and enact a Freedom of Information Law to establish the general principles of transparent government and citizens' right to information. Medium term efforts could focus on using a more decentralized, territorial approach for planning public investments to better integrate the country and organize its development around the growth potential and needs of the various regions; further V mobilizing civil society groups to monitor government performance in service delivery and economic governance through social accountability initiatives at the local and national level; and enhancing the role o f the Parliament in the economic governance agenda, most notably inbudget approval and control. 17. Donors' assistance for state building and institutional development will be important. Lessons from past experience in Haiti and other countries facing similar institutional challenges point to the need for such assistance to be better coordinated, focused on the broad priorities highlighted above and sustained over time. Notably, donor-financed activities will need to be more disciplined about not competing with government for essential staff, working through government channels, and helping to fix, rather than circumventing, what does not work. IV. BuildingInfrastructure to Last 18. Dramatic improvements in Haiti's currently poor levels of infrastructure quality and coverage are essential to restart economic growth and reduce poverty. Infrastructure reform must also be a central part o f any state building strategy, both because the provision of infrastructure services (as public goods) i s a central, defining function of the state and because the malfunctioning of current infrastructure institutions is a major source o f current state weakness. The electricity sector represents a massive drain on scarce government resources; the telephone utility has historically been an extra- budgetary source of financing for politicians; and poor security and management o f ports result in a significant loss of government revenues. 19. The underlying argument of Chapter four is that a two-pronged approach i s needed for infrastructure reform in Haiti: on the one hand swift execution of projects by donors to produce rapid results, in the form of actual service delivery to the population and, on the other, medium to longer-term improvements in public capacity to pursue not just sustainability of investments, but a more effective and transparent government role. For the shorter term, this chapter discusses the need for more effective donor involvement. While some important successes - such as the setting up of the road maintenance fund - were achieved recently, the lack of an integrated, strategic approach and poor coordination in implementation have slowed progress. The chapter also considers the need for urgent donor efforts, built around infrastructure, to address the socio-economic causes o f violence in Haiti's slums. This issue i s o f particular relevance both because of its implications for state building and national stability, and because, while slum upgrading was an expressed priority o f the ICF, little has been done. Efforts to strengthen the Road Maintenance Fund to ensure effective road rehabilitation and maintenance are also urgently needed. 20. The medium to longer-term emphasis on governance and capacity strengthening builds first on the main lesson of past donor interventions in infrastructure: that works are not sustainable without maintenance, which requires institutional capacity to organize and pay for it. The analysis also considers the need to strengthen the minimal functions required of the state in infrastructure: designing strategies and policies, setting and enforcing the rules for public and private operations, and coordinating national and international actors engaged in the sector. This does not imply a larger role for the state vi but a cleaner and more effective one thanplayed at present. Six specific priority areas are highlightedfor infrastructure reform: Improving donor involvement. This involves both better coordination and securing longer term donor commitment. There i s a need for better information sharing among donors and the government, through stronger central capacity to compile and disseminate this information. Reinforcing the coordinating capacity of the existing sectoral tables i s considered, as well as the option of introducing some kind of donor resourcepooling systemfor infrastructure. Slum upgrading to prevent violence. Cross-sectoral slum upgrading projects/programs can prevent violence through mainstreaming violence prevention into the design of infrastructure projects (e.g. including labor-intensive construction, street lighting, and community centers) and through providing a vehicle for specific anti-violence components, such as mediation services. They also provide a channel for community organization. Pilots should be attemptedbefore larger projects. Strengthening the functioning of the Road Maintenance Fund through: enhanced coordination on the programming of road maintenanceactivities; ensuringregular and adequate budgetary transfers to the FER account of proceeds from earmarked taxes for road maintenance; and developing expertise inFER to promote an efficient model for routine and emergency road maintenance, using either small firms or community- basedmicro-enterprises. Making the Ministry of Public Works, Transport and Communication (MTPTC) more effective: a new institutional diagnosis should be undertaken for the MTPTC, and a firm plan developed for implementing its recommendations. Greater sectoral specialization is neededto develop the Ministry's ability to make policy and strategy andplan investments, through a strengthening of the cadre technique and appointment of merit-based, non-political staffing. It needs to better oversee and coordinate with, andinterfere less, in state-ownedinfrastructure companies. Some decentralization of functions to regional offices would makethese more effective. Overhauling state utilities. There i s a need for strong government leadership of reform, in particular to implement the recommendations of the audits and organizational diagnostics underway on the key state enterprises. Funds must be channeled to the rehabilitation of existing dilapidated assets and improving service to existing customersbefore coverage can be expanded. Maintenance must be provided for by ensuringadequate funding and training staff. Transparency and accountability require better information systems, a more active role for boards of directors and open contracting, and regular public disclosure of information on levels and quality of services provided to the public and the performance of public utilities. And realistic strategiesare needed for involving the private sector, where possible. Enhancing the legal and regulatory framework for public and private service provision. This requires not just updating laws but actually enforcing them. Structures are needed in particular to improve oversight, promote competition where possible, and set the ground rules for all operators. vii V. InvestinginChildren 21. Education i s a well documented condition for growth and also has beneficial impacts on nutrition, health, empowerment of girls and poverty. It provides a means for avoiding the inter-generational transmission of poverty and thus offers hope that children will help families break out of the vicious circle of poverty. Educational access in Haiti has declined in recent years and disparities between urban and rural enrollment have worsened. The costs of education for poor families is a major deterrent that prevents them from enrolling their children in schools. In addition, the public sector suffers weaknesses similar to other sectors and as a result education services are offered primarily through a range of private providers with negative implications for equality, quality and efficiency within the sector. Ninety-two percent o f schools in Haiti are operated by a diverse group of private actors ranging from commercial entities to faith- based schools who have stepped in to "fill the gap" in the education sector. While it i s helpful that this gap has been filled, the sanitary conditions and the quality of teaching in several private schools are major drawbacks due to a weak institutional capacity of the Ministry of Education in performing its normative role, providing a policy framework and a strategy that engages public and private actors, and ensuring some minimum accountability. This along with the low levels of public spending for the sector and the constraint on the demand side has resulted in some of the poorest educational indicators inthe world. 22. Chapter five recommends that the sector agenda focus on a re-defined role for the state in education in order to attain the goal of having every Haitian child in school. As inthe infrastructuresector, this report does not advocate a larger role for the public sector but a more effective and facilitating one, whereby public agencies set the sector strategy, provide the rules, regulations and incentives for public and private actors, respond to private demand by providing sufficient inspectors to license and accredit non-public schools, and ensure that donor investments in the sector are used in a coordinated and transparent manner. Priorities for the short-term include: increasingthe share of public spending on education to strengthen the state's capacity, and investing inprimaryeducation; strengthening the newly established National Partnership Office (NPO) which was launched by the Transition Government in an attempt to improve dialogue, strategic planningand accountability inthe sector; and launching an effort to scale up enrollment of children by providing public transfers (including through a major scale up of donor assistance) to private schools; emphasis would be put on those schools which can increase enrollment of the poorest children and ensure quality improvements. 23. These efforts would then need to be expanded over the medium-term and be accompanied by actions to strengthen the capacity of the Ministry o f Education to design and implement sector policies, notably with a focus on providing incentives for public and private actors and effectively discharging its role o f sector regulator and accreditation. The chapter presents the results of a costing exercise to assess the financing requirements for enrolling all Haitian children in school and improving the V l l l ... quality of education over the next ten years, The chapter also discusses the option of introducing a pilot approach to targeting cash transfers to the poorest households that are least able to afford the costs of education in exchange for their commitment to send their children to school and having them vaccinated and well-fed. There is now significant experience in such cashtransfer type of programs inmany middleincome countries inthe region and new results also show that they can be effective poverty reducing programs in low income countries and fragile states. However, providing public transfers to schools or cash transfers to poor families will require that a solid fund channeling mechanism be set up as well as an information data base to ensure that internal and external accountability mechanisms guarantee that resources are well targeted and used transparently and efficiently. VI. Other Interventions 24. This report is focused on a select list of cross-sectoral binding constraints to growth and poverty reduction which also constitute the minimum functions that a state mustfulfill to provide itsbasic, purepublic goods. This meansonly that theseconstraints and these functions require priority attention, not that they should be addressed to the exclusion of others that are more sector specific. Indeed, policies aimed at improving additional aspects of the climate for private sector activity in agriculture, manufacturing and services, as well as enhancingthe provision of other social services, will need to be implemented, once the critical bottlenecks identified above have started to be tackled. The Interim Cooperation Framework, prepared in 2005 by the Transition Government with donor participation, identifies priority interventions in several sectors. Also, several recent studies have examined sector specific constraints to growth and poverty reduction which, therefore, have not been discussedin this report.' However, the recommendations of those studies need to be part of the elaboration of a comprehensive medium-term growth and poverty reduction strategy for Haiti. Table 1:Summaryof PolicyRecommendations Priority Recommended Policy Actions Medium-Term Reform Actions State Building, Security and Judicial Reform: Security and Accelerate reform of PNH (police), incl. Sustain police reform efforts, with Institutions establishing new rules and operating standards, elements to improve the relationship between vetting and training personnel, and improving the PNH and citizens, including greater oversight arrangements and career accountability of police officers for areas management. under their jurisdiction and professionalization and training of the force's Pilot slum upgrading projects with a focus on members. violence prevention (also see recommendations for Infrastructure). Reform the judicial and penal system with emphasis on personnel selection, monitoring Several of these studies are listed in the Bibliography; these include a review of existing constraints and priority interventions for rural development (World Bank, Haiti-Agriculture and Rural Development, 2005), an assessment of existing social protection mechanisms (World Bank, Haiti-Social Protection Policy Brief, 2005) and a review of Haiti's conflict poverty trap (Haiti-Country Social Analysis, 2006). i x and evaluation, and provision of personnel adequate salaries; provision of additional material and technical assistance; reform of legal codes. 0 Repeat, since most of the benefits have been lost, the 1990s program of prison and prison staff upgrading. Economic Governance: Focus on ensuring full implementation of the Continue to ensure full implementation of the new legal framework by building capacity of new legal framework, with expanded focus to central government agencies for revenue, financial management and planning functions expenditure and human resource management: in sectoral ministries: 0 Continue to keep unprogrammed expenditure 0 Ensure a more redistributive and efficient through "current accounts" at minimum level pattern o f public expenditures by basing and reduce administrative lags in expenditure Government budget on a multi-year processing; develop standard formats for budget framework, adopting procedures for preparation and reporting, train staff in their use, preparation and selection of investment and monitor compliance; strengthen CSCCA projects; and adopting program budget including support in preparation of budget approach for key ministries. execution reports for the new legislature. Institute procedures that embody the broad principles lacking in the Organic Budget Law (i.e., transparency, performance, responsibility and consolidated cash management); continue to strengthen the CSCCA. Strengthen tax administration with main focus on enforcement expanded to provinces and Implement tax policy reform to broaden ports - restoring security and presence of the tax base and enhance revenue collection customs in sea ports in and outside Port au by fighting evasion, reorganizing tax Prince will be critical for this. collection functions along functional lines and creating ownership for reforms. Creation of registry of state employees and of agency overseeing application and further Elaborate inventory of all positions, development of rules on appointment, elaboration of simplified job classification remuneration, evaluation and overall career scheme, standardized salary grid, description management; identify critical needs for new of duties and needsbasedtraining programs. hires and training. Ensure compliance with new procurement Improve legal framework for procurement. decree and promote training of staff and potential bidders. ServiceDelivery: Start to develop in core ministries systems for Develop, with donor assistance, tracking service delivery, starting with review of mechanisms for citizen collection of data or information already collected, specification of evaluation o f services; use citizen input to set services and qualitativelquantitative goals and new targets and agree with them and donors indicators, and determining additional on indicators and monitoring process. information needs; engage sectoral tables established under ICF in this effort. Pilot territorial approach to participatory planning, building on existing initiatives and use it as basis for planning public investments. X Transparency, Accountability and Participation: B Continue to mobilize civil society groups Share information collected as per above as a to monitor government performance through means of demonstrating performance to the social accountability initiatives at the local broader public. Focus on improving and national level. information systems on public expenditures and sectoral project databases. Encourage and support grassroots organizations to participate in local territorial Enact a Freedom of Information Law and of plans in a meaningful way. a Law on Asset Declaration and Disclosure. Support Parliament to ensure it can play its legislative and oversight role; including on budget approval and control. Infrastructure Pilot slum upgrading projects with a focus on Scale up slum upgrading interventions. violence prevention. Revamp MTPTC as per results of Emphasize strengthening of core institutional diagnosis, to depoliticize, infrastructure (with rehabilitation and adequate increase sectoral specialization and expertise, maintenance) before these are expanded; in and decentralize some responsibilities. particular, strengthen the functioning of the Road Maintenance Fund. Overhaul state companies by increasing transparency and accountability, involving Continue to improve transparency and private sector, and acting on accountability of public utilities. recommendations of audits for ANP, EDH, CAMEP and TELECO. Commission institutional diagnosis for MTPTC to enhance its capacity and improve Enhance legal and regulatory framework to efficiency. strengthen supervisory institutions and accommodate competition and private Develop realistic strategies for involving involvement. private sector in state companies. Develop coherent territorial strategies for Improve donor coordination by clarifying planning countrywide growth, to drive responsibilities within government, infrastructure investment. strengthening database, appointing and funding active coordinators for all sector tables. Continue to improve donor coordination. Education Start gradual increase in public resources Continue gradual increase in public allocated to primary education to reach EFA resources allocated .to primary education to medium-term targets. reach EFA medium-term targets. Reinforce Ministry of Education's normative Strengthen Ministry's capacity to assess functions of regulation and quality control and studentlschool performance by improving capacity for non-public school accreditation. administration and use of national exams (incl. public disclosure of results at school Establish National Partnership Office as level). public-non public partnership coordination body and National Partnership Fund as 0 Enhance capacity of in-service teacher financial mechanism to support improvements professional development centers around the inpublic and non-public schools. country to service both public and non-public teachers. Launch effort to scale up enrollment of poorest children and improve educational Gradually expand effort to scale up quality by: (i)providing public transfers to non- enrollment of poorest children and improve xi public schools; or (ii)designing a conditional educational quality through public transfers cash transfer mechanism to address education to non-public schools or a conditional cash demand constraints faced by households. transfer mechanism; complement these efforts through continued development of Increase donor financing to support local school management committees and implementation of above interventions. school-parent associations. Other Policies Maintain macroeconomic stability. Maintain macroeconomic stability, Strengthen institutional capacity for Strengthen institutional capacity for preventing and mitigating natural disasters. preventing and mitigating natural disasters. Reduce inefficiencies in remittance sending processes, leverage remittances as collateral and help disseminate information to migrants' hometown associations on community-level viable projects. xii CHAPTER 1 TRENDS,DETERMINANTSAND CONSTRAINTSOF HAITI'S ECONOMIC GROWTH Haiti's economy is in a long-term slump. During 1961-2000, Haiti's real per-capita GDP fell by an average of 1 percent per year, resulting in an overall decline of 45 percent over theperiod. Despite positive growth rates observed during the 1970sand, to a lesser extent, in the second half of the 1990s,these spurts in economic activity were not sustained. The relatively better growth peflormance in the 1970s is associated with growth in total factor productivity (TFP), while in the rest of the period of analysis (1960-1969 and 1981-2003) TFP's contribution to growth was negative. The above trends are linked to declining productivity in the agricultural sector, the unproductive nature of investment and measurement dificulties facing the country's national accounting system and the limited contribution of human capital throughout the lastfour decades. A detailed analysis of the constraints to growth in Haiti in a comparative perspective depicts it as one of the most difsicult environments in the world in which to do business. Inadequate infrastructure, political instability and ineficient bureaucracy, and poor access to and quality of education emerge as the binding constraints to sustained growth. Political instability weakens private sector confidence and the quality of public institutions: the latter afsects the business environment and the provision of basic services. Infrastructure limitations are significant and need to be addressed with a strong emphasis on rehabilitation and maintenance. Similarly, much needed improvements in human capital need to be accompanied by the lifting of other restrictions to growth, least a swift improvement in education will only result in increased brain drain. I.GrowthPerformanceduring1961-2000 25. Growth Trends. Haiti's economy is in a long-term slump. Between 1961 and 2000, Haiti's real per-capita GDP fell by an average of 1percent per year, resulting in a per capita real GDP reduction of 45 percent during the period (see Figure 1.1). Positive growth rates were observed during the 1970s. The 1980s and 1990sshowed two opposite trends: (a) inthe decade of the 1980s and the first half of the 1990sthe economy entered in a profound recession, with this trend deepening in the first half of the 1990s; and (b) after 1995, the economy recovered. Yet the per-capita real GDP growth remained negative for the decade of the 1990s. Also, the economic recovery o f the second half of the 1990swas short lived. The average growth rate in2000-2004was negative. 26. Haiti's economic performance has been one of the worst in the world (see Table 1.1). Ten-year averages of GDP growth per capita show that the L A C region converted a pattern o f negative growth in the 1980sinto positive growth in the 1990s. Incomparison, Haiti's average annual real GDP growth per capita, which was already under-performing incomparison to the L A C average duringthe 1960sand 1970s, showed a greater decline during the 1980s, and even a deeper fall duringthe 1990s--byfar the worst performer in 1 the region. Haiti's average annual real GDP growth per capita has also been significantly below that of Sub-SaharanAfrica in all decades with the exception of the 1970s. Figure 1.1:Per Capita Real GDP Growth, 1991-2004 10% 5% 52cn0% a a -5% n .-m I m Q 0 &lo% a -15% -20% 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 - 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Y f v '- FranpoisDwalier Jean-Claude Dwaiier Unestabie A rislide Preval Aristide Tr. period Gov Note: The permanent componentwas estimatedusingthe Baxterand King filter. Source: World Bank, WDI (2005). Table 1.1: Average Annual RealGrowth Rates of GDPper Capita, 1961-2000 (in percent) *Simple average, own calculations **Weighted average, n= 26. For2000-02, weighted average. n= 22 *** Weightedaverage, n= 109. For 2000-02, weighted average, n = 103 Source: Ianchovichinaand Kacker, 2005. 27. Historical analysis. President Francois Duvalier's dictatorship in the 1960s marks one of the saddest chapters in Haitian history with tens of thousands Haitians killed or exiled. During this period, the Haitian economy suffered from a near dearth of private sector investment, declining world market prices and simultaneous increases in taxation for agricultural products which greatly depressedagricultural production; several hurricanes aggravated this situation by affecting coffee (then the major export) and other agricultural crop production. The virtual isolation of Haiti from the world's community as a result of President Duvalier's brutal dictatorship meant that during this period the country received almost no external aid for development purposes. 2 28. Following the death of Francois Duvalier in 1971 and the succession of the presidency to his son, Jean-Claude Duvalier, the Haitian economy entered a period of rapid economic growth due to: (a) expansion of small scale export-oriented assembly industries; (b) increased public investment, including for several major infrastructure projects;2 (c) strong increases in private consumption and housing investment related to rapidurbanization; (d) an improvement in the country's terms of trade since 1975 due to a significant increase in coffee prices; and (e) the growth of t o u r i ~ m . ~O f notable importance, during this decade the manufacturing sector doubled its production mainly due to the growth o f the apparel assembly that took advantage of the proximity of Haiti to the US, low labor costs and tax incentive^.^ However, since agriculture maintained a dominant share in GDP and continued to be dependent on whether hazards, the year-to- year pattern of GDPgrowth inthe 1970s remained erratic. 29. This spurt ineconomic activity experienced in the 1970s was not sustained in the early 1980s as the Government increasingly intervened by employing fiscal and trade policies that were restrictive of the private sector and biased against exports, created monopolistic public enterprises and spent public funds without increasing the country's productive or absorptive ~apacity.~This was combined with unfavorable external developments in the early 1980s, including the 1981-83 U S recession and declining world coffee prices which affected Haitian assembly and coffee exports, the closure of a bauxite mine in 1982, and the negative impact of HIV aid on the tourism sector. Following the departure of President Jean-Claude Duvalier administration in 1986, the situation was characterized by political instability and civil unrest associated with attempted elections and changes in governments. Six different presidents followed each other between 1988 and 1991. This resulted in loss of private investors' confidence, work stoppages, unattended economic problems and a sharp drop in much needed external assistance. Thus, except for a short-lived improved economic performance during 1986-87, the decade of the 1980s is characterized by declining output, rising inflation, unsustainable fiscal and balance of payments deficits, and a breakdown in the provision of general services. 30. The economic decline gathered considerable momentum after 1991 following the overthrow of the democratically elected President Aristide by a military coup d'etat and the subsequent imposition of sanctions by the international community in the form of an embargo on most trade and financial transactions andthe suspension of most external aid. A major collapse of the economy ensued. Between 1991 and 1994, real GDP fell by around 20 percent and the rate of inflation rose from 12 percent to 51 percent. Loan disbursements were stopped and substantial arrears accumulated on external debt service payments to bilateral and multilateral institutions. Exports o f goods and services fell * These included the construction of the hydro-electric station at Peligre and improvements to the Port-au- Prince harbor and to the telephone system. Arrivals climbed from an average annual o f 50,000 during the 1960sto about 300,000 in the late 1970s. The number of enterprises operating inthe assembling sector grew from 13 to 127 between 1967 and 1979. Inparticular, the Government directed a significant share of public investment to uneconomic projects, including the establishment or takeover of five major industrial enterprises which were run inefficiently and represented a massive drain on public funds. 3 from US$224 million in 1991 to US$67 million in 1994, while at the same time imports of goods and services dropped from US$532 million to US$235 million. Investment as a whole decreased from 14 per cent of GDP in 1991 to 6 per cent in 1994, with the fall in public investment (resulting from the suspension of most external aid) accentuating the deterioration of Haiti's physical and social infrastructure. The manufacturing sector was hit the hardestby the 1991-1994 trade embargo, which limitedaccess to imported inputs and thus severely affected the productive capacity of the sector. The trade embargo spawned a 50 percent reduction of the sector's production. Following the lifting of the embargo in 1994, only the garment sector re-established itself and other assembly exports have virtually disappeared. As a result, Haiti's exports have become highly concentrated (with nearly 90 percent of these being apparelexports). 31. With the return to democracy and the end of sanctions in 1994, an economic reform program was initiated aimed at restoring macroeconomic stability, liberalizing trade and modernizing State enterprises. This paved the way for a substantial increase in financial flows to Haiti6 In 1996-97, the economy grew at 3.4 percent largely driven by construction and public works as a result of the repairs carried out on highways leading into some major agricultural production zones and other urban infrastructure works. The economic recovery was also the result of a significant increase in exports of textile and agricultural products resulting from improved private sector confidence. However, this recovery was very short lived and during 1997-1999, the growth rate of Haiti's economy showed a virtually steady decline and there was even a recession from 2000 to 2002. This was the result o f (i) political crisis (during 1997-99),7 which led to a reduction in international aid; and (ii)fall in coffee and cacao prices, leading to a decrease in export a earnings. The devastation caused to agricultural plantations by hurricane Georges in September 1998 aggravated the deterioration in economic performance. In 2000, agricultural production also sufferedfrom a serious drought. 32. The political deadlock following the disputed 2000 parliamentary elections again undermined private sector confidence and dampened investment, and led to a sharp cutback indonor assistance. The political polarization intensifiedin late 2003, leading to street demonstrations and increasing violence that culminated in an armed conflict and President's Aristide's resignation in early 2004. The May and September 2004 floods further aggravated the situation causing damage to the economy estimated at 5.5 percent of GDP. As a result, the period 2000-2004 was characterized by negative real GDP growth (-1.2 percent on average), highinflation, and large fiscal and external deficits. 11. Growthdeterminants the traditionalapproaches - 33. The traditional approach of the growth accounting framework, which breaks GDP growth down into growth of factor input quantities and the growth of total factor productivity (TFP), i s employed to analyze the sources of Haiti's poor growth performance. The methodology identifies the basic factors that explain growth: labor Net externalresourceflows amountedto aroundUS$650 million (or 25 percentof GDP) inFY1994/95, before leveling off to around US$300 million (7 percentof GDP) in FY1998/99. 7 In 1997, President Aristide broke from his OrganisationsPopulaires de Lavalas to create Fanmi Luvalas. The split ledto adeadlockedParliamentin 1997-98which was subsequentlysuspendedin 1999. 4 use, physical capital, human capital and total factor productivity (i.e. the unexplained residual after deducting the contribution of capital and labor). The growth of the latter is usually the result of the continuous process o f factor reallocation to activities of higher productivity as well as adoption of new technologies or better "quality" factors that boost production without commensurate increases infactor use. 34. Haiti's contribution of physical capital went from one of the lowest in the region in the 1960s to being at about the regional standards in the 1980s, and it declined afterwards. The contribution of human capital shows a slight upward trend since the 1960s. Remarkably, unlike what happened in the rest of the period of analysis, i.e. 1960- 2003, the relatively better growth performance in the 1970s is associated with a positive contribution of TFP. Inthe rest o f the period of analysis, TFP contribution to growth was negative, thus negatively affecting economic performance (see Figure 1.2). Comparing the Haiti TFP trends within the region, TFP contribution to growth was much lower than regional values, except duringthe 1970s (see Figure 1.3). Figure 1.2: Growth Accounting Figure 1.3: TFP contributionto Growth in Decomposition Haiti and LAC 5% ~," I 4% I 3% 3% 2% 2% f 2% :1% 2 - 1% 1% 5 : t P 0% L 0 % ; 0% dg 5 1% 1% B gB 2% ; 2 % 2% c E -3% 3% a -4% 19609 19705 19801 19901 1994.2003 OCap Sock GroMh DHvman Capital OIouth mTFP groWh 196- 187- 1880s 1PWI 1SS4-2W3 Source: Own estimates based on dataset compiled by Note: The boxplot representsTFPs of the LAC region. SandeepMhajan(2002) Source: Own estimates based on dataset compiled by Sandeep Mhajan(2002) 35. The above trends are related to three main issues affecting Haiti's economy: declining productivity in the agricultural sector, the unproductive nature of investment and measurement difficulties facing the country's national accounting system, and the limited contribution of human capital throughout the period. As shown in Avila and Evenson (2004), aggregate TFP index growth rate for Haitian agriculture was 0.6 percent during 1961-80 and -1 percent during 1981-2001 as compared to 1 and 1.7 percent for Sub-Saharan Africa and 1.5 and 2.4 percent for Latin America and the Caribbean during the same periods. Improvements in agricultural productivity have been constrained by the elevated fragmentation of rural property' combined with peasants' low incomes. Today, around 80 percent of land owners own less than two hectares and only two percent own more than ten hectares. Also, peasants' income levels have been so low that there has been virtually no peasant saving or investment. Hence, technology levels and labor productivity remain strikingly low: for the majority of peasants farm implements Land fragmentation i s due mainly to subdivision on inheritance. 5 and production techniques remain those o f two centuries ago (see Chapter 2).9 Agricultural productivity increases have therefore been minimal and have not kept pace with the growth of the population. Low technology levels have turned agriculture into a soil-use intensive activity, accelerating soil deterioration and further reducing the productivity of the sector. In the 1960s-1980s, Government policies compounded the situation with the levying o f high export taxes and the maintenance of export monopolies." Producer incentives were also eroded by high transport costs, lack of access to finance (see para. 59 ) and adverse natural events (notably hurricanes, floods and droughts). This resulted in peasants being trapped in a price-cost squeeze, with international relative prices for agricultural exports falling more rapidly than costs of production. Because of insufficient producer incentives, peasants have tended to increase food production, mainly of non-taxed subsistence crops, and cut down production for export. As a result, the exporting capacity of Haiti's agricultural sector has been reduced to a few products o f modest significance and o f a volatile nature. Agricultural production has contracted to the point that half of the food consumed in Haiti i s imported which generates significant pressures on the balance of payments. Figure1.4: Investmentto GDPRatioandPer- 36. High levels of investments capita GDPGrowth(1995-2003) have been common throughout the 1965-2004 period, interrupted only during the embargo of the first half Q o f the 1990s, when investment 0 I decreased by around 4 percentage points of GDP. After this decline, investment climbed back to levels above 20 percent o f GDP, attaining at some point levels o f more than 30 percent o f GDP. And yet, those high 0 0 levels of investment did not translate 4% into economic growth. Figure 1.4 LatinAmenca o Low incomecounlnes 6% .... .. . 1 0 .. ................ ...................... .. .. . .................................. ....................... .... ... ..... . ................... .. ....... illustrates the relationship between wb 5% 10% 15% 20% 25% 30% 35% 40% 45% M% the levels o f investment in terms of Inieslmenl to GDP (GrossCapitalFomatipn) Source: WDI (2005) GDP and the per capita growth rate inLAC and inlow income countries. 37. Haiti's paradox of high investment levels (in terms of GDP) accompanied by a sustained contraction of per capita GDP can be explained by the unproductive nature of investment and measurement difficulties facing the country's national accounting system. The unproductive nature of past investments relates to: (1) chronic lack of routine maintenance for infrastructure (see Chapter 4); (2) the uneconomic nature of some large In2002, inHaiti there was a tractor per 5,570 hectares- they were respectively 500, 323 and 403 hectares each per tractor in Nicaragua, the Dominican Republic and Bolivia - much higher than the LAC regional average of 84 hectares per tractor (CEPAL, 2005). loThemoststrikingexampleiscoffee, thechiefexportcrop. Throughoutthelastfourdecades,atleast one million people on the hillsides depended on it for their livelihood. Yet, coffee was subject to high export taxes (representing around 40 to 50 percent o f producer prices). 6 public investments;" and (3) the fact that investments were undertaken in an overall non- conducive environment. Poor past donor coordination and the high volatility of external aid have also affected the impact of investments in Haiti. Yet the above explanations alone are insufficient to explain such a paradox; with recurring negative returns on capital, a contraction of investment - and not a sustained increase like inHaiti - should be observed. This seems to indicate that the high levels o f investment in Haiti are due, at least in part, to measurement problems within the national accounts, a subject that i s discussed inBox 1.1. Box 1.1: The Enigma of High investmentas a Share of GDPinHaiti The system of national accounts in Haiti faces several challenges. First, the Haitian Institute of Statistics (IHSI) i s underfinanced and has weak institutional capacity. Second, the availability of data is very limited and outdated. National accounts are established following the old SCN 1968 methodology and the current base year i s 1986/87 (temporarily chosen as reference year to replace the original one of 1975/76, however migration to the temporary reference year was never completed). Currently, the IHSI is trying to apply the 1993 SCN recommendations and to use 199912000as the reference year, but there have been delays in this regard. The shortcomings o f the system of national accounts partly explain the investment enigma: investment in terms of GDP in Haiti seems to be extremely high taking in consideration the country's output performance. Besides the low quality andlor unavailability of data, the enigma can be explained by: Under-estimation of GDP: The base year currently employed (1986187) does not take into account the explosive increase of the informal sector experienced in recent years and the full effect of worker's remittances on the economy. The current methodology to estimate the value added of agriculture and construction takes in account only part of the informality in those sectors. In transport and personal services, the collection data does not cover informal activities. Overestimation of Investment: There are several factors affecting investment estimates, with some of them tending to underestimate the figures and others tending to overestimate them. The new effect i s unclear. Public investment figures are largely overestimated because several current expenditures are recorded as investment (this i s because donor financed programs are recorded as investment regardless of the final use of funds). For private investment, capital imports are largely overestimated because a large amount of consumption and intermediary goods are registered as capital goods-largely due to tax incentives favoring importing capital goods. With respect to the factors underestimating investment, such as smuggling and informality, these tend to underestimate GDP as well; thus, the net effect of these factors on the investment to GDP ratio i s diluted. Source: IHCE. 38. Comparing Haiti's growth performance to the Dominican Republic's, with which it shares some similarities and a common island, serves as a natural experiment where various factors are controlled, particularly the initial conditions and some external shocks (such as adverse natural events). From a starting point of identical levels o f real output per capita in 1960, the two countries have followed divergent growth paths (see Figure 1.5). Haiti suffered long periods of sustained real output contractions while the Dominican Republic experiences two golden decades o f growth (the 1970s and 1990s), 7 becoming one of the fastest growing economies in Latin America and the Caribbean (annual real GDP growth averaged 5 percent between 1970 and 2000). Figure1.5: GDPPer-CapitainHaiti, Figure1.6: GrowthAccounting DominicanRepublic andLAC DecompositioninHaitiandDominican ,. ReDublic 4500 Ip/< .................................. ......................... r ...... ....................................................... ~ 4000 .... _: ... I I I Hail OCapStockGrowth ~HumnCapitaIGrowthmTFPgrowth Source: World Bank, 2006. 39. These sharply diverging economic growth performances cannot be explained by geographic or climate differences, however. Both countries have a similar geographic fragmentation index (Luke et al., 2003), both have been repeatedly exposed to adverse natural events, and neither i s characterized by ethno-linguistic fragmentation. The U S i s the main trading partner for each, in part due to preferential access under the Caribbean Basin Initiative and its successor, the Caribbean Basin Trade Partnership Act. In different degrees, poor economic governance and weak institutions have severely affected both countries by raising the cost of doing business and limiting the impact of growth on poverty reduction.'* Corruption i s prevalent in both countries at the local and national scales. Finally, both countries have suffered from infrastructure shortages, in particular prolonged electricity crises and ineffective remedial measures. What then explains the two countries' diverging growth performances? A more enabling environment for private activities. The Dominican Republic experienced relative political stability, maintained a stable macro environment over prolonged periods and succeeded in attracting FDI (such as through legislation ensuring equal treatment of national and foreign investors) and remittances to help fuel private investment. Like Haiti, the Dominican Republic benefits from large remittances (over US$2 billion annually), but contrary to what happens in Haiti, about thirty percent of remittances is productively invested or saved as a result of the more attractive investment climate. In Haiti, the use of remittances remains overwhelmingly directed to consumption (due to both much higher levels of poverty and a less attractive investment environment). World Bank (2005), DominicanRepublic:PovertyAssessment. 8 0 Differences in productivity. Growth accounting analysis shows that dissimilarities in performance between the two countries are mainly associated with differences in productivity gains (see Figure 1.6). In Haiti, the dramatic income declines in the 1980s and 1990s are related to large TFP decreases, whereas the converse i s true in the Dominican Republic. A more outward orientedgrowth strategy. Long-run GDP growth inthe Dominican Republic has been largely associated with a shift away from traditional exports (sugar, tobacco, coffee, cocoa, and minerals) toward FTZs, tourism, telecommunications and financial services. Haiti tried to follow a similar strategy but failed in diversifying its economy into more dynamic sectors. The incipient industrialization of the early 1970s based on assembly products was interrupted by the 1991-94 embargo. Also, the precipitous liberalization of external trade in the 1990s in the absence o f an investment environment conducive to a supply response contributed to the collapse of production o f some agricultural commodities. Similarly, while the tourism sector in the Dominican Republic emerged as a substantial force, directly accounting for 6-7 percent o f economic output and employment, the development of tourism in Haiti has been erratic and never reached its full potential due to political instability, the negative impact of HIV/AIDS in the early 1980s and the lack of tourism infrastructure. 40. Loayza, Fajnzylber and Calderon (2005) use cross country growth regressions that included several variables for a sample o f LCR countries to evaluate the determinants o f their economic growth. Table 1.2 shows the results for Haiti's growth performance, comparing the 1990s to the 1980s. The observed sharp decline in growth rates in the 1990s as compared to the 1980s contrasts significantly with the improvement that would be projected by model estimates. Indeed, the difference between actual and projected rates i s about 3 percentage points. Structural policies have a large explanatory importance accounting for more than 2 percentage points of the growth difference; they were however offset by the worsening in stabilization policies. A comparison of performance in the 1980s with the 1970s highlights the sizable cyclical-reversion effect, confirming that the expansion of the late 1970s proved to be a transitory phenomenon. External factors also played against growth in the 1980s with respect to the 1970s. 9 Table 1.2: ExplainingChangesinGrowth BetweenDecades (changeinthe average growthrateof real GDP per capita) 1980scompared to 1970s 1990s compared to 1980s Actual change between decades -4.8% -0.6% Change explained by the regression 2.3% Transitional C~nvergence'~ -0.3% 0.5% Cyclical Reversion -1.2% 0.5% Structural Reforms 2.2% Stabilization Policies 0.4% -0.6% External Conditions -1.2% -0.4% Change not explained by the regression -2.9% Note: Transitionalconvergencerefers to initialGDPlcapitaandcyclical reversionto the initial output gap. Structural reforms include secondary enrollment,primary domesticcredit/GDP, structure-adjustedtrade volume/GDP, government consumption, maintelephoneslines per capita, the first principalcomponentof four indicatorsreportedinthe International CountryRisk Guide (prevalenceof law andorder, quality of the bureaucracy, absence of corruption, and accountabilityof public officials); external conditionsincludegrowth rate of terms of trade and periodshifts; and stabilizationpoliciesincludeinflation, standarddeviation of outputgap, index of exchange rate overvaluationandfrequency of years underbankingcrisis. Source: Loayza, FajnzylberandCalderon, 2005. 41. In addition, Haiti's volatility of real per capita output, lower than the LAC and SSA average during the 1970s and 1980s, was instead much higher than the LAC, SSA and world average during the 1990s (see Table 1.3). Haiti was one o f only four L A C countries (with Colombia, Grenada and St. Vincent and the Grenadines) for which volatility increased in the 1990s as long-run growth declined, quite an undesirable combination. Table 1.3: Volatility of Output per Capita (1961-2000) Source: Loayzaet al., 2005; andIanchovichinaandKacker, 2005. 42. Table 1.4 provides a comparison of Haiti's changes in growth with other L A C countries. It shows that a significant share of the change in real per capita growth in several other L A C countries i s also not explained by the traditional growth determinants and that there i s a large effect on growth not captured by the panel regression. Inmost of these countries, however, the effect of the factors not included in the regression on growth has been positive while in Haiti it has been negative-with the exception of Colombia and Jamaica. For Colombia and Haiti, the difference between actual and projected rates exceeded 3 percentage points. It i s doubtful, however, whether these disappointing cases can be used as evidence of policy reform failure. It i s more likely that the inability to explain Colombia and Haiti's growth deterioration resides on the empirical model's failure to consider the negative effects of political instability and violence. The experience of Colombia and Haiti can be regarded as the flipside of what l3 The "conditional convergence" hypothesis maintains that, ceterisparibus, poor countries should grow faster than rich ones becauseof decreasing returns to accumulable factors of production. 10 happened inEl Salvador and Nicaragua in the 1990s. These Central American countries were able to resolve the civil war and political strife that afflicted them in the 1980s and their growth improvement in the 1990s was well above what the model could project. Political instability and violence impairs a country's growth performance in ways that are not captured b y standard determinants, and, conversely, a recovery from them i s bound to be underestimated by them. Another factor not captured explicitly in the growth models but that is present inthe Haitian economy is the proceeds from the drug trade. The large unexplained residual inthe sources of growth analysis may be capturing this effect. Drug trade interacts with weak governance and prudential regulations, fueling consumption and services in the economy but negatively affecting political stability, crime and violence, governance and the rule of law. Because the proceeds from drug trade by their nature are not easily quantified, however, we cannot analyze their direct impact on growth. Table 1.4: ExplainingChangesinGrowthBetweenDecadesinHaiti and Other LAC Countries (changeinthe averagegrowthrate of realGDPpercapita -1990scomparedto1980s) Actualchange -0.59 4.40 4.14 3.80 3.49 3.44 -0.55 0.84 -1.86 ProjectedChange 2.34 1.84 2.09 1.13 2.54 2.42 2.11 0.82 -1.73 Transit. Convergence 0.49 0.67 -0.05 -0.19 0.11 -0.14 -0.32 0.04 -0.30 Cyclical Reversion 0.54 -0.97 -0.10 0.36 -0.02 0.46 0.15 0.25 -0.88 Structural Reforms 2.24 2.56 2.21 1.11 1.34 1.28 1.15 0.71 1.45 Stabilization Policies -0.56 0.18 0.41 0.15 1.70 0.48 1.47 0.16 -1.30 External Conditions -0.37 -0.60 -0.38 -0.31 -0.59 0.34 -0.34 -0.35 -0.70 Other factors -2.93 2.56 2.05 2.67 0.95 1.02 -2.66 0.02 -0.13 Source: Loayza, Fajnzylber andCalderon,2005. 43. Table 1.5 provides a comparison o f Haiti with SSA countries. It shows that a significant share o f the change in real per capita growth in several SSA countries is also not explained by the traditional growth determinants. In most of these countries, the effect of the factors not included in the regression on growth has been negative - as in Haiti. This i s in line with the above assessment since the SSA countries where the difference between actual and projected rates i s largest have also been affected by political instability and violence. 11 Table 1.5: Explaining Changesin Growth Between Decades inHaiti and Other SSA Countries (changeinthe average growth rate of real GDP per capita -1990scompared 1980s) to Actual change -0.59 -6.55 -0.31 -5.73 -1.45 -1.23 3.08 0.94 -1.41 ProjectedChange 2.34 -1.75 0.14 0.52 0.98 0.67 3.77 2.36 -3.00 Transit.Convergence 0.49 0.41 -0.32 0.22 -0.10 -0.12 0.23 0.20 -0.73 Cyclical Reversion 0.54 0.91 -0.45 -0.76 -0.08 -0.60 1.06 0.20 -4.76 StructuralReforms 2.24 0.93 -0.27 0.49 0.47 1.45 0.71 1.43 1.12 StabilizationPolicies -0.56 0.40 0.60 -0.45 0.21 0.02 1.27 0.41 1.17 ExternalConditions -0.37 0.09 0.58 1.02 0.48 -0.08 0.52 0.12 0.20 Other factors -2.93 -4.80 -0.45 -6.25 -2.43 -1.9 -0.69 -1.42 1.59 Source: Loayza, FajnzylberandCalderon, 2005. 44. Loayza, Fajnzylber and Calderon (2005) also use the estimates obtained in their cross-country panel regressions to simulate economic growth for the LAC countries, including Haiti, for the period 2000-2010. To do so, they consider two scenarios: a "continuous trend" scenario where the explanatory variables continue their recent past trends; and an "ambitious" scenario of quick and sharp progress in the conditions that drive growth (with policy determinants o f growth moving to the top 25 percent o f the L A C region). The results for Haiti are presented in Table 1.6. While these are only indicative and have limitations since they cover a period part o f which has already past, they show that under current policies, the average annual growth rate of real GDP per capita would remain negative. Under the reform scenario with policy determinants of growth moving to the top 25 percent of the LAC region and the world, the projected change would likely be significant (6 to 8 percent), with improvements in public infrastructure and education forecast to render the largest contribution to economic growth. As indicated above, these growth simulations may even underestimate the impact of improved political stability and reduction inviolence. 12 Table 1.6: Growth Simulations, 2000-2010 (change inthe average annual growth rate of real GDP per capita) Actual change in 1990s -2.91 0.33 2.67 3.48 1.53 3.75 0.72 0.12 -0.62 Projected Change- continuoustrends scenario -0.59 2.84 3.31 4.05 1.58 0.54 1.73 0.93 2.11 Projected Change- reform scenario (LAC) 6.02 5.06 1.86 1.46 3.98 -0.16 1.67 3.06 5.67 Structural Reforms Education 1.67 0.60 1.08 0.64 0.52 0.49 0.11 1.32 0.00 PublicInfrastructure 1.77 1.84 0.84 0.00 0.72 0.00 0.00 1.35 2.16 Potential Change- reform scenario (LAC) 7.84 6.45 3.65 2.80 5.31 1.57 3.32 4.54 7.04 StructuralReforms Education 2.05 0.99 1.46 1.03 0.90 0.88 0.50 1.71 0.35 PublicInfrastructure 2.37 2.44 1.44 0.40 1.32 0.59 0.60 1.95 2.76 Source: Loayza, FajnzylberandCalderon, 2005. 111. The BindingConstraintsto GrowthinHaiti 45. Inmany ways, Haiti has more incommon with the poorest African countries than it does with its neighbors in Latin America. It is one of the extreme cases of a country caught in a trap of persistent conflict, low growth, a state that suffers from great fragility and institutions that are incapable of delivering basic services to citizens. Within this context, the challenge i s to identify and then implement actions to defeat the legacy o f past decades and set out on a path o f sustained growth and state building. Almost everything needs to improve in Haiti; obstacles to growth are widespread and severe, and they are interrelated andfeeding off each other. This section will focus on identifying the critical constraints to economic growth in Haiti. The country's growth experience in the 1970s and, to a lesser extent, in the second half of the 1990s, shows that it is capable of experiencing rapid growth; the challenge for the country i s then how to retake a higher growth path and to ensure its sustainability. 46. Figure 1.7 presents the critical constraints to doing business in Haiti as identified by entrepreneurs operating in the country. Data is provided by the World Economic Report o f 2003-2004 whose global Competitiveness Index (GCI) ranked Haiti last among 102 countries across the world. Results show important points: tax rates and tax regulations do not seem to be critical constraints--typically this i s what i s most relevant in other countries--nor do labor regulations. Incontrast, inadequate infrastructure, access to 13 finance, political instability and inefficient bureaucracy, and an unskilled workforce emerge as the critical constraints. The rest of this chapter assesses how these factors constrain Haiti's growth. 47. Political instability is a major constraint to growth in Haiti. It weakensprivate Figure1.7: MostProblematicfactors for sector confidence and the quality of DoingBusiness According to Entrepreneurs public institutions. Today, the major OperatinginHaiti constraint to growth in Haiti i s investors' hadequateintrastMture uncertainty as to whether they can Access to ltnming obtain returns from their investment. averment i~tabihycoups heflklent bureatcracy This uncertainty stems mainly from Inadequat&ydtcatedwMorce political instability. Table 1.7 shows Comption that Haiti has had 15 administrations in hflatlon ForelmctrrercyrwatloM the last 20 years that lasted on average Polkylnstabillty less than a year and a half each. Such Cnmeardthefl Restmtivelatarrwations political instability has been Taxregulation accompanied by violence, corruption Taxrates and a continuous weakening of state Poorwrkethc o 2 4 8 8 a n s m 20 institutions, the rule o f law, the Source: GlobalCompetitivenessReport, 2003-2004 investment climate and macroeconomic Note From a list of 14 factors, respondents were asked to policies. All these factors severely limit select the five most problematic for doing business in the property rights, thus affecting the country and to rank them between l(most problematic) and 5. The bars in the figure show the responses weighted appropriability expectations of according to their rankings. entrepreneurs. Political instability has also resulted in a high volatile pattern o f external aidflows which has affectedpublic investments levels and sustainability and in a trade embargo inthe first half of the 1990s that crippled private sector activities. Table 1.7: Haiti's Governments, 1986-2005 Adm. Position and Name Periods served Time served Average per capita real GDP growth 1 President Henri Namphy [military gov.] 211986to 211988 2 years -2.0% (1986-1988) 2 President Leslie F.Manigat 211988 to 611988 4 months -1.5% (1988) 3 President Henri Namphy [military gov.] 611988 to 911988 3 months -1.5% (1988) 4 President Proper Avril [military gov.] 911988 to 411990 20 months -1.4% (1988-1990) 5 President Herard Abraham [military gov.] 411990 3 days -2.1% (1990) 6 President Ertha Pascal Trouillo 411990 to 211991 10months 0.1% (1990-1991) 7 President Jean-Bertand Aristide 211991to 911991 7 month inHaiti 2.3% (1991) 8 President JosephC. Nerette 1011991to 511992 7 months -6.3% (1991-1992) 9 No de-facto president 10 President Emile Jonassaint 611994to 911994 3 months -10.2% (1994) 11 President Jean-Bertand Aristide 1011994to 211996 16 months -4.9% (1994-1996) 12 President Rene Preval 211996to 212001 5 years -0.2% (1996-2001) 13 President Jean-Bertand Asistide 212001to 212004 3 years -3.1% (2001-2004) 14 President Boniface Alexandre 212004 to 512006 2 years -5.5% (2004) 15 President Rene Preval 512006 to date Source: Based on World Bank (1998) and media. 14 I S police is very weak: it i s being restructured but only 6,000 policemen cover the entire country. The problem i s aggravated by the weakness of the judicial system, with Haiti ranking last and second to last in the world in the ranking of judicial independence and efficiency o f the legal framework in the 2003-2004 CGR--although Haiti ranks better in terms of procedures, costs and time needed to enforce contracts (see Annex 1.2). Additionally, there are legal limitations for foreigners, including the fact that land and assets cannot be inherited, seized assets of foreign private banks must be sold in a short period (two years), and foreign private entrepreneurs are required to operate with a national partner, who, b y law, i s given a majority share o f the decision power within the firm. 51. Haiti's institutional weaknesses have also led to high levels o f red tape that are usually tied to widespread corruption: Haiti ranks 155th out 159 countries in the corruption perception index according to Transparency International (2005). According to private entrepreneurs, the required administrative procedures are not clear and the public official in charge has "veto" power on their completion. Private entrepreneurs also report that in some key public institutions (such as port administration, customs etc.) quasi-unions (groups of employees) arbitrarily control the make-or-break decisions related to private sector activities. Insuch a context, a significant constraint is the time it takes enterprises to complete needed administrative procedures. For example, in order to export a product from Haiti, 58 days are needed (compared to the 17 days needed in the Dominican Republic). Red tape problems do not stop here: they include the significant cost and the minimum capital required to start an enterprise. The minimum capital needed to start a company in Haitii s three times the country's per capita income, and an equivalent amount i s needed for completing the required administrative procedures (see Annex 1.2). Inthe Dominican Republic, on the other hand, the minimumcapital required i s only 1.2 percent of the per capita income and red tape costs only amount to 31 percent of the per capita income. These costs are particularly severe for small and micro businesses and encourage the growing informalization of the Haitian economy. In addition, it should be noted that the U S Congressional approval of the HOPE Act would provide a significant boost to Haiti's manufacturing sector. However, trade gains could be limitedby a weak institutional framework and investment climate. 52. Infrastructure limitations are identified by entrepreneurs operating in Haiti as the most problematicfactor for doing business in the country. Figure 1.9 shows that Haiti ranks last in the GCR indicators of infrastructure quality both in L A C and in the world. Particularly noteworthy are the poor performance o f the sea port and road infrastructure (which are critical for domestic and international trade) and o f electricity. Haiti's inadequate infrastructure clearly has a huge impact on the productivity of firms, and on the simple ability o f rural and urban households to generate income. 16 Figure 1.9: International Comparison of Key - Figure 1.10: International Comparisonof - Indicators on Infrastructure Cost of Electricity Meaco I Overall infraestructure quality Haiti I I Port infraestructure quality -- Guatemala I Honduras Railroad infraestructure deveiopment I _ . Nicaragua I Air transport infraestructure quality DominicanR... I - Costa Rlca I Quality of electricity supply L El Salvador Telephone lnfraestructure quality I Colombia -LAC -World 0 Haiti China 0 20 40 60 60 0 5 i o 15 20 Global Competitiveness Rank +,loo U.S. dollar cent per KWh Source: GCR (2003-2004). Source: IMF(2005): 53. The crisis i s especially acute in the electricity sector where only about 10percent of the population i s connected legally to the grid (mainly in the capital city) and electricity i s provided for only a few hours per day. Black-outs and rationing have forced entrepreneurs to have recourse to private (and costly) electrification systems; Figure 1-10 shows that the cost of electricity for the industrial sector i s relatively high (IMF, 2005). Electricity provision i s also imposing a huge fiscal burden, with public subsidies in 2005 amounting to US$47 million, or 7 percent o f the total budget. The lack of a reliable electricity supply curbs the possibilities for economic activity of all but the few who can afford their own generators, and imposes additional costs for those who can. Indeed, the electricity crisis represents a key obstacle to the country's short-term prospects for economy recovery, fiscal stabilization and long-term competitiveness. 54. The poor state of the sea port infrastructure is also a critical constraint. Ports are important in Haiti, given its geography (being an island) and its dependency on food imports and external assistance. Inaddition, ports are significant fiscally for the customs revenues collected. Port-au-Prince i s the most expensive port in the Caribbean (Lundanhl, 2004), to the point that some Haitian importers prefer to use ports in the Dominican Republic and transport their merchandise by road to the city of Port-au- Prince, thus increasing import prices. The poor state of port infrastructure also affects the competitiveness of Haiti's apparel industry. In addition to the dilapidated state of port infrastructure and inefficient operations, ports in Haiti (notably the one in Port-au-Prince) suffer from insecurity, corruption and governance issues. The road network in Haiti i s only a third of the one of the Dominican Republic. The problem i s exacerbated b y the lack of maintenance and the deterioration during the rainy season that cuts off entire agricultural regions; currently, only 5 percent of roads in Haiti are in good condition. The deteriorated road network impedes access to markets, especially in rural areas, and thus increases food prices; and prevents a more decentralized distribution of economic activities away from the capital city and the development of backward linkages with the rural economy. 17 55. Much needed improvements in human capital need to be accompanied by the lifting of other restrictions to growth. The average years of schooling in Haiti are among the lowest in the region. Only 55 percent of children aged 6-12 are enrolled in school; in ruralareas this indicator plunges to 23 percent. Children also tend to drop out of school very early. In addition to this, the quality of the education system in Haiti has been assessed as among the worst in the world (see Figure 1.11). Although illiteracy among the adult population decreased from 60 percent in 1990 to 48 percent in 2003, it remains the highest in the LAC region and is higher than the average for low-income countries. The average adult illiteracy rate in LAC and LIC countries i s 10 percent and 39 percent respectively (UNDP, 2005). The inability of Haiti's education system to provide an adequate education to future Haitian generationsjeopardizes medium-term growth, unless urgentmeasures to reversethis trendare taken. 56. Haiti is the developing country with the highest emigration rate of educated people. Orozco (2006) reports that 88 percent of the Haitians who emigrate do so because they cannot find work opportunities in Haiti. Indeed, the migration of skilled people (caused in the first place by political instability and the poor investment climate) has itself created a vicious circle through which the country has lost so many agglomeration externalities that attraction of productive investment (including FDI) and total factor productivity growth have become seriously constrained, fueling in turn continued migration of skilled people. In such a context, much neededimprovements in humancapital need to be accompaniedby the lifting of other restrictions to growth, least a swift improvement in education would only result in increased brain drain with a limited effect on growth. In turn, with an improvement in the investment climate, the potential contribution of the Haitian diaspora could go beyondjust sendingremittances- it could contribute to expanding foreign trade, capital and technology flows and to the important task of institutionbuilding. Figure 1.11: International Comparison of Figure 1.12: EmigrationRate of top 20 Key Indicators on Education Skilled Emigration Countries, 2000 (percentageif total educated labor) Dominican Tertiary enrollment ! Afganistan I Guatemala Quality of education system i Honduras Rwanda I1 I Vietnam Papua New Quality of public schools Cuba I Hong kong Nicaragua Sri Lanka El Salvador Somalia Angola Uganda LaoKenya PDR Mozambique Ghana -LAC-World Haiti I Hail! 0 20 40 60 0% 20% 40% 60% 80% 100% GbbalConpettiiveness Rank % of totaleducatedlabor Source: GCR (2003-2004). Note. It only includes countries with population above 5 million. Source: Docquier, F. and Marfouk, A. (2006). 18 57. Macroeconomic instability is currently not a binding constraint to growth, but it could easily become one due to high vulnerabilities. Inthe past, fiscal, monetary and financial policies in Haiti have often contributed to an unstable macroeconomic environment and financial and balance-of-payments crises. Also, high levels of inflation and sharp exchange rate fluctuations have had a significant negative impact on the poorest. By increasing uncertainty, they discouraged firm investment, increased societal disputes for the distribution of ex-post rents (for instance between firm owners and employees in the face of high unexpected inflation) and led economic agents to concentrate on managing high risks rather than productive activities. Macroeconomic instability has often been linked to political instability, due to the country's high dependence on external aid. While economic conditions deteriorated significantly during the early 2000s, since early 2004 the Transition Government has implemented macroeconomic stabilization policies that regained the support o f the international community. These macro policies were framed in two IMF emergency programs and the macro economic performance has remained on track since. Financial stability has been restored, inflation has declined and the exchange rate has stabilized. Fiscal accounts show some recovery and the external current account deficit has narrowed (see Table 1.8). Maintaining macroeconomic stability will be critical for economic recovery and maintaining a sustained growth path and should thus be the top priority for policy makers. Table 1.8: Selected Macro Economic Indicators (2000-2005) 2000 2001 2002 2003 2004 2005 (est.) CPI inflation (end of period) 15.3 12.3 10.1 42.5 22.5 15 Central Government overall balance (incl. -2.2 -2.4 -3.0 -3.5 -2.4 -0.6 grants) Total public debt (end of period) 31.3 35.8 38.7 47.8 39.5 34.6 External current account (incl. grants) -1.0 -2.0 -1.0 -0.1 0.4 0.5 Exchangerate (Gourdesper Dollar - end of 28.3 25.5 29.7 42.0 36.8 period) Liquid gross reserves (month of imports) 2.1 2.2 1.5 1.2 1.3 1.3 Source: IMF.2005. 58. Financing costs in Haiti are high and access to credit limited; however, the major problem is not one offinancing constraints, but of high levels of risks and low private returns to investment. As shown in Figure 1.13, Haiti i s ranked relatively well, 65th out o f 102, in terms of availability of national savings. This results from highlevels of workers' remittances. As mentioned above, these have continued to increase and have reached record levels by international standards (see Figure 1.14). Furthermore, returns on capital are low: real interest rates on deposits are positive and low by regional standard^.'^ Bank deposits have been increasing, doubling between 2001 and 2005, and numerous poor and middle class Haitians save in the formal banking sector. During the last five years public savings have been close to zero. Also, the current account o f the balance o f payments (including remittances) has registered only small deficits. l4 Dollar denominatedtime and saving interestrates are below 3 percentin the case of time depositsand below 2 percentin the case of savingdeposits. 19 Figure 1.13: International Comparison Figure 1.14: Top 20 Developing-Country of Key Indicators on Saving and Financing Recipients of Remittances, 2003 (percentage of GDP) Cape Verde Tajikistan Yemen, Kiribati Honduras Samoa Labanon Albania Dominican El Salvador Regulatory Quality - Nepal I - 1 Jamaica Bostia and Vanuatu West Bank Jordan Moldova Losotho Haiti -Caribbean -World Haiti Tonga 0 20 40 60 80 1 0% 10% 20% 30% 40% 50% Countws Percentile Rank (0-100) +- * % of GDP Source: GCR (2003-2004) Source: Ratha(2005) 59. Financial intermediation in Haiti is incipient and the formal banking sector has several shortcomings. The concentration of the banking system is at a record high in every possible dimension. In 2005, the first three banks represented 62 percent of the total assets of the financial system, captured 66 percent of total deposits and 70 percent of dollar deposits, and provided 61 percent of the total banking credit. About 68 percent of the total lending portfolio i s provided to 9 percent of the largest borrowers (see Figure 1.15). There i s also a high concentration of credit according to economic activity; bank credit i s directed primarily to consumption and services which absorb half of the total credits provided. Key sectors such as agriculture and transport literally receive no lending (see Figure 1.16)15. Credits are primarily short term; long-term credit i s not available. Figure 1.15: Concentration of LoanPortfolio Figure 1.16: Credit by Activity (% of total loan portfolio) 50% 45% 40% 35% 30% 25% 20% 5 I 15% ' 10% II 5% 0% 0% 5% 10% 15% 20% 25% 30% 35% Loans by size (in th of Oordeo)) %of 101d bans porlfollo Source: BRH (2004-2005) Source: BRH (2004-2005) l5The explicit prohibitioninHaitianlaw to use rural land as collateral to credit has made agricultural activity unattractive for credit granting. 20 60. Despite available savings, access to credit is very limited. In 2005, less than one percent (0.16 percent) of the adult population had access to bank credit. The severe credit access problem did not improve despite advances o f microfinance (see Box 1.2). Bank mortgage credits are almost inexistent and the real estate market i s extremely distorted. The bulk of real estate construction i s financed by private savings. Bank spreads are large, reflecting high levels o f uncertaintyhisk and high operative costs and the existence of market power. Bank spreads in Haiti are double those of other countries in the region, such as the Dominican Republic, Guatemala and Honduras (see Figure 1.17), this despite financial costs being low due to the fact that NPLs levels are low (around 10 percent). Most entrepreneurs rely on financing by local banks or their own sources, and only a minority of entrepreneurs can obtain loans abroad, and lending rates are high. Figure 1.17: Bank Spreads 61. Banks are liquid and bank Halt1 average ROES (returns on equity) are Costa Rica good, at around 20 percent; this Dornlnlcan R.. indicates that ex-post returns are high. Duafernala However, banks are extremely Honduras Nicaragua conservative in providing credit: Mexlco i investing in new potential borrowers is Chlna. rnalnland too risky given the high risks facing the El Salvador country. This risk-averse behavior is 0% 5 % 10% 15% 2 0 % 26% Source: IMF(2006) reinforced by the belief by banks that, in the case o f problems in the financial sector, the Government would not intervene to avoid a crisis because of lack of resources; the bankruptcy of the cooperative sector in 2002 i s a sharp reminder of this. These findings indicate that the major problem is likely to be not one of financing constraints, but of highlevels of risks andlow private returns to investment. Box 1.2: MicrofinanceinHaiti Since the mid 1990s, the provision of micro credit in Haiti has experienced a rapid expansion. Between 1994 and 1997, the number of micro lending institutions in Port-au-Prince expanded from four small NGOs to over thirty offices serving between 15,000 and 30,000 borrowers. B y September 2002, more than 60,000 micro borrowers in Haiti were served by 80 lenders (58 credit unions, 18 NGOs, associations and religious groups and four commercial banks). The fastest growing segment was that of the commercial banks, which at the time had 20 percent of the active clients and 30 percent of the total credit portfolio. Despite this rapid surge of microfinance institutions, less than 15 percent of poor households have access to microfinance; this deficit is more evident in the rural areas, where only 54,000 poor households benefits from a microfinance institution service. Two main obstacles are preventing the further development o f microfinance in Haiti: (i) political uncertainty.Microfinance clients live from day to day, and if their sale operations are hampered by strikes and other disturbances, they immediately get into repayment problems; and (ii)lack of appropriate collateral. Providing credit financing is risky because the majority of the borrowers do not have appropriate collateral ready to be usedas a bank guarantee. 62. The tax burden is not a constraint toprivate sector activity in Haiti, as taxes are low. On the contrary, revenue collection needs to be strengthened to mobilize public 21 resources for much needed social and infrastructure investments. Doing Business (2006) shows that the fiscal burden in Haiti i s equivalent to 32 percent of firms' gross profits, lower that the Latin American average of 53 percent and of the Sub-Saharan African average of 58 percent. Inthe GCR 2003-2004, taxes were pinpointed as one of the least relevant problems for entrepreneurs in Haiti, while they are indeed seen as one of the most problematic obstacles to business in other countries. 63. Tax revenues in Haiti are much lower than the levels observed in other low- income countries because of the country's narrow tax base and weak fiscal administration (IMF, 2005). Table 1.9 shows that the VAT tax rate is one of the lowest in the region. Additionally, the effectiveness of this taxation i s low - with tax rates similar to Haiti, Guatemala and Paraguay obtain a larger collection percentages; the Dominican Republic i s able to collect more, and with a lower rate. The Corporate Income Tax (CIT) rate i s close to regional standards but revenues from the CIT represent only 0.85 percent o f GDP, much lower than the 2 percent collection in low-income countries. Numerous fiscal exemptions and tax holidays are inplace -Haiti's investment code grants fiscal and customs' holidays of between 15 and 20 years. Moreover, Haiti grants VAT tax holidays, against international best practices. 64. I n the same vein, labor costs are low and labor legislation is flexible, thus Table 1.9: VAT in Selected Countries they do not represent a constraint to VAT growth. Haiti ranks 38th out of 151 Collection countries on the labor flexibility index Tax Rate to GDP CIT Rate reported by Doing Business 2006. Table Argentina 21.0 6.2 35 1.10 shows that difficulty and cost Bolivia 14.9 5.7 25 indicators related to the hiring and firing Brazil 20.5 8.6 34 of personnel are lower than the L A C and Chile 18.0 8.5 17 Colombia 15.0 4.9 35 SSA averages (see also Annex 1.2). As an CostaRica 15.0 6.5 30 example, to fire a person employed in the Dominican Rep. 8.0 3.1 25 formal sector in Haiti, the equivalent of a Ecuador 12.0 4.4 36.25 26 weeks salary needs to be paid in ElSalvador 13.0 5.3 25 compensation -- in Ecuador, Guatemala or Guatemala 10.0 3.7 31 Haiti 10.0 2.2 35 Bolivia, it i s 131, 100 and 96 weeks Honduras 12.0 3.8 25 respectively. Similarly, Haiti ranks 12th Mexico 15.0 3.2 33 and 18th among the 102 countries Nicaragua 15.0 10.0 considered in the Global Competitiveness Panama 5.O 2.0 30 Report 2003-2004 on hiring and firing Paraguay 10.0 4.5 30 Peru 18.0 6.3 30 practices and flexibility of wage Uruguay 23.0 8.4 35 determination respectively, obtaining a Venezuela 15.0 3.2 34 better position than countries like Brazil, Average (LAC) 14.3 5.3 Costa Rica, Mexico, the Dominican Source: Keen, et al. (2001), KPMG (2004) and IMF (2005) Republican and Thailand. Labor costs in Haiti are relatively low; as an illustration, labor costs in the assembling industry are the lowest in Latin America and the Caribbean, and comparable to those of China, Bangladesh and India (see Figure 1.18). 22 Table 1.10: Summary Indicators on Hiring & Figure 1.18: Assembly Sector Wage Costs FiringWorkers (inUS$ per hour) hdia Indicator Haiti LAC SSA Bangladesh Difficulty of Hiring Index 11 40.5 48.1 Hati (0-100) china (inland) Rigiditvof HoursIndex China (coast) co--i ooj 40 50.9 63.2 Nicaragua Difficulty of FiringIndex 20 29.5 47.8 Colorkia Honduras (0-100) Guatemala Rigidity of Employment 24 40.3 53.1 0 Satvador Index (0-100) DOminkanRapublk Hiringcost 9 15.9 11.8 Mexico (% of salary) Costa Rlca Firingcosts 25.6 62.9 53.4 0 1 2 3 (weeksof wages) Source: DoingBusiness2006 Source: IMF(2006) 65. Despite low labor costs, non-restrictive labor regulations and a relatively low tax burden, the majority of Haitian enterprises operate outside o f the formal economy and the level of unemployrnentl6 remains among the highest in the world. Clearly, political instability, excessive red tape and regulation, and poor rule o f law have discouraged the formalization of businesses, thus limiting the growth o f demand for formal labor. They are also prohibitive for the majority of the population and tend to act as a disincentive to small business initiatives, relegating large portions o f the population to the informal sector. Sixty-eight percent of Haitian enterprises operate outside of the legal framework and 99 percent are qualified as micro or small enterprises (CPED and ILD, 2001). Around 90 percent of the formal employment i s concentrated in Port-au-Prince and over halfof itis made upof the public sector andpublic enterprises. 66. Impact of natural disasters on growth. As indicated above, Haiti's growth prospects have been affected by the recurrence of adverse natural events. Haiti lies inthe middle of the hurricane belt and i s subject to severe storms from June to October. Their effects (flooding, landslides, wind damage and coastal surges) are exacerbated by environmental degradation, since only about 3.2 percent of Haiti's natural forests remains. Extensive deforestation and the presence of settlements in low areas and in floodplains of rivers are key factors in the intensity of the destruction. The most recent serious disasters in Haiti occurred in May 2004 and in September 2004 which, combined, affected over 400,000 people, destroyed 6,500 dwellings and claimed more than 4,000 deaths. Given the high levels o f poverty, the poor state o f infrastructure and lack of preparedness, disasters of lesser magnitude occurring on a frequent basis also cause immeasurable impacts on the population and on economic assets. However, a country's exposure to adverse natural events does not need to inevitably be a limiting constraint to growth. Indeed, the experience of the Dominican Republic shows that, with similarly l6Unemploymentis especially highin urbanareas (49 percentand 37 percentin metropolitan and other urbanareas respectively). 23 high risks of natural disasters, it is possible to secure long episodes of growth. This points to the importance of strengthening Haiti's institutional capacity for preventing and mitigating natural disasters. 67. The potential of the Free Trade Zones. As previous history suggests, the Free Trade Zones (FTZs), notably apparel assembly, have significant growth and employment generation potential in Haiti in light of the country's wage competitiveness, the proximity to the US and the Dominican Republic, and changes in international trade policy such as DR-CAFTA and the Multi-Fibre Agreement phase 0 ~ t . l In ~ particular, the WTO- mandated elimination of export promotion policies for the zonas francas in the Dominican Republic by 2010 (which would not affect Haiti because of its high poverty level) offers the potential to increase Haiti's share of the U S apparel market and prospects for co-production arrangements in which Dominican producers use Haitian labor for part of their production process. Co-production need not be limited to apparel assembly, as there may be other profitable export activities in which Haiti can become competitive within the new trade environment. To benefit from these opportunities, however, critical constraints to the expansion of FTZs in Haiti will need to be tackled, notably by addressing infrastructure bottlenecks, providing an adequate and stable investment regime, providing specific skills training, and promoting harmonious labor relations. A World Bank study exploring these opportunities and challenges i s planned for 2006. l7Fromthe Dominicanperspective, the zonas francas (which are highly concentratedin apparel production) are facing increasedcompetitionas a result of the Multi-Fibre Agreement phase out and increasedoverseas competition,pushingproducersto look towards new productsand new production opportunities (including co-productionarrangements) to remain competitive. 24 Annex 1.1: Growth Decomposition Real Output Decomposition by Sector (in percent) GDP Agriculture Manufacturing Services Average Growth 1965-1970 0.2 -1.6 1.4 1.1 1970-1975 2.9 2.5 6.8 1.4 1975-1980 4.8 1.5 10.6 3.7 1980-1985 -2.0 -1.1 -2.6 -2.2 1985-1990 1.6 0.9 -1.3 4.2 1990-1995 -3.8 -2.3 -8.0 -2.7 1995-2000 3.O -1.5 5.6 4.6 2000-2004 -1.2 -2.0 -0.8 -1.0 Average Participation 1965-1970 100 38 18 44 1970-1975 100 35 21 44 1975-1980 100 32 27 41 1980-1985 100 31 30 40 1985-1990 100 31 27 42 1990-1995 100 31 22 47 1995-2000 100 29 21 49 2000-2004 100 26 23 51 Source: The World Bank and IHCE. Real Growth Decomposition by Expenditure (in percent) GDP Consumption Investment Resource Balance Average Real Growth 1965-1970 0.2 -0.4 11.7 -11.5 1970-1975 2.9 1.7 14.3 -18.9 1975-1980 5.7 5.6 10.0 -241.5 1980-1985 -1.0 -0.6 2.9 26.8 1985-1990 0.2 2.0 -4.8 5.7 1990-1995 -3.8 2.1 0.1 29.2 1995-2000 3.0 8.4 11.3 18.4 2000-2004 -1.2 -1.2 0.3 -0.7 Average participation 1965-1970 100 90 6 4 1970-1975 100 87 12 2 1975-1980 100 85 19 -3 1980-1985 100 83 21 -4 1985-1990 100 86 21 -7 1990-1995 100 105 17 -22 1995-2000 100 138 26 -64 2000-2004 100 159 34 -93 Source: The World BankandIHCE 25 Annex 1.2: Rankingof Haiti inDoingBusinessIndicators Starting a Business PmCedunS I . hi,aliO" I cost + -4 Min Capld I Dealing with Licenses PmadUIe~ I i nme r I cost J Hiring & Firing Workers h?kW01Hinmhdsr Rigidlly d noun Index L-. -- ____ hmully d FlnngIndex r - Rlaldlly d Empbymenl Index ninng COP1 Finw CWP b Registering Property PmCedYDS L I Tlms I .I COS1 I I Getting Credit LwtRlghls Index 1 Credit 1"lCmatb"index 1 PUMO rw1stniooL8(age I 1 P"ie1ll bvrsa" cozrsga , Protecting Investors - D i i ~ b ~ uhdex r e I 1 hreolor LldbIlSY indm J Shuehddsr Suits Index hzslor Pld8ellan lndsx I 1 Paying Taxes Plymenl I nmll I Told lax payable I I I Trading Across Borders ~ U m B ( I 1 0 l o l B X w n t - S,Q"iU"IBS In exmn nme lor expm I W I . DOCYrnenlllor lmwn I - e I - I sgnaturss lor ,mmn W 1 rime br , m m ____--- i Enforcing Contracts PmOedUDl I - I Tlme L I COIl I I Closing a Business nme I COS1 t I RaOOut" I 1 8 1 0 20 40 60 80 100 120 140 Doing Business Ranking Haiti -Latin America and Caribian -World Source: WorldBank,DoingBusiness (2006) 26 CHAPTER2 POVERTYAND INEQUALITY 68. Haiti is the poorest country in Latin American and the Caribbean and among the poorest in the world, with more than half of the population (54 percent) living below US$l-a-day and 78percent living below the USS2-a-day line in 2001. The high rate of poverty in Haiti is manifested in poor access to infrastructure, minimal ownership of assets, low education and literacy levels, and high levels of child malnutrition. Haiti's pattern of socio-economic development has also been characterized by marked inequalities in access to productive assets and public services. Indeed, little variation is seen in the quality of living conditions across the poorest 80 percent of the population, while the richest quintile is markedly better along all dimensions. The best analysis possible with the existing impeij5ect data suggests thatfor the country as a whole poverty and inequality rates have not changed substantially since 1986. It appears likely that many Haitians have avoided falling into complete destitution by receiving remittances from abroad. Thepoverty profile outlined in this chapter, along with the constraints to growth discussed in the first chapter, point to the need to ensure that the benefits of growth are broadly shared through polices which, on the one hand, provide equal access to basic services for the poorest regions and households and, on the other, build institutions that deliver these services in a responsive, efective and accountable manner. I.APovertyProfileforHaitiin2001 Poverty 69. Overall, 54 percent of Haitians live below the US$1-a-day poverty line. There i s a sharp geographic contrast inpoverty between Port-au-Prince and the rest of the country. Poverty rates are highest in urban areas other than the capital (66 percent), followed closely by rural areas. The poverty gap and poverty severity indices are alternative poverty measures which, unlike the headcount, reflect the depth of poverty. Both the poverty gap and severity increase if the incomes of those living below the poverty rate fall, and the poverty severity index attaches extra weight to the poverty of those with the lowest incomes. The poverty gap and poverty severity indices show generally similar patterns to the headcount index. Figure 2.1 shows graphically poverty rates by all three measures. The same figure also shows poverty rates calculated with the higher US$2-a- day poverty line. By this higher poverty line, 78 percent of Haitians are poor. Box 2.1: HouseholdSurvey Data inHaiti A number of household surveys have been collected over time in Haiti. These include the Enqubte Budget et Consommation des Mtnages (EBCM I)collected in 1974, the ECBM I1collected in 1986, the EBCM I11 collected in 1999/2000, and the Enqubte sue le Conditions de Vie en Haiti (ECVH) carried out in 2001. Additionally, Demographic and Health Surveys (DHS) without consumption, expenditure, or income data were collected in 1994/95 and 2000. Data collection for a third DHS survey, which includes question on household expenditure, will be completed in 2006. The poverty profile presented here draws chiefly from the 2001 ECVH and the 2000 DHS. Data from all the surveys listed above are also used to examine changes inpoverty and inequality over time. 27 70. The fraction in poverty accords very closely with results from the socioeconomic self-categorization of a 2003 qualitative survey (Manigat, Lamaute-Brisson and Coulombe, 2003). Respondents self-classified themselves in 78.8 percent of cases as poor using the Haitian Creole word "malere," with an additional group of 1.8 percent classified as destitute, i.e. "miskreux" or "pbv". Similarly, in the 2001 ECVH survey, 85 percent of Haitians lived in households where respondents agreed with the statement, "We struggled with the strict minimumrequired to make ends meet." Thus, by both the US$2-a-day poverty line and two types of self-classification, roughly 80 percent o f Haitians are poor. In the qualitative survey, a mere 2.9 percent classified themselves as rich, and 17 percent characterized themselves as neither rich nor poor. Box 2.2: The Face of Poverty inHaiti Ina2003 qualitative study, a focus group inDesdunes described the poor as follows: One doesn't have to do more than look at their dilapidated houses, their state and their physical appearance. They are thin, sick. They have clothing that's dirty and in a bad state. One doesn't have to do more than look at theirface: the white mouth, the weak look because they are hungry and they have many children in the house insteadof in school. Source: Manigat, S., N.Lamaute-Brisson, et al. (2003). 71. Haiti has by far the highest poverty rate in Latin American and the Caribbean and one of the very highest in the world. Figure 2.2 shows poverty rates for countries categorized by the World Bank as "low income." Among countries for which recent estimates are available, only Burundi, Madagascar, Zambia, and Nigeria have higher poverty rates. Figure 2.1: Headcount Poverty Rates inHaiti Using US$1-a-Day Poverty Line Using US$2-a-Day Poverty Line 80%1 1 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Headcount PovertyGap Poverty Severity Headcount Poverty Gap PovertySeverity (FGTO) IFGT1) (FGTP) IFGTO) (FGT1) (FGT2) 10NationalPort-au-Prince0 Other Urban 0 Rural1 10National Port-au-Prince 0 Other Urban 0 Rural] Source: World Bank calculations based on the 2001 ECVH. 28 Figure 2.2: HeadcountPoverty Rates for Low Income Countries Worldwide Using US$1-a- Day Poverty Line Kyrgyz Republic Vietnam 1998 Tajikistan 2003 Cote d'iwire 2002 Yemen, Rep. 1998 Pakistan 2001 Cameroon 2001 Uzbekistan 2000 Moldom 2001 Ethiopia 2000 Mauritania 2000 Lao PDR 1998 Mongolia 1998 India2000 Bangladesh 2000 Malawi 1998 Ghana 1999 Burkina Faso 1998 Rwanda2000 Haiti 2001 Burundi 1998 Madagascar 2001 Zambia 1998 Nigeria2003 0% 10% 20% 30% 40% 50% 60% 70% 80% Source: World Development Indicators, 2005. Countries shown are those classified by the World Bank as "low income" that have at least one poverty estimate for the period 1998-2004. Poverty figures shown are from the most recent year available. 72. The stark difference inpoverty between Port-au-Prince and the rest of the country i s reflected in the regional poverty rates. Poverty (measured in terms of US$1-a-day) i s lowest inthe Ouest region (34 percent), which includes the capital of Port-au-Prince, and highest in the Nord-Est region (81 percent). Nonetheless, even in the Ouest region, poverty i s extremely high by international standards. The US$1-a-day poverty rate for the Ouest region, the richest inHaiti, is higher than that of any country in Latin America and the Caribbean. Figure 2.3 shows regional poverty rates and also the fraction of the nation's poor living in eachregion. Becausethe population as a whole i s concentratedin the Ouest and Artibonite, the poor population i s as well. Those two regions account for just over half of Haiti's poor. 73. Poverty rates andthe fraction of the national poor in various subgroups are shown in Annex Table A2. The table shows that the population is approximately evenly split between households headed by women and those headed by men, and poverty rates do not differ by the gender of the household head. Poverty rates also decline markedly with the education of the household head. Finally, there is a strong correlation between sector of household head's activity and poverty. Agricultural households have the highest poverty rates-89 percent living below US$2-a-day-while poverty rates are lowest amonghouseholds with heads working inthe public administration sector. 29 re are leiti's Poor' eruentage of the N I Poor Wing in Haitians survived on far less. Figure 2.5 shows income shares and means of per capita income by decile. Inequality in Haiti i s among the highest in the region, which in turn has higher levels o f inequality than any other area o f the world. Figure2.5: IncomeDistributioninHaiti Shares of National Income in Haiti Mean Per Capita Monthly Income By Income Decile By Income Decile - t 3o00, 2737 0 1 2 3 4 5 6 7 8 9 1 0 M e a n hcwne Decile D1 D2 03 D4 LE LX 07 DE W D10 Source: SEDLAC (CEDLAS andThe WorldBank)basedon microdatafrom the 2001ECVH). Vulnerability 76. The term "vulnerability" i s used in several different ways. On the one hand, it i s sometimes used to refer to the fact that the circumstances o f particular households and individuals may be highly volatile, so that even the non-poor may face a high risk of falling into poverty over time. On the other hand, the separate concept o f "vulnerable groups" refers to portions of the population that may face especially dire situations. 77. The concept of vulnerability in the sense of risk i s largely redundant in Haiti because so much of the population i s poor. By most definitions, all but a small portion of the population would be considered "vulnerable." Although hard data on changes that households experience over time i s scarce, it i s likely-given the country's political instability, the lack of security, frequency of natural disasters and the dearth of formal private or public insurance mechanisms-that most Haitians face high variability o f their living conditions over time. For most Haitians, however, this variability i s around levels of consumptionthat are already below those required to meet their basic needs. 78. The vulnerable groups concept i s more useful in highlighting the hardships of some elements of the population. Among the groups that might be considered vulnerable in Haiti are children, poor urban youth, and women. There are an estimated 3000 street children and large numbers left orphaned due to the death o f a parent infected with HIV. Poor urban youth who face minimal opportunities for education or employment are at highriskfor falling into drugs, crime, and gang violence.'* Demographics 79. Next, we consider the shape of poverty in terms o f household demographics. Much of the analysis in this section and the rest of the chapter i s presented in terms of '* WorldBank (2005b). 31 income quintiles. This mode of presentation i s preferred to the non-poor/poor/extreme poor categorization used for many poverty analyses because it allows for a broader view of variation in characteristics across the distribution. Readers who would prefer to consider the analysis in terms of conventional poverty categories should note that the US$2-a-day poverty headcount in Haiti i s 78 percent, while the "extreme poor" US$1-a- day poverty headcount i s 54 percent. Consequently, the richest quintile (Q5) corresponds roughly to the "non-poor," the next richest quintile (Q4) includes the "poor but not extreme poor," and the bottom three quintiles (41-43) are the "extreme poor," i.e. those living on less than US$1-a-day. 80. Household size in most countries i s larger for poorer households. The same pattern holds for Haiti. Household size, however, i s similar for all extreme poor households (those in the bottom three quintiles). Only households in the top quintile are notably smaller. Likewise, there i s very little variation in the number of children across the income distribution. In all quintiles, the average number of children is in the narrow range between 1.8 and 2.0. These household demographics are summarized in Figures 2.6. Dependency ratios-the number of elderly and children per working age adult-do vary more substantially by quintile. As shown in Figure2.7, the poorest households have more dependents per adult (2.5) than the wealthiest (1.5). Because this difference i s not driven by variation in the number of children, it must be due to the fact that there are larger numbers of older people in poorer households. The relatively high number of dependents in poor households places an additional burden on their limited resources. The mean dependency ratio of 1.9 in Haiti i s not, however, especially high relative to other countries in Latin American and the Caribbean. Many wealthier countries in the region have much higher dependency ratios. This provides some evidence that the burden of caringfor dependents i s probably not a key determinant of Haitian poverty. Figure2.6: HouseholdSize and Number of Children by IncomeQuintileinHaiti HouseholdSize Number of Children Under Age 12 2 00 - 195 I 9 0 g 185 180 175 Q1 Q2 C3 Q4 Q5 Mean Q1 Q2 Q3 Q4 Q5 Mean Quintile Quintile Source: SEDLAC (CEDLAS and The World Bank) based on microdata from the 2001ECVH., 2001). Quintiles shown are for equivalized income. 32 Figure 2.7: Dependency Ratio by Income Quintile inHaiti 1.5 2.0 1.5 I O 0 s 0.0 Q1 92 Q3 C4 QS Mean Source: SEDLAC (CEDLAS and The World Bank) based on microdata from the 2001 ECVH., 2001). Quintiles shown are for equivalized household income. 81. Haiti has a very young population. Actual and projected population pyramids for Haitifor 2000, 2025, and2050 are show inAnnex A9. Haiti's pyramidfor 2000 exhibits the bulging base typical of a country that has experienced high population growth in recent years. Nearly half (44.4 percent) of Haitians are under age 15, and almost two- thirds (65.7 percent) of Haitians are under age 25. Among the implications of such a population structure is that further population growth in the medium-term is inevitable. As the current large wave of young people reaches childbearing age, momentum will carry forward the country's population growth, even if fertility drops. Consequently, the projections show a significant expansioninpopulation over the next several decades. 82. As a result, Haiti will continue to experience pressure on its strained land resources and provision of basic services. Density doubled from the 1960sto the present, reaching the high level of 300 persons per square kilometer, and the projections show it topping 450 persons per square kilometer by 2050. These pressures highlight the need for acomprehensiveapproachto family planning services targetedto poor rural areas. Education and Literacy 83. Next, we consider education and literacy levels in Haiti. Education levels among older adults are quite low, and there are large gaps by gender. Among those age 41-50 (born 1951-60and thus of school age mainly in the 1960s and OS), men averagejust 4.2 years of education, while women average a mere 2.8 years. The age profile shows, however, that substantial improvement has been made in school attainment over time. Average schooling levels for Haitians age 21-30 (born 1971-1980 and thus in school chiefly inthe 1980sand early 1990s) are 7.1 years for women and 8.6 years for men. (See Figure 2.8)." 84. Because of the relatively rapid improvement in school attainment in Haiti, although average education levels for the entire adult population are low incomparison to other LAC countries, for younger cohorts education levels are higher than in many neighboring countries. For the population of all Haitians 25-65, average education levels are only 4.6 years. InLatinAmerican andthe Caribbean, only Guatemalahas lower adult l9 This refers to the highest grade attained and does not account for repetition. 33 education levels. For just adults age 21-30, however, Haiti looks much better. Average education levels are 7.8 years, higher than Guatemala, Honduras, and Nicaragua, and not far below Brazil. Figure 2.8: Average Years of Educationby Age and Gender 10 9 8 7 6 5 4 3 2 1 0 10-20 2130 31.40 41-50 5 i 6 0 61+ mFernale .Male Source: SEDLAC (CEDLAS and The WorldBank) basedon microdatafromthe 2001 ECVH., 2001). 85. The national averages hide the substantial inequality in education b y income. Among all adults, those in the top quintile have on average more than twice the number of years of education as those in the lower four quintiles. There is little differentiation in average education levels across the four poorest quintiles. Average years of education by quintile are show in Figure 2.9. 86. Inequality in education levels in Haiti, in fact, i s the highest in the region. Figure 2.10 shows a plot of mean years of education vs. the Gini coefficient for years o f education for countries in the region. Inequality and mean years o f education have a strong inverse correlation, and Haiti's education Gini o f 63.7 i s the highest for all L A C countries. Figure 2.9: Average Years of Educationby Income Quintiles (AdultsAged 25 to 65) lo1 8.6 Q1 Q2 Q3 Q4 Q5 Mean Quintile Source: SEDLAC (CEDLAS and the World Bank) basedon microdata fromthe 2001ECVH., 2001). 34 Figure 2.10: MeanYears of Educationvs. Inequality inYears of Education(Adults Ages 25 to 65) for Countriesin LatinAmerican and the Caribbean 87. The pattern for literacy rates shown in Figure 2.11 reflects that of educational attainment in general. Literacy rates are very low for adults as a whole. Only 39 percent of adult women and 53 percent of adult men in Haiti can read and write.20 Among the young cohort age 15-24, however, the situation is much improved, though still far from satisfactory, with 75 percent of women and 79 percent of men literate. Figure 2.11: Literacy Ratesby Age and Gender loo%, 15-24 25-65 65 + Age 1 Female 0 Male 1 Source: SEDLAC (CEDLAS and The World Bank) based on microdata from the 2001ECVH. 88. Figure 2.12 shows enrollment profiles by age and type of schooling for 2001. The figures show that many children start attending school late; at age 9, only half of children 2o Itshould be noted that literacy is self-reported in the data based on the 2001 ECVHand might therefore be somewhat inflated. The 2000 DHS has literacy figures for women based on literacy tests (asking them to read a card) and those figures are generally similar to the self-reported numbers. 35 are enrolled. The contrast between the poorest and the better off are stark. Very few children in the poorest quintile ever attend secondary school, and even at age 16-the peak age of enrollment, only 60 percent o f children are in school. Girls tend to have higher enrollment rates at younger ages, while boys have higher rates at older ages. There i s also a sharp contrast between urban and rural: urban enrollment rates are much higher at all ages. Figure 2.12: School Enrollment Pyramids by Age and Level of Schooling (a) All Children (b) Richest20% vs. Poorest20% Ail Richest 20% Poorest 20% = - - I I I I I I I I I 1 .75 .5 2 5Proportion .25 0 5 .75 1 Primary Secondary Post-secondary Primary Secondary Post-secondary (c) Male vs. Female (d) Urbanvs. Rural Male Female Urban Rural = 1 .75 5 25Proportion 25 0 - 5 75 1 - Primary Secondary Post-secondary Primary Secondary Posbsecondary Source: DHS 2001 89. Another way to consider recent schooling patterns i s to examine education attainment, i.e. the fraction of children who have completed particular grade levels. When we examine education attainment for teenagers in 1995-95 and 2001, we see that despite Haiti's continued economic problems, there has been progress in school attainment. Figure 2.13 shows the attainment profile-percentages of students reaching each grade level-by income group in the two years. The profile shows progress at nearly all points, particularly for students in lower grades in the poorest 40 percent of households. For example, while only 40 percent of 15-19 year-olds had completed 3'd grade in 1994-95, that figurejumped to 56 percent in 2001. However, while there seems to be an increase in school attainment, this does not suggest that the quality o f education 36 has improved. Since much of the new investment has been in low quality private funded schools, it i s likely that learningand literacy have not increasedsignificantly.*' Figure2.13: EducationAttainment Profileby Economic Status, Age 15-19, Haiti 1994-95and 2001 100% 90% I: 80% 70% -I-Richest20%,2001 60% --I.. 94-95 Richest 20%, 50% +Middle 40%, 2001 ..&.--Middle40%, 94-95 40% -O- Poorest 40%, 2001 --e- Poorest 40%, 94-95 30% 20% 10% 0% 1 2 3 4 5 6 7 8 9 Grade Source: DHS. Economic status groups (Richest 20%, Middle 40%, Poorest 40%) are derived from a ranking of households that is basedon household asset ownership and housing characteristics, 90. We also consider the private returnsto education in terms of wage earnings, self- employment earnings, and the sum of the two. Earnings are those reported during the 12 months prior to the survey in 2001. Only 17 percent of adults report any wage earnings at all duringthat period, while 47 percent report having some amount of self-employment earnings. The fact that such a small fraction of adults have such earnings complicates the analysis of the returns to education. 91. The small role for wage earnings i s also apparent in a breakdown of household income, shown in Table 2.1. For the average household, only 20 percent of income comes from wages, with an additional 37 percent from self-employment. Transfers- from within Haiti and abroad-make up another 25 percent, while own consumption (consumption of food produced by the household) accounts for 11percent. 21 It should be noted that, as noted inChaper 5, school attendancerates appear lower in 2000 than 1995, even though completed schooling levels of 15-19 year olds is higher in 2000. This could be the case if things were generally improving after 1995, but then 2000 witnesses a deterioration (which was the case given the significant political instability in that year). 37 Table 2.1: Household Income by Source (Means PercentagesAcross Households) Wage income 20% Transfers 25% Property income 3% Other 4% Self-employment 37% Own Consumption 11% Total 100% Source: Sletten and Egset (2004) based on analysis of ECVH. 92. The returns to education calculated here are conditionai on having some positive earnings (in the relevant category). The unconditional returns to education-which i s much more relevant to policy-involve both the probability of having earnings at all and the conditional returns to education. Separate analysis (not shown) indicates that education i s not strongly associated with the probability of having positive earnings. Ina multivariate regression analysis, each additional year o f education i s associated with just a 1percent increase in the probability of having some wage earnings.22 Each additional year of education i s associated with a 1.7 percent lower probability of having some self- employment earnings. Overall, those with more education are neither more or less likely to have positive total earnings. This suggests that the unconditional returns to education may not differ substantially from the conditional returns. 93. Results from a regression analysis of the returns to education are shown inAnnex 1, separately for each of the three earnings variables. Three different specifications for each earnings variable are shown. The first i s a version o f the standard Mincer form, with years of education, a quadratic in age, and a female dummy. The second includes interactions of all terms with a female dummy. The third adds a urban dummy and an urban*female interaction term. Overall, the estimates for the marginal return from an additional year o f education are 15.6-19.7 percent for wages, 11.2-13.6 percent for self- employment earnings, and 14.0-17.2 percent for total earnings. These estimates are similar to estimates for other countries in Latin American, which are typically around 15 percent (World Bank, 2006). Worldwide, the returns to education average roughly 10 percent (Pscharopoulos and Patrinos, 2004). Specific operational recommendations for the education sector are developed inChapter 5. Infrastructure 94. Next, we consider how access to infrastructure varies over the income distribution inHaiti. Figure 2.14 shows that dwellings for poorer households are much more likely to be made of low-quality materials, which include earth, wood or boards. 22These statements are based on marginal effects calculated at mean characteristics from probit regressions for having wagehelf-employment/total earnings on the same set of variables used in the long specification for the returnsto education analysis. 38 95. Access to infrastructure services varies almost entirely between the top quintile and the bottom four quintiles. Recall that the bottomfour quintiles encompass the group defined as poor in terms of the US$2-a-day poverty line. Access to safe water, electricity, telephone and hygienic restrooms by quintile is shown in Figure 2.15. Access to safe water i s rare even for those at the top end. Only 14percent of Haitians overall and 25 percent o f those in the richest quintile have piped water in their homes. Access to electricity is more common and highly differentiated between the poor and the better-off. Note, however, that because actual availability of electricity in Haiti can be sporadic, access to an electricity line does not mean that electricity i s consistently available. Telephone access i s virtually exclusive to those inthe top quintile, but only 17 percent of that group has a telephone. Likewise, hygienic restrooms (defined as having a toilet connected to a sewer system or septic tank) are only available to 12 percent of those in the top quintile and are essentially non-existent in poorer households. Specific recommendations on infrastructure are set forth in Chapter four. Figure 2.14: Share of Dwellings Made of Low-Quality Materials by Income Quintiles Q1 Q2 03 Q4 Q5 Quintile Source: SEDLAC (CEDLAS and the World Bank) based on microdata from the 2001 ECVH. Low- quality materials are defined as earth, wood, and/or boards. High-quality materials are concrete, brick, and stone. Figure 2.15: Infrastructure Access by Income Quintiles 56% 20% __. 01 a2 a3 04 a5 Mean ai a2 03 a4 Q5 Mean l m m e Wintile lmorneWintile Source: SEDLAC (CEDLAS and the World Bank) based on microdata from household surveys. Shares of population with coverage are reported. Water access is defined as having water piped into the living quarters, within the compound, or well on plot. Electricity indicates access to electricity within the living quarters. Telephone access includes either a fixed line or a cellular phone. A household has hygienic restrooms if it has a restroom with a toilet connected to a sewer system or septic tank. 39 96. With data from the 2001 Demographic and Health Survey, we can consider the geographic distribution of access to infrastructure. Figure 2.16 shows access by to various forms of infrastructure by region. For all three services-electricity, pipedwater, and pit latrines-access i s highest inPort-au-Prince. Figure2.16: Percentageof HouseholdswithAccess to Infrastructureby Region Electricity PitLatrine , Source: DHS 2001 Health and Nutrition 97. High levels of poverty and poor service delivery have resulted in dire health outcomes for Haiti's poor, as described in Box 2.3. Here, we consider measures of child health and nutrition. Infant (under age 1) and child (under age 5) mortality rates are very high in Haiti. Figure 2.17 shows mortality rates for 1994/95 and 2001 in Haiti and for 2002 in the Dominican Republic as comparison point. While the under-five mortality rates showed a slight decline between the two surveys, infant mortality increased. As Figure2.18 shows, while infant mortality is highinall regions of the country, it is highest inareas of the Ouestregion outside of Port-au-Prince. Figure2.17: Infantand ChildMortalityinHaitiandthe DominicanRepublic 140 f 120 100 de" 1 BO 4a 20 a Infant mortality rate Under-five mortality rate I I Haiti 1994/95=Haiti 2000 ODorninicanRepublic 2002 I Source: DHS. 40 Figure 2.18: Infant Mortality Rates in Haiti by Region, 2001 q:, fl 104.86- 107.66 90.45 104.86 87.78 90.45 72.82 87.78 --- 67.24 - 72.82 Source: DHS 2001. Rates shown are death per 1000 live births for the 10-year period preceding the survey. Box 2.3: A Personal Story About the Ties Between Health, Poverty and Deforestation A man fromBaie de Heme describes his situation as follows: I am a charcoal man like everyone else. Since that does not pay much, I am not in good health. It has become harder and harder to find wood in Ti Paradis. My house floods when it rains; I am not able to cover it. My children do not go to school. My wife and my children are not in good health. I went to the hospital with my child, and they wanted $600 for an operation. I do not have money. I cannot dig a latrine. There is no school and no clinic. Whena pregnant woman needs help, we put her on a boat to go 1to the Red Cross. Sometimes she dies along the way. Source: Manigat et al. (2003). I HIV/AIDS Epidemic 98. Haiti faces a severe HIV/AIDS epidemic. UNAIDS estimates that 5.6 percent of the adult populationhas HIV. B y these estimates, 260,000 adults and 20,000 children are living with HIV. The epidemic has been generalized since the late 1980s,spreading from highly vulnerable groups to the population at large. According to UNAIDS estimates, 70 percent of commercial sex workers were HIV-positive in 1998. 99. A number of factors make the further spread of HIV in Haiti likely. A 1993 survey showed high levels o f stigma associated with the disease. Fear and stigma can inhibit people from seeking testing, counseling, and information. Condom use is low, with just 12 percent of sexually active women having ever used a condom. The dangers that HIV will spread are also enhanced b y high levels o f male migration, the lack of 41 social ties and social control, and the high levels o f poverty that push some women into sex 100. Despite all the risk factors in Haiti, there is some evidence that prevalence rates have not increased dramatically in the last decade, as would be expected, and may have even declined. HIV sero-prevalence surveys at five sites was 6.296, 5.696, 5.696, and 3.1% in 1992, 1996, 2000, and 2004 respectively. UNAIDScredits the close public-civil collaboration and sustained political commitment for this trend. 24 CrimeandViolence 101. Crime and violence are woven into the fabric of everyday life in Haiti, particularly in the capital. The fear felt by residents of Port-au-Prince i s in marked contrast to the relative safety of rural life in Haiti. Sixty percent of those living in the capital report that they feel unsafe in their own homes "often" or "most o f the time", compared to 14 percent in the rest of the country. In Port-au-Prince, 25 percent of residents report having been the victim of theft, robbery, or burglaryin the two last years, compared to 16percent of those inother areas of the country (World Bank, 2006). 102. For the countr as a whole, crime and the resulting fear strike equally across the income distribution?' Crime victimization rates and self-reported levels of fear are similar for Haitians in the top income quintile and those in the bottom quintile. This uniformity is, however, in large part due to the fact that the poor are concentrated in the relatively safe areas outside the capital. Within Port-au-Prince, the poor experience greater feelings of insecurity. While levels of perceived insecurity are high across the socioeconomic spectrum in the capital, they are highest for the very poor. In Port-au- Prince, 29 percent o f those in the poorest quintile say they "never" feel safe at home, versus 13 percent of those inthe top quintile. 103. Gender-based violence i s undoubtedly one o f the most common forms of violence in Haiti. Gender based violence includes physical violence, emotional violence, sexual violence and economic violence based on gender or directed specifically towards men or women. Domestic violence and rape tend to be suffered by women in the majority o f cases, while gang violence i s targeted mostly to men. Box 2.4 describes the forms of violence and their impact affecting women in Haiti and Box 2.5 existing constraints to Haitian women's access to justice. 23 World Bank (2002) 24 http://www.unaids.org/en/Regions_CountrieslCountrieslhaiti.asp 25 Based on ECVH data analysis. 42 Box 2.4: Gender-Based Violence inHaiti Gender based violence against women in Haiti i s prevalent in both urban and rural areas. Data from a DHS violence module in 2000 shows that around 15 percent Haitian women who are currently married or living with a partner have experienced physical violence by such partner. Within the same group of women, about 19 percent have experienced sexual violence by their partners. DHS violence rates are usually a lower bound estimate as these surveys are conducted at home and there i s significant underreporting. Gender based violence has devastating consequences not only for victims but also for society as a whole. Studies for developing countries conclude that the health impact (especially for reproductive health) of gender based violence on women can be as high as some of the leading causes of injury. Previous studies have also documented the impact of gender based violence on earnings due to death and lost productivity, job loss, lost productivity of the abuser due to incarceration, and loss of tax revenues due to death and incarceration. In Haiti, the following health impacts have been attributed directly to experiencing physical violence by a partner: (1) women who suffer physical violence by a partner are 116 percent more likely to suffer from genital sores or ulcers; (2) women victims are 28 percent more likely to suffer from anemia; (3) children of women who suffer physical violence at the hands o f intimate partners are 32 percent more likely to suffer anemia; and (4) gender basedviolence also increases women's usage o f health services. Given these impacts and their costs, it is important to focus on the prevention of gender based violence, even more strongly than in the provision of services for victims. This requires women's social, economic, and legal empowerment. Prevention of all forms of gender based violence i s achieved through multi-sector approaches (health, education and judicial sectors) and policies working at the individual, community, and institutional levels. In the long term, violence prevention is enhanced through changes in norms and attitudes that foster violence at home, school, workplace and community. Box 2.5: Women and LegalInstitutions inHaiti Women's access to justice i s severely limited due to a series o f antiquated legal provisions in need o f urgent update as well as a systemic deficiency of the Haitian judicial system. Although Haiti is a signatory of the Universal Declaration of Human Rights, the CEDAW Convention for the elimination of all forms of discrimination against women and the Interamerican Convention for the elimination and punishment of gender-based violence (Belem do Para), the national legislation, particularly the civil and the penal codes, discriminates de facto women and does not protect them adequately, particularly in the case of domestic violence and rape. Rape i s classified as an offence against public decency. In addition, medical certificates are required to prove rape and are difficult or impossible to obtain. Adultery i s classified as a second-level crime and i s punishedwith two to three years of imprisonment. Inthe case the husband kills the wife upon discovering her and her lover injlagrante delicto, the judge may decide to excuse the murderer and condemn him to no more than two years of prison. Domestic abuse has traditionally been seen as a family issue and, as such, not to be penalized. Women are also discriminated in the exercise of their basic economic and property right. Although most of the families are informal and not legally sanctioned by marriage, the Haitian law precludes women living in informal marriages from the right of inheritance of her partner's properties and access to land titling. In addition, it excludes the possibility for a married man to recognize children born outside of the marriage, which contributes to increase the number of undocumented children and the number of children falsely registered as the offspring o f a married woman. These antiquated legal provisions are further compounded by the broad systemic deficiencies of the Haitian judicial system. As indicated in Chapter 3, at the lowest level, the Haitian justice system relies on the Judges of Peace ouges de paix), who have a broad range of functions in civil, criminal, and extrajudicial matters. However, as for other judges, the Judges of Peaceare underpaid, have limitedor no access to legal texts, and have little opportunity for professional development. This, coupled with the country's difficult conditions, and the problems faced by the national police, limits the access to, and the efficiency of, the judicial system, thus delaying the conclusion of a trial or promoting arbitrary decisions. 43 Agriculture and Rural Poverty 104. Inrural Haiti, the dominant activity is subsistence farming. Small-scale farmers suffer from land fragmentation, a problem which has increased as a result of rapid population growth. Declining land yields have pushedfarmers to move to less accessible land in the hills, feeding deforestation and erosion in a destructive cycle which has left rural Haitians in deep poverty. 105. Small farmers in Haiti have few assets and rely on primitive technology. While technology i s not available to even a majority of agricultural households in the top quintile, there is a sharp gradient by income in the use of irrigation, pesticideshsecticides, and fertilizer. As Figure 2.19 shows, the use of productivity- enhancing technology i s practically nonexistent among the poorest farmers, and i s much more common among better off farmers. Access to credit, however, does not vary by income; only 11-12 percent o f farming households ineach quintile has access to credit. Figure 2.19: Agricultural Technology and Credit Usedby FarmingHouseholds, By Income Quintile 40% 35% 30% 25% Poorest 20% 15% ,2,2 0Richest 10% 5% 0% Irrigation Pesticidesor Fertilizer Credit Insecticides Source: Verner (2005), based on data from ECVH (2001). 106. Verner (2005) shows with regression analysis that higher income per capita on farming households is associated with greater education levels, use of technology, more land, possession of a land title and access to infrastructure (water, roads and electricity). Although it is difficult to demonstrate causal effects with cross-sectional regressions, these findings are in accordance with those from studies in other contexts that generally find such variables to be associated with higher agricultural incomes. A 2005 World Bank report (Haiti - Agriculture and Rural Development) contains a detailed review of these constraints andpriority interventions for rural development. 11. Evolutionof Povertyand Inequality Over Time inHaiti 107. Examining changes in poverty and inequality over time in Haiti poses severe challenges. Haiti suffers from incomparable survey data, a problem common to many developing countries with weak statistical institutions. In Haiti, each of the three 44 expenditure or income surveys collected in recent years (1986, 1999, and 2001) has a very different design. As a result, the analyses of poverty done on the basis of those surveys differ in the estimates of incidence and trends (see the Annex for a review of those analyses). The Demographic and Health Surveys, designed to be comparable, are of high quality but unfortunately do not include the expenditure or income data needed for poverty estimates. 108. In this section, we consider changes over time in welfare in Haiti in three ways. First, we examine changes in socioeconomic indicators for which fairly consistent data exist over time. Second, we consider results from qualitative surveys in which people are asked about their subjective impressions of changes over time. Third, we present an econometric analysis designedto construct comparable welfare estimates over time. Changes inSocioeconomic Indicators 109. Child malnutrition i s one of the few indicators for Haiti that can be tracked with some reliability over a long period of time. Nutrition surveys carried out at various points since 1978 make it possible to consider changes over more than two decades. Figure2.22 shows various child malnutritionindicators for Haiti during the period 1978- 2000 and the Dominican Republic in 2002, as a point of comparison. Stunting is generally an indicator of long-term malnutrition, while wasting i s a sign of shorter-term hunger, e.g. due to famine. The condition of being underweight i s a hybrid of stunting and wasting. All three measures have declined over the long-term in Haiti. The stunting rate in 2000 was just over half the rate in 1978, and other figures show similar progress. Child malnutrition rates are still high, however, compared to those in neighboring Dominican Republic, where the stunting rate i s approximately equal to the average rate for the Caribbean as a whole. Haiti's stunting rate of 22 is similar to that of the Central American region (20.4) and substantially below the average inAfrica (35.2).26 Figure2.20: ChildMalnutritioninHaitiandthe DominicanRepublic 45% , 40% 35% 30% 25% 20% 15% 10% 5% 0% Stunted Wasted Underweight =Haiti 1978 =Haiti 1990 UHaiti 1994/95 OHaltl 2000 ODominican Republic2002 Source: DHS and WHO Global Database on Child Growth and Malnutrition. Figures shown are % of children under 5 years who are classified as undernourishedaccordingto three anthropometric indices of nutritional status: stunted(height-for-age), underweight (weight-for-age) and wasted (weight-for-height). ~~ 26Child stunting rates for Central America and Africa are estimates for 2000 from de Onis et al. (2004). 45 Subjective Impressionsof Changes Over Time 110. Respondents to a 2003 qualitative survey (Manigat, Lamaute-Brisson et al., 2003) were asked how they perceived changes in poverty for the country as a whole during the five years prior to the survey. More than half (52.5 percent) said they believed that poverty had much increased, and an additional 22.2 percent reported that poverty had increased somewhat. Perceptions were very similar across region, income group, gender and education level. 111. In another participatory study (Calpas and Dieusibon 2004), the authors quote a group o f women in Ilet Pierre Joseph of Anse d'Hainault, "Since 1986 our situation has grown worse. The reason i s that under the Duvalier dictatorship the political situation was stable enough that it did not adversely affect the economy." 112. For the 2001 ECVH, respondents were asked several questions about their subjective welfare. Asked "Since when has the economic situation of your household been so difficult?" 65 percent of households replied "5 years or longer," and only 5 percent said that their economic situation was not difficult. Comparison Over Time Basedon Econometric Poverty Projection 113. To address the survey incomparability problem, this section presents results from an analysis to compare welfare over time. The essential problem i s that in each survey with either income or expenditure information in Haiti, the data was collected with entirely different questionnaires. A wide variety of experience in other countries has shown that even small changes in the way such data i s collected can have a substantial impact on poverty estimates. 114. The procedure employed in this section involves estimating an econometric relationship between income and household characteristics with the 2001data, usinga set of characteristics common to the 2001 survey and an earlier survey. The estimated relationship is then used to project a simulated distribution o f income to the earlier survey.27 115. The primary weakness of the approach is that it relies on the assumption that the relationship between income and the household characteristics employed for the procedure i s stationary over time. It is difficult to evaluate the plausibility of this assumption. It would be more plausible if the household characteristics used for the analysis were ownership of consumer durables, if the prices o f such assets have not changed dramatically over time.28 Unfortunately, the 1986 survey in Haiti does not include asset ownership. As a consequence, the estimates for 1986 are based on an imputation using variables for household size, number of rooms, electricity access, toilet ''The procedure used here is very similar to that of Stifel and Christiansen (2006), drawing heavily,on the **workof Elbers, Lanjouw, and Lanjouw (2003). Imputations basedon assets alone are similar to various welfare indices based on assets. See for example Sahn and Stifel (2000) and Filmer and Pritchett (2001). 46 facilities and water facilities. Estimates for 1995 and 2000, however, do include assets in the regression equation. 116. Figure 2.21 shows summary figures for the main variables used in the projection analysis. Ownership of some assets-a television, refrigerator and bicycle-was slightly higher in 2000/2001 than in 1995. However, the data is ambiguous when it comes to radios, the only asset that i s owned by more than a quarter of the population and for which we have data over time. The 2000 DHS shows an increase in radio ownership relative to 1995, while the 2001 ECVH shows a decline. The figures for infrastructure also do not paint a clear picture. While the fraction of households with access to electricity has increased slightly since 1986, the fraction with a latrine has dropped from 78 percent to just over half. Figure2.21: HouseholdAssets and Services Over Time 1995-2001 Asset Ownership 1995-2001 Infrastructure 1986-2001 50%- 4, 90%, Radio TV Refrigerator Bicycle Motorcycle Car Electnuty Modemtalet Latnne Pipedwater Source: Staff analysisof householdsurveys. 117. The summary figures for trends in assets and infrastructure offer a rough preview of the results of the projection exercise. Because the poverty estimates produced by the projection exercise are based on the distribution of assets, infrastructure and a few other variables, in the absence of substantial shifts in those variables, the poverty estimates are unlikely to show substantial changes over time. 118. Figure 2.22 shows a plot of headcount estimates from the projection method over time. Confidence intervals at the 95 percent level are shown as error bars on the estimates. The point estimates show a small decline inpoverty over time nationally, from 60 percent in 1986 to 54 percent in 2001. It is important to emphasize, however, that due to the error inherent in the estimates, the difference between the two estimates for the two years i s not statistically significant. To emphasize this point visually, the 95 percent confidence interval on the 2001 estimates i s extended across the figure. In every case, the confidence intervals for a particular year's estimate falls within the confidence interval of the 2001 estimate. This indicates that statistically we cannot reject the hypothesis that poverty remained unchanged over time. Part (b) of the figure shows poverty trends separately for Port-au-Prince and the rest of the country. Estimates for both parts of the country show small declines in poverty. Note that because they are 47 based on smaller sample sizes, the confidence intervals on estimates at the subnational level are larger.29 119. Estimates based on the US$2-a-day poverty line are presented in the Annex. These show trends broadly similar to those for US$1-a-day poverty rates. The US$2-a- day headcount estimates show a decline from 84 percent to 78 percent. As with the US$1-a-day estimates, given the large margin of error in these estimates, the change is not statistically significant. 120. Despite the caveats and large confidence intervals associated with these estimates, they do provide some evidence that poverty may have not decreased or increased substantially since 1986.30 Figure2.22: US$1-a-DayHeadcountPovertyEstimates for 1986,1995,2000,2001: PointEstimatesand Confidence Intervals National Port-au-Princevs. The Rest of the Country 8oy01 ........................................................................................................................................................... I 80% ~ 70% 60% 67% 67% 57% 70% ..... 60% 53%54% 60% 50% 50%1 I+Non Port.au-Pr1ncel 40% - /+port J 40% - AU Prince 27% Source: Analysis of various surveys. Estimates for 1986, 1995, and 2000 are imputed from an econometric relationship. See text for details. Notes: Bars shown around each estimate show 95 percent confidence intervals. Dashed lines are confidence intervals on 2001 estimates. 121. The full set of poverty estimates along with estimates o f Gini coefficients are shown in the Annex. Trends over time for the poverty gap and the squared poverty gap are comparable to those for the poverty headcount. Inequality shows no statistically significant change over time. ~~~ 29 Itwould be desirable to have estimates at a more disaggregated level, e.g. rural, Port-au-Prince, and other urban. Unfortunately, there i s some evidence that the rural-urban categorization i s not consistent over time in surveys. The households classified as "other urban" inthe 2001 survey appear much worse off interms of infrastructure than those in all other surveys, suggesting that many are in fact in households classified as rural in previous surveys. For this reason, the "other urban" and "rural" households are considered together as "non Port-au-Prince" for the analysis of trends over time. 30 The main analysis was performed by estimating an econometric relationship with the 2001 survey and then projecting backwards into the 1995 and 1986 surveys. As acheck, the analysis was also performedin reverse, by estimating a relationship between explanatory variables in the 1986 survey and projecting forward to 1995 and 2001. Trends were similar with both methods. 48 111. The RoleofRemittances 122. The apparent improvement in education and child nutrition measures despite the country's economic stagnation and continued high poverty rates i s puzzling. The increase in remittances sent by migrants abroad may partially explain the improvements in some welfare measures. If remittance income is poorly measured in the household survey, and remittance income has gone in part to fund food and education, the rapid increase in remittances during the 1990s may be behind the observed welfare improvement^.^ ' Box 2..6: RemittanceFlows to Haiti Remittances, Aid and FDIFlows, 1971-2003 Migration i s one characteristic feature that has (current US%Der caeita) shaped Haiti's history and its people. However, over the past twenty years the influx of Haitians '*O abroad has increased in larger numbers. Today, Haitian expatriates are estimated to be about 2 million people, compared to 8.5 million living in Haiti. They reside primarily in the US, the Dominican Republic and Canada. These migrants have maintained ties to their families in Haiti and, as a result, the Haitian economy has become increasingly reliant on the money sent by the Haitian diaspora living abroad. The growth in remittance flows has accelerated more rapidly since the mid 1990s,rising from US$lOS million in 1995 to US$930 million, almost ten times that d r. a - amount in 2004 (Central Bank estimates). This m " m (0 E E E E m m N E m m In 00 E E E m m 0 7 N represents about a quarter of the country's GDP Iinvestment Foreigndirect (the second largest in the world as a percent of -Workers' ......, remittances, received GDP). Because Haiti has periodically Officialdevelopmentassistance and official aid I I Source: The WorldBank experienced political turmoil, economic crises and natural disasters, remittances have proved to be a steady source of income and a critical safety net for poor Haitians, consistently representing an amount o f money that has been many times larger that the total annual FDI to Haiti and annual revenue from merchandise exports, and more recently external assistance. I Source: Orozco (2006), Understanding the Remittance Economy in Haiti. 123. The Haitian economy has become increasingly reliant on the money sent by the Haitian diaspora living abroad. According to Haiti's Central Bank, Haitians remit almost one billion dollars a year, or more than a quarter of the country's GDP. The growth in remittance flows to Haiti has accelerated more rapidly since the mid-1990s (see Box 2.6). This hasbeen accompanied by an increase inconsumption (see Figure 2.23). The Haitian Central Bank estimates coincide with Haiti's 2001 Census which showed that one in five households receive remittances in the country. Using the Haiti Census and the average annual amount sent by remittance companies (US$2,000), the total amount yields approximately US$672 million in 2001, a figure similar to the Central Bank's report for that year. 31 This section draws on material from the remittances background paper prepared by Orozco (2006). \ 49 Figure 2.23: Consumptionand Workers 124. Inahouseholdsurvey conducted Remittances in 2001, 27 percent of households (1985-2003) reported receiving some form o f support from abroad (from relatives or others) in the previous 12 months. Among households with relatives abroad, sixty percent have relatives in the United States and twenty percent have family members in the Dominican Republic. WO,_,. la"."l."5a. *"dso-*n..!Lln 01 emoar-*.. Source: The World Bank, IHCEand IMF When asked about their relatives' motivations for migrating, seventy percent stressedwork as the primaryreason. DemographicProfileand Characteristicsof RemittanceRecipients 125. Inorder to analyze the characteristics and patterns of remittance recipients, a survey of 578 persons was conducted in 2006 in Haiti focusing on their demographic profile, remittance reception, access to and use of financial institutions, and contact with relatives. The survey was based on 578 remittance recipients living in the Western part of the country, predominantly inPort-au-Prince. 126. Beneficiaries of remittances have different backgrounds yet share similar characteristics, some of which are significantly different from other Latin American recipients. First, unlike people from other Latin American and Caribbean countries, half of Haitian remitters are women and 60 percent are under thirty years of age. Second, remittance recipients have a much lower income thantheir counterparts inLatin America, particularly because over half are unemployed. Moreover, twenty percent say they receive remittances from more than one relative, and at least half of recipients report receiving additional help other than remittances. Third, nearly half of recipients have at least one child to take care inthe household. 127. Households that received remittances in Haiti tend to be wealthier households. This is illustrated by the fact that when remittances are counted as part of income (as per usual), nearly half of households (48 percent) in the wealthiest quintile receive remittances, compared to just 14percent of those in the poorest quintile. The distribution of remittances by quintiles of income including remittances i s shown as the white bars in Figure 2.24. 128. The fact that wealthier households receive remittances i s due to the fact that remittances make those households wealthier. When remittances are excluded from income for the purposes of defining poorer vs. wealthier households, wealthier and poorer households have roughly the same likelihood of receiving remittances. The black bars in Figure 2.24 show the distribution of remittances when hypothetical income quintiles are defined, excluding remittances from the income variable. Inthe top quintile, 33 percent o f households receive remittances, compared to just 31 percent of households inthe poorer quintile. Remittances are slightly less likelyto be receivedby households in the middle three quintiles. 50 Figure2.24: Percentageof HouseholdsReceivingRemittancesby IncomeQuintile 6O% 50% 48% I 1 40% n 317'0 33% 30% 29% 25% 24% 27% 27% 20% 22% 200% 14% 10% 0% Poorest 2 3 4 Richest Mean Quintile of Income Per Capita 0Quintilesdefinedby income INCLUDING remittances I definedbyincomeEXCLUDINGremittances Quintiles 129. In terms of remittances received, recipients coincided in reporting the same average transaction amount that remittance senders mentioned, which i s US$150. Forty percent received less than US$lOO, but another forty percent received between US$lOO and US$300 (see Table 2.2). The frequency of receiving i s quite regular and, similar to what senders report, remittance transactions are made over ten times a year. Moreover, 47 percent have been receiving the money for at least five years. Twenty percent say they receive from more than one relative, while 60 percent o f remittance recipients receive the money from their parents, and 33 percent receive money from their siblings. The figure reported i s consistent with the age of the recipients, the majority of which are young people intheir twenties. Table 2.2: Amount ofRemittancesReceived Range % 100dollars or less 41 Between 101and 300 dollars 42 Between 301 and 500 dollars 9 More than 500 Dollars 8 NS/NR 1 Total 100.0 Source: Remittance recipient survey, Orozco (2006) 130. In terms of the expenditure of their money, Haitian recipients respond to a hierarchy of needs similar to other Latin American recipients: their money is spent on food, education, clothing and savings. As shown in Table 2.3, the expenses reported demonstrate a balanced expenditure on the different activities. For example, not all the money i s spent on food, but rather on a combination of items related to education and clothing. Second, the rate of savings of remittances i s among the highest in Latin America. 51 Table 2.3: How Do Remittance RecipientsUsethe Money? Use of Remittances % Food 81 Education 74 Clothes 64 Savings 34 Healthcare 28 Housing 27 Business 14 Other 11 Repayment of debts 9 Pavmentof debts 7 Source: Remittance recipient survey, Orozco (2006) As reported by Orozco (2006), worker's remittances in Haiti tend to stimulate more consumption that productive investment: while 81percent of income from remittances i s utilized for consumption goods like food, education and clothing. The expansion of consumption i s reflected in a growing trade deficit, which since 2000 has reached nearly 100percent of GDP. The trade deficit has been aggravated by the lackluster performance of exports, as explained above. This analysis highlights the high dependence of the Haitian economy on remittances, in addition to external aid. Leveraging Remittances for Development 131. Much attention has been given to the question of how policy can promote the use of remittances for long-run de~eloprnent.~~ A number of points from this work may be relevant to Haiti: 132. Importance of policy environment. When economic conditions in the receiving country are favorable to investment-i.e., when there are well-functioning public institutions and investment opportunities are available-remittances tend to go partially towards investment. Incontrast, when conditions are not ripe for investment, remittances tend to be spent largely on current consumption. This suggests that maximizing the effectiveness of remittances first requires improving the general investment climate in the country. 133. Lack of information on valid projects is a key barrier to migrants' support for their home communities. Migrants living abroad have increasingly pooled their resources to invest collectively (in addition to the person-to-person remittance flows) in development-related activities in their home communities, either through hometown associations (HTAs) or other migrant group schemes. HTAs have proliferated among the Latin American and Caribbean diaspora in Canada and the United States, and similar organizations exist in France, the United Kingdom, and Africa. A problem often raised by Haitian migrants living abroad is that they often have minimalinformation about valid 32See for example, World Bank (2006) and Orozco (2005). 52 projects that could be supported in the communities where their relatives leave andwhich the HTAs could support. Cooperative efforts could be made in tandem with the government, donor organizations and/or NGOs to help disseminate information on qualified projects. 134. Remittances can provide an entry to banking services. For some migrants, remittance flows come regularly enough that they could potentially be used to function as guarantees for loans or for accessing other banking services. If remittance transfer companies developed such lending instruments, they could provide the basis for greater investment. In this respect, it will be important to reduce inefficiencies in remittance sending processes. Examples in Latin America and other countries are available that could be explored for Haiti. 135. As indicated in Chapter one, with an improvement in the investment climate in Haiti, the potential contribution of the Haitian diaspora could go beyond just sending remittances-it could contribute to expanding foreign trade, capital and technology flows and to the important task of institution building. Implicationsfor BroadShared Growth 136. The growth and poverty determinants and trends analyzed inthe first two chapters point to the need to complement pro-growth polices with actions that ensure access to services for the poorest regions and households through institutions that are responsive, effective and accountable to their users and citizens. The following chapters provide a more detailed discussion of how this might be achieved through actions in the area of state and institution building as well as service delivery in the infrastructure and education sectors. 53 Annexes 1. PreviousPovertyEstimatesOver Time for Haiti 137. Pedersen and Lockwood (2001) of the FAFO Institute of Applied International Studies inNorway estimate poverty lines and poverty rates based on expenditure data in the two EBCM surveys. The Pedersen and Lockwood analysis generally follows the methodology typically used in World Bank poverty analyses. Because the consumption surveys are not identical in the two ECBM surveys, for purposes of analysis, it was necessary to combine some categories in each of the surveys to produce a fairly consistent basket across the two. The resulting food basket has 50 items in the EBCMI1 and 53 in the E B C M 111. Patterns of food expenditure in the second quintile were used to determine the weights for the food basket. The surveys show only minimal changes in expenditure patterns between 1986 and 1999/2000. The researchers chose to use expenditure patterns from 1999/2000 to determine the food basket. They calculated two different poverty lines, one based on 2900 calories per day per adult male equivalent, and another based on 2240 calories per capita. The cost of a food basket scaled up to these calorie levels was used to calculated the food poverty line. Non-food expenditures for the full poverty line were calculated by taking the non-food expenditures of households with total food expenditure equal to the food poverty line. The Pedersen and Lockwood analysis shows a decline in the general poverty rate from 59.6 percent in 1986 to 48.0 percent in 1999/2000. 138. Montas (2005) questions the Pedersen and Lockwood methodology for 1999/2000. He points out that the Pedersen and Lockwood poverty lines imply non-food shares of consumption of 70.2 percent in 1986 and 75.3 percent in 1999/2000.33 He recalculates the 1999/2000 poverty rate using a food share of 60.5 percent. It i s not clear what the basis i s for the selection of this share. With this food share he finds a general poverty rate of 59.6 percent in 1999/2000, exactly equal to the rate in 1986. He suggests that his figures indicate that poverty has not fallen. 2. Other Poverty Lines for Haiti 139. Different researchers have employed several slightly different methods to calculate poverty lines in Haiti. The US$l-a-day and US$2-a-day poverty figures presented here for 2001 were calculated using the World Bank's standard methodology and are identical to figures published in the World Bank's World Development Indicators database. This methodology i s described in Deaton (2003) and WDI(2004). For the US$1-a-day line, the Bank multiplies US$l.08 by the 1993 purchasing power parity (PPP) exchange rate and then by a price index in local currency to convert to a current value in local currency (Gourdes in Haiti). For the case o f Haiti, the Bank's $1-a-day 33Pedersenand Lockwooddetermine the non-foodshare inthe povertyline by estimating non-food expenditure for householdswith total expenditure correspondingto the food povertyline, which is a standardprocedureoftenusedby the World Bank. 54 poverty line on a yearly basis i s calculated as follows, where HPIt i s the Haitian price index inyear t: PWB =($I -08)(ppp1993)(HPI2000mpI1993) (365) (1) Egset and Sletten (2004a) andEgset and Sletten (2004b) also calculate US$1-a-day and US$2-a-day poverty lines, but use a slightly different procedure: Here USCPI, i s the U.S. Consumer Price Index in year t. UNDP(2003) produces $1- and $2-per-day povertylines and appliesthemto the 2001 ECVHusinga different procedure, multiplyingUS$1.08 by the PPPexchangerate for the year 2000, the year inwhich the data was collected: PUNDP =($1.08) (PPP2000)(365) = 2681 Gds Verner (2005) uses the US$1-a-day extreme poverty line determined b y UNDP(2003) but calculates the figures on a household rather than aperson basis, Le. determining the fraction of households inpoverty instead of the fraction of individuals inpoverty. 55 TableAl: HeadcountPovertyEstimatesfor Haiti Poverty Line inGourdes General Extreme (Extreme Personsor PovertyRate Poverty Rate Source households? (FGTO) (FGTO) Pedersen and Persons 59.6 44.5 Lockwood (2001) Pedersen and Persons 48.0 31.4 Lockwood (2001) Montas Persons (2001') I31.4 UNDP(2003) Persons 76 55 Verner (2005) Households 48.9 Egset and Persons 76 56 Sletten(2004) SEDLAC Persons 78.0 53.9 (2005) World Bank Persons 75.9 53.9 (2006) 56 Table A2: PovertybyPopulationSubgroups Contribution to NationalPoverty Poverty Headcount Index I% of Cateaorv) Yo of Yopopulation Yopopulation National below $2 a below $1 a below $2 below $1 Population day day aday aday Total Haiti 100.0 78 54 By geographical area: Port-au-Prince 9.2 51 29 6 5 Other urban 15.0 85 66 16 18 Rural 76.0 86 59 83 83 By gender of household head: Male 51.3 77 54 51 52 Female 48.9 79 53 50 49 By education of household head: Not completedprimaryeducation 89.9 83 59 96 98 Primary completed 7.9 55 32 6 5 Secondary completed 0.9 31 21 0 0 Tertiary completed 1.4 20 5 0 0 By sector of household head Agriculture 56.7 89 63 65 66 Low technology industry 3.1 68 40 3 2 Manufactureindustty 1.o 48 39 1 1 Construction 2.1 72 38 2 1 Commerce, hotels, restaurants 28.0 77 52 28 27 Gas, electricity,transports, communications 1.3 45 19 1 0 Banks, insurance,financial institutions 1.o 33 19 0 0 Public administrationand defense 1.1 27 17 0 0 Education, health, private services 4.9 46 27 3 2 Domestic services 0.5 77 65 0 1 Source: Bank staff anlaysis of ECVH2001 57 Table A3: Returnsto EducationinHaiti log(Wageearnings) log(Self-employmentearnings] (1) (2) (3) (4) (5) (6) Years education 0.187 ** 0.197 ** 0.156 ** 0.125 ** 0.136 ** 0.11 (0.010) (0.012) (0.012) (0.008) (0.010) (O.OO! Age 0.045 0.031 0.027 0.137 ** 0.142 ** 0.12 (0.035) (0.041) (0.033) (0.019) (0.025) (0.021 Age squared -0.0004 -0.0002 -0.0001 -0.0015 ** -0.0015 ** -0.001 (0.000) (0.000) (0.000) (0.000) (0.000) (O.OO( Female -0.024 -1.143 -0.941 -0.067 0.380 0.07 (0.105) (1.414) (1.330) (0.058) (0.689) (0.664 Female*Yearsed. -0.038 * -0.014 -0.025 * -0.01 (0.022) (0.021) (0.015) (0.014 Female*Age 0.102 0.088 -0.012 0.0c (0.077) (0.073) (0.036) (0.03! Female*AgeSq. -0.002 -0.001 0.000 0.0c (0.001) (0.001) (0.000) (O.OO( Urban 1.247 ** 0.77 (0.123) (0.084 Female*Urban -0.472 ** -0.16 (0.216) (0.11' Constant 5.911 ** 6.007 ** 5.700 ** 4.316 ** 4.117 ** 4.31 (0.665) (0.801) (0.649) (0.373) (0.502) (0.471 Observations 997 997 997 3129 3129 312 R-squared 0.31 0.31 0.4 0.15 0.15 0. Source: Analysis of 301ECVH. Notes: Dependent variables are shown above columns. Earnings are during the previous 12 months. Total earnings is the sum of wage earnings and self-employment earnings. Sample includes all individuals age 20-64 with positive earnings inthe relevant earnings category. 58 Table A4: Summary HouseholdCharacteristicsfor Various Years National 1986 1995 2000 2001 Household size (mean) 4.9 5.1 4.8 4.5 # of sleeping rooms (mean) 2.2 2.5 Household assets (% hholds) Radio 39.8 47.3 38.3 TV 17.0 22.6 21.1 Refrigerator 8.3 9.5 11.3 Bicycle 9.2 13.9 12.4 Motorcycle 1.o 1.2 1.o Car 3.3 3.6 3.1 Access to services (% hholds) Electricity 24.1 31.4 33.8 31.4 Modern toilet 1.7 4.4 3.5 3.7 Latrine 55.6 53.3 57.7 52.0 No toilet 42.7 42.3 38.8 44.3 Piped water (mainsource of water) 6.5 8.3 Piped water or bottled water (main drinking source) 12.3 11.5 9.8 Sources: EBCM I1(1986), DHS (1995 and 2000), ECVH (2001). All figures shown are national means across households (not individuals). Port Au Prince 1986 1995 2000 2001 Household size (mean) 5.1 4.8 4.6 4.4 # of sleeping rooms (mean) 1.9 2.3 Household assets (% hholds) Radio 73.0 81.6 69.5 TV 56.5 69.1 69.3 Refrigerator 27.1 29.3 37.2 Bicycle 6.0 8.2 7.0 Motorcycle 0.7 1.1 0.8 Car 11.3 10.0 9.9 Access to services (% hholds) Electricity 88.0 92.5 94.7 92.3 Modern toilet 6.4 16.9 10.7 12.7 Latrine 89.5 78.3 84.3 78.0 No toilet 4.2 4.8 5.o 9.4 Pipedwater (main source of water) 25.5 20.7 Piped water or bottled water (main drinking source) 33.1 25.6 24.8 Sources: EBCM I1(1986), DHS (1995 and ZOOO), ECVH (2001). All figures shown are national means across households (not individuals). 59 Table A4: (continued)SummaryHousehold Characteristicsfor Various Years NonPort Au Prince 1986 1995 2000 2001 Household size (mean) 4.8 5.2 4.8 4.6 # of sleeping rooms (mean) 2.2 2.6 Household assets (% hholds) Radio 29.6 34.9 28.9 TV 4.9 5.8 6.7 Refrigerator 2.6 2.4 3.5 Bicycle 10.2 16.0 14.0 Motorcycle 1.1 1.2 1.o Car 0.9 1.3 1.1 Access to services (% hholds) Electricity 13.0 12.7 11.7 13.2 Modern toilet 0.9 0.6 0.9 1.o Latrine 49.7 45.7 48.1 44.3 No toilet 49.4 53.8 51.0 54.8 Piped water (main source of water) 3.2 4.5 Piped water or bottled water (main drinking source) 5.9 6.4 5.3 Sources: EBCM I1(1986), DHS (1995 and 2000), ECVH (2001). All figures shown are national means acrosshouseholds (not individuals). 60 Table A5: US$1-a-Day Poverty Estimates for Various Years Based on Poverty Projection Exercise National 1986 1995 2000 2001 Poverty headcount (FGTO) 0.603 0.57 0.53 0.539 (0.029) (0.019) (0.024) (0.015) Poverty gap (FGT1) 0.351 0.33 0.298 0.307 (0.022) (0.017) (0.017) (0.012) Squared poverty gap (FGT2) 0.246 0.23 0.204 0.212 (0.019) (0.015) (0.014) (0.01) NonPort-au-Prince 1986 1995 2000 2001 Poverty headcount (FGTO) 0.665 0.673 0.638 0.633 (0.034) (0.023) (0.023) (0.016) Poverty gap (FGT1) 0.393 0.399 0.369 0.369 (0.026) (0.021) (0.019) (0.014) Squared poverty gap (FGT2) 0.278 0.281 0.256 0.258 (0.021) (0.019) (0.016) (0.012) Port-au-Prince 1986 1995 2000 2001 Poverty headcount (FGTO) 0.266 0.211 0.215 0.217 (0.042) (0.033) (0.051) (0.031) Poverty gap (FGT1) 0.119 0.09 0.091 0.094 (0.022) (0.017) (0.028) (0.017) Squared poverty gap (FGT2) 0.071 0.051 0.052 0.054 (0.016) (0.011) (0.019) (0.011) Source: Sources: EBCM I1(19861, DHS (1995 and 2000). ECVH (2001). Estimates for 1986, 1995, and 2000 are imputed from an econometric'relationship. See text for details. Standard errors shown inparentheses. 61 Table A6: US$2-a-Day Poverty Estimates for Various Years Based on Poverty ProjectionExercise National 1986 1995 2000 2001 Poverty headcount (FGTO) 0.836 0.799 0.774 0.78 (0.019) (0.013) (0.02) (0.011) Poverty gap (FGT1) 0.592 0.562 0.529 0.538 (0.021) (0.015) (0.019) (0.012) Squared poverty gap (FGT2) 0.468 0.443 0.41 0.419 (0.021) (0.016) (0.018) (0.012) NonPort-au-Prince 1986 1995 2000 2001 Poverty headcount (FGTO) 0.885 0.889 0.87 0.865 (0.021) (0.012) (0.014) (0.01) Poverty gap (FGT1) 0.643 0.65 0.621 0.618 (0.024) (0.018) (0.018) (0.013) Squared poverty gap (FGT2) 0.515 0.521 0.492 0.491 (0.024) (0.02) (0.018) (0.013) Port-au-Prince 1986 1995 2000 2001 Poverty headcount (FGTO) 0.563 0.485 0.495 0.489 (0.048) (0.044) (0.06) (0.035) Poverty gap (FGT1) 0.311 0.257 0.261 0.26 (0.036) (0.03) (0.045) (0.026) Squared poverty gap (FGT2) 0.211 0.169 0.172 0.173 (0.029) (0.024) (0.037) (0.021) Source: Sources: EBCM I1(1986). DHS (1995 and 2000). ECVH (2001). Estimates for 1986, 1995, and 2000 are imputed from an econometric'relationship. See text for details. ,I Standard errors shown in parentheses. Table A7: Gini Coefficient Inequality Estimates for Various Years Based on ProjectionExercise 1986 1995 2000 2001 National 0.659 0.703 0.692 0.687 (0.022) (0.018) (0.018) (0.016) Non Port-au-Prince 0.607 0.615 0.616 0.621 (0.025) (0.017) (0.014) (0.01) Port-au-Prince 0.617 0.636 0.635 0.627 (0.034) (0.024) (0.03) (0.025) Source: Sources: EBCM I1(1986), DHS (1995 and 2000), ECVH (2001). Estimates for 1986, 1995, and 2000 are imputed from an econometric relationship. See text for details. Standard errors shown in parentheses. 62 Table AS: US$l-a-Day Poverty and Inequality Estimates from a ProjectionForwardfrom 1986to 2001 Non Port-au- National Prince Port-au-Prince 1986 2001 1986 2001 1986 2001 Povertyheadcount 0.603 0.551 0.696 0.689 0.093 0.077 (0.035) (0.022) (0.041) (0.028) (0.034) (0.021) Povertygap 0.262 0.241 0.305 0.305 0.021 0.018 (0.02) (0.016) (0.024) (0.02) (0.009) (0.007) Squared poverty gap 0.144 0.132 0.168 0.169 0.008 0.006 (0.014) (0.011) (0.016) (0.015) (0.004) (0.003) Gini index 0.439 0.477 0.378 0.396 0.333 0.351 (0.02) (0.019) (0.024) (0.015) (0.024) (0.025) Source: Sources: EBCM I1(1986), DHS (1995 and 2000), ECVH(2001). Estimates for 1986, 1995, and 2000 are imputed from an econometric relationship. See text for details. Standard errors shown in parentheses. 63 CHAPTER3 STATEBUILDING,SECURITY AND INSTITUTIONS Contemporary Haiti lacks a set of public institutions capable of carrying out the basic statefunctions in a responsive, effective and transparentfashion. In recent years, public institutions have lost trained personnel and other resources, have not modernized their procedures and provide few incentives for staff pe$ormance. Political instability and budget constraints are part of the explanation, but the gap has its roots in Haiti's historical development. Overcoming these weaknesses will require an active and sustained commitment to building and working through institutions, and transparent rules and procedures to ensure pe$ormance and accountability. This is a priority in terms of building new legitimacy of the Haitian state with its citizens, reducing poverty through effective service delivery and addressing the deeply entrenched disparities that have haunted Haiti. Without this, it is difficult to imagine a scenario whereby Haiti could provide the pre-conditions for sustained and broad-based growth. Building on the ICF, recommended interventions include: (1)accelerating the reform of PNH, simultaneously strengthening other elements of thejudicial and penal system, and targeting slum areas with multi-dimensional development interventions; (2)addressing weaknesses in revenue, expenditure and human resource management to ensure an efficient and results-based use of domestic resources and external finances;(3) building capacity to plan and track basic service delivery, whether by public or private providers, with a focus on poor regions and households, and (4) increasing the dissemination of information on government outputs, as well as encouraging citizen groups to monitor and evaluate public institutions 'pe$ormance. I.TheNatureofStateBuilding 140. As is now widely recognized, institutions and governance matter, and the poor quality of both constitutes a major impediment to development efforts in all sectors. Of all the countries in the LAC region, Haiti faces the greatest challenges in these areas. If in a more exaggerated form, the country shares a basic problem with many of its neighbors - that o f weak institutions which seriously inhibit efforts to improve public service delivery, plan and oversee public investments, and guarantee that governmental and donor resources will be well spent. Assessments of state performance over the past decades amply demonstrate Haiti's unusually low performance in all areas. One particularly critical problem i s that the Haitian state has been unable to provide the minimal levels of political stability and public security needed for citizens to carry out their normal lives and for private, governmental and donor investments to be successful. 141. Over the past two decades, Haiti has experienced a chronic situation of internal conflict. As opposed to most LICUS, post-conflict or fragile states, Haiti has one advantage. Its current challenge i s purely state, and not nation, building.34 Haitians 34Haiti's nation-building challenge was addressedduring and in the decades immediately following independenceand can be consideredto have beenresolvedby the mid 19`hcentury with the separation of the Dominican Republic. However, even during the early decades, the disputeswere more over political 65 participating in or simply affected by the conflicts are not contesting their national identity, but rather how and by whom they will be governed. Box 3.1: Lessonsfrom donor interventions infragile states State building is a key challenge that needs to be recognized as a priority and addressed from the outset. The role of international community i s important but can also add to the fragility of states if not well coordinated. The international community can play a helpful role if it harmonizes around one set of rules, including procurement and financial management. Interventions must be prioritized so as not to overload governments and should generally provide some immediate results that help the state to build trust while developing institutional capacity in the medium-long term. All interventions, however mundane, if done in an inclusive, transparent manner and communicated effectively, can help to build trust in the new state. State building requires a longer term timeframe than a 5 year project cycle and donors must commit for the long haul. Enabling flexibility i s key and thus providing scenarios and a process that enables review and adjustments inpolicy dialogue is important. Public processes, such as budget and procurement must be understood as political as well as technical processes. Dependability in dialogue, avoiding too much rotation of staff, i s required. 142. State-building is not an easy task, and there are no quick fixes or standard formulas for how to accomplish it. There are many lessons that can be gleaned from experience elsewhere in the world, as shown in Box 3.1. Three are particularly relevant for Haiti. First, state building is a long term undertaking, but must be addressed strategically with an emphasis on advancing the basics first in a way that does not undermine longer term Second, although it i s incremental, some balance in progress i s essential, both as regards functions addressed and beneficiaries. The legitimacy gap poses a particular problem here as it inevitably originates in disparate and occasionally irreconcilable demands. Still, short-term responses can focus on some common denominators (e.g., security) and seek ways of satisfying less widely shared needs without exacerbating conflicts. Third, donors should support local institutions in ways that promote coordination among and between government entities and other stakeholders and result in a state that i s able to lead its own national development plan. This requires a long-term commitment. 143. State building i s essentially a political undertaking, but its goal can also be described in terms of the functions performed by an effective state. Definitions vary. Here we adopt the list used by the Overseas Development Institute (Ghani et al, 2005). Innoparticular order, the functions are: (1)legitimate monopoly onthe useofforce; (2) administrative control; (3) management o f public finances; (4) investment in human capital; (5) delineation and enforcement of citizen rights and duties; (6) provision of control than national "identity." Racial and class-basedexclusion has always been an issue, but not one giving rise, as it has in other countries, to secessionist movements. 35World Bank, 2005; 3-4. 66 infrastructure services; (7) market formation and regulation; (8) management of state assets; (9) international relations; and (10) rule of law. A well-functioning state must not only perform all these functions, but must do so effectively, in an integrated fashion, and ina way that engages and provides information to citizens. Stateswhich do not do so fall into what the OD1characterizes as vicious circles resulting in "the creation of contending centres of power, the multiplication of increasingly contradictory and ineffective decision-making processes, the loss of trust between citizens and state, the de- legitimization of institutions, the disenfranchisement of the citizenry and ultimately the resort to 144. The above description appears tailor-made for contemporary Haiti. Arguably the state has been 200 years in the making. Since independence, an inability to reach agreement on the basic rules of the game and an exclusionary treatment of the majority of citizens have placed the country in a perennial cycle of contested governments, political instability, popular uprisings and violent regime changes. Dictatorships, external interventions, and donors' channeling o f resources through nongovernmental actors, combined with substantial out-migration, have contained immediate pressures in a temporary fashion, but at the cost of preventing further development of state capacities and the emergence of a more inclusive governance system. Underlying structural problems - high levels of poverty, enormous inequality, and unsustainable patterns of economic growth -have increased the demands on the state and made their resolution more difficult. The new administration which will assume office in May 2006 may be facing the most difficult set of circumstances in Haiti's history. It will continue to confront possible political, social, and criminal violence, it will have to work through a state apparatus which remains inefficient, only partially under its control, and extremely limitedinits geographic penetration; and it will have to respond to a citizenry demanding visible progress in meeting a variety o f needs. Where to start i s a challenge, but strengthening state capacity to perform its most basic functions will clearly be critical. 145. All of the ten state functions require effective institutions, but some functions and institutions can be regarded as more critical to escaping the current vicious circle and addressing the ability of the Haitian state to provide core public goods.37 This means only that they require priority attention, not that they should be developed to the exclusion of all others. Taking some liberties with the OD1list, the suggested focus areas for institutional development are as follows: Capacity to guarantee public security within the rule of law - this requires both the expansion of the state's ability to protect citizens against crime and violence and of a counterbalancing set of institutions to prevent state abuses of these and its other powers. Although the World Bank is not working in either area in Haiti, it recognizes that breaching the security gap i s a precondition for significant progress in any other area. Capacity for economic governance - if weaknesses here are not as immediately threatening as those in security, few advances will be made in other areas absent the 36Ghaniet al., 2005; 9. 37WDR, 1995. 67 strengthening of budgetary institutions to guard against misuse of public funds or simpleinefficiency. Capacity to plan andoversee basic service provision and infrastructure development- whether the state provides services and infrastructure directly or works through NGOs andprivate firms, its ability to plan, coordinate andoversee the quality of basic services delivered mustbe enhanced. Capacity to foster greater transparency, higher levels of participation and more accountability for results. Although only implicit in several of the OD1functions, this i s a priority in Haiti. It does not require a separate set of institutions, but rather the inclusion of this capacity in core government functions and the ability of civil society to engage with government at the policy dialogue level. 146. Over the short to medium run, building institutional capacity in these areas is the most urgent priority. Before turning to some more specific suggestions, the rest of this chapter first looks in greater detail at the current situation and the factors contributing to its creation and perpetuation. It then discusses in more detail the priorities noted above and their relationship to a longer term reform program. Finally, it outlines some practical steps recommendedfor beginning to improve the country's institutions. 11. History of Victims, Victims of History: the PoliticalEconomyofHaitian Development 147. Haiti's current institutional challenges have their origins in the country's colonial history. The current overview will not reach back that far, but only note that, although Haitian independencebrought an abrupt end to a terribly exploitative colonial society, it did so without puttinga stable government inits place. 148. The revolutionary wars (from 1791-1804)built on a temporary alliance between a small Haitian elite contesting political domination by the white, and the enslaved majority, seeking freedom from bondage. The "slave revolt'' proved so threatening to France and other external powers as to provoke further attempts to reverse it, delay international recognition, and discourage trade relations for years after. 149. The independence period was followed by decades of internal conflict, the division and reunification (in 1820) of what would become modem Haiti, the annexation and then separation of the Eastern two thirds of the island, and a subsequent long history of coups, de facto regimes and external interventions. Between 1805 and 1987, Haiti had twenty-one constitutions and experimented with a variety of political regimes.38 Until well into the 20thcentury, presidents were usually chosen by the Assembly and relatively limited in their constitutional powers. However, disputes between the executive and legislative branch were periodically won by the former, with a consequent turn to more authoritarian control. Whoever prevailed, the impact on the majority of citizens was negligible at best. The country was predominantly rural, and for the mass of peasants the state's presence was felt only in the extraction of rents. Service provision and infrastructure development were minimal and virtually nonexistent for the rural majority. 38 BarthBlemy andCirault, 1993; 154. 68 150. While a rich colony (for the favored few), Haiti became a poor independent nation with the end of its slave based economy, the war-time destruction of the infrastructure built duringthe colonial period, and the severance of the ties the wealthy elite hadforged with European markets. Much of the country retreated to subsistence agriculture, a pattern not uncommon to other newly independent nations in the region. However, whereas other countries in the region eventually built alternative economic structures and corresponding state organizations encompassing increasing portions of their populations, Haiti's progress has been slower and beset by more setbacks. This has been the result of a class structure based on an extremely weak economic foundation and therefore lacking both a classical bourgeoisie and a large working class.39 These patterns of societal differentiation based on the emergence o f a more complex economy have been critical to forcing a broadening o f social and political participation elsewhere. Their relative absence in Haiti encouraged persistent and extraordinarily high levels of inequality, a politics of exclusion until well into the 20thcentury, and an explosive rather than gradual mobilization of a small middle class and the poor majority inthe 1970s and after.40 151. Known for their repressive policies, corruption, reliance on clientelism, and for delegating power to often corrupt and abusive state agents (e.g. the chefs de section and the tonton macoutes), the administrations o f Presidents FranCois and Jean-Claude Duvalier (1957-1986) witnessed the emergence o f new political actors.41 The problem, however, was that in the absence of accompanying economic development, the opportunities for individual advancement remained largely in access to public office and resources, in themselves still scarce, and thus provoked an ever fiercer competition for their control. Although the growing opposition, including the mobilized masses, succeeded in ousting President Jean-Claude Duvalier in 1986, it was unable to preserve its unity in forming a new government and soon provoked a military take-over, the military itself having emerged as one of the chief competitors for power. 152. Disunityremainedthe theme for the series of de facto and dejure governments to follow. It was exacerbated by two additional factors - external intervention and mass mobilization. Where internal competitors knew they could count on external allies, they were less amenable to forging domestic compromises and often quickly resorted to violence (as it i s outbreaks of escalating violence that seem most likely to provoke external intervention). With respect to the phenomenon of mass mobilization, the problem in Haiti has been the absence of broad based institutionalized political parties that have effectively represented their demands, and leaders representing the masses tended to do so through charismatic social movements. Other political actors, many of whom first emerged as backers of the old regimes (Le. demobilized military) or as allies of the Lavalas government (the chirndres), have also proved adept at instigating disruptive protests and violent criminality, especially in urban areas. A recent Bank 39 Fatton, 2002; 199. 40Fatton(1999; 56-64) describesthe role of PresidentJean-ClaudeDuvalierinprovokingthis mobilization through an effort to open the economy. His emphasison small industries also encouragedin-migration from rural areas and thus the creation and expansionof the urban slums surrounding Port-au-Prince. 41 Inthe former, the hithertoexcludedblack middleclass engaged politically and inthe latter a group of civil society organizations, again of largelymiddleclass origin, emergedoffering an incipientoppositionto authoritarian rule. 69 study calls them "entrepreneurs of violence," (World Bank, 2006b) seeing their emergence as another unfortunate consequence o f Haiti's apparent political impasse. 153. There are many reasons for the continuing growth and recent escalation o f crime and violence in Haiti. They include high rates of unemployment, the breakdown of ordinary social networks and controls in urban areas, the presence of a large group o f demobilized, but still armed former police, soldiers, and other repressive agents, and as discussed below, the state's incapacity to prevent or control them. However, the politicization of violence cannot be overlooked as a factor in its own right. Urban poverty and unemployment, especially o f young males, i s expected to raise the crime rate, but this impact i s increased b y the presence of interest and political groups prepared to mobilize them. Purposes may vary - from provoking foreign intervention, to making space for themselves at the table or simply discouraging policies interfering with their "business as usual,'742 to precipitating a government's downfall. This multiplicity of uses of violence i s ample evidence of its increasing importance as a political resource and one which is wielded by all sides of the political spectrum. Its ubiquity has negative consequences far beyond the immediate damage done. It further aggravates the polarization of political forces, discourages negotiated compromises, and undercuts the formation of more peaceful channels of interest representation. 154. Haiti's history of intense inter-elite battles and mass exclusion and the resulting two centuries of inattention to economic development, public service provision and infrastructure maintenance, have given way to a new political dynamic. Conflicts no longer are simply about who will control the state, but increasingly about what the state will do to respond to the escalating demands. Breaking with past practices -- the use of public employment to reward political followers (patronage) or for personal gain (rent- seeking), an emphasis on the state's provision of private as opposed to public goods (clientelism), and a zero-sum approach to politics -- will be difficult. With a static pie, the politics of patronage, clientelism, and rent-seeking reaches its limits rather quickly, pitting even potential allies against each other and so augmenting instability and the threats of violent ouster by crowds in the streets. This makes even more problematic the role of outside actors in attempting to mediate conflicts without creating dependencies or unsustainable compromises. 111. The PoliticalEconomy and itsImpact on State Development 155. InHaiti, the impact of political instability and perpetual violence has resulted in the state's inability to provide basic services and an environment conducive to attracting productive investment or effectively manage outside resources. Political instability and violence not only affect the life of ordinary citizens in profound ways, they also undercut political leaders' ability and will to create strong institutions. This situation can be summarized as a series of vicious circles: 42This business, unfortunately, is often illegal and increasingly drug-based, but Haiti is not the first country where the criminal class has usedpolitical violence to protect its turf. Brazil and Colombia are two infamous regional examples. 70 156. Vicious circle I: Political office i s the main means o f upward mobility. However, holding it i s so precarious that incumbents tend to use any means at hand to do so. Thereby they undercut the rules and institutions governing both theirs and others' actions. Also, there i s a premium on getting as much advantage as possible out of the temporary control of resources, thereby decreasing leadership's broader legitimacy and increasingthe chances of violent ousters. 157. Vicious circle 2: The formal rules for reaching agreements on policy are complicated and not well followed. Therefore, those trying to get things done will inevitably work outside them, thus further undercutting the formal processes and institution^.^^ 158. Vicious circle 3: Political alliances and organizations are transitory and leaders try to extend their direct control as far as possible, undermining the creation of more diffuse, institutionalizedcenters of power. 159. Vicious circle 4: Political leadership i s highly unstable and civil servants and those holding appointive office have few incentives to perform or to avoid their own rent- seeking strategies, given the uncertainty of rewards for good performance. 160. These vicious circles have resulted in the deinstitutionalization of Haiti's state structures and provided ground for corruption. Although most observers have focused on these, there is an additional negative impact in the lessened productivity of those institutions that remain in place. Haitihas a relatively small public sector which may not be of sufficient size or quality to carry out even basic functions. However, the problem may not only be attributed to the size and quality of the civil service but also to the failure to define work processes, supervise their execution and ensure that those carrying out tasks have the resources necessary to do them well. The underlying challenge i s far more complicated than fighting corruption or simply addressing employee absenteeism. Nonetheless, its origins lie in the same factors that encourage misuse of resources or employee absenteeism - a lack of emphasis on outputs and the role o f institutions in producing them, 161. Changing rules and adding resources (including intellectual capital via training) may be essential to altering behavior, but will not be sufficient. For that, two additional factors are needed: (i)leadership's commitment to working through institutions to produce desired outputs; and (ii)the creation and enforcement o f the necessary procedures to ensure that institutions function. While the former is an issue of political will, donors can help with the latter by providing assistance for the reengineering of processes and structures to enhance organizational/institutional productivity and greater responsiveness to citizen needs and demands. 162. The following sections look at the challenges and provide recommendations for the short and medium term in the four priority areas identified above, namely (1) public 43For a discussion of this phenomenon in the case of Argentina, see World Bank, 2006a. As the same study notes, such practices were also encouraged by the first generation macro-economic reforms which tended to be adopted via extraordinary legal measures. 71 security and rule of law, (2) economic governance, (3) service provision and supervision, and (4) transparency and participation. These areas can and should be pursued simultaneously as the first and most urgent steps in building a functioning state capable of serving the interests of all citizens. They are significant in their own right and collectively add up to far more than the sum of the parts, through their mutual reinforcement. IV. Public Security andRule of Law 163. This i s the area where state failure has been most critical, both for its direct consequences and for its impact on institution-building.44 Indeed, ensuring a dramatic improvement in security will be critical for the success of any other intervention in the country. Haiti's traditional mechanisms for insuring public order consisted of the army (Forces Arme`es d'Haiti, Fad'H), irregular agents of repression (tonton macoutes and military attach&) and local chefs de section who exerted boh policing and judicial powers. Unfortunately, the dissolution of the army was not accompanied b y a demobilization plan, and most members of the military were disbanded and sent home, still armed and able to engage in violence. The tonton macoutes began to disappear with the end of the Duvalier regimes which had depended on them to control the population. However, this tradition of the use of parallel and undemocratic forces of state violence continued with the Cedras military junta's attach&, the zenglendos of the early 1990s, and later the transformation of some organizations populaires, or popular organizations, into the chimeres (armed gangs). 164. Inthe 1990s, aseriesof UNpeacekeeping missionsworkedjointly withbilateral donors to rebuild the police force (Police Nationale Haitienne, PNH) and strengthen the institutions of justice, including the judiciary, public prosecution and prisons (rule o f law program). Efforts to rebuild the police absorbed a significant portion o f donor resources. The UN and bilaterals managed the selection and training of new recruits, funded equipment, and supported the drafting o f new operating and organizational rules. Unfortunately, this exercise was not accompanied by arrangements to finance the operating costs of the new force and thus the target of 10,000 recruits was never reached. The numbers peaked at 6,500 and have since declined. Although donor support lasted three years and created a new organizational structure, hindsight suggests that insufficient attention was paid to such factors as effective management, training, separation from the already overburdened Ministry of Justice (which i s also responsible for overseeing the judiciary, prosecution and prisons), the internal and external control apparatus, and the underdevelopment of the rest of the chaine pknale or criminal justice system (courts, prosecution, prisons, defense and the private bar). Major setbacks were the politicization of new recruits, police links to de facto armed bands (the chimdres) which eventually came to operate independently, and the entrance into criminal, often drug-related, activities by some police members. 44The areaof security andjudicial reformis explored in more depth by the Haiti Country Social Assessment. 2006. 72 165. Although recent World Bank opinion surveys done in Port-au-Prince's poor neighborhoods show a relatively high level of confidence in the concerns have been expressed about alleged police involvement in extra-judicial killings and other criminal activities in both poor and better off neighborhoods, their efficacy in preventing the escalation of violence, and continuing allegations of other human rights abuses. The inability of the police to contain instability and violence, following the departure of President Aristide in 2004, led to the creation of MINUSTAH, the UN peace-keeping force, to help improve the situation. 166. As mentioned above, a weak link in previous attempts to reform this sector was the inadequate attention paid to other elements of the justice system. In Haiti, these elements never operated well but, unlike the police, tended to be the victims of government neglect rather than of active intervention. The largest group of judicial authorities, thejuges de paix or justices of the peace, operated more in the ambit of local authorities (chefs de section) than under direct central control. Due to poor information and monitoring systems in the country, the Ministry of Justice, the official supervisor of the system, lacked much information on the identities ofjuges de p i x , their performance, or even their physical presence in office. Unlike the juges de paix, higher level judges and prosecutors had formal legal training and were more likely to experience interference from central authorities, but the entire system tended to function in a very decentralized and ad hoc fashion. As the economic situation worsened post-1980 and officials were paidpoorly, late, or not at all, the incentives for rent seeking increased considerably. The system arguably had a negative impact on government performance and was occasionally used to repress opposition, although it may have played some positive role in resolving local conflicts (through the juges de paix). First instance courts hearing criminal cases (courts d'assises) met infrequently, adding considerably to delays inthe administration of justice and to prolonged pretrial detention?6 Also, the poor functioning of the civil registry and the fact that over 40 percent of Haitians lacked identity records severely restrictedcourt access.47 167. While several donors (most notably USAID, CIDA, the French Cooperation Agency) as well as the UNDPsupportedjudicial reforms during the mid 1990s, the scope of the program was much narrower than that of the police. Although donor and local assessments revealed a wide range o f basic weaknesses (notably low salaries, inadequate preparation and equipment, poor demand for reform, alleged corruption, minimal oversight, and political interference), assistance programs focused largely on providing trainin?; some equipment and infrastructure, and reorganizing courtroom processing of cases. These programs have not proved very effective in improving performance even inless problematic settings, and where they have had any impact, have required far more time andresources as well as a fuller menu of complementary activities than provided to 45 World Bank, 2006b; 22. 46 Clouatre, 2002; 650. 47World Bank, 2006b. 48Onepositiveresult of donor attention was the productionof extensive documentationonjudicial problems. See Aucoin et a1(1997); Clouatre (2002); OASNN (1996); Saint-Louis (2002); and World Bank(2006b) for treatmentsand other sources. 73 Haiti.49Additional obstacles inHaitiwere the low priority assignedto these programsby the government and its continuing interference in the selection o f judges and in the conduct of specific cases. Donors tended to operate relatively independently of the Ministry of Justice as well as of the higher court andprosecutors offices, working largely with lower level courts (tribunuux de p i x ) . As a consequence, the problems of incentives or of external interference in system operations were never addressed. 168. As for prisons and the private bar, relatively little was done. Prison programs were in the nature of fixing the most egregious isolated problems (e.g. some improvements to existing infrastructure, some staff training, and the use of public defenders to check on pretrial detainees), but the improvements achieved were largely undone by subsequent inattention and the violence surrounding President Aristide's depart~re.~' 169. In order to improve public security, especially in the most violent areas, and provide ajudicial check to discourage abuses, there is a need to adopt a multidimensional approach which addresses key factors. Over the short term, this approach should include efforts to improve efforts to accelerate the reform of the PNH, re-launch judicial reform efforts based on lessons learned from the past, and target slum areas with multi- dimensional development interventions. Inaddition to these efforts to strengthen the `supply' of security and justice, the international community should strengthen local `demand' for reform by engaging withjustice reform stakeholders. Box 3.2: The Abolition of the Army inCosta Rica A small country in the conflict-ridden Central American sub-region, Costa Rica abolished its military after a civil war that ended in 1948. This decision was incorporated into the Constitution and marked the beginning of what has since become the most democratic and stable o f all the Central American states. The abolition of the military coincided with major investments in social and educational programs. Money otherwise spent on the military was funneled into these programs. Security treaties with neighbors as well as regional and international organizations were also established. The government also continually invested in its democratic system, through education, institutional support, media development and the support of basic freedoms. A 7,600 member national police service was given responsibility for law and border enforcement. Approximately 4,500 officers work in the rural areas, another 1,500 conduct border patrol, and other smaller units focus on immigration, drug enforcement, and intelligence. Much attention has been paid to creating a professional service free from a history of political cronyism in which service and promotions are based solely on merit and education. Officers must attend a six-month training academy that includes training on, among other things, human rights, democratic government and the role of a military versus a police force. Today, Costa Rica boasts one of the highest standards of living and largest middle classes in Central America. I t has remained a peaceful democracy throughout the difficult civil wars in its immediate neighborhood. Its experience shows that security is not only a military concept, but instead a comprehensive matter that includes political and economic variables. The absence of a military should thus not be seen as a loss of security, but instead as a redefinition of security. ~~ 49Here one might compare the Haitian program to that undertaken in El Salvador over a twenty-year period. El Salvador still has a flawed justice system, but improvements have been achieved with these and other interventions, including the gradual replacement of manyjudges and prosecutors, higher budgets and salaries, and more government and civil society involvement. See Hammergren (1998, and 2006). 50 Huntington el al, 2003 provides a description of these programs. 74 170. With respect to security forces, two things seem clear: (i) needs its own Haiti improved security force, and (ii) as long as it does not have one, it will require the continued presence of the UN forces. Moreover, reinstating the army would not fit the bill because it would either mean paying for two security forces, or substituting a militarized force for a civilian police. The first i s excessively costly; the second i s widely regarded by security experts as inappr~priate.~'Indeed, the experience of other countries shows that the absence of a military need not be seen as a loss of security (see Box 3.2 for Costa Rica). The current number of PNH personnel i s uncertain but official figures showed 4,500 officers in May 2005, assisted by 1,200 UN civilian police officers (UNPOL). Assuming that the force may have grown to 5,000 since then, the PNH remains one o f the most numerically weak police forces inthe world, with 63 officers per 100,000 citizens. This compares, for example, to 285 per 100,000 inthe LAC region. In addition to increasing the PNH size, it will be important to (i) improve coordination of the MINUSTAH and the PNH to extend their effective coverage and allow an eventual phase-out o f the latter, and (ii)accelerate efforts to reform the PNH, including establishing new rules and operating standards, vetting and training personnel, and improving oversight arrangements and career management. Were this to be done, the police force could be more efficiently deployed, and while it might not reach every village, it could at least guarantee greater security in urban areas, and in a good part o f the countryside. Until Haiti has a capable and effective police force, MINUSTAH'S presence and possible enlargement will be necessary. Police reform efforts by the UN and donors need to be sustained over the medium term, with elements to improve the relationship between the PNH and citizens by forming partnerships between the police, business and local communities, enhancing community policing efforts, including greater accountability of police officers for areas under their jurisdiction, and professionalization and trainingo f the force's members. 171. To ensure accountability of the security agencies, rules of engagement and levels o f force (i.e., how much coercive force i s justifiable in given situations) must be better developed and internal and external controls better defined. Part of the external controls should be vested in a reformed criminal justice system which will also be needed to regulate other types of activities and to process those detained by the police. On the basis of past failures in Haiti's rule of law programs, it i s recommended that more attention be given to the transparent selection, monitoring, and evaluation of judges and prosecutors (eliminating any political interference in either), that their salaries be raised and paid regularly, and that citizens be made aware of their basic political and civil rights. With regardto additional material and technical assistance, it would be important to prioritize, in conjunction with police reform, the criminal courts in Port-au-Prince and other urban centers. Over the medium-term, reform efforts should be expanded to other geographic areas and other jurisdictions (e.g. justices de p i x , civil courts) and involve substantial reform o f legal codes. There have been successful examples of the use of lay judges in other countries (e.g., Peru), as a means of resolving local conflicts via a mixture o f formal law and local custom. There are signs that some o f Haiti's juges dep i x have served this 51This is the unanimous view of external experts. See Mendelson (2005) who dismisses the idea out of hand. 75 end - expanding that potential through a program for regulating and improving their actions would enhance their impact. There is also a need to repeat, since most of the benefits have been lost, the 1990s program of prison and prison staff upgrading. Finally, the content and objectives of rule o f law programs should be discussed broadly and involve court officers, private lawyers, the business community, and civil society organizations. 172. Haiti's impoverished urban slums are a prime source o f violence and insecurity. It is well known that these areas are plagued by crime and controlled by gangs, and serve as a point for the mobilization o f violent political protests and criminal actions by "entrepreneurs of violence". There i s an urgent need to target urban slum areas, especially in Port au Prince, with a multidimensional approach specifically aimed at providing basic infrastructure, housing, and job creation (see details on these interventions in Chapter four) and increasing public security. Efforts to target youth at risk will also be important, mainly through youth development programs and efforts to ensure access to education for all children. Over the longer term, policies that promote pro-poor growth are needed to reduce poverty and hotbeds for violence and instability. Such policies include increasing human capital and creating an enabling environment for job creation and private investment. V. Economic Governance 173. As indicated in Chapter one, a critical challenge facing Haiti is to strengthen economic governance. An analysis of the country's public expenditure management practices and systems as of late 2003 reveals several weaknesses which impede the efficient use of domestic resources and external financing: (i)budget formulation, execution and reporting; (ii) public enterprise management; and (iii) resource human management. This i s accompanied by a very low yield of the tax system which has left the tax/GDPratio at a level that i s inadequate to support Haiti's development needs. This section reviews these weaknesses, progress made in the last two years to address them and priorities for reform going forward. Although economic governance weaknesses extend to all levels and sectors o f government, attention here i s focused on the functions and organizations most directly related to the programming and use of public resources- financial management inits most ample sense. 174. With respect to budget formulation and execution, the data for budget analysis and preparation was traditionally weak, and budget typically did not incorporate all expenditures and revenues. Their contents were not linked to programs or higher-level government goals but solely presented in line item form. The Treasury, which should manage liquidity and program disbursements in line with available funds, had become little more than a mechanism for endorsing checks to employees and suppliers and processing accounts on an accrual basis. Typically, deficit control was accomplished by tapping Central Bank reserves or accumulating internal and external arrears. During 1997-2001, the government operated without an approved budget and the three following budgets were approved late into their fiscal years. A significant share o f public resources (reaching 60 percent in 2002-03) was channeled through multiple "current accounts" held by individual ministries and used non-transparently at the spending agency level or was 76 assigned to a catch-all category called "other public interventions". While these practices evolved as an attempt to circumvent a complicated expenditure execution process, they resulted in lack o f transparency inpublic resource management and led to corruption. On the investment side, expenditures were not identified and planned in a coherent framework, projections were limited to a single-year timeframe and sector ministries' planning offices played little part inthe process (many had simply disappeared). 175. Internal controls were affected by the lack o f a well-structured accounting system, fragmentation of the accounting function, and the extensive use of "current accounts". External audits o f the government budget had not been done for several years. The supreme audit institution, the Cour Supe`rieure des Comptes et du Contentieux Administratif (CSCCA), focused nearly exclusively on ex-ante control, duplicating other agencies' functions and representing a conflict of interest given its ex-post control responsibilities. The posts of several of the CSCCA's ten members had remained vacant for several years and the agency lacked sufficient equipment as well as funds for training. Another usual source of external control, the Parliament, was not in operation between early 1999and mid-2000or after early 2004. 176. Untilrecently, public procurement operated under a 1989decree which had never been fully implemented and had several flaws.52 The system was decentralized, but the Commission Nutionale des Murche`s PubZics (CNMP), which was to oversee its implementation, had not existed since the early 1990s. The ad hoc arrangements used to fillthis void were neither efficient nor transparent and gave rise to allegations of corruption. Reduced operating budgets and salaries drove qualified procurement specialists to seek employment in the private sector. Sole-source contracts and unadvertised bidding had become the norm; there were no standardized bidding documents or manuals; and information on proposed or awarded contracts was not publicized or centrally collected. Chronically slow payments to contractors, combined with a lack of legal recourse, resulted in higher prices, liquidity problems for vendors, and the potential for additional abuses. 177. As a result of the above weaknesses, donors typically followed their own (different) procurement and disbursement procedures and practices to ring fence their funds and thereby short-circuit the national system. Thus, numerous special accounts were created, but without an automatic transfer o f information to the MEF. While the Ministry eventually received information on these accounts and disbursements, information on aid supplied in kind was typically not available. Donors also used their own salary scales for local hires and participants in projects they supported. Several donors have also taken more drastic measure to bypass the system altogether - simply providing funds outside the budgetary process and ensuring project management by nongovernmental actors. While this addressed the short-term governance concerns of donors, it contributed to the weakening of the state to plan, coordinate and oversee service delivery inkey sectors. 52Including its failure to regulate certain types of public procurement, overregulation of some areas and the provision for the CSCCA's prior approval of contracts, a clear conflict with its other roles. 77 178. Transparency International's Corruption Perception Index (CPI) has typically placed Haiti among the lowest rankings worldwide. While a study by TI'S Haitian branch indicated the private sector's belief that corruption is the leading constraint on economic growth and inve~tment,~~ it also showed that it was widely tolerated. Government's limited control of financial data and the lack of any law requiring information to be made public meant that citizens had little or no knowledge of how public resources were used. Media coverage remained uneven and civil society groups were more focused on basic service delivery than specialized advocacy. Within government, control mechanisms - the judiciary, public prosecution and police, as well as those mentioned above - remained weak, little trusted and plagued by their own internal problems. 179. On the revenue side, the central government's revenue/GDP ratio declined to around 7-9 percent over the past decade from a level of 10 percent in the late 1980s. Haiti's tax/GDP ratio i s well below that inother low-income countries, reflecting the low yield from income, excise and trade taxes and leaving the country at a level that i s inadequate to support the country's development needs. This weak revenue performance i s due to a combination of the following factors: the narrow tax base resulting from a sluggish economy and widespread exemptions and tax incentives; weak tax administration and enforcement, especially in the provinces, at borders and ports; and significant tax evasion, smuggling and fraud. The limited government control outside of Port au Prince, exacerbated by the recent security problems, have resulted in the diversion of trade to the provinces where customs controls are weak or inexistent result in further loss of revenues. 180. One of the greatest impediments to successful economic governance in Haiti is the inadequate quality and quantity of human resources - the absence of a sufficient number o f qualified personnel, within the MEF and the sector agencies, to make the system work and the lack of a good incentive system to attract, maintain and motivate civil servants. The legal framework (a 1982 law and the Constitution) called for a modern, reliable and competent public service based on equal access to employment and selection on the basis of merit. However, the cabinet-level office charged with overseeing the system was never created and the necessary regulations on personnel administration were never drafted. A series of donor-sponsored studies indicate that a majority of appointments were made on the basis of political and personal connections and that in some cases agencies (e.g., the public enterprises) have been captured by their own staff. For those that did get a job, by whatever means, incentives for good performance were scarce. Public sector salaries lost fifty percent of their value between 1995 and 2000, and dropped more rapidly afterwards. There was little funding for training and limits on the possibilities for promotion for good performance through a formal career system. Donors, NGOs, and the private sector often attract the most qualified staff. The concentration of staff in Port-au-Prince aggravates the problem with programs depending on decentralized execution or enforcement suffering. Reliable information on the number of public employees does not exist, let alone on their job 53 La Fondation HQitage pour Haiti, L'Etat des Liux de la Corruption en Haiti, 2003. 78 titles, duties, and performance; there i s no standardized salary scale and remuneration continues to be determined inan arbitrary and nontransparent fashion. 181. The conditions described above were those confronting Haiti in early 2004 when the Transition Government joined forces with the donor community to prepare the Interim Cooperation Framework (ICF). The ICF identified strengthening economic governance and institutional development as a priority area. In line with the priority interventions identified in the ICF, in 2004-05 the Government took a series of steps to begin to address the above weaknesses. Many have involved updating the legal framework. Among those are the passage of a new Organic Budget Law (OBL, prepared and enacted under the previous Government, and publishedin August 2004) and a new Procurement Decree (published in February 2005). The former, in addition to establishing the outlines of a new budgetary process, also mandates the creation of a new accounting system, creates the position of contrdleur financier (internal ex-ante control) and a new internal auditing office (the Inspection des Finances), and removes the CSCCA' s ex-ante control functions. The Procurement Decree creates a Commission Nutionale des Marchb Publics (CNMP) with a revised mandate and full-time members selected transparently with input from both the public and private sectors. The decree also requires that competitive bidding practices replace direct contracting. A new civil service law and a new law for the organization of the central administration were drafted and enacted by decree in July 2005. Both the CSCCA and the CNMP have drafted and had enacted additional decree-laws and regulations to permit the implementation of their revised functions. The government i s also endeavoring to implement a new budget cla~sification~~and a chart of accounts that were approved by the previous government after many years o f preparation. 182. Institutional changes have produced additional progress. The vacant positions in the CSCCA have been filled. The CNMP became operational and standard bidding documents were prepared; the CNMP i s helping to introduce more effective procurement controls in the public administration. The recent hiring of an international procurement consulting firm will allow it to perform an extended range of its mandated functions, in particular those associated with strengthening procurement capacity in line ministries. Finally, advances in transparency have been achieved through the publication of lists of government contracts awarded and a supplier database, both available through the CNMP website. The FY2004-05 and 2005-06 budgets were prepared on the basis of the requirements of the 2003 Budget Law and ICF priorities, presented in draft to the donor community and civil society organizations for consultations, and approved by the Council of Ministers before the beginning of the fiscal year. The approved budgets and key budget execution data have been disseminated publicly. 183. Improvedbudget execution monitoring resulted in the significant reduction of the use of discretionary accounts to less than 10 percent of cumulative non-salary current public expenditures during October 2004 to early 2006. Since 2003, an automated 54 The current budget classification used in the budget law and in the SYSDEP system does not have a program classification (the current budget classification has a functional one, but the SYSDEP system does not), which makes it difficult to introduce the program budget approach desired by the authorities, and the hierarchy: program3sub-program+activity or project. 79 financial management system, SYSDEP, has been installed in seven ministries. While it so far only tracks expenditures, and then incompletely, it has greatly improved the timely input of financial information. The MEF is aware of SYSDEP's limitations and has sought donor assistance to expand its coverage and add additional modules to cover other functions. 184. A Coordination Unit in the Prime Minister's Office has taken on the human resource functions as a first step to overseeing the implementation of the new civil service decree. The unit has drawn up an action plan for meeting short and medium term human resource needs. Civil service salaries were increased by 30 percent in 2004, in consultation with the IMF, and an additional 15 percent increase is under The Ministry of Economy and Finance also established an Anticorruption Unit (the Unite` de Lutte Contre la Corruption, ULCC). As part of its efforts to develop and implement a strategy to improve governance and fight corruption in the public sector, this Unit has conducted a detailed comprehensive diagnostic survey (covering households, public, private and the NGO sectors) on the state o f governance and the perception o f corruption in Haiti. The Unit is also developing a staff training plan that focuses on international best practices. 185. As part of its efforts to improve the management and transparency of key public enterprises, with donors' support the Government hired international audit firms on a competitive basis to undertake financial audits for the national port authority (APN), the telecommunications utility (TELECO), the national electricity company (EDH); and for the financial audit, administrative audit and technical review of the metropolitan water authority (CAMEP). The Government also undertook a comprehensive accounting rehabilitation o f TELECO and EDH. The Fonds d'Entretien Routier (FER), created by law in 2003 to finance road maintenance, was made operational in 2005 (see Chapter 4 for details). 186. Despite the progress made in the last two years, much remains to be done. Reforms inthis area, as inthe previous section, need to be incremental and sustained over time. In a situation characterized by considerably dysfunctional systems and institutional weaknesses affecting all levels and sectors of government, reform priorities must be set. Even though the legal framework recently introduced does not embody all modern budget management principle^,^^ it represents a clear improvement over its predecessor and a first step ina multi-phase, incrementalprogram of reforms. 55While this improved the situation, even after the 30percent increase government salaries for technical directors were roughly one-third and one fourth, respectively, of those paid by NGOs and international organizations. Haiti, Bureau du Premier Ministre, "Renforcement des Capacite'sHumaines pour la Mise en Oeuvre du CCI, October, 2004. 56Such as those related to accountability, the medium-term fiscal framework, and the transparency of fiscal targets and performance. The OBL does not require the government to include a medium-term budget strategy in budget; calls for a budget based on inputs rather than results and does require performance indicators; it does not specify the responsibilities o f payment authorization officers vis-h-vis the parliament, nor those of managers of funds allocated to the ministries. 80 187. Itis recommended that the immediate focus of Government efforts be on ensuring the full implementation of the new legal framework, starting with central government agencies and then gradually extending coverage to offices handling financial management in sectoral ministries. Ensuring these core institutions function appropriately and in coordination is key for encouraging both a steady flow of resources (including those of donors) and their efficient use. This will require considerable effort to establish new functions and units described in the laws and meet their minimum staffing levels through the hiring and training o f core staff. It will also be necessary to develop more precise procedures for new functions and offices so that staff has adequate operational guidance and can be evaluated on its performance; develop standard formats for budget preparation and reporting, train staff in their use and monitor compliance. Strengthening revenue collection agencies will also be critical for the ability of the government to mobilize domestic resources for development priorities. It will also be important to explain the content and importance o f these efforts to nongovernmental constituencies and to buildsupport for the program of reforms from the broader public. 188. Budget formulation. Over the short term, the following priority actions are needed to enhance the quality of the budget. First, the normal schedule mandated by the law, under which the ministries have adequate time to prepare their annual budgets, needs to be applied, the MEF should state the indicative annual expenditure ceilings earlier and sufficient time for budget negotiations be allowed. Second, the "public interventions" item in the budget should be eliminated and financial expenditure from the ministries' own resources and the extra-budgetary accounts included. Third, revised procedures and manuals concerning budget preparation should be prepared. Over the medium-term, the government budget should be based on a multi-year framework. A functional budget classification, which i s currently missing, should also be introduced in accordance with international standards (IMF GFSM). More importantly, efforts should be made to achieve a more redistributive and efficient pattern of public expenditures. While the allocation of public expenditures i s inadequate for several sectors (notably education, health, infrastructure and rural development), this i s more the result of limited resources across the board than of serious inter-sectoral expenditure misallocations. Despite the limitedroomfor redistribution across sectors, a programbudget approach would increase the efficiency and accountability of public expenditures and help improve intra-sectoral allocations. The program budget approach will be a long-term objective in Haiti, but some preliminary steps to reach that goal could be taken now (see Box 3.3). The Government should also institute new procedures that embody the broad principles lacking in the OBL-Le., transparency, performance, responsibility, and consolidated cash management. 189. Expenditure execution. The Government should continue to keep unprogrammed expenditures through the "current accounts" at a minimum level and reduce the number of such accounts. In order to ensure that the recent reduction in the misuse of current account advances i s lasting, it i s recommended that the Government continue its efforts to reduce administrative lags in processing expenditure requisitions- which can be as long as two to four months. In order to implement the O B L and the proposed enabling decrees, there is an urgent need to prepare revised procedures and manuals concerning budget execution, update the internal regulations of the MEF's 81 departments, and train accountants and budget operators. Also, the Government should continue to expand SYSDEP's coverage to all spending ministries and government agencies, add additional modules to cover other functions (particularly a "government accounting'' module, a debt module and a human resources management module), and ensure that it can be accessed publicly, thereby making it a tool for transparency and governance. 190. Incorporation and monitoring of public investmentprojects. Although loans and grants from some donors were incorporated in the last two budget laws, this information was not comprehensive. Projects financed by donors within the framework of the ICF are not identified by the new budget classification, although this i s essential for implementation monitoring. Moreover, information on investment implementation i s scattered. Over the short-term, responsibilities o f the MPCE, MEF, spending ministries and the Strategic Coordination Unit of the Prime Minister's Office regarding project preparation, execution, and financial and physical monitoring should be clarified and coordination enhanced, and assistance provided to enhance capacities. For this to be effective, it would require donors to provide more comprehensive and timely information than they have to date-which the Government should pressure for. This would include furnishing monthly status reports on budgetary assistance provided to the government and on disbursements from their bank accounts. To enhance the current project information system, alternatives such as the model of the Development Assistance Database (DAD), which has been set up by governments with UNDP support in countries such as Afghanistan, Sierra Leone, Sri Lanka and the Maldives, or the Aid Management Platform, a web-based aid information-sharing system piloted by Ethiopia could be considered. Over the medium term procedures for project preparation and selection need to be developed and training for planning staff in sectoral ministries provided. 82 Box 3.3: Short-Term Priorities for Moving Toward a Program Budget Approach The Haitian authorities have expressed an interest in implementing a program budget approach over time. This can be defined as an approach to the budget process whereby: (i) i s a strong link between budgeted there funds and the results to be produced; (ii) is a system of information on performance; and (iii) there objectives interms of improving effectiveness are defined and are accompanied by measurable indicators. The transition to a program budget approach assumes the fulfillment of certain prior conditions, notably a stable macro-fiscal framework; technical capacity to define programs, establish performance indicators and assess performance; and clear procedures and respect of prevailing regulations. These prior conditions have not yet been met in Haiti, thus the full introduction of program budgets will be a long-term goal. However, several ministries, and particularly the ministries of Agriculture, Education, Health and Population, and Public Works, Transportation and Communications, are capable of formulating medium-term objectives, along with easily quantifiable indicators, and of attempting to incorporate them into budget programming. Thus, despite the current weaknesses of the budget system, preliminary steps towards a program budget approach could be taken. Needed actions that can be achieved over the short-term include: (i) better budget preparation (more realistic expenditure projections, definition of unit costs, etc.); (ii) description of macro and fiscal objectives, as well as the identification of priority sectors; and (iii) better control of budget execution. Also in the short term, pilot ministries could be designated for the definition o f programs and key performance indicators (including physical objectives to be achieved) over a 1%-monthperiod; the sectoral tables set up under the ICF provide a context in which this exercise can be done including other stakeholders such as donors and civil society. A pragmatic approach seems preferable to one that allocates resources to meet a predetermined target. Examples are provided below: The Ministry of Health and Population, to which 4 of the 8 Millennium Development Goals are relevant, can measure improvements in rates of infant and maternal mortality, for example. It can also monitor a number of objectives: numbers of healthcare workers, healthcare facilities per population, etc. The Ministry of Education, which enrolls barely 15 percent of the school-age population, could focus on program monitoring and school supervision. The Ministry of Agriculture can pursue the same goal in terms of infrastructure development or soil conservation management. The Ministry of Public Works, Transportation and Communications can use numbers o f kilometers of roads maintained and rehabilitated. The ministries could sum up the main components of their programs and performance indicators in a brief document for each ministry, and append it to the draft budget law. This approach would enable the CSCCA to assess the effectiveness of Government's action even in the absence of program budgets. Source: IMF,2005. 191. In the area of procurement, focus in the short-term should be on ensuring compliance with the new procurement decree across all ministries and public enterprises and enhancing compliance tracking mechanisms; development of training programs for sector procurement staff in standardized bidding documents; and the education o f potential bidders and the general public on procurement processes. Over the medium term, further improvements inthe legal framework for procurement will be required. 192. The Superior Court of Accounts and Administrative Disputes (CSCCA) should continue to be strengthened to play its new core mission, i.e., ex post controls of the Iegality o f the accounts and the effectiveness o f the government's expenditures. Inorder to enable the Chamber of Deputies and the Senate to exercise their parliamentary oversight, the CSCCA must prepare annual budget execution reports for the legislature. In its turn, the Government should adhere to the timetable of the OBL to compile and submit the accounts and the Budget Review Law. 83 193. Haiti's Parliament has a critical role to play in the economic governance agenda, most notably in budget approval and control. In other countries with new parliaments, donors have funded courses and workshops to enhance legislators' understanding of budgetary issues as well as help create permanent specialized staff and reference services. Beyond this, the MEF and other economic governance agencies should consider their own creation of congressional liaisons. Efforts in the future should thus incorporatethese elements as the economic governance reforms move forward. 194. Priority attention should also go to the revenue side (DGI and AGD). A recent IMFreview (2005) highlights the following channels for increasing revenue collection, which will be critical for mobilizing additional public resources for meeting the country's development needs. In the short term, the existing tax administration should be strengthened, with a main focus on enforcement being expanded to the provinces (about 70 percent of total VAT collection inHaiti takes place at the borders) and ports. For this, restoring security and the presence of customs in sea ports in and outside Port au Prince will be critical to avoid a diversion of trade activities to where the presence of the State is weakest. Second, tax administration could be improved by strengthening programs to fight tax evasion, improving collection procedures, and reorganizing tax agencies along functional lines. Over the medium term, progressive tax policies should be considered, including increasing specific excises, property taxes above some relatively high threshold, comprehensive coverage of personal income tax in the tax net, broadening of the VAT base (notably by repealingexemptions) and increasingthe VAT.57 195. Over the short term, the following priority actions are recommended to strengthen human resource management. First i s the creation of a permanent registry of all state employees and of an agency to oversee the application and further development of the rules regarding their appointment, remuneration, evaluation and overall career management.58 This agency should also guide the creation of human resource units in every ministry, public enterprise and other state agencies. Second, increased emphasis on merit appointments and promotions i s required. A pragmatic way to move in that direction would be to delineate clearly which positions should be awarded only on merit and which may also include political criteria59, and define rules for new merit appointments and promotions. Third, in order to better control the wage bill and eliminate wage arrears, personnel in spending ministries needs to be better managed, by immediately notifying the Treasury of personnel hired and departing from the civil service and using reliable monthly establishment data maintained at the level of each spending ministry as well as within the MEF; also, wage bill projections need to correspond to the exact number o f staff already employed. A fourth immediate task is to evaluate current staffing needs in terms o f a minimum and optimal scenario and to prioritize new hires and training of existing staff. As budget restraints may impede the filling or even creation of some positions dictated by key new laws (e.g., the creation of financial controllers and inspectors, staffing up of the new CNMP), it i s urgent that 57 At 10percent it is at the lower end of the range among low-income countries. 58 This could be achieved by expanding the human resources unit in the Prime Minister's office into a real Human ResourcesDepartment. 59 Normally, the line is drawn at the level of secretariesor directors with all positions below that level subject only to merit selection. 84 further analysis be done to: (1) determine where needs are critical; 2) where existing staff can be redistributed and trained for new functions; and 3) where new hires are most essential. Box 3.4: PublicSector EmploymentinHaiti The small size and very limited capacity of Haiti's public sector government contrast with the massive development challenge facing the country. The ratio of Haiti's public employment to population has remained below 1 percent as compared to an average of 2 percent in Africa, 2.6 percent in Asia and 8-9 percent in small island developing countries. As a percent of total population, public employment in Haiti ineducation, healthcare and police is roughly a third or less than that in low-income countries. The public sector wage bill in Haiti is also very low. Between 1998 and 2003, the public sector wage bill fell from 4.5 percent of GDP to 3.3 percent of GDP, well below the international average of 5.4 percent of GDP. During the same period, public sector wages in Haiti fell by one third in real terms. However, the public wage bill takes up a significant portion of the government budget in Haiti (33 percent of government revenues and 26 percent of expenditures), reflecting very low overall revenue and expenditure levels. Also, public sector wages have fallen in real terms. The small size of the public service has severely constrained the government's ability to enforce the rule of law, the tax system and property rights, as well as to assure physical security and provide basic social services. Low public sector wages have also encouraged rent- seeking and corruption, and diminish civil service productivity. Civil service compensation should be broadly comparable with that of the private sector. Comparability- adjusting for security o f employment and differences in other non-wage benefits of employment-is important to allow the government to recruit qualified staff and avoid corruption. However, according to a recent survey conducted by the Prime Minister's office, public sector salaries in Haiti appear to be significantly below private sector averages. Employees in higher skill managerial positions earn 40 percent of their private sector counterpart's salaries, while employees inlower skill positions earn 70 percent. An efficient public sector requires more than adequate staffing and pay. Efficiency typically also requires: (i)makingthecivilserviceaffordableandinlinewiththeroleagovernmentassignsitself; (ii) providing the incentives, skills and motivation to civil service employees to perform; and (iii)enhancing civil service management and accountability at all levels. As discussed above, the size of the civil service in Haiti is very low even taking into account the country's level of development and size. This raises questions about the ability of the government to deliver essential public services, notably security, education, health and basic infrastructure. The low level of public wages also raises doubts about the government's ability to attract and retain qualified staff and avoid corruption. Dealing with these issues in a way that supports sustained public sector efficiency will require increasing government revenue and implementing reforms involving systemic restructuring of the civil service. Source: IMF(2005). 196. Developing capacity in the ministries and agencies requires seriously and systematically addressing the difficulty o f attracting capable and qualified staff to the public sector, given that the state cannot pay competitive wages now or inthe foreseeable future. Emigration has depleted the local talent pool and international donors and NGOs pay much higher salaries, which enables them to attract much of the expertise that remains. Salary complements paid by donors to staff within project management units have had mixed results--they have a positive effect on this staff`s motivation and time spent at work but have limited, or even a negative, impact on other civil servants. In addition, it has led to the emergence of project implementation units staffed with specialized consultants that, while reducing the risk of mismanagement of funds, have not 85 allowed for the creation of stable institutional capacity. Over the short term, it i s recommended that the donor community agree with the Government to regulate the criteria and pay levels for salary complements they provide to civil servants. This should start with a complete inventory of the positions that are being paid by donor programs and followed with the issuance of regulations to establish a transparent way to top up salaries which would allow for civil servants to perform. A monitoring agency or joint commission should be set up to oversee the implementation of this process. Also, the creation of PIUs should be limited and, eventually, phased out. As a guiding principle, PIUs should be created only where the project's executing agency does not have the capacity to implement complex projects, or where the implementation of the reforms supported by the projects depends on a swift and politically-neutral decision making process. Whenever created, the PIUs' terms of reference should be very precise and, while focusing on the implementation of project activities, they should also promote the creation of permanent capacity within, and the transfer of institutional and technical know-how to, ministries and agencies. 197. Overt the medium to long term, addressing the issue of how to attract and retain sufficient numbers of qualified personnel and avoid conuption will require increasing Government revenues, gradual increases in salaries consistent with macroeconomic stability and implementing reforms involving systemic restructuring o f the civil service. The latter would involve inventorying of all positions and the elaboration of a simplified job classification scheme, a standardized salary grid, a description of duties and of needs based training programs within the civil service, and exit strategies for remaining PIUs. After a first rough cut at the universe of employees, further refinements could be done on an agency-by-agency basis. VI. Service ProvisionandSupervision 198. Building ministerial capacity to oversee and evaluate basic service delivery, whether by public or private providers, will be critical for Haiti's development. This section will examine the role of the state in delivering basic services as well as the importance of a more decentralized, territorial approach to the programming of these services. 199. Delivery of basic services in Haiti has two characteristics: it is either dominated by private actors with potentially negative impacts on quality, access and effectiveness; or, it i s undertaken by inefficient public enterprises. In the education sector, non-public actors dominate. Haiti ranks second in the world in terms o f proportion of students enrolled in non public schools. These schools constitute 92 percent of all schools in Haiti, the large majority of which do not receive public subsidies (see Chapter 5). Their dynamism is largely due to a gap left by the public sector. As Box 3.5 shows, it i s not unusual infragile states to find that such basic services are delivered by the private sector andsuch arrangements serve to fill important gaps, increase access and provide enhanced accountability through community participation and continuity during crisis periods. However, there are also severe potential negative impacts in this private delivery approach including low quality services, limited access due to the inability of poor households to pay fees or service related costs, lack of transparency about costs, poor 86 coordination between actors, lack of financial sustainability and inequities across the sector. As such, the entire concept of public goods or services i s lost. Box 3.5: Contracting health care to NGOsin Cambodia C`ambodia experienced decades of conflict resulting in a depleted human resource base and no functioning public service infrastructure, extremely poor health indicators, a rudimentary public health care system and very low usage of facilities. Despite this, levels of private expenditure on health care were relatively high, without translating into good health because much of the expenditure was on inefficient use of health services and purchasesof pharmaceuticals. [n order to address the problem of poor service provision by the state, a government study (supported by the Asian Development Bank) piloted two options. In the first case, health services were "contracted out" to NGOs or private organizations that were then held accountable for performing the contract. In the second case, NGOs were "contracted in", Le., based within the Ministry of Health but responsible only for providing management support to civil service health staff, whilst recurrent operating costs were provided through normal government channels. Several "control" regions, where the state continued to operate without NGO involvement, were also included in the study. Quantitative evidence two and a half years after the NGOs started working showed the greatest improvement (according to a number of pre-defined indicators, essentially contact rates in both curative and preventative services) in those districts where NGOs were contracted out, followed by areas where they were contracted in, and finally for the control group. In addition, the contracted-out model was the most efficient model, requiring the least out-of-pocket expenditure by users (especially low-income users), and also providingthe greatest equity and access for the poorest groups. The study concludes that not only is government contracting of the provision of health services to non governmental entities feasible, but it can increase coverage of health services in a short time, improve equity and access, deliver interventions faster than conventional government service delivery mechanisms, and improve efficiency. However, the study stresses that there are four key conditions to make this approach successful: pre-determined and objectively verifiable performance indicators measured prior to contracting, coupled with well defined and contracted performance targets; political support for contracting at both central and local levels; civil service arrangements that allow government health care professionals to work for NGOs at market wage rates; and "management by results", wherein contractors are given maximum latitude to achieve the predefined indicators. Source: Bhushan. I. Keller. S and Schwartz, B (2002). 200. In these situations, transition strategies that find a more appropriate role for the state in partnering with private actors are necessary. For some years, the Haitian state has recognized the importance of forming public-private partnerships in core areas such as health and education. In the health sector, the state has come to agreements with private actors around the provision of child immunization programs and pre-natal care. Inthese cases, the state has provided vaccines or medicineshupplements, but service delivery has remained the responsibility of the non state service providers. Inthe education sector, the state has tried to carve out a role for itself as a regulator through the accreditation 87 system6', and by maintaining control of examinations. At present, only students in accredited schools have the right to take exams, a leverage that the state tries to use to motivate non-public schools to register. However, most schools are able to circumvent this restriction through a variety of ways, and there are major bottlenecks in the accreditation system. Private providers are also conscious that no penalty action will be taken if they do not register. As a result, less than one quarter of all non public schools in Haiti is accredited, mostly all in urban areas, creating serious challenges for any state initiative to set quality standards or coordinate the system. The Haitian state i s responsible for carrying out inspection visits to all schools and health providers, both public and non state, in order to ensure standards are being observed. However, service providers complain that officials are rarely available and do not visit the schools and clinics. 201. Both the education and health sectors have national strategies that recognize the need for a stronger role for the state to oversee and provide incentives and support to private actors (see Chapter 5 for education). Although attempts to do this under the transition government in the education sector through the National Partnership Office (ONP) have not advanced as quickly as hoped, the concept behind the ONP remains a valid one and could provide a model for other sectors. However, as a review of LAC experiences of public-private partnerships notes, there are indications that in the longer run the sustainability and full effectiveness of public-private partnerships depend on whether the rest of the state apparatus improves its capacity and becomes more performance oriented.61 202. In the infrastructure sectors, the Haitian state has played a more direct role in service delivery through public enterprises but with unsatisfactory results. There the performance of state owned enterprises is characterized by lack of maintenance, poor quality services and a significant drain on scarce public resources, and the opportunity for illegal and corrupt practices. Thus, despite the stronger presence o f the state in this sector, as opposed to the social sectors, service delivery i s also weak. Chapter four and five discuss specific options for public-public partnershipsininfrastructure and education andoffer short and mediumterm recommendations for pursuingthem. 203. One way to enhance development efforts in Haiti i s through a more decentralized, territorial approach to the programming of basic services. This would help better integrate the country and organize its development, building on the growth potential of the various regions. Despite several attempts in the past, Haiti has no such strategy. Untilnow, the capital has taken the lion's share of investmentresources, contributing to an unbalanced development pattern, with underdeveloped secondary cities and largely neglected rural areas. Haiti's regions are heavily populated (the country has among the highest population densities in the world). Some regions have strong growth potential, particularly in the agriculture sector.62 Nevertheless, economic growth is constrained by the lack of geographic cohesiveness and the fragmentation of the country. Territories 6o The Ministry of Health would like to implement an accreditation system along these lines. World Bank, 1995. 62 A 2005 World Bank study has identified a number of regions with growth potential for specific agricultural products, including coffee, mangoes and vetiver. 88 remain poorly integrated with limited rural-urban linkages, notably due to the advanced deterioration o f means of communication, and roads inparticular. 204. A territorial approach to planning could identify growth poles (eg., urban centers that are important consumption, processing or trading places, as well as rural production areas) and set out ways to integrate them. Such a strategy could also promote social cohesiveness and help restore hope by addressing the needs of both the urban and rural poor. There is thus a need to take stock of all the various existing initiatives to help design a broader scale framework for territorial planning. This would help define a coherent and integrated development strategy for the various regions, prioritize among investment alternatives and determine the critical mass needed for combined investments (or the "pump priming" i n ~ e s t m e n t ~to~ )have an impact on growth and poverty reduction. Several initiatives have been launched to define development strategies at various sub-national levels (departements, "micro-regions", communes). However, a common framework at the national, regional and local level i s needed to better fit together such strategies. In the longer run, this could prepare the ground for a decentralization process, although a significant strengthening o f the institutional capacity of government institutions i s needed as a pre-condition for that. 205. Haiti now has an institutional capacity gap for planning between the departments and the communes, despite this being the appropriate territorial level for growth-based strategic planning of infrastructure investments. Existing administrative sub-divisions (departements, arrondissements, communes, communities) often do not reflect the country's economic reality. Communities do benefit from the dynamism of grassroots civil society organizations, which have become an effective substitute for formal governmental administrations. But these have a limited effect on growth because of their small size and relative inefficiency (lack of economies o f scale). On the other hand, departments, as the largest sub-national divisions, are too big to be fully accountable to economic actors, particularly ruralproducers, because of the lack of territorial integration and the poor state of communications infrastructure. The right level for growth-based planning of infrastructure investments is usually some ad hoc intermediate arrangement, such as groups o f communes, in which no institutional capacity for strategic planning exists. Municipalities (communes) remain very weak although some have received donor assistance to prepare communal development plans. The arrondissements barely exist in practice. 206. Local and regional stakeholders still have limited participation in strategic decisions regarding the development of their territory. The alignment of public investments with local needs, with the prospects of removing bottlenecks to growth, i s highly dependent on the ability of all key stakeholders to express their needs. Currently, there are few fora in which stakeholders can contribute to territorial development strategies, except at the grassroots level within Community-Driven Development (CDD) initiatives. At the level of departements, structures called tables de concertation 63A "pump priminginvestment" is one undertaken,usually by a government,to stimulateeconomic activity ina depressedregion. 89 departementale, under the authority of a Delegate appointed by the central government, are being set up. But these are currently too weak to engage in strategic planning. 207. A new Bank-financed project will implement on a pilot basis a new approach to territorial development planning, structured around a transport investment in two Haitian micro-regions (see Box 3.6). This acknowledges that both local and regional stakeholders should be involved in the strategic planning process for infrastructure. Local stakeholders (mayors, community organizations, small entrepreneurs, cooperatives) have a better understanding o f the key impediments for income-generating activities. However, they may miss the broader perspective (e.g., potential for export, situation in neighboring territories, long-term risks such as environmental degradation) due to limited access to information. This can be brought by regional stakeholders (traders' associations, regional representatives from line ministries) or, for certain topics (e.g., market studies) by national or external experts. A strategic planning process, with the participation of local and regional stakeholders and with the proper support from the donor community can thus help bridgethat gap. Box 3.6: A pilotfor strategicplanningof transportandterritorialdevelopment Within the strategic objective of restoring hope for the poor, the Transport and Territorial Development Project (TTDP) was designed by the Government and the Bank in early 2006 to identify and implement a combination of basic infrastructure investments that could promote growth in selected micro-regions. A micro-region is a territory, usually a combination of communes, with economic coherence in terms of growth potential. The TTDP is intervening on a pilot basis in two micro-regions (Dondon-St. Raphael in the north and Thiotte-Anse h Pitre in the south-east). Ineach territory, a pump priminginvestment (a main road) has been identified, the rehabilitation of which would make a critical contribution to improving accessibility and linking agricultural production areas to markets. A methodology was then developed to prioritize and implement other smaller-scale basic infrastructure investments that could enhance impact on growth. This "territorial development process" is organized in two stages: local and regional. At the regional level, the process is led by a subset of the table de concertation departementale. A comprehensive monitoring and risk mitigation framework was also established. 208. Developing a territorial approach to local development will be a gradual process. Any strategy for territorial development planning should be designed to take account of the various initiatives underway, and the characteristics and needs of the country. Ultimately, a sound territorial planning framework could become the platform on which donors develop and coordinate their interventions, particularly in the area o f basic infrastructure. Proper attention should also be paid to promoting sustainable growth and to the monitoring of possible adverse impacts of investments. Guiding principles for implementation should include: work at a territorial level intermediate between the departement and the commune; promote coordination with the planning processes engaged at other territorial levels (e.g., CDD initiatives, communal development plans, regional plans like schemas directeur d 'amknagement du territoire, national plans and strategies when available); 90 0 design a methodology to prioritize investments that have the greatest impact on poverty reduction and growth and that can be sustainably operated and maintained; 0 ensure that key local and regional stakeholders are involved in the process to prioritize infrastructure; 0 make sure that investments are sufficient to overcome possible "threshold effects" and to reach the critical mass that will have an impact (either a pump priming investment like the improvement of a key transport corridor or a combination of bundled infrastructure investments); and 0 monitor adverse impacts and pay due attention to sustainability and risk mitigation. 209. Encouraging citizen involvement inevaluating andtracking services can also help improve service delivery. Citizen involvement creates demand for reforms as well as provides important feedback to government ministries on how services are being delivered as perceived by their users. More specifically, tools such as citizen and community scorecards used in many countries are now leading to the establishment o f service benchmarks by ministries with incentive systems that reward those who improve their services most, based on user feedback. 210. That said, agencies must still develop their own ability to monitor delivery of basic public services, whether the direct providers are public or private entities, This can be seen as an extension o f the economic governance function--to ensure that public resources are spent where intended but also that they produce outputs of a reasonable quality. Since many o f these services are provided outside of the direct vision of the central agencies, this means developing a capacity to evaluate results at a distance through the collection of information and periodic on-site supervisions. In Haiti, the situation i s complicated by the array of private service providers with no direct relationship with the state. The setting up of monitoring and evaluation systems that are linked to programmatic budgets are a medium term goal but, as indicated in Box 3.3, preliminary steps towards establishing some basic service monitoring could be take. VII. Transparency,ParticipationandAccountability 211. As noted above, there is limited experience at inclusive, transparent or responsive governance in Haitian history. There i s plenty of evidence that points to the positive links between inclusive governance and sustainable, long-term pro-poor growth (Keefer and Kray, 1998). Engaging citizens in decisions that affect their lives i s often described as an end in itself, although it i s now also widely recognized to be good for development effectiveness, service delivery and providing checks and balances on potential abuses (World Bank, 2000 and 2003). Inclusive governance i s of particular importance in Haiti which i s plagued by deep inequalities and institutions that are too often ineffective at providing services to the poor and captured by powerful groups. Governance i s often analyzed by looking at the both the supply side, or public action, and the demand side, or citizen action. 91 212. On the supply side, it is clear for many of the reasons discussed in the Political Economy of Haitian Development section that in the past the Haitian state has not provided information or the opportunity for citizens to participate in policy dialogue. This is in part because the politics of exclusion has not put a high value on collecting information, programmatic planning, and monitoring and evaluation in the public sphere. Thus, by moving towards a more efficient results-based approach to economic governance, opportunities for participation of citizens in the policy process will inevitably increase. 213. On the demand side, it is also true that there has not been a well orchestrated insistence by citizens that traditional governance mechanisms be changed. One reason for this may be that the clientelistic logic extends to many citizens, who may be more interested in what their contacts within the state apparatus can get them, than in broader governance issues. Another reason may also be that Haiti's legal framework for civic engagement i s somewhat limited. Haiti,for example, still has no Freedom of Information Act (FOIA) nor much identified demand for the types of information it might furnish to the population. A third reason is that the vast majority of Haitian CSOs i s engaged primarily in service delivery andhas had less opportunities to work with the state around large scale, policy issues. It i s thus not surprising that there is little capacity among Haitian civil society for analyzing important information such as budget performance, or limited opportunities for policy dialogue at the national level. This low capacity, along with the state's institutional weaknesses, constitute tremendous barriers to changing the situation, but untilthey are overcome, the legitimacy gap will remain, or will be breached only by opportunistic, andinthe end, self-defeatingpopulism. 214. Exiting from this vicious circle thus requires cultural changes among the citizenry at large and within the public sector, as well as capacity building on both sides. Public demand for information will need to be created along with the ability to use what i s disseminated. Here intermediaries -- NGOs, the press, research institutes - will be important on both counts. They can help citizens understand the importance of access to information on state actions and assist them in interpreting the information they receive. InHaiti, there is currently an enormous enthusiasmamong a few NGOswho havebegun to engage in monitoring efforts around the economic governance reforms (see Box 3.6) and who are interested in engaging in more specific initiatives around budget planning, expenditure analysis and monitoring, and evaluation of service delivery. However, enthusiasm not withstanding, these organizations now need to be exposed to specific tools that exist to undertakethese initiatives and to ensure that such exercises are shared by a broad, and not exclusive, group of civil society actors. Without this gradual growth indemandanduse, the state's responseis likely to belimited. 215. On the supply side, the first step is encouraging the collection of informationfor internal and external use. What i s not collected cannot be distributed, but what is not collected also cannot be used to improve performance. Both require the introduction of mechanisms to allow the collection of basic data on state activities and outputs. Also, opening the budget process to some input and oversight by citizens will improve its relevance and transparency. Similarly, donors can promote transparencyby ensuringthat any assistance generates and disseminates information on outputs, both of the donor 92 financed program and of the entity receiving the assistance. Support for the passage of a Freedom of Information Law or Act (FOIA) should also be considered as well as providing access to information for poor and rural communities through community radio and other forms of mass media. The passage of a Law on Asset Declaration and Disclosure would also be important to ensure transparency and accountability o f public sector employees. 216. The next step on the supply side i s the creation of channels for receiving and responding to citizen reactions (voice), either to general policies or to the quality of specific outputs. In some countries, this has been done b y starting at the lowest levels of local government, where citizens are better able to assess performance and policies. While this i s complex, due to the constitutional provisions for the creation and staffing of local governments, there are some positive experiences in community development approaches to local development around the world. Another venue for giving local communities voice in decisions that affect the quality of their lives i s through the territorial approach to the programming o f investments described above. At the national level, the sectoral tables established under the ICF provide a venue for collaboration between government entities, civil society and donors. While some of these sectoral tables have not been as active in the past as they might, they provide an useful multi- stakeholder venue to support a move towards a more inclusive and results-based approach to service delivery. Such structures have worked well in countries like Honduras where they have served as effective ways of coordinating international and national actors around common objectives in, for example, reconstruction efforts. A first step towards improving the functioning of these tables would require funding and appointing active and effective coordinators to those that currently lack them, and developing guidelines for the tables' responsibilities and reporting requirements. Also, participation of a wider range of stakeholders would enhance their f ~ n c t i o n i n g . ~ ~ 217. In 2005, the Government established a civil society oversight mechanism linked to the Ministry of Finance to provide transparency and accountability over the implementation of the economic governance reform agenda. As shown in Box 3.7, this mechanism produces regular reports monitoring key reforms; these reports are discussed between civil society and the government and, upon approval, are released to the public through the MEF's website. This mechanism, while young and still gaining capacity, provides another means for the new administration to involve civil society more broadly in oversight activities and to support their strengthening by linkingto groups outside the capital. Involving citizens as monitors o f government output i s one of the several mechanisms that might be used to increase participation, improve communication, and encourage greater transparency. Given the low development of any o f these functions in present-day Haiti and the many other developmental priorities, it bears mentioning that the goal i s not an immediate transformation to a fully participatory society, but rather the gradual creation of a series of viable mechanisms for advancing the process and the creation of some societal checks on what would otherwise be a largely top-down effort. 64 For further ways to enhance the functioning of the sectoral tables, see COCCI, 2005. 93 Box 3.7: Citizen Oversight of Economic Governance Reforms Haitian civil society groups monitor and report results on the implementation o f the key components of the government's Economic Governance Reform program. The monitoring mechanism consists of an independent Oversight Committee made of twelve notables representing eight sectors of Haitian civil society as well as an operative structure consisting of a technical Secretariat linked to the Ministry of Finance and five Civil Society Organizations (CSOs). The CSOs carry out the monitoring of the five areas of reform: (i)budget preparation and execution, (ii)anti-corruption and public procurement, (iii) road maintenance fund and public enterprises, (iv) public-private partnership in health and education, and (v) communications. Upon termination of the monitoring reports, these are submitted to the Oversight Committee, if deemed technically sound by the Secretariat. The members o f the Committee discuss the monitoring results with the government before releasing them to the public through the Ministry of Finance's web-page. Due to little experience with social accountability exercise, creation and setting up of the mechanism faced several delays. However, since December 2005 the mechanism is fully functional. In the period January-April 2006 three reports were produced summarizing the monitoring reports prepared by the CSOs. After a steep learning curve, quality of these reports has improved over the last two months. The following are recommendations to further strengthen this mechanism and its impact: Link indicators to outcomes: So far, monitoring exercises focus on processes foreseen under the economic governance reform agenda (passing of laws, decrees, launching o f bids, signing of contracts, etc.). Revising indicators based on output would guarantee a more performance oriented monitoring of public services. 0 Capacity building: Training for members of the Consultative Committee as well as the monitoring organizations exposing them to methodologies to establish indicators in their respective sectors, and to monitoring tools to track these would enhance reliability and effectiveness of the monitoring exercise. 0 Expand to the local level: Expanding the mechanism to the local level would draw more attention to areas where the state, at present, i s virtually absent and provide an opportunity for more diverse civil society participation. Dissemination: Impact of the monitoring exercise would be increased if results were disseminated more broadly through local media, the radio, etc. Also, once in session, members of Parliament should be informed by the monitoring mechanism and appropriate channels forged Institutionalization and Sustainability: An institutional mandate would ensure the mechanisms' functioning and sustainability in the long-term, and provide for a space for civil society participation in policy dialogue. 94 CHAPTER4 BUILDINGINFRASTRUCTURETO LAST Dramatic improvements in Haiti's currently poor levels of infrastructure quality and coverage are essential to restart economic growth and reduce poverty and inequality. The underlying argument of Chapterfour is that a two-pronged approach is needed for infrastructure reform in Haiti: on the one hand swqt execution of projects by donors to produce rapid results, in theform of actual service delivery to thepopulation and, on the other, medium to longer-term improvements in public capacity to pursue not just sustainability of investments, but a more effective and transparent government role. For the short term, this chapter discusses the need for more effective donor involvement. While some important successes - such as the setting up of the road maintenancefund - were achieved recently, the lack of an integrated, strategic approach and poor coordination in financing and implementation have slowed progress. The chapter also considers the need for urgent efforts, built around infrastructure, to address the socio- economic causes of violence in Haiti's slums. Efforts to strengthen thefunctioning of the Road Maintenance Fund to ensure effective road rehabilitation and maintenanceare also urgently needed. The medium to longer-term emphasis on governance and capacity strengthening builds first on the main lesson of past interventions in infrastructure: that works are not sustainable without maintenance, which requires institutional capacity to organize and pay for it. The analysis also considers the need to strengthen the minimal functions required of the state in infrastructure: designing strategies and policies, setting and enforcing the rulesfor public and private operations, and coordinating national and international actors engaged in the sector. I.Theurgentneedforinfrastructureimprovement 218. Haiti's infrastructure indicators are the worst in the Western Hemisphere and among the lowest in the world, usually below Sub-Saharan African averages (see Table 4.1). Less than a third of households have access to electricity, while around 5 percent of roads are in good condition. Only 11percent of households have piped water. There are just 150,000 fixed telephone lines for more than 8 million people (penetration of 1.8 percent). There are no sewerage systems or wastewater treatment facilities and few arrangements for solid waste disposal. Poor quality exacerbates the situation: electricity i s only available for four hours per day, on average, and roughly half the telephone lines do not work. 219. Dramatic improvements in infrastructure are essential if economic growth i s to resume and be sustainable, as present conditions seriously limit output potential. As discussed in Chapter 1, Haiti's inadequate infrastructure clearly has a huge impact on the productivity of firms, and on the simple ability of rural and urban households to generate income. The lack of a reliable electricity supply curbs the possibilities for economic activity of all but the few who can afford their own generators, and imposes additional costs for those who can. The deteriorated road network impedes access to markets, especially in rural areas. Expensive and inadequate port and airport services hamper trade. And poor telephone and internet services prevent communication with suppliers and clients. 95 Table4.1: Haiti's InfrastructureCompared Latin America and Caribbean A rica 1,503 +--t+ Sub-Saharan I Haiti I Electric power consumption (kwh per capita) Improved water source (% pop. with access) Total 89 58 67 Rural 69 45 55 Urban 96 82 88 Improved sanitation facilities (% pop. with access) 74 36 34 Total 44 26 23 Rural 84 55 52 Urban Fixed and mobile phone subscribers (per 1,000 people) 497 84 64 Internetusers (per 1,000 people) 115 20 59 220. Many studies find that infrastructure improvement drives economic growth (see Calderh and Servh, 2004) as well as poverty reduction. In addition, infrastructure investment has diminishing marginal growth returns (Bricefio-Garmendia, Estache, and Shafik, 2004), which implies that a country with infrastructure as poor as Haiti's stands to gain the most, in terms of growth, from any improvement. While economic growth i s in itself a powerful force for poverty reduction, infrastructure improvement tends to be pro- poor in its impact. Infrastructure development in poor regions reduces production and transaction costs (Gannon and Liu, 1997). And in poor rural areas, it expands job opportunities by lowering the costs of accessing product and factor markets (Smith et al., 2001). Those who gain access from any service expansion tend to be poorer than those who already hadit, particularly inrural areas. 221. Infrastructure services are also significant economically for the direct contribution of infrastructure companies to GDP and for the employment that infrastructure projects and services can offer. Rapidjob creation is a central aim of many infrastructure projects under the InterimCooperationFramework inHaiti, inthe face of the country's extremely high unemployment. However, and particularly in the case of public infrastructure companies, the aim of job creation must be balanced with the need to pursue greater operating efficiency. Indeed, Haitian public companies are generally overstaffed, largely as a consequence of political pressure to use these firms both as a social safety net and as a channel for political patronage. 222. Improving Haiti's infrastructure i s also essential to promote health and safety and raise living standards. Access to clean water and sanitation curbs disease, which reduces child mortality, and raises attendance at schools and workplaces. The United Nations 96 estimates that around 80 percent o f disease in developing countries i s due to contaminated water. A recent study in Brazil (PEA 2005) found that increases of 1 percentage point in access to treated water and sewerage reduce child mortality by around 3 and 6 percent, respectively. Haiti's child mortality rate o f 118 deaths per 1,000 i s far higher than the regional average o f 33, andcloser to sub-Saharan Africa's o f 171.65 223. Strengthening the state's capacity to provide infrastructure services i s a core element of state building in Haiti. This i s bothbecausethe provision of public goods i s a central, defining, state function, and because the current condition of infrastructure institutions seriously weakens the current state in other ways. The electricity sector i s a massive drain on state resources, which undermines attempts to pursue fiscal sustainability (as well as leaving fewer resources for other basic services). On the other hand, the telephone utility TELECO (much less profitable now than in the past) long served as a cash cow, reducing the government's reliance on taxation and budgetary processes, and consequently its accountability. The poor functioning of ports and airports hampers revenue mobilization efforts, increases transport costs and facilitates illegal drug traffic. And corruption and rent-seeking have long been rife in infrastructure companies and contracting. 224. Significantly, durable improvement in infrastructure-and really building the state-therefore requires that infrastructure aid, projects and policies focus more on institutional and governance aspects, and seek to engage and repair (or eliminate if appropriate) the flawed structures and organizations that already exist for infrastructure provision. Such a task i s complex, as the example of the ports sector illustrates (Box 4.1). 652003 data, from WDI indicators. 97 Box 4.1: The multiple challenges of reformingHaiti's portssector Ports are critically important in Haiti, due to its geography (on an island) and its dependency on external support. Most trade transits by sea, with 80 percent of port activity concentrated in Port-au-Prince. Ports are also vital for domestic transport, as coastal navigation (cabotage) i s often easier than using the 3eteriorated roads. Indeed, sea transport accounts for 18 percent of the total traffic of goods within the :ountry, a high proportion by international standards. Ports are also significant fiscally, for the customs revenuescollected. However, the port sector is uncompetitive, inefficient, politically dominated and riddled with corruption. The obstacles to reform are considerable. Physical infrastructure i s inadequate. Facilities are seriously dilapidated and much port equipment is obsolete or out of service. The national port authority (APN - AutoritC Portuaire Nationale) cannot afford to upgrade or replace it. The IDB is commissioning a study of the physical infrastructure needs of the Port-au-Prince facility, with a view to addressing them. Political pressure has led to frequent management changes and serious overstaffing. APN had four Directors General in 1988, a particularly unstable year, and a further seven in the period 1992-1997. The body now employs roughly 1,800 people (three quarters in the port of Port-au-Prince), up from around 450 in 1985, despite a drop in cargo handled of approximately 30 percent over the period. While competitive ports abroad function around the clock, normal operating hours at Port-au-Prince are 7.30 am to 4.00 pm. Given existing traffic levels (the port of Port-au-Prince handles between one and two million tons per year), labor productivity i s extremely low, at about 1,250 tordemployee. In comparison, an average-to-high performance port like Puerto Quetzal in Guatemala has a productivity of 8,250 tordemployee. But charges are among the highest in the Caribbean, which drives many importers and exporters to use the more competitive ports of the neighboring Dominican Republic instead. The basic port charge per container is US$350-370 in Port-au-Prince, the highest in the Caribbean (where the average i s around US$250). There are other serious governance issues. While APN officially administers all Haitian ports, numerous private ports and operators have sprung up, over which APN has little control or enforcement capacity. Indeed, the Maritime Agencies (private transport companies) are heavily indebted to APN. Few safety standards are in place at the ports. APN's port management and financial systems are not computerized, and procedures involving transporters are also entirely manual. Not only is this slow and cumbersome but it hampers transparency and creates greater opportunities for fraud, which APN management concedes to be a serious problem. GoH has launched a financial audit of APN, under the Bank-supported Economic Governance Reform Operation, which is a first step toward reform in this area. The ports labor union is powerful, highly politicized and resistant to reform. In addition, the atmosphere at the Port-au Prince port is violent and volatile, with a number of shooting incidents reported recently. This security situation makes physical access to the port difficult and dangerous, which impedes the effective engagement of donors and other external advisors in port reform, as well as tariff collection. 11. Lessonsfrompast infrastructureprojects: maintenance, capacity building and political will 225. Years of heavy donor inflows have had little lasting impact on the quality of Haitian infrastructure. Particularly in the case of road projects, this has been due largely to a lack of maintenance, which in turn i s due to the incapacity of public sector institutions in the country to plan, fund and implement regular maintenance activities once project funding dries up. The World Bank approved eight road projects from 1956 to 2002, with most funds going to the National Road Nl. Nonetheless, by 2002, following the complete breakdown of the regular road maintenance program in the late 98 1980s, it still took 7-8 hours to cover the 250 km between Port-au-Prince and Cap Haitien, while 55 km were nearly impassable (World Bank, 2002). Less than 10 percent of Haitian roads receive any kindo f regular maintenance. 226. At the project level, this failure to build capacity i s reflected by an almost total lack of sustainability. A 2002 Operations Evaluation Department (OED) assessment of World Bank activities in Haiti found that, of 18 Bank projects rated for institutional development impact (IDI) (all sectors), 3 had substantial ID1and only 1 was rated as sustainable. Overall, the report concluded that the Bank's assistance program "did not tackle the critical institutional and governance impediments to effective development and aid, and almost none of its objectives were achieved." It added that "the large amount of aid that poured into Haiti in the mid-1990s accomplished little and may even have had a negative impact by overwhelming government capacity and creating opportunities for rent seeking." 227. B y not fully engaging and attempting to strengthen the public sector, in the past donors have exacerbated the problem of public sector weakness, as the Interim Cooperation Framework (ICF) June 2004 summary report acknowledges: "The donors recognize a lack of coordination, o f consistency and of strategic vision in their interventions. These donors have often set up parallel project implementation structures that have weakened the State, without, however, giving it the means to coordinate this external aid and to improve national absorptive and execution capacities." 228. If current and future donor involvement is to have a significant and lasting positive impact on the quality and coverage of the country's infrastructure, donors must seriously help address the issue of buildingpublic sector capacity. This goes far beyond providing for maintenance. Much infrastructure i s in the hands of (dysfunctional) state companies - and i s likely to remain so. And the government should have an effective role in setting and enforcing the overall framework andrules for public and private operators, incoordinating donors, developing long-term strategy and pursuingsocial goals. This is not to advocate an expansion of government ownership and control. Rather, private participation should be fostered where possible and the state should remain in a limited role, but a clearer, more transparent and more effective one than at present. 229. Institution-building i s a gradual and difficult process, and can only be achieved with strong and sustained commitment from the government and donors. A lack of both has doomed past efforts at capacity strengthening in Haiti's infrastructure sectors. Government support i s so challenging to maintain because much of the task at hand involves reducing the influence of politicians, over policies, payrolls, and revenues inkey ministries and state companies. However, it must be hoped that the current crisis in infrastructure provision (which has reduced the potential for rent-seeking and patronage), the urgent need to meet the public's demands for services, and the imperative of reviving economic growth will steel the resolve of political elites. On the donors' side the horizon for the provision of development assistance tends to be short-term. And past assistance inHaitihasbeenvolatile due to political issues. However, donors needto adopt a longer- term approach and accept that the country will need external financial and technical assistancefor many years to come. 99 230. Capacity-strengthening must therefore be part of a two-pronged approach: renewed efforts in this area must work beside more traditional infrastructure projects that can produce quick results interms o f service delivery. 231. To reflect the lessons learnt, the Interim Cooperation Framework seeks to follow this two-prong approach in the infrastructure sector: identifying and addressing immediate needs for services and jobs, while also seeking to strengthen institutions and government involvement in the longer term. Mechanisms for aid coordination and monitoring are also envisaged. In this context, the ICF identifies the following main objectives related to infrastructure: 1) improving and modernizing the management o f public enterprises in key sectors (electricity, communications, water and sanitation, ports and airports); 2) restoring the health of the electricity supply and expanding access; 3) rehabilitating transport infrastructure to improve access and strengthening transport ` institutions to make targeted interventions sustainable; 4) increasing water production capacity in the short term to supply water four hours a day, then improving service provided, especially to vulnerable groups, as well as strengthening sector capacity and studying how to improve service and coverage in the longer term; 5) ensuring collection of solid waste in urbanized zones and its elimination, responding to social and environmental standards; and 6) improving living conditions in slums, along with participatory preparation of a national program for the improvement of urban living conditions andinstitutional strengthening of relevant agencies. 232. The ICF was effective for identifying infrastructure needs in key sectors but it lacked an integrated, strategic approach to infrastructure. And implementation has been disappointing. While economic governance reforms are a central pillar of the ICF, there i s no coherent approach to improving institutions and governance in infrastructure, and the few measures in this area do not go far enough. In addition, implementation has fallen well short of goals: by September 2005, by the ICF's own evaluation (COCCI 2005), progress of only 21 percent had been made towards meeting targets for the axis of "Improving Access to Basic Services", excluding food security. In slum improvement, no targets had been met at all. However, there have been some successes in other areas, such as the establishment of the Road Maintenance Fund (see below), as well as initiatives not included in the ICF. 111. Six priorities for action 233. Drawing on the lessons of past interventions in Haitian infrastructure, and considering efforts already underway through the ICF, the remainder of this chapter sets out six priority areas for future action, identifying specific measures to address each one. The aim here i s not to revisit the ICF's work in identifying needs. Rather, the focus i s on how to address those needs more effectively-by strengthening the institutional arrangements for infrastructure, to improve operation and oversight, and the implementation and sustainability o f projects, as well as raising Haiti's capacity to absorb foreign aid. In the short term, this requires improving donor involvement, ensuring road maintenance and rehabilitation, and finding effective ways to provide services to the country's slum areas, both to improve conditions for the poor and address the socio- economic causes of violence. (The topic therefore has broader relevance for the themes 100 o f state buildingand violence reduction). In the medium to longer term, measures seek to enhance the state role in service provision, makingstrategy and policy, setting the ground rules for operators and coordinating interventions. Specifically, the six priority areas are: 1)improving donor involvement; 2) focusing on slums, to ,deliver services and prevent violence; 3) strengthening the functioning of the RoadMaintenance Fund; 4) makingthe MTPTC more effective; 5) overhauling state utilities; and 6) enhancing the legal and regulatory framework for public and private service provision. 1. Improving donor involvement 234. Because efforts to promote sustainability through institution building will take time to have real impact, and external funding will be needed for years to come, this section will consider how to make donor involvement more effective inthe short term. 235. The ICF systems have proved more effective for infrastructure needs assessment and mobilizing donor funding commitments than they have for implementation and coordination going forward. The ICF aimed to set up a coordination structure, including "tables" for infrastructure sectors or themes, led by a Joint Committee for the Implementation and Monitoring of the ICF, chaired b y the Prime Minister, with the participation of other representatives of the government, civil society and the international community. The ICF's sectoral tables have worked better for some infrastructure sectors than others, and, in any case, have served more as a forum for communication than for actual coordination in implementation. There are four tables for infrastructure (roads and transport, energy, water and sanitation, and urban services). The table for energy has been the most successful, largely due to the presence of a proactive coordinator, now paid for by CIDA. Donor cooperation has worked best overall for energy (mainly electricity). In October 2005, the Government, EDHand principal donors signed the Brussels Memorandum of Understanding ("Proposition de Protocole daccord pour une Sortie de Crise du Secteur de L'ElectricitC") committing themselves to a series of measures to strengthen the governance and increase the efficiency and transparency of investments in the electricity sector. The M o U also included the commitment of the donors to finance key investments and technical assistance to allow for an effective response to the difficult challenges inthe sector. 236. As discussed in Chapter three, a first step towards improving donor coordination would be to improve the current system of sectoral tables. This would involve funding and appointing active and effective coordinators to those that currently lack them and providing guidelines for the tables' responsibilities and reporting requirements. Also, participation of a wider range of stakeholders would enhance their functioning.66 237. Better donor coordination also requires more effective government involvement and the establishment of a thorough information system on infrastructure activities 66 For further ways to enhance the functioning of the sectoral tables, see COCCI, 2005. 101 containing detailed, comparable data on projects, planned and in implementation. The two issues are related: donors active in the infrastructure sectors currently deal mainly with the Ministries of Public Works and Economy and Finance, not the Ministry of Planning and External Cooperation, which houses a unit for coordinating and monitoring the ICF, despite being less central to interaction on operations. Meanwhile, "strategic coordination" has been undertakenby the Prime Minister and a unit within his office (the Cellule de Coordination Strategique) which serves as the secretariat for the ICF. The government therefore needs to clarify responsibilities for coordination-and provide the tools andchannels of communication required. 238. A more ambitious move to improve donor coordination would be to introduce some kind of pooling system for infrastructure, based on donor contributions into a trust fund managedby a lead donor. This would be narrower in scope than multi-donor trust funds set upinmany recentpost-conflict or disaster situations, buthave some of the same advantages. Donor contributions go into a common pool, which prevents earmarking and allows for coherent, larger-scale planning and execution of works. This should be simpler and cheaper for overall implementation than the current fragmented arrangements. Pooling of donor resources also facilitates strategic planning of investmentswithin and across sectors. And arrangements can be set up to cover recurrent costs (including salaries) over a longer period and in a more uniform way than individual projects or PIUs. 239. However, such a fund would only be advisable in Haiti if it were administered in a way that contributed to institution-building in the government, and if it had the backing andparticipation of most major donors. While a fund would need a certain critical mass to be effective, it should not try to incorporate existing projects if they are working well without it. It would also have to be set up so that the implementation body was within the existing government structures rather than a separate aid management agency, integrated enough with regular government functions to contribute to capacity building, but still able to operate efficiently. The right balance would also need to be struck, in ensuring sufficient government participation and ownership, while leaving donors with enough oversight. 240. For aid to be effective in infrastructure sectors a more stable source of funding also needs to be in place for at least the next decade. Aid flows for infrastructure, as for other sectors, have been extremely volatile in Haiti (figure 4.1) which has seriously undermined their overall effectiveness (see chapter one for a discussion of this issue). 102 Figure4.1: Total donor disbursementsfor infrastructure 100 , 90 80 70 A 60 e3 E tn 50 2. 40 30 20 10 0 I 95 96 97 98 99 00 01 02 03 04 05 Sources: Haiti Redbooks, 2001-2002 and 2002-2003 (2004 draft), prepared by the World Bank Note: 2005 figure is approximate, calculated from ICF aggregate data for July 2004 to December 2005 2. Focusingon slums, to provide services and prevent violence 241. Improving infrastructure services inurban informal settlements i s most urgent and important from a humanitarian and poverty reduction perspective, given dire existing conditions. But slumupgrading should also be seen as a way to prevent violence. 242. A significant reduction in slum-based urban violence and associated risk factors is essential for economic development, state-building and political stability in Haiti. Violence in Haitioriginates mostly from its major slums, particularly those of the capital: CitC Soleil, Bel Air, Saline, and Carrefour Feuille. Armed gangs control these areas and use them as a base for their activities across the city. The state barely has a presence in these areas - and certainly does not have a monopoly on the use of force there. Long- term political stability also requires neutralizingthe threat of social and political upheaval originating among the politicized gangs of the informal settlements of the capital and other major cities. While President PrCval apparently has a strong base of support in these areas, expectations are correspondingly high, and results will be needed if the current relative calm i s to endure. 243. Tackling violence in Haiti's slums requires measures to address the socio- economic roots of gang membership and violence, in addition to a direct focus on law and order. The causes and risk factors include: high unemployment especially among young men; weak community structures, organizations and links (social capital) due to the transience of these communities; and miserable, overcrowded living condition^.^^ 244. To really address these socio-economic causes and risks, more focused interventions are needed. There are numerous social assistance programs already 67See World Bank (2006) for analysis of the socio-economic, political and institutional causes of violence inHaiti. 103 underway in the major slums of Port-au-Prince, but these are usually small in scale and narrow infocus, with little coordination between them. InCitC Soleil, which has roughly 350,000 residents, at least 144 programs are underway or planned, involving UN agencies, U SAID, the Red Cross, and other international and local organizations. They provide food, water, health and social services, education, employment, infrastructure and sports and social facilities. The UNi s setting up a task force on CitC Soleil, with the participation of NGOs and other donors, which should improve coordination. 245. Programs that provide basic infrastructure services can help prevent violence and mitigate the associated risks both by addressing socio-economic causes and by providing a vehicle for specific anti-violence activities. Comprehensiveurbanupgrading programs, providing access to a range of infrastructure services and including a specific focus on crime and violence prevention, provide a promising model for Haiti. Inaddition to major works for water, sanitation, roads and so on, the infrastructure components of such programs can include crime prevention elements, typically based on Crime Prevention through Environmental Design (CPTED) principles (see Box 4.2). The Bank i s currently piloting projects of this nature in Brazil, Jamaica and Honduras. (See Box 4.3 for the Jamaicanexample). Box 4.2: CrimePreventionthroughEnvironmentalDesign(CPTED) A key cross-sectoral methodology for urban violence reduction is "Crime Prevention through Environmental Design" (CPTED). The fundamental concept i s that the "physical environment can be changed to impact on criminal behavior in a way that will reduce the incidence and fear of crime and improve the quality of life" (Cooke, 2003). CPTED techniques have been particularly popular since the 1970s in North American and European cities, where they have achieved some success. They have more recently been adapted to African and Latin American contexts, including South Africa and Brazil. Early CPTED projects focused on reducing the opportunities for crime through "situational prevention" actions (including provision of street lighting, access control, security, surveillance and landscaping). More recent interventions (second generation CPTED) also seek to address the underlying social causes of crime (Le. "social prevention") through more integrated community-building activities, including the construction of community centers and other safe urban meetingplaces, as well as involving local residents and groups inproject design and implementation. The World Bank has recently included second generation CPTED principles in urban operations in Brazil, Honduras, and Jamaica. Adapted from Fay (2005) and Cleveland and Saville (1997). 246. With the difficult security situation on the ground and the challenges involved in implementation, it would be essential to advance slowly in Haiti. Pilot projects could initially be attempted in one or two selected communities in CitC Soleil or perhapsjust Bel Air (which i s somewhat safer and more stable). If these are successful, a larger intervention could be tried, targeting more communities inthe capital and perhapsviolent secondary cities, such as Gondives. 104 247. Learning from international experiences and analysis, and considering Haiti's specific needs and situation, development interventions in slum areas should aim to include the following elements: A package of cross-sectoral interventions concentrated on a few "hot spot" communities, for maximum synergies and overall impact. This should include elements designed to produce quick results for the population as well as medium and longer term programs. High job creation impact: the use of labor-intensive construction and service provision. Crime prevention considerations mainstreamed into the choice and design of infrastructure works. While most infrastructure works should focus on basic infrastructure (particularly water, sanitation, and roads), "situational prevention" (e.g. street lighting) and "social prevention" (e.g. the provision of community centers) should also be included. CPTED tools can be useful here. A specific anti-violence component. This should include both short term measures to mitigate and resolve conflict, such as mediation services, andmedium-term social prevention and capacity enhancement interventions, such as skills development and family support programs. Implementation through NGOs and other groups already active in these areas. This i s essential to gain trust and acceptance. It also recognizes that no government agency currently has sufficient capacity or presence in the slums. A USAID project that provides street-sweeping and other services, and provides jobs in the process, i s apparently being successfully implemented by the Pan-American Development Foundation (PADF) inCitC Soleil. Working closely with communities. Community participation helps ensure that interventions reflect the priorities of local people, as well as building a sense o f ownership and commitment, which i s vital for project sustainability. The Bank's pilot Community-Driven Development (CDD) project has had some success in building social capital around effective service provision in rural areas. CDD elements should therefore be considered for any urban upgrading project. However, slum areas are generally more challenging for CDD than rural ones, in particular because social organizations are weaker. In addition, CDD interventions tend to be slower to implement and less easy to scale up. Strengthening capacity and organizations. Whether or not the CDD model i s used, any project should include training and a strong role in implementation for local individuals and organizations, with aim of developing useful skills and experience. Close coordination with government and other donor interventions to ensure complementarity of programs. The limited execution capability of the MTPTC, and the existing relations NGOs have built up in the slums arguably make these a better choice for implementation. However, there i s still a strong role for state companies to play - particularly the water utility (CAMEP), which already has a significant presence in slums through water kiosks. Close cooperation of MINUSTAH and police. Infrastructure projects are only one element of an anti-violence strategy. Appropriate accompanying actions are required from MINUSTAH (and the local police, if feasible). This might involve the 105 establishment of permanentMINUSTAHposts, which has apparently worked well in BelAir. This approachis to bepreferredover simply increasingthe sheer numbers of MINUSTAH tanks, armored cars and personnel in the slums, given that the stabilizationforces are apparentlymistrusted. Box 4.3: Urbanupgrading to prevent crime and violence: the Jamaica ICBSPproject The Jamaica Inner Cities Basic Services for the Poor project (ICBSP) is an urban infrastructure and public safety enhancement operation targeting inner-city and informal communities, prepared at the request of the Government of Jamaica (GoJ). It focuses on 12 inner-city communities, selected on the basis of high poverty indicators, `high priority' ratings for public safety by the Ministry of National Security, and the need for geographic and political diversity. Jamaica's gang-controlled inner-cities are at the center of the island's growing problem of violent crime. The project has three components: 1) Access to services, which will provide a demand-driven bundle of investments for construction and maintenance o f basic infrastructure, improved access to microfinance services in project areas and land tenure regularization; 2) Public safety enhancement and capacity building, which will finance integrated packages of consultant services, training and technical assistance focused on both short-term mitigation and conflict resolution and medium-term social prevention; and iii) Project management. Crime prevention influenced the design of the first component, as well as the second. Inaddition to water, sewerage and road works, infrastructure sub-projects include "situational prevention" elements, such as street lighting and zinc fence removal, and "social prevention" ones, including the construction of community centers (safe public meeting spaces that facilitate community organization). In addition, the expansion of microfinance i s to foster self-employment among at-risk youth and other groups. The second component, specifically for public safety enhancement and capacity building, includes: alternative livelihoods and skills development; mediation and conflict resolution; family support programs; youth education and recreation; and community capacity building and public awareness. Local residents have participated in the identification of sub-projects. The project has recruited community organizers to maintain close links with every community throughout implementation. The project has been designed and will be implemented in cooperation with the Jamaican police and authorities, as well as with other international development partners. It will be implemented by the Jamaica Social Investment Fund, an offshoot of the GoJ. The project has a total value of US$32.8 million, of which US$29.3 million is financed by the World Bank. It was approved by the Bank Board of Directors inMarch 2006. Implementation will start in mid-2006. Source: ICBSP Project Amraisal Document. March 2006. World Bank. Washington, DC. 106 3. Strengtheningthe functioningof the RoadMaintenanceFund 248. The Haitian Road Maintenance Fund (Fonds d'Entretien Routier or FER) was created by law in 2003, following the collapse of the previous Haitian road maintenance system-to the detriment of road quality-in the late 1980s. The fund i s an autonomous institution, financed through specific earmarked taxes, including a gasoline tax of one gourde per gallon. The World Bank estimates that annual revenues could reach US$5 million, but may be much less. FER will finance and oversee road maintenance contracts, procured by MTPTC, for priority road segments. A board including representatives of MTPTC, the Ministry of Economy and Finances (MEF), local governments, oil importers and roadusers supervises its operations. 249. The Government, with the help o f the donor community (particularly the Bank, through the Economic Governance Reform Operation), has taken measures since 2004 to strengthen FER'Sinstitutional capacity. These include the appointment o f its five Board members and the recruitment of staff, including three professionals: a Managing Director, an Administrative and Financing Officer and an Operations Officer. The first road maintenance contract was signed inJanuary 2006. 250. To really establish a credible road maintenance system in Haiti, further steps are required to strengthen the FER. First, coordination on the programming of road maintenance activities should be enhanced and formalized between key actors: FER, MTPTC, MEF and CNMP. Second, regular and adequatebudgetary transfers to the FER account of proceeds from earmarked taxes for road maintenance should be ensured. Third, the FER should develop expertise that will allow it to promote an efficient model for routine and emergency road maintenance, using either small firms or community- based micro-enterprises. 4. Making the Ministry of Public Works, Transport and Communications (MTPTC)moreeffective 251. A single ministry (the MTPTC) has responsibility for most infrastructure sectors, but lacks technical expertise, resources, and capacity for strategizing, planning and execution. Inkey sectors, there i s not even an attempt at strategizing and planning within the MTPTC, independent of resource availability. The staffing structure i s also apparently inefficient, with too many unqualified people in ineffective roles, and a shortage of qualified, technical staff, particularly in the regional departments. Many employees are poorly motivated, and low salaries require many to hold secondjobs. The existence of two project management units and other less-formal project implementation arrangements within the ministry complicates the structure. 252. A thorough institutional diagnosis should be undertaken for MTPTC, to examine its structure, the functions of every department and post (and whether incumbents are actually on the job) and provide concrete recommendations for enhancing capacity and improving efficiency. The study should establish a clear vision for the role that MTPTC 107 should play in the infrastructure sectors68. It should involve staff in a participatory way, to identify bottlenecks from their perspective. And then a firm plan should be developed for implementing the recommendations. The recommendations below should therefore be considered as preliminary, and subject to refinement and revision on the basis o f future analysis. 253. MTPTC should ideally have the capacity to make policy, strategize, plan and coordinate investments and play some role in managing and implementing projects. It should also have the capability to execute small works and perform some routine maintenance activities (MTPTCdoes still execute some works under force account using the Centre National de 1'Equipement or CNE). It i s not realistic or desirable to aim to return MTPTC to the more active role that it played in the 1980s, when it directly implemented most donor-financed projects. It would be more reasonable, but still difficult enough, to attempt to rebuild or replace some of the capacity - including sectoral divisions, information systems and planning units - that existed then, but was subsequently lost, amid increasingpoliticization of ministry staffing after 1991. 254. To improve coordination, oversight and policy-making, MTPTC's relationship with state-owned infrastructure companies should be redefined. Currently, the ministry does not have control of the port authority (APN) which comes, illogically, under the Ministry of Economy and Finance (MEF). And while most other main state enterprises do fall more directly under the supervision o f the MTPTC, their communication with and support from it vary greatly. It might be advantageous for each utility to have a high-level technical representative at MTPTC to enable strategic coordination and monitoring, and resolve key issues that require political attention. And while MTPTC should pressure utilities to develop the basis for monitoring (targets, information systems and so forth), utilities should also be protected from arbitrary interference b y the ministry (see next section for more on state utilities). 255. More effective policy-making and strategizing also requires greater sectoral specialization and de-politicization within the ministry, through a strengthening of the cadre technique (i.e. skilled technical staff). Currently decision-making i s concentrated inthe Minister and his cabinet, each member of which may have more than one sector to focus on (for example, a single director covers both telecommunications and energy). This stretches or exceeds the expertise of the individuals involved, and implies insufficient high-level time and focus for each sector. The current structure i s also excessively politicized, which affects decisions and creates instability as senior posts change hands frequently. A return to the previous system o f having secretaries of state or at least sector directors for each sector, each with their own department, might be advisable, especially if the new government would commit to merit-based non-political staffing of these. (Indeed, a Secretary o f State for Telecommunications, Energy and Miningwas recently appointed, but the post was eliminated in2005). Inthe longer term, greater technical specialization might favor the creation o f a separate ministry for energy, The study shouldbuildon the analysisof the 1998Rapport sur la Reforme Administrative en Haiti, undertakenby the Commission Nationale pour la Reforme Administrative, and financedby UNDP. This maderecommendations,apparentlynot implemented,for clarifyingand strengtheningthe functions of the MTPTC. 108 buildingon the existing Bureau des Mines et de 1'Energie (BME). This is attached to the MTPTC and has some capacity for strategizing butno supervisory authority. 256. Within MTPTC, the regional offices (Directions D&partamentaZes)are most in need of capacity-building, but this must be driven by a gradual decentralization of some roles and responsibilities from the center, with support for basic equipment and staffing. Regional MTPTC offices currently have little capacity to implement, coordinate or gather information. Most settlements outside the capital have no effective urban development frameworks or systems within which to coordinate individual projects. It i s especially difficult to recruit qualified staff to go to the provinces. Regional offices are even worse equipped than the center. Some may havejust one engineer for an entire region, as is the case in Grand'Anse. He/she may be expected to work without a vehicle, computers or much electricity. Reflecting both MTPTC's current centralization and staffing and resource constraints, MTPTC does almost nothing in some provinces. If it i s to establish even a minimal effective presence in these areas, capacity needs to be strengthened and certain responsibilities and functions decentralized. The regional offices need to gain a role inproject implementation, and over time, as capacity i s gradually built, procurement capabilities, which are currently concentrated at the national level. 5. Overhauling state utilities 257. The state i s the main provider of electricity, water and fixed telecommunications, and the country's dismal service indicators for all of these are a reflection of the poor performance o f the utilities involved6'. The principal utilities are: ElectricitC d'Haiti (EDH); Centrale Autonome MCtropolitaine d'Eau Potable (CAMEP), the capital area drinking water company; the national drinking water company, Service National d'Eau Potable (SNEP); and TELECO, for fixed-line telephone service. The state also controls ports and airports through the AutoritC Portuaire Nationale (APN) and the AutoritC ACroportuaire Nationale (AAN).7o 258. State enterprises suffer from many o f the staffing and management problems discussed in the context of MTPTC. Staffingi s politicized, and management has limited autonomy. There i s a shortage of skilled personnel, and uncompetitive salaries make recruitment difficult. Overstaffing i s extreme: with 1,675 regular staff and 724 temporary employees, EDHhas 71 employees per 1,000 connections (efficient Caribbean electric utilities have fewer then 10). TELECO has more than 3,500 employees, which yields a productivity ratio of about 32 lines per employee, a third of what it should be. CAMEP has about 750 employees, or 14 per 1,000 water connections, more than twice the efficient level (see Box 4.1 for information on APN). 259. The crisis i s especially acute in the electricity sector. EDH is technically bankrupt, as its liabilities are worth more than its assets. The company imposes a major fiscal burden: government support in 2005 was US$43 million, or 7 percent o f total 69 Water coverage indicators would be much worse without the activity of numerous NGOs and private local providers. ' OThis section focuses more on the main utilities as, while many of the messages apply, APNand AAN have different structures and functions. 109 expenditures. These subsidies are largely to cover two power purchasing agreements (PPAs) signedwith private independent power producers (IPPs) as well as the cost of the diesel for these-and EDHitself-to generate electricity with. Under both PPAs, electricity must be paid for even if it is not pr~duced,~'as it sometimes has not been, when the government has fallen behind in paying for fuel, or whether for technical reasons or other, the firms have not been able to produce. EDH's own generation capacity i s in a dire state: in April 2006 only 9MW from installed capacity of 179MW in Port-au-Prince were available, although part of the shortfall is due to the dry season, which reduces hydroelectric output. In March 2006, electricity was provided for four hours per day nationwide. And while 31 percent of households report electricity access, only about 10 percent have legal, active connections, according to EDH72.EDHbills less than half the electricity it produces. In 2002, 51 percent of electricity was lost in transmission and distribution, by far the highest level inthe 260. The case o f TELECO presents additional governance challenges. The company was used by previous administrations to provide jobs for supporters and hard currency, thanks to its profitable monopoly on international telecommunications traffic, with opaque financial management practices and lax corporate governance. However, in recent years, the company's profitability has suffered amid international deregulation of markets, which has slashed international tariffs, as well as intensifying competition from mobile operators and internet service providers. On rough preliminary estimates, TELECO's profits fell to US$4 million on revenues of US$20 million in 2005, from around US$15 million on US$68 million in2002. 261. There are many initiatives underway to reform state enterprises, but these have advanced slowly. The ICF designated as a priority, for improving economic governance, the "cleanup and modernization of the management of public enterprises in key sectors". TELECO i s undergoing a positive management transformation and a strategic repositioning, involving the appointment of a well respected professional manager, the replacement of most of the staff in the finance department, the installation of a modern financial management system, and accounting rehabilitation (also underway in EDH), conducted with the support of local management consulting and audit firms. International audit firms have been contracted to undertake audits and make recommendations for TELECO, EDHand CAMEP. 262. To increase the momentum and effectiveness of the reform process, and to really bringabout the thorough overhaulthese companies need, several elements are needed: 263. More government leadership. As discussed above, it is crucial to secure high- level government backing for the reform process, which must include a firm commitment to leaving state companies free from undue political interference. Restructuring the 71The PPA with Sogener, to serve the provinces, is entirely "take or pay". The other arrangement, with Alstom for Port-au-Prince, was modified in December to reduce the take or pay portionto 30 MW, leaving 20 MWon a standby basis. 72The remaining 21 percent includes includes illegal connections, inactive ones and those on which no consumption is registered, often due to broken meters. l3According to World Bank WDI data. 110 MTPTC to enhance technical, policymaking and strategizing capacity would helpbut firmbacking from the highest political level is also needed. The government established the Counsel of Modernization of Public Enterprises (CMEP) in 1996 to oversee the reform process, but the body has only recently been reactivated and has yet to play its full role. In particular, CMEP and other government bodies should, with company management, develop a plan for implementing the recommendations of the company audits underway. Political leadershipas well as donor coordination-has been more effective in the electricity sector, including in planning interventions in EDH. But there i s apparently little political focus on the need to improve the water companies, despite their inefficiency. 264. Emphasis on rehabilitation and maintenance. Most physical infrastructure is obsolete or seriously dilapidated, as there has been almost no maintenance of state-owned infrastructure assets in the last 20 years. Haiti i s one o f the few countries in the world without a high-voltage electric transmission grid, while low and medium voltage distribution networks are in critical condition. In general, core networks need to be strengthened before these can be expanded.74 Providing adequate service to those already connected i s also a precondition for pursuing greater cost recovery. There i s a willingness to pay for adequate service, as the existence of private providers demonstrates. But electricity and water utilities are in a vicious circle. Poor service quality leads to non-payment and low revenues for the company, which in turn compromises its ability to improve service. Many donors are concentrating on rehabilitation in projects and plans, notably in electricity, where ADF, CIDA, the European Union, the IDB, USAID and the World Bank all have ongoing or planned projects focusing on rehabilitation. But more must be done to put in place arrangements for regular maintenance, including funding and training of the technicians involved. There is a need to develop among utility management and politicians the conviction that maintenance must be a priority, as these still tend to favor new projects over rehabilitation and maintenance. 265. Greater transparency and accountability. Transparency requires that more information be gathered and made available. Modern, computerized systems must be put inplace to enable proper monitoring and accounting, as well as billing. Such information i s also essential for regulation, as discussed in the next section. Some efforts are underway, but at least for CAMEP, SNEP and EdH, these do not go nearly far enough due partly to insufficient funding. In CAMEP, for example, the billing system has recently been computerized. However, the system i s not functioning accurately, in part because of a lack of skilled personnel. A similar situation applies to TELECO, where the recent upgrade of the financial information system led to a deterioration in the financial information produceddue to a lack of skilled staff to runit. 266. Contracting by state utilities or ministries on their behalf should be made more open and public, and competitive where possible. In the electricity sector, the most urgent task is to increase the transparency and efficiency of the purchase of electricity ~ 74 However, some expansion of service in the most needy areas, particularly urban slums, should also be considered (see section above on slum upgrading). 111 (through PPAs with private investors) and of fuel (through competitive processes with the MEF,which handlesthese transactions). The Brussels MoUcommits the governmentto using competitive procurement for any future power purchasing agreements. However, the government has so far not opened any such process, despite the electricity shortage. Inaddition, pastcontracts shouldbemadepublic,includingtheexisting PPAs. 267. State utilities must be made central to efforts to fight corruption. Anti-corruption initiatives in these areas should have high political visibility. As indicated in Chapter three an Anti-Corruption Unit (ULCC) was established in 2004 as an autonomous entity under the Ministry of Economy and Finance, but this will require high-level support and resources to give it "teeth." 268. Utilities' boards of directors should be resurrected and empowered, to give them an active role incompany oversight and to promote accountability. The challenge will be to appoint competent individuals with sufficient autonomy and political neutrality. While suchboards formally exist, they do not usually meet. The board of EDHmet for the first time last year in response to pressure from donors. Neither SNEP nor CAMEP has a board that meets regularly. 269. Greater attempts should also be made to strengthen the pressurefor accountability from other channels: Parliament, civil society and the media. Requiring utilities to release more financial and operating information to these on a regular basis would be a first step here. Using the internet i s one form but has obvious limitations in Haiti and thus community radio and local media should also be considered. In addition, piloting mechanisms such as user scorecards, that provide a channel for public enterprises to improve service delivery by incorporating feedback from the user's of these services would serve a double purpose of not only improving the quality of services as well as the transparency with which they are delivered. Finally, the ComitC Consultatif linkedto the Ministry of Finance and set up to provide oversight over the economic governance reform program (see Box 3.6 in Chapter 3), could also play a role in creating demand for reform and improving accountability over service delivery in these sectors. A first step would be to increase the communication between the CC and the relevant agencies and initiate a discussion on the indicators presently being used to monitor reforms in these sectors. 270. Realistic strategies for involving the private sector. Given the size of the task ahead, and the limited capacity of Haitian utilities to transform themselves from the inside, private participation shouldbe sought where possible. However, in addition to the utilities' needs, such strategies need to take proper account of what the private sector would be willing to do, and what would be politically acceptable in Haiti. A failure to accommodatethese aspects underminedprevious privatization efforts.75 271. While nothing should be ruled out, conventional full privatization of the main utilities (through sale to strategic investors) i s likely to be difficult. Privatization i s a 75For example. the 2002 OED Evaluation states that, while the 1990 Restructuringand Development Project was revised in 1990 to support the privatization of state owned utilities, this failed, due to "lack of political will at the highest level and the weakness of the judicial framework." 112 politically charged issue in Haiti and opposition could be expected. In addition, the investment needs and organizational challenges of the utilities involved are vast. Meeting them would require considerable capital outlays and technical expertise, which probably rules out domestic investors. International appetite for infrastructure investments in low-income countries worldwide has shrunk in recent years. And Haiti's high political risk, weak governance, lack of local financing sources, and the limited resources o f both the government (for subsidies) and consumers (for tariffs) make its utilities particularly unpalatable. TELECO remains the most attractive company, with an as yet unexploited cellular license, and a still profitable monopoly on international communications. While the global environment may not be as conducive as five years ago for a traditional strategic investor privatization, with increasing competition from cellular and Internet services, TELECO could potentially be rendered attractive for a private investor. But it would first require a strategic repositioning, by focusing on building a national backbone and access to international cables (including a link to the Dominican Republic) to act as a wholesale carrier, while at the same time capitalizing on its cellular licenses through Rectel. 272. Structures for private involvement that require less capital commitment and protect external operators from some risks are more credible alternatives, as long as the investment capital required i s provided by donors or other sources. These structures include management contracts, which are being considered for both EDH and CAMEP76 (the ICF aimed to have prepared both o f these by March 2006). However, management contracts have had a disappointing track record for Caribbean utilities, particularly where political interference i s possible (see Jha 2005 for examples). One alternative i s the greater use of twinning or other arrangements for comprehensive technical assistance - including administrative, financial and technical aspects - with international utilities. The Jacmel Power System Rehabilitation Project, through which Canadian utility Hydro- QuCbec worked closely with local EDHoperations, provides a model that could be used inother regions or sectors (see Box 4.4). 273. Another option i s to decentralize administration of regional operations before pursuing private involvement, whether through technical assistance, management contracts, concessions or outright sales, especially for electricity distribution but also perhaps for water. A limited form of decentralization, covering operations, administration, billing and tariff collection, has been central to the success of the Jacmel case, but continuing central government control over tariffs (and delays in raising these) have created difficulties for it. Haiti i s probably too small a country to take this to the extreme of dividing utilities into separate companies for different geographic areas (i.e. horizontal unbundling). But this can have several advantages: it cuts out the organizational problems and huge debts of the center; it should increase the efficiency and responsiveness of administrative and management functions through decentralizing these; and it creates units of a more manageable size for integrated reform efforts to tackle. 76A PPIAFstudy is underway to assess the viability of a managementcontract for EDH. 113 274. Local companies should also be involved, if possible, in any attempt to bring the private sector into state utilities. Domestic investors have a better understanding o f local conditions and are less exposed to currency or political risk. Smaller regional contracts or units for electricity distribution or even water would be more within the reach of domestic companies and investors. The issue of allowing private participation in distribution will be considered inthe next section. I Box 4.4: Jacmel Power System Rehabilitation Project I Financed by the Canadian International Development Agency (CIDA) and the Government of Qutbec and with the close involvement of Hydro-QuCbec, this program provides a comprehensive package o f assistancethat has allowed the EDHoffice in Jacmel to produce and distribute electricity in the area. The program provided equipment (such as generators and lines) and Hydro-QuCbec employees worked to upgrade equipment and provide technical, administrative and financial assistance and training to EDH employees. EDHagreed to decentralize control over operations, administration, billing and tariff collection to its Jacmel office. Community meetings were organized from the start, to build awareness and support and attempt to counter the culture of non-payment. Consumers were not charged for electricity until full service was returned. When it was, higher tariffs were to be charged. Rehabilitation took priority over coverage expansion, but demand for connections increased steadily as service improved, prompting a significant expansion in access. As a result of the project, Jacmel became the only area in Haiti with 24-hour electricity. The payment rate rose to 92 percent of billing, while the non-technical lossesfell to 20 percent. The goal was for EDH Jacmel to charge and receive cost recovery tariffs for the constant provision of electricity. The project has not achieved cost recovery or self-sustainability as government resistance delayed tariff hikes, and necessitated further support from CIDA. But the Jacmel project has otherwise been largely successful. CIDA aims to establish a CAN$20 million fund with other donors (and full government support) to apply the model to selected secondary cities, including Gonayves. Inthe longer term, it estimates that the whole country apart from Port-au-Prince could be electrified in this way, at a cost of CAN$50-60 million. 6. Enhancing the legal and regulatory framework for public and private service provision 275. The state has an essential role to play in any country in setting and enforcing the rules of the game for public and private infrastructure providers. But the Haitian state does not fulfill these functions. The legal and regulatory framework i s outdated and needs to be updated. For example, despite the technological changes that have revolutionized the sector, the main legislation for telecommunications i s a 1977 decree- law that regulates the sector and in particular TELECO's monopoly. Updating the framework would also be necessary to implement reforms (as discussed above) to the MTPTC and state utilities. Enforcement can be strengthened by the establishment of appropriate structures. 276. However, the issue of political will is, once more, central. The policy-making capacity of the Haitian state i s currently weak, but this has not been the main constraint to 114 updating the legal framework in recent years, so much as the lack of follow up by the Executive in drafting new laws and implementing ones that have been passed. Where laws have been drafted with government involvement, these have not always reachedthe legislature (one example i s the 1998 draft water law and the major updating attempts made for telecoms in 1999 and 2000, with the support of a World Bank project). Even if laws have been enacted, they have frequently not been implemented, through the issue of the appropriate regulatory decrees, or have been ignored in practice. Attempting to update the main laws for infrastructure is only worth pursuing if the new government i s committed to bothenacting the laws andputting theminto practice. 277. With that caveat, the main objectives in a serious drive to update the legal and regulatory framework should be: 1) improving supervisory institutions; and ii) accommodatingcompetition andprivate participation. Improving supervisory institutions 278. There is a pressing need across the board for more effective and transparent regulation. Establishing proper information and financial systems within utilities, and promoting greater transparency, will be important elements towards achieving this, as supervision and monitoring are impossible without proper information. But changes are also requiredon the legal and institutional side. 279. Only in the telecommunications sector i s there a regulator, but this needs strengthening. The National Council for Telecommunications (Conatel) was established in 1969 by decree and redefined by a decree-law of 1987. While on paper an independent regulatory authority that largely mirrors the structure of the U.S. FCC, with a five-member Board of Directors, Conatel is, in practice, under the control of the MTPTC as its Board has never even been appointed. The institution as a whole lacks the staffing, technical and financial resources necessary to reassure investors in this increasingly private sector-oriented infrastructure sector. As part of an overall legal and regulatory reform of the sector, Conatel needs to be transformed into a truly independent regulatory authority, with capacity to enforce the current legislation, as well as to supervise the use of resources such as the radio spectrum. 280. In the water and sanitation sectors, the supervisory function also needs to be strengthened. However, trying to set up a regulator i s probably not the answer, given how difficult this has proven in other small countries with much greater institutional capacity. A stronger apex authority i s certainly needed - whether this i s a sector department within the MTPTC, as discussed above, or outside it - to lead the policy dialogue, oversee the functioning of utility boards, and manage the existing water information system, WASAMS, which currently has no institutional home. Institutional responsibility for sanitation also needs to be established, as there i s currently no entity in charge of this. Inthe longer term, a new sector framework law ( h i cadre) i s also needed. The 1998 draft law provides a usefulbasis, but would need some revision. This aimedto merge CAMEP, SNEP and the small rural water unit of the Ministry of Health, giving the new entity responsibility for both water and sanitation. It also provided for the gradual transfer of the responsibility for water and sanitation to municipalities (if these have the 115 capacity required), which in turn could delegate services to the private sector or water committees. 281. The electricity sector also needs some authority to regulate the sector. Such an authority would play an essential role in supervising the execution of PPA and fuel procurement contracts and in ensuring that future PPAs are awarded through transparent competitive processes. Accommodatingcompetitionandprivate involvement 282. Overall, the state needs to set a level playing field for private and public operators, encouraging competition and private sector involvement where possible. The Haitian private sector is dynamic in many respects, partly because it has learnt to cope in a risky environment and with a weak state. But the government still needs to lay down basic rules and standards. Specific legal changes will also be needed to allow private sector participation in state companies, as discussed above. 283. In telecoms, the priorities in this area should include the licensing of new operators and promotion of private investment in the provision of new services and introduction of new technologies, particularly new broadband wireless internet services, as well as the deployment of fiber optic networks and access to international submarine cables. Such access i s necessary to reduce the prices and improve the quality of international communications. The telecom sector already has the most competition and private participation at the national level, with two existing mobile phone companies and a third preparing to enter the market. Nevertheless, neither TELECO nor the private operators have any significant presence beyond the main urban centers, and are unlikely to do so in the medium term, given large pent-up demand in urban areas. It i s therefore important to consider a universal service funding mechanism to mobilize competitive private investment towards rural areas within the context of an output-based approach. 284. Electricity generation should become more competitive. There are already two private generating companies, as discussed above. But given the high prices paid, the lack o f transparency about the arrangements in place and Haiti's extreme shortage of electricity, there is a convincing case for further liberalization in this area. It i s not enough to commit to competitive procurement for PPAs, as the government has already done under the Brussels MoU,if no such competitive processes are opened. Inthe longer term, EDH's monopoly on distribution could be reconsidered, to allow for some private sector distribution. However, it i s not efficient to have more than one distribution network in any given area, and allowing private operators to skim off lucrative large consumers would further endanger EDH's survival. 285. Poor coordination and a lack of supervision and regulation among private water operators should be addressed via SNEP. While the water sector i s the least suitable for opening up to competition from private operators at the national level as it i s not efficient to construct parallel water systems, rural water supply in many areas i s largely in the hands of NGOs and water committees. Among these providers there is almost no coordination, regulation or supervision, and levels of price and quality vary. SNEP, with 116 World Bank support, i s developing a strategy for strengthening its rural unit to coordinate and supervise the work of the many NGOs and donors in rural water supply. This relationship aims to be a participative and complementary one, with SNEP seeking to provide organization and assistance and uphold some standards, rather than replacing the private sector. As such it could represent a model for other sectors, beyond infrastructure, when operational. 117 CHAPTER 5 INVESTING INCHILDREN Education is one of the main assets enabling Haitians to move out of poverty and is also critical in reducing the risk that Haiti will relapse into conflict. In addition, it provides an opportunity for the Haitian state to deliver core public services that matter to Haitian citizens, thereby re-establishing some legitimacy. Yet today, almost one-half of children 6-12 are not in school, and of those enrolled, only one-third reachfifth grade. Public education spending (1.7percent of GDP) and the capacity to deliver education services are low even by Sub-Saharan standards, such that about 90percent of primary education schools are non-public. Because non-public schools charge fees for enrollment, this constitutes a tremendous constraint on access - with average school fees representing over 20 percent of incomefor the poorest income quintile, families report that the main reason they do not send children to school is cost. Quality is also a problem - 70percent of schools.lack accreditation and 60percent of teachers are unqual@ed. Many children come to school hungry, a serious impediment to learning. In addition to providing a rationale for increasing expenditures in the education sector, Chapter 5 develops cost scenario and a discussion of alternative means for reducing the cost of schooling, for example by: (i) channeling resources through the National Partnership Office to non- public schools that agree to reduce or waive fees for poor students and adhere to minimum quality standards, or (ii) giving cash-transfers directly to poor households conditional on their children attending school. This will require a diferent, more effective and facilitating role for the state whereby it sets policy and regulatory guidelines, enforces quality control standards and ofsers incentives to other partners to offer quality education to all. I.Growth,Poverty,InequalityandEducation International evidence 286. International evidence shows that investing in children's education i s one of the best ways to address the goals of poverty reduction, economic growth and improved equality over the medium-term. Education promotes labor productivity and labor mobility, key ingredients for growth. Education is a key condition for achieving sustainedgrowth; no country has reachedsustainedeconomic growth (5 percent per year, net of population growth) without first having reached universal or close to universal primary education. Poverty reduction i s also stronger when growth i s labor-intensive, which implies the need for a labor force with a minimum base of human capital to respond to economic opportunities. But beyond traditional growth theory, improving equality of educational access i s an effective instrument for increasing equality more generally, which in turn raises the positive effect of growth on poverty. Primary education has many additional well-documented beneficial effects on nutrition, health, fertility levels andempowerment of girls. 118 Education, growth, poverty and inequality inHaiti 287. Wage and producer income analysis shows that education is the most important determinant of poverty in Haiti, controlling for other factors.77 The behavior of poor Haitian families confirms this; they make great sacrifices to keep their children in school because of the future pay-off to the family and associated higher incomes, migration and remittance^.^' Corroborating these findings, the World Bank Haiti Rural Growth Study (2005) highlights education as the most important factor in determining income of rural Haitians (the majority of the population), and recommends greater investments in education to stimulate economic growth in rural areas. Unfortunately, income poverty in rural areas i s compounded by virtually absent public social service delivery, in particular education provision. 288. The Haiti Social Protection Strategy Brief prepared by the Bank in July 2005 also highlights education as the single most important investment in preventing household risk, and in helping households manage those riskshhocks when they occur. Shocks to poor Haitianfamilies frequently result in the withdrawal of some or all o f their children from school, because of difficulties in paying school fees, uniforms, books and other expenses. This results in irreversible losses of human capital, raises household vulnerability to risks and detracts from broader future economic growth. 289. Households with less education earn less income, and education levels in Haiti are so low that the majority of the population lives in poverty. As seen in Figure 5.1 below, over half of breadwinners in the poorest decile (Dl) have no education at all, and even in the middle and upper-middle income deciles education endowments are low. The inter- generational transmissionof poverty and inequality through the educational system i s clear. Poverty headcounts are 90 percent for household heads with no education and 43 percent for those who finished primary education, as compared to 25 percent for household heads who finished secondary education. Malnutrition and poor health reinforce the constraints to acquiring education and engaging productively in the labor market - as severely malnourished children cannot learn and may suffer from impaired de~elopment.~~ 77 ''World Verner, 2005. Bank. 2004. HaitiRisk and Vulnerability Assessment. 79 Approximately40 percent of Haitianchildrensuffer from a caloricdeficit of an average of 460 kilo- calories per day. 119 Figure 5.1: Educational Attainment of Primary Breadwinner by Income Decile EducationalAttainmentof Primary Breadwinner by IncomeDecile 0Secondary Degree IPrimaryDegree No Schoolina D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 Income Decile Source: Staff Calculations, ECVH 2001. 290. Finally, education is a principal driver of the Haitian economy in its own right. The education sector employs 90,000 teachers (over 7,000 in the public sector, and over 82,000 in the non-public sector), making it the single largest employment sector after agriculture.80 Also, private household spending on education i s estimated at more than 4 percent of GDP. Inother words, investing in long-term human capital formation in Haiti i s also a short-term driver of employment and growth. 291. The many causal linksbetweenpoverty, equity, education and growth point to the need to invest in short- and long- term poverty reduction and humancapital formation. A strategy of investment in children, which reduces household costs for schooling and improves provision of basic education services, could simultaneously offer `quick wins' andtackle the deeper, long-term drivers of poverty. Quick wins also offer gains insocial stability; international experience in post-conflict countries shows that getting children into school i s a key factor in regaining normalcy. By contrast, conflict and instability are major constraints to the investment Haiti desperately needs to embark on a steady growth path. 292. Tackling constraints to acquiring human capital is fundamental to poverty reduction in Haiti. Efforts to expand the supply of and demand (capacity to pay) for services are critical. While the bulk of donor activity in education under the Interim Cooperation Framework has focused on supply side issues, efforts have fallen short as they are centered by and large on the public sector which services only 20 percent of enrolled primary students. Tackling demand-side constraints in Haiti has received less attention from the donor community, and is an area where the World Bank has a comparative advantage and extensive experience. MENJS, Direction de la Planification et de la CoopCration Externe, Recensement Scolaire 2003 120 11. The Provisionof EducationinHaiti: Current Constraints and Challenges Access to Education 293. Educational trends. With only 55 percent of children aged 6-12 enrolled in school, and only 32 percent of primary education students reaching fifth grade, Haiti i s far from the objectives of Education for All reaffirmed in Dakar in 2000. Haiti i s unlikely to reach those objectives without a major effort from government, households and the international community. But as will be seen further below, households are already doing all they can to invest in their children's education, meaning the State and international partners need to dramatically raise their support for education if Haitian households are to break free from the "education trap". 294. A comparison of household survey data from 1994 and 2000 permits the analysis of trends of key education indicators. The essential message from this analysis i s that educational access in Haiti declined over the period. Disparities between urban and ruralenrollment havebeenworsening, further fueling inequality.81 Table 5.1: TrendsinKeyEducationIndicators, 1994-2000 Source : ORC Macro. 2004. Profil de 1'Education de DHS Ed Data (Haiti DHS). 295. Equity inAccess. Gender differentials inaccess to schooling are apparent. Up to the age of 10, the percent of females attending school i s higher than it is for males, but the trend reverses between the ages of 14 and 24. Households are more likely to continue investing in the education of their male children than in their female children. But it i s interesting to note that the net enrollment rate at the secondary level i s higher for females than for male in urban areas, while it i s lower in rural areas. Geographically, access i s unequal across the nine regions of Haiti at primary and secondary levels, ranging from 41 to 72 percent at the primary level, and from 8 to 33 percent at the secondary level. The South-west, North-west and Center regions are 81ORC Macro. 2004. DHS EdData Education Profile: Haiti. 82In2000, 86percent of grade six students were over-age. Thisreflects sporadic attendance andrepeated grade repetition, linked to household inability to pay fees on a regular basis. 121 especially disadvantaged in this regard, suggesting that the geographic targeting of education benefits to these regions would be well-suited. 296. Public and Non-Public Education. Education inHaiti i s dominantly non public- sector provided. The latest education census (2002-03) showed that 92 percent of schools are privately operated and financed. In terms of student enrollment, the State provides for just 18 percent of primary school enrollment, while non-public education services make up 82 percent of primary enr~llment.~~ Non-public schools are a heterogeneous group, encompassing international-quality schools attended by the country's elite and extremely low quality schools attended by the country's poorest.84 The majority of non- public schooling i s affiliated with religious associations (Catholic, Protestant, Presbyterian and the like), as seen in Figure 5.2 below, followed by independent secular schools. Figure5.2: Non-PublicPrimaryEnrollmentby Type ofProvider Non-Public Primary Enrollment by Type of Provider IIndependent IReligious PDistrict Community ISource: MENJS, Direction de la Planification et de la Coopiration Externe, Recensement Scolaire, 2002-2003. 297. Schooling and Income. The 2001 Household Living Conditions Survey provides a wealth of information on schooling and income. Not surprising, schooling for primary and (especially) secondary levels varies significantly by household income quintile, as seen in Figure 5.3 below, because households must pay the costs of schooling for their children. Lower income households find it more difficult to finance their children's education, inrelative and absoluteterms. 83MENJS, Direction de la Planification et de la CoopirationExterne, Recensement Scolaire 2002-2003 84According to the 2003 School Census, 225 non-public schools were functioning under a tree and over 2,200 non-public schools were located inresidential houses. 122 Figure 5.3: Enrollment by Income Quintile Enrollmentby Income Quintile r-- IPrimary ISecondaw Q1 Q2 Q3 Q4 Q5 Income Quintile Source: ECVH2001, 298. The primary reason stated by households for not enrolling their children in school AND for their children leaving school after they are enrolled is the high cost of schooling (Table 5.2 and Figure 5.4). Consistent with this logic, a higher percentage of poorer households, compared to wealthier households, indicated that the high cost of schooling is the primary factor constraining their children's enrollment in school. Table 5.2: Top Five Reasons for Not Attending School, inOrder of Importance, by Quintile r Q1 Q2 448.2 3 4 4 Q5 Total Too costly 50.0 42.0 40.9 33.7 43.4 Lack of Incentives 6.5 7.2 8.2 8.6 7.1 7.6 Health reasons 3.5 3.4 3.8 6.0 4.1 4.1 Too young 2.8 3.1 2.9 2.8 2.0 2.8 Repeatedfailure 1.5 2.2 2.1 3.2 5.0 2.6 Other reasons 34.5 42.1 34.8 38.5 48.1 39.5 Total 100.00 100.00 100.00 100.00 100.00 100.00 Source: Staff calculations from the ECVHsurvey 2001. 123 Figure 5.4: Stated Reason for Leaving School among 6-24 year olds I Stated Reason for Leaving School among 6-24 year olds 45 40 35 1 :: 30 15 10 5 0 High Cost Lack of Motivalon Poor Health Repeated Failure Source: ECVH2001 EducationSpendinginHaiti 299. Public financing. The domestically financed allocation for the education sector in the FY2005/06 budget (Gdes. 4 billion, or US$89 M) is just 1.7 percent of GDP, the lowest in the L A C region (average 5 percent o f GDP) and low by Sub-Saharan standards, as well (average 3.9 per~ent)'~.If all external financing i s included in the calculation, public financing for education rises to 2.1 percent of GDP, still dismally low. When combined with estimated private (household) spending on education (Gdes. 7.4 billion, see below), total spending rises to 5.2 percent of GDP, a major increase but still too low to dramatically improve educational access and quality (and largely a reflection of Haiti's extremely low GDP). 300. Besides low levels relative to GDP, public recurrent education spending (primary and secondary, exclusive of external financing) in FY2005/06 i s less than 11percent of the total government budget, or about half of what i s considered as the standard qualification benchmark (20 percent) for the Education For All Fast Track Initiative, and well below the average of 25 percent for other low income countries. This is a slight decline from FY2003/04 and FY2004/05, when public education spending accounted for 12 and 18 percent, respectively, of total public spending. Although Haiti's prospects for growth and public revenue mobilization are expected to severely constrain public spending levels, it is important for the Haitian Government to do all it can to demonstratethepriority given to education by increasing its share in public spending. 301. The FY2005/06 education budget i s composed of recurrent (Gdes. 3.3 billion) and investment (Gdes. 0.7 billion) spending. Eighty-five percent of spending goes to pay teachers' salaries, leaving little to finance qualitative inputs such as textbooks, supervision, testing, and the like, not to mention expansion in capacity. In absolute terms, non-personnel recurrent expenditures were just US$ 6 million in FY2003/04 to serve 2.1 million students, or just $3 for each student. 85 EducationFor All inAfrica: Pavingthe Way for Action 2005, UNESCOBREDA. 124 302. The education budget of US$89 million for FY2005/06 serves about 2.1 million students enrolled in primary and secondary education, which translates into US$45 for each student each year, extremely low even by sub-Saharan African standards (estimated US$110). In terms of composition, just 38 percent of total public spending on education is allocated to primary education, an insufficient amount given Haiti's abysmal enrollment rates and in comparison to Sub-Saharan African francophone comparators (average 49 percent). 303. Unitcosts for public schools are at least two times those for non-public schools at the primary level. The main reason for this i s that average public school teacher salaries (estimated at US$2,765 in 2003,7 times GDPper capita) are more than four times higher than non-public school teacher salaries (estimated at US$589, or 1.6 times GDP per capita). 86 304. Household Spending On Education. In contrast to state spending, household spending on education i s extremely high relative to incomes. On average, households spend Gds. 3605 (roughly US$120) for each child per year on primary education (including fees, books, transportation, uniforms and food), with upper quintile households spending three times more than lower quintile households. This same pattern i s repeated at the secondary level, with the highest income quintile spending more than three times per child than the lowest income quintile. School fees are the primary cost, followed by educational materials and uniforms. Households in Port-au-Prince spend on average almost four times more than households in rural areas (Gds.8,618 compared to Gds.2,242). School fees for each child represent over 20 percent o f household income for the lowest income quintile; a poor household with three children thus spends more than 60 percent of householdincome on education, a truly remarkablesacrifice. Figure5.5: HouseholdSpendingonPrimaryEducationPer Childby IncomeQuintile Household Spending on PrimaryEducation Per Child by IncomeQuintile 7000 6000 86 c 4000 5 3000 2000 1000 0 Q1 Q2 Q3 Q4 Q5 Income Quintile Source: Haiti ECVH, 2001, 86 Enseignmentde BaseinHaiti, WorldBank,2003, p. 10. 125 Figure 5.6: Average Primary and Secondary Education Spending Among All Households, as a percent of average householdrevenue, by quintile I !I of told household Secondary education i expenditures8s a sha 01 Q.? a3 a4 0 5 Income QulnUle Source: Staff assessment from the ECVH2001. 305. Figure 5.6 compares average household expenditures across all income levels for both primary and secondary education to average household revenue by income quintile. While the richest quintile of the population spends only 3 percent of their average revenue on primary school costs, the lowest quintile spends 67 percent of its average revenue. The situation i s even worse for secondary education, in which case average household expenditures exceed by far the revenue of the poorest quintile of the population, which explains why virtually no extremely poor children of secondary school age are enrolled at that level (as pointed out inthe Chapter on Poverty andEquality). 306. In the context of Haiti's predominantly private sector market for education, it is fair to assume that higher spending per child reflect higher school fees and higher educational quality. Linking this section with the previous analysis of educational attainment, compared to their peers from upper income quintiles, Haitian children from lower income households get less and worse education. They are less likely to enroll in school, and more likely to drop out. They are less likely to gain degrees at each educational level, and at each educational level more likely to attend schools of lower quality (using spending per child as a proxy for educational quality). Thus, the primary constraint to more and better education for the large majority of Haitian households, but particularly those from lower income quintiles, i s their ability to pay for schooling. Impact on Educational Quality 307. Lack of resources and weak capacity in the Ministry of Education (MENJS) have resulted in distortions inthe education sector and hampered the MENJS from fulfilling its normative functions of regulation and quality control. As result, quality in education has sharply deteriorated: 126 (i) a high percentage of schools do not meet basic sanitary and safety standards (16 percent are houses, 33 percent are in churches, and 9 percent are open-air); (ii) percentofschoolsarenotaccreditedbytheMENJS(10,932 outofatotal 70 of 15,664 primaryschools); qualified ; (iii)60 wercent of teachers in the non-public sector are not appropriately (iv) only 360 school inspectors are responsible for accreditation, pedagogical supervision and administrative support, or an average o f one inspector for every 6,000 students. Also, inspectors are poorly equipped and trained; and (v) because of the increasing number of non-public schools and inadequate regulation, the education sector is increasingly segmented, with little standardization or curricula, pedagogy and/or assessment, making systemic qualitative improvements difficult to achieve, much less measure. Only a minority of non-public education providers belong to formal school associations (for example FONHEP and CONEP) andaim to provide the conditions needed for quality education. A formal link with the public sector is missing to harmonize quality standards and coordinate education policy and reforms in the education sector. Institutional Capacity of the Ministry of Education 308. An exhaustive analysis of the Ministry of Education's capacity was carried out in 1995 by the UNDP and UNESCO, in collaboration with the Ministry of Education itself. Its findings then are equally relevant today: MENJS senior decision-makers are absorbed in daily operational issues, which prevents them from assuming their strategic responsibilities; basic supervisory and regulatory capacity of the MENJS i s ineffective, procedures for dialogue and consultation with key non-public education stakeholders are unclear; there i s an absence o f coordination between pedagogical supervision and curriculum development, for public and non-public schools; MENJS rules, norms and administrative procedures are simply not applied in many cases, within the MENJS and externally with schools and parents; basic management information systems are weak; and there i s a lack of qualified personnel and minimumlogistical capacity.'* "RecensementScolaire,Ministirredel'Education,CIDA,2003. RenforcementInstitutionneldu Ministbrede 1'Educationet de sa Capacitt de Gestion, MENJSPNUDiUNESCO,DCcembre 1995. 127 309. The UNDPAJNESCO study made a series of recommendations for strengthening basic capacity for administration, strategic planning, decentralization and partnership. Some improvements were made, for example in carrying out a School Census and restructuring the human resources department of the Ministry. But many other initiatives were simply not carried out because of lack of resources, staff turnover and lack of qualified personnel to carry out strategic programs. Lessons of Experience from the Interim Cooperation Framework and other Post- Conflict Countries 310. The Interim Cooperation Framework (ICF) includes an ambitious combination of educational investments, focused on four priority objectives: 1. Improving governance and institutional capacity of the sector 0 develop instrumentsand capacity to regulate and control quality of public and non-public schools 0 develop mechanisms for accreditation, partnership and financing of non- public education 0 develop procedures and mechanisms for expanded parental and community participation in school decision-making and management 0 finalize the national strategy of EducationFor All 2. Improvingaccess and learning conditions 0 expand supply capacity inbothpublic and non-public schools 0 support educational demand from most vulnerable and disadvantaged groups 0 prepare implementation of third cycle of basic education 0 support higher education 3. Improving Educational Quality 0 improve andexpand teacher pre-service and in-service training 0 review and revise curricula 0 improve and expand pedagogical supervision 4. Improve access, relevance and quality of vocational education and training 311. In the ICF implementation, the bulk of investments has gone into improving educational quality and access in public primary schools, with the European Union and Inter-American Development Bank as the largest external financiers. Classroom rehabilitation, teacher training and supply of materials for public schools have been foremost. Demand-side interventions included financing of scholarships during the school year 2005/06 by USAID for 15,000 children negatively affected by political instability and environmental disasters, implemented through the National Scholarship Fund (Fonds National de Parrainage), an NGO. USAlD also began assistance to the MENJS" for accreditation of non-public schools. Meanwhile, CIDA financed a student *'MENJS,Cadre ~ de CoopBrationInternationale,Jan 2004-Decembre2005 128 loan program to help poor families pay for primary schooling of 30,000 children in 200906, implemented through a Canadian NGO. CIDA also assisted in institutional strengthening of the Departmentof Planning of MENJS. UNESCO and the World Bank focused on governance issues and collaborated with the MENJS and non-public education stakeholders in the feasibility study for the establishment of the National Partnership Office and in the finalization of the Education For All national strategy. Total disbursementsfor education during the ICF period as of December 31, 2005 were US$22 million, out of a total of US$86 million committed by external donors." 312. While still early, some conclusions and lessons can be drawn from this experience. The ICF attained most of its short-term goals of providingrelative stability in the education sector (notably schools opened on time and remained open, teachers were paid, educational materials were distributed to schools, and school feeding programs continued). However, the medium-term objectives of improved institutional capacity of the MENJS, expanded access, and improved quality remainto be achieved, despite some important progress made during the ICF period. On the positive side, ICF activities demonstrated the feasibility of demand-side financing to expand educational access and quality, and the process of monitoring the ICF greatly improved coordination and policy dialogue between donors andthe MENJS. 313. Lessons from Post Conflict Countries. Key findings from a study of education in 52 post-conflict countries recently published by the World Bank" highlight the important role of education as a critical element in the strategy to regain normalcy and reducethe risk of conflict (or relapseinto conflict). 314. More specifically, the study suggests four important starting points relevant for Haiti: 0 focus on the basics to get the system working, get the children and youth inschool as a "peace dividend"; 0 acknowledge the symbolism of a functional education system, and ensure some bold symbolic actions to signal the education systemi s moving forward; 0 build recognition that education reform must be led from within the country, as consensus develops on the wider development vision of that society; 0 focus from the beginning on building capacity for reform, which includes supporting participation of communities, local authorities and other stakeholders. 315. More operational lessonsfrom this study that apply to Haiti include: 0 building partnerships between the public and private sectors, and government (central, regional and local); 0 give priority to primary education, within a system-wide approach, as provisionof services at that level benefits the greatest number of households, particularly among the poor (which need to see a stake in stability); sa MENJS, Cadre de CoopCrationIntCrimaire, Bilan de la Mise en Oeuvre, Juin 2004- DCcembre 2005. Reshapingthe Future,Education and PostConflict Reconstruction,World Bank, 2005. 129 e build capacity of central authorities to ensure an enabling environment for decentralization. Given the predominance of the private sector in Haiti, particularly outside of Port-au-Prince, this means using incentives to improve enrollment and quality in non-public schools, while strengthening the State's normative and supervisory roles and capacities to enforce quality control; e international support for both recurrent and capital expenditures are needed, since Haiti lacks quick access to domestic revenue neededto keep the system running; e reduction of household costs of education i s an effective strategy to promote access and equity, but it needs careful implementation so as not to compromise quality. 316. The Bank study also specifically addresses the case o f education in Haiti, indicating that as the country struggles to recover, it faces two critical questions concerningprivate education. First,it needs to allocate public spending on education in a way that builds on the strengths and dynamism of the private system rather than simply expands public schooling. Second, it needs to address the regulation of the private education sector, without introducing bureaucratic controls that will further hinder the provision of quality education by public and non-public schools. A range of strategies that channel public resources to rivate education within a well-regulated framework are among the options facing Haiti.9 P 111. A Strategy for ImprovingEducation Service ProvisioninHaiti 317. Improving the education system and broadening access to education, especially that of the poor, requires a series of actions and reforms. The four most important elements of this strategy are: (i)increasepublicresourcesallocatedtoeducation; (ii) strengthen the capacity of the State to fulfill its policy, coordination and normative responsibilities; (iii)adopt pro-active reforms to help households finance schooling; and (iv)increase donor financing to support implementation of this strategy. The ultimate goal should be no less than universal primary educationby 2015, as promotedby the Education For All Fast Track Initiative (EFA-FTI) program, which i s supported by a wide range of national and international agencies. Increase PublicFunding for Education 318. The EFA-FTI has developed guidelines for assessing a country's eligibility for entrance into the program. Guidelines include a wide array of access, quality, institutional and spending indicators. Regardingthe latter, public spending for education should aimto meet the following targets: 92 Ibid., pp. 72-73 130 Table 5.3: Haiti's CurrentPublicSpending(2005/06) and EFA Medium-TermTargets for Education 319. Inother words, over the medium term public education spending inHaiti needsto roughly double relative to the FY 2005/06 allocation, as a percentage of GDP and as a percentage of government recurrent spending. In absolute terms, this would be equivalent to raising annual education spending by US$90 million, an ambitious but not impossible target. Most of this increase should go to the primary level, to raise its share of education spending from 35 to 50 percent, and to enable increased non-salary spending on educational quality. These resources should increasingly be targeted on the poorest households, least able to finance their children's educational costs. 320. A number of key policy decisions need to be made by Haitian leaders to achieve education for all Haitian children within realistic time-frames, financing scenarios, and institutional capacities. To inform these policy decisions, several partial scenarios have been simulated and their cost implications calculated. A description of these scenarios and their projected costs (Table 5.4) follows below. More fully developed cost simulations would need to be carried out to further develop implementation details of the proposed strategy. 321. Scenario 1- Public Financing; and Public Provision: All 500,000 currently unenrolled children and the additional children entering the age range o f 6-12 would be gradually enrolled inpublic schools, at public school unit costs (US$120), achieving EFA by 2011. N o child currently enrolled, besides those already attending public schools, would be supported; this would be the household's responsibility. This would require massive public school classroom construction, a very expensive and time-consuming undertaking. All primary school children would receive school feeding for 120 days of school year. 322. Scenario 2 - Public Financing and Non-Public Provision: All 500,000 currently unenrolled children, the additional children entering the age range 6-12, and the poorest 20 percent of currently enrolled children would be fully subsidized by the State so they could enroll in non-public schools, at the average non-public school unit cost (US$60). Expansion of enrollment would be rapidbut gradual, reaching EFA in 2011. Under this scenario, no public classroom construction would be needed, as it would be left to the private market to respond to this demand. Only accredited non-public schools would be 131 eligible to enroll subsidized students, which would require a major effort on both the public and the non-public sector. The poorest 20 percent of primary school children would receive school feeding for 120days of school year. 323. Scenario 3 -PublidHouseholdFinancing and Non-Public Provision: All 500,000 currently unenrolled children, the additional children entering the age range 6-12, and the poorest 20 percent of currently enrolled children would be subsidized by the State so they could enroll in non-public schools, at 50 percent of the average non-public school unit cost (US$60). Households would need to pay the other 50 percent of this cost. As in Scenario 2, no public classroom construction would be needed. The poorest 20 percent of primary school children would receive school feeding for 120days of school year. Table 5.4: ProjectedCosts of Achieving EFA by 2011(US$M) SCENARIO 1(US$M) - - - - 2007 2008 - 2009 2010 2011 - Total Enrolling all children6-12 9.8 16.3 19.9 23.6 14.2 83.8 PublicClassroomConstruction 30.0 49.0 58.6 68.2 40.3 246.0 Accreditationof Non-PublicSchools 0.0 0.0 0.0 0.0 0.0 0.0 QualitativeImprovements 5.0 15.0 15.0 10.0 10.0 55.0 School Feeding 65.8 71.1 77.4 84.7 89.0 387.9 TOTAL 110.6 151.4 170.8 186.5 153.5 772.8 SCENARIO 2 (US$ M) 2007 2008 2009 2010 2011 Total Enrolling all children6-12 22.2 28.9 39.0 52.8 53.9 196.9 Public ClassroomConstruction 0.0 0.0 0.0 0.0 0.0 0.0 Accreditationof Non-PublicSchools 0.1 0.3 0.3 0.2 0.2 1.1 QualitativeImprovements 5.0 15.0 15.0 10.0 10.0 55.0 SchoolFeeding 13.2 14.2 15.5 16.9 17.8 77.6 TOTAL 40.5 58.4 69.8 80.0 81.9 330.6 SCENARIO3 (US$M) 2007 2008 2009 2010 2011 Total Enrolling all children 6-12 11.3 14.8 20.2 27.7 28.5 102.4 Public ClassroomConstruction 0.0 0.0 0.0 0.0 0.0 0.0 Accreditationof Non-Public Schools 0.1 0.3 0.3 0.2 0.2 1.1 QualitativeImprovements 5.0 15.0 15.0 10.0 10.0 55.0 School Feeding 13.2 14.2 15.5 16.9 17.8 77.6 TOTAL 29.5 44.3 51.0 54.8 56.5 236.1 324. Note that all three scenarios include additional financing improvements in educational quality (textbooks, furniture, equipment, buildingrepair, latrines, etc.), using a very conservative cost of US$lO,OOO per school. 325. Clearly, Scenario 3 i s the most affordable for the State (and perhaps the most realistic in terms of likely financing from external and household sources). However, 132 Scenario 2 i s the most equitable and targeted on the poorest households (whether their children are currently enrolled or not) for maximum enrollment impact. Scenario 1 appears least attractive, as it would entail both higher unit costs and huge school construction costs, making this both prohibitively expensive and probably unrealistic in terms of implementation capacity. 326. To assess the financial realism of achieving EFA under one of these scenarios, Table 5.5 estimates total public primary education financing capacity, total required primary financing under Scenario 2 (described above), and the financing gap which would need to be closed by external donors or private households. The analysis suggests that if the government maximizes its efforts to invest in primary education, and adopts the most cost-effective approaches for expanding access, reducingdropout andimproving quality, the gap financing required may be within the realm of external donors' capabilities. Table5.5: EstimatedPublic FinancingCapacity andFinancingGapfor AchievingEFA UnderScenario2 Numberof Additional Children (US$M)\3 \l:projectsa10percentannualincrease \2: assumesprimary spending as a share of total educationspendingincreasesgradually to 50 percentby 2011 \3: FY2005/06primaryeducation financingplus incremental cost of Scenario 2 327. Finally, a more conservative and perhaps more realistic estimate o f the education system's capacity to expand and absorb all children aged 6-12, which retains the goal of achieving EFA by 2015, would lower these costs. Simply estimated, stretching out the program's objectives by five years to 2015 would mean roughly halving the annual costs of the three scenarios estimated above inTable 5.4 for the period 2007-2011. Inturn, this would half the financing gap estimated in Table 5.5, which would appear to be feasible given the interest of external cooperation agencies inthe sector. Selectively Strengthen the Capacity of the State inthe Education Sector 328. The predominance of the private sector in education, in financing as well as the provision of education services, i s primarily a reaction to the limited presence of the State. Moreover, weak State capacity to assert its most basic policy, norm-setting, supervisory and assessmentroles constitutes a major constraint to the education sector. 133 329. As seen in Figure 5.7 below, the vast majority of non-public schools (almost 11,000) are not accredited nor licensed by the Ministry of Education. Technically, these schools operate illegally, cannot benefit from MOE supervision or subsidies, and cannot offer the national exam (to get around this unlicensed schools send their students to licensed schools to take the exam). This means there i s almost no quality control of the non-public education offered in Haiti, and the State i s unable to fulfill its most basic educational responsibilities in relation to its citizens. Accreditation is the major area where the Ministry of Education's legal, normative, and supervisory roleneedsto be reinforced. Figure5.7: MENJS LicensingbyType ofNon-PublicSchool MENJS Licensingby Type of Non-PublicSchool 8000 7000 - 6000 u) 2 5000 4000 n b 5 3000 z 2000 1000 0 I Independent Religious Communate ccmrunity I Source: MENJS, Direction de la Planification et de la Cooptration Externe, Recensement Scolaire, 2003 330. One of the mandates of the Department for Private Education and Partnership (DAEPP)inthe Ministry of Educationi s to accredit non-public schools. The DAEPPhas lacked the resources, clear policy guidance from senior decision-makers, and transparent guidelines for licensing non-public schools. USAID i s supporting the DAEPP, but much more needs to be done. (Estimated costs for this are in Table 5.4) And, if public financing for non-public schools was conditional on accreditation, or at least on an agreed program to upgrade the school to meet certification requirements, this could be a powerful mechanismfor improving non-public school quality. 331. State-building in the sector i s needed to address a host of other supply-side constraints, in both public and non-public schools, low quality being the foremost. Low quality is most closely linked to poor teacher qualifications and practices. To address this, the MENJS has begun setting up in-service teacher professional development centers (EFACAPs) around the country to service both public and non-public teachers, with the support of international partners (IDB,EU, USAID, CIDA). It is important that this work continue. Limited capacity for pre-service teacher trainingkertification i s a major 134 supply-side constraint to expanding both the public and non-public sectors, which deserves both increased State and donor support. Strengthening the State's ability to ensure the supply of educational materials is another priority; this could be largely addressed by raising public funding for non-salary related spending, as mentioned above, andusing the National PartnershipFunddiscussedbelow. 332. In addition, to improve quality the MENJS needs to strengthen its pedagogical supervision capacities, for both public and non-public schools, with more staff, skills upgrading and logistical support. Finally, the MENJS needs to strengthen its capacity to assess student (and school) performance by improving its administration and use of national examinations (including public disclosure of results at the school level). The dominance of non-public service providers means parents need information regarding national exam results by school, so they can decide where to enroll their child and so that schools (particularly those which might receive public subsidies) are accountable to both parents andthe MENJS. ExpandEducationalAccess by Addressingthe MainConstraint: HouseholdAbility to Paythe Costsof Schooling 333. As indicated in Section 2, the primary constraint to improved educational access i s household ability to pay. There are at least two ways this constraint can be addressed: lowering the costs of schooling for households or raising household income to meet these costs. Both options are discussedbelow inthe context of institutional mechanisms which could be established. The MENJS and its partners need to consider the pros and cons of both mechanisms, choose one and then invest inmaking it fully functional. 334. National Partnership OfficelNational Partnership Fund. In accordance with the recommendations of the Basic Education National Program (PNEF), Haiti should promote quantitative and qualitative improvements in the education sector by improving its mechanisms for partnership with the non-public sector. To this end, the Ministry of Education formally proposedin early 2006 the establishment of the National Partnership Office (NPO) inthe Ministry of Education, along with its companion entity, the National PartnershipFund(NPF). 335. The NPO i s intended to be an institutional space for policy dialogue, strategic planning, research and operational collaboration between public and non-public stakeholdersinthe education sector. It is a critical mechanismfor establishing the State's legitimacy vis-&vis the non-public sector. The NPO i s also designed to promote the Education For All (EFA) agenda and strategic plan, and to generate consensus and adherenceto minimumeducational quality standardsinpublic and non-public schools. 336. The NPO and NPF would be governed by a 7-member Board of Directors, with representationfrom the MENJS, non-public education providers, and the main teachers' union. The Board would be the decision-making entity of the NPONPF, establishing strategic goals, basic objectives, operational guidelines and procedures, and other activities for promoting partnershipbetweenthe public and non-public sectors. Its multi- stakeholder composition would promote transparency and equity, further enhancedby the 135 NPO's mandatory Annual Partnership Conference, to which donors, a broad range of non-public education service providers, parent associations, student associations, teacher unions, etc, would be invited. The NPO would submit to conference attendees for review its annual report, proposed future policy directions, and objectives, and financial information. Over time, it i s expected that the NPO would set up regional offices as part of the decentralization agenda, although this would not be part of its first phase of development.93 337. The recommendation to create the NPO was first made in the late 1990s and a multi-stakeholder commission to assess its feasibility was established at that time. Political and institutional problems arose in 2000, and the concept was shelved. In 2003 the World Bank's "Enseignment de Base" education sector study re-proposed this idea, which the Transition Government and donors embraced in the Interim Cooperation Framework (ICF) strategy in July 2004. Under the ICF, an in-depth feasibility study for the establishment of the NPO and NPF was financed by the World Bank, with results shared with key education stakeholders. Legislation to create the NPO and NPF was prepared by the Ministry of Education, in collaboration with key non-public education actors (for example FONHEP), which was submitted to the Council for Ministers for approval in January 2006. Should it be approved, US$200,000 in World Bank LICUS grant funding i s available immediately to support its immediate start-up activities. In addition, the 2005/06 MENJS budget includes approximately US$lOO,OOO in financing for the NPO and NPF. 338. The National Partnership Fund i s designed to be a financial mechanism for supporting improvements in educational access and quality innon-public schools (and, in certain cases, inpublic schools). For example, the NPFcould channel public resourcesto non-public schools which agree to cut school fees charged to poor families and adhere to minimum quality standards (or commit to attain those standards, for example, by improving teachers' qualifications). The Board of the NPF would review proposals submitted by non-public schools, based on social fund-type standardized proposal templates and evaluation criteria. Alternatively, under procedures which would need to be worked out, the NPF could issue vouchers to poor Haitian parents which they could use to choose which accredited school they wished their children to attend. Vouchers could be for the full average unit cost of non-public schooling, or a percentage of that cost, which schools could thenredeem for payment fromthe NPF. 339. Scenarios 2 and 3 presentedinTable 5.4 above in fact assume the existence of the NPO/NPF, or some similar institution. Basic calculations under scenario 3 suggest that over a five-year period (2006-2011) the NPO and NPF could progressively expand coverage and subsidies to non-public schools (equivalent to US$30/student) which agree to waive school fees charged to the poorest 20 percent of children, to reach up 1,000,000 students (both currently unenrolled and enrolled), by 2011. This is well within the 93The full 120-pagefeasibility study for the establishment of the NationalPartnership Office was carried out by internationaland nationalconsultants in2005, under the supervision of UNESCO, with LICUS funding fromthe World Bank. Ref: << Etude Relative6 laMise en Place du 1'Office Nationaldu Partenariat et du FondsNationaldu Partenariat D, MENJS/UNESCO/BanqueMondiale ;See also, Manuelde ProcCduresAdministrativeset Financieres<(, MENJS/UNESCO/BanqueMondiale 136 financing parameters indicated in the section above which calls for increased public funding for education. 340. The NPO and NPF offer several advantages. First, they aim to stimulate educational access and quality by building on the strengths of the non-public sector (lower costs, ability to mobilize household contributions, greater supply response and flexibility). Second, the NPO and NPF would retain the State's normative, regulatory and supervisory role, and raise the State's capacity to fulfill those roles and improve educational quality, notably by stipulating that only accredited schools could get subsidies. Third, they would aim to complement, not substitute entirely, household education spending such that the subsidy would enable the poorest households to keep their children in school or sendtheir children to school for the first time. This would be a much more cost-effective way of expanding access than simply financing access inpublic schools. Fourth, for the first time, they would increase synergy and collaboration between the public and non-public sectors, improving information-sharing, strategic planning and resource allocation in the sector. Indeed, experience in other countries indicates the viability and cost-effectiveness of public funding for non-public education service provision, as describedin Box 5.1 below. Box 5.1: Demand-side financing can be done via a number of different interventions: 1. block-grantsto schools - which inturnagree to lower school fees; 2. targeted vouchers - cash payments given directly to families or schools; 3. stipends - via a conditional cash transfer program or scholarship; 4. targeted bursaries - cashpayments earmarked for specific purposes, such as book purchases; 5. student loans - to help defray costs; and 6. community grants - funds given to a community and linked to enrollment in community schools. Relevant examples of such demand-side financing analogous to the National Partnership Office include: In Haiti, the National Sponsorship Fund (Fonds National de Parrainage) finances student scholarships, with funding from the private sector and USAID. In the Dominican Republic, the Ministry of Education contracts with private secondary schools on a per student basis to provide education services, through its Department of Private Education. In Jamaica, the Ministry of Education finances the base salary of teachers working in accredited schools owned and operated by communities and/or NGOs, which then charge fees to top-up salaries and finance school maintenance and qualitative inputs. In Colombia, the Ministry of Education offers vouchers on a lottery basis to poor students to attend non-public secondary schools in areas where the public system is operating at full capacity. InBangladesh, the Ministry of Education subsidizes 80 percent of base salaries of secondary level teachers working in community-managed not-for-profit schools. Such schools represent 90 percent of all secondary schools, similar to the situation in Haiti. Schools follow state criteria for recruiting students, and state pays for 80 percent o f the costs of hiring a new teacher for every 35 students enrolled. 341. A Complementary Way to Expand EducationAccess: the Applicability of ConditionalCash Transfer Programsto the HaitianCase. As seen inBox 5.1 above, 137 there i s a vast experience in developing interventions that tackle the demand constraints faced by households in accessing education (that is, their ability to pay and the opportunity cost of children attending school). These interventions are designed to strengthen the social safety-net and counter negative coping mechanismsthat households undertake as survival strategies during tough times - such as withdrawing kids from school or neglecting to access health care services. These include, among others, targeted fee-waivers to allow poor families to access basic social services, and providing cash transfers to families conditional on keeping their children in school (following the Oportunidudes model developed in Mexico which has been widely reproduced in Latin America and elsewhere around the world). Conditional Cash Transfer (CCT) programs have been particularly ,successful in expanding access to education as well as reducing poverty (via the cash transfer that i s given to households) and empowering women, who are often the recipients of the transfer, but have by and large only been implemented in middle-income countries. Preliminary results for Cambodia suggest that it may be possible to implement an adapted version of such a program in a low-income, post- conflict setting such as Haiti, although a feasibility assessment would be needed to confirmthis (see Box 5.2). Box 5.2: Cambodia's Cash Transfer Program Cambodia i s recovering from years of internal and external strife, with severe repercussions for all aspects of the economy and society, including education. As a result, while most Cambodian children attend some schooling, a large share only completes a few grades and there is a particular problem that too few girls progress to secondary school. Cambodia's Japan Fundfor Poverty Reduction tackles this problem by providing a cash transfer of US$45 to poor families whose daughters enter the first year of secondary school. The value of the scholarship is large-in 2002, mean per capita GDP inCambodia was approximately US$300. While the benefit i s called a "scholarship", it is given directly to parents at a public school ceremony, which also honors the girls who have been selected according to transparent eligibility criteria. Despite governance challenges, the program has been successfully run by the Ministry o f Education. While beneficiary selection originally took place at the schools, this function has been centralized at the ministry to reduce discretion in applying selection criteria. A recent impact evaluation o f the program shows that it has a large positive effect on the school enrollment and attendance of girls of 30-43 percentage points, and the impact appears to have been largest among girls with the lowest socioeconomic status. Adapted from Filmer and Schady, 2006. GettingGirls Into School: Evidencefrom a ScholarshipProgram in Cambodia. World Bank. 342. The cost of such an initiative in Haiti would dependprimarily on coverage goals (how many children would benefit from the program). As an example, assuming a US$10 per month per child income transfer, US$ 11.5 million per year (including administrative costs of 15 percent) would cover 100,000 children. This i s equivalent to roughly 20 percent of primary school-aged children who currently do not attend school. For an introductory consideration of implementation issues that would need to be consideredfor designing such a program, see Annex 1. 343. Finally, efforts to increase public and external donor spending for education, strengthen the capacity of the MENJS, and develop demand-side financing mechanisms 138 for expanded access need to be complemented by continued development of local school management committees and school-parent associations. Several donors have exciting initiatives in this area (EU, CIDA) in Haiti which need to be expanded. International experience (e.g. EDUCO in El Salvador) has repeatedly shown that increased community and parental involvement in school management leads to improved educational outcomes for children of those schools. The commitment of Haitian households to the education of their children suggests this is a very promising area for productive investment by external donors, the MENJS and the communities themselves. 139 Annex 1: Cash Transfer ProgramImplementation Considerations 344. Drawing on international experience, a successful CCT program would call for the following basic elements: simple and objective beneficiary targeting and selection methodology and mechanism; reliable supply of the service which constitutes the conditionality (usually education or health services); capacity to verify fulfillment of conditionality (for example school attendance or receipt of vaccines); payment system; and active beneficiary database for entry/exit of beneficiaries, and for program monitoring and evaluation purposes. 345. Usually CCT programs are centralized, technocratic structures with skilled staff, strong information and communication technology systems, links to the national budget office (for payments) and the sectoral ministry responsible for providing or overseeing provision of services supported by the cash transfer (in this case, education). Such a structure may be difficult to set up in Haiti, and for such a program to work, alternative mechanisms for ensuring the above-mentioned parts of a CCT program need to be found. 346. Targeting: could be based on the recently-released Haiti poverty map and zones where poverty is highest, for families with children between the ages of 6-12. Working with community-based associations and local government authorities, contracted non- governmental organizations would identify qualifyingfamilies and inform them about the program, its benefits and their responsibilities. 347. Service Delivery: could be through public and non-public education service providers. 348. Verification: could be through certification by the education service providers, submitted on a quarterly basis to the program manager, with funds to make transfers included as part of the project disbursements for the provision of education services. Program supervision would include random spot checks of providers to ensure correct application of program regulations. 349. Pavment System: managed by contracted firms (e.g. remittance agencies) or NGOs, to which beneficiary would take a "voucher" for payment issued by the contracted agency. 350. Beneficiary Database: contracted service providers would be responsible for maintaining the database of their clients, and for communicating this information to central program database management. Supervision visits would be conducted by government or non-governmental organizations to verify beneficiary participation and to instruct service providers of the need to remove families from the program (for example, 140 when children graduate from primary school, or fail to comply with their co- responsibilities). 141 REFERENCES Banquede la Republique d' Haiti. 2004. Rapport Annuel. Barriers. Research Paper 2003/32. Research Paper Series. Globalization, Productivity andTechnology. Bravo-Ortega, C., and Lederman, D. 2004. Agriculture and NationalWelfare aroundthe World: Causality and International Heterogeneity since 1960. Bricefio-Garmendia, C., A. Estache, and N. Shafik. 2004. "Infrastructure Services in Developing Countries: Access, Quality, Costs and Policy Reform." Policy Research Working Paper 3468. World Bank,Washington, D.C. Calderbn, C., and L.Serve'n. 2004 "The Effects of Infrastructure Development on Growth and Income Distribution." Policy Research Working Paper 3400. World Bank, Washington, D.C. COCCI. 2005. "Rapport d'Evaluation de la Mise en Oeuvre du Cadre de Cooperation Interimaire." COCCI. Cooke, G. R. 2003. "CPTED makes a comeback." www.vcnet.com/exDert/library/cDted gc.html. Collier, P and A. Hoeffler. 2002. "Aid, Policy and Growth in Post-Conflict Societies." Policy ResearchWorking Paper 2902. World Bank, Washington, D.C. D e Ferranti, D., Perry, G., Foster, W., Lederman, D. and Valdes, A. 2005. Beyond the City. The Rural Contribution to Development. World Bank Latin America and Caribbean Studies. Advance Edition. World Bank. Docquier, F. and Marfouk, A. 2006. International Migration B y Education Attainment, 1990-2000. In: International Migration, Remittances & the BrainDrain. Economist Intelligence Unit 2006, Haiti, Country Report, February 2006, The Economist. Fisher, C. et al. 2006. Haiti: Rapport de la Mission Multisectorielle, Fonds Monetaire International, 14fevrier 2006. Fay, M.(Ed.). 2005. "The UrbanPoor inLatin America." World Bank, Washington, D.C. Fay, M. and M. Morrison. 2005. "Infrastructure in Latin America and the Caribbean: Recent Developments and Key Challenges" World Bank, Washington, D.C. Gannon, C., and Z. Liu. 1997. "Poverty and Transport." TWU (Transportation ,Water and UrbanDevelopment) DiscussionPaper 30. World Bank, Washington, D.C. 142 Gajdeczka, P., Jaramillo, L., Everaert, G.,Sancak, C., Dalsgaard, T., and Maticen, J. 2005, Haiti, Selected Issues, International Monetary Fund. Girma, S., Henry, M.,Kneller, R. andMilner C. Threshold andInteraction Effects inthe Openness-Productivity Growth Relationship: The Role of Institutions and Natural Barriers. Giustinaina. A. 1999. Stability and Structural Reforms. DominicanRepublic: Selected Issues. IMF staff Country Report No 99/177. October 1999. Hausmann, R. Rodrik, D., and Velasco, A. 2004.Growth Diagnostics .Working paper, Harvard University, October 2004. ICEA and EFC (consultants for World Bank) 2005 "Current situation and analysis of strategic options for telecommunications sector policy reform". ICFRCsumCedu rapport du groupe thkmatiqe ENERGIE(ElectricitC) 27 May 2004 IMF. 2005. Staff Report for the 2005 Article IV Consultation and Review of the Program Supportedby Emergency Post-Conflict Assistance. April 29,2005. IMF. 2001. The Dominican Republic: Stabilization, Reform, andGrowth. Washington D.C. Institut Haitien de Statistique et d'Informatique, Enquete sur les Conditions de Vie en Haiti (ECVH-2001). Instituto de Pesquisa EconBmica Aplicada, (PEA) Texto para discussiio No 1081, SAUDE E SANEAMENTO NO BRASIL Mkio Jorge Cardoso de MendonGa Ronald0 Seroa da Motta, Rio de Janeiro, abril de 2005. Instituto Libertad y Democracia. 2001. The Capitalization of the Poor in Haiti, Urban Real Estate andBusness Formalization", Volume I, report. Final InterimCooperationFramework (ICF). Summary Report. July 2004 Jha, A. K. (Ed.). 2005. "Institutions, Performance and the Financing of Infrastructure Services inthe Caribbean." World Bank- Working PaperNo. 58, Washington D.C. Justesen, M., Verner, D. 2005. Haitian Youth: Overview and Analysis of Risk and Protective Factorsfor Youth inHaiti. DRAFT,World Bank. KPMG. 2004. Corporate Tax Rates Survey. January 2004. Luke, J., Gaviria, A., Lora, E. 2003. I s Geography Destiny? Lessons from Latin America. Inter-AmericanDevelopment Bank. 143 Lundahl., M. 2004. Sources of Growth in the Haitian Economy. Inter- American Development Bank. Lustin, D. 2005. La Micro-Finance et son Role Potentiel dans 1' Allegement de la PauvretC et le Developpement en Haiti. CEPAL. Ministry of Agriculture andWorldbank. 2005. Diagnostic and Proposalsfor Agriculture and RuralDevelopment Policies and Strategies, October 5, 2005. Nexant andEconergy (for USAID).Haiti-Energy Sector Assessment, March 2005. Orozco, M. 2006. Understanding the remittance economy in Haiti, Inter- American Dialogue, World Bank Ratha, D. 2005. "Workers' Remittances: An Important and Stable Source of External Development Finance". In: Remittances. Development Impact and Future Prospects. Editedby Samuel Munzele Maibo andDilipRatha. Saville, G and G. Cleveland (1997) cr2ndGeneration CPTED: An antidote to the Social Y2K Virus of UrbanDesign." Schiavo-Campo, S. 2003. Financing andAid ManagementArrangements inPost-Conflict Situations. World Bank - CPR Paper No. 6, Washington D.C. Schwartz, J., S. Hahn, I.Bannon. 2004. "The Private Sector's Role in the Provision of Infrastructure in Post-Conflict Countries: Patterns and Policy Options." Social Development Papers: Conflict Prevention and Reconstruction Paper No. 16. World Bank,Washington, D.C. Smith, D.,A. Gordon, K. Meadows, and K. Zwick. 2001. "Livelihood Diversificationin Uganda : Patterns and Determinants of Change across Two RuralDistricts." FoodPolicy 26: 421-35. Verner, D. 2005. MakingPoor Haitians Count, Labor Markets and Poverty inRuraland UrbanHaiti, Basedon the FirstHousehold Survey for Haiti, October 25, World Bank. World Bank. 1998. Haiti: The Challenge of Poverty Reduction, Volume I.PREM. World Bank. 2000. Measuring growth in total factor productivity. PREM notes No 42. September 2000. World Bank. 2001. Dominican Republic- Informesobre la Pobreza: La Pobrezaen una Econom'a de Alto Crecimiento (1986-2000). Direcci6n Regional de Reducci6n de la Pobrezay Gesti6n Econ6mica. 144 World Bank. 2005. "Diagnostic and Proposals for Agriculture and Rural Development Policiesand Strategies." WorldBank, Washington, D.C. World Bank. 2005. DominicanRepublic-Review of Trade and Labor Competitiveness. Direcci6nRegionalde Reducci6nde la Pobreza y Gesti6nEcon6mica. World Bank. 2005. World DevelopmentReport. 2005. A better InvestmentClimatefor Everyone,World Bank and OxfordUniversity Press ,Washington, D.C. World Bank. 2006. Dominican Republic: Country Economic Memorandum, The Foundationsof Competitiveness, PREM. World Bank. 2006a. "Urban Violence in Haiti: A rapid assessment of socio-economic, political and institutional causes of urban violence in Haiti." Draft. World Bank, Washington, D.C. World Bank 2006b. "State Failure and Social Resilience in Haiti: A Country Social Analysis." ReportNo. P095864-HT. World Bank. 2006c. Jamaica: Inner City Basic Services for the Poor, Project Appraisal Document.World Bank,Washington, D.C. World Bank 2006d Haiti: Electricity Loss Reduction Project Appraisal Document. (DecisionMeetingDraft),World Bank, Washington, D.C. World Bank 2006e Haiti: Transport and Territorial Development Project Appraisal Document.World Bank, Washington, D.C. World Economic Forum. 2004. The Global Competitiveness Report. 2004-2005. PalgraveMacmillan. New York. Young, P., Dunn, D.,Giustiniani, A,, Nadal-De Simone, F., Tanner, E., and McHugh, J. 1999. DominicanRepublic: SelectedIssues, International Monetary Fund, Washington, D.C. 145