HAITIAN CITIES: ACTIONS FOR TODAY WITH AN EYE ON TOMORROW HAITIAN CITIES: ACTIONS FOR TODAY WITH AN EYE ON TOMORROW Nancy Lozano-Gracia & Marisa Garcia Lozano Editors © 2017 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The bound- aries, colors, denominations, and other information shown on aTny map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissem- ination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522- 2625; e-mail: pubrights@worldbank.org. Cover photo: © Ingrid Nelson. Further permission required for reuse. Cover design: © Ingrid Nelson. Further permission required for reuse. TABLE OF CONTENTS ACKNOWLEDGEMENTS i ABBREVIATIONS iii OVERVIEW 1 “We are like a reed; we bend but we don’t break” 2 Urbanization in Haiti: Unserviced cities growing in a fragile and risky 3 environment Planning: A shift toward resilient urban planning is needed to address current 5 infrastructure deficits and prepare for future urban growth Connecting: Better within-city connectivity and accessibility are achievable through 8 improved motorized transport and enhanced coordination between land use and trans- port investments Financing: Strengthening municipal finances is essential to close the urban infra- 12 structure and services gap, and to accommodate the growing urban population “Live today, but think about tomorrow - Wè jodi a, men sonje demen” 15 References 19 CHAPTER 1. LAYING DOWN THE FACTS ON HAITIAN URBANIZATION 23 Urbanizing while grappling with widespread fragility and disaster risk 25 Urbanizing under an unfavorable labor market and high levels of poverty 34 Urbanizing under crowding conditions and with no supporting infrastructure 40 Reviewing the challenges ahead 44 References 45 SPOTLIGHT 1: WHAT IS URBAN AND WHAT IS AT RISK 49 References 58 CHAPTER 2 – MOVING FROM RECONSTRUCTION TO RESILIENT URBAN PLANNING 63 FOR A BRIGHT FUTURE Why resilient urban planning? 63 Cities in Haiti are marked by basic service deficits and high exposure to natural 64 disaster risks Weak land administration, inappropriate regulation and information gaps hamper 72 effective decision making and exacerbate planning challenges Despite recent efforts, governance challenges remain a hurdle to long-term resilient 78 urban growth Instruments that can help initiate change today, with an eye on building stepping 82 stones for tomorrow References 94 SPOTLIGHT 2: PATTERNS OF LAND USE WITHIN CITIES 99 References 107 CHAPTER 3: SHAPING LABOR MARKETS: CONNECTIVITY, JOBS AND RISKS 111 Connectivity in Haitian urban areas 111 Urban transport in Haiti is slow and unaffordable to many 113 Structure of the urban areas of Port-au-Prince and Cap-Haïtien: Where are the people? 120 Where are the opportunities? Fragmented labor markets - Few commuters not travelling far 129 Transport interventions that focus on speed and resilience can reduce spatial 136 mismatch Matching people and jobs – Overcoming the challenges 142 References 149 SPOTLIGHT 3. INTER-CITY CONNECTIVITY: WHY IT ALSO MATTERS FOR HAITI’S 151 PRODUCTIVITY 158 References 163 CHAPTER 4: FINANCING HAITIAN CITIES 165 The process and progress of decentralization in Haiti 174 Incomplete decentralization and a weak legal framework for municipal finance con- found responsibilities Limited sources of municipal revenue hamper the capacity to provide services 176 Lack of transparency and limited reliability of the transfer systems exacerbate 190 financial constraints A path to strengthening municipal finances 194 References 199 ANNEXES 201 ACKNOWLEDGEMENTS The Haiti Urbanization Review was prepared by a team led by Nancy Lozano-Gracia (Senior Economist GSU10–Task Team Leader), and consisting of Sarah E. Antos (IT Officer, Data and Information Management, ITSOP), Paolo Avner (Urban Economist, GSUGL), Andrea Colombo (Consultant, GSU10), Chandan Deuskar (Consultant), Marisa Garcia Lozano (Consultant, GSU10), Alexandra Panman (Consultant), Jonas Ingemann Parby (Senior Urban Specialist, GSU10), Joseph Denis (Consultant, GSU10), Claudia Soto (Disaster Risk Management Specialist, GSU10), and Benjamin P. Stewart (Geographer, GGSCE). Claudia P. Pacheco Florez (Program Assistant, GSU10) provided overall administrative assistance. The work in Chapter 3 constitutes a collaboration between the World Bank, Flowminder Foundation, the WorldPop Project, and Digicel Haiti. The Flowminder team included Guilherme Augusto Zagatti, Miguel Gonzalez Canudas, Chris Brooks, Maximilian Albert, Elisabeth zu Erbach-Schoenberg, Alessandro Sorichetta, Simon Dutka, Priya Burci, Andrew Tatem, Erik Wetter, and Linus Bengtsson. The team is grateful for peer review input from Roger Gorham (Transport Economist, GTI04), Augustin Maria (Senior Urban Development Specialist, GSU11), and Michel Matera (Senior Urban Specialist, GSU13). Additional contributions were provided by Roland A, Bradshaw (Senior Disaster Risk Manage- ment Specialist, GSU10), Lauren Nicole Dauphin (Consultant, GGSCE), Katie L. McWilliams (IT Officer, Data and Information Management, ITSOP), Emilie Perge (Economist, GPV04), and Franck Taillandier (Senior Urban Transport Specialist, GTI08). The team benefited from the technical guidance from Catalina Marulanda (Practice Manager GSU12) and Pierre Xavier Bonneau (Program Lead LCC8C). Valuable comments and guidance were provided by Judy Baker (Lead Urban Economist, GSU10), Trevor Monroe (Senior Operations Officer, GTKM1), and Raju Singh (Program Lead LCC8C. The report was prepared under the overall guidance of Ming Zhang (Practice Manager, GSU10). The team also recognizes the support of the Country Director for Haiti, Mary A. Barton-Dock. The overview was edited by Communications Development Incorporated, and editorial support for the entire report was provided by Jean-Dany Joachim. The book’s design is credited to Ingrid Nelson. The policy discussion in this report has benefited from discussions with high-level government officials and development partners including representatives from: Comite Interministeriel d’Ame- negament du Territoire (CIAT), Centre National de l’Information Géo-Spatiale (CNIGS), Institut Haïtien de Statistique et Informatique (IHSI), Ministère de l’Intérieur et des Collectivités Territori- ales (MICT), Ministère de la Planification et de la Cooperation Externe (MPCE), and Ministère des i Travaux Publics, Transports et Communications (MTPTC). Special thanks go to Michèle Oriol and Rose-May Guignard from CIAT for their overall guidance and contribution to this work. The work in this report has been made possible thanks to the financial contribution from three grants: support from the Global Facility for Disaster Reduction and Recovery (GFDRR) TF0A2693; a grant from the World Bank’s Jobs Umbrella Trust Fund, which is supported by the Department for International Development/UK AID, and the governments of Norway, Germany, and Austria, the Austrian Development Agency, and Swedish Development Agency SIDA, TF0A2893; and a grant from the Innovations in Big Data Analytics program, under the Global Data and Text Analytics Operations unit in the Global Themes Vice Presidency of the World Bank. ii ABBREVIATIONS ASEC Assemblée de la Section Communale (Communal Section Assembly) CAP Cap-Haïtien CASEC Conseil d’Administration de la Section Communale (Communal Council) CDD Community-Driven Development CFPB Contribution Foncière des Propriétés Bâties (Tax on built properties) CIAT Comité Interministeriel d’Aménagement du Territoire (Inter-Ministerial Committee for Territorial Development) CNIGS Centre National de l’Information Géo-Spatiale (National Center for Geospatial Information) DGI Direction Générale des Impôts (General Tax Office) DHS Demographic Health Surveys DINEPA Direction Nationale de l’Eau Potable et de l’Assainissement (National Drinking Water and Sanitation Directorate) DRM Disaster Risk Management EC European Commission ECLAC Economic Commission for Latin America and the Caribbean ECVH Enquête sur les conditions de vie en Haïti (Survey of living conditions in Haiti) ECVMAS Enquête sur les Conditions de Vie des Ménages Après Seisme (Survey of households’ living conditions in Haiti) EDH Electricité d’Haiti (Electricity of Haiti) FGDCT Le Fonds de gestion et de développement des collectivités territoriales (Local Government Development Fund) FY Fiscal Year GDP Gross Domestic Product GHSL Global Human Settlements Layer GUF Global Urban Footprints GVA Gross Value Added HTG Haitian Gourde (currency) IADB Inter-American Development Bank iii IHSI Institut Haïtien de Statistique et Informatique (Haitian Institute of Statistics and Information) IMF International Monetary Fund ILO International Labor Organization JRC Joint Research Centre LAC Latin America and the Caribbean MARNDR Ministère de l’Agriculture, des Ressources Naturelles et du Dévelopment Rurale (Ministry of Agriculture, of Natural Resources, and Rural Development) MDE Ministère de l’Environnement (Ministry of Environment) MEF Ministère de l’Economie et des Finances (Ministry of Economy and Finance) MSPP Ministère de la Santé Publique et de la Population (Ministry of Health) MICT Ministère de l’Intérieur et des Collectivités Territoriales (Ministry of Interior and Local Authorities) MPCE Ministère de la Planification et de la Cooperation Externe (Ministry of Planning and External Cooperation) MTPTC Ministère des Travaux Publics, Transports et Communications (Ministry of Public Works, Transport and Communications) ONACA Office National du Cadastre (National Agency for Cadastre and Land Registration) PaP Port-au-Prince PDNA Post-Disaster Needs Assessment PFM Public Financial Management PRAFIPUM Programme d’Amélioration des Finances Publiques Municipales (Program for the Improvement of Municipal Public Finance) UN WUP United Nations World Urbanization Prospects USAID United States Agency for International Development USD United States Dollars WDI World Development Indicators WHO World Health Organization iv OVERVIEW NORTHERN PART OF CAP-HAÏTIEN, NORD. PHOTOGRAPHED BY REMI KAUP, 2006 SOURCE: WIKIMEDIA, CREATIVE COMMONS LICENSE OVERVIEW Today, more than half of Haiti’s population calls cities and towns their home, in a major shift from the 1950s when around 90 percent of Haitians lived in the countryside. Urbanization is usually paired with economic growth, increased productivity, and higher living standards, but in Haiti it has taken a different course. Potential benefits have been overshadowed by immense challenges, all of which require immediate action. To better understand the factors that constrain the sustainable and inclusive development of Haitian cities, this Urbanization Review organizes the challenges along three dimensions of urban development: planning, connecting, and financing. Planning reviews the challenges in supporting resilient growth to create economically vibrant, environmentally sustainable, and livable cities. Connecting focuses on the obstacles of physically linking people to jobs and businesses to markets, while financing focuses on identifying the key capital, governance, and institutional constraints that are hurdles to successful planning and connecting. Along these lines, the analysis suggests three main challenges for Haitian cities: PLANNING: Resilient urban growth is hindered by wide gaps in basic services, increasing exposure to natural disasters, and ineffective land use planning. CONNECTING: Poor connectivity within Haitian cities hampers integrated labor markets and access to economic opportunities. FINANCING: The ability of local governments to plan, service, and connect cities and towns is heavily constrained by limited resources at the municipal level. To respond to these challenges, the Urbanization Review proposes three broad strategies (encap- sulated in the Summary Matrix, Table O.1): PLANNING: A shift toward resilient urban planning is needed to address current infrastructure deficits and prepare for future urban growth. This includes investing in basic service deficits, lever- aging information for decision making, and strengthening property rights. 1 CONNECTING: Better within-city connectivity and accessibility are achievable through improved motorized transport and enhanced coordination between land use and transport investments. This entails investing in and improving efficiency, increasing affordability, and strengthening coordination of land use and transport investments. FINANCING: Strengthening municipal finances is essential to close the urban infrastructure and services gap, and to accommodate the growing urban population. This requires consolidating, harmonizing, and enforcing existing frameworks; building capacity and expanding financial opportunities; and expanding and leveraging the local revenue base. “WE ARE LIKE A REED; WE BEND, BUT WE DON’T BREAK” Proverbs are integral components of Haitian culture and speech. Along with metaphors, imagery, and storytelling, they are a traditional form of communication used to pass down knowledge and wisdom from one generation to another. Though the findings and messages of the Urbanization Review are technical, they resonate with a well-known Haitian proverb that reflects Haitians’ day-to-day struggles and hopes for a brighter future. Nou se wozo; nou pliye nou pa kase—Haitian Creole for “We are like a reed; we bend, but we don’t break.” This proverb captures Haiti’s long history of resilience in the face of slavery, colonialism, political oppression, widespread destruc- tion from natural hazards, social exclusion, inequality, and poverty—all which have shaped the country’s urbanization, a process that determines its current challenges to development, but most importantly, the opportunities that lie ahead. Historically, fragility1 in Haiti has been driven by political violence and instability (World Bank 2015b).2 Such instability has weakened state institutions, the rule of law, and the investment climate, leading to violence and distrust in public authorities (Singh and Barton-Dock 2015). Despite some improvements in governance indicators, the country ranks the lowest in Latin America and the Caribbean (LAC) in control of corruption. While Haiti’s fragility can be traced to the Duvalier regime (1957-1986), subsequent political and institutional instability have exacerbated volatility in the country’s recent history. Between 1986 and 2014, Haiti witnessed eighteen changes of government and over twenty major cabinet changes. In the second half of the 20th century, factors such as faulty agricultural policies and overexploitation of land deteriorated the rural economy and fueled a massive migration into urban areas of peasants seeking security, opportunities, and access to services, particularly in Port-au- Prince, the capital. But the provision of basic services did not expand to meet the new population pressures and cities were unable to meet the demands of the incoming population. 1 The World Bank Group classifies Haiti as a “fragile” state due to its low Country Policy and Institutional Assessment ratings for economic, social, and public sector polities and institutions. Broadly, fragility is defined as the weakness of institutions and vulnerability to instability, conflict, and violence. 2 In less than thirty years (1986-2014) the country was led by a succession of eighteen short-lived governments and suffered repeated delays in elections, which led to Parliament’s dissolution in 2015. 2 The destructive impacts of the earthquake in 2010, coming on the heels of a devastating hurricane season in 2008, and followed by the even more devastating Hurricane Matthew in 2016, further depleted the resources to generate greater prosperity throughout the country, with most efforts geared toward recovery and reconstruction. Despite these obstacles, as the proverb says, Haitians “bend but don’t break.” While facing political instability and immense losses from disasters, the country has also taken important steps toward development. Key milestones on the social and economic fronts have been achieved, including a reduction in extreme poverty and expansion of education and health services. Today, 90 percent of children attend school, and infant mortality decreased by 9 percent between 2005 and 2012. The country’s major cities are now all connected to the main road network, tourism has increased, and access to finance—in particular micro credits—has expanded (World Bank 2015b). There is a renewed sense of optimism for the future. Haiti is now at a decisive point in its history, as the country shifts from a focus on reconstruction to long-term development and forward-thinking planning. As another proverb states, Wè jodi a, men sonje demen, or “Live today, but think about tomorrow.” This proverb is a manifestation of Haiti’s culture, one that stands strong and determined to act today, with tomorrow in mind. The Haiti Urban- ization Review aims to contribute solutions for living through the problems faced by cities today and provide recommendations to help build the “thinking” about solutions for a better tomorrow. URBANIZATION IN HAITI: UNSERVICED CITIES GROWING IN A FRAGILE AND RISKY ENVIRONMENT Around the world, urbanization has often had a positive effect on economic growth. The strong link between urban levels and income has been well-documented. Historical data between 1996 and 2015 for over 180 countries show incomes rising as the share of population living in urban areas increases. Densities found in cities promote productivity and offer opportunities to improve people’s livelihoods and quality of life, eventually helping lift many out of poverty. For firms and workers in cities, proximity makes skills matching and job searching more efficient. For govern- ments, basic public services and infrastructure can be provided at lower cost due to economies of scale. The relationship between urbanization and economic growth, however, is not always linear. High densities of people alone are not enough to create the agglomeration economies often attributed to cities. When densities are poorly managed, externalities such as congestion, pollution, and high crime rates can overshadow the benefits of urbanization. In Haiti, urbanization has not gone hand in hand with economic growth. Gross Domestic Product (GDP) per capita remained stagnant and in fact dropped from USD 757 in 19963 to USD 727 in 2013, even when urban rates increased from 33 to 58 percent. Unlike similar countries in LAC, Haiti has not benefited from urbanization. Widespread fragility and costly natural disasters may have under- mined the benefits of the urbanization process. A closer look at the country’s system of cities, the deficits in urban infrastructure and services, and the limitations in governance and urban financing at the local government level helps explain why Haiti has urbanized without economic growth. 3 At constant 2010 prices. 3 Since 2000, Haiti has urbanized rapidly, albeit later than many of its LAC peers. In 1950, roughly 10 percent of the country’s population lived in urban areas, but from this decade until the 1980s, the number of urban dwellers increased at four times the rate of the rural population. In the 1980s, the urban population increased at a faster rate than the total population, reaching 30 percent by the mid-1990s. According to national official statistics, 52 percent of the popula- tion in 2015 resided in urban areas. The United Nations (UN) World Urbanization Prospects put this figure at 57 percent for the same year, with an average annual urbanization rate of 5 percent between 2000 and 2015. In 15 years, Haiti’s urban population grew 3.6 percentage points faster than the average Caribbean country and doubled in size from just over 3 million people to nearly Figure O.1. PROPORTION OF POPULATION IN HIGH-DENSITY AND URBAN CLUSTERS (WORLDPOP VS. UN WUP URBANIZATION RATE) Source: Deuskar, Stewart, and Lozano-Gracia (2016) based on WorldPop (2015), European Commission thresholds for urban areas, and UN World Urbanization Prospects (2015). 4 6 million. Each year, as many as 133,000 Haitians are becoming city dwellers (World Bank 2015b). Ten years from now, the number of urban dwellers is expected to increase by almost another 2 million people, and could surge to around 11 million by 2050, for a 76 percent urban rate. Satellite imagery to update what is urban in Haiti suggests that the urban population may be higher than suggested by official statistics. Definitions of urban areas in Haiti are outdated and have used unclear criteria. To address this limitation, this report uses a different measure of “what is urban” based on gridded population estimates and population density thresholds. By producing an urban vs. non-urban classification at high resolution (100m x 100m cells) and identifying urban built-up areas, the analysis suggests that Haiti has an urban population of over 6 million people, or about 64 percent of the total population—a marked difference from the 52 percent reported by official figures based on projections from the latest census. This makes Haiti the fourth most urbanized country in LAC, just after Puerto Rico, Trinidad and Tobago, and Mexico, instead of ranking 20th based on the United Nations World Urbanization Prospects (UN WUP) data (Figure O.1). PLANNING: A SHIFT TOWARD RESILIENT URBAN PLANNING IS NEEDED TO ADDRESS CURRENT INFRASTRUCTURE DEFICITS AND PREPARE FOR FUTURE URBAN GROWTH Haitian cities are not supported by adequate urban infrastructure and basic services, under- mining productivity and livability. Rather than benefiting from high densities, cities in Haiti today are overcrowded places with wide gaps in infrastructure and services. Resilient urban growth is hindered by these gaps and by increased exposure to natural disasters, and by ineffec- tive land use planning. As many as 35 percent of urban residents do not have access to improved water, and the share of families with water connections inside their dwelling or with access to a public tap fell sharply between 2000 and 2012 (respectively, from 24 to 3 percent, and from 65 to 21 percent). Two-thirds of urban residents lack improved sanitation, and an estimated 8 percent of urban residents practice open defecation. Haiti has the lowest collection rate of solid waste services in LAC (12.4 percent), ranking far behind the next lowest country in the region, Paraguay (57 percent), and behind low-income African countries such as Senegal, Benin, Mali, and Ghana, with collection rates of 21, 23, 40, and 85 percent, respectively (Hoornweg and Bhada-Tata 2012). Cities are also growing in an uncoordinated and unregulated manner, heightening their exposure to natural disaster risks. Haiti is one of the world’s most exposed countries to multiple natural hazards. Over 93 percent of its surface and more than 96 percent of its population are at risk of two or more hazards (World Bank and ONPES 2014). Between 1976 and 2012, hydro-me- teorological events cost the economy nearly 2 percent of GDP a year (World Bank 2016). The 7.3 Richter-magnitude earthquake in 2010 inflicted massive economic losses, representing 120 percent of GDP (World Bank 2015a). Built-up areas are particularly vulnerable, as they are disproportionately concentrated in high seismic hazard zones (60 percent), and around half are considered at risk for flooding. How urban areas expand in the country, and how buildings and infrastructure are built, are vital when vast amounts of land are exposed to different types of natural hazards. 5 Further, weak land administration, information gaps, and inappropriate regulation hamper effective decision making and exacerbate planning challenges. Information on land is limited and out of date, affecting the overall quality of land administration in Haiti, which scores the lowest in LAC according to the 2017 Doing Business index. Out of 190 countries, Haiti ranks 180th on ease of registering property and 166th on obtaining a construction permit.4 Construc- tion permit fees represent 15 percent of the total cost of construction, far higher than the average 2.5 percent in LAC and 7.6 percent in Sub-Saharan Africa. 5 Costs and time constraints are associated with informal and unplanned development, and it is calculated that 60 percent of Haitian households do not have any formal document of property ownership (USAID 2010). The lack of a national land cadastre is a major obstacle to effective property rights, which are pivotal to carrying out large-scale urban housing or infrastructure investment programs. Efforts are underway to introduce such a cadastre, but the process is highly challenging due to the fragmentation of the land registry system. Governance challenges remain the overarching hurdle to long-term resilient urban growth. Since 2010, the government has focused on reconstruction activities. Recently, however, it is undertaking a broader effort, transitioning to comprehensive forward-looking urban planning. The government is promoting decentralization to increase local governments’ role in urban planning through the Strategic Development Plan of Haiti, which emphasizes territorial reform as a gateway to achieving the country’s development objectives. Strategic plans have been developed to guide decision making across different levels and sectors of government. However, their effective implementation to shape Haiti’s urban areas faces two major constraints: plans may exist in law, but are not implemented in practice; and, where plans are developed, there is a gap between expectations set out by the plans and the financial and technical capacity to implement them. Resilient urban planning is central to Haiti’s social and economic development. Forward- thinking planning must be embraced to steer urban growth in the direction of rising incomes and economies of agglomeration, and away from congestion and heightened exposure to risk. The shape and form that cities take can have real impacts on productivity and livability. Smart and targeted investments are needed to reap the benefits of urbanization and control the associ- ated economic, social, and environmental costs that it may bring. Planning for resilient devel- opment is about supporting coordinated action to help shape urban growth so that it supports a country’s (and its cities’) development objectives, while managing natural disaster risk to protect hard-won advances in living conditions. As the pressure of urban population growth intensifies, laying the foundations for cities to work through better planning will be required. The shift toward resilient urban planning requires actions in the following three areas to address current infrastructure deficits and prepare for urban growth. 4 Refer to the “Doing Business 2017 – Equal Opportunity for All” report to see how Haiti ranks in other indicators that shed light on the country’s business environment. 5 In less than thirty years (1986-2014) the country was led by a succession of eighteen short-lived governments and suffered repeated delays in elections, which led to Parliament’s dissolution in 2015. 6 Invest to address the basic service deficits Big gaps in urban services such as water, sanitation, and solid waste collection call for large investments immediately. In the short term, Haiti can invest in basic services by lever- aging community engagement and by improving service management and delivery through local government capacity building. Community engagement and participatory approaches are key to successfully upgrade access to services in areas where development has taken place in an unregulated manner, as they are linked to improved confidence in government and long-term sustainability of urban development. Haiti’s experience in community-driven devel- opment (CDD) projects attests to their potential, but their design could be improved to better address particular urban challenges such as high levels of violence, criminal activity, and social exclusion. In the near future, the government can consolidate basic service delivery by building on “what works”; in other words, improvements in basic services can be viewed as a ladder by which each modernization effort builds capacity and paves way for new and more advanced initiatives led by local governments. The national government can help build capacity and improve local service provision by providing the right incentives (such as grants, subsidies, or transfers to municipalities based on specific outcomes or performance in service delivery). Leverage information to facilitate coordinated decision making To manage unregulated growth and minimize the risk exposure of Haitian cities, households, firms, and local governments need to be provided with relevant land use planning and risk infor- mation. Specifically, resilient planning in cities can be achieved by disseminating risk analysis insights to support decision making, placing trunk infrastructure for basic services ahead of devel- opment, and integrating hazard risk knowledge into transparent urban infrastructure investment decision making. Information on risk in urban areas must be made publicly available to support non-structural measures for protecting people from risk, such as emergency planning and infor- mation-based campaigns to encourage flood risk mitigating behavior. Following the 2010 earth- quake, Haiti developed tools to strengthen disaster risk management information in planning, such as multi-hazard risk assessments, seismic zonation mapping, and location of exposed assets. Technological innovation can provide new opportunities to engage citizens and disseminate infor- mation on risks. In Tanzania, unmanned aerial vehicles—drones—were used to map floodplains in Dar es Salaam, the country’s largest city. The information was used to plan and predict how water would move in the event of a flood. This cost-effective approach could be considered in Haiti. It is a more expensive and complicated task to provide basic services to unplanned areas than it is to put in place trunk infrastructure ahead of development. Planning ahead will save financial resources. Although plans can be an effective tool in anticipating urban growth, local actors sometimes lack the funds or incentives to implement the decisions. In this case, simple, disseminated plans can be highly effective in guiding new development. Compliance with plans can be encouraged by making clear and credible information available to households and firms so that they can make better, informed decisions. Such is the case of Tunis, Tunisia where, rather than restricting urban expansion into unplanned areas, households were provided with clear infor- mation on future infrastructure expansion plans, which helped to secure rights of way for future 7 investment (World Bank 2014). To minimize risk exposure, important win-wins could be achieved by integrating flood risk management with broader development objectives, such as in the northern corridor. The north and northeast areas of Haiti experience population growth pressures, deficits of basic services and transportation, and significant flood risk. Inaction in these places will lead to increasing numbers of people at risk. Strengthen property rights and promote institutional reform for improved governance Resilient planning is a long-term effort that demands institutional reform and strengthening, but stepping stones can be put down today. The government can begin with specific actions, including strengthening property rights with dispute-resolution mechanisms and promoting municipal cooperation in Haiti’s largest cities. The establishment of a single and accurate record of land ownership is vital for resilient development, but there are many challenges to cadastral reform and the establishment of an effective land registration system. In Haiti, legal uncertainty over property rights stands out as one of the main hurdles, and any effort to establish formal land titles depends on the broader institutional structure for property rights. Initial efforts, however, can be made by strengthening alternative dispute-resolution mechanisms to help solve the backlog of unresolved property rights disputes, which may facilitate the creation of an official land registry. Also, building frameworks for municipal cooperation becomes increasingly important in light of Haiti’s continued progress toward political and fiscal decentralization. Effective service provision is a priority for institutional reform, yet coordination across municipal, and even departmental, boundaries is necessary to avoid duplication of activities or contradictions in policies. This can be achieved by developing coordination frameworks to promote cooperation. CONNECTING: BETTER WITHIN-CITY CONNECTIVITY AND ACCESSIBILITY ARE ACHIEVABLE THROUGH IMPROVED MOTORIZED TRANSPORT AND ENHANCED COORDINATION BETWEEN LAND USE AND TRANSPORT INVESTMENTS Today, poor connectivity within Haitian cities hampers integrated labor markets and access to economic opportunities. Accounting for within-country variation, accessibility to economic oppor- tunities in Haitian cities is poor. Public transport in urban areas is unaffordable to many, limiting their access to economic opportunities, especially among the poorest households.6 Tap-Taps are the most widely used form of public transportation in Port-au-Prince, covering 56 percent of the market.7 Although regulated by the Ministry of Social Affairs, Tap-Tap fares are unaffordable to many poor households. Based on current fare costs, if a Tap-Tap journey was repeated twice a day, five days a week, transport expenditures would represent anything between 25 and 73 percent of per capita expenditures in the lowest quintile. The proportion of households that spent nothing on transport in 2011–2012 was 57.1 percent, according to the Haitian Household Expenditure Survey (IHSI 2012), meaning that just under 60 percent of households used no motorized transport whatsoever. Haiti is 6 Most information on urban transport patterns in the metropolitan area of Port-au-Prince was extracted from a report “Urban Transport in Port-au-Prince” prepared by Kopp and Prud’homme (2011) for the Inter-American Development Bank. 7 Tap-Taps are converted pick-ups often imported from the United States and Canada, and can seat 10-14 people but often accommodate up to 20. These informal minibuses are the dominant collective transport service in urban areas. 8 COMMUTERS DURING DAYTIME FOR WORK-RELATED ACTIVITIES (LEFT) Figure O.2. AND DURING THE EVENING (RIGHT) Source: Authors’ elaboration using Digicel data. in the process of eliminating subsidies on fuel prices, which could worsen the affordability of public transport without compensatory measures. Though well-intended, because they are highly regres- sive, the removal of these subsidies may result in higher costs to Tap-Tap owners if they must absorb the fuel cost increases. Any increase in transport fares intended to offset extra costs would exacer- bate the unaffordability and further exclude the poor from accessing economic opportunities. This report is the first to use mobile phone call records to understand commuting patterns and the degree of spatial mismatch between jobs and homes in Haiti. A good understanding of how workers move in and around the largest cities, where the job centers are and how accessible they are for different segments of Haitian urban society, and what the most critical road segments are to ensure that job accessibility not be affected by a disaster, can provide valuable information for evidence- based decision making. Absent up-to-date censuses and travel surveys, accessibility to opportuni- ties in Haitian cities was measured using mobile phone data. Call data records8 were utilized to track locations of users throughout the day, and even at night. This information was plotted into maps, which showed where jobs were geographically concentrated in the two main urban centers of Port-au-Prince (Figure O.2) and Cap-Haïtien. The overall picture is one of concentration toward the city center during daytime, where most jobs are located, and inversely one of diffusion toward the outskirts during the evening. 8 Digicel, the main mobile phone subscription provider in Haiti (with close to 80 percent of the market), granted the team access to a sample of de-identified call records. This technique was used to overcome data scarcity in Haiti. 9 An analysis of commuting patterns shows that labor markets are fragmented in Haiti. Port-au- Prince and Cap-Haïtien suffer from low employment accessibility as measured by commuting patterns. Data show that only a small share of people in both cities travel to work and the distances traveled by these commuters are short. Only 42 and 40 percent of the population are considered commuters in Port-au-Prince and Cap-Haïtien, respectively, traveling beyond their home cluster (1km radius). These patterns are indicative of local job matching, meaning that access to a large array of economic opportunities is low. While short commutes are not negative in themselves, they may reflect the difficulty and unaffordability of traveling. Motorized travel is slow and lengthy in Port-au-Prince, mainly due to a combination of lack of road infrastructure, poor road maintenance, and suboptimal use of public street space. Based on travel logs of various vehicles, speeds in Port-au-Prince are as low as 10.9km/hour on average. Public transport travels at slower speeds, with Tap-Tap users reporting an average on-board travel time of 44 minutes for an average commute distance of 5.9km, suggesting average speeds for Tap-Taps of 8km/hour. Part of the problem is lack of road space, but Kopp and Prud’homme (2011) argue that congestion can mainly be attributed to the suboptimal use of street space and poor road maintenance. Encroachment of the public realm occurs when sidewalks are used by merchants for commercial purposes, forcing pedestrians to walk in the traffic lane. In addition to pedestrians’ use of roads, road space is also compromised by parked vehicles. Together, these factors lead to considerable loss of speed and heightened safety risks for pedestrians. Cities should become integrated labor markets that provide opportunities to residents, allowing them to choose jobs from larger pools and thus leading to increased welfare (Bertaud 2014). Integrated labor markets exist when it is possible for an individual to reach a large share of the employment opportunities within a city at a reasonable cost or within a reasonable period. Large and integrated labor markets support improved matching by increasing the number and diversity of employers and job seekers, which makes the best of their skills and aspirations. When access is good, firms also benefit from the proximity to product and labor markets that the density of cities allows. Conversely, when accessibility is limited the likelihood of finding a good match is smaller, because firms and households must select from a smaller pool of workers and employment options. Jobs outside high-density economic clusters tend be scarcer, more informal, and lower paying. Low accessibility levels also pressure families to locate closer to jobs, which can turn out to be a disadvantage given that land and housing are more expensive, in turn forcing these families to live in basic conditions and fueling the growth of central slums found, for example, in many African cities (Antos, Lozano-Gracia, and Lall 2016). Connectivity and accessibility are necessary—but not sufficient—conditions to achieve efficient urban labor markets. Creating jobs and achieving efficient labor markets demand a multidimensional solution to overcome many obstacles, ranging from the lack of a financial and banking system to create businesses, to low education levels, and a costly regulatory framework. Improving accessibility will not solve all these issues, but failing to address urban accessibility will impede progress in productivity and livability. To address connectivity challenges in Haitian cities, actions along the following three lines are essential. 10 Improve travel speeds and quality of service through more investments and enhanced efficiency in urban areas There are various avenues to increase speeds for motorized transport in urban areas, which could improve accessibility and help cities become better matchmakers. Investing heavily in roads and public transport is certainly a way to do this, although such an option would require large financial resources and is unlikely to be very effective before the chronic challenges on the current network are addressed. Therefore, a more immediate solution is to focus first on improving the operation and maintenance of the current network. These two options are not mutually exclusive, but the most effective sequencing would be to start by improving the current network. Less expensive alternatives, such as improved traffic management, road space allocation, and road maintenance can yield significant results. Improving road space requires freeing up road lanes for circulation rather than to accommodate street parking or pedestrians. At the same time, sidewalks must prioritize pedestrian mobility over street vendor activity9 and ensure comfort and safety of those commuting by foot. Road maintenance— repairing potholes and uneven road surfaces—can save vehicle maintenance costs and travel time. A first step would be to replenish Haiti’s existing road maintenance fund. In the longer term, collective transport lanes could be a promising approach to reduce travel times in urban areas. Increase the affordability of collective transport for inclusive matching of opportunities Reductions in operating costs could help lower transport fees and improve affordability, via several means. First is to increase speeds on the road network, through interventions on the network and by rationalized Tap-Tap routes to allow Tap-Tap drivers to complete more round trips in a given time. This option would increase the revenues and margins of Tap-Tap operators and could lead to lower fares, and is the most promising. Another approach would be making Tap-Tap vehicles – often operated for more than 25 years (Kopp and Prud’homme 2011) – more fuel efficient to lower the volume of fuel required. Public interventions to scrap old, fuel-inefficient, informal minibuses and subsidize the purchase of more efficient vehicles were adopted in Senegal and the Dominican Republic, and the lessons learned from these experiences could help in Haiti. The urban area of Dakar (Senegal), for example, created incentives for informal car rapide (minibus) operators to buy more fuel-efficient minibuses from 2003 to 2008, providing subsidized loans that covered around 75 percent of the purchase cost of the vehicle (Kumar and Diou 2010). In exchange, car rapide owners had to retire their old vehicle and formalize their activities. Using the lessons from Senegal, a suitable model for the Haitian context could be designed, such that it can be negotiated with local operators. An ongoing technical assistance of the World Bank is exploring different mechanisms to offset fuel-cost increases, and the results will inform whether scrapping old Tap-Tap vehicles is a viable option. In the longer term, carefully targeted transport subsidies could be directed toward the poorest households to ensure they obtain or retain access to opportunities. 9 Digicel, the main mobile phone subscription provider in Haiti (with close to 80 percent of the market), granted the team access to a sample of de-identified call records. This technique was used to overcome data scarcity in Haiti. 11 Strengthen coordination of land use and transport investments for improved access and increased resilience Interventions aimed at coordinating land use and transport reduce the disconnect between residential areas and employment opportunities and help build resilience in the wake of natural hazards. Two main ways of improving accessibility to opportunities in urban areas have been identified: increasing speeds and reducing distances. The first calls for investing in the connective network and making motorized transport more affordable, while the second entails reducing the fragmentation of the urban footprint by incentivizing density of people and opportunities and better integrating land use and transport. At present, population densities in Port-au-Prince and Cap-Haïtien are high, so there is limited room to reduce the distance between people and economic opportunities by increasing densities further. However, modifying the spatial layout to encourage land use clustering can increase accessibility within a given period. There is also considerable room to advance accessibility by planning for urban expansion while reducing exposure to natural hazards. Both cities show examples of urban development, either in safer but poorly connected areas or closer to economic opportunities but in riskier locations. It is important to avoid these trade-offs and carry out investments that prioritize both measures. A first step in building a strategy for increased resilience against natural disasters is under- standing which road sections are the most critical links in the network. Based on a criti- cality analysis, the most urgent road segments for intervention in Port-Au-Prince’s network are the Route Nationale 2 that connects downtown to Carrefour and beyond to the west, RN1 connecting the downtown and north areas of the capital, an isolated link between downtown and Pétionville, and a couple of links connecting Canaan to the rest of the network. FINANCING: STRENGTHENING MUNICIPAL FINANCES IS ESSENTIAL TO CLOSE THE URBAN INFRASTRUCTURE AND SERVICES GAP, AND TO ACCOMMODATE THE GROWING URBAN POPULATION Given its current rapid urbanization, Haiti faces big challenges in strengthening public finances. Today, the ability of local governments to plan, service, and connect cities and towns is heavily constrained by limited resources at the municipal level. Over the past two decades, the gap between the funding capacity of Haiti and the pace of urban growth has led to a constant deficit of urban infrastructure and services in cities and towns. The progressive decline of international aid further widens the funding gap. According to World Bank (2016), Haiti faces critical challenges, including adapting to financial reductions, raising more revenue internally, and making better use of existing funds. Despite improvements in the country’s fiscal revenue, from less than 10 percent of GDP in 2004 to 12.6 percent of GDP in 2014, Haiti remains the poorest performer in revenue mobilization in LAC (World Bank 2016). This greatly hinders the country’s ability to carry out much-needed devel- opment spending in infrastructure, health, education, and other key sectors. As cities expand in size and population, the challenge to finance sustainable and inclusive urban development grows. Municipal governments are unable to provide adequate infrastructure and services due to incom- plete decentralization and a weak legal framework for municipal finance. The Constitution of 1987 (including the 2012 amendment) and the Presidential decrees of 2006 establish the fiscal and 12 financial autonomy of communes, the decentralization of public services provision, and the institu- tionalization of municipal revenue. But while the decentralization framework is in place, effective devolution of key expenditure and revenue functions to municipal governments has not yet taken place. A fragmented municipal finance framework hampers the ability of local governments to raise revenue to finance service provision. Thus improving local government capacity for public financial management remains key to successful devolution of responsibilities to communes and to effective fiscal decentralization. Only 0.6 percent of GDP is currently spent at the communal level, and total municipal revenue makes up only 1.7 percent of total revenue. It is important to address the inconsis- tencies in devolution and decentralization so that function follows finance. Limited and unpredictable sources of municipal revenue undermine capacity for planning, budgeting, and delivery of much-needed services in urban areas. Local governments have four main sources of revenue: transfers from the central government; taxes collected on behalf of the communes by the General Tax Office (DGI); duties and royalties collected by the communes; and other external sources (such as development partners). However, local governments are dispro- portionately dependent on national transfers, mainly the Local Government Development Fund (FGDCT). Except for Port-au-Prince, Pétionville, and Delmas, it is the main source of income for communes, typically ranging between 80 and 95 percent. The transfer system, however, lacks trans- parency and is unreliable. Only half of the funds designated for the communes are transferred, and a significant share is transferred to inactive or nonexistent structures. Additionally, national transfers reduce the incentives for improving public financial management in local governments, as these are neither linked to improved performance in service delivery nor are they entirely needs- based. The financial capacity of local authorities is further curtailed by their limited tax collection capacity. Municipalities often only collect a fraction of their revenue potential. Own-source revenue is highly concentrated in major cities, with the five communes of the Port-au-Prince metropolitan area collecting 80 percent of all Haitian communes’ own resources. Better planning, connecting, and servicing cities and towns in Haiti require substantial inter- vention to review, revise, and scale their financing arrangements. Enhancing transparency in fiscal transfers, developing more dynamic sources of local revenue, and strengthening municipal finances are crucial steps in guiding urban growth away from crowding and toward healthy and productive densities. Present levels of resources fall extremely short of current demand for services. There is an urgent need to improve the volume, predictability, timeliness, and management of finances, and to identify additional mechanisms for generating own-source revenue. This requires a systematic effort to adjust and implement reforms aimed at improving national and local government management and the oversight of resources. Since May 2014, the government has developed a comprehensive public financial management reform strategy and action plan aimed at ensuring a financial system that promotes transparency, accountability, fiscal discipline, and efficiency in using and managing public resources by growing revenue from taxes and tariffs, thus increasing local government autonomy. However, limited progress has been made to date. The following paragraphs highlight the key actions needed to strengthen municipalities’ finances and their capacity to provide much-needed local services. 13 Consolidate, harmonize, and enforce the legal and regulatory framework for municipal financing Policy options should first focus on fixing the gaps in the institutional, regulatory, and financing framework for local governments. Haiti’s urban development is taking place in a context of incom- plete decentralization and an unclear legal framework for municipal financing. A regulatory framework that provides clarity of roles and resources is crucial for effective decentralization. Steps in this direction include reviewing the normative framework of municipalities, as stated in the decrees of 2006, and identifying actions for implementation; formalizing the taxation functions and responsibilities of municipal governments; and reviewing legislation and regulations, particularly those related to property tax and business tax. Together, these actions seek to clarify the respon- sibilities, systems, incentives, and accountability relationships for the financing and delivery of services, and the capacity of local governments to manage and allocate increased funding. Recent reform work has opened the door to opportunities to deepen decentralization efforts. A draft law on the financial autonomy of communes and communal sections is on the legislative agenda. Strengthen the system for municipal finance to build capacity and accountability, and expand financial opportunities A stronger municipal financial system is needed to increase the financial autonomy of local govern- ments. As mentioned, cities are heavily constrained by limited revenue sources and are highly reliant on transfers from the national government. The solution may differ between small and large cities. For smaller cities, efforts can be focused on enhancing the management, oversight, and transparency of the FGDCT, including fund mobilization, allocation, transfer, expenditures, and accounting, and using the fund as an opportunity to build local capacity for implementation. Larger cities, on the other hand, must prioritize building capacity for own-resource revenue collection, management, and spending. Expand and leverage the local revenue base Municipalities need more revenue autonomy through access to strong and broader sources of local revenue. The financing system in place is not working to the advantage of local governments: not only are revenue levels too low to meet the demand of public services and infrastructure, but the options to increase revenue are flawed and leave little room for accessing adequate financial resources. For municipalities to generate and collect own-source revenue, efforts must be geared toward strength- ening the planning and budgeting capacity of municipalities, including forecasting of revenue. To improve the financial management capacity of local governments, the Ministry of Economy and Finance (MEF) and the Ministry of Interior and Local Authorities (MICT) could develop a local government public financial management manual, which sets the basic standards and procedures in budgeting, accounting, reporting, procurement, and audit. This improvement could be achieved through capacity-building programs that focus on four main areas: strengthening the administrative capacity of the municipal financial units; strengthening the capacity of municipalities in project management for timely disbursements of FGDCT funds allocated to them; increasing municipality revenue mobilization capacity including enhancing technical competencies of staff; and providing municipalities with incentives to explore alternative financing mechanisms. 14 “LIVE TODAY, BUT THINK ABOUT TOMORROW” (WÈ JODI A, MEN SONJE DEMEN) The challenges discussed in this report need action today. But, as the proverb says, policymakers need to think about tomorrow. Not everything can happen today, so defining what must take place now and how that will open the way for much-needed changes is essential. Table O.1 summarizes policy recommendations and puts forward specific actions that can be taken in the short, medium, and long terms, distinguishing between high-, medium-, and low-pri- ority actions, as well as identifying the institutions leading the actions. Most actions require the engagement of more than one institution, thus underscoring the importance of collaborative efforts in achieving urban development that benefits all. 15 Table O. 1. SUMMARY MATRIX OF POLICY RECOMMENDATIONS: PLANNING, CONNECTING, AND FINANCING CITIES IN HAITI Key to Table PRIORITY LEVEL TIME HORIZON H High M Medium L Low S Short-term M Medium-term L Long-term (next 12 months) (next 1–3 years) (next 3-5 years) MOVING FROM RECONSTRUCTION TO RESILIENT URBAN PLANNING FOR A BRIGHT FUTURE A shift toward resilient urban planning is needed to address current infrastructure deficits and prepare for future urban growth TIME PRIORITY LEADING BROAD HORIZON LEVEL INSTITUTION(S)/ ISSUE RECOMMENDATION SPECIFIC ACTIONS (S, M, L) (H, M, L) CHAMPION(S)* High deficits in basic Invest in resilient Upgrade and extend access to services; leverage S H MTPTC urban services, and infrastructure to address past community-driven efforts, but at the same Communes current deficits in basic time build capacity now for the long term limited investment in urban services. infrastructure to meet Improve basic service management and S H MICT growing population delivery while building and consolidating local Communes government capacity through performance- needs based mechanisms. Disseminate risk-analysis insights to support M H MICT-DPC informed decision making and vital non- CNIGS structural measures to protect people and assets MTPTC from risk. Uncoordinated Leverage information Use information to align incentives: inform S M MPCE growth of cities with to facilitate coordinated citizens about future plans and risks so that Communes decision making among they can make better decisions. insufficient regard to households, firms, and natural disaster risk government. Integrate flood risk knowledge with M M MTPTC transparent urban infrastructure investment MARNDR decision making. Conduct vulnerability assessment of critical S H MTPTC public infrastructure. Weak land Strengthen property rights Strengthen property rights with dispute L M CIAT administration, and promote institutional resolution mechanisms. reform for improved opaque information governance. For Haiti’s largest cities, build frameworks for L M MICT on land ownership, municipal cooperation. MPCE and inappropriate land regulation SHAPING LABOR MARKETS: CONNECTIVITY, JOBS, AND RISKS Better within-city connectivity and accessibility are achievable through improved motorized transport and enhanced coordination between land use and transport investments TIME PRIORITY LEADING BROAD HORIZON LEVEL INSTITUTION(S)/ ISSUE RECOMMENDATION SPECIFIC ACTIONS (S, M, L) (H, M, L) CHAMPION(S)* Fragmented labor Increase speeds and Better manage road and sidewalk space for S H MTPTC markets caused by the improve quality of transport increased speeds, more pedestrian comfort, Communes through more investments and decreased road accidents. spatial mismatch and enhanced efficiency. between economic Guide urban expansion toward accessible and S H MPCE opportunities and safe locations, and secure rights of way for Communes future infrastructure investments. residential locations Invest in road maintenance for lower future S H MTPTC costs of repairs and increased speeds. Unaffordable and slow Leverage information Build Tap-Tap stops and dedicated public M M MTPTC public transport to facilitate coordinated transport lanes for increased speeds and Communes decision making among accessibility and lower costs for operators and system households, firms, and fares for users. government. Promote retirement of fuel-intensive Tap-Tap M M MEF vehicles to lower operators’ costs and travel fares and to reduce vulnerability to an increase in fuel prices. Limited coordination Strengthen coordination of Build resilience of the transport network S H MTPTC between land use and land use and transport in- by identifying the most critical links and DPC vestments for better access upgrading or investing in redundancy. transport planning and resilience. reduce accessibility In parallel, enforce building codes to minimize L M MTPTC and increase the impact of natural hazards such as Communes vulnerability of the earthquakes. network Develop registers and statistical systems L M IHSI for targeted demand-side public transport subsidies for the poorest and most vulnerable. * LEADING INSTITUTION is shown bolded. The institutions are abbreviated as follows (in alphabetical order): DPC Directorate for Civil Protection MARNDR Ministry of Agriculture, of Natural Resources, and Rural Development DGI General Tax Office MEF Ministry of Economy and Finance IHSI Haitian Institute of Statistics and Information MICT Ministry of Interior and Local Authorities CIAT Inter-Ministerial Committee for Territorial Development MTPTC Ministry of Public Works, Transport and Communications CNIGS National Center for Geo-spatial information MPCE Ministry of Planning and External Cooperation FINANCING HAITIAN CITIES Strengthening municipal finances is essential to close the urban infrastructure and services gap, and to accommodate the growing urban population TIME PRIORITY LEADING BROAD HORIZON LEVEL INSTITUTION(S)/ ISSUE RECOMMENDATION SPECIFIC ACTIONS (S, M, L) (H, M, L) CHAMPION(S)* Incomplete Consolidate, harmonize, Review the normative framework of the ter- S H MICT decentralization and and enforce the legal and ritorial collectivities as established in the five regulatory framework for decrees of 2006 and identify possible actions a fragile and municipal financing. for implementation. fragmented legal framework that Formalize the taxation competencies entrusted S H MICT to the municipalities as stipulated in Article MEF-DGI governs municipal 142 of the decentralization framework. finances Revisit municipal tax laws, particularly those M H MEF-DGI related to property tax and business tax. MICT Limited sources of Strengthen the system Strengthen tools available for linking invest- S M MICT municipal revenue for municipal finance ment planning, budgeting, and execution for Departments and expand financing local governments. Communes opportunities. Build local capacity for budget managing for M M MICT timely execution of budget and better service Departments provision. Communes Carry out property assessments in all S H MEF-DGI municipalities and update the property tax Communes registry accordingly to broaden the tax base. Lack of transparency Conduct a diagnostic of the inefficiencies in L M MEF and limited reliability the FGDCT and agree on an action plan. MICT of the transfer system Implement an action plan for enhancing L M MEF FGDCT (allocation, management, transfer, MICT and monitoring and evaluation) and initiate drafting of intergovernmental fiscal strategy. * LEADING INSTITUTION is shown bolded. The institutions are abbreviated as follows (in alphabetical order): DPC Directorate for Civil Protection MARNDR Ministry of Agriculture, of Natural Resources, and Rural Development DGI General Tax Office MEF Ministry of Economy and Finance IHSI Haitian Institute of Statistics and Information MICT Ministry of Interior and Local Authorities CIAT Inter-Ministerial Committee for Territorial Development MTPTC Ministry of Public Works, Transport and Communications CNIGS National Center for Geo-spatial information MPCE Ministry of Planning and External Cooperation REFERENCES Antos, S. E., N. Lozano-Gracia, and S. V. Lall. 2016. “The Morphology of African Cities.” Policy Research Working Paper 7911, World Bank, Washington, DC. Bertaud, A. 2014. “Cities as Labor Markets.” Working Paper #2, Marron Institute on Cities and the Urban Environment, New York University, New York. Deuskar, C., B. P. Stewart, and N. Lozano-Gracia. 2016. “Defining Urban Areas in Haiti.” Working Paper, World Bank, Washington, DC. Hoornweg, D., and P. Bhada-Tata. 2012. “What a Waste: A Global Review of Solid Waste Manage- ment.” Urban Development Series Knowledge Paper No. 15, World Bank, Washington, DC. IHSI (Haitian Institute of Statistics and Informatics). 2012. Survey on the Living Conditions of Households after Earthquakes (database). Port-au-Prince: IHSI. Kopp, P., and R. Prud’homme. 2011. Urban Transport in Port-Au-Prince. Washington, DC: Inter-American Development Bank. www.rprudhomme.com/resources/Rap+2011+Port+au+P rince+$28BID$29.pdf. Kumar, A., and C. Diou. 2010. “Renouvellement Du Parc d’autobus a Dakar : Avant - Apres.” 11. Sub-Saharan Africa Transport Policy Program (SSATP) Discussion Paper. Washington, DC: World Bank. Singh, R. J., and M. Barton-Dock. 2015. Haiti: Toward a New Narrative: Systematic Country Diagnostic. Washington, DC: World Bank. UN World Population Prospects. 2015. United Nations, Department of Economic and Social Affairs, Population Division. 2015. World Population Prospects: The 2015 Revision, Key Findings and Advance Tables. Working Paper No. ESA/P/WP.241. USAID. (2010). Land Tenure And Property Rights In Haiti The Importance Of Land Tenure And Property Rights Issues and Post-Earthquake Recovery In Haiti. Property Rights And Resource Governance Briefing Paper 6. ———. 2014. Tunisia Urbanization Review: Reclaiming the Glory of Carthage. Washington, DC: World Bank. ———. 2015a. Building Regulation for Resilience: Managing Risks for Safer Cities. Washington, DC: World Bank. ———. 2015b.  Haiti: Country partnership framework for the period FY16-FY19. Washington, DC: World Bank. ———. 2016. Better Spending, Better Services: A Review of Public Finances in Haiti. Washington, DC: World Bank. ———. 2017. Doing Business 2017: Equal Opportunity for All. Washington, DC: World Bank. World Bank and ONPES (Observatoire National de la Pauvreté et de l’Exclusion Social). 2014. Investing in People to Fight Poverty in Haiti: Reflections for Evidence-based Policy Making. Washington, DC: World Bank. WorldPop. 2015. Database. Southhampton, UK: GeoData Institute, University of Southhampton. www.worldpop.org.  19 CHAPTER 1 LAYING DOWN THE FACTS ON HAITIAN URBANIZATION Andrea Colombo GONAIVES, ARTIBONITE. PHOTOGRAPHED BY ERIK BARKER, 2008 SOURCE: WIKIMEDIA, CREATIVE COMMONS LICENSE CHAPTER 1 – LAYING DOWN THE FACTS ON HAITIAN URBANIZATION Around the world, urbanization has been the level of urban population increased to 58 accompanied by growth, with incomes rising percent, but the country’s GDP per capita as the share of population in cities increases. remained stagnant, and in fact dropped to Historical data between 1996 and 2015 for USD 727. By way of comparison, in the Latin over 180 countries suggest a positive relation- American and the Caribbean (LAC) region, ship between Gross Domestic Product (GDP) Honduras and Guatemala had urban levels per capita and levels of urbanization (Figure similar to Haiti in 1996, but their GDP per 1). Generally, as countries become more capita increased respectively by 41 and 28 urbanized, their economies tend to perform percent as they became more urbanized better. The links between GDP growth and (Figure 1). Fragility and economic shocks urbanization have been highlighted broadly from natural disasters have played a large role by previous work, and international evidence in hampering the benefits of urbanization shows that most countries reach middle-in- in Haiti. Among fragile countries, Haiti and come status only after a significant popula- Côte d’Ivoire show similar patterns of urban- tion shift into cities (WDR 2009). The density ization without growth (Figure 1). However, found in cities provides fertile ground for the experience of Côte d’Ivoire also suggests economies of scale and higher productivity, that as stability is built, cities, if well managed, increasing opportunities that may eventually can be leveraged as engines of growth. help lift many people out of poverty.1 Contrary Understanding current urban challenges to international trends, Haiti has urbanized is a key step into long-term planning for without GDP growth. According to the World economic success. What is different in Development Indicators (WDI), 33 percent of Haitian cities? Why are they not engines of Haitians lived in urban areas in 1996. In that growth? What are the key bottlenecks that year, the country’s GDP per capita averaged prevent transforming Haiti’s story of urban- at USD 757 (in constant 2010 prices). In 2013, ization into one of growth and prosperity? 1 See, for instance, Annez and Buckley (2009) and the World Bank’s World Development Report (2009).areas. *This chapter benefited from comments by Nancy Lozano Gracia, Claudia Soto, Olivia D’Aoust, and Paolo Avner. The analysis of the size and expansion of Haitian urban areas draws from Deuskar, Stewart, & Lozano-Gracia (2016) and was made possible thanks to the assistance of Sarah E. Antos and Katie L. McWilliams. Thanks to Emilie Perge for providing data from ECVMAS 2012 and ECVMAS 2013 for the analysis of Haitian households. 23 A POSITIVE CORRELATION BETWEEN GDP PER CAPITA AND Figure 1. URBANIZATION LEVELS PREVAILS ACROSS THE WORLD, BUT NOT Figure 1. FOR HAITI (1996-2010) Source: World Development Indicators, own calculations. 24 This chapter aims to answer these questions and security concerns have added further by looking at the trends of urbanization and constraints to sustainable and productive examining the key characteristics of cities urbanization. Between 1986 and 2014, Haiti across Haiti, accounting for the elements of witnessed eighteen changes of government fragility and disaster risk. First, the chapter and over twenty major cabinet changes. focuses on Haiti’s urban growth in terms of Empirical research has shown that such insti- population and land, as well as the challenges tutional volatility is detrimental to growth; related to how urban areas are defined. It then recent estimates suggest that Haiti would have provides an overview of the system of cities grown 1.2 percent faster if it had achieved in the country, looking at their economic and an average level of stability across years (See labor market characteristics. The chapter Aisen and Veiga 2013, and Singh, Bodea and moves on to discuss some of the challenges of Higashijima 2016). Further, growing levels urban poverty in its monetary and non-mon- of crime and violence, particularly in urban etary forms, with a focus on lack of basic agglomerations, hamper investment and services and infrastructure. It concludes by growth as businesses are usually forced to linking the challenges of Haiti’s urbanization shut down or move operations elsewhere due to three policy pillars that are the subject of the to high security costs. remaining chapters: Planning, Connecting, Exposure to multiple natural disasters and Financing. throughout most of the territory has made the road toward development steeper and gravely URBANIZINGWHILE GRAPPLING WITH WIDE- limited Haiti’s economic growth potential. SPREAD FRAGILITY AND DISASTER RISK Most of Haiti’s territory is exposed to at least One cannot narrate Haiti’s story of urban- one kind of disaster (Figure 2), with cities at ization without considering the country’s state larger risk for disaster loss given that they of fragility.2 Like other fragile states, Haiti has increasingly concentrate people, assets, and grappled with deep poverty and inequality, infrastructure. Between 1971 and 2013, the economic decline and unemployment, insti- economy was hit – to various degrees – by tutional weakness, and violence (World Bank the occurrence of disasters almost every year 2007). For many years, poor governance led to (Singh and Barton-Dock 2015). It is calculated minimal investments in basic infrastructure that hydrometeorological events have caused and deterred an environment favorable for the country an average of USD 150 million growth. Haiti’s long history of government per year in losses and damages, or 1.7 percent neglect of basic services provision and poor of GDP (1976-2014). While localized floods expenditure on key infrastructure explains have a limited economic impact (less than 2 today’s state of cities: unplanned, unser- percent of GDP), Hurricane Jeanne in 2004 viced, and overcrowded. Political instability and major cyclones registered in 2008 (Faye, 2 The World Bank Group classifies Haiti as a ‘fragile’ state due to its low Country Policy and Institutional Assessment (CPIA) ratings for economic, social, and public sector polities and institutions. Broadly, it is defined as the weakness of institutions and vulnerability to instability, conflict, and violence. For more information, see the Country Partnership Framework for the Republic of Haiti for the Period FY16-FY19: https://openknowledge.worldbank.org/handle/10986/23127 25 Gustave, Hanna, and Ike), caused losses and Fragility and natural disasters have also damages equivalent to 7 and 14.6 percent of shaped the social capital in Haiti in different GDP respectively.3 Seismic disasters are much ways. Social capital is traditionally measured less frequent, but could result in larger human via indices of political activism, interpersonal and economic losses.4 The 7.3 Richter-magni- trust, and trust in institutions.8 Post-earth- tude earthquake that struck Haiti on January quake surveys in Haiti suggest an increasing 12, 2010, followed by at least fifty aftershocks, proportion of Haitians have become more resulted in massive economic and human active in the political life of the country in losses, with economic costs representing 117 the aftermath of the disaster.9 Citizens have percent of national GDP. become increasingly involved in community During the first weeks after the earthquake, activities, with the proportion of residents large population movements out of the capital participating in some kind of association were reported by the United Nations Popula- within affected municipalities increasing 5 tion Fund. Using mobile phone network data, from 34.9 percent in 2008 to 45.6 percent in it was estimated that 630,000 people present in 2010. The same survey suggests that Haiti may Port-au-Prince (PaP) the day of the earthquake have one of the highest participation rates in left the capital within 19 days after the disaster. the entire LAC region, with almost 80 percent Overall, 20 percent of the PaP pre-earthquake of the surveyed population declaring to have population moved out of the epicenter’s area, taken part in at least one civic association seeking shelter in neighboring municipalities, twelve months before the survey. However, including Les Cayes and Saint-Marc. Seven building trust appears to remain a challenge. months after the disaster, 1.5 million people According to the Latin American Barometer, were living in 1,555 temporary camps.6 Three Haiti now has the lowest rate of interper- and a half years later, 172,000 people were still sonal trust in the region. Between 2008 and living in crowded conditions in 306 camps. 2010, the proportion of interviewees trusting Those who had left the camps did not neces- their peers decreased from 41 percent to 33 sarily find permanent housing.7 Today, there percent. Crime rates also increased consid- remain many buildings that have yet to be erably in affected municipalities after the reconstructed and tents in the center of PaP earthquake (26.4 percent in 2010, an increase (see Spotlight 2). of 10 percentage points compared to 2008). 3 See Matera, M., Ishizawa, O.A., Van del Borght, R., Nsimba, E., Simon, I., Dorsaint, W., and Surin, R. (2016) for further infor- mation on the fiscal impact of natural hazards in Haiti. 4 There is no unanimity about the death toll of the earthquake. The Haitian government reports a death toll between 200,000 and 316,000, while other sources (see, for instance, Kolbe, Athena R. et al. [2010] and Daniell, J. E., B. Khazai, and F. Wenzel [2013]) suggest a number of victims anywhere between 100,000 and 160,000 victims. 5 See, for instance, UNFPA-Haiti (2010). 6 Figures are reported by Saint-Macary and Zanuso (2016). 7 Estimations are retrieved from the International Organization for Migration’s 2013 report. 8 See the seminal work by Putnam, Leonardi, and Nanetti (1994). 9 The impact of the earthquake on the social capital in those municipalities affected by the 2010 earthquake was measured through three waves of surveys conducted before and after the disaster within the Latin American Public Opinion Project, Latin American Barometer, and USAID. For more detailed information, see Zéphyr, Córdova, Salgado, and Seligson (2011). 26 MOST OF HAITI’S TERRITORY IS EXPOSED TO MULTIPLE HAZARDS Figure 2. Figure O.1. Source: Atlas des menaces naturelles en Haïti (2015). Urban growth: A story of rural push and high In the mid-1980s, 80 percent of Haiti’s popula- fertility rates tion remained in rural areas, but the number of In 1950, less than 10 percent of Haiti’s total residents in cities and towns was increasing at population of just over 3 million lived in urban four times the rate of the rural population since areas. In the next twenty years, the number of the 1950s. By 1982, over a million Haitians urban dwellers grew on average 5 percent each were living in urban areas. Most urban popula- year, and eventually doubled in number at the tion growth took place in Port-au-Prince, with start of the 1970s, reaching just over 700,000.10 as many as 70 percent of urban residents living 10 Figures are extracted from the 1971 and 1982 census and can be retrieved from IPUMS. For 1950 demographic data, see World Bank (1985). 20 See Tobin (2013) and Box 1 for more details. 21 Data on country-level fertility rates were retrieved from the World Development Indicators. For more details on the Haitian fertility rate at the subnational level and in urban areas, see Chahnazarian (1992). 27 BOX 1 – POLITICAL AND ECONOMIC DRIVERS OF HAITI’S URBANIZATION: A HISTORICAL ACCOUNT The first seeds of the cultural dichotomy between urban and rural people (in Creole, moun lavil, literally “downtown people,” and moun deyo, literally “people from the outside”) were sown at the dawn of Haiti’s independence from the French Republic in 1804. As the country’s agricultural sector transitioned from relying on a large-scale plantation economy to a small-hold and leasehold production, farmers improved their living standards and began to aspire to an urban life and political representation. Elites in cities curbed their ambitions by blocking rural populations’ access to urban areas, controlling people’s movement on roads, and limiting investment in any kind of connecting infrastructure. The 1915-1934 US occupation reversed this trend. During this period, Haiti’s transport infrastructure was modernized, the ancient regime institutions were abolished, and Haiti underwent an initial urbanization push. Port-au-Prince increasingly assumed the leading role as the country’s capital. According to the first census conducted in 1950, more than half of the residents in Port-au-Prince were born elsewhere, and their number grew by approximately 2,000 people per year. By 1982, 70 percent of urban residents lived in the country’s capital. During the second half of the twentieth century, impoverished farmers regularly chopped down trees to obtain their cooking fuel (charcoal). Deforestation and soil erosion followed, causing the loss of fertile topsoil, and worsening the productivity of rural lands. Further, erosion increased vulnerability to landslides and floods, which nowadays often disrupt the country’s economy. The massive migration from rural to urban areas–especially to Port-au-Prince–in the 1960s and throughout the 1980s has institutional and economic roots, as described above. The reluctance of cities to welcome rural migrants (a hesitancy also observed in the history of other Latin American countries)11 impacted the potential leverage of increasing densities to boost urban productivity. As more Haitians flocked into PaP, the provision of water, sanitation, electricity, and other basic services did not expand accordingly, as further discussed in this and the following chapters. In 1998, two-thirds of the population of Port-au-Prince was estimated to be concentrated in teeming slum districts.12 Sources: Tobin (2013), Yarrington (2015). 11 Feler & Henderson (2008) discuss the case of Brazil, where during the dictatorship, undesired migrant households were prevented from accessing basic services, eventually shaping the geography of Brazilian cities. 12 See World Bank (1998) for more details about the living conditions in slums in Haiti back in the Nineties. 28 in the capital in 1982. The early primacy of 2000 and 2015. These statistics may, however, PaP dates back to the early second half of the underrepresent the size of the urban popula- twentieth century when factors such as faulty tion in Haiti. As outlined in Spotlight 1: agricultural policies, over-exploitation of land, “What is Urban and What is at Risk?” there and a general cultural and political bias toward are a number of difficulties in estimating the cities deteriorated Haiti’s rural economy.13 population from census data in Haiti. With over 80 percent of government revenue Today, as much as 64 percent of the popula- drawn from direct taxation of farmers, while tion of Haiti may be urban, making it the policies favored urban commercial develop- fourth most urbanized country in the LAC ment and assembly plants especially around region. In this report, we employ an innovative Port-au-Prince. Haitians flocked to the capital methodology for measurement that draws on in search of better economic opportunities and gridded population estimates and population higher living standards. density thresholds (see Spotlight 1). The results Estimates suggest that by the mid-1990s, are comparable with other countries across the 30 percent of Haitians lived in cities. LAC region, irrespective of differences between Approximately 1.4 million people lived in countries in the way they define ‘urban’ in the capital, and 100,000 people lived in the their national census.16 Using this approach, second largest city, Cap-Haïtien (CAP). we estimate the urban population of Haiti Existing literature and household surveys to be around 6,179,000 people. According to trace the latest urbanization trends back to these estimates, it is one of the most urbanized demographic dynamics. For instance, in the countries in the region, following only Puerto 1990s, Haiti’s fertility rate was twice as high Rico, Trinidad and Tobago, and Mexico. as the average in the LAC region (5 births per Regardless of the definition of “urban” used, women), and in Port-au-Prince, women gave urbanization trends are set to increase in the birth to four children on average.14 coming years. According to official statistics, The combination of these demographic Haiti’s urban population grew at a rate of 3.6 changes pushed Haiti well over the 50 percent percentage points faster than the average urbanization mark at the beginning of the Caribbean country and nearly doubled its 21st century. According to national statistics, urban population, from just over 3 million 52 percent of the population lived in urban in 2000 to nearly 6 million in 2015. Ten years areas in 2015.15 The United Nations World from now, the urban population is expected to Urbanization Prospects (UN WUP) indicate a increase by almost another 2 million people higher number – 57 percent – and an average and could surge to around 11 million by 2050, annual urbanization rate of 5 percent between for a 76 percent urban rate. 13 See Tobin (2013) and Box 1 for more details. 14 Data on country-level fertility rates were retrieved from the World Development Indicators. For more details on the Haitian fertility rate at the subnational level and in urban areas, see Chahnazarian (1992). 15 Institut Haitien de Statistique et d’Informatique (IHSI) 2015 population estimates. 16 The challenges arising from using these conflicting definitions for comparison between countries or for global aggregation of data are not unique to Haiti and have been widely discussed in previous work (Satterthwaite et al. 2007; World Bank, 2009; Dijkstra and Poelman, 2014). 29 URBAN SIZE, SHARE OF OVERALL URBAN POPULATION, AND SHARE Figure 3. OF URBAN POPULATION IN HIGH DENSITY CLUSTERS FOR SELECTED ARRONDISSEMENT Sources: Own computations based on Landscan gridded population data, and Deuskar, Stewart, and Lozano-Gracia (2016). An evolving urban landscape: Changes in followed by Nord (66 percent) and Artibo- population and built-up areas nite (57 percent). Haitian départements are Haiti’s territory is divided into ten regions divided into 42 provinces (arrondissements). (départements), the largest being Ouest and The capital’s arrondissement comprises eight Artibonite, located in the west and center- municipalities and is home to 2.7 million north of the country, respectively. The Ouest residents – comprising 45 percent of the département, where the capital Port-au-Prince country’s urban population and 29 percent of is located, is the most urbanized (87 percent), the total population (Figure 3).17 17 The metropolitan area encompasses the municipalities of Carrefour, Cité-du-Soleil, Delmas, Gressier, Kenscoff, Petionville, Port-au-Prince, and Tabarre. 30 EVOLUTION OF URBAN POPULATION DISTRIBUTION BY CLASS Figure 4. OF CITIES Figure O.1. Sources: Own calculations based on Landscan gridded population data and night light data for 1996, 2000, and 2015. Only agglomerations in which light emission was registered in at least one of these three years were included. Haiti has a diversified portfolio of cities. cities between 100,000 and 300,000 inhabi- For this report, we further extract informa- tants. This is represented visually in Figure tion about urban population and the extent 3. Details of the methodology used for this of cities using nighttime lights. The results analysis and a full list of classification of indicate that 40 percent of the population in cities can be found in Spotlight 1. cities is located in the metro area of Port-au- The composition of city categories has Prince. The second and third largest areas shifted significantly since 2000. Cities have are Cap-Haïtien in the Nord département, moved significantly between city categories and Léogane in the Ouest département . based on important changes in popula- Cap-Haïtien has 320,000 inhabitants, tion size between 2000 and 2015. At the representing 6 percent of the urban popula- turn of the century, only Cap-Haïtien had tion and 5 percent of the overall population. more than 200,000 inhabitants, but was Léogane has an urban population of 317,000 joined by Gonaïves fifteen years later. The inhabitants. Gonaïves, Saint-Marc, and number of cities with more than 100,000 Dessalines follow with 261,000, 240,000, inhabitants increased as well from three to and 214,000 inhabitants, respectively. Over seven, including the cities of Ouanaminthe 19 percent of the population is located in and Port-de-Paix. Several new “medium- 31 Figure 5. GROWTH RATE OF BUILT-UP LAND AND MARGINAL INCREASE, BY ARRONDISSEMENT (1975-2011) Source: Deuskar, Stewart, and Lozano-Gracia (2016), based on WorldPop size” cities appeared.18 The weight of each people and more than 1,500 people per class of cities in the urban system has also sq. km.19 In Port-au-Prince, density levels changed. Figure 4 represents the evolution reach as high as 32,500 people per sq. km, of the share of urban population by category between 1 and 2 kilometers away from the of city, from 1990 to 2015. Marché en Fer (Iron Market). This is much The majority of urban residents in higher density than the center of African all cities live in high-density neighbor- cities with similar levels of per capita hoods. All cities in the portfolio have over income, such as Nairobi which has 21,700 60 percent of their population living in inhabitants per sq. km in the center of the high-density neighborhoods, defined as city (see Spotlight 2 for further discussion within-city areas with more than 50,000 of land use within the city). 18 Since no new smaller cities were added, the number of small cities slightly decreased from twelve in 2000 to nine in 2015. 19 See Box 2 for more details about the clustering methodology and the definition of differently dense urban clusters. 32 Port-au-Prince dominates the urban Just as population concentrates in and system. The administrative, economic, and around Port-au-Prince, so does Haiti’s political primacy inherited from the Duvalier economic activity. The metropolitan area regime makes Port-au-Prince the major pole of has the highest economic density in the attraction for job and business opportunities. country, with an estimated GDP of USD 5 As more and more Haitians migrated to the billion – 41 percent of the country’s overall capital, many settled on the outskirts of the output in 2006. Delmas and Croix-des-Bou- metropolitan area. We estimate that in 2015, quets are the next largest contributors to 111,000 Haitians live in highly dense neigh- the country’s GDP. Other large cities like borhoods from the neighboring arrondisse- Cap-Haïtien, Gonaïves, Port-de-Paix, and ment of Croix-des-Bouquets. In fact, certain Saint-Marc together contribute an estimated areas of Croix-des-Bouquets form, along 11 percent to GDP (USD 1.3 billion). Les with the arrondissement of Port-au-Prince, Cayes (large city) and Jacmel (medium-size the “Greater Port-au-Prince” (“Greater PaP”). city), in the south, also show some concen- This urban conglomerate (a de facto city), tration of economic activity, underscoring although not framed within official adminis- the potential role of connections between trative boundaries, provides jobs and services remote rural areas and larger cities in the 20 to almost 3 million urban residents. Haitian urban system. Both population and It is crucial to acknowledge the real extent economic density, however, contributed of the metropolitan area of Port-au-Prince little to boosting Haiti’s economic growth. when planning for services and infrastruc- The next sections will highlight some of ture provision for all the workers that might the socio-economic obstacles that prevent potentially commute every day to the capital. the country from leveraging its cities’ high From 1975 to the early 1990s, the “Greater densities and benefiting from urbanization. PaP” expanded faster than the national In addition to Port-au-Prince, our estima- average (6 percent annually); in 1990, it tions suggest that there are five other large covered over 100 sq. km, more than double urban clusters in Haiti. The arrondissements the surface from 15 years before. In line with of Cap-Haïtien (Nord) and Léogane (Ouest) the national trend, growth of built-up land are the next largest in terms of urban popula- in the metropolitan area slowed down in tion. Cap-Haïtien has 320,000 inhabitants, the following twenty years, but grew higher representing 6 percent of the urban popula- in magnitude (8 sq.km on average annually) tion and 5 percent of the overall popula- than any other urban agglomeration in Haiti. tion, with as many as 74 percent living in Figure 5 shows the growth rate of built-up high-density neighborhoods. areas for selected arrondissements in each of the départements, together with the value of their marginal increase.21 20 At a more aggregate-level, the Ouest département - where the metropolitan area is geographically located - hosts 55 percent of the total urban population and 40 percent of the overall population. 21 For details on the methodology, see World Bank’s background note for the Haiti Urbanization Review (2016). 33 URBANIZING UNDER AN UNFAVORABLE of the construction industry after 2010, and LABOR MARKET AND HIGH LEVELS OF 9 percent of urban residents in 2012 were POVERTY employed in the construction business. Haiti’s economy has been recovering Industrial activities in urban areas play a slowly since the 2010 earthquake, when it secondary role in the economy, but have great shrank by 5.5 percent. In 2015, annual GDP potential for future growth. Manufacturing growth was estimated to be 1.2 percent, accounts only for 8 percent of the country’s higher than the LAC average, but more than 3 GDP (estimations from the Economist Intel- percentage points lower than the average Low ligence Unit 2016). However, export-oriented Income Country (Singh and Barton-Dock garment businesses can leverage on recent 2015). Inflation has been reduced to single investments in the Port-au-Prince and digits and external debt shrank following Cap-Haïtien seaports, which now handle 90 debt cancellation. But the diminishing percent of Haiti’s international trade.24 The inflow of post-earthquake foreign aid, as well SONAPI industrial park, located within 5 km as political instability at the beginning of from Port-au-Prince’s main port, is one of 2016, led to the depreciation of the country’s the largest garment clusters in the country. currency.22 Most recent economic growth, Other smaller garment factories are on the between 2005 and 2015, has been mainly outskirts of smaller cities like Ouanaminthe driven by the rise in sectors leveraging urban- (Grupo M, 6,500 workers) and Cap-Haïtien ization forces, namely industry and services. (S&H Global, 2,500 workers). Nevertheless, According to survey data, the urban employment in the manufacturing industry economy is dominated by the tertiary is still relatively low, accounting for 4 sector: 38 percent of the 1.4 million workers percent of workers in PaP and 3 percent in in cities have a job in commercial activi- smaller cities.25 23 ties. Consumer services in Port-au-Prince Outside the capital, agriculture-related account for one-sixth of the capital Gross jobs are still widespread. The agricultural Value Added. Beyond trading, 40 percent of sector, highly vulnerable to shocks and urban workers – especially in Port-au-Prince scarcely contributing to the country Gross – are employed in other service sector activ- Value Added, still employs most workers (45 ities, such as transport, education, finance, percent) nationwide. The primary sector and tourism. Post-earthquake home recon- is the largest in Grand’Anse (66 percent of struction and some advances in government the labor force), Centre (62 percent), and infrastructure projects sustained the growth Sud-Est (59 percent). Agricultural employ- 22 Data and estimations from The Economist Intelligence Unit, Country Report (December 2016). 23 Data is obtained from the 2012 “Enquête sur les Conditions de Vie des Ménages Après Seisme” – ECVMAS survey. These figures match to some extent the 2016 estimations by the Economist Intelligence Unit and the 2012 approximations by Oxford Economics. 24 Investments in port infrastructures have pushed Haiti among the 10 (out of 32) best Latin American countries and 2nd (out of 29) best among the poorest countries in “trading across borders.” See the 2015 World Bank report “Country Partnership Framework for the Republic of Haiti for the Period FY16-FY19.” 25 On the low employment rates in the apparel industry, see also International Monetary Fund (IMF), 2015; Hornbeck, 2011. 34 ment, though, is not confined to rural areas. on rural employment can be traced to a few It involves 15 percent of the labor force from specific phenomena related to age groups and smaller urban agglomerations and rural gender. First, the employment increase was towns. These urban clusters could therefore very pronounced among the younger cohort be functioning as a gateway to mediate the of rural individuals (15-25 years old). Second, flow of labor and products between rural female employment rates in rural areas hinterlands and urban centers.26 increased across all age groups. However, the expansion in employment rates among young Unemployment, underemployment, and men and women, and for women of all ages, little productivity is almost entirely explained by an increase According to recent estimates, 75 percent in individuals reporting being unpaid family of the working age population in Haiti is workers working a few hours a week – reducing part of the labor force.27 Almost 4 out of five the real increase in rural employment (Scot Haitians are either employed or willing to and Rodella, 2016). start a job on short notice if offered one. Despite an overall decline, unemployment National unemployment rates have declined remains an urban issue. Survey data from 2012 steadily since 2001, from 27 percent to 12 indicates that 40 percent of the urban labor percent in 2012. At present, unemployment force does not have a job. The probability of in Haiti is much higher than LAC’s average being employed in cities was between 4 and of 6 percent, but comparable to other nations 5 p.p. lower than it had been in 2001, while in the Caribbean region.28 On the other hand, the same likelihood in rural areas was almost the number of underemployed people has 20 p.p. greater. The urban labor market can increased since 2007.29 In 2012, eight out of nonetheless rely on a younger labor force than ten workers in the country earned less than the rural one: in 2012, 57 percent of residents the national minimum wage (250 gourdes in cities and towns were between the ages of 15 per day – approximately, USD 6 at the 2012 and 49 years. Across urban areas, workers in exchange rate). smaller cities are 3 percent more likely to find Rural areas, more than cities, have contrib- an occupation than those living in Port-au- uted to the fall of unemployment. In 2012, Prince, conditional on their characteristics. the probability of being employed in rural Underemployment and informality are areas was almost 20 percentage points (p.p.) two other characteristics of the urban labor greater than in 2001. The increasing figures market. The most recent statistics suggest 26 Working in non-farm activities is correlated with a significantly lower probability of being poor, highlighting the importance of fostering non-primary jobs even in traditionally rural areas (Scot and Rodella, 2016). 27 The labor force is the supply of workforce available for producing goods and services in an economy. It includes people who are currently employed and people who are unemployed but seeking work, as well as first-time job-seekers. Not everyone who works is included, however. Unpaid workers, family workers, and students are often omitted, and some countries do not consider members of the armed forces. Figures are elaborated by the International Labour Organization (ILO). 28 In particular, the Bahamas and Belize (14 percent each) and Barbados (12 percent). See Scot and Rodella (2016) for further details. 29 Underemployment is defined as the share of workers receiving less than the minimum wage. 35 BOX 2 – A SNAPSHOT OF HAITI’S ECONOMIC SECTOR COMPOSITION The service sector contributes 57 percent to the Gross Value Added (GVA). From 2005 to 2015, it grew by 2 percent annually. Construction services surged given an increasing demand for homebuilding and earth- quake and hurricane reconstruction. Consumer services account for around 30 percent of the national GDP. Other important tertiary subsectors are: wholesale trade, transport and communication services, public services, and financial and business services, in order of relevance.30 Between 2000 and 2015, the secondary sector increased by 4 percent annually, with manufacturing growing at an average annual rate of 2.6 percent. Yet industry accounts for only 21 percent of the GVA and has progressively shrunk over the years. In 1990, it represented 18 percent of Haiti’s GDP, but decreased following a United States embargo that lasted from 1991 to 1994 (Singh and Barton-Dock 2015).31 However, the increasing request for apparel assembly, primarily from the US, boosted manufacturing and industry, especially in urban areas. Textiles represent 88 percent of the country exports (20 percent of GDP) and are expected to expand further into markets in North America, thanks to the HOPE II Act.32 Agricultural production accounts for 22 percent of the GVA, but has been stagnant for the past fifty years and declining after 2010.33 The 1991-94 embargo hampered access to important agricultural inputs and contrib- uted to the decline of productivity in the sector. High fragmentation of land, legally insecure land rights, credit shortage, low levels of technology, and soil deterioration have also constrained agricultural productivity (Singh and Barton-Dock, 2015; WTO, 2015). Food production is constantly put at risk by natural disasters: in the past years, droughts have affected around 50 percent of crops (USAID, 2011). Deficiencies of the primary sector together with an increasingly urbanized population force Haiti to import 55 percent of the country’s food needs. This number explains Haiti’s trade imbalance: the country imported 50.6 percent of its GDP in 2016. Sources: Tobin (2013), Yarrington (2015). that, on the demand side of the market, the not registered or do not keep formal accounts share of non-agricultural workers without – decreased by 8-10 percent between 2006 written contracts and social protection and 2012. However, this trend reflects the amounts to 80 percent in the capital and increase in the share of workers employed 90 percent in other cities (See Box 3 for an in the public sector and nongovernmental example of women in the informal economy). organizations (NGOs), rather than a decrease On the supply side, the probability of workers in the share of informal employees. In fact, being employed in the urban private informal in 2012, the labor force employed in small non-farm sector – i.e., unincorporated enter- informal enterprises was 50 percent in PaP prises, mostly household business, that are and 65 percent in other cities, respectively.34 30 Ibid. 31 World Development Indicators and the Economist Intelligence Unit forecasts. 32 Other products exported are metals (3.1 percent of overall export, mostly iron and copper) and essential oils (2.2 percent). Data extracted from UN Comtrade and referred to 2014. 33 Ibid. 34 Figures are based on the authors’ analysis of the 2012 ECVMAS survey. See Scot and Rodella (2016) for further details. 36 As Haiti becomes more urbanized, jobs indeed slowed down and has led the Haitian are created in low-productivity non-tradable Central Bank to cut its 2016 GDP growth services (like trade and construction) rather forecast.37 More decisive policies should be than in higher productivity tradable activi- adopted to shift the economic system toward ties (like manufacturing and services). The more environmentally and socially sustain- precarious growth of Haitian cities seems to able manufacturing and services. Otherwise, be driven by domestic consumption, rather future shortages of the exceptional resources than production. Although hard evidence of income that Haiti relies on, might plunge (and data) is not yet available, consumption the country and its urban areas even further in turn might be driven by remittances and into poverty. foreign aid. In fact, survey data suggests that, as of 2012, over 35 percent of urban house- Vulnerability and chronic poverty remain a holds and 20 percent of rural households challenge, but poverty in cities has declined receive remittances. Monetary transfers are Extreme poverty has declined in Haiti in often large in value and their contribution the first decade of the 2000s, especially in to total income (an average of 24.5 percent urban areas. At the national level, extreme in the country) is larger than that of in-kind poverty decreased from 31 percent in 2000 gifts (12.2 percent). Foreign aid contributed to 24 percent in 2012. Most of its reduction to increasing labor demand for post-2010 took place in the metropolitan area (from earthquake reconstruction and to the 20 to 5 percent) and in smaller cities (from growth in the share of construction services 21 to 12 percent). Despite this progress, to local GDP – especially in Port-au-Prince. Haiti remains the poorest country in the Based on evidence from other countries, LAC region and one of the poorest in the such as Nigeria and Angola, consump- world. Today, around 59 percent of Haitians tion-driven urbanization curbs cities’ live in poverty. Nearly 6.3 million Haitians 35 productivity even further. Moreover, it cannot meet their basic needs and 2.5 ties them to the volatility and precarious- million cannot cover their food needs. 40 ness of external sources of income, like Around 695,000 poor people live in Port-au- remittances and foreign aid. While future Prince (16 percent in extreme poverty) remittance inflows are difficult to forecast, and 1.4 million reside in smaller cities (24 aid flow has been in decline and is expected percent in extreme poverty) (Figure 6). to continue to fall, negatively affecting one Recent work using data from two of the country’s main economic activities Demographic Health Surveys (DHS) for (i.e. construction).36 In the past years, the 2006 and 2012 further suggests that living growth of the construction industry has standards improved the most in urban 35 Gollin, Jedwab, and Vollrath (2016) and Jedwab (2013) discuss the rise and issues of consumption in cities. 36 In the aftermath of the 2010 earthquake, international donor assistance represented nearly 17 percent of GDP, but fell to 7 percent in 2014 (World Bank 2016). 37 Based on the Economist Intelligence Unit’s 2016 forecasts. 40 See Singh and Barton-Dock (2014). 37 BOX 3 – “MADAN-SARA”: A WELL-ORGANIZED CHAIN OF INFORMAL WORKERS Key protagonists of Haitian informal trade are the “madan-sara.”41 “Madan-sara” – named after a local migratory bird – buy in bulk from producers or intermediaries in rural areas to sell wholesale in urban centers. They are therefore a crucial intermediary connecting producers in remote areas with large markets in cities. In September 2015, The New Yorker covered the stories of Haitian business-women periodically jumping on crowded boats for a six-hour overnight journey between Marigot – in the South – and Anse-à-Pitres, along the border with the Dominican Republic. Once or even twice a week, a “madan-sara” reportedly pays around USD 5 (250 Gourdes) to travel to Anse-à- Pitres. She then crosses the border to the Dominican Republic and trades goods, e.g. clothes for other primary goods. She then pays around USD 5 per carrier bringing the merchandise back to the Haitian side of the border, and USD 30 per bag of freight loaded onto the boat back to Marigot. Back home, she would also hire Tap-Tap drivers to carry the goods into storage. For each successful trip, each of these women can make about 10,000 HTG – nearly USD 200. Source: Jelly-Schapiro, 2015. areas.42 While the proportion of rural house- Despite an overall reduction, chronic holds that were poor or very poor remained poverty remains widespread in Haiti.44 unchanged between this six-year period, Using panel data with information on about 80 percent of the urban households consumption levels registered in urban reported to be somewhat better off. The data areas in 2012 and 2013, and ad hoc income suggests that households in the Port-au- thresholds, Figure 7 shows: (a) the share of Prince metropolitan area, for instance, had the poor population; (b) the proportion of better access to consumption goods and people who, although not poor, are vulner- 43 services and to improved housing. able and likely to fall back into poverty; and 41 To our knowledge, there are no official statistics in this sense. However, local newspapers and magazines – like Le Nouveliste and Woy magazine – often report “madan-saras” to be usual victims of road accidents. 42 The Demographic Health Surveys (DHS) and households’ surveys used above also ask questions about households’ poverty, although measured according to two slightly different methodologies. The DHS conducted by USAID do not collect infor- mation about income and capacity of consumption of households, but rather on the characteristics of their houses and their access to consumption goods and services. The resulting wealth index is then used to classify households along five quintiles of wealth: poorest, poorer, middle, richer, and richest. The wealth index built by the DHS, however, do not capture the main determinant of poverty in developing countries: the spending capacity of families. Moreover, it does not track the same households across waves; through the DHS it is therefore difficult to understand how many households have been stuck in a poverty trap and the proportion of those that emerged from poverty or reverted into it. 43 See Singh and Barton-Dock (2014) for a more extensive analysis. 44 Chronic poor are those households identified poor both in 2012 and 2013; this means that their consumption was always below the poverty line. The transitory poor are those households who spend only one period (either 2012 or 2013) in poverty. See Perge and Scot (2016) and the literature mentioned therein for a more detailed discussion about these definitions. 38 Figures 6 & 7. FIGURE 6. POVERTY INCIDENCE BY FIGURE 7. PROPORTION OF PLACE OF RESIDENCE (PROPORTION HOUSEHOLDS BY POVERTY STATUS IN AND ABSOLUTE NUMBERS) 2012 AND 2013 Source: Own calculations based on ECVMAS 2012 and Source: Perge and Scot (2016) based on ECVMAS 2012 and thresholds computed by Perge and Scot (2016). ECVMAS 2013. (c) the share of non-vulnerable individuals.45 is in line with the trend prevailing in the In 2013, the proportion of poor residents LAC region.46 At the same time, 40 percent decreased from 40 to 34 percent with respect of individuals in urban areas are vulnerable to the previous year. However, 22 percent of to poverty, higher than what was observed residents that were poor in 2012 remained at the national level (28 percent).47 This so the following year; this figure is half than fact highlights the fragile conditions that what was observed at the national level and still affect those that have to some extent 45 See Perge and Scot (2016) for more technical details. 46 As in Vakis, Rigolini, and Lucchetti (2015). Notice, however, that the poverty line used in that study is the USD 4 a day (2005 PPP), significantly higher than Haiti’s median poverty line. 47 Between 2012 and 2013, 14 percent of formerly poor persons increased their wealth and became less – but still – exposed to poverty; 8 percent of non-vulnerable residents in 2012 became so in 2013. 39 overcome poverty; in particular, transi- URBANIZING UNDER CROWDING CONDI- tory poor in urban areas face shocks linked TIONS AND WITH NO SUPPORTING INFRA- to urban labor markets, such as company STRUCTURE failure or loss of income from non-agricul- Poverty is not only defined in monetary tural services. terms; lack of infrastructure can represent Income inequality is stagnating and a major obstacle for urban households to remains the highest in the region, although escape poverty traps.49 This is particularly improvements have been registered in cities. true in areas that combine the highest popula- The Gini coefficient remained constant at tion density and where access to clean water, about 0.6 between 2001 and 2012; since then, improved sanitation and electricity is limited inequalities have persisted. Haiti remains or lacking. In such cases, density leads to the most unequal country in LAC and one of overcrowding, unhealthy living conditions, the most unequal in the world. In 2012, the and marginalization, rather than becoming top ten percent of the population controlled an asset for growth. As further discussed in almost 50 percent of the country’s resources. the next chapters, the infrastructure gap in At the subnational level, between 2002 cities is large in terms of basic services and and 2012, inequality decreased the most in other infrastructure such as roads. urban areas (from 0.64 to 0.59), whereas it increased in rural areas (from 0.49 to 0.56). Access to improved water and sanitation The widening of inequality in rural areas only reaches a few can be explained by repeated weather-re- More than one third (35 percent) of urban lated shocks that undermined agricultural residents do not have access to an improved production, and by inflation that fueled water source (WDI 2015), and trends show higher prices of food; both phenomena that, due to recent urban growth, rates are affect the rural poorest disproportionately.48 declining.50 The 2012 DHS survey shows that the proportion of families with water piped into their dwelling or with access to a public tap decreased from 24 to 3 percent and from 65 to 21 percent, respectively between 2000 and 2012. Similarly, in smaller cities and towns, the proportion of households with piped water into their dwellings decreased from 20 to 2 percent over that period. In 48 See Singh and Barton-Dock (2014) for further details on poverty diagnostics in Haiti. 49 Evidence on lagging access to infrastructure and basic services is drawn from own computations based on 2000, 2006, and 2012 Demographic Health Surveys. 50 Overall, only 58 percent of the Haitian population had access to drinkable water from an improved source. This figure places Haiti 25 percentage points below the second worst-performing country in the LAC region (the Dominican Republic) and among the 10 worst-performing low-income countries (slightly better than Eritrea, Niger, and Tanzania). Improved sources of water include piped water into dwelling, to yard, or to the neighbor; public tap water; tube well or borehole; protected dug well; protected spring; rainwater; bottled water or sold by company. 40 addition to this, there is important variation at night (Tilmans et al. 2015). Box 4 discusses in access rates across and even within urban informal workers, referred to as bayakou, who areas. While data to assess access to services help deal with the lack of adequate sanitation within individual cities is hard to find, a infrastructure. recent profile of Cap-Haïtien suggests that access is geographically uneven, with only 20 Access to electricity is unequal across urban percent of the country’s communes having areas satisfactory levels of running water service With regards to electricity, only 38 percent (UN Habitat Cap- Haïtien Profile 2012). of Haitian households were estimated to have Smaller cities also have lower overall water access in 2012. No other country performed access rates – 55 percent in Milot, for example, as poorly as Haiti in the LAC region, where and rural towns tend to have low household on average 96 percent of the population is connection levels to the public network – as covered. However, when compared to low-in- low as 5 percent (Habitat 2012 Milot Urban come peers elsewhere, Haiti is among the top profile; Brault, Sanz, and Le Bansais 2014). ten in terms of access. Its access rate is similar The challenges are further exacerbated by to Gambia (35 percent of the population individual household constraints to ensure covered), Eritrea (36 percent) and Benin (38 that their water is safe; in smaller cities 45 percent). Within Haiti, electricity coverage percent of families lack the tools to boil water is unequal: in 2012, half of the population for cooking or washing hands (DHS 2012). living in small cities and towns had access Current sanitation systems are inadequate to electricity, in contrast to nearly 90 percent to serve the urban population. Two-thirds (66 of households in the capital. Coverage in percent) of urban residents lack improved “secondary” urban areas, nonetheless, slowly 51 sanitation (WDI 2015). In urban areas, caught up in the past decade, while it had shared facilities are common. The 2012 been gradually decreasing in Port-au-Prince. DHS indicates that 48 percent of residents Differences between urban and rural areas in Port-au-Prince and 41 percent of house- are more striking, with access in rural areas holds in second-tier urban agglomerations being as low as 15 percent. use pit latrines with slab.52 At least 8 percent Despite coverage in urban areas being of urban residents practice open defeca- above 50 percent, Haiti has one of the lowest tion (WDI 2015), and research suggests that electricity consumption rates worldwide. this figure may be higher given that urban At the beginning of 2010, despite heavily residents that rely on public toilets may resort subsidized residential tariffs, the per capita to open defecation to meet sanitation needs electricity consumption was equal to 51 kWh, 51 Improved sanitation facilities include non-shared toilets flushing to piped sewer system, to pit latrine, and to septic tank, pit latrines improved by ventilation or with slab, and composting toilet. The overall access rates in Haiti are 50 p.p. lower than the average country in the LAC region and 5 p.p. lower than the average low-income country. 52 Pit latrine with slab is a dry pit latrine whereby the pit is fully covered by a slab or platform that is fitted either with a squatting hole or seat. The platform should be solid and can be made of any type of material (concrete, logs with earth or mud, cement, etc.) as long as it adequately covers the pit without exposing the pit contents other than through the squatting hole or seat. 41 Figure 8. ACCESS TO WATER AND SANITATION (A) Access to sources of drinking-water, by place (B) Access to sanitation services of residence Source: DHS 2000, 2006, 2012 forty times lower than the LAC average.53 megawatts (MW), outdated and deteriorated Under-consumption is notably driven by the equipment constrain the available capacity to shortages in the supply, monopolized by the only 244 MW. public provider, Electricité d’Haiti (EDH). Low access to electricity cripples local urban For those who have a connection, electricity is productivity. In large urban areas, obtaining available only at limited times during the day: an electric connection for businesses is very about 15 hours in Port-au-Prince, and between costly. Since 2009, tariffs for industrial and 5 to 9 hours elsewhere, on average. While the commercial consumers (especially the indus- installed capacity ranges between 250 and 400 trial and transportation sectors) cost USD 0.36, 53 Lucky et al. (2014) discuss the electricity market in Haiti in further details. Concerning the residential tariffs, Haiti has the third lowest ones (USD 0.16 per kWh, on average) compared to other Caribbean country pairs: Trinidad (USD 0.5), Suriname, Dominican Republic, Nevis, Belize, St. Lucia, Bahamas, Guyana, Barbados, St. Kitts, St. Vincent, Grenada, Dominica, Jamaica, Antigua (USD 0.37). Data come from the World Development Indicators. 42 BOX 4 – BAYAKOU: A SCORNED YET VITAL JOB WHEN SANITATION INFRASTRUC- TURE SANITATION IS LACKING Among Haitian households without access to working sewers, only a few can afford to dig cesspools in their backyards for collection and disposal of human waste. Once every year or so, these households hire informal laborers known as bayakou to come in the middle of the night to empty these cesspools. Bayakou, then, play a key role in the functioning of cities in the absence of sewerage infrastructure. Keeping human waste out of water bodies and crops is crucial to avoid the spreading of infectious diseases; when lacking proper sanitation services, bayakou are the only way to keep households safe. To hire a bayakou, a homeowner must first negotiate with a middleman, who is in charge of arranging the assignment. During the following nights, a team of two or three bayakou enters the homeowner’s backyard and pours a bottle of floor cleaner into the cesspool in order to soften the excrement and cut the smell. Once the floor-cleaning solvent has soaked, one member of the team climbs into the latrine and scoops out the human waste with a bucket. Usually, workers carrying out this activity are not wearing any clothes, making them exposed to possible cuts and injuries and thereby infections. There are no designated safe areas to dispose of the human waste, so bayakou usually dump it on the ground, in ravines, or in rivers. This only transfers the risk and exposure to disease from one area to another. According to some estimates, a bayakou team can empty a fifty-cubic-metre cesspool in two to three nights. In spite of its reputation and danger entailed, the job can be financially attractive to many: through the inter- mediary, a bayakou can make just over 3,000 HTG (about USD53, using the official exchange rate of 1 HTG = USD 0.016 as of August 31, 2017) per latrine – the equivalent of a month’s salary sewing T-shirts in a garment factory. To partially solve for bayakou’s precariousness and possible environmental costs, some aid groups have entered the market and have been hiring them, introducing some minimum safety standards, and overseeing their work directly. The National Drinking Water and Sanitation Directorate (DINEPA) has also been running public health programs directed to training bayakou and testing small pumps. Sources: Curnutte, 2011; Katz, 2014; Wilentz, 2010; and discussions with the Inter-Ministerial Committee for Territorial Devel- opment and DINEPA. on the higher end of the Caribbean distribu- Costly procedures to access to electricity tion. The ineffective provision by EDH in its affects the business environment and job grids, moreover, prompts many factories and creation in the secondary and tertiary sector.55 businesses to generate electricity with their own private diesel and polluting generators, further inflating their production costs and the import of fuels.54 54 See Lucky et al. (2014) and the 2016 Doing Business index for more detailed data. 55 The secondary sector includes garment factories and other manufacturing activities. The tertiary sector is made of all transport, touristic, and financial services, and – most of all – trade activities. 43 REVIEWING THE CHALLENGES AHEAD policymakers identify priorities for invest- Today, as Haitian policymakers begin ments from a jobs perspective. Hence, a to shift their focus from addressing the good understanding of how workers move challenges of reconstruction to planning for in and around the largest cities, how acces- a brighter, more resilient and sustainable sible jobs are, and what are the most critical future, a better understanding of the key road segments to ensure job accessibility is bottlenecks that are driving crowding and not affected in the event of a disaster, can the economic stagnation of cities in Haiti provide valuable information for evidence- is needed. This chapter has highlighted based decision making. This is the focus of the key challenges that the urban system Chapter 3. faces today. First, it examined the difficul- But to address the key infrastructure ties brought by high population density bottlenecks that cities face today, and and limited infrastructure in Haitian cities. prepare to build resilient cities for a brighter Urban areas in Haiti today are crowded and future, strengthening cities’ financing mecha- growing in an uncoordinated manner, with nisms will be essential. Chapter 4 explores insufficient regard to risks. Second, despite the key bottlenecks that municipalities face population density, cities are not generating in financing their needs. Cities must be economies of agglomeration nor becoming equipped with the financial tools and means centers of economic growth. Instead, infor- to ensure they can respond to the increasing mality and poverty prevail. Third, cities demand for infrastructure and services. are growing with limited service provision, Governance and institutional arrangements which coupled with the growing population on the structure of local finances can either densities, are consumed by the negative strengthen or weaken the tools that munici- effects brought by the concentration of palities have at hand to respond to urbaniza- people and prevent them from leveraging tion pressures. This final chapter provides an the potential economic benefits that such overview of municipal finances in Haiti and concentration can bring. charts a road of action that can help build Better planning for resilient cities can local financial sustainability. help address current deficits and prepare for future urban growth. Chapter 2 explores how land management and coordinated decision making can help address the needs of today while looking for ways to unleash the potential of cities in the future. But for economic activity to thrive in Haitian cities, connecting people to jobs, and firms to input and output markets, will be essential. A better-integrated labor market can help match talent and skills with enterprises’ needs. Identifying the key challenges related to within-city connectivity can help 44 REFERENCES Aisen, A., and F.J. Veiga. 2013. How does from Indian districts. The World Bank political instability affect economic Economic Review. growth? European Journal of Political Cayemittes, M., M. Fatuma Busangu, J. de Economy, 29, 151-167. Dieu Bizimana, B. Barrère, B. Sévère, Annez, P., and R.M. Buckley. 2009. Urbaniza- V. Cayemittes, and E. Charles. 2013. tion and growth: setting the context. In Enquête Mortalité, Morbidité et Utilisa- M. Spence, P. Annez, and R. M. Buckley, tion des Services, Haïti, 2012. Calverton, Urbanization and growth (pp. 1-45). Maryland, USA: Ministère de la Santé Washington, D.C.: The World Bank. Publique et de la Population - MSPP/ Azevedo, J. P., G. Inchauste, S. Olivieri, J. Haïti , l’Institut Haïtien de l’Enfance - Saavedra, and H. Winkler. 2013. Is labor IHE, and ICF International. Retrieved income responsible for poverty reduction? from http://dhsprogram.com/publica- A decomposition approach. World Bank tions/publication-FR273-DHS-Final-Re- Policy Research Working Paper 6414. ports.cfm#sthash.bJDFmHEa.dpuf Bengtsson, L., X. Lu, A. Thorson, R. Garfield, Chahnazarian, A. 1992. Hausse récente de and J. Von Schreeb. 2011. Improved la fécondité en Haïti: un nouvel engoue- response to disasters and outbreaks by ment pour la vie en union? Population, tracking population movements with 47(3), 583-616. mobile phone network data: a post-earth- Christianen, L., J. De Weerdt, and Y. quake geospatial study in Haiti. PLoS Todo. 2013. Urbanization and poverty Med, 8(8). reduction: the role of rural diversifica- Berdegué, J. A., J. Escobal, and A. Bebbington. tion and secondary towns. Agricultural 2015. Explaining spatial diversity in Latin Economics, 44(4-5), 435-447. American Rural Development: Struc- Curnutte, M. 2011.  A Promise in Haiti: A tures, institutions, and coalitions. 73, Reporter’s Notes on Families and Daily 129-137. Lives. Vanderbilt University Press. Brault, J-M., Z. Sanz Uriarte, and B.J. Le Daniell, J. E., B. Khazai, and F. Wenzel. Bansais. 2014. Delivering water supply and 2013. Uncovering the 2010 Haiti earth- sanitation services in fragile states: profes- quake death toll.  Natural Hazards and sionalizing drinking water service delivery Earth System Sciences Discussions,  1(3), in small towns of Haiti. Water and sanita- 1913-1942. tion program field note. Washington, D.C.: Deuskar, C., B.P. Stewart, and N. Lozano- World Bank Group. Gracia. 2016. Defining urban areas in Busso, M., and L. Madrigal. 2013. Produc- Haiti. Mimeo. Washington D.C.: World tivity and resource misallocation in Bank. Latin America. The BE Journal of Macro- Dijkstra, L., and H. Poelman. 2014.  A economics, 13(1), 903-932. harmonised definition of cities and rural Cali, M., and C. Menon. 2012. Does urban- areas: the new degree of urbanisation, ization affect rural poverty? Evidence European Commission  (No. 01). Regional 45 Policy Working Papers. bution of economic activity. The Open Dorosh, P., and J. Thurlow. 2014. Can cities Geography Journal, 3(1). or towns drive African development? Gollin, D., R. Jedwab, and D. Vollrath. 2016. Economywide analysis for Ethiopia and Urbanization with and without Indus- Uganda. World Development, 63, 113-123. trialization. Journal of Economic Growth, Dorvilier, F. 2010. Les causes de la crise de la 21(1), 35-70. transition démographique en Haïti: une Green, M., and D. Hulme. 2005. From analyse néo-institutionnelle. Centre de correlates and characteristics to causes: recherches en Démographie et Sociétés, 1-28. Thinking about poverty from a chronic The Economist Intelligence Unit 2016. Haiti poverty perspective. World Development, Country Report. 33(6), 867-870. Ellison, G., E.L. Glaeser, and W.R. Kerr. Henderson, J., A. Storeygard, and D.N. Weil. 2010. What causes industry agglomer- 2012. Measuring economic growth from ation? Evidence from coagglomeration outer space. American Economic Review patterns. The American Economic Review, 102(2), 994-1028. 100(3), 1195-1213. Hornbeck, J. F. 2011. The Haitian Economy Feler, L., and J. Henderson. 2008. Exclu- and the HOPE act. Current Politics and sionary policies in urban development: Economics of South and Central America, How under-servicing of migrant house- 4(3), 565. holds affects the growth and composition Hsieh, C-T., and E. Moretti. 2015. Why do of Brazilian cities. National Bureau of cities matter? Local growth and aggregate Economic Research No. w14136. growth. National Bureau of Economic Flowminder Foundation, Digicel Haiti, and Research Working Paper, No. w21154. World Food Programme. 2016. Haiti: Hulme, D., and D. Shepherd. 2003. Concep- Hurricane Matthew. Estimated Population tualizing chronic poverty. World develop- Movements as of 22 November 2016. Mimeo. ment, 31(3), 403-423. Freire, S., K. MacManus, M. Pesaresi, E. IMF. 2015. Haiti. IMF Country Report No. Doxsey-Whitfield, and J. Mills. 2016. 15/158, Washington D.C. Development of new open and free Institut Haïtien de Statistique et d’Informa- multi-temporal global population grids tique. 2015. Population totale, population at 250 m resolution. Proceedings of 19th de 18 ans et plus, ménages et densités estimés AGILE Conference on Geographic Informa- en 2015. Retrieved from http://www. tion Science . Helsinki. ihsi.ht/pdf/projection/Estimat_PopTo- Gelting, R., K. Bliss, P. Molly, G. Lockhart, tal_18ans_Menag2015.pdf and T. Handzel. 2013. Water, sanitation Jedwab, R. 2013. Urbanization without struc- and hygiene in Haiti: Past, present, and tural transformation: Evidence from future. The American journal of tropical consumption cities in Africa. Working medicine and hygiene, 89(4), 665-670. paper . Washington, D.C.: George Ghosh, T., R.L. Powell, C.D. Elvidge, K.E. Washington University. Baugh, P.C. Sutton, and S. Anderson. Katz, J. (2014, March 12). Haiti’s shadow 2010. Shedding light on the global distri- sanitation system. The New Yorker. 46 Retrieved from http://www.newyorker. Perge, E., G. Fleurantin, C. Kiyomi, M. com/tech/elements/haitis-shadow-sani- Muller, and O. Calvo-Gonzalez. 2016. tation-system Haiti – Poverty and Equity Note #2. World Kolbe, A.R., R.A. Hutson, H. Shannon, E. Bank. Trzcinski, B. Miles, N. Levitz, M. Puccio, Philippe, F., and M. Culot. 2009. Household L. James, J.R. Noel, and R. Muggah. solid waste generation and character- 2010. Mortality, crime and access to istics in Cape Haitian city, Republic basic needs before and after the Haiti of Haiti. Resources, Conservation and earthquake: a random survey of Port-au- Recycling, 54(2), 73-78. Prince households.  Medicine, conflict and Putnam, R. D., R. Leonardi, and R.Y. Nanetti. survival, 26(4), pp.281-297. 1994.  Making democracy work: Civic tradi- Krugman, P. 1991. Increasing returns and tions in modern Italy. Princeton University economic geography. Journal of political Press. economy, 99(3), 483-499. Ravaillon, M., S. Chen, and P. Sangraula. Lall, S. V. 2013. Planning, Connecting, and 2007. New evidence on the urbanization Financing Cities—Now: Priorities for City of global poverty. Population and Develop- Leaders. Washington D.C.: World Bank ment Review, 33(4), 667-701. Publications. Rodrik, D. 2013. Unconditional convergence Lucky, M., K. Auth, A. Ochs, X. Fu-Ber- in manufacturing. The Quarterly Journal taux, M. Weber, M. Konold, and J. Lu. of Economics, 128(1), 165-204. 2014. Haiti Sustainable Energy Roadmap: Saint-Macary, C., and C. Zanuso. 2016. Build Harnessing Domestic Energy Resources back better? Long-lasting impact of the to Build an A ordable, Reliable, and 2010 Earthquake in Haiti. Working paper. Climate-Compatible Electricity System. Satterthwaite, D. 2007.  Adapting to climate Washington, D.C.: Worldwatch Institute. change in urban areas: the possibilities Maynard-Tucker, G. 1971. Unions, Fertility, and constraints in low-and middle-income and the Quest for Survival. Social Science nations (Vol. 1).2nded. & Medicine, 43(9), 1181. Scot, T., and A-S. Rodella. 2016. Sifting Matera, M., O.A. Ishizawa, R. Van del Borght, through the Data: labor markets in Haiti E. Nsimba, I. Simon, W. Dorsaint, and through a turbulent decade (2001-2012). R. Surin. 2016. Diagnostic sur l’Impact Policy Research working paper no. WPS Économique et Budgétaire des Désastres 7562. Washington, D.C.: World Bank en Haïti. World Bank. Group. Moomaw, R., and M.A. Alwosabi. 2004. An Singh, R. J., and M.A. Barton-Dock. 2014. empirical analysis of competing expla- Haiti. Toward a new narrative: systematic nations of urban primacy evidence from country diagnostic. Washington, D.C.: Asia and the Americas. The Annals of World Bank Group. Regional Science, 38(1), 149-171. Singh, R. J., and M.A. Barton-Dock. 2015. Perge, E., and T. Scot. 2016. Vulnerability Haiti - Toward a new narrative: systematic and Poverty in Haiti. Mimeo. Washington country diagnostic. Washington, D.C.: D.C.: World Bank. World Bank Group. 47 Singh, R., C. Bodea, and M. Higashijima. nytimes.com/2010/11/26/opinion/26wi- 2016. Guns, books, or doctors? Conflict lentz.html and public spending in Haiti: lessons World Bank. 1985. Haiti - Policy proposals for from cross-country evidence. Policy growth. Research working paper no. WPS 7681. ———. 1998. Haiti - The challenges of poverty Washington, D.C.: World Bank Group. reduction. Tobin, K. A. 2013. Population density and ———. 2007. Social Resilience and State housing in Port-au-Prince: historical Fragility in Haiti. Washington, DC: World construction of vulnerability. Journal of Bank. Urban History, 39(6), 1045-1061. ———. 2009. World Development Report 2009: Tilmans, S., K. Russel, R. Sklar, L. Page, Reshaping Economic Geography. S. Kramer, and J. Davis. 2015. Contain- ———. 2010. Haiti earthquake PDNA (Post-Di- er-Based Sanitation: Assessing Costs and saster Needs Assessment) : assessment of Effectiveness of Excreta Management damage, losses, general and sectoral needs. in Cap-Haïtien, Haiti. Environment & Washington, DC: World Bank. Urbanization 27(1): 89-104. ———. 2015. Country Partnership Framework United Nations. 2014. World Urbanization for the Republic of Haiti for the Period Prospects: The 2014 Revision. Department FY16-FY19. Washington, DC: World of Economic and Social Affairs, Popula- Bank. . tion Division. ———. 2016. Better Spending, Better Services: UNFPA-Haiti. 2010. Enquete sur les migra- A Review of Public Finances in Haiti, tions internes: phase I: Aire metropol- Overview. Washington, DC: The World itaine de Port-au-Prince, draft report. Bank Group. Port-au-Prince: UNFPA-Haiti. WTO. 2015. Trade Policy Review: Haiti. World USAID. 2010. Housing for Haiti’s Middle Class. Trade Organization. Post-Earthquake Diagnosis and Strategy. Yarrington, L. 2015. The Paved and the Washington D.C. Unpaved: Toward a Political Economy of USAID. 2011. Feed the Future Strategy for Infrastructure, Mobility, and Urbaniza- USAID in Haiti (Fiscal Years 2011-2015). tion in Haiti. Economic Anthropology, 2(1), Washington D.C. 185-204. Vakis, R., J. Rigolini, and L. Lucchetti. Zéphyr, D., A. Córdova, H. Salgado, and 2016.  Left behind: chronic poverty in M.A. Seligson. 2011. Haïti en détresse: Latin America and the Caribbean. World impact du séisme de 2010 sur la vie et les Bank Publications. opinions des citoyens. Nous. Walsh, B. M. 2000. Urbanization and the regional distribution of population in post-famine Ireland. Journal of European economic history, 29(1), 109. Wilentz, A. (2010, November 25). In Haiti, waiting for the Grand Bayakou. The New York Times. Retrieved from http://www. 48 SPOTLIGHT 1: WHAT IS URBAN AND WHAT IS AT RISK Sarah E. Antos* Urban definitions can vary widely from one country to another. But how urban areas are defined can impact a broad range of issues that span from our understanding of the evolution of a country and its cities to the investment decisions that a country takes. Hence, understanding “what is urban” and what is not is of great importance, as it can influence the decisions that are taken today in terms of basic service provision, education, and health, among others. But defining what is urban is not an easy task in a country like Haiti. While the census has historically divided the country into urban and rural enumeration areas, the criteria used for such classification are not clearly recorded in official documents. Even if such criteria were better documented, they would be outdated today, given they were made during the last census in 2003. Alternative methods that do not rely on population measures but rather on observation of built-up area can be an alternative to circumvent these challenges. Lastly, comparison of urbanization rates across countries is also challenging because each country uses its own criteria, so finding a common definition across countries can help in benchmarking Haiti’s urbanization process to that of its Latin American peers. For this report, we apply three alternative methods to reach a better assessment of what is urban today in Haiti. Details of the approach used in these three different methodologies are provided in Box 1. The results are a definition of urbanization levels that is comparable with other countries across the LAC region, irrespective of differences between countries’ own national definitions of urban.1 Our analysis suggests that as much as 64 percent of the population lives in urban areas, making Haiti one of the most urbanized countries in the LAC region. Using a cut-off point of 5,000 people as the minimum threshold to identify an urban cluster, we estimate the urban population of Haiti to be around 6,179,000 people. This is equivalent to 64 percent of the total population, and is higher than the Institut Haïtien de Statistique et Informatique’s (IHSI) estimate for 2015 (52 percent) and slightly higher than the UN World Urbanization Prospects (UN WUP) estimate for 2015 (57 percent). 1 The challenges arising from using these conflicting definitions for comparison between countries or for global aggregation of data are not unique to Haiti and have been widely discussed in previous work (Satterthwaite, 2007; World Bank, 2009; Dijkstra and Poelman, 2014). *The spatial analysis was led by Sarah E. Antos but relied heavily on analysis performed by Nancy Lozano-Gracia, Chandan Dueskar, and Benjamin P. Stewart. Furthermore, the section benefited from the valuable support of Lauren Nicole Dauphin and Katie McWilliams as well as data generously provided by the DLR Earth Observation Center, the Global Facility for Disaster Reduction and Recovery (GFDRR) and Haiti’s National Center for Geospatial Information (CNIGS). 49 Furthermore, more than half of this urban population (36 percent of the total population) lives in high-density clusters. High-density clusters are groups of contiguous cells with more than 50,000 people and with more than 1,500 people per sq. km. Figure 2 provides a detailed overview of the share of population in urban clusters, compared with other Latin American and Caribbean countries, based on our estimations. The share of urban population as reported by the UN WUP is also reported for comparison. A look at the system of cities suggests that while Port-au-Prince dominates the Haitian urban system, there are six other sizeable cities in the country. A full list of classification of cities based on the gridded population data is provided in Annex 1. The results indicate that 40 percent of Figure 1. HAITI – URBAN AND HIGH-DENSITY CLUSTERS USING THE EUROPEAN COMMISSION’S METHODOLOGY ON WORLDPOP DATA (2015) Source: World Bank analysis using WorldPop data 50 the population in cities is located in the metro area of Port-au-Prince. The second largest city is Cap-Haïtien in the Nord département. Other large cities include and Léogane in the Ouest département. Gonaïves, Saint-Marc and Dessalines, each with over 200,000 individuals. Over 19 percent of the population is located in cities between 100,000 and 300,000 inhabitants. Medium and small-size cities, with respective populations of 50,000-100,000 and 10,000- 50,000, have shown positive growth since 1990, with some cities averaging between 8 and 9 percent of growth each year. PROPORTION OF POPULATION IN HIGH-DENSITY AND URBAN Figure 2. CLUSTERS (WORLDPOP VS. UN WUP URBANIZATION RATE) Source: Deuskar, Stewart, and Lozano-Gracia (2016) based on WorldPop (2015 data), European Commission thresholds for urban areas, and UN World Urbanization Prospects (2015). 51 BOX 1 – THREE METHODS TO ESTIMATE URBAN AREAS This work compares the results obtained from three alternative methods for identifying “what is urban” in Haiti. The results of these three different measures or proxies for level of urbanization for most communal sections in the country, allowing comparisons in cases where there are discrepancies between the three methodologies. The steps followed for each of the methods used are outlined below. The first methodology involves updating the IHSI classification of communal sections into five classes, based on the proportion of each communal section’s population that is classified as urban in IHSI’s 2015 population projections. The data is extracted from the IHSI publication, available online (Institut Haïtien de Statistique et d’Informatique, 2015). This classification follows the scheme previously used by IHSI on 2003 census data, in which the level of urbanization of a communal section is defined either as très fort if 75-100 percent of its population is recorded as urban, fort (50-75 percent), moyen (25-50 percent), faible (10-25 percent), or très faible (0-25 percent). The assignment of each communal section to an urban class provides the official characterization of urban areas in Haiti. This is “what is urban” according to official statistics. A second methodology uses WorldPop high-resolution census data to model population at a 100x100 meters grid cell resolution.2 These data rely on the 2009 demographic estimations by the IHSI, combined with a wide range of other sources, including data on built-up areas, topography, and locations of hospitals and schools. We then use the European Commission’s (EC) “degree of urbanization” approach, which applies population size and population density thresholds to the above gridded population data. We first identify cells with a popula- tion density higher than 300 people per sq. km, group them into clusters of contiguous cells, and define those clusters with more than 2,000 or 5,000 people as “urban clusters,” depending on the threshold. Those with more than 50,000 people and made of cells with more than 1,500 people per sq. km are labeled as “high-den- sity clusters.” The third methodology uses data on built-up area, a proxy for urbanization that does not rely on census data. We combine the Global Human Settlements Layer (GHSL) and the Global Urban Footprints (GUF) layer. GHSL is produced by the EC’s Joint Research Center (JRC) and is based on optical imagery (LandSat); it provides an historical perspective over the proportion of the area of communal sections that were built-up.3 The GUF is based on radar data from satellites, with 84m resolution, and is therefore more accurate. Provided that the GUF is only available for circa 2011-2012, a combination of GHSL and GUF will provide a conservative estimate of built-up areas in Haiti, rather than a comprehensive one. These results are sensitive to the empirical methodologies, data, and assumptions used, and should therefore be taken with caution. However, the analysis confirms that Haiti is significantly urbanized, and that estimates of “what is urban” leveraging on high-resolution data can allow a more informed conversation about urbanization in Haiti. The results from these methodologies should be taken as an effort to have a better understanding of what is urban in Haiti and not as a final count of urban population. Ultimately, there is no substitute for a good census. Sources: Tobin (2013), Yarrington (2015). 2 Full metadata on the WorldPop Haiti layer can be found online at http://www.worldpop.org.uk/data/WorldPop_data/AllCon- tinents/HTI-POP_metadata.html. 3 Full GHSL dataset can be retrieved at http://ghslsys.jrc.ec.europa.eu. GUF data can be browsed and requested at http://www. dlr.de/eoc/en/desktopdefault.aspx/tabid-9628/16557_read-40454/. 52 GROWTH OF BUILT-UP AREA IN HAITI, 1975-2015 Figure 3. Note: Data from 2015 is from the 2015 GUF, while all other data is from GHSL trimmed with the 2015 GUF.8 Source: Authors’ calculations. As the urban population has grown, so too has the extent of the built-up land area. The analysis indicates that built-up area in Haiti roughly quadrupled between 1975 and 2015 (Figure 3). Between 1975 and 1990 only, the land area expanded an average annual growth rate of 5.1 percent, doubling from 87 sq. km. The growth rate slowed, as would be expected given the larger base. The built-up areas grew at an average rate of 1.4 percent a year between 1990 and 2000, and then at an average of 0.8 percent a year until 2015 when the total urban land area reached 363.6 sq. km. The majority of the built up area is concentrated in the Ouest départment. In keeping with population distribu- tion, nearly 53 percent (192.5 sq km) of the total built-up area in the country was concentrated in that départment as of 2015. Artibonite and the Nord follow with 152.6 sq. km. of built-up area (14.5 percent of Haiti’s built-up area) and 28.7 sq. km. (9 percent of Haiti’s built-up area) respectively. 8 Marconcini, M., S. Üreyen, T. Esch, A. Metz, J. Zeidler, and D. Palacios-Lopez. 2017. “Outlining the urban side of the Earth – the GUF+2015,” Scientific Data (in preparation). 53 But the way cities expand also has important implications for resilience: the analysis for this report suggests that the majority of land in and around urban areas is highly exposed to multiple hazards.4 Analysis of satellite data indicates that 94 percent of built-up areas in 2011 are considerably vulnerable to earthquakes.5 Nearly all of the land area of Haiti (97 percent) is exposed to “medium” or “high” seismic hazard, but built-up areas are disproportionately concen- trated in high seismic hazard zones.6 Seismic hazard exposure is a particularly serious problem in Nord-Est, Nord, and Ouest départements,7 but the majority of land in cities of all sizes – from Port-au-Prince and Cap-Haïtien to smaller cities like Ounaminthe, Mirebalais, Fort-Liberté, and Léogâne – are considered high risk. The widespread nature of this risk underlines the importance of incorporating mitigation methods into all urban construction and infrastructure development. This is discussed further in Chapters 2 and 3. Exposure to erosion and landslides is increasing as urban growth continues to occur in high-risk areas. Overall, we estimated that 58 percent of built-up areas are at risk from flood events 9, and 24.2 percent are exposed to elevated, serious, or very serious erosion.10 Total built-up area exposed to 4 Spatial data on exposure to natural hazards (earthquake, flooding, landslides, and soil erosion) was obtained from the Haiti Data website (www.haitidata.org), an online repository of spatial information for Haiti. These layers are available only at one point in time; thus, in this analysis we must assume that the level of exposure to hazards is static over the time period being evaluated. The combined layer was then intersected with the various risk layers listed above to find the built-up area exposed to risk in each year. When the input risk layer is a vector layer (flood, erosion), the resulting output layer has the same resolution as the combined built-up layer (84m). When the input risk layer is a raster layer, the resulting output layer will have the lower resolution of the two layers used as input. 5 This raster layer shows the 2 percent probabilities of exceedance in 50 years of a given peak ground acceleration (PGA, color coded, from 0 to 180 = 1.8g), taking into account the soil amplification. The 50-year period corresponds to the average life span of a building. A 2 percent probability over 50 years is equivalent to an annual probability of 1/2,5000. The probable ground acceleration levels have been converted into Modified Mercalli Intensities, in accordance with the criteria established by the USGS project known as Shakemap (http://earthquake.usgs.gov/eqcenter/shakemap/). Potential damage ranges from zero to moderate for inten- sities 1 to 5 (maximum PGA of 0.092 g), from strong to very strong for intensities 6 to 7 (maximum PGA of 0.34), and from severe to extreme for intensities 8 and above. Categorization thus helps identify three areas of seismic hazard for Haiti: low, moderate, high. The influence of Haiti’s most important identified tectonic fault is clear, particularly the Presquîle du Sud Fault, where intensities might reach at least 9 degrees. Most of the territory is exposed to intensities of at least 6 degrees. It was created by the NATHAT Project, in May, 2010.” http://haitidata.org/layers/cnigs.spatialdata:hti_hazardseismic_intensitymodified2percentprob_raster062010 6 While only 33 percent of the country faces high seismic risk and 65 percent faces medium risk, 60 percent of the built-up areas are actually in the high-risk zones compared to 34 percent in medium-risk zones. 7 Where 87, 79, and 76 percent of built-up areas face high seismic risk. Only 31, 13, and 60 percent of their total land areas are prone to high seismic risk, which again indicates that their built-up area is disproportionately concentrated in the riskiest zones. 8 Three separate layers were used, showing “probable” and “propitious” (potential) flood zones as polygons, as well as “frequent” flood zones (only for Port-au-Prince): (i) Probable flood zones: “This map layer shows Haiti’s region of probable (frequent) flooding. It was created by NATHAT Project, using Google Maps, digital terrain models, and field observations, in May 2010”; (ii) Propitious flood zones: “This polygon vector layer shows areas of propitious flood areas for Haiti. It was created by United Nations Institute for Training and Research (UNITAR), in May 2010”; (iii) Frequent flood zones: “This map layer models areas of frequent flooding for Port-au-Prince region. It was created by United Nations Institute for Training and Research (UNITAR), in May 2010” (Not available for the rest of Haiti). See Haitidata.org. 9 “This polygon vector layer shows Haiti’s areas most susceptible to erosion…It is taken at a 1:300,000 scale. It was created by the ‘Secrétairerie d’Etat au Plan’ and CNIGS in April, 1998.” http://haitidata.org/layers/cnigs.spatialdata:hti_environment_erosion_polygon_042008 10 Three separate layers were used, showing “probable” and “propitious” (potential) flood zones as polygons, as well as “frequent” flood zones (only for Port-au-Prince): (i) Probable flood zones: “This map layer shows Haiti’s region of probable (frequent) flooding. It was created by 54 PORT-AU-PRINCE EROSION AND FLOOD RISK Figure 4. (A) Erosion risk and built-up areas in (B) Flood risk and built-up areas in Port-au-Prince Port-au-Prince Sources: NATHAT, DLR, JRC Sources: UNITAR, NATHAT, DLR, JRC floods increased from 122 sq. km in 2000 to 211 sq. km in 2015. In 2000 51.4 sq. km were exposed to elevated, serious, or very serious erosion, and in 2015 it has increased to 87.2 sq. km of built-up areas exposed. The amouvnt of built-up area exposed to medium, elevated, or high landslide suscepti- bility increased from 8.8 to 22.7 sq. km. Within the cities, landslides tend to occur in irregular neighborhoods where buildings have been precariously constructed on steep slopes. Indeed, it is worth noting that in Port-au-Prince, there have been 11 major liquification incidents since 1994. Of those slides, only three did not take place in a zone deemed precarious by the National Center for Geospatial Information (CNIGS) or in a neighborhood classified as irregular by the imagery. As the amount of built-up areas exposed to different natural hazards have steadily increased, so has the number of people. In 2015, over 44 percent of the built-up area in the country is exposed to at least one type of hazard, classified as “high exposure.” The total population at risk of “excep- tional” floods in Haiti has increased approximately by 300,000 individuals, from 2.6 million to NATHAT Project, using Google Maps, digital terrain models, and field observations, in May 2010”; (ii) Propitious flood zones: “This polygon vector layer shows areas of propitious flood areas for Haiti. It was created by United Nations Institute for Training and Research (UNITAR), in May 2010”; (iii) Frequent flood zones: “This map layer models areas of frequent flooding for Port-au-Prince region. It was created by United Nations Institute for Training and Research (UNITAR), in May 2010” (Not available for the rest of Haiti). See Haitidata.org. 55 Figure 5. CAP-HAÏTIEN FLOOD AND EROSION RISK (A) Erosion risk and built-up areas in (B) Flood risk and built-up areas in Cap-Haïtien Cap-Haïtien Sources: UNITAR, NATHAT, DLR, JRC Sources: CNIGS, DLR, JRC 2.9 million, between 2000 and 2015 (see Annex 2). With regards to erosion, the urban population exposed to “elevated,” “serious,” and “very serious” erosion has increased from approximately 750,000 people to 1.12 million in the same 15-year period. To better understand how urban growth intersects with risk exposure, an in-depth analysis of landcover was conducted for Port-au-Prince and Cap-Haïtien. Imagery covering the city footprint of Port-au-Prince and Cap-Haïtien was acquired at two different time periods. These high resolu- tion (50cm) scenes were then transformed into landcover maps. Further details on the classifica- tion approach are provided in Spotlight 2. The dominant risk in the Greater Port-au-Prince area is earthquakes, but flood risks and erosion are also present. The majority of land (78 percent of built-up area in 2011) is exposed to high seismic risk. Given that out of the total area within the boundaries of PaP, 66 percent is exposed to high seismic risk, it is clear that built up intensity is higher in high-risk areas. This can be seen in Figure 4 (A). Built-up land is concentrated in flood-risk areas, given that one-third is at risk of floods (37 percent) compared with only 13 percent of the broader area. There is also a high proportion of elevated erosion risk (27 percent of built-up area). The new areas of expansion in the north and east of the city face flood risks (Figure 4 [B]). Overall, in PaP, 47 percent of the built-up area within 5 km of the city center is exposed to at least one hazard (high exposure), while 59 percent of the built-up land within 10 to 20 km of the city center is exposed. In Cap-Haïtien, the patterns are slightly different: 73 percent of built-up areas within 3 km of the city center are in areas with high exposure 56 to at least one hazard, while only 14 percent of built-up land within 5 to 10 km of the city center have the same level of exposure. The analysis shows that residential growth in Cap-Haïtien continues to occur in high flood- prone areas. Flooding is a recurrent problem in Cap-Haïtien, due to the frequency of overflowing of the Haut du Cap river that traverses through the city and the river’s estuary (Bassin Rhodo). Satellite images from December 12, 2005 and April 15, 2015 reveal high levels of inundation along the southern and eastern side of the estuary. Despite this flooding at both time periods, construc- tion of homes here continues. The figures below show the expansion of settlements into the Bassin Rhodo estuary. Note that the dark land in the center of the 2015 image represents standing water, and the white land in the south of both images represents areas that were previously inundated, but now dry and left with dry dirt and sediment from the water. Irregular growth is also occurring in areas that experience high exposure to erosion. Risk of erosion in Cap-Haïtien is heavily concentrated in the north of the city. Figure 5 (A) shows the composition of the buildings that have been built on land that is at high risk for erosion. Almost 50 percent of the buildings located in high-risk erosion areas are located in an irregular neighborhood. This is propor- tionally much higher than the cities’ overall percentage of 27 percent. The data can shed further light on how urban areas change in response to natural disaster-induced shocks. It has long been expected that disasters have shaped urban demography in Haiti. As discussed in Chapter 1, recent analysis using mobile phone data has provided new insight on the number of people that were displaced in the aftermath of the 2010 earthquake and Hurricane Matthew in 2016. Satellite data measurement provides some new additional insights on these trends. For example, a comparison of images from Port-au-Prince between 2007 and 2015 show that over such period, there was large new expansion in the north, such as the area known as “Canaan” (discussed further in Spotlight 2). This area is itself subject to high seismic risk. Cities in Haiti are hence expanding and increasingly growing into risk areas. Information, coordination, and investments are needed to guide development toward resilience. The expansion of built-up areas presents both new opportunities and challenges for policymakers. On the one hand, these areas can have positive dividends for growth, and firms and households can take advantage of larger employment opportunities and access to goods (see Chapter 3). On the other hand, there is a need for coordination among municipalities to manage this process effectively and to take advantage of the potential economies of scale in providing basic services to these areas (see Chapter 2). Further discussion of the pattern on growth within cities and agglomeration is provided in Spotlight 2. 57 REFERENCES Deuskar, C., B.P. Stewart, and N. Lozano-Gracia. 2016. Defining Urban Areas in Haiti. Mimeo. Washington, DC: World Bank. Dijkstra, L., and H. Poelman. 2014. A Harmonised Definition of Cities and Rural Areas: The New Degree of Urbanisation, European Commission (No. 01). Regional Policy Working Papers.marcon Institut Haïtien de Statistique et d’Informatique. 2015. Population totale, population de 18 ans et plus, ménages et densités estimés en 2015. Retrieved from http://www.ihsi.ht/pdf/projection/ Estimat_PopTotal_18ans_Menag2015.pdf Marconcini, M., S. Üreyen, T. Esch, A. Metz, J. Zeidler, and D. Palacios-Lopez. “Outlining the urban side of the Earth – the GUF+2015”, Scientific Data (in preparation) Satterthwaite, D. 2007. Adapting to Climate Change in Urban Areas: The Possibilities and Constraints in Low- and Middle-Income Nations (Vol. 1). 2nd ed. World Bank. 2009. World Development Report 2009: Reshaping Economic Geography. 58 CHAPTER 2 MOVING FROM RECONSTRUCTION TO RESILIENT URBAN PLANNING FOR A BRIGHT FUTURE Alexandra Panman Nancy Lozano-Gracia and Claudia Soto* PORT-DE-PAIX, NORD-OUEST, PHOTOGRAPHED BY DAVID MCINNIS, 2010 SOURCE: FLICKR, CREATIVE COMMONS LICENSE CHAPTER 2 – MOVING FROM RECONSTRUC- TION TO RESILIENT URBAN PLANNING FOR A BRIGHT FUTURE WHY RESILIENT URBAN PLANNING? imagery. As outlined in detail in Spotlights 1 Planning for resilient urbanization is about and 2, we use these new data to explore how taking coordinated action to help shape urban areas have expanded over time, how land urban growth, with the aim of supporting a use within cities has changed, and what the country and cities’ development objectives and implications of these patterns of growth are for managing natural disaster risk. As discussed exposure to natural hazard risk. We also review in the previous chapter, the form that cities the current governance framework for urban take can have very real impacts on urban planning in Haiti, noting important advances productivity and livability. This form emerges made in recent years and the many challenges from the interaction of decisions taken by that remain. The objective of this chapter is to firms, households, and the government. Firms identify where the current weaknesses of urban decide where to produce, buy their inputs, and development lay and what is needed to build a sell their outputs; households choose where brighter future for resilient cities. to live and where to work; and governments There are three key findings and messages make decisions that range from where to locate from this chapter. The first is that urban infrastructure investments to defining zoning residents live in crowded, unserviced, and regulations. Effective coordination of the unsafe housing and neighborhoods. There are actions of these three actors is therefore key in many negative externalities associated with shaping a city’s form and, through that, influ- these conditions, which can undermine the encing its future. economic benefits of density. It is key for basic In this chapter, we highlight social, environ- service infrastructure investment to catch up mental, and economic dimensions of the with the reality of urban expansion. current form of urban development in Haiti. Second, Haitian cities are growing in an The analysis draws on household surveys uncoordinated manner, with insufficient and other existing data collection exercises, regard to natural disaster risk. New infrastruc- complemented with insights from satellite ture can influence the decisions of households *This chapter draws on background notes prepared by Chandan Deuskar, Benjamin P. Stewart, Nancy Lozano-Gracia, and Sarah E. Antos. The authors are also grateful for comments from Roland Bradshaw and Harley Etienne (University of Michigan). 63 and firms on where to locate. Given that much strengthening government capacity. In this of the land around Haitian cities is highly way, urban planning can help ensure that cities hazardous, infrastructure investment decisions grow as economically vibrant, environmentally have important implications for the number sustainable, and livable places. In line with the of people and value of assets that are exposed Haitian proverb mentioned in the opening of to natural disasters. It is therefore paramount the report, this chapter provides options that that risk assessments, risk optimization strate- look at today’s problems, but with an eye on gies, and land use planning are integrated into setting stepping stones for a brighter future. urban investment decisions. Third, Haitian cities are marked by weak CITIES IN HAITI ARE MARKED BY BASIC land administration. Opaque information on SERVICE DEFICITS AND HIGH EXPOSURE land ownership and poorly functioning land TO NATURAL DISASTER RISKS regulation hampers efforts to deliver basic High levels of population density are not services and integrate hazard risk into land supported by basic service infrastructure use planning and building standards and can In Haitian cities, most residents live in leave poor households vulnerable to eviction. crowded conditions. Although precise data A well-functioning land market is also central on overcrowding is not available, on average, to harnessing the agglomeration benefits of urban households of between 4 and 7 members urbanization, as it is the mechanism through will share accommodation with only two which land is allocated for its most productive bedrooms (DHS 2012). As outlined in Spotlight uses. Long-term efforts to improve the quality 1, settlement patterns are comparatively dense, of land administration are thus central to the even in smaller cities. This density, however, is resilience of urban development. not supported by basic service infrastructure. How can Haiti address these challenges? Urban areas have developed with inadequate Fundamentally, improving resilient urban basic services. As discussed in Chapter 1, this planning capacity is a question of governance. under-provision of basic services has come Strong institutional frameworks are needed about as a result of specific historical, political, to guide decision making among the many and financial conditions. It has important players whose choices impact urban outcomes. implications for current and future living The good news is that carefully prioritized and standards. sequenced short-term projects can help build Current water supply and sanitation (WSS) confidence in change and promote a virtuous systems are inadequate to serve the urban cycle of governance (World Bank 2011b). This population. As highlighted in Chapter 1, more chapter identifies a number of tools that can than one-third (35 percent) of urban residents simultaneously help address immediate urban do not have access to improved sources of challenges and also support long-term objec- water (WDI 2015), and trends show that rates tives of building confidence in collective action, are declining.1 Overall, only 58 percent of the supporting state-society engagement, and Haitian population had access to drinkable 1 Improved sources of water include piped water into dwelling, to yard, or to the neighbor; public tap water; tube well or borehole; protected dug well; protected spring; rainwater; bottled water or sold by company. 64 water from an improved source. This figure Senegal, Benin, Mali, and Ghana, with collec- places Haiti 25 percentage points (p.p.) below tion rates of 21, 23, 40, and 85 percent, respec- the second worst performing country in the LAC tively (Hoornweg and Bhada-Tata 2012). As region (the Dominican Republic) and among indicated in Figure 1 (A), it also lags far behind the 10 worst-performing low-income countries other countries in the Caribbean. Furthermore, worldwide (slightly better than Eritrea, Niger, although information is scarce, it is thought and Tanzania). Two-thirds (66 percent) of that collection rates vary substantially within urban residents lack improved sanitation (WDI the country. Figure (B) presents information 2015).2 The overall access rates in Haiti are 50 on collection rates from 2001; it shows that p.p. lower than the average country in the LAC there is only one area of more than 100,000 region and 5 p.p. lower than the average low-in- inhabitants where more than half of the waste come country. The 2012 DHS indicates that produced is collected. 48 percent of residents in Port-au-Prince and In addition to this, none of the waste that 41 percent of households in second-tier urban is collected in Haitian cities is disposed of in agglomerations use pit latrines with slab.3 At a sanitary landfill. The most common form least 8 percent of urban residents practice open of disposal is use of open dumpsites, which defecation (WDI 2015); and research suggests accounts for 62 percent of waste disposal in that this figure may be higher given that urban the country or about 1.2 million tons of waste residents that rely on public toilets may resort per year. This is the second largest share of to open defecation to meet sanitation needs at dumping in Latin America, close to Guatema- night (Tilmans et al. 2015). la’s 69.8 percent, and ahead of Nicaragua’s 59.3 Low levels of solid waste removal services percent. A large portion of waste in major cities exacerbate flood and disease risks. Solid waste is disposed of in water sources, exacerbating management is central to ensuring productive challenges of urban flooding and the associ- urbanization, since effective removal of waste ated toll of diseases.5 In addition to this, there is vital for a healthy urban environment. Haiti are “congestion effects” of litter, uncollected has the lowest collection service coverage in garbage, and other signs of poor cleaning and 4 the Latin American and Caribbean region. maintenance. Looking ahead, these challenges With an overall waste collection rate of 12.4 are only set to increase. According to the World percent, Haiti lags far behind the next lowest Bank’s “What a Waste” report, estimated country in the region, Paraguay, which collects solid waste production is likely to skyrocket in 57 percent of waste produced and behind countries such as Haiti, from an estimated 3,233 other low-income African countries such as tons per day today to 11,152 tons per day by 2025. 2 Improved sanitation facilities include non-shared toilets flushing to piped sewer system, to pit latrine, and to septic tank, pit latrines improved by ventilation or with slab, and composting toilet. 3 Pit latrine with slab is a dry pit latrine whereby the pit is fully covered by a slab or platform that is fitted either with a squatting hole or seat. The platform should be solid and can be made of any type of material (concrete, logs with earth or mud, cement, etc.) as long as it adequately covers the pit without exposing the pit contents other than through the squatting hole or seat. 4 Comparison of data from Regional Evaluation of Solid Waste Management in Latin America and the Caribbean, 2010, and data for Haiti in 2012 taken from the “L’évolution des conditions de vie en Haïti entre 2007 et 2012” (IHSI, IRD, Dial, Nopoor, ANR. 2014). 5 See Hoornweg and Bhada-Tata. (2012). 65 Figure 1. SOLID WASTE COLLECTION RATES ARE LOW (A) Waste collection rate (B) Waste collection rate, cities of 100,000 or more in Haiti Source: World Bank calculations based on (i) L’évolution des conditions de vie en Haïti entre 2007 et 2012. IHSI, IRD, Dial, Nopoor, AN 2014 (Haiti); (ii) Jamaica Population and Housing Census 2011 (Jamaica); (iii) Censo Población y Vivienda 2010, Volumen 2, pg. 470 (Dominican Republic); (iv) UNSTAT 2013 (Dominica); (v) EVAL 2010 (Belize); (vi) INTEGRATED SOLID WASTE MANAGEMENT PROJECT – GRENADA, Caribbean Dev Bank, Appendix 2.3. 2014 (Granada); (vii) Oficina Nacional de Estadística e Información, Chart 2.48 (p. 53). (Cuba); (viii) SIDSDOCK 2015 (St Kitts/Nevis); and (ix) IADB 2015 Capacity Building workshop on Solid Waste Management in Barbados. (Knowledge Sharing Programme KSP-IADB). 2015. Source for cities within Haiti: What a waste, 2012. Annex G. Collection Data for Cities over 100,000 people (Data date: 2001) 66 There is urgent need for additional basic nate potential sources of drinkable water service infrastructure investment to meet (Graham and Polizzotto, 2013). Further- growing needs. Current infrastructure more, workers that manually empty latrines deficits in Haiti can be attributed to a number in Haitian cities – known as bayakou – have of factors, including financial and human been recorded dumping collected waste on capital resource constraints and structural the ground, in ravines, and even in rivers deficiencies in the management of existing (Katz, 2014). As further outlined in Box 4, networks. At the national level, WSS services bayakous have appeared as a response to are heavily dependent on external financing, the wide gaps in the availability of adequate with 61 percent of the National Drinking sanitation infrastructure. The lack of infra- Water and Sanitation Directorate (DINEPA) structure and the unsafe conditions in which operating expenditures and 95 percent of they are forced to work enable the spread of investment costs financed by development diseases and can increase health challenges 6 partners. In urban areas, only 54 percent related to flooding. Indeed, it is notable that of the operating expenditures (excluding 42.5 percent of deaths in Haiti are attrib- depreciation) of urban water operating units utable to communicable diseases and that are covered by water revenues.7 The pace water-borne diseases are a leading cause for of urban growth is adding pressure to this mortality of children in Haiti (World Bank situation, with the experience of medium- 2014). sized cities being a case in point; medium- The current pattern of urban growth sized cities are grappling to provide services exacerbates basic service delivery challenges. in response to sudden increases in popula- The urban development patterns of large tion (Country Partnership Framework). Haitian cities such as Cap-Haïtien and There are high social, economic, and Port-au-Prince create additional obstacles environmental costs to these basic service for basic services. In overcrowded settle- deficits in dense urban areas. Poor quality, ments in Port-au-Prince, households often dense sanitation increases exposure to do not have sufficient room in their houses communicable diseases such as diarrhea, for private sanitation solutions8; while some typhoid, dysentery, and cholera. In crowded settlements in Cap-Haïtien are in areas cities, shared latrines are associated with where pit latrines cannot be dug, because higher exposure to health risks (Heijnen the buildings are constructed on compacted et al. 2014; Fuller, Clasen, Heijnen, and solid waste, above land that is otherwise Eisenberg 2014). If pit latrines are close unstable due to sea water encroachment to groundwater wells, they can contami- (Tilmans et al. 2015; Pelling 2011). In the 6 Source: FY 2013/14 DINEPA Budget. The IADB and the AECID are the major providers of funding and technical assistance for DINEPA, with the World Bank, UNICEF, the Swiss government, the US CDC, and other organizations also providing assistance. 7 PAD: Sustainable Rural and Small Towns Water and Sanitation Project (P148970), World Bank 2015. 8 It is notable that in their report of a pilot Container Based Sanitation project, Russel et al. (2015) note that one-third of initially selected participants were screened from the project due to having insufficient space in their dwelling for household level sanitation facilities. 67 BOX 1 – URBAN FLOODING IN FOCUS Our analysis of satellite data indicates that more than half (51 percent) of built-up areas in Haiti are exposed to flood events. The area exposed is rising: the data indicates that built-up areas exposed to flooding increased by 87 sq. km between 1990 and 2011. Furthermore, the proportion of urban land exposed to floods is greater than non-built-up land: one-fifth (20 percent) of urban land is exposed to floods, compared with only 6 percent of Haitian land in general. Unregulated growth increases exposure to flood risks, as can be seen in Cap-Haïtien. Analysis of satellite imagery suggests that pockets of irregular settlements have appeared in different areas of the city, in both the city center and the outskirts of the city (see Spotlight 2 for further details). We estimated that Figure 2. CAP-HAÏTIEN FLOOD AND EROSION RISK In 2010 there were 1,714 rooftops within 50 meters from the Haut du Cap’s Bassin Rhodo by 2015 that number jumped to 2,274, a rise of 32 percent. Of those 2,274 rooftops, a high proportion of them are part of an irregular residential neighborhood. More specif- ically, 55 percent of the rooftops located in this highly flood-prone land appear irregular. Share of Buildings within 50 meters of the river basin 2015 0 0.5 1 Note: Rooftop points created in 2010 and provided by CNIGS offered a baseline. New structures that appeared in the 2015 image were manually added. 68 about 72 percent of Cap-Haïtien’s buildings in 2015 had been constructed on flood prone land.9 Of the buildings located in high flood risk areas, 22 percent of them are located in neighborhoods that have been classified as irregular, and therefore structurally vulnerable, using semi-automated methods for satellite imagery classifica- tion.10 Furthermore, construction in these areas is ongoing: despite the fact that the images from December 12, 2005 and April 15, 2015 reveal large amounts of inundation along the southern and eastern side of the Haut du Cap river estuary (Bassin Rhodo), the number of houses in this area increased by 32 percent over this time period. This can be seen in Figure 2. Urban form matters for the severity of flood incidence, and urban planning tools can help mitigate damage. Flood risk is driven by exposure to weather events and the physical vulnerability of cities, which is often impacted by a range of public policy decisions, including watershed management and deforestation. In urban areas, the challenges of flooding are often exacerbated by poor planning practices. For example, as paving and other impermeable surfaces increase, the importance of well-functioning drainage for storm water run-off increases. Vulnerability to urban flooding can be addressed through a combination of structural and non-structural measures. Structural measures are those that help control the flow of water – for example, through investment in drainage and water barriers. Non-structural measures include information dissemination and evacuation plans, which help keep people safe from flooding. The case of Sao Bernardo do Campo, highlighted in the recommen- dations section of this chapter, is an example of an approach that integrates both structural and non-structural measures to address flooding, while also integrating flood management with other development objectives. case for solid waste management, narrow and even features of the urban built environ- roads impede collection vehicles from ment such as narrow winding streets and reaching houses. The urban form thus social wellbeing. This is because the design adds both technical and financial strains to of built environment can influence the ease efforts to meet service deficits. with which interaction in public spaces can The current pattern of urban growth also be monitored by other users and the extent undermines living standards by creating to which different groups such as people of situations opportunities for violence. As different ages or gender feel ownership and noted in Chapter 1, urbanization and poverty capacity to use them (World Bank 2011). reduction are closely connected. Yet the shape Litter and solid waste dumping can also that cities take has important implications contribute to the perception of lawlessness, for social risks such as exposure to crime. affecting both opportunistic crime and the There is a link between lack of public services sense of security felt by a victim.11 9 For this calculation, the city of Cap-Haïtien consists of 4 sections: Bande du Nord, Haut du Cap, Petite Anse, and Basee Plaine. 10 This “irregular” label can be considered a proxy for relatively lower income neighborhoods and from a remote sensing/ technical perspective means the area is characterized by small, unorganized buildings. 11 The impact of physical disorder such as litter on community decline is based on the Broken Window theory (J. Q. Wilson, G. L. Kelling, 1982) which suggests that signs of disorderly and petty criminal behavior trigger more disorderly and petty criminal behavior, thus causing the behavior to spread. This may cause a development sequence in a neighborhood leading in the medium and long term to decay and deterioration of the quality of life of its inhabitants. 69 Table 1. THE PROCESS TO REGISTER PROPERTY IN HAITI IS CUMBERSOME STEP TO REGISTER TRANSFER OF LAND ACTORS ESTIMATED COST ESTIMATED TIME 1. Permission to Commissaire du 2 months survey land gouvernement (in commune) 2. Survey land* Surveyor HTG 15,000 1 month 3. Prepare sales Notary 1 percent of sale price; VAT 2-3 weeks (simulta- agreement (percent varies by property type) neous with surveying) 4. Obtain ‘avis de DGI Various fixed fees and supple- 1 day cotisation’ and pay mentary taxes and stamp duty, as well as fees for registration, for registration transcription (based on percent- age of property price) 5. Register sale Local community 6-9 months office of DGI Source: Doing Business 2017 * land surveying is required every ten years and can be no older than five years at time of transaction The pattern of urban growth in Haiti also alone are estimated at an amount equivalent increases exposure to natural disaster risk to almost 2 percent of GDP per year.12 In Haiti is considered one of the world’s most addition to the immediate costs of humani- exposed countries to multiple natural hazards, tarian disasters, there are often hidden longer- including hurricanes, floods, erosion, term costs of natural disasters in urban areas. droughts, earthquakes, and landslides. For example, in the aftermath of the 2010 Natural disasters can wipe out advances in earthquake, increased reliance on off-grid living standards in urban areas and further energy supply such as diesel generators or exacerbate basic service deficits. Overall, biomass burning, as well as demolition and historical data for the period of 1976-2012 construction efforts, may have contributed to indicates that average damages and losses rising levels of harmful air pollution in cities associated with hydrometeorological events (Davis and Rappaport 2014).13 12 Diagnostic sur l’impact économique et budgétaire des désastres en Haïti, World Bank 2014 13 Although there is no publicly available data on air quality in Haiti, insights from independent research indicate that air pollution may be a serious concern in large cities, particularly for households and informal vendors in densely populated settlements. Davis and Rappaport (2014) sampled the air quality in Port-au-Prince and Cap-Hatien in 2012 and 2013. The PM2.5 levels recorded in 70 BOX 2 – CORRUPTION IN HAITI: A THREAT TO DEVELOPMENT AND LIVES According to the 2016 Corruption Perceptions Index (CPI) of Transparency International, Haiti ranks 159th out of 176 countries for perceived levels of public sector corruption. It obtained a score of 20 of a maximum of 100, along with countries like Burundi, the Central African Republic, Chad, and the Republic of Congo. Corruption thwarts economic growth, discourages foreign investment, and reduces resources for infra- structure, public services, and anti-poverty programs (Robinson 1998; Ugur 2014). In 2004, the Haitian government established the Anti-Corruption Unit (L’Unité de Lutte Contre La Corruption [ULCC]), an autonomous agency under the Ministry of Economy and Finances tasked with fighting corruption in all its forms within public institutions. Despite some improvements, Haiti still ranks lowest in control of corrup- tion and government effectiveness compared to its LAC peers. Corruption in the country takes form in the corruption of institutions responsible for rule of law, embezzlement of public funds by political and private organizations, payments to government-associated individuals for goods not provided and services not rendered, and abuse of discretionary accounts by government officials, among others. Corruption has also been associated with the impact of natural hazards in Haiti. Ambraseys and Bilham (2011) observed the link between corruption and deaths in earthquakes and calculated that 83 percent of all deaths caused by building collapse since 1980 took place in countries scoring consistently low in the CPI. They point out that compliance to earthquake building standards depends on responsible gover- nance, which, among other factors, can be undermined by corruption. Corrupt practices in the construc- tion industry, such as bribery in the form of awarded construction contracts and corrupt inspection practices, among others, are a major contributor to the death toll in the wake of a natural disaster. Such is the case in Haiti, where an estimated 200,000 people lost their lives as a result of the 2010 earthquake. Sources: Ambraseys and Bilham, 2011; Robinson, 1998; Singh and Barton-Dock, 2015; Ugur, 2014. Analysis of satellite data indicates that are built-up areas less prone to risks than the vast majority of built-up areas are other parts of the country, but there are still vulnerable to natural hazards. As discussed areas within cities that are highly exposed. in depth in Spotlight 1, most land in all Urban expansion continues to occur cities is considered highly exposed to earth- in hazardous areas, resulting in growing quake hazard. Furthermore, built-up areas numbers of people exposed to disaster risk. are disproportionately concentrated in The proportion of built-up area exposed to high seismic hazard zones. One quarter of risks has remained steady over time in Haiti, the country’s land (26 percent) is exposed suggesting that the pace of growth has been to erosion, and more than half of all urban just as strong in risk areas as in areas with less areas are considered at flood risk (see Box risk (see Spotlight 2 for further details). This 1). Only in terms of erosion and landslides pattern of growth increases the number of the Port-au-Prince sample sites would be considered “hazardous” (18 percent of cases) or “very unhealthy” (41 percent of cases) by US Environmental Protection Agency standards if they were observed over a 24-hour period. The PM2.5 levels recorded in Cap-Hatien traffic were higher than levels published for any other city in a developing country other than Nanjing, China. 71 WEAK LAND ADMINISTRATION, INAPPRO- people exposed to risk.14 Indeed, it is notable PRIATE REGULATION, AND INFORMATION that of the 113 landslides that have occurred GAPS HAMPER EFFECTIVE DECISION since 1994 (and mapped by CNIGS), almost MAKING AND E XACERBATE PLANNING half of them have taken place in densely or CHALLENGES intermediately densely populated areas. This underlines the urgency of incorporating risk Vital building regulations are not enforced information into land use planning, as will The frequency and intensity of disaster be discussed further below. events in Haiti make land use zoning and Current growth patterns also lead to building codes vital. Building codes are very greater exposure of assets, such as water important in Haitian cities, given the high and sanitation infrastructure and education percentage of urban areas that are exposed facilities. The government of Haiti to natural disaster risk. Haiti has developed estimated that the 2010 earthquake resulted several tools to strengthen the housing sector in USD 15 million in damages to pre-ex- since the 2010 earthquake, including: (i) the isting water supply infrastructure, including National Building Code, which integrates reservoirs and pipes. In the capital, one of retrofitting (2012); (ii) guidelines for repair the five buildings of the Autonomous Metro- and construction of small masonry buildings politan Drinking Water Plant (CAMEP) was (2011); (iii) the National Housing Policy (2013); destroyed, and 15 percent of the 600 private and (iv) a communication strategy to promote water sales outlets in the metropolitan area better building processes in Haiti. The World were damaged (World Bank, 2010). Overall, Bank is also supporting the Ministry of Health the housing sector has incurred the highest to reinforce their “safe hospital” unit and the share of damages and losses in recent major Ministry of Education on safe school guide- hydrometeorological events, 28 percent of lines and has supported the Ministry to build total damages and losses for hurricanes safe community schools in underserved areas Jeanne (2004), Faye, Gustav, Hanna, and (World Bank 2015a). Challenges remain, Ike (2008), and 31 percent for Hurricane however, and many buildings damaged in Matthew (2016). central Port-au-Prince during the earthquake have yet to be repaired. Data from the 2012 DHS survey indicates that 54 percent of respondents in the metropolitan area of Port-au-Prince said that their house was damaged during the earthquake. Of these, 63 percent said that the damage was evaluated by a team of experts, but only just under 40 percent of those that had been evaluated confirmed their buildings had been completed or were in the process of implementing the needed repairs. 14 By some estimates, increased density of population in exposed areas in Port-au-Prince and Gonaïves in the second half of the 20th century has led to between a doubling and quadrupling of risk associated with tropical cyclones (Klose 2011). 72 There are, however, fundamental – but there is a need to explore how lessons challenges in translating building codes into from rental support cash grants can be trans- common practice, including both financial formed to support long-term improvements constraints and the difficulty of attracting in the rental sector.16 and retaining qualified personnel to oversee Alongside limited enforcement of vital and enforce them. Enforcement is particu- building codes, land administration is larly challenging (see Box 2). In addition to cumbersome. It is expensive and time-con- this, affordability remains a major concern suming to register property and gain for many households. Sustainable construc- building permission. Haiti ranks 180th in tion practices require expensive materials the World Bank’s Doing Business ranking and qualified workmanship. The majority in terms of ease of registering property and of Haitians live in self-produced housing 166th for getting a construction permit. (i.e., informal housing that is built without The process for registering a land trans- the appropriate technical expertise) and action is conducted in accordance with is incrementally expanded in line with stipulations set by a law from 1890, and households’ needs and available resources. there is no mechanism to file a complaint Housing finance is almost nonexistent, against a mistake made in the transaction and less than 15 percent of the population registration process. The steps are set out even have an account at a financial institu- in Table 1 below. Although the number of tion (WDI 2014). Many urban residents are steps itself is comparable to that of OECD tenants (USAID 2016).15 A number of innova- countries, the process takes nearly 14 tive efforts to increase the supply of safe times longer. 17 Construction permit fees, rental housing were piloted in the aftermath in turn, are estimated to be as high as 15 of the 2010 earthquake. These have met with percent of the total cost of construction, short-term success and have been replicated which is much higher than the average in other countries – such as in the Philip- of 2.5 percent in the Latin America and pines in the aftermath of Typhoon Haiyan Caribbean region. 15 According to the Haitian statistics agency, 53 percent of Port-au-Prince residents in 2010 were renters. This number increases to 65 percent if households rent the land that the house they “own” is built on. During the 2010 earthquake, Oxfam America estimated that 75 percent of internally displaced persons in camps were renters (Etienne 2012). 16 Cash grants empower recipients to prioritize their own needs and make decisions about where to locate. The project design included a provision known as “Keep the Change” to mitigate any potential upward pressure on rental prices. This provision encouraged households to negotiate their rent with landlords, by allowing the tenant to keep the difference between the rental allowance and the agreed rent paid. The inclusion of a verification visit to ensure that the rental housing met minimum safety standards acted as an incentive for landlords to improve the quality of housing offered. Further efforts are needed to explore the long-term sustainability of innovative programs such as this. While the short-term impacts were hailed as highly successful, some estimates indicate that as few as one quarter of beneficiaries renewed their contract at the end of the year, with insuffi- cient funds being cited as the main reason for moving (Phelps 2016). It has also been suggested that some households used their grant money to purchase land in irregular areas. 17 On average the process takes 312 days compared to 22 days in OECD countries and 68 days in the Latin America and Caribbean region average. It is also estimated that the process costs 7 percent of the value of the property, compared to 6 percent in the LAC region, and 4 percent among OECD countries. 73 The high costs of formal land development the responsibility of registering land trans- impact adherence to vital regulations such actions belongs to the General Tax Office as building codes. Inappropriate regulation (Direction Générale des Impôts [DGI]) of the drives up the costs of land development and Ministry of Finance who notes land trans- is a disincentive to formal development of actions by hand in a chronological book of land. As such, it does little to generate govern- records dating back to 1824 (IMF 2015). The ment revenue while undermining compliance lack of clear land records and necessary with vital safety standards (GFDRR 2011). planning tools, as well as fragmented gover- Furthermore, across the world, inappropriate nance, have many negative repercussions for land regulation processes have been linked resilient urban planning. to informal development, as households are Lack of transparent and accessible land thought to be pushed out of formal urban ownership records is a constraint on urban housing and into the informal sector. Although housing and infrastructure investment it is difficult to provide reliable estimates of programs. Water, sanitation, or transporta- housing needs, most attempts suggest that tion infrastructure projects face extensive between 300,000 and 400,000 units are needed implementation delays when the design is – while the formal system has never produced based on out-of-date maps. Furthermore, more than 4,000 housing units per year (Hoek- lack of clear land ownership records can Smit 2013). Estimates suggest that 60 percent undermine large-scale urban investment of Haitian households do not have any formal projects (see Box 3), as lack of trust in the document of ownership (USAID 2010). fairness of the proceedings can cause urban public-private partnership projects to become Opaque land ownership is a constraint to bogged down in costly delays and controver- infrastructure and housing investment sies (UN Habitat 2011). Land records are also The quality of land administration in essential for an effective property taxation Haiti is low (Figure 3). Largely inspired by system, which in turn has implications the French system, the current system of for local municipal finances and thus the property registration relies on judgment of capacity of local government (as discussed a restricted number of public notaries and further in Chapter 4). In addition to this, in surveyors, which are positions appointed by the past, it has been an avenue for political the President. In addition, the rights to the corruption (see Box 4. ). profession are usually inherited, and the Uncertainty over property rights has been procedures to integrate new professionals a constraint on the development of safe, lack transparency (IMF 2015 and Oriol et al. affordable housing. A number of donors and 2017). Furthermore, the land registry system NGOs have developed projects to provide a is fragmented. The Office of the Cadastre range of housing solutions, including new is housed under the Ministry of Public infill houses, new finished houses in green Works; geospatial imagery is kept in the field developments, and multifamily units. National Center for Geospatial Information Homes are often subsidized through a highly (Centre National de L’information Geospatiale subsidized model of leasing-to-own (the cost [CNIGS]) in the Ministry of Planning; and of the new houses ranges from USD 12,000 74 BOX 3 – THE UNCERTAINTY IN LAND LAWS AND ITS IMPACT ON PEOPLE’S LIVES Investment in infrastructure within and between cities is seriously undermined by the uncertainty of Haiti’s land laws – ultimately affecting real people. In 2013, Reuters reported the case of the National Road No. 7, a 56-mile road project meant to connect Les Cayes – a port city in the south – with Jérémie – a city in Grand’Anse, one of Haiti’s poorest départements. The USD 100 million project was announced in 2008 and was backed by the Canadian International Development Agency and the Inter-American Development Bank. Shortly after beginning the works, the bid-winning company abruptly abandoned the construction site. The project had run into parcels of land whose ownership was unclear and for which displaced residents had not been compensated. The company left behind incomplete infrastructure works, roadside homes seriously damaged by its trucks, and households with demolished homes. National Road No. 7 remained then as a poor-quality, single lane, and dangerous road, especially in most remote areas. Source: Ferreira, 2013. to USD 40,000 per unit and occupiers are direct records to measure the extent of land expected to make payments of between 1 conflicts between citizens, they are thought to and 5 percent of that value). Yet the absence be increasing as competition for land intensi- of secure titles that can be collateralized are fies with urbanization (Etienne 2012; USAID a major constraint to affordability of these 2010). Furthermore, as USAID stressed, in projects, as credit is only available through the aftermath of disasters, households with expensive short-term loans (Hoek-Smit 2013). weak or contested land tenure claims find Lack of clarity of land ownership adds themselves increasingly vulnerable to efforts uncertainty for citizens, can be a source of by actors with the ability to take advantage tension between citizens and the state, and of the disruptive circumstances to concen- even with emergency relief and other non-gov- trate their land holdings (USAID 2010). ernment organizations (NGOs). Although Eviction can even take place as part of public most households lack property documenta- programs: Amnesty International recorded tion, there exist a combination of formal and that hundreds of families were evicted from informal processes that – although complex downtown Port-au-Prince to make way – allow households to inherit, manage, lease, for construction of public administration and transfer land securely (Tarter et al. 2016). buildings (Amnesty International 2015). Nonetheless, mechanisms to address griev- Furthermore, lack of clarity over ownership ances or conflicts are lacking. Land arbitra- has hampered emergency response in the tion processes are opaque, time-consuming, aftermath of the 2010 earthquake, as lack and highly variable from one commune to of clear property rights was a constraint on another (see Annex 3 for details of the legal NGOs ability to support households with framework for managing property disputes). financing to recover and/or repair property On average, legal cases take five years to (International Housing Coalition 2011). In resolve (OAS 2010). Although there are no addition to this, there were instances where 75 Figure 3. OVERALL QUALITY OF LAND ADMINISTRATION IS POOR COMPARED WITH LAC PEERS Note: this index is comprised of information on reliability of infrastructure, transparency of information, geographic coverage, equal access to property rights, and land dispute resolution. The Index is scored from 0 to 30. The higher the score, the higher the quality of land administration. The dotted line represents the average LAC. Source: Doing Business 2017. lack of clarity over land ownership under- the loss of land records that occurred with mined trust in the work of some aid efforts, the destruction of the DGI building in the as there were cases where NGOs became 2010 earthquake. The Haitian government, implicated in land conflict after being under the leadership of the Inter-Ministe- granted permission to establish operations rial Committee for Territorial Development in areas they were unaware were the subject (Comité Interministériel d’Aménagement du of ownership disputes (Etienne 2012). Territoire [CIAT]) is currently proposing to Efforts are underway to introduce a land update the legal framework for land tenure cadastre.18 Efforts to modernize this system systems, modernizing land administration have faced numerous challenges, not least tools, and elaborating a methodology for 18 A land cadastre is a public inventory of land. The core information that it usually contains on all properties includes: bound- aries; ownership or interests (rights, restrictions, and responsibilities); improvements, in the form of buildings and infrastruc- ture; and an estimation of the value. The information in a land cadastre is methodically arranged and displayed in maps. 76 BOX 4 – PROPERTY RIGHTS, INSTITUTIONAL CAPACITY, AND STATE FRAGILITY The World Development Report of 2011 stressed that legitimate institutions are the “immune system” that helps defend countries against the internal and external stresses that result in conflict and violence. State, market, and social institutions that provide security, justice, and economic opportunities are thus central to peace, stability, and development. Property rights and broader land market institutions are examples of such institutions. They are particularly important where rapid urbanization weakens social cohesion and informal dispute-resolution mechanisms. The literature on land in conflict environments underlines this message further. Many studies have stressed that where land ownership records are not transparent and publicly accessible, they can be exploited by political factions to buy support for either their government or rebellion (Global Land Tool Network, online). Lack of transparency over land rights and use can even be used as a form of gerrymandering: changes to land use rights motivated by a desire to influence settlement patterns and thereby shape electoral outcomes (de Waal 2009). Public land assets are substantial in many countries across the world, and self-interested public authorities can capture private benefits of this land by amending the user rights associated with the land or by selling off these public assets at below market rates to allies. Indeed, it is notable that in Haiti it has been alleged that the use of land to curry political favor was common under both Presidents Duvalier and Aristide (Etienne 2012). How can property and land market institutions be strengthened? The World Development Report highlights several key messages for transforming institutions in fragile situations. Institutional reform is never easy, and it is further complicated where there is a legacy of violence that can undermine trust in government and hinder cooperation. Reforms in themselves may be blocked or derailed by actors who risk losing out from changes: any significant shift in the status quo is likely to create both winners and losers, and if the losers are well-organized they can form a powerful lobby against reform. Thus, for example, experience suggests that actors who benefit from the current system of property regis- tration may resist reform to the system out of self-interest. Furthermore, in contexts where there is lack of trust between citizens and state, individual households may be strongly suspicious of government efforts to collect cadastral information. Nonetheless, global experience suggests that institutional strengthening can be attained, through an approach that combines confidence building, participation, and careful prioritization of reform. As will be discussed in greater depth below, these are principles that can also help guide institutional strengthening for resilient urban planning. In short, there are potentially strong mutual benefits between Sources: Ambraseys and Bilham, 2011; Robinson, 1998; Singh and Barton-Dock, 2015; Ugur, 2014. 77 the establishment of a “pre-cadastre,” which of a broader effort to transition from recon- draws on geo-referenced data on land tenure struction to comprehensive forward-looking to link parcels and land ownership. Initial planning. This approach has combined insti- piloting stages have been completed. Box 5 tutional decentralization, as well as signif- highlights some of the findings of one such icant efforts to provide strategic vision to pilot. Although the process is likely to be coordinate activity in specific localities and challenging to complete, progress in this area sectors. can have many important long-term benefits. Urban planning responsibilities are These could also extend to improved disaster formally divided among a number of different risk management, as land ownership maps bodies. These are highlighted in Figure 4. are also a key tool for disseminating infor- The three main bodies with planning respon- mation and enforcing regulation on hazard sibilities are: The Ministry of Planning and areas; effective property taxation can be a External Cooperation (MPCE); the Ministry tool to incentivize compliance with building of Interior and Local Authorities (MICT); norms19; and current and accessible land and the Ministry of Public Works, Transport, records can also help authorities to react in and Communications (MTPTC). Four the aftermath of a disaster.20 sector-specific ministries also conducting important activities that shape urban DESPITE RECENT EFFORTS, GOVERNANCE form include: The Ministry of Agriculture CHALLENGES REMAIN A HURDLE TO (MARNDR), the Ministry of Health (MSPP), LONG-TERM RESILIENT URBAN GROWTH the Ministry of Environment (MDE), and the Since 2010 government activities in urban Ministry of Finance. These ministries house areas across all levels have been strongly agencies that have roles that are central to focused on reconstruction activities. This has effective planning, such as Office National included the challenging task of coordinating du Cadastre (ONACA, housed under the the work of many non-government organiza- Ministry of Public Works), the Centre tions (NGOs) and international development National de l’information Geospatiale organizations whose activities also shape the (CNIGS, under the Ministry of Planning), urban space.21 At present, the government and the Direction Générale des Impôts (DGI, is undertaking a series of reforms as part under the Ministry of Finance). The Comite 19 In Turkey, houses that abide by regulation and pay taxes are eligible to participate in an earthquake insurance fund (the Turkey Catastrophe Insurance Pool), a mechanism that resulted in insurance coverage rocketing from 600,000 to 3.5 million in the year it was established (GFDRR 2011). For further discussion on the state of taxation collection in Haiti, please see chapter 4. 20 For example, the existence of documented and up-to-date public cadastral records and urban hazard maps, as well as inventories of public roads and infrastructure was a vital tool in the aftermath of Hurricane Katrina hitting New Orleans in 2005. The New Orleans authorities salvaged these legal records and used the information to help inform relocation responses, provide data for insurance companies and banks to respond, and to plan the rebuilding of basic service infrastructure (World Bank blog, 2016). 21 NGOs build houses, deliver vital services, and help with disaster risk management assistance in Haiti. These activities shape urban space and require considerable coordination and oversight. Although a “cluster system” was introduced to coordinate the international humanitarian response to the 2010 earthquake, it is widely believed that only a fraction of the estimated 10,000 NGOs present in Haiti are registered under this system. http://blogs.worldbank.org/latinamerica/what-haiti-taught-us-all 78 BOX 5 – AN EFFORT TO DOCUMENT THE LAND USE AND OWNERSHIP PATTERNS IN PAP As part of the efforts around implementing the Plan Foncier de Base in Haiti and collect accurate and up-to-date information on land use and ownership, a pilot was implemented in 2013 in part of the district of Bas Peu-de-Chose, outside the boundaries of the colonial city of Port-au-Prince, in a neighborhood called Le Bas Peau-de-Chose. This effort included surveying a total of 997 plots, with 798 parcels covering 26 hectares. Analysis from the information collected suggests that 41 percent of parcels are between 100 and 250 square meters. Most of these plots (82 percent or 73 percent of the surveyed land) is controlled by private owners. State-owned land represents 8 percent of the parcels and 27 percent of the surveyed area; it is characterized by large land lots: 22 percent of state-owned parcels cover 89 percent of the state-owned area). Overall, the neigh- borhoods of Bas Peau-de-Chose investigated remain mainly a residential area with 48 percent of plots being residential; another 41 percent is either devoted to commercial activities or occupied by public services, while the remainder is a combination of residential and commercial use. A considerable number of owners do not live in their property (29 percent). Moreover, even though such private parcels have been purchased by individuals (36 percent) or are held under undivided ownership, en indivision (31 percent), property titles were collected only for 31 percent of them; 77 percent of these documents were notarized. Further, the information collected for this area suggested that only 1 percent of all the parcels in the area analyzed are in tenure conflict, suggesting ownership conflicts are not of big concern in the area studied. In general, key challenges observed in the pilot neighborhoods of Bas Peu-de-Chose in regards to the misuse of land stem from the nearly complete absence of rules of urban planning and mismanagement of land. Source: Contributed by CIAT based on CIAT, 2017. Les Cahiers du foncier du CIAT. Le Plan Foncier de Base à Bas Peu-de-Chose. Les leçons apprises. Secrétariat Technique du Comité Interministériel d’Aménagement du Territoire. No. 2, July, 2017. Interministeriel d’Amenegament du Terri- government. The municipalities are then toire (CIAT), which was created in 2009 and organized into 42 arrondissements, which, in is headed out of the Prime Minister’s office, groups of three to seven, finally composed 10 has overall coordinating responsibility for all departments. In theory, municipalities have a ministerial initiatives regarding territorial share of the responsibility for urban planning planning. activities, such as the provision of water and Local governments are also important sanitation services. For further details on the actors in urban development. The constitu- roles and responsibilities of these munici- tion of Haiti defines the country as having “a palities, arrondissements, and departments, decentralized form of government” organized please see Chapter 4. along three layers (Constitution of Haiti, In practice, there is a gap between the 1987). There are 570 “communal sections,” structure on paper and the day-to-day the smallest political subdivision. They functioning of government. Across the are distributed across 146 municipalities world, the daily functioning of government is (communes), the intermediate level of local shaped by both written and unwritten rules. 79 Figure 4. THERE ARE MANY GOVERNMENTAL BODIES WITH RESPONSIBILITY FOR URBAN PLANNING AND DEVELOPMENT Source: Authors’ elaboration Unwritten rules, bargaining, and power that local bodies are unable to fulfill respon- dynamics between interested actors all have sibilities in practice. This is discussed in important impact on how activities take greater depth in Chapter 4. place in practice (World Development Report In recent years the government has under- 2017). Although there is a lack of research taken a number of initiatives to promote that rigorously documents these dynamics decentralization, which can help improve in Haiti, there is evidence of duplication accountability by bringing service delivery and ambiguity over responsibility for urban responsibilities closer to the population that planning decisions and implementation. benefits from it. The Strategic Development For example, it can be seen in the uncer- Plan of Haiti emphasizes the importance of tainty over the allocation of funds to support territorial reform to achieving the country’s urban planning activities such as basic development objectives, and there is a nation- service provision. Thus, in the case of water al-level commitment to decentralization provision, legal responsibility is divided in through the establishment of regional and an unclear and overlapping manner between local government offices. Important efforts national, departmental, and communal in technical assistance for capacity building bodies; while unpredictable financial flows and public investments have taken place in (both in terms of quantity and timing) mean line with this vision. The plan also outlines a 80 vision of regional growth pole development, These plans help to fill an important infor- which is expected to help counterbalance the mation gap, but implementation remains a political and economic dominance of Port-au- challenge. There are two major constraints Prince. Other notable developments include to the effectiveness of these plans in shaping the election of mayors for all municipalities the development of Haiti’s urban areas in across the country in 2016, for the first time practice. First, many planning instruments in ten years. Nonetheless, many challenges exist in the law but are not implemented in remain. Municipal capacity varies substan- practice. For example, the law on ‘organisa- tially, even between municipalities within tion de la Collectivité territoriale de Section the larger metropolitan areas. Constraints Communale’ of 1996 and the ‘Décret portant include human resource limitations in the sur l’organisation et le fonctionnement form of insufficient number of qualified staff des Sections Communales’ of February and municipal revenue. Efforts to devolve 2006 specify a number of instruments that power to local authority remain incomplete, have rarely been used (see Annex 3 for and only about half of the funds designated further details). Secondly, where plans are for communes are transferred in practice (see developed, there remains a challenging gap Chapter 4 for further details). between the expectations set by the plans Important efforts have also been undertaken and the financial and technical capacity to help guide decision making across different to implement their recommendations in levels and sectors of government through the practice. There remains lack of clarity development of strategic plans. Plans play a over the division of responsibilities across vital role in resilient urbanization, since they different levels of government, uncertainty can provide a framework to leverage the value over financing for activities, and some of investments by integrating development confusion over the legal status of plans. objectives across sectors and different levels Effective plans must be backed up by the of government. The Government of Haiti has capacity to guide the behavior of house- taken considerable efforts to establish effective holds, firms, and other government actors and coordinated decision-making processes to abide by the plan. Haiti is not alone in in recent years. One key element of this has grappling with this cwhallenge. Across been the development of national, sectoral, Sub-Saharan Africa, for example, many and local plans. As indicated in Annex 4, this cities have detailed plans that set high includes a national development plan and standards for urban development. These comprehensive policies for the housing, water master plans are expensive to produce, and sanitation, and disaster risk management requiring months, if not years, of consci- sectors. As also indicated in the annex, there entious work. Many, however, have little have been almost a dozen efforts to develop an impact, as they set planning ambitions that effective master plan for Port-au-Prince and are difficult to enforce. Where they set exces- 22 Cap-Haïtien in recent years. sively high standards for development, they 22 This includes UN Habitat supported forums in 2011 (which included consultations with 600 representatives of the private sector, civil society, academics and professionals of architecture and planning, community leaders, and also local and municipal technicians) and the 2014 First National Urban Forum. 81 are linked to rising informality and reduced ening capacity for resilient planning in the overall investment in urban development future. In the following section, we discuss (Lall, Henderson, and Venables 2017). a number of policy actions that can help Global experience shows that plans alone address a pressing urban challenge and also cannot guide urban development effectively; help stimulate a virtuous cycle of institu- institutional capacity is needed to ensure tional strengthening. The actions outlined that coordination and cooperation across below draw on tools to help consolidate key agencies takes place in practice. In the institutional features that distinguish fragile following section, we therefore turn to policy and violent situations from stable develop- tools that can help empower local actors and ment environments, such as: confidence strengthen coordination across the many building and bottom-up support for state-so- different sectors of government, to improve ciety engagement, which is vital to ensure capacity for resilient urban planning. that key actors will collaborate in collective action; initiatives that leverage transparency INSTRUMENTS THAT CAN HELP INITIATE of information to stem illegal financial flows CHANGE TODAY, WITH AN EYE ON BUIL- and strengthen accountability; and insti- DING STEPPING STONES FOR TOMORROW tutional strengthening in priority areas of The previous analysis has shown that justice and security.23 Haitian cities face a number of challenges. The following recommendations have Urban areas are crowded, rather than dense; been sequenced in light of the need to they are growing in an uncoordinated address urgent short-term challenges manner, with insufficient regard for risks; and build momentum for vital long-term and are hampered by lack of clear, authori- improvements in resilient urban planning tative, and accessible property rights. What and service delivery capacity. The first set of can be done to ensure a brighter future for policies are those that can help nudge urban Haiti? Cases from around the world point growth away from crowding and towards the to specific initiatives that can address each kind of density that can help a city thrive. of these challenges. Ultimately, however, Specifically, different approaches to meet sustainable improvements will require current basic service needs and build resil- improved institutional capacity for resilient ience, through measures that help address planning and effective delivery of services. current challenges associated with the This is a long-term agenda, but steps can be absence of effective planning in the past. taken today to build strong roots for future These may be termed “corrective” measures transformation. to address basic service deficits and vulner- Policymakers have several tools at their ability to hazards. The second group are disposal to address the urgent challenges efforts that leverage information as a tool of today, while also progressively strength- to improve land use management; these 23 The initiatives discussed below are aligned with recommendations from the World Development Report on Conflict and Development (2011), as well as the World Development Report on Governance and the Law (2017) on reinforcing governance (World Bank 2017a). 82 are examples of cases in which new sources in conflict-affected and fragile states, service of information are used to help coordinate improvements can send a strong signal of decision making across different government change, help build confidence in government, departments and to build support for reforms and extend the reach of the state into urban among urban citizens. These may be under- areas, and thus form part of a positive cycle stood as more “preventive” measures to help of improved governance and stability (World guide the pattern of future growth. Finally, Bank 2011b, p131-2). specific institutional strengthening initia- tives are highlighted. Continue to empower communities and strengthen local government through basic In the short term, invest to address basic service initiatives service deficits Community engagement and empower- Basic urban services such as water, waste ment is vital to successfully upgrade access collection, and maintenance of order in public to services in areas where development spaces are the front-line of citizen interaction has already taken place in an unregulated with the state. These services are spaces in manner. Upgrading existing unplanned which the state is most visible (Denney et al. areas presents a number of challenges: 2015; Jones and Howarth 2012). They can these are technically complex projects that be a visible testimony to the presence and require area-specific knowledge, as well effectiveness of government. Studies from a as ingenuity and patience (Arnold 2015). number of post-conflict countries have found Haiti has almost ten years of experience that there is a strong statistical relation- implementing Community Driven Devel- ship between participation and perceptions opment (CDD) projects. 25 Overall, this of government.24 Participation is linked to record testifies to the potential of CDD as improved trust between citizens and the a tool address urgent needs by empowering state, and among citizens themselves, and communities. Reconstruction activities as such may have implications for long-term such as the Port-au-Prince Neighborhood sustainability of urban development. Indeed, Housing Reconstruction Project PREKAD26 24 Research from Nepal, Pakistan, Sri Lanka, Uganda and Sierra Leone have found that users opinion and trust in government is linked to user experiences of services, and suggest that poor performance can be a driver of grievances. Indeed, findings suggest that the relationship between participation and perception of government may be more important than the quality of service delivery itself (Denney et al. 2015). 25 CDD is “an umbrella term for projects that actively include beneficiaries in their design and management” (Mansuri and Rao 2004). The objective is to ensure that local people have agency and voice in addressing local problems by working in partner- ship with government and other organizations in the design and implementation of development projects. It operates on the principles of transparency, participation, local empowerment, demand-responsiveness, greater downward accountability, and enhanced local capacity (World Bank, online). 26 PREKAD is a USD 65 million project supported by the Haiti Reconstruction Fund implemented between 2011 and 2016. The objective of the project was to help earthquake-affected residents of selected Port-au-Prince neighborhoods to repair and/or reconstruct their houses and/or return to improved housing conditions and improving basic community service infrastruc- ture. The project included debris removal, housing repair and reconstruction components, as well as community service infrastructure repair/improvement, and support for capacity building. 83 and the Haiti Urban Community Driven community engagement. Overall, experience 27 Development Project PRODEPUR in the Haitian urban context has demon- projects in particular have demonstrated strated that CDD can be the entry point the potential of CDD to support housing for community-based crime and violence reconstruction and improvements in basic prevention activities, as communities can be community service infrastructure. In mobilized around small-scale infrastructure addition to this, experience has shown that provision. Careful facilitation of this engage- delivery of services such as access to water ment is key (Word Bank 2013a). It may be and solid waste management are central to necessary to ensure that the design of projects conflict resolution in communities defined includes efforts to ensure that gang members by complex territorial groups or “Bases.”28 are aware of the project activities to prevent There is room to do more: through careful interference in the implementation process. design, the benefits of CDD can be further Furthermore, facilitation may need to balance leveraged. Emerging global best practices concerns that the presence of criminal actors suggest that there are specific challenges may create barriers to other specific groups with community engagement in urban areas, of community members – such as women – and thus room to further improve the design attending meetings (Arnold 2015). As such, 29 of CDD projects in cities. Specifically, three project design should reflect the principle of key elements can be strengthened in order to building “inclusive-enough” coalitions.30 leverage the full potential of CDD in Haiti. Secondly, ensure that projects are better Firstly, refine facilitation of community integrated with local government processes engagement. The size of the target and help to build the accountability of insti- community is typically larger in urban areas tutions to deliver the services over time. In than in rural areas and the population is often the context where local government capacity more heterogeneous. It is also often more is weak, CDD projects often need to rely on challenging to encourage engagement in civil society organizations to mobilize and projects, as people living in cities tend to value support community engagement (Arnold time for wage-earning jobs more than their 2015). Yet local authorities have a vital role to rural counterparts, and they may have a less play in creating a permissive environment for strong sense of local “community” (Arnold community projects to succeed. Past experi- 2015). In addition to this, in contexts where ence of NGOs in Haiti has found that even the violence and criminal activity is prevalent, lack of explicit formal approval from govern- there can be specific challenges to facilitating ment can constrain community engagement 27 PRODEPUR is an urban community-driven development project whose objective is to improve access and satisfaction with basic and social infrastructure and services, and income-generating opportunities for residents of targeted disadvantaged urban areas. 28 Bases combine local leadership, political affiliation, cultural expression, and criminal activity. 29 As stressed in a recent effort to compile lessons learned from CDD projects in seven different countries (Indonesia, Vietnam, Benin, Morocco, Kyrgyz Republic, Tanzania, and Haiti), most CDD projects across the world have been implemented in rural areas and thus the literature on urban CDDs is still in its infancy. 30 The WDR 2100 defines “inclusive-enough” coalitions as those that include the parties necessary for implementing the initial stages of confidence-building and institutional transformation; but they do not need to be “all-inclusive” (p12). 84 and willingness to act (Pelling 2011). Further- aim to address social exclusion often need more, as current experience in Haiti indicates, to be designed to stretch across adminis- the long-term sustainability of the projects can trative boundaries in urban areas. Those benefit from improved engagement of local that focus on specific settlements such as government. Recent assessments of PREKAD “slum upgrading,” need to carefully weigh and PRODEPUR projects noted that invest- how the intervention may impact different ments could have been better linked to local groups, such as renters, who may not actively development plans, and that there is a need to engage in the project but could be harmed secure long-term commitment for operation by price effects associated with infrastruc- and maintenance of the services from relevant ture improvement. CDD project design local government authorities. Indeed, as must be flexible, as social dynamics that discussed in Chapter 4, in these projects and regulate community engagement are likely other cases where local governments partner to differ across city center, peri-urban areas, with NGOs and cities in other countries, the and small towns, and even to be impacted programs have relied on delegated implemen- by specific events or shocks such as natural tation rather than strengthening local govern- disasters (Arnold 2015). ment delivery capacity. CDD approaches can help build local Build on what works to consolidate basic government capacity, if they are designed to service delivery support local governments to take respon- Improvements in basic services can be sibility for service provision in the medium regarded as a ladder, whereby each successful term. CDD projects are often designed to modernization effort also builds capacity, build capacity among local community which in turn makes it viable to undertake implementation teams to conduct activities new, more advanced, initiatives led directly such as procurement, as well as to establish by local governments. Improving basic processes that deter fraud and corruption. service management and delivery requires These include transparency of budgeting, a comprehensive effort to reform organi- internal and external third party monitoring, zational structure, build capacity in local and establishment of grievance redress governments, and raise awareness. Attempts mechanisms. Similar tools can be built to plug leaking services with piecemeal inter- into CDD projects to help strengthen local ventions risks trapping Haitian cities in a government engagement and performance low-level equilibrium of poor services and (World Bank online). Looking ahead, it will high costs. Yet it is unrealistic to expect that be important that CDD projects are designed complex and large-scale reform can be done to support eventual increased responsibility all at once. As such, it is important to prior- and effective delivery of services by munici- itize short-term initiatives that are linked to palities. long-term gains in the resilience and quality Thirdly, because benefits and costs spill of services. across administrative boundaries, coordi- The Government of Haiti (GoH), through nation is key to maximize the positive the Ministry of Interior and Local Authori- impacts of interventions. CDD projects that ties (MICT), has made important progress 85 in pushing the decentralization reform by strengthening. For example, the Jamaica directly empowering municipalities and Social Investment Fund (JSIF) provides developing various local support initia- community groups results-based support to 31 tives. As opposed to channeling funds maintain a clean community; environmental through community-based organizations to wardens have been established to enforce respond to basic service delivery needs, the local littering and ensure maintenance of the GoH is directly empowering municipalities community; and training opportunities with to finance and operate local investments in the goal of improving community participa- accordance with local and sectorial develop- tion and pride and the sense of security in ment plans. Nevertheless, the participatory urban spaces. approaches utilized in previously imple- For municipalities that are taking on mented CDD projects remain relevant to management services for the first time, initial ensure community participation, transpar- objectives could concentrate on building ency, and accountability. the basis for service provision. These could Programs that encourage good perfor- include, for example, procedures to plan mance on service delivery by providing grants service delivery in line with budget plans and or financing based on results or outputs can execution, manage procurement processes, improve service provision in the short term and communicate effectively with citizens, and contribute to building capacity in the long as well as key skills related to infrastructure term. National governments can use their planning and operations and maintenance transfer systems to provide financial resources systems. One example of such an initiative to municipalities in the form of grants or other is the “Proyecto de Desarrollo Municipal” transfers to execute their duties and provide (PRODEM) project in the Dominican basic services provided they meet perfor- Republic. This project worked with small mance criteria reviewed annually. When municipalities (populations between 2,000 these projects are framed in terms of broader and 50,000 people) in the three poorest programmatic objectives, they can effectively provinces of the country. The project provides tackle immediate challenges while building comprehensive capacity building, including capacity for the longer term. Solid waste reorganization of staff, financial management services can provide a good example where training and software, creation of municipal service performance-incentivized improve- development plans, infrastructure planning, ments can both help to address service participatory budgeting, and transparency deficits and contribute to local government and reorganization and optimization of basic 31 Local development support programs are coordinated by Ministry of Interior and Local Authorities (MICT). Recent municipal development activities build on two flagship initiatives from MICT in the Nord and Nord-Est départements of the country, the Programme d’Intervention Nord /Nord-Est (PINNE), which aims at strengthening municipal administrative structures, and the Appui à la Gouvernance et à l’Investissement Local en Haiti (AGIL) funded by the European Union in sixteen municipalities in the Nord and Nord-Est départements. The AGIL aims to empower municipalities in managing resources for local service delivery. Lokal + is a USAID-funded project that supports local governance and decentralization in Haiti, including revenue collection support. 32 The approach was successful in 28 of the 31 municipalities: they reached these three levels of performance and rewarded with a variety of works including parks, recreational facilities, cemeteries and fire stations (a total of 85 were constructed). 86 services. As a reward for reaching each of three noted above, resilient planning is fundamen- defined levels of performance, the municipali- tally about guiding the decisions taken by ties are provided a public worker to help them households, firms, and government, in order 32 support their ability to execute their services. to minimize risk exposure and to ensure In areas where some service structures are that the resulting urban form is supportive already in place, initiatives can be undertaken of broader development objectives. As the to help consolidate capacity and better leverage examples below highlight, the very act of existing resources, such as output-based aid. making information public can be a tool to Output-based aid uses performance-based catalyze citizen engagement in collective subsidies to improve delivery of services in action and build trust in government. underserved areas, sectors, or households. Output-based aid ties the disbursement of Disseminate risk-analysis insights to support public funding in the form of subsidies to the informed decision making achievement of clearly specified results that Accurate, accessible information is directly support improved access to services, needed to support non-structural measures including improved water supply and sanita- for protecting people from risk. There have tion and access to services such as energy, been important advances in understanding health care, education, solid waste manage- the risks that urban areas in Haiti face. ment, and transportation. For example, Several knowledge tools were developed in the Southern West Bank, subsidies are after the earthquake to inform reconstruc- being paid to solid waste service providers in tion processes and strengthen DRM infor- response to independently verified improve- mation necessary for planning. These ment in cleaning, collection and disposal include information on (i) hazards and risks services, and improved financial sustain- (multi-hazard risk assessments, hazard ability. The program improved the cost-re- atlas, and historical data on damages and covery mechanisms with over 90 percent of losses from PDNAs), (ii) Seismic Zonation the service providers achieving these outputs Mapping by MTPTC; (iii) location of and receiving the corresponding subsidy. In exposed assets (georeferenced critical infra- Nepal, a similar project is under implementa- structure like schools, hospitals, and roads); tion that is expected to benefit 800,000 people (iv) high-resolution satellite imagery and in five participating municipalities. Lidar for the country; and (v) full diagnostic of the fiscal and economic impacts of In the medium term, leverage information disasters in Haiti.33 Given the high level of to facilitate coordinated decision making exposure and vulnerability of Haitian cities One way to guide decisions is by making to multiple natural hazards, it is an imper- relevant information available to households, ative to leverage this information to reduce firms, and different local governments. As risk through structural corrective measures 33 For further details, see Analysis of Multiple Natural Hazards in Haiti (NATHAT) and the Guide Méthodologique Réduction des Risque Naturels en Zone Urbain en Haiti (Government of Haiti/UNDP [2015]). See also World Bank 2015b. 87 and land use planning preventive measures, The cost of mapping by drones is much as well as to improve disaster preparedness lower than aerial photography and can and response capacity of national and local bring additional benefits. Drones fly at authorities and urban communities. around 100 meters above the ground and Publicly available risk information can always within contact by remote control. be used to support vital non-structural They are thus not affected by cloud cover flood risk measures. These non-struc- and can be deployed even in the immediate tural measures include: (i) emergency aftermath of natural disasters to assess planning, such as the development of damage, as was demonstrated in Haiti in the flood evacuation plans and alert systems; aftermath of Hurricane Matthew. Beyond and (ii) information-based campaigns data collection, it is important to note that designed to encourage flood risk mitigating the introduction of this new technology can behavior, such as minimizing flood risk also present an opportunity to engage local by keeping drains clear and adjusted solid officials and citizens on risks. Specifically, waste management practices (Jha et al. it is important to leverage opportunities for 2012). Past experience in Haiti testifies data sharing, as the technology creates new to the effectiveness of such initiatives. In avenues to communicate the information areas such as Camp Perrin, in Les Cayes collected, which can in turn be a catalyst arrondissement, civil protection initiatives for behavior change. proved to be lifesaving during Hurricane In Tanzania, drones were used to map Matthew in 2016. Two months prior to the flood plains in Dar es Salaam, the country’s hurricane 100 families in the most at-risk largest city. The information was used neighborhoods took part in a disaster to plan and predict how water will move simulation exercise, and it is notable that in the event of a flood. A team of local these families were all kept safe during the researchers, local government officials, real-life devastation of the hurricane that and land surveyors were trained in the followed (UNDP 2016a). Furthermore, the use of the new technology. This training civil protection volunteer group (brigadiers) can be completed within a period of one sprang into action ahead of the disaster – or two weeks. In addition to this, over spreading information about the storm and the course of the project, it became clear preparing evacuation centers – as well as in that the excitement generated by the new the immediate aftermath by clearing access technology helped to build communication to hospitals (UNDP 2016b). There is room and engagement between local government to replicate this experience in urban areas and communities in the mapped hazard more broadly. Technological innovation areas. Maps can be printed out for discus- is providing new opportunities to engage sion with the community, the informa- citizens and disseminate information on tion can be updated simultaneously on a risks. In Haiti and in other countries across computer, and the corrected final product the world, Unmanned Aerial Vehicles (UAVs can then reprinted for verification before or drones) are increasingly being employed being incorporated into land records. As a to provide information on disaster risks. result of these positive initial experiences, 88 a much larger-scale project is underway to the incentives to implement the decisions. Yet use drones for cadastral mapping in the compliance can also be encouraged another island of Zanzibar.34 way: by making clear and credible informa- tion available to households. Use information to align incentives: aim Simple plans – that are disseminated – can for the best but prepare for the worst be highly effective in guiding new develop- Basic infrastructure investment is urgent ment. One example of this is the approach and costly. There are economies of scale in used in Tunis, Tunisia. By making vital urban service provision: the cost of providing information widely available to households, piped water in cities is estimated to be about the local government was able to guide their three times lower per capita than in sparsely choices. This approach aims for the best, 35 populated areas. But it can be much more but prepares for the worst (acknowledging expensive and complicated to provide basic that its vision for urban development faces services to unplanned areas than it is to put short-term financial constraints). Rather in place trunk infrastructure ahead of devel- than trying to restrict urban expansion into opment. By some estimates, investing in unplanned areas, the government instead basic service infrastructure ahead of devel- decided to provide clear and transparent opment is two to three times cheaper than information to the public on the future infra- “slum upgrading” (Akibo 2007), and every structure expansion plans. Households, who dollar spent on disaster risk mitigation saves are settling in what today are unplanned and society four dollars (Multihazard Mitigations unserviced areas, can use this information Council, 2005). Planning ahead can help save to make sure that rights of way are left clear financial resources. for this future investment. This benefits the Yet how can governments effectively government by reducing the costs of invest- plan ahead without the financial or insti- ment, but it also benefits the households, as tutional capacity to fully implement plans? they are less likely to be adversely affected by As discussed above, urban plans can be an future interventions. effective tool to anticipate urban growth and The example of Tunis shows that by infrastructure needs. Many important efforts disseminating information on future public to develop plans have been undertaken in investment, public authorities can help Haiti in recent years. There is, however, a guide urban expansion. This simple regula- discrepancy between the visions set out in tion can reduce infrastructure investment the plans and the reality on the ground. One costs in the long term and helps maintain the major constraint to effective use of plans is presence and capacity of the state in areas of that local actors have neither the funds nor urban expansion. In addition to this, given 34 There are important considerations around the use of new technology such as drones that need to be carefully considered. For one, few countries have established a functioning regulatory framework to govern the use of drones. The imagery produced by drones has a resolution of around three centimetres per pixel, which is sufficiently high resolution that potential privacy issues should be carefully considered. For further details on this project, see World Bank (2016a; 2016b). 35 The price is approximately USD 0.70 to USD 0.80 per cubic meter to provide piped water in urban areas, versus USD 2 in sparsely populated areas. 89 that much of the land around Haitian cities in Brazil. Sao Bernardo is one of 39 munic- is prone to natural disasters, it can also be ipalities that make up the Sao Paulo Metro- a good opportunity to guide urban develop- politan Region. The municipality is located ment to be more resilient. Specifically, the next to the Billings Reservoir of the Alto Tiete government can use hazard risk maps not watershed, and it has experienced rapid and only to inform decisions about the location informal population growth of largely poor and of future infrastructure, but also to communi- marginalized communities along the water’s cate information about risks with the public. edge. This pattern of growth has presented the There is thus the opportunity not only to save municipality with many serious challenges. costs by preserving rights of way for infra- For one, the Billings Reservoir is the primary structure investment, but also to help reduce source of water for nearly 5 million people and exposure to hazard by ensuring that risk a constitutive part of a wider watershed system information is reflected in both public and that supplies 70 percent of the vast metropol- private investment decisions. itan region’s 20 million inhabitants. Informal population growth has been associated with Integrate flood risk knowledge into trans- a rapid decline in the quality of the water, as parent urban infrastructure investment untreated sewage, solid waste, and storm water decision making runoff have increasingly polluted the reservoir. Important win-wins could be achieved by For another, the ground has become increas- integrating flood risk management informa- ingly impermeable and subject to flooding. tion with the northern-corridor development A careful program of data collection and objectives in Haiti. As recent planning efforts community engagement created the oppor- highlight, the North and Northeast areas of tunity for coordinated action on flooding the country are marked by population growth and build coalitions for collective action. The pressure, deficits of basic services and trans- municipality of Sao Bernardo started by identi- portation infrastructure, and significant flood fying at-risk informal settlements, located in risks (World Bank 2017b). The costs of “doing fragile watershed areas (World Bank 2013b).36 nothing” in this area are high: continued Armed with empirical evidence on social, development in this line will lead to increasing environmental, and economic conditions in numbers of people at risk of floods, as well as these settlements, the municipality was able to growing environmental pressures. Case studies establish a system of prioritizing investments from around the world indicate that real gains that met both the utility company’s concerns can be attained by leveraging information on over water quality in the watershed and local flood risks to guide coordinated, integrated residents’ needs for improved services. Invest- action to build resilient and sustainable cities. ments in new sewage network connections, One successful example of this comes from storm water drainage infrastructure, and the municipality of Sao Bernardo do Campo, public transportation services were combined 36 The municipality identified 261 precarious and informal settlements, of which 151 were in the environmentally fragile watershed area, and 65 were considered to be at high risk of natural disasters. They developed a transparent system of prioritization based on a combination of social, environmental, and financial considerations to identify 52 settlements for intervention. 90 with public meetings and information dissem- efforts must be carefully sequenced and ination in schools to encourage environmental prioritized, since – as Haitian reformers behavior change. New parks were created to experienced in the early 2000s – too much serve the tripartite purpose of providing “green institutional reform, too quickly, can overtax areas” to absorb storm water, a buffer between and undermine appetite for change (World urban growth and the reservoir, and public Bank 2011b, p145). space with amenities for outdoor activities and promotion of social inclusion. All the informa- Strengthen property rights with dis- tion was made public through a custom-made pute-resolution mechanisms online mapping system,37 which municipal The establishment of a single authorita- authorities believe helped build trust in the tive, transparent, accessible, and accurate government in marginalized communities. record of land ownership is vital for resilient Furthermore, it is likely that this focused urban development. The lack of transparent project has spillover benefits: by helping to land ownership records leads to inefficien- build institutional experience of integrating cies in basic service investment, opens workflows across different government author- avenues for debilitating corruption, and ities, it provides the foundation for future makes the urban poor vulnerable to eviction. integrated policy development in other sectors In addition to this, clear land records are and areas. needed to effectively integrate risks maps and other knowledge on DRM information In the long term, strengthen property – such as insurance coverage levels, exposed rights and promote institutional reform values of assets, information on the impact for improved governance of former disasters – into effective land use The government of Haiti has made planning and disaster risk management. commitments to long-term projects that The establishment of a working cadastre is, can fundamentally improve resilient urban however, a long-term project: past experience planning, such as reforming property rights in Haiti, and comparable experience from and decentralization. These are important around the world, highlight that there are and challenging commitments. Progress many challenges to cadastral reform. For the should be contextualized in the broader new system to be successful, the records must effort toward institutional transformation be accurate, legitimate, and easily accessible. in Haiti, which has included, inter alia, Sound expectation and management of costs reforms to the justice, electoral, revenue is also important, as cadastres are expensive collection, and anti-crime sectors (World to establish and maintain.38 Furthermore, Bank 2011b). Decisions over the pace, focus, there are specific challenges to transforming and sequencing of institutional reform institutions in fragile and conflict-affected 37 This system is known as “HABISP” (http://sihisb.saobernardo.sp.gov.br). 38 The “one-off ” cost of establishing a cadastre are often borne by public finances. In many countries, the running costs – including updating records – are covered through user fees (Hawerk, online). It is important to consider how cost-recovery models may affect accessibility and perceptions over the transparency of the cadastre. 91 states.39 Indeed, property rights reform and processes. ADR is a method that can vary titling initiatives can in themselves become from facilitated direct negotiations between a source of conflict; they can even result in two interested parties to efforts that more increased vulnerability among the urban closely resemble courtroom processes, and poor by introducing new procedures that it has been adopted for a wide range of they are disadvantaged to navigate (DFID contexts, from the Democratic Republic of 2002; Payne, Durand-Lasserve, and Rakodi Congo to Chile (Herrera and da Passano 40 2009). One approach may be to prioritize 2006; UN Habitat 2012; UN Habitat 2013; improvements to titling and registration of and Vlassenroot 2012).41 In some cases, it plots for new housing development in order relies on local leaders with high levels of to lower costs (Hoek-Smit 2013). social recognition, who, as research from S equencing matters: effective Mali and Kenya suggests, may be regarded dispute-resolution mechanisms are a key as better placed to solve land disputes than foundation for broader reforms. At present, the official court systems (World Bank Haitian courts are burdened with a backlog 2011b, p155). of unresolved property rights disputes. These disputes may intensify with efforts For Haiti’s largest cities, build fram e - to reform land tenure and indeed in the work s for m uni c ipal cooperation creation of an official registry of land; The footprint of urban economic activity it is therefore important that conflict- is often much wider than traditional and dispute-resolution mechanisms are administrative boundaries. As highlighted strengthened. Alternative Dispute Resolu- in Spotlight 2, Haitian urban areas are tion (ADR) mechanisms can be useful in expanding into larger agglomerations. helping to reduce the pressure on courts, As Haiti continues in the process toward resolve conflicts effectively, and even help political and fiscal decentralization, it will build confidence in formal land property be important to consider that coordina- 39 The WDR 2011 defines institutional transformation as “[d]eveloping over time ‘rules of the game’ that increase resilience to risks of violence, including laws, organizations, norms of behaviour, and shared beliefs that ensure that the benefits from individuals choosing to act peacefully and lawfully exceed the costs.” 40 The national housing plan of 2013 notes that the state will adopt a real estate policy that supports equitable and fair property rights, including land use tenure. To date there have been a number of pilot projects to this end, including a USAID-funded pilot project to map 10,000 plots in the Port-au-Prince neighborhoods of Delmas 32 and Carrefour-Feuille (USAID 2016). This initiative aims to record information on land tenure and housing ownership. Habitat for Humanity created the Haiti Property Law Working Group in 2011 focused on “longstanding land tenure issues.” 41 In Congo, the process for land conflict resolution mediation is composed of ten main steps: 1) request for mediation from the land mediator; 2) analysis of the context, scope, object, and causes of the conflict; 3) invitation of parties, witnesses, or resource people (this invitation may be made directly or through an intermediary; 4) exchanges between the parties or witnesses; 5) review and analysis of facts to identify the implications of the conflict; 6) analysis of documents or components of the file/tenure certificate or title deeds; 7) listening to witnesses (chiefs, administrative authorities, neighbors or residents, etc.; 8) visits to the area to assess the conflict, to understand the respective claims of the parties; 9) exploration of possible solutions, assisting the parties to draft agreements and signing the arrangement; 10) collective monitoring and implementa- tion of agreement (UN Habitat 2012; UN Habitat 2013; and Vlassenroot 2012). 92 tion across these adjacent municipalities is regional or national authorities to provide often beneficial. Job growth in one munici- disposal (Peru). In the case of Colombia, the pality is likely to attract workers from a wide framework leverages the financial incentives catchment area and rely on transporting provided by direct disposal facilities while inputs and outputs across large distances. No regulating the nature of disposal services municipality can single-handedly support to ensure jurisdictional boundaries do not these processes and manage the associated limit the access to service. As Haiti moves challenges, such as pollution and congestion. toward implementing decentralization As discussed in Chapter 4, in the context of objectives, it will be important to learn from limited financial resources, creation of new these experiences and develop an effective local entities may lead to increased strains means of incentivizing coordination for in resources. Yet without coordination, it service delivery. is likely there will be wasteful duplication of activities or policies may be undermined because they are contradicted by policy choices of neighbors (Samad, Lozano-Gracia, and Panman 2012). As such, it is important to consider how coordination across existing municipal boundaries can be facilitated. A number of initiatives are currently underway to encourage coordination across municipal and even departmental bound- aries, focusing on waste services. 42 Yet, without strong institutional frameworks in place to promote cooperation, these efforts face many challenges. Indeed, international experience suggests that countries that have been successful in establishing multiple shared facilities have done so within a national framework to facilitate coopera- tive agreements. There are multiple ways to develop such a framework In some Latin American countries, access to finance is the main incentive (Argentina and Brazil), while others have established legal mandates for 42 In the metropolitan region of Cap-Haïtien, the Association Intercommunale de Traitement des Ordures Ménagères Le Marien was created. This association includes the Cap-Haïtien, Quartier Morin, and Limonade municipalities, which are located in two separate departments (North and Northeast). The association is financed by AGIL and the AFD. Similarly, efforts are underway to forge cooperation between the municipalities of Caracol, Trou du Nord, Terrier Rouge, and Limonade (financed by IADB). 93 REFERENCES to reconstruction blocked tenure disputes. Reuters. Retrieved from by land http://www.reuters.com/article/us-hai- Abiko, A., L.R. de Azevedo Cardoso, R. ti-land-idUSBRE90P0BM20130126 Rinaldelli, and H.C.R. Haga. 2007. Basic Fuller, J. A., T. Clasen, M. Heijnen, and J.N. costs of slum upgrading in Brazil. Global Eisenberg. 2014. “Shared sanitation Urban Development Magazine, 3(1). and the prevalence of diarrhea in young Ambraseys, N., and R. Bilham. 2011. Corrup- children: evidence from 51 countries, tion Kills. Nature 469: 153-155. 2001-2011. The American journal of tropical Amnesty International. 2015. “Ten Facts medicine and hygiene, 91(1), 173-180. About Haiti’s Housing Crisis,” from GFDRR. 2011. Turkish Catastrophe https://www.amnesty.org /en /latest/ Insurance Pool: Providing Affordable news/2015/01/ten-facts-about-haiti-s- Earthquake Risk Insurance, from http:// housing-crisis/ w w w. g fd r r.o rg /s i te s /g fd r r.o rg /f i le s / Arnold, M. 2015. World - Participatory and documents/DFI_TCIP__Jan11.pdf Community Driven Development in Urban Government of Haiti. 2013. Politique Areas. Washington, DC: World Bank. Nationale du Logement et de l’Habitat Davis, M., and A. Rappaport. 2014. Air (PNLH) Quality in Developing World Disaster ———. (2010). Analysis of Multiple Natural and Conflict Zones – The Case of Hazards in Haiti (NATHAT) Post-Earthquake Haiti. Science of the Total Government of Haiti/UNDP. 2015. Guide Environment 496: 22-25. méthodologique Réduction des Risque DFID. 2002. Better Livelihoods for Poor People: Naturels en Zone Urbain en Haiti. The Role of Land Policy. Consultation Graham, J. P., and M.L. Polizzotto. 2013. Pit Draft. latrines and their impacts on ground- de Waal, A. 2009. Why Humanitarian water quality: a systematic review. Organizations Need to Tackle Land Environmental health perspectives, 121. Issues. In S. Pantulizano (Ed.), Unchar- Hawerk, W. Cadastral Systems in Devel- tered Territory: Land, Conflict and oping Countries - Technical Options, Humanitarian Action. Warwickshire, from https://www.fig.net/organisa- UK: Practical Action Publishing. tion/comm/7/activities/reports/events/ Denney, L., R. Mallett, and D. Mazurana. penang97/penang9710.htm 2015. Thematic Paper on Peacebuilding Heijnen, M., O. Cumming, R. Peletz, G. and Service Delivery. United Nations Ka-Seen Chan, J. Brown, K. Baker, and University. T. Clasen. 2014. Shared sanitation versus Etienne, H. F. 2012. “Land Rights, Land individual household latrines: a system- Tenure, and Urban Recovery: Rebuilding atic review of health outcomes. PLoS One, Post-Earthquake Port-au-Prince and 9(4). Léogâne,” Oxfam America Research Herrera, A., and M.G. da Passano. 2006. Land Backgrounder series. Tenure Alternative Conflict Manage- Ferreira, S. (2013, January 26). Haiti’s road ment. FAO Land Tenure Manuals 2, 94 Rome, FAO. Critical Review. The World Bank Research Hoek-Smit, M. 2013. Options for Expanding Observer 19(1): 1–39. Access to Housing Finance in Haiti. Multihazard Mitigations Council (MMC). Discussion Note. 2005. Natural Hazard Mitigation Saves: Hoornweg, D., and P. Bhada-Tata. 2012. An Independent Study to Assess the What a Waste: A Global Review of Solid Future Savings from Mitigation Activi- Waste Management. Urban Develop- ties. Volume 1 – Findings, Conclusions ment Series; Knowledge Papers No. 15, and Recommendations. World Bank, Washington, DC. OAS (2010). Foncier Haiti: Modernization of IMF. 2015. Haiti – Selected Issues, IMF Cadastre and Land Rights Infrastructure Country Report, May 2015 in Haiti. An Inclusive Approach. ——— 2014. Poverty Reduction Strategy Oriol, M., B. Jacquet, and A. Touati. Imple- Paper. mentation challenges of land adminis- International Housing Coalition. 2011. Haiti tration in rural areas of Haiti: from the Shelter Sector Assessment: One Year After elaboration of a pre-cadaster methodlogy the Earthquake. International Housing to the land tenure reform. Paper prepared Coalition: Washington D.C. for 2016 World Bank Conference on Land Jha, A. K., R. Bloch, and J. Lamond. 2012. and Poverty. World Bank. Washington, Cities and flooding: a guide to integrated DC, March 20-24, 2016 urban flood risk management for the 21st Payne, G., A. Durand-Lasserve, and C. century. World Bank Publications. Rakodi. 2009. The Limits of Land Titling Jones, S., and S. Howarth. 2012. “Supporting and Home Ownership” Environment and Infrastructure Development in Fragile Urbanization 21 (2): 443-62. and Conflict-Affected States: Learning Pelling, M. 2011. Urban Governance from Experience,” UKAID Report. and Disaster Risk Reduction in the Katz, J. M. (2014, March 12). “Haiti’s shadow Caribbean: The Experiences of Oxfam sanitation system.” The New Yorker. GB. Environment & Urbanization 23(2). Klose, C. 2011. Evidence for Higher Tropical Samad, T., N. Lozano-Gracia, and A. Storm Risks in Haiti Due to Increasing Panman. 2012. Colombia Urbanization Population Density in Hazard Prone Review: Amplifying the Gains from the Urban Areas. Environmental Research Urban Transition. Directions in Develop- Letters 6(4):1-4. ment. Washington, DC: World Bank. Lall, S.V., J.V. Henderson, and A.J. Venables. Singh, R. J., and M.A. Barton-Dock. 2015. 2017. Africa’s Cities: Opening Doors to the Haiti – Toward a new narrative: system- World. Washington, DC: World Bank. atic country diagnostic. Washington, DC: Lherisson, G. 2015. Principes, règlements World Bank. & outils opérationnels de la législation Solid Waste Association of North America haïtienne sur le développement urbain, (SWANA). 2010. “Municipal Solid Waste Port au Prince. Collection Needs in Port-au-Prince, Mansuri, G., and V. Rao. (2004). Commu- Haiti” Position Paper. nity-Based and Driven Development: A Tarter, A.M., K.K. Freeman, K. Sanders. 95 2016. A History of Landscape-level Land Participatory Approach. Report on Management Efforts in Haiti Lessons the seminar held in Brussels on 20-21 Learned from Case Studies Spanning Eight September 2012. Decades. Washington, DC: World Bank. WHO/PAHO. (31 July 2013). “Addressing Tilmans, S., K. Russel, R. Sklar, L. Page, S. Waste Management in Haiti,” Online Kramer, and J. Davis. 2015. Contain- Blog. er-Based Sanitation: Assessing Costs and World Bank. 2010. Haiti Earthquake PDNA: Effectiveness of Excreta Management Assessment of damage, losses, general and in Cap-Haïtien, Haiti. Environment & sectoral needs. Washington, DC: World Urbanization 27(1): 89-104. Bank. UNDP. 2016a. “Haiti Civil Protection Volun- ———. 2011. Violence in the City: Under- teers Put Others First,” from http:// standing and Supporting Community stories.undp.org /haiti- civil-protec- Responses to Urban Violence. tion-volunteers-put-others-first Washington, DC: World Bank. UNDP. 2016b. “Haiti Rising After the Storm,” ———. 2011b. World Development Report 2011: from http://stories.undp.org/haiti-rising- Conflict, Security and Development. after-the-storm. Washington, DC: World Bank. UN Habitat. 2011. Blog: “Port-au-Prince ———. 2012. “Integrated Urban Water hosts ‘the city we want forum,’” from Management Case Study: Sao Paul.” http://mirror.unhabitat.org /content. from http://siteresources.worldbank. asp?cid=10674&catid=153&. o r g / I N T L A C / R e s o u r c e s /2 5 7 8 0 3 - ———. 2012. “Cap-Haïtien Urban Profile.” 1351801841279/SaoPauloCaseStudyENG. ———. 2012. “Milot Urban Profile.” pdf. ———. 2012. “Toolkit and Guidance for ———. 2013. Indonesia: Evaluation of the Preventing and Managing Land and Urban Community Driven Development Natural Resources Conflict.” Program - Program Nasional Pember- ———. 2013. Guide to Land Mediation: dayaan Masyarakat Mandiri Perkotaan Based on the Experience in the Eastern (PNPM-Urban). Democratic Republic of the Congo, ISBN ———. 2013b. Blue Water Green Cities Interna- No. (Volume): 978-92 -1-132559-1. tional Workshop, December 4–6 2013, Sao USAID. 2016. “Haiti Housing & Settlements Paulo, Presentations by Amauri Pollachi Fact Sheet.” and Tassia Regina, from http://web. ———. 2010. Land Tenure And Property worldbank.org/WBSITE/EXTERNAL/ Rights In Haiti The Importance Of Land C O U N T R I E S / L AC EXT/0, ,c o n t e n t- Tenure And Property Rights Issues and MDK:23328153~pagePK:146736~piP- Post-Earthquake Recovery In Haiti. K:146830~theSitePK:258554,00.html. Property Rights And Resource Gover- ———. 2014. “Haiti Conference on Clean nance Briefing Paper 6. Water, Improved Sanitation and Better Vlassenroot, K. (Ed.). 2012. Dealing with Health.” Working Paper. Land Issues and Conflict in Eastern ———. 2015a. Haiti Country Partnership Congo: Towards an Integrated and Framework, FY 2016-2019. 96 ———.8 January 2015b. “What Haiti Taught Us All.” World Bank Blog. http://blogs. worldbank.org/latinamerica/what-haiti- taught-us-all. ———. 2016a. “Drones Offer Innova- tive Solution for Local Mapping.” Online Feature Story, from http:// w w w.w o r l d b a n k .o r g /e n /n e w s / feature/2016/01/07/drones-offer-innova- tive-solution-for-local-mapping. ———. 2016b. “Tanzania: Using Drone Technology to Secure Land Rights.” Online; h t t p s : //i r e v o l u t i o n s . org/2015/08/19/world-bank-using-uavs/. ———. 2017a. World Development Report 2017: Governance and the Law. Washington, DC: World Bank. ———. 2017b. Project Appraisal Document: Municipal Development and Urban Resilience Project. ——— Online. CDD Toolkit. Governance and Accountability Dimensions. Guidance note: Enhancing Governance and Accountability in Community-Driven Development-Based Operations. An Overview of Risks and Responses. Yazdani, M., D. Bercovitch, and J. Charles-Voltaire. 2014. Knowledge transfer on urban violence: from Brazil to Haiti. Environment and Urbanization, 26(2), 457-468. 97 SPOTLIGHT 2: PATTERNS OF LAND USE WITHIN CITIES Sarah E. Antos* Improved urban policies and investments require more, better quality and up-to-date informa- tion on the way cities are growing. Recent studies suggest that the pattern and form of urbanization may be as important as the speed of growth (Christiaensen, Weerdt, and Todo 2013; Christiaensen and Todo 2014). In fact, as cities grow, large infrastructure investments will be necessary to facil- itate the moves of goods and services, support the exchange of knowledge and ideas, and provide good quality of life for its residents. But in order to maximize the social and economic returns of these investments, infrastructure provision must be tuned with the spatial layout of people and land. Otherwise, there is the risk of resources being wasted and large populations remaining under- served, disconnected from jobs and markets. To understand the way cities are growing, we turn to very high-resolution satellite images to draw a characterization of land use patterns in Haitian cities. We use high-resolution imagery to identify numerous land cover types, using a semi-automated algorithm. These high-resolution (50cm) scenes are transformed into land cover maps using the methodology developed by Graesser et al. (2012). This process allows us to use both spectral and textural information from the images to distinguish between different uses, such as forested areas, residential buildings, and industrial warehouses.1 Originally created to accurately detect shanties in major cities throughout the world, this method has been proven effective in a diverse set of cities (Kandahar, Kabul, Caracas, and La Paz).2 Furthermore, it has been shown to be effective at capturing land cover change in five primary cities in Africa by Antos, Lozano-Gracia, and Lall (2016).3 Since its creation, it has been adopted by the US Census Bureau, US Department of Energy’s Oakridge Laboratory, and The George Washington University.4 The methodology is used to derive land cover maps for nine cities5 in Haiti over two points in time. For further details of the methodology, see Box 1. 1 Exum et al. (2005). Estimating and Projecting Impervious Cover in the Southeastern United States. US Environmental Protection Agency. 2 Graesser et al. 2012 Image-based Characterization of formal and informal neighborhoods in an urban landscape. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 5 (4) August:1164-1176. 3 Antos, Sarah Elizabeth and Lall, Somik V. and Lozano-Gracia, Nancy, The Morphology of African Cities (December 9, 2016). World Bank Policy Research Working Paper No. 7911. Available at SSRN: https://ssrn.com/abstract=2883394 4 High-Resolution Urban Image Classification Using Extended Features: Data Mining Workshops (ICDMW), 2011 IEEE 11th International Conference Dec. 2011. Author: Vatsavai, R.R. Published by IEEE. 5 These include: Cap-Haïtien, Port-au-Prince, Quanaminthe, Fort Liberté, Jérémie, L’Estere, Mirebalais, Jacmel and Miragoane. *The imagery classification and analysis was led by Sarah E. Antos, with the valuable support of Lauren Nicole Dauphin. The analysis was enriched by local expertise and auxiliary data provided by Haiti’s National Center for Geospatial Information (CNIGS) and the general guidance of World Bank’s Geospatial Operational Support Team (GOST). 99 The analysis reveals that Port-au-Prince is crowded and increasing in density over time. Being the largest urban agglomeration in the country, Port-au-Prince has over ten times the built-up area of any other urban area in the country, and more built-up area than the rest of the country combined. Port-au-Prince is also very densely populated, with approximately 32,500 people per sq. km between 1 and 2 kilometers from the city center (taken as the Marché en Fer – Iron Market). These are density levels that are comparable with certain areas of Manhattan, despite there being very few high-rise buildings. The density progressively declines the farther away from the market you go, but increases once more between the sixth and seventh kilometers from the center, where two of the densest municipalities are located – Petionville and Delmas. It is notable, as well, that the ratio of growth between Port-au-Prince and Croix-des-Bouquets remained constant between 1975 and 2011, indicating that the center of the city continued to densify overtime. BOX 1 – THE UNCERTAINTY IN LAND LAWS AND ITS IMPACT ON PEOPLE’S LIVES A semi-automated classification approach can be used to examine the texture and structural composition of various neighborhoods, grouping land with similar patterns into a single class.6 For this work, such an approach was used to detect roads and ultimately combined with a vector road network from CNIGS.7 This vector line was then converted to a raster and merged into the classified layer so a final land cover map could show primary, secondary, and residential roads. Particular attention was given to dividing up the “residential”-looking neighborhoods into regular and irregular, and for Cap-Haïtien a residential sparse class was added. Residential sparse was defined as having low-density small structures surrounded by lots of vegetation, while residential regular land had larger, more tightly packed rooftops. As rule of thumb, if the rooftops were spaced out more than two rooftops away from another building, the area was considered residential sparse. These neighborhoods are more likely to occur near the periphery of the city and in new settlements. Note that this class was only used in Cap-Haïtien due to the noticeable growth of new low-density homes in the periphery of the city. In contrast, the growth occurring in Port-au-Prince appeared much denser, with new, often tightly packed rooftops appearing throughout the city. The truck/tent class was created to avoid confusion between small rooftops and other small objects. The size of the rooftops in the residential irregular class were so small (often 10-20 sq. meters) that the algorithm started misclassifying groups of tents, market stalls, trucks in parking lots, cargo containers and mausoleums as irregular residential. Therefore, this experimental class was made to represent areas with groups of objects smaller than homes, but larger than personal cars. The commercial/industrial class represents areas that have buildings that are considerably larger than a single-family home. These buildings have long, linear features or distinct geometric curves; they are often factories, retail buildings, or manufacturing plants, typically surrounded by paved parking lots or major roads.8 6 The road layer came from the Centre National de l’Information Géo-Spatiale (CNIGS), 2016 created as part of a rural accessibility index study. 7 Ibid. 8 The road layer came from the Centre National de l’Information Géo-Spatiale (CNIGS), 2016 created as part of a rural accessibility index study. 100 It should be noted that large apartment buildings are difficult to distinguish from large office buildings and are thus frequently classified as commercial/industrial. In addition, small businesses, such as ones run out of homes or in structures the size of single-family homes will be classified as residential. Due to such challenges, the importance of confirming on the ground the results of the classification and complementing the analysis with local knowledge cannot be emphasized enough. Lastly, the algorithm was trained to detect roads, but only the major thoroughfares were successfully identified. As a result, a vector road network from CNIGS was modified to accurately reflect the vintage of each image.9 This vector line was then converted to a raster and merged into the classified layer so final land cover map could show primary, secondary, and residential roads. Training sites are areas selected by the analysis as “stereotypical” of the specified land cover class to be identified. These sites will then be used to “teach” the classifier so that it can identify similar areas throughout the image. The training sites used in this study were drawn by remote sensing experts who have experience classifying numerous cities, relying upon the physical characteristics visible in the imagery when drawing the polygons for the training sites. Examples of training sites used for the different land cover classes used in this work are shown below. To strengthen the change detection analysis that looks at two images at different points in time, whenever possible the same training sites were used to classify both images. To ensure comparability between classes over time, training sites were first drawn over each of the cities earlier imagery (2005 for Cap-Haïtien and 2006/7 for Port-au-Prince). Then, whenever possible, the training sites were used to train the second, more recent image. Occasionally, land cover changed and sites needed to be modified, but 95 percent of the training sites remained the same. The model, built by sampling the training sites, also did not alter between time periods. Screenshots of training site polygons. 1.Water 2.Vegetation 9 Different from the data we used to classify the Haitian system of cities, information about within-city distribution are drawn from the US Census Demobase (Landscan). This is a high-resolution gridded population map based on a combination of 101 3.Residential Regular 4.Residential Irregular 6.Commercial Industrial 7. Truck/Tent 8.Sparse Residential 102 PORT-AU-PRINCE LAND CHANGE AND DENSITY Figure 1. (A) Port-au-Prince 2016, change in land cover with distance (B) Density compared with African cities from city center Source: Authors’ elaboration using 2016 50cm WorldView2 imagery. The spatial structure of Port-au-Prince resembles that of comparable African cities, but density levels are higher. When compared to African cities with similar levels of per capita income, Port-au- Prince shows remarkably higher density levels. As shown in Figure (B), the pattern of population distri- bution closely resembles that of Nairobi, the capital and largest city of Kenya, and Addis, in Ethiopia.10 In all three cities, we see a high concentration of population living within just 1 kilometer of the central business district, followed by a sharp and steady decline of population density as one moves to the periphery. However, Port-au-Prince is denser: 28,700 inhabitants per sq. km live within 1 km from the main market, whereas 21,700 inhabitants per sq. km do so in Nairobi. This pattern of growth likely reflects topographical constraints: Port-au-Prince is flanked by mountains to the south and coastal areas prone to flooding in the north. As indicated in images (A) and (B) in Figure 2, growth has been constrained largely within areas that already had some settle- ments, except for new residential buildings in the northwest and a large area of irregular settlements to census data and satellite image analysis more precise than WorldPop. It allows for a comparison between Port-au-Prince and African cities similar to the Haitian capital in terms of income per capita. Based on Oxford Economics 2012 estimation, GDP per capita in Port-au-Prince amounted to USD 1,547. In Addis Ababa was estimated to be around USD 727, in Nairobi to USD 2,592, in Kigali to USD 1,380, in Dar es Salaam to USD 2,915. 10 This calculation is based on population estimates derived from 2015 (UN adjusted) WorldPop data and only includes the area of Sections Haut du Cap and Petite Anse. 103 Figure 2. GROWTH IN PORT-AU-PRINCE (A) Port-au-Prince – Dec 2006/Jan 2007 (B) Port-au-Prince – Feb 18, 2016 Source: Authors’ elaboration using 2006 50cm QuickBird2 imagery and 2016 50cm WorldView2 imagery. the north. This includes the area known as “Canaan”: a settlement of approximately 200,000 that did not exist before the earthquake of January 2010. In March of that year, a presidential decree declared the area to be for “public utility.” Thousands of households rushed to occupy and build on the land, even though the second part of the same decree forbade any construction, subdivision, or transactions in the area. In Cap-Haïtien, residents are also crowded in the central areas of the city. This is depicted visually by the graph in Figure 3. Cap-Haïtien has a population of over 200,000 people, and its urban footprint spills over municipal boundaries. The city’s average population density is estimated at around 7,800 people per square kilometer.11 As indicated in Figure 6, residential land is most concentrated in the center of the city and makes up about 20 percent of the land in between 2-4 km from the city center. This pattern is strikingly different to other cities of similar sizes in Central and South America, where residential areas have remained centrally concentrated and account for less than 5 percent of total land after 3 km from the city center. Cap-Haïtien is particularly noteworthy for residential settlements extending beyond 4 km from the city center: the expansion toward the northwest of the city is limited by typographical constraints, given the presence of tall hills (Morne du Haut du Cap). 11 Habitat 2013. “The Relevance of Street Patterns and Public Space in Urban Areas.” UN Habitat Working Paper. UN Habitat. 104 CAP-HAÏTIEN CHANGE IN LAND COVER WITH DISTANCE FROM Figure 3. THE CITY CENTER (2015) Source: Authors’ elaboration using classified layers created using 2015 50cm GeoEye1 imagery as input Satellite imagery does, however, reveal sparse residential housing emerging in other peripheral areas. The efforts to classify satellite imagery also allow looking at how much land is dedicated to road infrastructure in each city. A comparison of this indicator is displayed in Figure 4, showing Haitian cities at the bottom end of the distribution and considerably behind even some African cities. While in Accra, about 11 percent of land within the city is dedicated to roads, PaP has less than 5 percent of land dedicated to roads. 105 Figure 4. SHARE OF LAND ALLOCATED CITY ROADS, A COMPARISON BETWEEN HAITI’S CITIES AND OTHER CITIES IN THE WORLD Sources: Cities in blue are authors’ calculations from the classified layers. All other numbers are taken from UN Habitat 2013.11 106 REFERENCES Christiaensen, L., J. Weerdt, and Y. Todo. 2013. Urbanization and poverty reduction: the role of rural diversification and secondary towns. Agricultural Economics, 44(4-5), 435-447. Christiaensen, L., and Y. Todo. 2014. Poverty reduction during the rural–urban transformation– the role of the missing middle. World Development, 63, 43-58. Graesser, J., A. Cheriyadat, R.R. Vatsavai, V. Chandola, J. Long, and E. Bright. 2012. Image based characterization of formal and informal neighborhoods in an urban landscape. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 5(4), 1164-1176. Antos, S.E., N. Lozano-Gracia, and S.V. Lall. 2016. “The Morphology of African Cities.” World Bank Policy Research Working Paper, no. 7911. 107 CHAPTER 3 SHAPING LABOR MARKETS: CONNECTIVITY, JOBS, AND RISKS Paolo Avner Guilherme Augusto Zagatti Nancy Lozano-Gracia Miguel Gonzalez Canudas Sarah Elizabeth Antos and Linus Bengtsson* PORT-AU-PRINCE, OUEST, CAPITALE PHOTOGRAPHED BY JAMES G. PINSKY, 2009 SOURCE: WIKIMEDIA, PUBLIC DOMAIN CHAPTER 3 – SHAPING LABOR MARKETS: CONNECTIVITY, JOBS, AND RISKS CONNECTIVITY IN HAITIAN URBAN AREAS access to jobs, sometimes isolating entire Motorized transport is expensive with neighborhoods from the rest of the urban respect to households’ budgets in urban Haiti. economy. Improvements in the connectivity Thus, a large share of the urban population system are required to increase resilience and cannot afford to travel by any other means promote economic growth and job creation. than by foot. These commuters are left with While large investments and policies could the choice of an exhausting and lengthy trip lead to better coordination of transport and to the main employment centers within the land use, these are likely out of reach in Haiti urban area or of settling for a job close to their presently. Instead, incremental improvements homes, which might not be the best fit. For and investments aimed at better utilizing road commuters that chose to ride on Tap-Taps on space and orienting future urban develop- a regular basis, they will have access to a larger ment toward safe and well-connected areas pool of jobs, but at the cost of large travel could increase speeds on the roads, lower expenses, which will reduce their available fuel consumption, and improve the resilience budget for food, clothing, and shelter. In of the network. As will be discussed in detail both cases, accessibility is poor, leading to in this chapter, such solutions would consti- fragmented labor markets – where matching tute a first step toward a more dynamic and will happen at a local scale – and are unlikely integrated labor market capable of spurring to lead to agglomeration economies that have economic development and job creation. spurred much of the economic growth in developed and emerging economies. Cities as matchmakers between people In parallel, Haiti’s vulnerability to natural and jobs hazards adds supplementary challenges When accessibility is good, cities can for the urban transportation networks. Any become integrated labor markets that provide disruption to specific links will have ripple opportunities to residents, allowing them to effects on the economy in the form of reduced choose jobs from larger pools and leading to * The team wishes to thank Digicel for granting access to the CDR data. The authors thank Katie L. McWilliams, Benjamin P. Stewart, and Lauren Nicole Dauphin for providing important help in running the network analysis and the calculation of transport times in Port-au-Prince and Cap Haïtien. Pierre Xavier Bonneau provided crucial guidance and, together with Malaika Becoulet and Franck Taillandier, helped the team navigate the issues of urban transport in Haïti. Emilie Perge’s in-depth knowledge of the ECVMAS survey was key to understand transport expenditures and to the writing of the corresponding sections in this chapter. 111 increased welfare (Bertaud 2014). Integrated City management can help provide labor markets exist when it is possible for an residents with good accessibility levels. individual to reach a large share of the employ- Through efficient planning, interventions ment opportunities within a city at a reason- aimed at coordinating land use and transport able cost or within a reasonable timeframe. can reduce the disconnect between residential Large and integrated labor markets support areas and employment opportunities by short- improved matching by increasing the number ening travel times. Reductions in transporta- and diversity of employers and job seekers, tion costs have been found to have positive which makes the best of their respective impacts on the employment prospects of skills and aspirations (see Box 1 in Annex 5 the youth in Ethiopia (Franklin 2015) and in for an academic perspective on accessibility France (Le Gallo, L’Horty, and Petit 2017), and productivity). Consider, for example, a while land use interventions also carry the schoolteacher or carpenter in Port-au-Prince potential to increase employment accessi- looking for a job. They could apply for ten bility (Avner and Lall 2016; Peralta Quirós positions, and if they can effectively travel to and Mehndiratta 2015). Conversely, inade- all ten daily, they would be able to choose the quate urban policies can foster disconnection best one offered in terms of salaries, topics, within urban areas. The massive sprawl and or desirable location within the urban area. low-density housing development in Mexican When access is good, firms also benefit from cities was partly the result of the housing the proximity to product and labor markets policy reform of 2000. In Mexico City low-in- that the density of cities allows. When opening come households living in the peri-urban a position, an employer would want to be able areas can spend an additional four hours to choose from the largest possible pool of commuting per week compared to low-income candidates to select the one that is the fittest families residing in more central areas (Kim for the job. This matching is made possible and Zangerling 2016). And nowhere is the through connectivity: the ability to learn impact of policies on the shape of cities and about suitable opportunities and travel to residents’ accessibility as visible as in South them regularly. Conversely, when accessibility Africa, where apartheid policies produced is limited, the likelihood of finding a good an entirely segregated spatial structure with match is smaller, since firms/households must Coloureds and Africans living in peripheral select from a smaller pool of workers/employ- and areas of Cape Town mostly disconnected ment options. Jobs located outside high-den- from employment opportunities. As a result, sity economic clusters tend be scarcer, more Africans on average incurrTed commuting informal, and lower paying. Low accessibility times that were 70 percent higher than for levels also pressures families to locate closer Whites in 1998 (Rospabe and Selod 2006). to jobs, which can be a disadvantage given Connectivity and accessibility are that land and housing are more expensive, necessary – but not sufficient – conditions forcing families to live in basic conditions, to achieve efficient urban labor markets. thus fueling the phenomenon of central slums Creating jobs and achieving efficient labor found in many African cities (Antos, Lozano- markets demand a multi-dimensional Gracia, and Lall 2016). solution to overcome many obstacles, ranging 112 from the lack of a financial and banking lated activities to the national GDP (IHSI 2007) system to create businesses, low education and the cost structure of various transport levels, and a costly regulatory framework. services. This study is, however, outdated. Improving accessibility will not solve these A more recent effort led by the Inter-Amer- issues, but failing to address urban acces- ican Development Bank (IADB) produced sibility challenges will impede progress in a report documenting urban transport terms of productivity and livability. patterns in the metropolitan area of Port-au- Prince in 2010-2011 (Kopp and Prud’homme Where disasters are common, matchmaking 2011). Although this report focuses only on is harder Port-au-Prince, it is rich with information Haiti’s extremely high exposure to natural and is therefore useful to summarize its main hazards puts the economy at great risk. As findings here, with the assumption that the mentioned in Chapter 2, Haiti is one of the documented challenges are likely to apply countries most exposed to natural hazards in to other major urban areas in Haiti such as the world. Ninety-three percent of its surface Cap-Haïtien, albeit to a lesser extent. and more than 96 percent of its population Motorized transport is dominated by is at risk of two or more hazards (World Bank Tap-Taps in Port-au-Prince. Tap-Taps are and ONPES 2014). As much as 56 percent converted pick-ups, often imported from the of the country’s GDP is linked to areas United States and Canada, which can seat up exposed to risk from two or more hazards, to fourteen people but often accommodate and therefore every event, whether hurricane, up to twenty. Alongside Tap-Taps, there are flood, earthquake, landslide, or drought, minibuses and buses, which can seat between has economic consequences (World Bank twelve and thirty people and sixty people, and ONPES 2014). According to the Poverty respectively. Moto-taxis are an increasingly Assessment completed by the World Bank popular means of transport in Port-au-Prince, and the Government of Haiti in 2014, vulner- as their small size allows them to partly escape ability is extensive in the country. One million the high congestion in the metropolitan area people live slightly above the poverty line and (Ryko 2014). Private cars and two-wheelers could be pushed below it by a shock. Nearly represented in 2011 a marginal share of 75 percent of households are economically transport demand at 7 percent, the rest being impacted by at least one shock every year, public (93 percent). By comparison, public with weather-related disasters having great motorized transport is dominated by Tap-Taps disruptive potential. (56 percent) and minibuses (24 percent). Buses add an extra 8 percent while moto-taxis had a URBAN TRANSPORT IN HAITI IS SLOW 3.7 percent market share in 2011, which is likely AND UNAFFORDABLE TO MANY to have grown in recent years. Data on mobility patterns of Haitians living A massive share of people walk in Haiti and in cities and urban transport systems is scarce. motorized transport uptake is low. The share of In 2004-2005 the Institut Haïtien de Statistique households with positive spending on any type et d’Informatique (IHSI) conducted a survey of transport in 2011-12 is as low as 46 percent at documenting the contribution of transport-re- the national level, meaning that 54 percent of 113 households did not use any kind of motorized only at the Port-au-Prince metropolitan area, transport (Cadena and Perge 2017). Looking the same figures show a higher daily/regular only at “regular transport,” the transportation transport uptake, with the share of house- that is used for commuting and ordinary trips, holds that never use motorized transport for the share of households with positive expen- commuting purposes comprising between 66 ditures is even lower at 26.6 percent. In the and 80 percent for the first three quintiles, but absence of a travel survey that could decipher decreasing to around 49 percent for the richest trip mode shares, we interpret this figure as a quintile. proxy for the proportion of people who take Natural disasters, by destroying the motorized transport in cities on a regular connective infrastructure, can exacerbate the basis; the remaining 73.4 percent either do not challenges of accessibility to opportunities travel or walk everywhere. Figure shows that in urban areas. Whereas large disasters such between the 1st and the 3rd quintiles, the share as earthquakes carry the potential to disrupt of households that either do not commute or urban areas overall, lower magnitude events commute by foot could be anywhere between such as floods can nevertheless disrupt local 79 and 92 percent at the national level. Looking economies by destroying transportation Figure 1. MOTORIZED REGULAR/DAILY TRANSPORT UPTAKE AT THE NATIONAL AND METROPOLITAN LEVELS ARE LOW Source: Authors’ elaboration using 2006 50cm QuickBird2 imagery and 2016 50cm WorldView2 imagery. 114 AMONG THOSE THAT USE REGULAR MOTORIZED TRANSPORT, Figure 2. THE MAJORITY DOES SO DAILY Source: From Cadena and Perge (2017) using ECVMAS (2012) Note: Frequency of usage for regular transport users. networks, making commuting and exchanges present chapter will help identify key corridors extremely difficult. To increase the resilience that require robustness or redundancy in order of urban areas it is essential to understand the to ensure access to opportunities in the event vulnerability of the transportation network to of a disaster. low-magnitude yet disruptive natural disasters. A little less than one-third of households in One way of measuring the vulnerability of the the metropolitan area use motorized transport transport network is to calculate the accessi- every day while 57 percent never use it. Figure bility consequence of the destruction of specific 2 focuses on households that declare positive links of the network. Where Chapter 2 investi- expenditures on regular transport and inves- gated the consequences of lack of planning or tigates the frequency of this transport service implementation of planning documents on the use. In the metropolitan area of Port-au-Prince increased vulnerability to natural hazards, the the choice to resort to regular transport is 115 binary: either households use it every day or at diture was approximately HTG 9,700 or USD least a few times a week (83 percent of users or 230. So, if a Tap-Tap trip was repeated twice a 36 percent of total households) or they do not day, five days a week, transport expenditures use it at all. Irregular use of transport is low would represent anything between 25 and (16 percent of users or 7 percent of total house- 73 percent of per capita expenditures in the holds). Also, except for the first quintile, which lowest quintile. This would leave very little for has a lower share of daily regular transport fundamental expenses such as food, housing, users, usage frequency is quite steady across and clothes. the income distribution with 72-78 percent of As a consequence of unaffordability, households using regular transport every day transport budget shares are low. Instead, versus 61 percent for the first quintile. At the households prioritize food and housing. The national level, infrequent users of transport are ECVMAS (IHSI 2012) survey conducted in more widespread. This appears to be mainly 2012 demonstrates that the poor spend very due to the inclusion of residents of rural areas, little on average on regular/daily transport which have lower needs for regular transport. even in proportion to their total expenditure, Unaffordability of motorized transport around HTG 1,139 a year per household or appears to be the main reason for high walking a little above USD 27 (2012 exchange rates). shares. The Ministry of Social Affairs (MAST) Based on a Tap-Tap trip costing on average oversees prices for public transportation USD 0.35, this budget could afford less than and petroleum products. Tap-Tap fares are forty round trips a year. Non-poor households therefore regulated to increase the affordability spend on average seven times more than poor of motorized transport. The rule for defining households on regular/daily transport. This these fares is unclear, however, and so is the is again indicative that motorized transport degree to which fare regulation is enforced usage is highly dependent on income levels. in practice. Despite this fare regulation, poor Given the unaffordability of motorized inhabitants of Port-au-Prince cannot afford transport for many, transport (and regular to travel using Tap-Taps on a daily basis. transport in particular) expenditure shares Kopp and Prud’homme (2011) estimated the are low. The left pane of Figure 3 shows that unit cost of riding Tap-Taps – the cheapest these oscillate between 1.5 and 2.1 percent of motorized transport mode – to be around household expenditures for the three lowest USD 0.07 per kilometer, which would approx- quintiles in the metropolitan area and are imately amount to USD 0.35 for a 5km trip. lower in other geographic areas. The richest More conservative anecdotal evidence suggests quintile spent 2.5 times more than the lowest that Tap-Tap fares are lower, around HTG quintile relative to their total expenditures. 5 or USD 0.12 with 2012 exchange rate.1 In These low figures can largely be explained by parallel, the lowest quintile per capita expen- low usage of regular transport. When focusing 1 A study investigating Tap-Tap fares is currently ongoing. Initial results report an average fare of HTG 11-12, which would correspond to USD 0.17-0.19 using June 2, 2016 exchange rates. These prices cannot be easily compared to 2012 expenditures, however. There is also a large variation in Tap-Tap fares depending on the circuit as can be seen from official tariffs: http:// www.haitilibre.com/article-13327-haiti-avis-nouveaux-tarifs-du-transport-en-commun-zone-metropolitaine.html. These official tariffs seem generally to be consistent with Tap-Tap fares, according to the ongoing survey. 116 THE SHARES OF EXPENDITURE ON REGULAR/DAILY TRANSPORT OUT OF TOTAL EXPENDITURE BY INCOME QUINTILE AND GEOGRAPHIC AREA FOR ALL HOUSEHOLDS (LEFT) AND FOR REGULAR/DAILY TRANSPORT Figure 3. USERS (RIGHT) ARE LOW Figure 3. The shares of expen- diture on regular/daily transport out of total expenditure by income quintile and geographic area for all households (left) and for regular/daily transport users (right) are low Source: Based on Cadena and Perge (2017) using ECVMAS (2012) Source: Based on Cadena and Perge (2017) using ECVMAS (2012) specifically on regular transport users (Figure diesel (World Bank 2015). On average, the share 3, right pane), the picture is very different. of subsidies as a proportion of fuel costs was as Expenditure shares are between 6.4 and 7.8 high as 29 percent when considering volumes percent of total expenditures in the metro- of diesel and gasoline. These subsidies had politan area, with the lowest income quintile limited fiscal impact in 2010 (0.34 percent of having the highest expenditure share. While GDP), but as the price of oil products increased these numbers are not high in absolute terms and the Haitian Gourde depreciated, they grew when compared to other countries (around 15 to represent as much as 2 percent of GDP in percent in Argentina and France), they are for 2014. By comparison, fuel subsidies represent a low-income country such as Haiti. To put on average 0.9 percent of GDP in the rest of these results into perspective, expenditures the Caribbean countries and as little as 0.1 on rice represent up to 7 percent of household percent in the Dominican Republic (World expenditures for the poorest income quintiles Bank 2015). The high share of GDP committed (Q1 and Q2). to fuel subsidies is crowding out important Fuel prices have historically been subsi- public spending, such as on health or social dized, but the government of Haiti is in the protection, which amounted in 2015 to only 0.8 process of eliminating the subsidies through percent and 0.3 percent of GDP, respectively. an automatic adjustment of fuel pump prices. Finally, the fuel subsidies are highly regressive Fuel subsidies in 2014 represented 44 percent in Haiti with approximately 93 percent of the of the fuel price for gasoline and 18 percent for subsidies benefitting the richest 20 percent of 117 the population, who are more likely to own cars ties. Potentially, it would also indirectly harm and motorbikes (World Bank 2015). In light of Tap-Tap operators through lower occupation these figures, and the unsustainable nature rates of their vehicles. of the subsidies, the government of Haiti is Due to this situation, combined with low looking to implement an automatic adjust- affordability of motorized transport, it is ment mechanism of fuel prices, which would critical to devise compensation mechanisms adjust according to international fuel prices that could ensure that the subsidy removal and exchange rate movements. This automatic does not translate into higher fares or loss of adjustment mechanism has been planned by a activity for Tap-Tap operators. Compensatory 1995 Legislative Decree, but was only activated mechanisms are needed to maintain or even for the first time in 2003 and has been removed increase Tap-Tap ridership despite higher fuel and reactivated on multiple occasions between prices. Many avenues to do so are possible. 2008 and 2014. Firstly, increasing speeds on the network, The removal of the fuel subsidies and the through interventions on the road network implementation of an automatic adjustment and by rationalizing Tap-Tap routes, so as mechanism is likely to spur tension and to allow Tap-Tap drivers to complete more worsen the unaffordability of public transport round trips in a given time. This option would in the absence of compensatory mecha- increase the revenues and margins of Tap-Tap nisms. Because Tap-Tap fares are regulated, operators and could lead to lower fares, and is increased fuel costs would have to be absorbed the most promising. Another approach would by Tap-Tap operators as a non-marginal loss. be making Tap-Tap vehicles, often operated Preliminary results from an ongoing survey for more than 25 years (Kopp and Prud’homme on the cost structure of Tap-Taps indeed 2011), more fuel efficient to lower the volume of indicate that fuels represent 34 percent of fuel. Public interventions to scrap old, fuel-in- cost spending, while previous studies place efficient, informal minibuses and subsidize these as high as 49 percent (World Bank the purchase of more efficient vehicles were 2015). Previous attempts to implement the implemented in Senegal and the Dominican automatic adjustment scheme failed in part Republic and could provide a blueprint because of opposition from students, trans- for doing so in Haiti (as discussed further porter unions, and popular community below). This option would require more in organizations taking to the streets. The last depth analysis to uncover whether a suitable attempt to reinstate the automatic adjust- model can be identified and negotiated with ment mechanism in August 2016, led to a the operators. An ongoing technical assis- 20 percent increase in fuel prices and was tance of the World Bank is exploring different suspended immediately after transporter mechanisms to offset fuel cost increases and unions mobilized. Conversely, if the public the results will inform whether scrapping old transport fare could increase to compensate Tap-Tap vehicles is a viable option. Tap-Tap operators for the fuel price increase, Motorized travel is slow and lengthy this would exacerbate the unaffordability in Port-au-Prince, so even those who can of public transport and further exclude the afford it will incur high time costs. Kopp poor from accessing economic opportuni- and Prud’homme (2011) used Inter-Amer- 118 BOX 1 – FUEL PRICES POLICIES AND THEIR IMPLICATION ON CONNECTIVITY AND HOUSEHOLDS’ HABITS Since 2003, Haiti has subsidized fuel prices at high socio-economic and environmental costs. With subsidies representing a high share of total revenues (15 percent), Haiti has underspent in key sectors such as health (0.8 percent of GDP) and social protection (0.3 percent). Subsidies are also regressive since poor households spend very little on fuel products (0.11 percent of the average household’s budget). Artificially low fuel prices propped up the demand in fuel products nonetheless, which increased levels of outdoor pollution and lowered the incen- tives for industries to invest in more efficient technologies. In 2014, the Haitian government began to gradually phase out fuel subsidies; the consequences of this policy on the Haitians’ welfare is, however, ambiguous. The complete elimination of subsidies is not expected to have large direct effects on the real consumption of house- holds in the bottom 40 percent of the welfare distribution given that they spend less than 0.06 percent of their annual budget in fuel-related products, including cars, motorbikes, and generators for electricity. Eliminating subsidies would result in a 29 percent increase in the cost of fuel, but the real consumption of the bottom 40 percent would decrease by only 0.02 percent. However, indirect effects through transport and food expenses could be larger – especially for the poor and vulnerable in urban areas – for three main reasons. First, the vulner- able and the poor represent 40 and 32 percent, respectively, of transport services customers, including Tap-Taps and public buses. Vulnerable households, in particular, spend around 4 percent of their total budget on daily and school transportation. An increase in transport tariffs due to higher fuel prices would negatively affect poor households’ income. Second, increasing transportation costs could impact the share of food expenses in households’ budgets (60 percent on average). Urban households are particularly concerned, since they usually buy fresh products transported by traders from rural areas. At the national level, a 30 percent increase in fuel prices would trigger a 1 percent increase in food prices. Third, most rural households and a portion of urban households (26 percent) rely on kerosene for cooking and lighting their homes, which may become costlier with higher fuel prices. Users will react to higher transportation fares by walking longer to reduce their expenditures on transport; they will then be more exposed to pollutants from vehicle fumes, which may lead to fatigue and underperformance at school or work. As Tap-Tap owners would try to maximize their profits by limiting their circuits and overcrowding their vehicles, the quality of transport services would decrease and will ultimately become a less attractive option for customers in the long run. Increased food prices may cause reductions in poor people’s daily food intake. In urban areas, where cheaper food is more easily available, households would consume lower quality meals, with consequences for their health. Finally, an increase in the price of kerosene would push households to reduce the overall hours of house lighting and to shift core activities into daylight times. This may have implications for opening hours for business, night schools, and study time at home. The reduction in kerosene for cooking may put pressure on the environment as consumers will rely on other more polluting types of fuel – e.g. charcoal. All in all, a reduction in fuel subsidies could increase poverty by 1.1 percentage point, with vulnerable households suffering the most. These results suggest that considering secondary effects is essential to understand the full impact of reductions in fuel subsidies and that special measures may be needed to minimize the impact on the poorest of the poor. Source: “Haiti: Steering towards effective fuel subsidy reform” (Perge et al. 2017), based on ECVMAS 2012 data and Haiti Public Expenditure Review (World Bank, 2015a). 119 ican Development Bank vehicles’ travel logs space is occupied by parked vehicles. Taken to assess the speeds of various vehicles in together, nearly half of the devoted road is no Port-au-Prince in 2010. In general, speeds longer available for traveling vehicles, leading were as low as 10.9km/hour on average. The to considerable loss of speed and risking speed of public transport is even lower because pedestrians’ well-being by forcing them off the Tap-Taps stop regularly to pick up and drop off sidewalk. The frequent stops of Tap-Taps and passengers. Tap-Tap users on average reported other minibuses with no dedicated space also an on-board travel time of 44 minutes per trip contributes to congestion. Finally, poor road with an average speed and distance of 8km/ maintenance forces vehicles to brake regularly hour and 5.9 km. If travel time and distance to avoid potholes or to absorb the shock. to the Tap-Tap route are included, as well as waiting time, the average trip duration STRUCTURE OF THE URBAN AREAS OF becomes 76 minutes and distance about 7.2km. PORT-AU-PRINCE AND CAP-HAÏTIEN: Assuming a round-trip a day, that would mean WHERE ARE THE PEOPLE? WHERE ARE that a Tap-Tap user would on average travel for THE OPPORTUNITIES? 152 minutes, i.e. above 2.5 hours.2 Cars, taxis, Traditional data needed to characterize moto-taxis, and motos travel at slightly higher the level of accessibility to opportunities in speeds, 11km/hour at least, so a similar trip to Haitian cities is unavailable or outdated. that of a Tap-Tap user would be 32 minutes In order to understand the degree of spatial instead of 44 minutes. mismatch between economic opportunities Slow speeds of motorized transport are and housing and assess the level of labor caused by a combination of lack of road market fragmentation, it is important to infrastructure, poor road maintenance, but document where economic opportunities are above all by suboptimal use of street space. in relation to residential areas. Such a task There are some narrow streets in Port-au- is usually tackled by using a combination of Prince, but many others are wide enough to up-to-date population and firm censuses. accommodate two-way traffic, so that the slow Population censuses would provide a highly speeds described above can only be partially disaggregated image of where residential areas explained by lack of road infrastructure. The are, while the economic census information other factors leading to congestion, Kopp and would provide insights about the location of Prud’homme argue, are the competing uses of jobs. Similarly, understanding travel patterns, road and street space and poor road mainte- including commuting within the urban areas nance. In particular, sidewalks are used by of Port-au-Prince and Cap-Haïtien, usually merchants for commercial purposes leaving implies the existence of personal travel no other option to pedestrians than to walk in surveys. None of these exist in Port-au-Prince the traffic lane. In parallel, a large share of road and Cap-Haïtien.3 The analysis is focused on 2 This is a conservative estimation, as we disregard any supplementary trips that Tap-Tap users might undertake during the day. 3 The last population census dates back to 2003 (IHSI 2003) and does not capture the impacts of the 2010 earthquake likely to have substantially impacted households’ locations. Similarly, although a firm census was conducted in 2012 (“Répertoire des Entreprises”: http://www.haitibusiness.com/), the coordinates of the firms are not directly available. Finally, to the best of our knowledge, no travel surveys exist for major urban areas in Haiti. 120 RESIDENTIAL POPULATION IN PORT-AU-PRINCE, A STAR-SHAPED CITY Figure 4. WITH A DOMINANT CENTER Source: Authors’ elaboration using Digicel data. Port-au-Prince and Cap-Haïtien because they used mobile operator data for develop- are the two largest urban areas of Haiti and ment of applications (Tatem et al. 2009) play major economic and cultural roles. to predict population displacement after Big data innovations involving mobile the 2010 earthquake (Lu, Bengtsson, and phone data were used to fill this knowledge Holme 2012), as well the spread of cholera gap and contribute to a better understanding in its aftermath (Bengtsson et al. 2015), this of commuting challenges in Haitian cities. is the first time CDR data is used to identify For this report, a sample of de-identified job location in Haitian cities. Several billion Call Detail Records (CDR) collected over voice call event records were included in 3-month period from March 1 to May 30, the dataset. CDRs contain the geolocation 2016 was analyzed for this work.4 This work information of the cell tower to which the was possible thanks to the partnership built subscriber is connected at the start of a voice between Flowminder and Digicel, the largest call. This information can be used to measure mobile network operator in Haiti with over the collective mobility patterns of popula- two-thirds market share in mobile phone tions over time, as well as the local popula- subscriptions in the country (CONATEL tion densities at various times of day (see 2016). While Flowminder had previously Box 2 for a summary of the methodology). 4 Data analysis was carried out by the Flowminder team, with whom the authorship of this chapter is shared. 121 BOX 2 – IDENTIFYING JOBS AND RESIDENTIAL LOCATIONS USING MOBILE PHONES The identification of jobs and residential locations follows a two-step approach: identification of “meaningful places” and categorization of such places. This approach, and the complete analysis of the CDR, was devised and carried out by Flowminder, who are the co-authors of this chapter. First, it is necessary to ascertain the locations that structure callers’ days and nights. Cell towers are ranked by the number of days a subscriber was connected to it. Since a subscriber’s location is not static, and because cell tower coverage can overlap, towers are clustered together based on their call-days and the proximity between them. A weighted average of cell tower location provides the best estimation for a subscriber’s “meaningful place.” Call-days are aggregated across clusters, and only those with a minimum of activity are considered for further analysis. Figure 5. IDENTIFYING MEANINGFUL LOCATIONS – HARTIGAN CLUSTERING ALGORITHM Towers and number of days user connected to 1. Algorithm selects tower with highest number each tower (call-days) of call-days as the first centroid 2. Algorithm scans the space for the next tower within 3. Algorithm finds new centroids and continues the threshold (1km) and re-calculates the centroid table scan 122 4. Algorithm finishes scan and finds all centroids 5. Algorithm selects only the centroids with call-days above nine as the set of meaningful places Source: Authors’ elaboration. Second, “meaningful places” are labeled as “home” or “job” locations based on their call pattern. This is done with the assumption that callers are likely to spend most of their early morning, evening, and night time at “home,” and conversely, most of their day-time at “work.” Another important difference is people’s behavior on weekends where they are less likely to work. In order to be able to identify “home” or “job” locations, Flowminder built a scoring system that attributes different scores to calls made during different times of day and days of week. On the first row, a score of -1 represents an evening time call, while a score of 1 represents a call passed during work hours. On the second row, a score of 1 represents a called passed during a work weekday and a score of -1 represents a weekend call. The overall result approximates both the location and importance of areas for jobs and residences. This novel methodology necessarily relies on a certain number of assumptions. These are discussed in detail in a background technical report that accompanies the Haiti Urbanization Review. Let us, however, list two main caveats in the analysis: the resolution for identifying meaningful places and the labeling of these places as home or job. LABELING MEANINGFUL LOCATIONS AS “HOME” OR “JOB” ACCORDING Figure 6. TO A SCORING CRITERIA Source: Authors’ elaboration. 123 The identification of clusters is dependent on the distance buffer employed to cluster close call locations together. Low distance buffers would lead to numerous clusters, each with low call-days, presenting a picture without much structure. At the other end, high distance buffers would lead to a loss of information, as short trips would not be picked up. The distance buffer retained for this study aims to be a compromise: identifying the main structure of meaningful places by limiting the number of clusters, yet still capturing the most important travel patterns. It is, however, possible that the study fails to capture the movement of people who work very close to or from their home. Second, the labeling of locations as “job” or “home” relies on the assumption that people will tend to work during daytime. The study will, therefore, miss people with unusual work patterns such as night or weekend workers. While all necessary methodological precautions have been taken to limit these cases, in partic- ular by carrying out in-depth sensitivity analyses of the results to the variation of these criteria, this study remains an approximation and could usefully be complemented by a firm and travel survey. Port-au-Prince – a star-shaped city with a Carrefour. The east of Carrefour is the most dominant city center populated part of the neighborhood with Port-au-Prince has a dominant center population densities slightly lower than the with three subcenters. There are 3.5 million center of Pétionville. To the northeast of the people living in Port-au-Prince5, and Figure National Palace, population is concentrated 4 shows their residential distribution across around Delmas. Past the airport, high relative the metropolitan area (identified from densities are observed in Croix-des-Bouquets evening calls, see Box 2). The population of to the east and along Route Nationale 1 until Port-au-Prince is scattered in a three-pointed the intersection with Route Nationale 3. star-shape with its center at the National Around this intersection, one finds Canaan Palace and the edges reaching Carrefour with around 10,000 to 15,000 people/km2 in to the west, Pétionville to the southeast, its densest part. This is one of the most recent and Canaan and Croix-des-Bouquets in the additions to Port-au-Prince, formed from northeast. The center of Port-au-Prince sees temporary camps set up post-earthquake. the highest population densities, reaching up The city center of Port-au-Prince attracts to 60,000 people/km2. The density around workers6 traveling beyond their home cluster the center, which includes neighborhoods (1km). A clearer understanding of the main such as Portail Leogane, Turgeau, and Fort day and night destinations can be gleaned National, can be above or around 50,000 from Figure 7, which focuses exclusively people/km2. Pétionville is the second most on the relative distribution of commuters populated area in Port-au-Prince, reaching at different times of the day. It shows the population densities of up to 50,000 people/ percentage of commuters with respect to km2 in its center. To the west of the National the total local population during daytime Palace high population density is observed (left) and nighttime (right). Unsurprisingly, along Route Nationale 2, which leads to downtown Port-au-Prince appears as a prime 5 We are looking at a rectangular surface given by a set of coordinates, so when we speak of the metropolitan region, it might include a larger area than the administrative boundaries. 6 Commuters in this study are defined as people that have distinct home and work clusters, which means that work is situated at a distance greater than the 1km threshold used as a buffer to limit the number of meaningful call locations in the sample of callers. 124 COMMUTERS DURING DAYTIME FOR WORK-RELATED ACTIVITIES Figure 7. (LEFT) AND DURING THE EVENING (RIGHT) Source: Authors’ elaboration using Digicel data. CUMULATIVE DISTRIBUTION OF POPULATION AS A FUNCTION OF DISTANCE TO CITY CENTER IN PORT-AU-PRINCE FOR VARIOUS TIMES OF THE DAY (DAYTIME/EVENING) AND FOR VARIOUS CATEGORIES OF Figure 8. POPULATION (COMMUTERS/NON-COMMUTERS) Source: Author’s own elaboration using DIGICEL data. 125 attractor during daytime with commuters shown in Figure 9. The highest evening to the downtown area representing up population density can be found around La to 72 percent of the local population. In Fossette, a popular low- to middle-income the evening, the trend is reversed when residential neighborhood to the west of the commuters to downtown represent only Mapou River and close to a bridge linking 40 percent of the local population. In the both sides of town. Densities in these regions residential areas of Carrefour and Canaan, can reach about 40,000 people/km2. On the we see an opposite trend, with low shares of other side of the bay, densities are highest commuters going there during the day, but around Petite Anse where they oscillate many returning in the evening after work. around 20,000 people/km2. Population Another interesting finding from Figure 7 densities decrease as one travels south along is that we see numerous commuters repre- the Mapou River. Cap-Haïtien is not as big senting large shares of local population as Port-au-Prince, and population density is alongside Route Nationale 1 and 8 leading, high only within a narrow 2km strip west of respectively, to Canaan and Croix-des- the Mapou River and 6km from the mouth of Bouquets, indicating that these might host the river until Haut-du-Cap. Outside of this commercial activities during daytime. region, density drops dramatically, reaching The overall picture is one of concentra- 500 people/km2. Outside of Cap-Haïtien, tion toward the city center during daytime population is relatively higher along Route and, inversely, one of diffusion toward the Nationale 1 from Vaudreuil to Moustique to outskirts during the evening. Figure 8 shows the southwest, in Quartier Morin, Limonade, that total population within 5km of the and Trou-du-Nord to the southeast, and in city center is about 5 percent higher during Milot to the south. Arguably, most of those daytime than in the evening. If the focus is regions are not part of the metropolitan exclusively on commuters, then the picture area of Cap-Haïtien. Population density is even more striking, with 46 percent of in those areas fluctuates between 500 and commuters within 5km of the city center 1,000 people/km2 and is particularly higher during daytime versus less than 37 percent in in Vaudreuil and Trou-du-Nord. Outside of the evening. those satellite regions, density drops dramat- ically, reaching below 500 people/km2. Cap-Haïtien – the center has it all The business district in the center of Population in Cap-Haïtien is mainly Cap-Haïtien strongly dominates any other concentrated in a strip along the Mapou daytime destination. During the day, River. Cap-Haïtien hosts approximatively commuters can represent nearly 70 percent 500,000 people.7 The vast majority of people of the local population in the city center of live closer to the center of town on the west Cap-Haïtien (Figure 10). This figure drops to side of the bay and Mapou River and along just a quarter of the total population during the south bay, east of the Mapou River as the evening, meaning that most people in 7 As in Port-au-Prince, we are looking at a rectangular surface given by a set of coordinates and thus when we speak of the metropolitan region it might include a larger area than the administrative boundaries. 126 RESIDENTIAL POPULATION DISTRIBUTION IN CAP-HAÏTIEN, A Figure 9. DOMINANT CENTER Source: Authors’ elaboration using Digicel data. the city center are there for work or other Cap-Haïtien is a small and very concen- purposes during the day and at night return trated urban area with a highly dominant to their homes, located outside the city center. Figure 11 shows that close to 60 percent center. The other densely populated area of of its population is concentrated within 5km Cap-Haïtien, Petite Anse, sees an opposite of the city center. The attraction exerted by trend with incoming commuters in the the city center is powerfully illustrated by morning representing 30 percent of the local the fact that during the day 40 percent of all daytime population, but outgoing commuters commuters can be found within 1km of the being close to 40 percent. city center and nearly 80 percent within 5km. 127 Figure 10. COMMUTERS DURING DAYTIME FOR WORK-RELATED ACTIVITIES (LEFT) AND DURING THE EVENING (RIGHT) Source: Authors’ elaboration using Digicel data. CUMULATIVE DISTRIBUTION OF POPULATION AS A FUNCTION OF DISTANCE TO CITY CENTER IN PORT-AU-PRINCE FOR VARIOUS TIMES Figure 11. OF THE DAY (DAYTIME/EVENING) AND FOR VARIOUS CATEGORIES OF POPULATION (COMMUTERS/NON-COMMUTERS) Source: Author’s own elaboration using DIGICEL data. 128 FRAGMENTED LABOR MARKETS – FEW commutes are not negative per se – actually, COMMUTERS NOT TRAVELLING FAR they have many benefits such as low fuel Only a small share of people travel to consumption – if they correspond to self-selec- work. Only 42 percent and 40 percent of the tion processes whereby people settle near the population are considered to be commuters job they want, they probably reflect the diffi- in Port-au-Prince and Cap-Haïtien, respec- culty and unaffordability of travelling. In this tively, meaning that they travel beyond their sense, generalized short commutes are indic- home cluster (1km radius). The remaining 58 ative of local, rather than integrated, labor percent and 60 percent either work from home markets in the two urban areas. Fragmented 8 or in its immediate vicinity or do not work. labor markets are unlikely to act as match- Ideally, this figure would be compared to the makers, decreasing the probability of effec- distribution of population in Port-au-Prince tively pairing employers and employees. and Cap-Haïtien by age group. In the absence Central residents tend to commute less of this information, we can compare to the than people living farther from the city center. average at the country level. Given that 64 As numerous opportunities are available very percent of the Haitian population is of working locally, there is less of an incentive to travel age (Singh and Barton-Dock 2015), commuters far from home. Commuters that live the 9 in the two urban areas represent a maximum farthest from the city center (beyond 5km) of two-thirds of the active population. in Cap-Haïtien and Port-au-Prince have the The distances travelled by commuters in longest commutes, as they are more isolated Port-au-Prince and Cap-Haïtien are short, from economic opportunities and have to which is indicative of local matching. In incur longer trips to reach them. However, Port-au-Prince, it is estimated that the median the increase in the median distance travelled trip is 1.1km and 3.1km if only commuters are remains limited to 1.5km and 2.5 km compared considered. In Cap-Haïtien, the corresponding to the most central locations in Port-au-Prince statistics are slightly higher at 1.6km and and Cap-Haïtien, respectively. This pattern 3.3km, respectively (Table 1). In both urban is again indicates fragmented labor markets; areas, these figures are low. For example, even although job densities in the vicinity are much focusing on commuters only, these median lower, travel distances do not increase in trips represent less than an hour of walking. commensurate proportions. Consequently, the access to a large array of A non-negligible share of commuters travels economic opportunities is low (Tables 2 and long distances. Figure 12 shows the distribu- 3). By comparison, commuter trips in Buenos tion of distances travelled by commuters that Aires, Argentina are between 7.5km and 10km, live (left column) and work (right column) depending on the gender of the commuter in each distance from the city center buffer and the existence of children (Peralta Quirós, in Port-au-Prince. What is striking is that Mehndiratta, and Ochoa 2014). While short although there are high concentrations 8 Or have other very local regular activities such as attending school, for example. 9 It is possible that part of the people identified as commuters are, in fact, going to school rather than to work or a center of opportunity. This would bring the share of commuters down further. 129 Table 1. TRAVEL BEHAVIORS IN PORT-AU-PRINCE AND CAP-HAÏTIEN: FEW COMMUTERS NOT TRAVELLING FAR PORT-AU-PRINCE CAP-HAÏTIEN Total Population 3.5 million 0.508 million ALL COMMUTERS ALL COMMUTERS Mean Trip 2.5 km 4.5 km 2.8 km 4.7 km Median Trip 1.1 km 3.1 tkm 1.6 km 3.3 km NON-COMMUTERS COMMUTERS NON-COMMUTERS COMMUTERS As percentage 58.14% 41.86% 60.51% 39.49% of total Live less than 1km 4.13% 2.86% 12.02% 6.13% from the center Mean Trip — 3.37 km — 3.73 km Median Trip — 2.11 km — 2.52 km Live less than1km 4.13% 31.67% 46.87% 56.81% from the center Mean Trip — 3.29 km — 3.09 km Median Trip — 2.52 km — 2.55 km Live less than1km 61.67% 65.27% 40.94% 36.98% from the center Mean Trip — 5.12 km — 7.24 km Median Trip — 3.66 km — 5.84 km Source: Author’s own elaboration using DIGICEL data. 130 ACCESSIBILITY TO OPPORTUNITIES IN PORT-AU-PRINCE FOR VARIOUS TRAVEL TIME THRESHOLDS IN PORT-AU-PRINCE Table 2. SHARE ACCESSIBLE DEPENDING ON TRANSPORT MODE USED CARS TAP-TAP + WALKING WALKING ONLY < 30 min 24% 7% 3% < 45 min 43% 16% 7% < 60 min 61% 27% 12% Source: Authors’ elaboration using Digicel data. of commuters that travel short distances, can access respectively 24, 43, and 61 percent there are also fat tails indicating that some of total opportunities10 in the urban area commuters incur long trips, up to 20km. within a 30, 45, and 60 minutes in congested Finally, Figure 12 also reaffirms that the city conditions. For Tap-Tap users the situation center is a main attractor, as there are approx- is drastically different, as on average they imately 4.5 commuters travelling to the city can only access 7, 16, and 27 percent within center for each commuter that leaves the city a 30, 45, and 60 minutes. These “accessi- center for work purposes. We find similar bility to opportunities” figures for collective patterns in Cap-Haïtien (Figure 13), with an transport are low and likely optimistic, as even larger share of people commuting long they do not account for waiting times.11 They distances to the city center, particularly from are in line with figures for Nairobi (Avner and Trou-du-Nord. Lall 2016), which is notoriously congested, having ranked as one of the cities where the Low accessibility limits the potential for journey to work is the lengthiest (IBM 2011). matchmaking By comparison, in the metropolitan area of Accessibility, measured by the share of Buenos Aires, an urban area that has four opportunities that can be reached within a times more population, accessibility figures given timeframe, is low in Port-au-Prince. using public transportation are 7, 18, and 34 On average, car users within Port-au-Prince percent for the same time thresholds (Peralta 10 Our assumption is that “opportunities” mainly correspond to jobs or occupations capable of generating income, although opportunities could also correspond in a limited number of cases to repeated social gathering and other non-job related activ- ities. The term “job” should be interpreted widely as an activity that takes place regularly in a specific location. Opportunities are identified, as described in Box 2 of this chapter, as “locations where people frequently spend time during the day.” 11 Waiting times could add on average 12.5 minutes to a Tap-Tap trip based on figures from Kopp and Prud’homme (2011). 131 Figure O.1. Figure 12. DISTRIBUTION OF DISTANCES TRAVELLED FOR EACH DISTANCE TO THE CITY CENTER BUFFER IN PORT-AU-PRINCE Notes: Home to work trips are represented on the left (people living in the buffer) while work to home trips are to the right (people working in the buffer). The first row shows results for locations within 1km from city center, the second row for locations between 1km and 5km from the center and the last row for locations beyond 5km and within 25 km from the center. The red shade indicates trips that originate and end in the same buffer while the blue shade indicates that the commute spans over two or more distance buffers. Source: Authors’ elaboration using Digicel data. 132 DISTRIBUTION OF DISTANCES TRAVELLED FOR EACH DISTANCE TO Figure 13. THE CITY CENTER BUFFER IN CAP-HAÏTIEN Note: Home to work trips are represented on the left (people living in the buffer) while work to home trips are to the right (people working in the buffer). The first row shows results for locations within 1km from city center, the second row for locations between 1km and 5km from the center and the last row for locations beyond 5km and within 25 km from the center. The red shade indicates trips that originate and end in the same buffer while the blue shade indicates that the commute spans over two or more distance buffers. Source: Authors’ elaboration using Digicel data. 133 Figure 14. SHARE OF ACCESSIBLE OPPORTUNITIES IN PORT-AU-PRINCE WITHIN ONE HOUR BY CAR (LEFT PANEL) AND BY TAP-TAP (RIGHT PANEL) Source: Authors’ elaboration using Digicel data. Quirós 2015). And in Greater Dakar, an urban ties and live in an area where the road and the area roughly equivalent in size to Port-au- Tap-Tap network are the densest. Conversely, Prince, with a population above 3 million accessibility sharply drops as one moves away (World Bank 2016), the share of accessible from the city center, albeit more gradually jobs within one hour is 52 percent, nearly for car users whose vehicles can compensate twice the level in Port-au-Prince (Stokenberga longer distances with higher speeds. 2017). Pedestrians in Port-au-Prince have While we do not have the Tap-Tap network lower accessibility levels with a maximum in Cap-Haïtien, so cannot compare for this of 12 percent of opportunities that can be transport mode, accessibility is higher for car accessed within one hour. As discussed in the users and pedestrians than in Port-au-Prince. previous sections, given that a large share of A smaller urban footprint and lower popula- the population cannot afford to commute by tion means that it is easier to commute to Tap-Tap, accessibility levels will remain low, jobs and other opportunities, even the most closer to the levels displayed for pedestrians. distant ones, including by walking (Table 3). Central residents of the Port-au-Prince The share of opportunities that can be urban area have better access to opportuni- accessed within thirty minutes by car is nearly ties. Figure 14 shows the spatial distribution twice as high in Cap-Haïtien (42 percent) of accessibility to opportunities for people than in Port-au-Prince (24 percent). This commuting by car and by Tap-Tap. Unsur- can be explained easily by a smaller urban prisingly, central residents benefit from footprint in Cap-Haïtien, which mechani- higher accessibility levels, as they both are cally reduces distances to jobs, markets, and physically closer to a large share of opportuni- other opportunities (Figure 15). 134 ACCESSIBILITY TO OPPORTUNITIES FOR VARIOUS TRAVEL TIME THRESHOLDS AND MODES ARE LOW IN CAP-HAÏTIEN Table 3. SHARE OF CAP-HAÏTIEN’S OPPORTUNITIES ACCESSIBLE WITHIN A GIVEN TIME-FRAME DEPENDING ON THE TRANSPORT MODE USED CARS WALKING ONLY < 30 min 42% 8% < 45 min 52% 12% < 60 min 63% 18% Source: Authors’ elaboration using Digicel data. SHARE OF ACCESSIBLE OPPORTUNITIES IN CAP-HAÏTIEN WITHIN Figure 15. ONE HOUR BY CAR Source: Authors’ elaboration using Digicel data. 135 Surprisingly, however, despite the smaller roads beyond the core center of Cap-Haïtien, urban footprint of Cap-Haïtien, the accessi- as well as the low intensity of land develop- bility within one hour is similar to Port-au- ment, which means that opportunities are Prince (63 vs. 61 percent). This can be sparse beyond the urban core as seen in explained by the decrease in space devoted to Chapter 2. Table 4. IMPACTS OF INCREASING SPEEDS ON THE TRANSPORT NETWORKS IN PORT-AU-PRINCE, COUNTERFACTUAL SCENARIOS ACCESSIBILITY IN PORT-AU-PRINCE FOR VARIOUS SPEEDS ON THE NETWORK WITHIN ONE HOUR CARS TAP-TAP Baseline: 50% of official speeds 61% 27% 40% of official speeds 48% 21% 60% of official speeds 71% 33% 70% of official speeds min 80% 40% 80% of official speeds 85% 45% Official speeds 91% 55% Source: Authors’ elaboration using CNIGS data on road and Tap-Tap networks and regulatory speeds. TRANSPORT INTERVENTIONS THAT FOCUS In Cap-Haïtien we find similarly strong ON SPEED AND RESILIENCE CAN REDUCE impacts of increasing speeds. We display the SPATIAL MISMATCH results for car users. Improving speeds on the network Increasing speeds in the urban areas of Increasing speeds in Port-au-Prince is Port-au-Prince could help cities become one way of achieving higher accessibility, matchmakers to boost economic growth. especially for Tap-Taps. If current speeds There are many options to do so. The most could double and reach the regulatory speeds obvious one would be to invest heavily in of the network, then average accessible oppor- roads and public transport. However, such an tunities using the Tap-Tap network would option would require large financial resources double and reach 55 percent. and is unlikely to be very effective before the 136 VISUAL DEPICTION OF INCREASING THE SPEEDS ON THE NETWORK Figure 16. Figure O.1. FOR TAP-TAP ACCESSIBILITY IN PORT-AU-PRINCE (A) Baseline: 50% of official speeds (B) Official speeds (C) 70% of official speeds (D) 80% of official speeds Source: Authors’ elaboration using CNIGS data on road and Tap-Tap networks and regulatory speeds. 137 Table 5. IMPACTS OF INCREASING SPEEDS ON THE TRANSPORT NETWORKS IN CAP-HAÏTEN, COUNTERFACTUAL SCENARIOS ACCESSIBILITY IN PORT-AU-PRINCE FOR VARIOUS SPEEDS ON THE NETWORK WITHIN ONE HOUR CARS Baseline: 50% of official speeds 63% 40% of official speeds 55% 60% of official speeds 72% 70% of official speeds 78% 80% of official speeds 83% Official speeds 90% Source: Authors’ elaboration using CNIGS data on road and Tap-Tap networks and regulatory speeds. chronic challenges on the current network reach, but in central areas of Port-au-Prince it are addressed. A more immediate solution often only means going from 5km/hour (barely consists on first improving the operation and quicker than walking) to around 10km/hour. maintenance of the current network. Both options are not mutually exclusive, but the Road network vulnerability to natural most effective sequencing would be to begin hazards: accessibility implications improving the existing network. Road space Natural hazards have both direct and management and road maintenance can go indirect costs. They have obvious disastrous a long way to increasing speeds, and they are humanitarian consequences such as loss both achievable. Road space management of life, injuries, and grief. But by damaging implies ensuring that roads are used for circu- infrastructure such as housing, roads, and lation, rather than to accommodate on-street electricity production facilities, they also parking or pedestrian movement. In turn, this impose high direct costs of repairs and means securing the sidewalk for pedestrians rebuilding. Finally, the disruption to the so that they do not have to step on and off the network infrastructure, typically roads and sidewalk. Road maintenance could also pay electric transmission lines, also imposes high high dividends as drivers would no longer indirect costs to society through ripple effects, have to slow to a near halt to avoid a pothole, where people are no longer able to reach their for example. Doubling speeds can seem out of jobs or by disrupting supply chains. 138 VISUAL DEPICTION OF INCREASING THE SPEEDS ON THE NETWORK Figure 17. Figure O.1. FOR TAP-TAP ACCESSIBILITY IN CAP-HAÏTIEN (A) Baseline: 50% of official speeds (B) Official speeds (C) 70% of official speeds (D) 80% of official speeds Source: Authors’ elaboration using CNIGS data on road and Tap-Tap networks and regulatory speeds. 139 Figure 18. ROAD NETWORK IN PORT-AU-PRINCE AND IDENTIFICATION OF MOST CRITICAL ROAD LINKS Source: Authors’ elaboration using CNIGS data on road and Tap-Tap networks. Identifying the most critical links in the as a consequence of hurricanes and river road network is the first step to address these flooding (World Bank 2010). These repeat- potential vulnerabilities and build resilience edly damage the road network, entailing in the face of disruption. As mentioned longer commuting times at best and, in some in Chapter 2, Haiti is highly vulnerable to instances, the disconnection of entire areas disasters; over the course of the last century of the country, particularly urban. Under- up to 2016, a minimum of 100 hazard events standing which road sections are the most took place, according to the Emergency critical links in the road network is a first step Events Database (Guha-Sapir, Below, and in building a strategy to increase resilience Hoyois 2017). Geophysical, hydrological, to these low-key yet costly disruptions. The and meteorological events alone affected a analysis presented below focuses on Port-au- minimum of 11.7 million people directly or Prince, because of its importance in terms of indirectly. While the 2010 earthquake was population, wealth production, and exposure by far the deadliest and most disruptive, to floods, but could be extended to Cap-Haï- Haiti also experienced lower magnitude tien and other frequently exposed urban yet recurring flood episodes, frequently areas in the future. 140 TWO EXAMPLES OF THE LOSS OF ACCESSIBILITY WITHIN AN HOUR COMPARED TO THE BASELINE TRIGGERED BY THE DISRUPTION OF ROAD LINKS TOWARD CARREFOUR AND CANAAN. THE IMAGE SHOWS CONSIDERABLE LOCALIZED LOSSES IN ACCESSIBILITY FROM DAMAGE Figure 19. TO SPECIFIC ROAD LINKS. Source: Authors’ elaboration using CNIGS data on road and Tap-Tap networks. 141 The most critical links in the road network hazards, such as floods, by implementing of Port-au-Prince are (1) the segments of measures that would make them less the RN2 (Route Nationale 2) that connect vulnerable. This could involve, for example, downtown to Carrefour and beyond to the elevating a road where floods are most likely west, (2) the RN1 going from the downtown to occur. The second primary option is to to the north of Port-au-Prince, (3) an isolated invest in redundancy. By building alternate link that connects downtown to Petionville, routes, the impact of a disruption to a specific and (4) a couple of links that join Canaan link would remain small, as residents, to the rest of the network. The most critical workers, and enterprises would have an links, in red in Figure 18, are identified based alternative option to bypass the bottleneck. on the impact that their removal has on Deciding which of these options makes most average accessibility within the urban area sense should be done on a case-by-case basis of Port-au-Prince.12 When a specific link is and goes beyond the scope of the paper. The disrupted, it forces users to choose an alter- analysis presented here, however, provides a native route which will lengthen the trip and first step to investigate further. potentially lead a number of jobs to become out of reach within specific timeframes. The MATCHING PEOPLE AND JOBS – OVER- accessibility impacts of the road network COMING THE CHALLENGES disruption ignore second-order effects such The previous sections have highlighted as increased congestion on alternative roads. that the two main urban areas in Haiti While average loss of accessibility suffer from low employment accessibility, remains moderate when particular road which hampers the potential for growth links are disrupted, specific urban areas and inclusion. The spatial distribution of can be strongly impacted, losing access to opportunities (mainly jobs) resembles that up to 80 percent of opportunities.13 This is of homes, meaning that there is limited typically the case in Canaan and Carrefour, clustering of economic activity occurring. for example, as shown by the maps in Figure This outcome is both a consequence of low 19. The loss in accessibility to the west of accessibility and constrained mobility and a Carrefour is much more severe, as indicated cause for fragmented and local labor markets by the dark brown shade; however, it is less unlikely to benefit from economies of concen- widespread and affects many less people tration. Of course, increasing accessibility is than in Canaan. only one step in creating jobs and achieving The identification of the most critical efficient labor markets, but it is a key part of links in the network is a first step to build the solution. In parallel, the high exposure resilience into road infrastructure. There are of Haiti to natural hazards imposes the need two main options to do so. The first involves to think about vulnerabilities and the resil- better protecting the road links from natural ience of the road network. There are three 12 The criticality analysis has been carried out without the use of a traffic assignment model. It therefore cannot account for increases in congestion levels, and thus speed reductions, on alternate links when one route becomes unavailable. In consequence, the figures displayed here represent a lower bound of the impacts of road disruptions; real impacts are likely to be higher. 13 See appendix for more information on the criticality analysis. 142 broad avenues for improvement: increasing the pavement. It should be noted, however, travel speeds and improving comfort through that simply banning street vendors from more investments and improved efficiency in sidewalks would carry potential negative Port-au-Prince and Cap-Haïtien; increasing social impacts and should only be done if the affordability of collective transport for accompanied by a well-designed strategy inclusive matching (including through travel to offer alternative options and transitional speeds increases); strengthening coordina- support for street vendors. An experience tion of land use and transport investments in Panama City is of interest here, as close for improved access and increased resilience. to 300 street vendors were relocated to an We will discuss these three broad priorities in underutilized property, de facto creating detail below, laying out the possibilities for what became a popular market. Progress improvement and drawing, where available, was also made in Delhi, India to safeguard from successful examples around the world. pedestrians while providing vendors, an integral part of urban streets in Delhi, a Increase speeds and improve comfort separate space for commercial activities through more investments and enhanced (WHO 2009). The solution was to identify efficiency the road segments with the most fatalities One area for improvement is investing and to create a protected pedestrian lane. in and better managing sidewalk space for Sidewalk investments and management will pedestrians. As argued in this chapter, a not only speed the trips of pedestrians, but large share of the population in the urban equally those of cars and Tap-Taps that will areas of Port-au-Prince and Cap-Haïtien are no longer have to stop or swerve to avoid pedestrians. It is very important to make sure pedestrians. that their travel conditions are as good and More efficient use of road space can go a safe as possible, as approximately a quarter long way in reducing travel times. Evidence of road fatalities worldwide are pedestrians on Port-au-Prince suggests that lack of urban (WHO 2015). Ensuring comfort and security road space is not the current main cause of of those commuting by foot should be a congestion and low motorized travel speeds priority, particularly where sidewalks do (Kopp and Prud’homme 2011). Rather, it is not exist in urban areas. Where sidewalks the sub-optimal use of road space. Lack of exist in Port-au-Prince, evidence suggests dedicated parking space results in stationed that they are used mainly for commercial vehicles occupying one of two lanes where purposes by street vendors who occupy two-lane roads exist, both ways. As street most of the available space and frequently vendors force pedestrians off the sidewalks, force pedestrians onto the pavement where these can be pushed farther onto the street they are in danger of getting hit by a vehicle by stationary vehicles. Besides the risks, this (Kopp and Prud’homme 2011). Besides the also means that a sizeable portion of street risks associated with sharing road space with space is not used by traffic. Finally, the lack vehicles, competing uses of sidewalks can of dedicated stops where Tap-Taps could also lengthen trips made on foot by forcing pull over to pick up and drop off customers pedestrians to constantly step off and on slows traffic considerably. Identifying areas 143 on main streets with enough space for a Tap-Tap fares, rendering motorized transport Tap-Tap stop to be added would be a low-cost further out of reach for a large share of the option to relieve traffic congestion. Similarly, urban population. There is thus a need to where possible, it would be important to think of mechanisms that could offset rises identify space that could be used for parking, in Tap-Tap fares or even reduce them. We will therefore freeing up a large chunk of road explore below a number of avenues for doing space. To be effective, parking enforcement so while acknowledging that more careful will be key. analysis is needed, in particular by better Road maintenance can save both vehicle understanding the economics of Tap-Tap maintenance costs and travel time. Potholes operations. A World Bank study tackling and uneven road surfaces force vehicles these questions is currently underway. What to slow down. They also damage vehicles, most of the explored options have in common increasing repair and maintenance costs. is that they seek to reduce costs for Tap-Tap Regular road maintenance could reduce operators, crucial actors of urban mobility, both these effects. A road maintenance fund likely resulting in lower transport fares. exists and should be properly supported. Improved transport speeds, already In the longer term, collective transport discussed in the previous section, would lanes could be a promising approach to allow Tap-Tap drivers to increase the reduce travel times in urban areas. Tap-Taps number of trips that they can make within mainly run on roads that have two lanes in a given timeframe. This would also result in each direction. If road space can be freed up more paying customers and higher profits. for traffic as argued above, and given current Therefore, all options that can reduce travel low motorization rates, it would be sensible times through road space management to reserve space for collective transport. would participate in lowering operational costs. It cannot be emphasized enough: Increase affordability of collective transport increasing travel speeds will enhance both for inclusive matching accessibility and affordability of motorized Affordability of collective transport is a key transport. constraint for urban areas in Haiti to act as Boosting the energy efficiency of matchmakers, better connecting people with Tap-Taps, while likely increasing the initial opportunities. While a pedestrian in Port-au- capital investment in the vehicle, would Prince can access on average 12 percent of reduce the cost burden of fuel and thus the opportunities within one hour, a Tap-Tap cost of operating the vehicle. Options to do so rider can access 27 percent. Yet affordability involve either the purchase of higher capacity of collective transport is further threatened vehicles, which would have a lower fuel by the need to eliminate fuel subsidies, consumption by passenger, or the purchase which are both unsustainable from a fiscal of more fuel-efficient vehicles of similar and macroeconomic perspective and highly size. Both these options appear reasonable regressive, overwhelmingly benefiting the and warrant further investigation. The hilly richest households. The removal of fossil fuel landscape of some Haitian cities (in partic- subsidies, however, will likely result in higher ular Port-au-Prince), as well as the existence 144 of narrow roads, deserve careful thought In the longer term, carefully targeted when advising for the purchase of larger and transport subsidies could be directed toward less maneuverable vehicles. As for incentiv- the poorest households in urban areas in izing the purchase and operation of higher Haiti. These targeted demand-side subsidies fuel efficiency vehicles, experiences in other are widely recognized as the most efficient countries can provide some guidance based mechanisms to ensure that the poorest get or on lessons learned. The authorities in the retain access to opportunities without exces- urban area of Dakar, for example, created sively burdening the local or national budget incentives for informal minibus operators to (Estupiñán et al. 2007). Such subsidies, purchase more fuel-efficient vehicles. From however, require detailed registries that aptly 2003 to 2008, operators were provided a subsi- capture households’ or individuals’ socio-eco- dized loan for the purchase of more fuel-ef- nomic conditions. While these seem out of ficient minibuses, which covered nearly 75 reach in Haiti at present, in the longer term, percent of the purchase cost of the vehicle they could be an interesting option. (Kumar and Diou 2010). In exchange, car rapides14 owners had to retire their previous Strengthen coordination of land use and vehicles, become a member of a transporta- transport investments for better access and tion association (thereby formalizing their resilience activity), and accept to run a specific route There are two main ways of improving for a fixed set of tariffs. The success of the accessibility to opportunities in urban areas: operation is still debated; while evidence increasing speeds and reducing distance. suggests fares were reduced, usage appears These two opposite options are famously illus- to remain low (Kumar and Diou 2010). In trated by Atlanta, on one hand, and Barcelona, particular, the assessment of this project on the other (Bertaud 2002). The first one is to concludes that leasing mechanisms can be increase travel speeds, primarily by investing effective in replacing aging public transport in the connective network, but also by making fleets, but that their success depends on motorized transport more affordable so that “operator inputs at the design stage, technical a larger proportion of the population can assistance to professionalize operators and travel at higher speeds. The second option drivers, and restructuring of the network of is to reduce the distance between homes informal transport operators.” This option and opportunities. This entails reducing the would require more in-depth analysis to fragmentation of the urban footprint, incen- uncover whether a suitable model can be tivizing density of people and opportunities identified and negotiated with the operators and better integrating land use and transport. in Haiti. An ongoing technical assistance of Population densities in Port-au-Prince and the World Bank is exploring different mecha- Cap-Haïtien are high (up to 90,000 people/ nisms to offset fuel cost increases, and the km2 in Port-au-Prince) suggesting that the results will inform whether scrapping old potential to increase accessibility by reducing Tap-Tap vehicles is a viable option. distance is limited. Chapter 2 on planning 14 Informal and colorful paratransit vehicles in Dakar. 145 actually suggests that these urban areas are vulnerable to river floods as it is situated on a crowded, in the sense that higher densities slope. However, it is situated at a long distance have not been supported with adequate infra- from the center of Port-au-Prince and is structure. In fact, urban growth has continued connected by few roads, limiting accessibility at the same pace in central Port-au-Prince to economic opportunities (as can be seen in as in the more peripheral area of Croix-des- Figure 18). In Cap-Haïtien, as documented in Bouquets, indicating continued increases the planning chapter, urban growth has taken in population in an area that benefits from place close to the city center and, thus, close highest accessibility. This pattern contrasts to the economic opportunities, but at the cost to that observed in many other major cities of high risk, as urbanization between 2005 around the world where more rapid growth and 2015 encroached on the riverbed by 21 occurs in the outskirts of urban cores (Angel percent. Careful planning should avoid these et al. 2011). In the short term, therefore, trade-offs and incentivize settlements that there is only limited scope to reduce distance are in areas relatively safe from multiple risks between people and economic opportunities (or can be protected more easily) and benefit by increasing densities further. To be able to from proximity to economic opportunities, do so without worsening living conditions either through physical proximity (infill rather would require the construction of taller multi- than leapfrog urbanization) or through access family buildings. In the medium term, such a to transportation options. solution should be encouraged, but it requires Planning itself is not sufficient; imple- the emergence of a skilled construction sector mentation is key. The planning chapter has able to enforce building codes to help units documented that informal and uncoordi- withstand seismic hazards. nated urban growth has happened despite a There is, however, considerable scope to wealth of national, sectoral, and local plans promote accessibility by planning for urban for both Port-au-Prince and Cap-Haïtien. expansion, while simultaneously reducing Three recommendations appear to be relevant exposure to natural hazards. Planning for here, too. First, increase coordination and urban expansion should have a double focus cooperation across implementing agencies in Haitian urban areas: limit the exposure of to maximize the chances of success of these people to natural hazards and promote acces- plans. Second, set achievable goals which sibility to economic opportunities. Examples are more likely to be taken seriously and less in Port-au-Prince and Cap-Haïtien show that likely to trigger unintended consequences.15 urban growth can achieve one of these, but Third, “plan for the best, but prepare for the not both simultaneously. In Port-au-Prince, worst.” Tunis, Tunisia provided clear and recent development north of Croix-des-Bou- transparent information on where future quets, in Canaan, came in the aftermath of infrastructure would be laid out. This did not the 2010 earthquake. It appears to be less prevent unplanned urbanization from taking 15 A recent World Bank report describes the consequences of ill-adapted land use regulations to a country’s development state. In some cases well-intended regulations such as minimum lot size, which aim at providing a basic living space per person, can be unaffordable for residents and force them into informality (Lall, Henderson, and Venables 2017). 146 place, but it had two advantages: it effectively main options. The first is to retrofit/improve guided the new informal settlements into the road links to increase their resilience to areas where the government had planned certain types of disasters. This could involve expansion, and it incentivized households to improving the drainage system of the road or leave the rights of way that would be servicing elevating the road. The second option is to these areas in the future (with roads, sewer invest in network redundancy. This option systems, etc.). By doing so, the government can be more favorable when the cost of reduced construction costs as retrofitting is improving the resilience of the road segment typically more expensive. is high or when it is difficult to assess the risks. Creating functional land markets is a notori- Considering that risk is a function of ously arduous task. Yet it is essential if land is to physical vulnerability and level of hazard be used at its highest value. Throughout cities exposure, improving infrastructure and across the globe, high-value land typically promoting urban development in less-ex- promotes density, as it supports accessibility posed areas are complementary measures that to economic opportunities through reduced contribute to building resilience and reducing distances. The planning chapter has provided disaster risk. The rehabilitation or protection some ways forward. They do not always involve of critical infrastructure such as roads, as a formalizing land titles, but they should allow corrective measure, will decrease the vulner- for their transferability. ability of the infrastructure, thus reducing In a resource–constrained country such as disaster risk. Proactively guiding urban devel- Haiti, there is an even greater need to make opment and new transport investments in the most efficient use of funds devoted to road areas that are less exposed to natural hazards maintenance/improvement. Given the high through land use planning and the use of risk exposure of the country to natural hazards, information, as a preventive measure, will attention should be paid to building the also reduce the overall risk. resilience of the network. Indeed, recurrent However, improving or protecting infra- localized natural hazard events (such as structure as a risk-reduction measure can be floods) can lead to the isolation of neighbor- extremely expensive for local and national hoods, which impacts commuting patterns governments. Large scale, city-wide infra- and disrupts economic value chains. One structure investments for flood protection or method for identifying the most urgent measures to make roads, ports, and power areas of intervention is through a criti- generation facilities more resilient to extreme cality analysis. This chapter has provided a events may be necessary in many cities in roadmap for road work/maintenance prioriti- the short term, but they are expensive. For zation. While incomplete, as it does not build example, the cost of protecting the 100km in congestion costs, it still identifies the most coastline of Dar es Salaam with a sea wall critical road segments. would be USD 270 million. Therefore, it While the criticality analysis presented is imperative to develop accountable and in this chapter identifies priorities, the type responsive governance systems that can of intervention needed will warrant a more build long-term resilience through capacity in-depth economic analysis. There are two building, and land use planning. 147 The Can-Tho Urban Development and Urban resilience Project in Vietnam combines both approaches to reduce the risk of flooding and build long-term resilience. The project aims to improve flood risk management and environmental sanitation through priority flood control infrastructure investments in the urban core, including surrounding the embankment, building tidal gates/valves, and improving the drainage system. It also promotes the development of new low-risk urban growth areas by enhancing connectivity with the city center through the construction of transport investments on higher grounds. The new and improved transport infra- structure aims to increase accessibility and connectivity, land values, and investment opportunities. Project activities are supported by the development of management systems to improve spatial planning, flood risk management, and transport, including a web-based geospatial database that serves as a single platform for spatial data and is intended to be used across departments for purposes of spatial planning and infrastruc- ture development. These lessons can be applied to a Haitian context. 148 REFERENCES Franklin, S. 2015. “Location, Search Costs and Youth Unemployment: The Impact of a Randomized Transport Subsidy in Urban Angel, S., J. Parent, D. L. Civco, and A. M. Blei. Ethiopia.” CSAE Working Paper WPS/2015-11. 2011. “Making Room for a Planet of Cities.” Guha-Sapir, D., R. Below, and Ph. Hoyois. 2017. Policy Focus Report, Lincoln Institute of Land “EM-DAT: Emergency Events Database.” Policy. Brussels, Belgium: Université Catholique de Antos, S.E., N. Lozano-Gracia, and S.V. Lall. 2016. Louvain. http://www.emdat.be/. “The Morphology of African Cities.” World IBM. 2011. “Global Commuter Pain Survey: Bank Policy Research Working Paper, no. 7911. Traffic Congestion Down, Pain Way Up.” Avner, P., and S.V. Lall. 2016. “Matchmaking in http://www-03.ibm.com/press/ us/en/pressre- Nairobi: The Role of Land Use.” World Bank lease/35359.wss. Policy Research Working Paper, no. 7904. IHSI. 2007. “Enquête Transport 2004-2005.” Bengtsson, L., J. Gaudart, X. Lu, S. Moore, Institut Haïtien de Statistique et d’Informa- E. Wetter, K. Sallah, S. Rebaudet, and R. tique. Piarroux. 2015. “Using Mobile Phone Data to ———. 2012. “Enquête Sur Les Conditions de Vie Predict the Spatial Spread of Cholera.” Scien- Des Ménages Après Seisme (ECVMAS).” tific Reports 5 (1). doi:10.1038/srep08923. Institut Haïtien de Statistique et d’Informa- Bertaud, A. 2002. “The Spatial Organization of tique. Cities: Deliberate Outcome or Unforeseen Kim, Y., and B. Zangerling, eds. 2016. Mexico Consequence?” World Development Report Urbanization Review: Managing Spatial 2003, Background Paper. http://siteresources. Growth for Productive and Livable Cities in worldbank.org/DEC/Resources/spatialorg- Mexico. The World Bank. http://elibrary. city.pdf. worldbank.org/doi/book/10.1596/978-1-4648- ———. 2014. “Cities as Labor Markets.” Marron 0916-3. Institute on Cities and the Urban Environment, Kopp, P., and R. Prud’homme. 2011. “Urban New York University. Working paper, no. 2 Transport in Port-Au-Prince.” prepared for (February). IADB. www.rprudhomme.com/resources/ Cadena, K., and E. Perge. 2017. “Transport and Rap+2011+Port+au+Prince+$28BID$29.pdf. Poverty in Haiti Existing Evidence from Kumar, A., and C. Diou. 2010. “Renouvelle- Household Survey Data.” Draft report. World ment Du Parc d’autobus a Dakar : Avant Bank. - Apres.” 11. Sub-Saharan Africa Transport CONATEL. 2016. “Tableau de Bord Du Secteur de Policy Program (SSATP) Discussion Paper. La Téléphonie Mobile.” Conseil National des Washington, DC: World Bank. Télécommunications. http://conatel.gouv.ht/ Lall, S.V., J.V. Henderson, and A.J. Venables. 2017. node/283. Africa’s Cities: Opening Doors to the World. Estupiñán, N., A. Gómez-Lobo, R. Muñoz-Raskin, Washington, DC https://openknowledge. and T. Serebrisky. 2007. “Affordability and worldbank.org/handle/10986/25896. Subsidies in Public Urban Transport: What Le Gallo, J., Y. L’Horty, and P. Petit. 2017. “Does Do We Mean, What Can Be Done?” World Enhanced Mobility of Young People Improve Bank Policy Research Working Paper Series, no. Employment and Housing Outcomes? 4440. 149 Evidence from a Large and Controlled – Toward a New Narrative: Systematic Experiment in France.” Journal of Urban Country Diagnostic.” Washington, DC: Economics 97 (January): 1-14. doi:10.1016/j. World Bank. http://documents.worldbank. jue.2016.10.003. org /curated/en /64 21314679 92805241/ Lu, X., L. Bengtsson, and P. Holme. 2012. Haiti-Toward-a-new-narrative-systemat- “Predictability of Population Displacement ic-country-diagnostic. after the 2010 Haiti Earthquake.” Proceedings Stokenberga, A. 2017. “Urban Accessibility of the National Academy of Sciences 109 (29): Analysis for Dakar, Senegal.” Draft. World 11576-81. doi:10.1073/pnas.1203882109. Bank. Peralta Quirós, T. 2015. “Mobility for All: Getting Tatem, A.J., Y. Qiu, D.L. Smith, O. Sabot, A.S. Ali, the Right Urban Indicator.” Connections - and B. Moonen. 2009. “The Use of Mobile Transport & ICT, no. Note 25 (November). Phone Data for the Estimation of the Travel http://pubdocs.worldbank.org/pubdocs/publi Patterns and Imported Plasmodium Falci- cdoc/2015/11/67931446737085683/Transport- parum Rates among Zanzibar Residents.” ICT-Newsletter-Note25a-Nov-highres.pdf. Malaria Journal 8 (1): 287. doi:10.1186/1475- Peralta Quirós, T., and S.R. Mehndiratta. 2015. 2875-8-287. “Accessibility Analysis of Growth Patterns in WHO. 2009. “Global Status Report on Road Buenos Aires, Argentina: Density, Employ- Safety: Time for Action.” Geneva: World ment, and Spatial Form.” Transportation Health Organization. Research Record: Journal of the Transporta- ———. 2015. “Global Status Report on Road Safety tion Research Board 2512 (November): 101–9. 2015.” Geneva: World Health Organization. doi:10.3141/2512-12. World Bank. 2010. “Analysis of Multiple Natural Peralta Quirós, T., S.R. Mehndiratta, and M. Hazards in Haiti (NATHAT).” http:// Catalina Ochoa. 2014. “Gender, Travel and reliefweb.int/report/haiti/analysis-multi- Job Access: Evidence from Buenos Aires.” ple-natural-hazards-haiti-nathat. Perge, E., M. Muller, G. Fleurantin, and K. ———. 2015. “Better Spending, Better Services – a Cadena. 2017. “Haiti: Steering towards Review of Public Finances in Haiti.” Public Effective Fuel Subsidy Reform.” Draft report. Expenditure Review. Washington, DC: World Bank. World Bank Group. Ryko, J.M. (2014, December 3). La problématique ———. 2016. “Perspectives Urbaines: Villes de la moto-taxi dans la zone métropolitaine Émergentes Pour Un Sénégal Émergent.” de Port-au-Prince. Le Nouvelliste. Retrieved World Bank. http://documents.worldbank. from http://lenouvelliste.com/lenouvelliste/ org /curated/en /6 6 69114679 95082776/ article/138681/La-problematique-de-la-moto- Perspectives-urbaines-villes-émergentes- taxi-dans-la-zone-metropolitaine-de-Port-au- pour-en-Sénégal-émergent. Prince World Bank, and ONPES. 2014. “Investing in Rospabe, S., and H. Selod. 2006. “Does City People to Fight Poverty in Haiti : Reflec- Structure Cause Unemployment? The Case of tions for Evidence-Based Policy Making.” Cape Town.” Poverty and Policy in Post-Apart- World Bank. http://documents.worldbank. heid South Africa, 262-87. org/curated/en/222901468029372321/Reflec- Singh, R.J., and M.A. Barton-Dock. 2015. “Haiti tions-for-evidence-based-policy-making. 150 SPOTLIGHT 3. INTER-CITY CONNECTIVITY: WHY IT ALSO MATTERS FOR HAITI’S PRODUCTIVITY Marisa Garcia Lozano* GAINS FROM INTER-CITY CONNECTIVITY Chapter 3 focused primarily on assessing the state of Haiti’s intra-city connectivity, and its impact on employment generation, accessibility, and productivity. Inter-city connectivity, however, also plays a pivotal role in harnessing the benefits of urbanization and economies of agglomera- tion. How cities, regions, and ports are connected proves critical in fostering economic prosperity, reducing poverty, and promoting livability (Ellis and Roberts 2016). So that the benefits of increased economic density are more widely shared in rapidly urbanizing countries such as Haiti, connective infrastructure is key (World Bank 2009). Connecting cities is the springboard to increasing the growth potential and productivity level of city markets – including markets of labor, goods, and services. Investing in connecting infrastruc- ture is expensive (at times representing more than 15 percent of a developing country’s GDP), but it reduces trade costs and facilitates more efficient allocation of resources (Ellis and Roberts 2016). Inter-city connections benefit both producers and consumers; they give producers access to input (including labor) and output markets in other cities and regions, and offer more options and better prices for consumers (World Bank 2013). With better connections, businesses have the possibility to relocate when land is too costly and hence, grow their profits by moving to other areas. When access to cities is not adequate, firms are left with no option but to bare these costs. In addition to enhancing productivity, investing in spatial connectivity also translates into higher livability levels, as measured by access to non-network basic services, such as education and healthcare. The quality of life of populations residing in lagging areas improves as they become well-integrated with places that provide these services. There is evidence pointing to the impact of distance required to travel to schools on attendance rates. A survey in Sri Lanka indicated that long distances to travel or lack of adequate transport to schools were given as reasons for non-en- rollment by children (17.8 percent). Further, a study conducted in Ghana (Vuri 2007) showed that the farther a school was located from a child’s home, the less likely child was to attend school by 0.03 percentage points for each additional minute of travel. *The author acknowledges Sarah E. Antos and Katie L. McWilliams for their valuable input. 151 A World Bank report1 points to three steps policymakers can take to identify the most effective improvements to connective networks. 1. Value the city’s external and internal connections. For external connections, this means comparing a city’s transport costs – and data on density, quality, and capacity of roads, railways, waterways – with those from similar cities, to identify the biggest improvement needs. 2. Coordinate among transport options and with land use policies and related infrastructure plans. This calls for identifying the modal mixes that a city demands and encouraging compe- tition to reduce the gap between transport prices and costs. 3. Leverage investments that will yield the highest returns for cities – collectively and individu- ally. City leaders must identify connective investments in areas with the highest demand of inter-city infrastructure and transport services and in corridors that yield the highest returns for efficiency and equity. HAITI’S PERFORMANCE IN CONNECTING ITS CITIES As mentioned above, connectivity and access to cities brings numerous gains to different stake- holders. But how well are Haitian cities connected? We can see how cities are performing by assessing access to markets, as well as access to services that raise living standards. One approach to measuring physical accessibility is looking at the current road network. Physical accessibility to markets was calculated for 138 cities and towns in Haiti using the road network and weighting by the size of markets.2 A cost matrix function within the network analysis toolset in ArcGIS was used to measure market accessibility. This function uses the road network to determine the cost (time) to get from an origin point to a destination point. First, a grid of 27,077 origin points was created across the entire country using a resolution of 1km , and destinations were identified as 238 market towns. Travel time calculations resulted in over 3 million routes through the transport network, connecting the origin points with each destination market. As shown in Figure 1, areas with higher levels of accessibility are depicted in red while areas with low levels of accessibility to markets are in blue. The map below suggests that while market accessibility in Port-au-Prince and its surrounding areas is high (in red), there are some valleys of limited access in the north and southwest (in blue), as well as in some areas in the middle of the country (yellow). Particularly striking are the low levels of accessibility in the south and northwest areas of the country. Access to schools, hospitals and health facilities was measured across urban and rural populations based on travel times by road and walking (see Figure 2 [A-C] below). Travel times were measured based on the speed limits in the Haitian CNIGS road dataset, and a 4km/hour walking speed. The CNIGS dataset was improved around Port-au-Prince and other small cities in Haiti by using satellite imagery and/or OpenStreetMap to identify roads that needed to be added to the dataset. Roads were then individually digitized to create a complete network. Areas were defined as either urban or rural 1 World Bank, 2013. Planning, Connecting, and Financing Cities – Now: Priorities for City Leaders. The report discusses connections both within and between cities, but this piece brings in information related only to inter-city connections. 2 Market accessibility measures were calculated using the following formula: Accessibility of any given Origin Point X = Popula- tion of Market Town Y *e [(-Time between X and Y) / (2* a2)]) where Y is the destination point, e is a constant fixed at 2.71, and a is the maximum distance across the country. 152 MARKET ACCESSIBILITY IN HAITI Figure 1. Source: World Bank Urbanization Review Team’s calculations based upon the European Commission designations: densely populated areas, intermediate density areas, and thinly populated areas. The intermediate density areas and thinly populated areas were consolidated into a “rural” categorization, and the densely populated areas into an “urban” categori- zation. Hospitals and other health facilities, and schools (provided by the Haitian government) were then determined to be either “urban” or “rural.” Catchment areas around the locations of the schools, hospitals, and health facilities were built based on the following parameters: • Hospitals: Urban – within 30 minutes of driving; Rural – within 60 minutes of driving • Other Health Facilities: Urban – 15 minutes driving; Rural – 30 minutes driving • Schools: Urban – 30 minutes walking; Rural – 60 minutes walking Catchment areas were created by linking the hospital, health facility, and school locations to the closest point along the road network (with a maximum snapping distance of 5km). All hospitals and health facilities were within this threshold of a road, but there were 152 rural schools that were not. In such cases, it was confirmed with satellite imagery that there were no apparent roads in the vicinity. 153 Figure 2. ACCESS TO EDUCATION AND HEALTH SERVICES ACROSS HAITI (A) Access to Schools (B) Access to Hospitals (C) Access to Health Facilities Source: World Bank Urbanization Review Team, with data from the Haitian CNIGS road dataset and population data from Haiti WorldPoP (www.worldpop.org). 154 COMPARISON OF INTERNATIONAL PRICES OF TON-KM (USD) Figure 3. Source: Authors’ elaboration using Digicel data. For these unconnected schools, the catchment areas were created by building a 4km buffer around each school (which approximately equals one hour of walking time). Once all the catchment areas were created, 1km buffers were added around the periphery of the road-based results in order to better account for people who may be walking to the road network. Finally, zonal statistics were calculated using Haiti WorldPop3 data to determine the number of people within each of the catchment areas. According to the results, access to health facilities shows the largest gap, with as much as 3.8 million people not being near a facility. Rural populations are farther away from hospitals compared to urban residents, with an average travel time of thirty minutes longer. Access to schools is more evenly spread across the country, but rural populations remain at a far greater disadvantage than urban dwellers. 3 Based on IHSI 2015 estimations. 155 The key to stronger connectivity is a well-functioning and efficient transport network.4 Haiti’s transport sector plays an important role in its economy, contributing 12 percent of the national GDP. Terrestrial transport moves an estimated 80 percent of goods and people across the country, and hence the performance of the trucking industry provides a good indicator of connectivity between cities. In terms of prices for transport services, moving goods around Haiti is highly costly. The price per transported ton-km is USD 0.43, the highest in LAC, and among the highest when compared to countries in other regions. In the group of countries shown in the figure below, only Tunisia (USD 0.45), Japan (USD 0.67), and South Korea (USD 0.81) have a higher haulage price per ton-km than Haiti. Fuel, maintenance and tires, and labor are the three largest components of operating costs for all Haitian operators, representing 44, 27.9 and 22.1 percent, respectively. Road quality is a major factor contributing to the high expenses operators allocate to tires. They spend as much as 15 percent of their operating costs, compared to a visibly lower average of percent in Central America. According to the survey, the “road condition” variable increases the price per ton-km by 25 percent on average. Some progress has been made to improve and expand the road network, and hence improve overall country connectivity. But greater investments are needed to enhance mobility of markets and produc- tivity, especially for rural populations. Around 50 percent of the country’s territory is poorly connected, affecting 3.6 million people, including 3.2 million in rural areas. An overview of Haiti’s inter-city connectivity shows that physical access to markets and basic services such as schools, hospitals and health facilities, is lagging. This is particularly true for cities and towns outside the Port-au-Prince metropolitan area, which tends to be better connected. To foster productivity and enhance livability, greater links between cities and between rural and urban areas are crucially needed. While more work is required to better understand the gaps in connectivity, pointing to areas for improvement, such as roads, is a starting point. Identifying the bottlenecks in the north and southwest regions of the country is also key to developing effective policies that contribute to improving overall connectivity. WHAT HAVE OTHER COUNTRIES DONE – OR ARE DOING – TO INCREASE INTER-CITY CONNECTIVITY? The Republic of South Korea. The Government invested heavily in a major road infrastructure program to increase connectivity between provinces and cities. The construction of the highway system yield great economic gains; between 1995 and 2010, levels of gross regional domestic product increased substantially, especially in areas with better connectivity. The country’s spatial evolution of the economy also changed with a more integrated transport infrastructure. It enabled the development of new towns and the deconcentration of manufacturing jobs from Seoul and Busan into secondary cities. 4 This section draws heavily from the 2014 World Bank report on the trucking industry in Haiti. The report discusses the results of a trucking industry survey developed and implemented by the Etude Economique Conseil (EEC) Canada between March and May 2014. Its purpose was to develop a better understanding of inter-city and trans-border transport costs in Haiti. It is based on 280 origin-destination combinations obtained from 100 respondents, including individual Haitian truckers, medium and large Hatian companies, as well as operators transporting from the Dominican Republic into Haiti. 156 The People’s Republic of China. Since the late 1980s, the country has heavily invested in physical infrastructure to connect cities and regions across the country. The construction of highways and integration of transport systems has substantially improved connectivity and facilitated mobility of labor, especially from rural to urban areas (World Bank 2014). More recently, between 2006 and 2012, China built 780,500 kilometers of roads, 50,860 kilometers of expressway, 65,230 kilometers of rural highways, 20,900 kilometers of railway, 2,3631 kilometers of high-grade inland waterways, and 41 airports (World Bank 2014). Thanks to the new expressways and high-speed railways, inter- city travel time has been cut by 50-70 percent. Enhanced connectivity has also improved accessi- bility to and from second- and third-tier cities, and finance, information technology, tourism, and manufacturing sectors have gained from facilitated connections between firms. Sri Lanka. The Government of Sri Lanka carried out major investments to develop its road sector in order to increase connectivity for social integration and economic development. With assistance from the World Bank, and other development partners, Sri Lanka launched the Road Sector Assistance Project. The national road system was made more efficient through investments in road rehabilitation and maintenance. Improvements in rural roads benefited communities in terms of access to public amenities, schools, health care, business centers, and markets, thus promoting spatial equity. Some noteworthy changes include: agricultural sales points increased by 143 percent; 14 percent of students moved from poorly resourced schools to schools in towns with better facilities and higher quality teachers; and school attendance increased by an average of 45 percent (World Bank 2016). India. In 2001, India launched the Golden Quadrilateral (GQ), a 6,000 km roadway connecting the four major Indian industrial and cultural centers: Delhi, Mumbai, Kolkata, and Chennai. The GQ had improved the connectivity and market accessibility of districts closely located outside these four cities compared to those farther away. For instance, districts 0-10 kilometers away from the GQ network show a significant entry of new manufacturing firms and increases in productivity (Ghani, Goswani, and Kerr 2013). England. The UK government has put forward a “Northern Powerhouse” policy agenda to reduce the productivity disparities between the north and south regions of England. In the last 30 years, GVA per capita in the north has averaged 25 percent below than the rest of England (SQW 2016). Initially introduced in 2014 by the then Chancellor of the Exchequer, the Northern Power- house – home to 10.7 million people – seeks to strengthen the economic and physical connections of six main city regions: Hull, Leeds, Liverpool, Manchester, Sheffield, and the North East. The results of an independent economic review concluded that improved transport connections across the north, along with better skills, innovation, and inward investment, could increase GVA by 15 percent (approximately USD 124 billion) and create an additional 850,000 jobs by 2050. Transport for the North (TfN), a partnership of elected and business leaders, is currently working on an ambitious vision for a rail network that will link the six city regions and the regions’ largest airport. 157 REFERENCES Ellis, P., and M. Roberts. 2016. Leveraging Urbanization in South Asia: Managing Spatial Transfor- mation for Prosperity and Livability. Washington, DC: World Bank. Etude Economique Conseil – EEC. 2014. “Report on the Trucking Industry in Haiti – 2014.” Ghani, E., A.G. Goswani, and W.R. Kerr. 2013. “Highway to Success in India: The Impact of the Golden Quadrilateral Project for the Location and Performance of Manufacturing.” Policy Research Working Paper #6320. World Bank. Institut Haïtien de Statistique et d’Informatique. 2015. Population totale, population de 18 ans et plus, ménages et densités estimés en 2015. Retrieved from http://www.ihsi.ht/pdf/projection/Estimat_ PopTotal_18ans_Menag2015.pdf SQW. 2016. “The Northern Powerhouse Independent Economic Review: Final Executive Summary Report,” from http://www.sqw.co.uk/files/5414/6723/8824/16987_-_TfN_-_NPH_IER_-_Executive_ Summary_-_Final_24_June_2016.pdf. UNICEF. 2013. “Out-of-School Children in Sri Lanka: Country Study” from http://www.uis.unesco. org/Education/Documents/srilanka-oosci-report-2013.pdf. Vuri, D. 2007. “The Effect of Availability and Distance from School on Children’s Time Allocation in Ghana and Guatemala.” Understanding Children’s Work working paper series. Washington, DC: World Bank. World Bank. (May 5, 2016). Feature Story: 10 Years of Better Roads in Sri Lanka. http://www. worldbank.org/en/news/feature/2016/05/05/10-years-of-better-roads-in-sri-lanka. World Bank. 2004. Sri Lanka – Reshaping Economic Geography: Connecting People to Prosperity. Colombo: World Bank. ———. 2009. World Development Report 2009: Reshaping Economic Geography. World Bank. ———. 2013. Planning, Connecting, and Financing Cities - Now: Priorities for City Leaders. Washington, DC: World Bank. World Bank. Development Research Center of the State Council, the People’s Republic of China. 2014. Urban China: Toward Efficient, Inclusive, and Sustainable Urbanization. Washington, DC: World Bank. 158 CHAPTER 4 FINANCING HAITIAN CITIES Jonas Ingemann Parby and Joseph Denis* GONAIVES, ARTIBONITE. PHOTOGRAPHED BY COMMUNICATION SPECIALIST 2ND CLASS, ERIK BARKER, 2008 SOURCE: WIKIMEDIA, US NAVY PUBLIC DOMAIN CHAPTER 4 – FINANCING HAITIAN CITIES As mentioned in earlier chapters, cities urbanization, the ability of local governments become vibrant and livable places for both or communes to finance basic public services people and firms when they offer adequate and infrastructure is increasingly pressing. coverage of basic services – such as water, Better planning, connecting and servicing sanitation, electricity, roads, and solid waste cities and towns in Haiti require significant collection. But like all large investment interventions to review, revise and scale up projects, upfront capital is needed to put in existing financing arrangements. The ability place the proper infrastructure to support of municipal governments to provide services such services. With the existing gaps in efficiently is heavily constrained when local infrastructure and services, and the flow of revenues are limited, remain unchanged, or aid progressively declining, Haiti faces big do not grow a la par with increased population challenges in strengthening public finances, and responsibilities. In Haiti, due to financial specifically adapting to the reductions, raising and technical constraints, municipalities are more resources, and making better use of unable to carry out all the functions estab- existing funds (World Bank 2016). Despite lished under the 1987 Constitution and other improvements in Haiti’s own fiscal revenue, decrees related to decentralization (IMF 2015), from less than 10 percent of GDP in 2004 to as it will be further detailed in this chapter. 12.6 percent of GDP in 2014, Haiti remains Current levels of resources fall extremely short the poorest performer in revenue mobilization of the requirements and demands for services, in Latin America and the Caribbean (World and coordination failures increase the risk of Bank 2016). This greatly hinders the country’s suboptimal use of resources. ability to carry out much-needed development Cities need to have resources to address spending in infrastructure, health, education, the deficiencies across major urban services, and other key sectors. The infrastructure including drainage and sanitation, solid financing gap remains very wide and signifi- waste, transport and spatial planning, and cant amounts of capital are needed to narrow management for future urbanization. Master this gap. In Haiti’s current framework of plans for many cities remain unfinanced decentralization, and in a context of rapid and out of sync due to delays in implementa- * The team wishes to thank Digicel for granting access to the CDR data. The authors thank Katie L. McWilliams, Benjamin P. Stewart, and Lauren Nicole Dauphin for providing important help in running the network analysis and the calculation of transport times in Port-au-Prince and Cap Haïtien. Pierre Xavier Bonneau provided crucial guidance and, together with Malaika Becoulet and Franck Taillandier, helped the team navigate the issues of urban transport in Haïti. Emilie Perge’s in-depth knowledge of the ECVMAS survey was key to understand transport expenditures and to the writing of the corresponding sections in this chapter. 163 tion. Intra-urban transport in Port-au-Prince municipal governance. Public financial requires investments to upgrade, diversify, management capability is weak in municipal and scale public transportation systems. Some governments, and budget execution is ineffi- missing road links need to be built to connect cient, resulting in poor service delivery. In key cities to stimulate domestic economic addition, budget reporting is very limited, integration. This requires improving the and classification of expenditures is not volume, predictability, timeliness, and homogeneous. Fiscal transfer mechanisms do management of finances for infrastructure not sufficiently provide incentives for public and maintenance, as well as harnessing the financial management improvement in local potential additional mechanisms for revenue government, as fiscal transfers are not linked generation. It also requires a systematic effort to performance in services delivery and to adjust and implement existing reforms financial management, and current transfer aimed at improving national and local govern- principles are not adequately needs-based. ment management and oversight of resources, Therefore, strengthening public financial along with a thorough review of existing management capability of municipal staff local government support grants and their should be of continuous importance and management. These efforts may substantially well-coordinated between the main stake- contribute to help Haiti reap the benefits of holders, Ministry of Finance and Ministry devolution and bringing services closer to of Interior and Local Authorities, to ensure 1 citizens, as well as addressing inequity. harmonization of the PFM reform agenda Public financial management (PFM) at between government agencies. the subnational government level is vital Policy options should first focus on fixing for successful decentralization. Financial inconsistencies and gaps in the institu- management is an important competence for tional, regulatory, and financing framework municipalities as “it enables local government for all municipalities. This means to first to plan, mobilize, and use financial resources address the inconsistencies in devolution in an efficient and effective manner, as well and decentralization so that functions as fulfill its obligation to be accountable to its follow finances and basic minimum human citizens” (Farvacque-Vitkovic and Kopanyi resources capacity is in place within subna- 2014). Effective devolution of key expen- tional entities. Second, it will be important diture and revenue functions to municipal to review, revise, and strengthen the fiscal governments has not taken place yet. Critical transfer systems in key areas, specifically reforms first need to be implemented to addressing the gaps and inconsistencies in enhance equalization and transparency the allocations, transfers, and monitoring of in fiscal transfers, develop more dynamic the Local Government Development Fund sources of local revenue, and strengthen (FGDCT). Third, the national government 1 This chapter is based on the work conducted under two separate studies, one looking at municipal financing in Haiti as a whole, and the other a case study analyzing the financial situation of six communes in the North Department (Municipal Finance in Haiti [2017, working paper] and Diagnostic des communes de l’agglomérat du Cap-Haïtien [World Bank 2017, working paper]). Please refer to Annex 2 for specific consideration and limitations regarding methodology and data. 164 should focus on boosting own-source revenue available can contribute to transparency and collection, including expanding the registra- build long-term trust. A participatory process tion of taxpayers, upgrading the cadastral that involves citizen from early planning register, as well as revising the formulas for stages can ensure that actions respond to distribution and allocation of taxes collected local needs. at national level to increase transparency and In the following sections, this chapter objectivity. In addition, it will be important reviews the state of municipal finances within to support and enable better collaboration the current decentralization process and between departments, communes, and the highlights key areas that need strengthening, utilities to generate economies of scale in providing specific recommendations for public service delivery. action in the short, medium, and long term. Improved planning and accountability and enhanced transparency can help increase THE PROCESS AND PROGRESS OF DECEN- citizen confidence in local government. While TRALIZATION IN HAITI urban local governments in general have been Urbanization in Haiti brought about two better able to collect own-source revenue, significant challenges in local public invest- the demand for services is also much more ment. The first is to improve the level of urban expansive. In a context of high urbanization, infrastructure in cities to enable economic the development and management of cities activities and reduce urban inequality. Over play a decisive role in the development of the the past two decades, the gap between the country. Unfortunately, local governments funding capacity of the country and the pace struggle to provide adequate services due to of urban growth has led to a constant deficit a chronic shortage of resources. At the local of urban infrastructure in towns and cities, level, insufficient attention is paid to citizens’ as covered in earlier chapters. The other needs in the planning process, and account- challenge is to finance urban development to ability and transparency in the use of public “keep pace with urban growth” as cities sprawl funds are critically lacking. and expand in size and population. As shown Scaling up upward and downward in Table 1 below, access to basic services accountability can, in turn, improve local remains limited. service provision in the long term. Building The Constitution of 1987, and its subse- on lessons from local programs and projects quent 2012 amendment, anchors the decen- that have managed to successfully respond tralization process in Haiti. In its preamble to citizens can shed light on how to build and Articles 81, 83, 87, and 217 – which the necessary systems for ensuring stronger explicitly outlines the autonomy of the local accountability. Existing efforts to strengthen communes, the decentralization of public the capacity of local governments to take services, and the establishment of a framework charge of their development would need to be for local finances – Haiti laid down the basis accompanied with efforts to bolster financial of decentralization. As part of the democratic management, as well as capacity for internal reform process following the departure of the monitoring and control. Further, efforts to Duvalier regime, the 1987 Constitution lays make information on local finances publicly the foundation for good local governance 165 Table 1. ACCESS TO BASIC SERVICES COVERAGE RATES (2001-2012) (IN PERCENT) INDICATOR NATIONAL URBAN RURAL 2001 2012 2001 2012 2001 2012 School-age children in school 78 90 84 93 74 87 Under-5 mortality (per 1’000 live births) 137.7 92 111.7 88 149.4 99 Children (12-23 months) fully vaccinated 33.5 45.2 33.6 2012 2001 2012 Access to improved drinking water sources WHO definition a — 53 — 55 — 52 Access to tap water (in house) 7 11 13 18 3 5 Expanded definition b — 73 — 91 — 56 Treated water purchased — 20 — 36 — 4 Access to energy c 32 36 62 63 11 11 Rate of open defecation d 63 33 44 11 76 53 Access to improved sanitation e — 31 — 48 — 16 Habitat, nonhazardous building materials 48 60 71 81 33 41 Sources: ECVH 2001; ECVMAS 2012; World Bank and ONPES (2014). Note: — = not available. WHO = World Health Organization.2 a According to the international definition (WHO), access to improved drinking water is the proportion of people using improved drinking water sources: household connection, public standpipe, borehole, protected dug well, protected spring, rainwater. b The expanded definition includes the international definition (WHO), plus treated water (purchased).  Includes electricity, solar, and generators. c d  Rate of open defecation refers to the proportion of individuals who do not have access to improved or unimproved sanitation. This indicator is part of the Millennium Development Goals (MDG) and is a key element of discussion for the post-2015 agenda. The open defecation rate declined from 63 to 33 percent nationwide between 2000 and 2012, reflecting gains in both urban and rural areas.  Improved sanitation is access to a flush toilet or an improved public or private latrine.générateurs. e 2 International WHO definitions used. 166 with an emphasis on bringing public services and 570 communal sections (Figure 1). There closer to citizens. In moving forward with a are thus four administrative levels of the State. decentralization process, the national govern- Three levels for the local authorities are set ment recognized that good local governance out in the constitution and in the subsequent could become an important avenue toward decrees: the department, the commune, the revival of the national economy, as well as and the communal section. Executive and for improving and expanding the provision of deliberative bodies, govern each body. In the basic services. case of the communes (hereafter referred To ensure the effectiveness of government interchangeably as local governments), it is interventions, the 1987 Constitution and the important to note that the powers given to laws3 of the country organized Haiti in such them by the law are specified in title III of the a way as to encourage and foster the partici- decree5, defining the operating and organizing pation of local communities in public affairs. principles of the territorial collectivities. Each As such, Haiti is a decentralized unitary state. commune has a municipal council (conseil Article 61 of the 1987 Constitution defines municipal) consisting of three members local and regional authorities at three levels: elected by the population of the commune for the department, the commune, and the a four-year term. The municipal council is led communal section (section communale). by a president – the mayor. Each commune They are made up of deliberative organs (the has a municipal assembly (assemblée munic- assemblies) and executive structures. These ipale) whose members assist the council in its structures enable local governments to have work. Assembly members are also elected for decision-making power and management four years. autonomy. Since local governments are closer The Constitution aimed at building local to communities than the national govern- and regional authorities as powerful instru- ment, they are in the best position to decide ments for decentralization and with the on the selection of projects to meet the devel- ultimate objective of improving the effec- opmental objectives of the communes and tiveness of public service provision. By their improve the well-being of their constituents. mandate and functions as stipulated in the As of 2017, Haiti has 10 departments, which Constitution, local and regional authori- are further subdivided into 42 arrondisse- ties have a wide range of competencies and ments4, 146 communes (6 communes have responsibilities in the provision of services been added in the last 2 years, but are still (see Box 1 which outlines the main mandates subject to clarification of their boundaries), of communes). They are created to contribute 3 For example, the law on the organization of communal sections (1996), the creation of the Management and Development Fund of the Territorial Collectivities (FGDCT) (1996), the publication of the five decrees related to the charter of territorial collectivities: Organization of the departments, organization of the communes, organization of the communal sections, and the framework decree of decentralization (2006), and the creation of the Local Development and Land Use Planning Fund (2007). 4 Arrondissements are group of municipalities. They have only administrative functions and most often are not well established structures. They do not have any executive authority. 5 Decree defining decentralization framework, as well as the organization and operating principles of the territorial collectivi- ties, passed on February 1, 2006 and published by the official gazette Le Moniteur on June 14, 2006. 167 Figure 1. TERRITORIAL DIVISION OF HAITI MUNICIPAL RESOURCES Source: Author’s elaboration. to the improvement of the living conditions basic and vocational levels. In theory, they are of the population at all levels. Local govern- empowered to collect taxes to finance their ments are involved in several key areas of operations and investment projects. The legis- service delivery, including: environmental lative framework also confers management protection, land use planning, watershed and administrative autonomy to local author- management, solid waste management, ities. Local and regional authorities also have drinking water distribution, parks and recre- the ability to appoint individuals for positions ations, and the safety of populations and at different levels of the state, in particular in provision of education services at both the the local justice system.6 6 As such, local and regional authorities have competencies to designate judges for the Justices of the Peace, the Courts of First Instance, and those of the Courts of Appeal, while the councils of the commune sections are legally empowered to prepare the list for members of a jury and ensure compliance with laws and regulations. 168 BOX 1 – FUEL PRICES POLICIES AND THEIR IMPLICATION ON CONNECTIVITY AND HOUSEHOLDS’ HABITS Below, the key mandates for local governments included in the 1987 Constitution are summarized according to the main functions. Territorial Management: roadway construction; development and implementation of subdivision plan, after approval of the supervisory authority; issuing building permits; issuing of compliance certificates; regula- tion of urban traffic; designation, construction, and maintenance of the sites of railway stations and parking lots; numbering of the houses, road signs, and naming of the streets; construction of public spaces; construc- tion of recreational areas; construction of sanitation infrastructure. Land Management and Registration: allocating plots and issuing operating titles related to the communal property sector or the parts of the national land that have transferred management to their benefit; tax collec- tion, taxes, tickets, and rights and royalties linked to real estate and land ownership. Environmental and National Resource Management: sanitation and treatment of liquid pollutants; fight against squalor and pollution; removal of solid waste. Health and Hygiene: construction and participation to the management of first health level structures; imple- mentation of hygiene service and sanitary police; inspection of food products’ quality; inspection of expiry date of medicines; regulation and action taken regarding hygiene, squalor, and illness prevention; water quality control. Education and Vocational Training: promotion of universal enrollment; localization, construction, and involvement in the management of public secondary schools or high school; promotion of vocational and technical schools. Culture and Sports: construction and management of cultural, sports, and youth infrastructure; valoriza- tion of historical, natural, archaeological, cultural, and artistic potential; promotion of cultural, sports, and youth activities; construction and management of museums and libraries; management of sites and historic monuments; creation of green spaces. Civil Protection, Assistance, and Relief: contribution to the organization and relief management for the benefit of vulnerable group and victims; management of asylums; management of public orphanages and youth rehabilitation centers; participation to the organization of civil protection and firefighting. Funeral Homes and Cemeteries: cemeteries development; issuance of exhumations permits; oversight of compliance with regulation in terms of funeral operations and transfer of human remains; construction, maintenance, and management of funeral homes. Water and Electricity: production and distribution of drinking water; drilling and management of wells and publics standpipes; development and implementation of the water supply scheme; management of energy infrastructure; installation and management of the street lighting system. Markets and Slaughterhouses: construction and regulation of markets, slaughterhouses, and slaughter spaces; organization of fairs; management of municipal markets. Public Safety: participation in the municipal security council. Source: Authors’ elaboration based on the Constitution of Haiti, 1987. Source: Authors’ elaboration based on the Constitution of Haiti, 1987. 169 While a decentralization framework limited own-source revenue mobilization is in place, several obstacles still impede (even in large urban areas) contribute to the effective devolution of competencies poor service delivery at the local level. The and the implementation of public services second obstacle is related to the devolution under a decentralized model. The first of competencies. While some progress has obstacle is systemic and affects the delivery been made, the framework for devolution of local services. Limited and unpredictable continues to feature contradictions and financing, delays in central government overlaps in the existing legislation and in the transfers, and the lack of transparency application of the existing mandates, leading about the application of regulations for to confusion between ministries and local the key national transfers combined with governments about financing and service REPORT ON EXPENDITURES OF THE COMMUNES COMPARED TO Figure 2. THE GDP OF THE SELECTED COUNTRIES FROM LATIN AMERICA, THE CARIBBEAN, AND AFRICA Sources: ECLA and World Bank. 170 REVENUE PARTICIPATION OF LOCAL GOVERNMENTS IN THE Figure 3. COUNTRIES OWN TOTAL REVENUE Sources: ECLAC, MEF, DGI, and MICT. delivery mandates. This means that devolu- vein, the share of revenue collected at the local tion in some cases is not complete, or the level also remains low in Haiti compared to related financing for a specific function is not other countries in the region. transferred to the local governments, hence Despite the legal framework providing an creating inefficiencies and unpredictability assignment of competencies between the in the provision of services. central government and local government, the In comparison with other countries in the delineation of responsibilities is not always region, the relative share of expenditures and clear. Beyond the Constitution and other legal resources spent at the decentralized level documents such as the organic law of the remain limited in Haiti. Only 0.6 percent of General Tax Office (DGI) and the Ministry of GDP is spent at the commune level, and total Economy and Finance (MEF), a series of five municipal revenue makes up only 1.7 percent decrees from February 2006 cover important of total revenues (see Figure 2 above and Figure aspects related to municipal resources 3 below). Participation of local governments and some legal prerogatives granted to the is very small compared to other countries in commune.7 The five decrees lay the general the region. Thus, urban local governments in and comprehensive legislative framework of Haiti face massive constraints in addressing decentralization, along with the organization increased service delivery demands in the and operating principles of the Haitian terri- context of rapid urbanization. In similar a torial collectivities.8 171 Lack of clarity in the current legal in the law and legal frameworks. De jure, framework leads to confusion and overlap of municipalities are responsible for providing responsibilities. According to the provisions basic services, but their weak planning and of the Constitution and the 2006 decrees, financial capacity constrain them from several areas of responsibility are being fulfilling this function in its entirety. In most shared by both levels of governments (central cases, the national government steps in and and local). In some sectors, the delineation takes charge of these responsibilities. There is clear, while in others, it is opaque. As are, however, some overlaps among agencies highlighted earlier, local governments have a of the national government in the de jure broad mandate, covering several sectors (see responsibilities. This creates deep coordina- Box 1). In the organization of the territory, tion challenges, not only between national the communes play a central role in land agencies, but also with international property management, including markets, agencies and NGOs that also enter the field fairs, slaughterhouses, cemeteries, and to fill the void of service provision. Figure 4 trash-collection being the responsibility of below illustrates the point of duplication of local governments. However, although the functions and institutional fragmentation legal framework provides a broad definition using the water sector as an example. of the roles and duties of local governments, lack of clarity leaves room for broad inter- pretation, and increases the risk of dupli- cation of functions in some cases, and lack of a leading actor in other instances (for example, in the cases of the transport sector, education, water, and sanitation). As a result, both citizens and local and national authori- ties do not have a clear understanding of the responsibility and role of local governments with regard to the delivery of services and local economic development. De jure, overlapping responsibilities create confusion; de facto, limited local capacity leads to local governments not being able to meet their mandates in the provision of public services. Responsibili- ties overlap both in practice and as written 7 The five decrees included provisions for the organization of the various levels of local government: (i) Organization of the departments, (ii) Organization of the communes, (iii) Organization of the communal sections, (iv) Territorial civil service, and (v) provided a framework decree for decentralization. 8 Refer “Le Moniteur” No 57 of Wednesday June 14 of 2006) and the one setting the organization and operation of the commune (ref. “Le Moniteur” No 2 of 2006). 172 OVERLAPPING OF RESPONSIBILITIES IN WATER GOVERNANCE – NATIONAL, Figure 4. DEPARTMENTAL, AND COMMUNAL LEVELS DEMAND AND PLANNING ENFORCE REGULATORY CREATE REGULATORY DESIGN OF POLICIES IDENTIFICATION OF QUALITY CONTROL INFRASTRUCTURE OPERATION AND COORDINATION PROVISION OF MAINTENANC ADMINISTRA- ENTITY ENTRUSTED NORMS NORMS TIVE LEVEL RESPONSIBILITY National CIAT Management of water resources Ministry of Management of Environment/ water resources Ministry of Agriculture, of Natural Resources, and Rural Development MARNDR Irrigation DINEPA, Ministry of Drinking water and Public Works, Transport sanitation services [1] and Communications Regional Regional Office for the Drinking water and management of water sanitation services and sanitation services (OREPA) of the DINEPA Departmental Unité Rurale Drinking water and Departemental (URD) of sanitation the DINEPA Municipality Technical Operations Drinking water and Center (CTE)[1] for sanitation urban areas, “Comités d’Approvisionnement en Eau Potable et Assainissemen”t (CAEPA) and professional operators (OP) for small towns and denser rural areas, “Comités de Point d’Eau” (CPE) for rural areas “Technicien en Eau Drinking water and Potable et Assainissement sanitation Communal” (TEPAC) Irrigation association Irrigation [1] The 2009 Act gives DINEPA broad powers and responsibilities in the field of drinking water, through the Ministry of Public Works: tariffs, water quality, licensing, monitoring and evaluation of water quality and system performance, approval of infrastructure projects and mediation between contractors. Source: Authors’ elaboration using information from Ryan Stoa (2015). Water Governance in Haiti: An Assessment of Laws and Institutional Capacities. Available at: http://ecollections.law.fiu.edu/faculty_publications/97 173 INCOMPLETE DECENTRALIZATION AND A eighties, moving from building to actually WEAK LEGAL FRAMEWORK FOR MUNICIPAL implementing a legal framework has been FINANCE CONFOUND RESPONSIBILITIES difficult, leading to what today is an incomplete A fragile and fragmented legal framework decentralization. Political instability under- governs municipal finances in Haiti and mined the actual and full implementation does not facilitate revenue collection by and enforcement of decentralization. In many local governments. Municipal finance is cases, this has left local governments inactive, governed by several laws and regulations, dysfunctional, and in some cases, without an including, (i) the 1987 Constitution (and 2012 official mandate due to substantial delays in amendment), (ii) a set of laws and Presiden- holding local elections. Furthermore, the lack tial decrees including the organic law of the of both financial and human resources at the General Tax Office (DGI), (iii) the organic local level have constrained implementation. law of the Ministry of Economy and Finance Serious efforts from the central government (MEF), (iv) several outdated tax legislations, to equip local governments with the tools and as well as (v) a law creating the Local Govern- technical competencies they need to deliver ment Development Fund (Fonds de Gestion et on their mandates have been limited and de Développement des Collectivités Territoriales sporadic. In addition, the magnitude of central [FGDCT]). Article 217 of the Constitution government transfers is marginal with respect stipulates that “the finances of the Republic to the needs of local authorities, which affects have two components: National Finance and their capacity to provide service delivery at the Local Finance. They are managed by bodies local level. Estimates suggest that resource and mechanisms provided for this purpose.” needs are at least five to eight times larger than However, new rates and new sources of local existing transfers. In 2015, the per capita alloca- taxation can only be created by Parliament, tion of the FDGCT ranged between USD 0.3 not by the Communes. Article 218 of the and USD 2.23 per capita among Hait’s depart- Constitution further contends that “no tax ments, compared to estimated needs of about for the benefit of the State may be estab- USD 10 per capita, that would be required to lished except by law. No charge, no taxation, bridge infrastructure gaps. whether departmental, municipal, or Since their establishment in the 1987 communal, can be established only with the Constitution, local and regional authorities consent of these Territorial Collectivities.” have never been fully operational in accor- In the absence of a robust legal framework, dance with the law. Communal and depart- municipalities are not empowered to substan- mental assemblies have never been constituted tially increase their tax base, imposing new due to successive political and electoral crises local taxes and enforcing collections. and lack of institutional capacity, thereby The decentralization system mandated hindering accountability in local govern- by the 1987 Constitution has never been fully ments. The election of new mayors in all implemented due a combination of factors, communes in 2016 marks a new opportunity including political instability, lack of resources, for decentralization in Haiti, which now and complex political economy issues. Despite requires consolidation via the establishment efforts on decentralization starting in the late of the communal sections to constitute the 174 councils (this is expected in 2017). These delib- decrees may have a positive contribution in the erative authorities have the task of guiding the decentralization process moving forward, due decisions of the executive authorities at the to the institutional innovations they bring to level of the communal sections, communes, bear.9 These decrees lay the foundations of the and departments. Each level has its own three levels of local government by providing executive and deliberative governing body, a clear definition of their mission, functions, with each commune being ruled by a munic- and operations. ipality. Members of the communal section Among the provisions introduced in governing bodies, the Communal Council 2006 is the definition of the main sources of (CASEC), the Communal Section Assembly funding for local governments. Articles 133 (ASEC), and the Municipal Council are all to 140 of the decree provide the framework elected by popular vote. The rest of the local on decentralization and set out the various government structures are elected indirectly. types of revenues for the territorial collec- The Municipal Assembly is elected indirectly tivities: regular revenues and extraordinary by the communal sections, and in turn, the revenues. The regular revenues include tax Municipal Assembly elects the members of revenues, municipal royalties, user fees, the Departmental Assembly. Further, the and central government transfers to the Departmental Assembly selects, among its territorial collectivities. The extraordinary members, three people to sit in the Depart- revenues include loan proceeds, temporary mental Council. The Departmental Assembly or occasional revenues, grants, and public also appoints one member to integrate the or private subsidies. Article 142 of the same Inter-Departmental Council (CID), who decree grants to the commune competen- participates in the Council of Ministers as an cies in terms of mobilization and collection assistant in decentralization activities and of the property tax (CFPB), mobilization represents the interests of the departments and collection of business tax receipts, and and the communes. the creation of some duties and municipal After the Constitution, the most important royalties. As to the decree on organization and changes to the legal framework came with five functioning of the communes, Section V on decrees in 2006. Since the 1987 Constitution, local finance (Articles 157 to 202) details the few legislative improvements had been made general framework and mechanisms to ensure for the implementation of the decentralization efficient management of municipal finance. process until the introduction of five decrees Despite the steps taken by the 2006 decrees, in February 2006. The five decrees focused on the current legal framework remains incom- a framework for decentralization, organization plete and contains contradictory provisions and functioning of communes, and provisions regarding decentralization of responsibilities. for grants to communes. The 2006 decrees Most of the institutions and mechanisms are considered an important breakthrough to proposed in the 2006 decrees have not been improve the legal framework for decentraliza- established yet or their operation is still in tion, and despite their imperfections, these early stages, so their effectiveness cannot yet 9 E.g. refer Paul and Charles (2014). 175 be assessed. Further, to some extent the 2006 LIMITED SOURCES OF MUNICIPAL REVE- decrees contradict certain provisions of the NUE HAMPER THE CAPACITY TO PROVIDE Constitution regarding taxation, such as the SERVICES creation of tax by municipalities. Local government financial resources Effective decentralization of fiscal resources are limited with municipalities outside remains the major impediment to effective of Port-au-Prince and surrounding cities implementation of decentralization in Haiti. (Petionville and Delmas) being highly The principle of fiscal and financial autonomy dependent on national transfers. Local of the local government is stated in the 1987 governments in Haiti have four main Constitution and its 2012 amendment, but sources of revenue: transfers from the local communities lack resources that enable central government, mainly the Local them to perform the functions entrusted Development Fund (FGDCT); taxes to them by the legislation. This situation collected on behalf of the communes by the is even worse in major cities (notably in National Tax Authority (DGI); duties and Port-au-Prince and Cap-Haïtien) where local royalties collected by the communes; and authorities lack resources to solve the serious other external sources of income, such as infrastructure deficits (including housing, those from development partners. Apart water and sanitation, drainage of waste water) from the three big cities in the metropolitan and are unable to properly provide basic urban area, the majority of communes depend services such as garbage collection or trans- heavily on the FGDCT as their main source portation. Additional options for infrastruc- of income (typically ranging from 80-95 ture financing and the implementation of the percent), while other types of revenue decentralization legal apparatus are needed to remain low (almost negligible). See Table capitalize on the benefits of decentralization 2 below outlining the resource distribution and bring services closer to citizens – and to between central and local governments. address both spatial and social inequity. Taxes collected by the DGI are second In a context of limited financial resources, in importance as a source of local revenue. creation of new local government entities The agency is placed under the supervision may lead to increased strain over resources. of the Ministry of Economy and Finance The recent creation of new local government (MEF), which is entrusted with the govern- entities needs to be carefully balanced to avoid ment’s fiscal policy. The operating mode of misalignment with resourcing, further compli- DGI is governed by the decree of September cating the implementation of the decentral- 28, 1987, which sets out in Article 2, among ization framework. For local authorities to other things, the following powers: enforce- be operational, they must have the human ment of tax laws and duty collection, as resources and financial means to meet their well as other public revenues. DGI is the needs. The recent creation of six additional authority that receives all tax income of the communes without the necessary financial central government and territorial collec- and human resource allocations is of concern, tivities and has administrative bodies that and the additional breakup of entities that are allow it to collect local taxes established by already small may create further inefficiencies. the law. These devolved bodies are the tax 176 RESOURCE DISTRIBUTION BETWEEN THE CENTRAL GOVERNMENT AND Table 2. THE COMMUNES (2014-15) MILLIONS OF HTG MUNICIPAL RESOURCES TOTAL RESOURCES CENTRAL FGDCT DGI DUTIES & TOTAL GOVERNMENT ROYALTIES 76,639 647 942 60 1,649 Source: Own figures based on data from the MEF, DGI and MICT. centers that operate in the major towns of torate of the Treasury. Lack of inter-gov- the departments and districts, and then the ernmental coordination leads often to local tax agencies operating in small rural delays in the transfer of these tax receipts. governments. In theory, the local taxes are Municipalities receive the total amount of regularly transferred to the accounts of the property tax collected by the DGI and 80 local governments on behalf of which these percent of the total amount of Business Tax collections were made. However, cumber- Receipt collected. Other local collections some administrative procedures cause (for example, building permits, markets delays in the release of funds collected fees, facilities rental, payments for use of by DGI to local governments. In some cemeteries) are generated at the munic- cases, delays can range from three to six ipality level with little transparency and months, thereby negatively impacting accountability in the use of these proceeds service delivery and daily operation of (see Table ). According to the Ministry of the commune. There is an urgent need to Interior and Local Authorities, property tax engage all stakeholders in identifying and accounts for about 86 percent of the total removing the key administrative bottle- amount of revenue, while the business tax necks, so that funds are released promptly (patentes) is 10 percent. to municipalities. The concentration of economic activities Even though overall tax collection in the metropolitan area of Port-au-Prince remains under the control of the DGI, is reflected in the large share of revenues municipalities are responsible for collecting collected at this level. Data gathered from local taxes and for negotiating the rules of the Directorate of the Treasury confirm local transfers. Taxes collected by DGI are the dominant role of Port-au-Prince and banked into temporary collection accounts its surrounding areas as the economic at the central bank and then transferred to center of the country since 98 percent of the the account of the commune at the national taxes collected by DGI for the communes bank following authorization of the Direc- are concentrated in the Port-au-Prince 177 Table 3. MAIN REVENUES COLLECTED IN FY 2013 (HTG) AND LEGAL REFERENCES LEGAL TOTAL ANNUAL AMOUNT TAXES REFERENCE COLLECTED Property Tax April, 5 1979 287,450,480.00 Business Tax Receipt September, 28 1987 214,700,154.00 Building Numbering Permit February 18, 1971 235,257.00 Construction Permit April 5 1978 Setback Regulations August 10, 1961 1,016,963.00 Right of Way Regulations August 7, 1913 497,875.00 Cattle Sales Permit August 2, 1950 438,469.42 Materials and Commodities September 9, 1918 60,639.00 on the right of way Tent Permit August 7, 1913 756,792.15 Taxe Calibration October 7, 1975 212,753.50 Burial Permit Septembre 28, 1938 1,199,868 Others 10,731,884.38 TOTAL 517,301,136.56 Source: DGI 178 metropolitan area. As depicted in Table DGI account for only 4 percent of total 4 below, the commune of Delmas has the revenues collection. Recent experiences highest level of revenue collected by DGI with various donor-funded projects aimed (dominated by the patente and CFPB with at mobilizing local revenues have shown 37 percent, followed by Petionville and that the communes could have increased Port-au-Prince with 21 and 20 percent, the amount of property tax collected if they respectively. Due to anemic economic had control of the collection process and growth in the other regions and a limited received technical assistance to build local local private sector, the amount of tax capacity. For instance, the municipality of generated through the business tax receipts Carrefour had more than quadrupled their remains limited outside of the metropol- revenue collection, from USD 309,000 in itan area. Other local taxes collected by 2011 to USD 1.8 million in 2012.10 REVENUES COLLECTED BY DGI ON BEHALF OF COMMUNES IN THE METROPOLITAN AREA OF PORT-AU-PRINCE (IN HTG), OCTOBER Table 4. 2013-MARCH 2015 TOTAL COMMUNES Delmas 264,280,509.89 SHARE Petionville 149,189,782.23 37% Port-au-Prince 142,013,642.25 21% Table 4: Revenues Collected by DGI on Behalf of Communes Tabarre 64,411,172.31 20% in the Metropolitan Area of Carrefour Port-au-Prince (in HTG), 36,687,347.95 9% October 2013-March 2015 Cité Soleil 27,786,339.02 5% Source: MEF. Croix-des-Bouquets 14,732,771.62 4% Total 699,101,565.27 2% Source: MEF. 98% 10 USAID Fact Sheet: Revenue Collection Provides Opportunity. One of the projects is Lokal+ financed by USAID. The project was designed to improve local governance and support the decentralization process in Haiti by strengthening the capacity and transparency of local governments and improving their ability to provide goods and services to their communities 179 Municipalities often only collect a fraction 15 percent being transfers from DGI from of potential revenues, and they lack the revenues collected on behalf of communes. required capacity to effectively use resources However, in FY 2015 FGDCT annual alloca- for the delivery of local services. Except for tion ranged from a mere 22 HTG or USD the municipalities in the metropolitan area, 0.3311 per capita in the metropolitan area to more than eighty percent of the overall 163 HTG (around USD 2.60) in the North- revenue of local governments come from West Department. At the same time (FY15), central government transfers, whereas on total tax revenues for the country’s 140 average only about five percent of revenue municipalities totaled only 1,250 million were own-source revenue collected by HTG – roughly USD 20.1 million. About 93 communes themselves with the remaining percent of this amount (1,166,000,000 HTG) Source: “Haiti: Steering towards effective fuel subsidy reform” (Perge et al. 2017), based on ECVMAS 2012 data and Haiti Public Expenditure Review (World Bank, 2015a). TOTAL TAX COLLECTION IN FY 15 ACROSS PORT-AU- Figure 5. PRINCE, DEPARTMENT CAPITALS, AND OTHER COMMUNES (IN GOURDES) Sources: MICT, PRAFIPUM. 11 Exchange rate on 1/1/2015 was 46.86 HTG:USD 1 180 BOX 2 – EXAMPLE OF REVENUE COLLECTION IN ACUL DU NORD IN THE NORTHERN DEPARTMENT, 2013-2016 (IN HTG) ACUL DU NORD REVENUE 2013–14 2014–15 2015–16 Property tax (CFPB) 97,378 180,797 139,884 Business tax (Patente) 1704 1120 2952 Total amount received from DGI (CFPB+ Patente) 123,822 213,167 197,736 Transfer from central government (FGDCT) 5,713,875 5,713,875 5,713,875 Extraordinary or ad hoc Transfers 285,000 175,000 510,000 Revenue collected by the commune 347,163 436,050 293,925 Total revenues 6,568,942 6,720,009 6,858,372 As seen in the table, annual central government transfers account for about 90 percent of the commune total revenues for the three years covered, and overall revenue remains limited in volume (around USD 105,000). The share of central transfers as share of total commune revenue declined slightly from 87 percent in FY 2014 to 85 percent in FY 15 and 83 percent in FY16. Revenue collected directly by the commune increased from HTG 347,163 (about USD 5,500) in tFY 14 to HTG 436,050 (about USD 6,900) in FY 15, but declined in FY 16 to stand at HTG 293,925 (around USD 4,600). There is not enough data to identify the reason behind this decline, but the fluctua- tion may have highlighted the existence of untapped revenue potential. Revenue generated from the business tax remained limited, representing less than 1 percent of total revenue. This may be the result of the limited economic opportunity in the commune, but may also highlight a deeper issue: the low level of voluntary tax compliance. Property tax, which is usually a leading source of municipal revenue, is also limited in the commune of Acul du Nord, with a value ranging from 1.45 to 3 percent of the total revenue of the commune. 181 is collected by only 10 municipalities, while manner to finance current and capital expen- the other 130 communes collected only 7 ditures. The amount allocated for opera- percent of these resources, about 83.3 million tional expenditures represent an estimated 15 HTG (See Figure 5). In other words, average to 20 percent of its total transfers. In the case annual tax revenue for a municipality among of project funding, in the FGDCT budget the rural and small communes is only 641,180 there is an amount of around 671 million HTG (or about USD 10,100), which is about HTG whose distribution is the responsibility 53,432 HTG per month (approximately USD of the MICT. There are great differences in 847). It is worth noting that this amount is the allocations provided to communes across insufficient to cover the salary of two mayors regions, and across rural and urban areas. In of a municipality of that size. Limited collec- FY 2015, the average FGDCT allocation was tion capacity of municipalities is reflected about 59 gourdes (or USD 1) ranging from a in collection levels with own-source revenue mere 22 HTG per capita (or USD 0.35) in the highly concentrated in major cities. The five metropolitan area to 163 HTG (USD 2.6) in communes of the Port-au-Prince metropol- the North-West Department. Similarly, the itan area collect more than 80 percent of all northeastern department receives four times the Haitian communes’ own resources. more resources than the Artibonite depart- To further illustrate the ranges of revenues ment. available for most local governments in Haiti Finally, several local governments also outside of Port-au-Prince and a few other benefit from external support for local major cities, a detailed example is provided development. Municipal governments are in Box 2 below. Box 2 shows the trends in eligible to receive funds from various sources revenue collection, indicating the overall including, from international cooperation. reliance on national transfers, the low and Some local governments have either direct unexploited level of local revenue collection collaboration with NGOs and partner cities (around 4 percent of total revenues), and (such as in France) or are benefitting from the insufficient level of overall volumes of support through donor financed projects on financing available to enable communes to local development, local governance, and deliver services. It also provides the trend in infrastructure, among others. In general, revenue collection for a small municipality these programs do not transfer resources in the Northern Region with a population directly to the communes but instead have of about 55,000 people, predominantly rural used a principle of delegated implementa- (76 percent of the population is estimated to tion, often with NGOs, due to the limited reside in rural areas). capacity of the local governments (for Municipalities also receive other support example, in the case of the World Bank-fi- through central government transfers and nanced projects PRODEP and PRODEPUR). grants. This includes subsidies and miscella- The support can be both in the form of neous grants for financing local government financial assistance for service delivery, as spending. As highlighted above, the national well as technical assistance and capacity government through MICT provides regular building in various areas (including revenue transfers to communes in a discretionary mobilization, financial management, devel- 182 opment planning, etc.). In 2016, there were at Taxes collected on behalf of the communes least seven agencies working on local devel- by DGI opment support, including but not limited The Haitian law identifies taxes under the to: the Canadian International Development responsibility of communes – a total of 10 taxes Agency (CIDA), the United States Agency with property taxes and business licenses as for International Development (USAID), the most important ones12. Nevertheless, the the African Development Bank (AFD), the communes are not collection agents (it is European Union, the United Nations Stabi- DGI) nor do they have provisions to set the lization Mission in Haiti (MINUSTHA), the tax rate which falls under the authority of the National Democratic Institute (USA), and central government. Initiative Développement (France). Property tax and business licenses The volume of financing from develop- account for almost all own-source revenue ment partners varies substantially, but in collected. The only taxes with weight from general, local development programs have a quantitative standpoint are property taxes not been national, but rather applied a and business licenses, which account for regional or departmental focus or covered 98.1 percent of total collection (62.5 percent limited technical areas. For example, there of property tax and 35.6 percent of business are several specific support programs licenses) for FY13. A survey of the built targeting revenue collection, including environment (see Figure 6 below) carried LOKAL+ (USAID), which covers about nine out by MICT in 2015 has identified twenty communes for collection of local revenue. communes with high potential to increase Similarly, a program supported by CIDA short-term revenue derived from CFPB. targets the Palme region. In general, most of There has been a significant increase of the programs are oriented towards capacity collection in recent years, albeit from low levels, building and have more limited financing and mainly in the communes of the larger cities available for capital investments. However, (particularly in the Port-au-Prince agglomerate). because of the fragmentation of efforts, it is Overall own-source revenue collection levels difficult to obtain a more precise estimate almost doubled from 517 million HTG for the of the level of financing provided to local year 2010-11 to 757 million HTG (2012-13) and governments through these channels. Going 942 million HTG for the year 2014-15. forward, and with the commitments of the There are vast differences in collection recently installed government, it is expected levels between predominantly urban and rural that more development partners will engage areas. In the Port-au-Prince metropolitan area or re-engage on decentralization and local and Nippes, the share of revenue of business government support, in combination with licenses reaches 35 percent of the total, but support for the reconstruction efforts in the for the rest of the country it is not higher than Grande Anse, following damage caused by 20 percent of the total (see Figure 7). On the the Hurricane Matthew in 2016. other hand, outside Port-au-Prince, the CFPB 12 Property taxes, business licenses, livestock certificate of sales, alignment duties, calibration, cemetery plots, wreck tax /sale, small shop, arbor, ajoupa, character reference, and tax on materials and food products on public road. 183 Source: “Haiti: Steering towards effective fuel subsidy reform” (Perge et al. 2017), based on ECVMAS 2012 data and Haiti Public Expenditure Review (World Bank, 2015a). Figure 6. COMMUNES WITH HIGH POTENTIAL TO INCREASE REVENUE FROM CFPB Sources: MICT, PRAFIPUM. accounts for 80 percent of collection. Never- to increase their revenue collection levels. The theless, it should be noted that these percent- collection level in the Port-au-Prince metro- ages are strongly influenced by the situation of politan area is not comparable to any of the the Port-au-Prince metropolitan area, where 81 other regions. The data shows that there are percent of the total is collected with 22 percent real possibilities to increase revenues in most of the population. Four departments: West of the country’s departments. For example, (including the metropolitan area of Port-au- the differences in collection of revenue per Prince), the South, the North, and Artibonite. capita between the Nippes and South regions The analysis of municipal revenues per are almost a factor of fifty, yet their level of capita (Figure 8) demonstrates the vast differ- development does not explain this situation. ences in collection levels between depart- Hence, there is ample opportunity to expand ments, and the potential for emerging cities the revenue collection levels in Nippes. 184 Source: “Haiti: Steering towards effective fuel subsidy reform” (Perge et al. 2017), based on ECVMAS 2012 data and Haiti Public Expenditure Review (World Bank, 2015a). TAXES COLLECTED BY DGI BY DEPARTMENT (FY 2012-13) Figure 7. Sources: MICT, PRAFIPUM. Finally, it is important to provide further government action program for PFM reforms. information on the legal framework and the A short presentation of the constraints and current application of the property tax system key issues associated with the current tax in Haiti, given its relative importance for local regime is presented below. The government revenue. The ambiguity of the framework and should consider a review of the current law application of the code and regulations for and the way it is applied in a selected number property tax is likely to increase the risks of of communes. This review could then inform losses and inefficiencies in revenue collection, further actions to be taken to make the and therefore requires attention as part of a property tax system function more effectively. 185 Source: “Haiti: Steering towards effective fuel subsidy reform” (Perge et al. 2017), based on ECVMAS 2012 data and Haiti Public Expenditure Review (World Bank, 2015a). Figure 8. REVENUE COLLECTED BY DGI BY INDIVIDUAL AND BY DEPARTMENT IN GOURDES (2012-13) Source: Authors’ figures based on MICT, DGI, and IHSI. Note: The data from the five Port-au-Prince communes are not included. The amount from the Nippes department is so small that it does not appear on the graph. 186 BOX 3 – PROPERTY TAX IN HAITI – THE CONTRIBUTION FONCIÈRE DES PROPRIÉTÉS BÂTIES (CFPB) The CFPB is the leading source of the municipalities’ own resources collected by the General Tax Office (DGI). It is therefore important to understand its main characteristics and limitations. According to Article 1 of the Decree of 1979 on property tax13, the Contribution Foncière des Propriétés Bâties “is a real municipal tax calculated on the rental value of a building. The rental value is the price at which the immovable property is rented or the one to which it can be rented.” In most countries, this tax is computed per the estimated market value of the builtproperties and not based on its rental value. The levy on the rental value is similar to a tax on income derived from renting the property, rather than a tax on capital. Another limitation of the CFPB is that it is restricted to the built environment and not extended to all proper- ties, whether built or not. Thus, this tax leaves out empty lots, which are great source of municipal revenue in other countries. Furthermore, the Article is not clear in terms of the calculation of the rental value and provides a lot of ambiguity. The 1979 Decree proposes to calculate rental value in cases where the property is not rented as 3 percent of the value of the property (about 0.25 percent per month). Compared to international practices where rental values vary in the range of 4 to 6 percent, this percentage is very low in a country with limited housing stock. Article 1 also states that the CFPB is a communal tax even though it is set at the national level by the central government. In practice, this means that the revenue collected is transferred to the municipality. The decree also defines a series of rebates (deductions) related to the policy of urbanization and others that are not clear. In all cases the deductions are high and, in general, these types of deductions create risks for graft and accumulation of tax arrears. Articles 5, 6, 9, and 10 are relevant examples: — Article 5: Any new construction of buildings in cities other than Port-au-Prince and Pétionville will benefit from a tax deduction of the CFPB over four years until full tax payment is required — Article 6: Furnished apartment buildings will benefit from a deduction of the CFPB in proportion to the value of the furniture with deduction not to exceed the third of the annual amount of the tax owed. — Article 9: Exceptionally, any building with multiple apartments (housing units) rented will benefit from a reduction of the CFPB in the order of fifty percent if these apartments are rented furnished and thirty percent if they are rented unfurnished. Under the terms of this decree, the word “apartment” means a unit comprising at least one bedroom, a dining room, and a bathroom. — Article 9: Buildings erected in industrial parks, designed to serve as a place for industrial or artisanal operations, will benefit from total CFPB exemption for the first year of operation and a gradual one for subse- quent years, based on the location of industry with different exemptions for industries inside and outside the metropolitan areas (up to 21 years’ exemption and deductions) — Article 10: buildings belonging to entities are exempt from the CFPB: The State and the communes, religious institutions and nonprofit associations, the Catholic Church or other religious groups recognized by the State and not producing income, and non-income producing congregations, unions, or cultural associations. 13 Article 1 of the Decree published in the official journal Le Moniteur, No. 32-A, April 19, 1979. 187 Taxes collected by the communes The duties and royalties represent a tax The third source of income for the type that is not fully institutionalized, lack communes are revenues generated from duties control and oversight, and is characterized and royalties, collected by the communes by limited transparency and accountability. themselves. The duties and royalties are There are multiple duties and royalties in the set by the municipalities and constitute an various municipalities of the country. In the uncertain source of revenue. In this case, the case of six northern communes14, reviewed in statistical information is not aggregated in the case study prepared for this report, a total financial reporting or at the central level and of 25 duties and royalties have been identi- must be found in each commune. fied. Often, these taxes are not collected or Figure 9. SHARE OF LOCAL GOVERNMENTS EXPENDITURES IN TOTAL PUBLIC SECTOR SPENDING (2015) Source: ECLAC. 14 The six communes covered in the study are Cap Haitian, Limonade, Quartier Morin, Acul du Nord, Plain de Nord, and Milot. 188 only at a minimal level. The most common tory framework that the country is unable to duties and royalties according to the survey attain yet. As can be seen in Figure 9, relative undertaken in the six communes of Cap-Hai- to the region, the participation of municipali- tian are the following: construction permits, ties in government expenditures is very small burial permits, garbage collection charges, in Haiti. advertising, and certificates of attestation. Regarding Haitian budgetary planning, The relative importance of duties and it is done based on the distinction between fees have been low, but in recent years, it operating expenditures and capital expendi- appears that the level of revenue generated tures. In the case of the budget of the central from these sources has increased, and may government, the share of operating expen- potentially continue to do so. Estimates made ditures is estimated at around 76 percent by MICT indicate that this source is used in and the capital expenditures at around 24 various ways depending on the municipality percent. Wages constitute 40 percent of considered, and it represents in the national total spending. MICT does not have aggre- total a value fluctuating around four percent gated data on the distribution of municipal of the total resources of the municipali- expenditures, although it is estimated that ties. However, the estimate of four percent operating expenditures represent between 60 is only representative of the smaller and and 70 percent of the total. In the distribu- rural communes, i.e. Type 3. In department tion of resource programming from FGCDT capitals and in the metropolitan area of Port for fiscal year 2014-15, 846 million HTG is au Prince, it appears that these resources allocated for municipalities and communal have become increasingly important over the sections’ operations, and 670 million HTG past few years. In the case of Cap-Haïtien, for for projects’ funding (i.e., approximately 56 example, duties and fees comprised almost percent for operating expenditures and 44 thirty percent of the commune resources over percent for capital expenditures). Unfortu- the period of 2013-16. nately, there was no information available concerning actual expenditure. However, in Local government expenditures and the Cap-Haïtien agglomerate, the expendi- controls ture distribution across the six communes Spending of local governments remains at covered in the study is very different (see levels considerably lower than those of other Figure 10) – on average, the capital expendi- Latin American countries. The spending of tures represent only 20.6 percent of the total the territorial collectivities represents only spending. If we differentiate between the 3 percent of total public spending, with, 2.7 Cap-Haïtien commune and the five other percent being at communes’ level and 0.3 smaller, less urbanized communes, one percent at the municipal sections. From observes different results as evidenced in this purely quantitative standpoint, it is the figure. Given the data available (absent obvious that the decentralization process expenditure data at the national level), it is in Haiti is very weak in connection to the likely that the actual distribution across the rest of the countries of the continent. This country is closer to the levels seen in the situation coexists with an ambitious regula- region of Cap-Haïtien. 189 Source: “Haiti: Steering towards effective fuel subsidy reform” (Perge et al. 2017), based on ECVMAS 2012 data and Haiti Public Expenditure Review (World Bank, 2015a). Figure 10. EXPENDITURES DISTRIBUTION IN THE CAP-HAÏTIEN AGGLOMERATE Source: Municipalities’ accounting department. LACK OF TRANSPARENCY AND LIMITED and the law of May 28, 1996 constitutes RELIABILITY OF THE TRANSFER SYSTEMS the legal basis for fiscal transfers from the EXACERBATE FINANCIAL CONSTRAINTS central government to municipalities. The Challenges from high dependency FGDCT was created to facilitate access by from transfers are exacerbated by lack local and regional authorities to financial of transparency and limited reliability resources that can enable their develop- of the existing transfer system. In Haiti, ment. The law established the fund, its the decentralization framework decree of funding sources, and its management February 1, 2006, the law of August 20, 1996 structure. The framework decree calls for on the Local Government Development ordinary and extraordinary fiscal transfers, Fund (Le Fonds de Gestion et de Développe- but the modalities of these transfers are ment des Collectivités Territoriales [FGDCT]), not well-defined, nor are they publicized 190 on a regular basis, and statistics on actual accordance with the provisions of the law, and transfers from the central government to the management of the FGDCT is therefore communes remain limited and opaque. the de facto responsibility of the MICT. Further, the management of FGDCT is not The allocation and disbursement of central fully in compliance with the legal framework, transfers is not implemented in full accor- and the overall level of fund mobilization dance with existing regulations. Only half could be improved. According to the legal of the funds designated for the communes framework, the fund management board are transferred, and a significant share of is composed of the three members of the funds is transferred to structures that are not Inter-departmental Council, assisted by a active. The amount of the FGDCT disburse- representative of the MEF, a representative ment foreseen for the 2014-15 budget was of the Ministry of Planning, and a represen- 2,300,293,375 HTG (around USD 37 million). tative of MICT. The objective of the fund is to Given that total internal revenues anticipated ensure the efficient functioning of the terri- in the FY15 budget was 60,130,190,000 HTG, torial authorities. According to the law, this it appears that fund disbursement represents fund is financed by tax collections by DGI only 3.8 percent of the total internal revenues on a series of products or activities: sale of of the central government. Out of the total cigarettes, insurance premiums, international expenditure forecast of the FGDCT, only phone calls, customs invoices, international 1,129,501,955 HTG (49.1 percent) are directly travel ticket purchases, tax withholding on assigned to the territorial collectivity bodies payroll, withholding of net taxable income of as specified in the law (see Table 5). It is any taxpayer, tax withholding on lottery wins worth noting that 170,130,000 HTG (around and other games.15 These levies are designed USD 2.7 million) are assigned to bodies that to finance operations of five local govern- have not been functioning for several years ment bodies specifically mentioned by the or have never existed (the interdepartmental law: organization of meetings of the territorial council, which is assigned HTG 6,000,000). collectivities, communal sections (CASEC), The basis for allocating central govern- municipalities/communes, the departmental ment transfers needs review. Lack of councils, and the interdepartmental council. predictability in central government Out of the five bodies, there is one that has transfers impedes on planning, budgeting, never been created to date (the interde- and service delivery in local government. partmental council), and two that are not FGDCT allocations to local governments are operational (the assemblies of the territorial disbursed monthly and are used primarily to collectivities and the departmental councils). cover operational costs (63 percent). Salary This means that, currently, the management accounted for 76 percent of the operational board of the FGDCT is not established in costs. Only 23 percent of the fund expendi- 15 As the 1996 law prescribed, the FGDCT resources come from (a) 20 percent of the cigarette tax, (b) 5 percent of the tax on insurance policies, (c) 8 HTG per international mobile telephone call, (d) 2 percent of the customs tax (with some excep- tions), (e) 25 HTG per international airplane ticket, (f ) 1 percent of the income tax on salaries greater than 5,000 HTG, (g) 1 percent of the net revenue from all taxpayers, and (h) 5 percent of the winnings from lotteries and other games. 191 Table 5. DISTRIBUTION OF THE FGDCT ALLOCATIONS LINE AMOUNT (HTG) AMOUNT (USD) Expenditures of the bodies of the territorial collectivities 1,129,501,955 18,217,773 Support to current expenditures 77,000,000 1,241,935 Project Tablefunding 4: Revenues Collected 699,891,420 11,288,571 by DGI on Behalf of Communes Subsidies to territorial collectivities association 40,500,000 653,226 in the Metropolitan Area of Port-au-Prince Other expenditures (in HTG), 267,000,000 4,306,452 October 2013-March 2015 Source: Other MEF. bodies 86,400,000 1,393,548 Total 2,300,293,375 37,101,506 Source: National Budget 2014-15. Supplementary Documents. tures were spent on investment projects (see of the fund and transform it into an effective Table 6). In the absence of the interdepart- fiscal transfer mechanism to finance service mental council, allocations decisions rest delivery at the local level. solely with the central government through Lack of prudent, effective, and efficient the MICT, without consultations with local financial management of the local govern- governments. In theory, 40 percent of funds ments meager resources further impedes in the FGDCT are earmarked for capital service delivery provision. In most investments, but there is no clear guidance communes, the budget lacks credibility, as and criteria for the selection of these projects. it is considered more as a legal requirement According to data received from the Ministry to receive funds from the central govern- of Interior, average total central government ment rather than a management tool. This transfers to local government for all 140 has usually resulted in poor expenditure municipalities are estimated to be less than controls, inadequate reporting, and an 600 million HTG. This amount is grossly incentive to spend the budget allocation as insufficient to cover the costs of operation of soon as possible, thus lacking focus on actual local government, let alone provide needed service delivery. As such, revenue forecasting financial resources to invest in local infra- is not realistic and expenditures often structure. There exists a window of opportu- exceed revenue collection by more than ten nity to review current management practices percent, leaving some communes in a state 192 DISTRIBUTION OF FGDCT EXPENDITURE IN 2015-16 (FY), IN HTG Table 6. OVERALL EXPENDITURE AMOUNTS TOTAL TOTAL (USD) CATEGORY CATEGORY Operations Cost Salary 891,716,950 1,161,551,950 18,405,720 Rent 31,800,000 Commune 77,425,000 Holidays (Patronales) Other Subsidies 160,610,000 Capital Projects 415,005,663 415,005,663 6,576,097 Expenditure Others Recurrent Costs 67,000,000 260,899,400 4,134,160 Subsidies to May- 38,500,000 ors Associations Others 155,399,400 Totals 1,837,457,018 29,115,976 Source: MICT. of continuously accumulating payments lack of external controls increases the risks arrears to service providers. In addition, for accountability gaps and potential misuse salary and other fixed administrative costs of funds. dominate the expenditure patterns of local The government has recognized the gaps, governments. While some monthly financial and fiscal decentralization is an important reporting is required by law, municipalities part of the government’s reform program do not produce annual financial reports and on public financial management (PFM). the Cour Supérieure des Comptes (Supreme If implemented effectively, this program Audit Institution) lacks both the financial could help the government in addressing and human resources capacity to carry out some of the key deficits. Since May 2014, the annual audits of the 146 communes. This government has developed a comprehen- 193 A PATH TO STRENGTHENING MUNICIPAL sive PFM reform strategy and action plan, FINANCES organized around six main pillars related to budget management, treasury management, There is a clear need to strengthen internal and external controls, revenue municipal finances for cities to be able to mobilization, financial decentralization, and close the urban infrastructure and services financial information systems. The vision gap, as well as to accommodate the growing of the PFM reform strategy is to ensure a urban population. As detailed in Chapters 1, public finance system that promotes trans- 2, and 3, Haitian cities are marked by high parency, accountability, fiscal discipline, deficits of basic services, including supply and efficiency in the management and use of of water, sanitation, waste collection, and public resources for improved service delivery electricity. Poor city management is another and economic development. The program defining feature, which negatively impacts the aims at increasing revenue from taxes and way urban areas are expanding, since local tariffs, thus increasing the autonomy of local governments are insufficiently equipped to government entities in decision-making as guide growth away from crowding and toward well. During this period, important regula- “healthy” and productive densities. Based on tory measures have also been implemented the review of municipal finances in Haiti, to improve financial management at the this section offers a set of priority measures local government level, such as standardiza- to mitigate the identified shortcomings tion of accounting processes in local govern- described above. ment, regulation of communal own-source Despite efforts by the government to revenue collection, review and amendment address fiscal decentralization in the broader of the local finance law, creation of an inter- PFM reform action plan, many weaknesses in communal fund, etc. The implementation local government PFM remain. The Govern- of a fiscal decentralization action plan has ment will need to review the existing roadmap not progressed at the desired pace due to for reforms to address PFM at the local level, a combination of financial, logistical, and including prioritizing areas that see limited human resources issues, as well as the period progress, as well as potentially reconfig- of transition in 2015-2016. Therefore, to date, uring and updating the existing roadmap Haiti has made limited progress in terms of by considering the progress made in some increasing fiscal and financial autonomy of areas. In addition, it is imperative to review local government entities. the existing legal and regulatory framework to clarify the roles and responsibilities of each level of government for the delivery of services and the provision of infrastructure at the local level and to set out actions to address weaknesses in systems, capacities, 16 PMAC – Programme de Modernisation des Administrations Locales. This is a program conceived by MICT and currently under implementation. 194 incentives, and accountability that currently tax) is obsolete and outdated, and therefore impede service delivery. Since intergovern- impedes the capacity of municipalities to mental transfers (particularly the FGDCT) increase revenue collections. A key require- have not been based on a transparent formula ment for improved local service delivery and clear objectives and criteria, the govern- is that local governments receive greater ment should consider reviewing the existing funding for basic service provision. In turn, regulatory framework for the FGDCT and this means that all local governments have develop a framework for financing local basic budgeting, accounting, reporting, and government provision of services and infra- procurement functions in place so that they structure. This should include outlining the can effectively use this funding to improve principles for strengthening the management service delivery. Increased funding must also and operations of FGDCT and considering go hand-in-hand with improved account- other options (such as additional grants) ability for the use of funds. Therefore, a that need to be put in place to adequately review and clarification of the current local finance local service delivery. Finally, it is finance legislation in Haiti could pave the 16 advisable to strengthen the PMAC in order road for boosting tax collection. to provide a human resources framework for local governments that articulates a set of Consolidate, harmonize, and enforce the principles, systems, and practices that should legal and regulatory framework for munic- be developed and followed to ensure that local ipal financing governments are adequately staffed. Haiti’s urban development is taking A more efficient, accountable, transparent, place under a context of incomplete decen- and fiscally responsible decentralization tralization and an unclear legal framework system will need municipalities with stronger for municipal financing. The basis for the and broader sources of own revenue. There effective implementation of decentraliza- is a lack of significant revenue autonomy at tion, one that provides clarity of roles and the municipal level and a large dependency resources, is a strong legal and regulatory on central government transfers. However, framework. Three main courses of action evidence suggests that current levels of have been identified. First, it is necessary to transfers remain extremely inadequate given review the normative framework of the terri- the broad responsibilities of municipalities. torial collectivities as established in the five This creates a situation whereby local author- decrees of 2006 and identify possible actions ities are held responsible by their constitu- for implementation. Second, the govern- ents for mandates that are not adequately ment must formalize the taxation functions funded. Legislation and administrative and responsibilities entrusted to the munic- requirements underpinning the collection of ipalities as stipulated in Article 142 of the municipal taxes (property tax and business decentralization framework.17 Thirdly, there 17 Framework for decentralization passed in 2006. Article 142 of the same decree grants to the commune competencies in terms of mobilization and collection of the CFPB, participation in the mobilization and follow-up on business licenses, and the creation of some duties and municipal royalties. 195 is an urgent need to undertake a review of to advocate for a special law on local finance key legislation and regulations, particularly and measures to increase tax and non-tax those related to property tax and business revenues, accruing to local government. tax (for example, address the issues around In addition, the government is currently rate setting and the fact that many of the tax piloting a series of programs designed to rates are outdated and not in line with current strengthen administrative decentralization market trends). Together, these actions seek (such as PMAC [Programme de Modernisation to clarify the responsibilities, systems, incen- des Administrations Locales]) and enhance tives, and accountability relationships for revenue collection at the communal level. the delivery of services, the financing needed for services, and the capacity of local govern- Strengthen the system for municipal finance ments to effectively use increased funding, and expand financing opportunities and so lead to improved service delivery. As emphasized throughout the chapter, The lack of detailed regulations about cities are constrained by limited revenue the different local government levels has sources and are highly reliant on transfers limited the effectiveness of the decentral- from the central government. Reduced ization reforms, but a new emerging legal financial autonomy of local governments reform work map opens further opportuni- calls for a stronger municipal financial ties for deepening decentralization efforts. system. For smaller cities, efforts can be While the innovation introduced by the five focused on fixing the transfer system and decentralization decrees adopted in 2006 using it as an opportunity to build local was commendable, other pieces of legislation capacity for implementation. For larger important for implementation of the decen- cities, steps are needed in the direction of tralization structures are critically lacking. building capacity for own-source revenue Regulations about the operation procedures collection, management, and spending. for both the delegations of the communes and First, the government can focus on enforcing the departments would need to be enacted. A the regulatory framework of the FGDCT to draft law on decentralization and autonomy enhance the management, oversight, and of communes and communal sections (so that transparency of the fund, including fund local governments can become financially mobilization, allocation, and transfer, as well autonomous) is currently in the legislative as expenditures and accounting. Specifically, agenda. The passage of this bill will be critical it must ensure that the funds are mobilized for further consolidation of decentralization. and allocated based on the criteria stated Also, the government is considering a law in existing legislation; review the param- amending the decree establishing the general eters for allocation of funds to communes framework for decentralization and the and other local authorities; and develop principles of organization of the Haitian terri- tracking mechanisms to ensure that the torial communes (on staffing). It is important FGDCT is adequately replenished or collects that decision makers seize the opportunity to all mandatory contributions. To address specify as clearly as possible the functions and the lack of timeliness and predictability in competencies of local government, as well as finance levels and volumes, special attention 196 should be given to improving the coordina- including forecasting of revenue. The first tion mechanisms between the communes, set of activities to improve the financial DGI, Ministry of Finance, treasury, and management functions of local govern- the central bank for tax collection to reduce ments is to design and roll out a Local delays in releasing the funds to municipali- Government PFM Manual. The Ministry ties. Finally, it is also important to clarify the of Finance and Economy and the Ministry status and future implementation of other of Interior should develop the manual to set types of government funds, including the out the basic standards and procedures for Fonds de Développement Local et Aménagement local government financial management, du Territoire (FDLAT). including budgeting, accounting, reporting, procurement, and audit requirements. This Expand and leverage the local revenue base could be achieved through capacity-building There is a lack of significant revenue programs that focus on four main areas: autonomy at the municipal level and a large strengthening the administrative capacity of dependence on the transfer system, which is the financial units in municipalities; strength- embedded with structural problems and is ening the capacity of municipalities in project reliable only to a certain extent. Additionally, management for timely disbursements of municipalities often collect only a fraction FGDCT funds allocated to them; increasing of their revenue potential and are without municipality revenue mobilization capacity, proper capacity to effectively use resources including enhancing staff technical compe- for the delivery of local services. A more tencies; and providing municipalities with efficient, accountable, and fiscally responsible incentives to explore alternative financing decentralization system will need municipal- mechanisms, including partnerships with the ities with stronger and broader sources of local private sector and hometown associa- local revenues. Currently, the main source tions of the diaspora. of revenue in municipalities (the property In terms of improving tax collection tax rate and business tax receipts) are not capacity, the government’s computeriza- reflecting current market trends. In addition, tion of systems can improve collection rates. collection of user fees and other duties The government could consider progres- collected directly by municipalities critically sively automating nationwide tax collection lack transparency, and mechanisms to track functions by deploying CIVITAX18 in munic- the collection and account for its use are ipalities, taking advantage of all the function- nonexistent. alities offered by CIVITAX, ensure the In moving toward municipalities that progressive characteristic of property taxes are able to generate and effectively collect (with a higher scale for the properties of greater their own revenue sources, efforts must be market value), carry out property assessments geared toward strengthening the planning in all municipalities, and update the property and budgeting capacity in municipalities, tax registry accordingly to broaden the tax 18 CIVITAX is a comprehensive local tax management software system created by the Haitian firm Solutions S.A., in collabora- tion with the LOKAL project, the MICT, the DGI, and other local partners. 197 base. Efforts in this direction can be taken while at the same time facilitating legal and technical mechanisms for municipalities to carry out a census and appraisal of proper- ties. In the communes where a census was carried out, the collection increased consid- erably. The MICT PRAFIPUM program (Program for the Improvement of Municipal Public Finance) is a step in the right direction. However, detailed revenue mobilization action plans with pragmatic revenue targets need to be developed in the twenty communes with the highest potential to increase revenue derived from property tax. Also, to combat tax avoidance, attention should be paid to the taxpayer’s registry and efforts should be devoted to constantly update it. Finally, it is also important to minimize taxpayer exemp- tions and deductions. Improvements in regulation and efforts to follow up and ensure compliance can improve control over municipal budgets and increase efficiency in the allocation of resources. To exert greater control over the municipal budget and ensure that it is allocated to the most efficient use, the following steps can be taken: establishing a regulatory mechanism to limit budget spending commitments and adhere to available resources to ensure the smooth implementation of the budget and avoid the accumulation of arrears; ensuring that municipalities adhere to the regulatory deadlines for drafting and approving the budget; reinforcing existing mechanisms to engage citizens and local civil society in the budgeting process; and encouraging and supporting municipalities to devise a communal development plan accompanied by multi-annual investment plans, which can feed easily into the budgeting process. 198 REFERENCES Port-au-Prince, Ministère de l’Intérieur et des Collectivités territoriales/USAID LOKAL. Beaudin, R. 1998. Séminaire atelier sur la Moniteur. (2006, June 2). Le décret du 1er Planification Décentralisée: La fiscalité février 2006 fixant l’organisation et le locale en Haïti. Miméo. fonctionnement de la collectivité territoriale Chery, F.G. 2009. Le financement de la décen- municipale dite commune ou municipalité. tralisation et du développement local en Moniteur. (2006, June 14). Le décret du 1er Haïti. Secrétariat Technique d´élab- février 2006 fixant le cadre général de la oration du Document de Stratégie décentralisation ainsi que les principes Nationale pour la Croissance et la d’organisation et de fonctionnement des Réduction de la Pauvreté (DSNCRP). collectivités territoriales haïtiennes. Denis, J. 2014. Haiti, Local government Paillant, J. 2012. Le code fiscal haïtien, Henri system fiduciary assessment, Haiti: Deschamps, Port-au-Prince. USAID. Paul, B. and C. Charleston. 2014. Les Denis, J. 2016. Public financial manage- collectivités territoriales : analyses du ment: Constraints and opportunities processus démocratique bloqué en Haïti. for CDCS. Haiti: USAID PRIMATURE/OMRH. 2013. Programme- Deshommes, F. 2003. Décentralisation et cadre de la réforme de I ‘Etat: Réforme collectivités territoriales en Haïti, un administrative et décentralisation- état des lieux, Port-au-Prince, Editions Octobre 2012-Septembre 2017, Port-au- Cahiers Universitaires. Prince, Office de management des Farvacque-Vitkovic, C.D. and M. Kopanyi. ressources humaines (OMRH), 2014. Municipal Finances: A Handbook Primature, République d’Haïti. for Local Governments. Washington, Privert, J. 2006. Décentralisation et collec- DC: World Bank Group. tivités territoriales (contraintes, enjeux et Grand-Pierre, C. 2015. Gouvernance de défis), Edition Le Béréen. proximité et collectivités territoriales en USAID. 2015. Fact Sheet: Revenue Collec- Haïti, C3 Editions, Port-au-Prince. tion Provides Opportunity. Available Le Nouvelliste (July 2014). Bilan de l’écon- at https://www.usaid.gov/results-data/ omie haïtienne (Année 2012-2013 Le success-stories/revenue-collection-pro- mot de l’Association Haïtienne des vides-opportunity. Accessed on 3 July, Economistes-AHE), publié le 01-07- 2017. 2014 World Bank Group. 2016. Better Spending, MICT. (May 2007). La municipalité haïtienne Better Services: A Review of Public et les entités de l’administration publique. Finances in Haiti. Washington, DC: MICT. (June 2008). Projet de réforme des World Bank Group. finances municipales : Propositions pour une meilleure mobilisation des recettes communales (version préliminaire). MICT. 2011. Recueil de textes normatifs annoté, entourant l’action locale, 199 ANNEX 1 – CLASSIFICATION OF CITIES BASED ON GRIDDED POPULATION DATA AND NIGHT-LIGHT DATA CITY CATEGORIES # OF CITIES DESCRIPTION Metropolitan Area 1 Moving beyond official administrative boundaries allows us to identify Port-au- Prince as a large metropolitan area that comprises some of the municipalities in the arrondissement of Croix-des-Bouquets. In line with previous results, the “Great- er PaP” remains by far the largest urban cluster, accommodating 2.5 million inhab- itants – equivalent to 43 percent of Haiti’s population in cities. Between 1990 and 2015, the urban population of the “Greater PaP” increased each year on average by 3 percent. Very Large Cities 2 Cap-Haïtien and Gonaïves are “very large cities” with urban populations exceed- (200,000-300,000) ing 200,000 inhabitants. Between 1990 and 2010, the share of urban residents in- creased by 4 percent in Gonaïves, and by 2 percent in Cap-Haïtien. Large Cities 5 “Large cities” with more than 100,000 inhabitants are Port-de-Paix and Saint- (100,000-200,000) Marc, which expanded at a yearly average rate of 4 percent since 1990. Among these third-class cities, we find the agglomeration of municipalities Cayes-Tor- beck-Chantal-Camp-Perrine in the Sud département, and Ouananaminthe in the Nord-Est département. Finally, as the municipality of Léogâne forms together with Gressier – officially, part of Port-au-Prince – another large conglomerate, we ob- serve that the metropolitan area stretches eastward, as well.1 Medium-size Cities 3 Petite-Goave, Arcahaie, and Jacmel are “medium-size” cities. In the case of Pe- (50,000-100,000) tite-Goave, annual growth averaged at 9 percent between 1990 and 2015. Small Cities 6 These are municipalities with positive growth rates in their urban share. For in- (10,000-50,000) stance, in Anse-à-Pitre (Sud département, on the border with the Dominican Re- public), the urban population increased by 8 percent on average every year since 1990. After the 2010 earthquake and the 2015 deportation of Haitians from the Dominican Republic, informal settlements were precariously established, leaving thousands of households without water, access to basic services, and eventually exposed to cholera outbreaks. Small towns 2 Two of the smallest urban clusters have experienced population loss since 1990 – (1,000-10,000) like Petite-Bois, in Croix-des-Bouquets, possibly suffering competition from other neighborhoods closer to the metropolitan area. 1 It is important to note that while the last two categories are not officially classified as “cities,” they are clearly shown to be important areas in terms of night light emission, population concentration, and urban footprint. 201 ANNEX 2 – POPULATION (URBAN VS RURAL) EXPOSED TO EROSIONS, LANDSLIDES, AND FLOODS, 2000 AND 2015 Erosion 2000 Erosion 2015 Population Exposed Population Exposed Landslide 2000 Landslide 2015 Population Exposed Population Exposed Flood 2000 Flood 2015 Population Exposed Population Exposed 202 ANNEX 3 – CLASSIFICATION OF CITIES BASED ON GRIDDED POPULATION DATA AND NIGHT-LIGHT DATA LAND DEVELOPMENT, CONSTRUCTION PERMITS, AND GUIDING PRINCIPLES OF URBAN DEVELOPMENT 1923 Law pretaining to roads and public ways. 1924 Law introducing construction regulations in cities. 1937 Decree initiating special rules on housing and town management. 1963 Establishes land use regulations and regulation for the management of urban areas. 1971 Decree of March 23 modified Article 29 of Décret-loi of July 22, 1937 with respect to construction permits (and introduced public use zones). 1977 Decree of April 6, 1977 does not modify the procedures for allotment established under the laws of 1937 and 1963, but adds procedures for land development realized by a syndicate or society/collective of prop- erty owners. 1982 Decree of January 6 on subdivisions, plots, and parcels pertaining to size, conditions, and procedures. LOCAL PLANNING INSTRUMENTS 1996 “Organisation de la Collectivité territoriale de Section Communale,”, introduced “local urban plans” and commune development plans. There are no examples of these plans having been implemented in practice (Garry Lherisson 2015). 2006 Decree providing for local territorial management plans and local development plans. There are a handful of cases where the latter plan has been developed where financing and techni- cal assistance were available (Garry Lherisson 2015). LEGAL FRAMEWORK FOR EXPROPRIATION OF LAND 1842 The principle of expropriation has been enshrined in every Haitian constitution since 1843, with the exception of the constitution of 1888. 1951 Law of September 1951. 1979 Law of November 1979, Article 10 (defines instances of “public good”). 1987 The constitution establishes the process for expropriation and compensation. Stipulates property be returned to the original owner if the project is abandoned (Articles 36-1). LAND DISPUTE RESOLUTION 1987 The constitution mandated the Conseils d’Administration des Sections Communales (CASECs) to deal with land disputes; high court (Court de Cassation) and magistrates (Justice de Paix) can call tribunals for these conflicts. National Agrarian Reform Institute (Institut National de la Ré- forme Agraire [INARA]) also engages in rural land issues. There is no official channel to address errors in land registration system, and no data on number of disputes at present. 203 ANNEX 4 – POPULATION (URBAN VS RURAL) EXPOSED TO EROSIONS, LANDSLIDES, AND FLOODS, 2000 AND 2015 PLAN NAME TYPE SCOPE Strategic Development National Plan This plan builds on the national reconstruction plan of March Plan of Haiti 2010, with long-term vision of development that looks ahead (PSDH, 2012) to 2030. National Housing Sectoral Plan This plan aims to support access to safe, serviced housing. Plan (2013) Identifies major challenges including lack of: building con- struction norms and enforcement; oversight and regulation of individual construction efforts; skilled labor; urban planning; and housing finance. Transport Sector Sectoral Plan Aimed to provide strategic direction for short-, medium-, Strategy (2006-2011) and long-term transportation policy to support national economic development Water and Sanitation Sectoral Plan Strategy developed by National Drinking Water and Sanitation Sector Strategy 2010 Directorate (DINEPA) MASTER-PLANNING EFFORTS FOR PORT-AU-PRINCE 1976 Plan de développement de Port-au-Prince” 1987/8 “Plan Directeur d’Urbanisme de Port-au-Prince, phase I” 1982-1994 Urban growth management studies for the metropolitan area (no specific plan realized). 2001-2003 ‘Schéma Directeur d’Aménagement de la Zone Métropolitaine de Port-au-Prince’ and a new ‘Plan Stratégique de Développement’ (financed by the Inter-American Development Bank) 2010 Onward, many studies and plans were realized with support from NGOs (e.g., “Plan for the Reconstruction of the Centre of Port-au-Prince,” the “Plan d’Aménagement et de Développe- ment Durable,” and a “Plan Local d’Urbanisme du Nouveau Port-au-Prince”). 2011-2014 Two UN Habitat-supported participatory planning forums, the latter of which resulted in the “Nation- al Urban Forum Declaration.” 204 LOCAL SECTORAL PLANS 2010 Plans communautaires pour l’aménagement post-séisme de quartiers de l’Airemétropolitaine de Port- au-Prince. This plan provided micro-zoning. 2011-2013 Micro-zoning for Port-au-Prince (UNDP-supported initiatives) (i) Location of exposed assets (georeferenced critical infrastructure like schools, hospitals, and roads) (ii) GIS Maps developed by ministries of education, health, and public works in conjunction with CNIGS MASTER PLANNING EFFORTS FOR CAP-HAÏTIEN 1982 Urban Growth Study 1994 Urban Growth Study 1996 -1998 UNDP-supported project or prepared background studies for master plans for Cap-Haïtien, Fort-Lib- erté, Port-de-Paix, Gonaïves, Saint-Marc, Hinche, Miragoâne, Jacmel, and Cayes et Jérémie 1997 “Schema Directeur d’Amenagement Urbain” of Cap-Haïtien (financed by ACDI-Canada, executed by the consulting firm DESSAU) 2012 /3 The “Plan d’Aménagement du Nord/Nord-Est: Couloir Cap – Ouanaminthe, “ developed by CIAT 2012 /3 “Plan Stratégique Multisectoriel d’Aménagement” (MPCE with support from UNDP) 2012 /3 “Cap-Haïtien – Ouanaminthe Development Corridor: Regional Comprehensive Plan” (developed with support from the IADB) LOCAL SECTORAL PLANS 2008 Tourism plan for Cap-Haïtien. Developed by the Ministère du Tourisme, this plan links tourism with urban devleopment (for example, suggesting land use admendments, a metropolitan road network to structure the future urbanization, waterfront redevelopment according to different zones, redevel- opment of the estuary of Bassin Rodo into a touristic circuit, and treatment of the banks of the main rivers as linear parks). 2007-2009 Strategic plans for the water sector (co-funded by the ACP-EU, Oxfam-UK, PROTOS, and GTIH) 2011 The Plan de Financement de Services Publics Communaux (output of the USAID LOKAL+ project 2011) 2011-2015 “The Conception et Aménagement d’un Centre de Gestion Intégrée des Déchets Solides (CGIDS)” (supported by AFD) 2015 Northern Corridor Sustainable Mobility Plan (IADB) 2015 The Northern Development Corridor Haiti: Urban Development and Climate Change Study (ERM with IADB, February 2015) 205 ANNEX 5 – CONNECTIVITY ON THE COSTS OF FRAGMENTED LABOR MARKETS AND THE BENEFITS OF ACCESSIBILITY Cities are often referred to as centers of opportunity because they reduce the economic distance between workers, employers, buyers, and sellers. This density reduces economic distance, decreases the unit cost of supplying infrastructure such as transport networks, allows ideas to circulate, grow, and mature, and promotes economic specialization whereby people can focus on what they do best or enjoy the most (Glaeser and Kahn 2004; Ciccone and Hall 1996). These advantages of cities explain why urban residents are generally more productive than workers outside urban areas, what economists commonly call agglomeration economies that stem from proximity and exchanges (Duranton and Puga 2004; Rosenthal and Strange 2004). Cities are not always associated with high levels of employment accessibility. Although cities concentrate on a restricted area of land, large numbers of households, and employment opportu- nities, this simple geographic proximity does not imply that all households have an equal or even good access to jobs. In Mexico, for example, urban sprawl and massive population decentralization into single-use residential and peripheral settlements combined with uncoordinated urban growth has widened the distance between jobs and housing. This undermines effective matching of skills BOX 1 – AN ACADEMIC TAKE ON ACCESSIBILITY, PRODUCTIVITY, AND WELL-BEING – GATHERING THE EVIDENCE Prud’homme and Lee (1999) demonstrate that worker productivity increases with the share of jobs in the urban area which are accessible within a 30-minute trip. Their case study investigated this relationship in 23 French cities and showed that a doubling of the share of jobs that could be accessed within that timeframe was associated with a 15 percent increase in productivity (an elasticity of 0.15). From an employer’s point of view, Cervero (2001) shows that in the San Francisco Bay area there is weaker yet, nonetheless, positive evidence that the number of workers that can access work within a defined peak period ranging from 30 to 60 minutes (labor markets) positively influences worker productivity levels. Melo, Patricia C., Daniel J. Graham, David Levinson, and Sarah Aarabi (2013) show for a sample of US cities that increasing accessibility to jobs results in increased productivity as measured by real wages. The authors report that a doubling of the number of jobs accessible per worker within 20-minute thresholds result in an average increase in real wages of 6.5 percent. Aslund, O., J. Osth, and Y. Zenou. (2010), using a natural experiment in Sweden, also find that the proximity to jobs positively impacts employment prospects even in the long term. Venables (2017) identifies that commuting costs and thus average accessibility matters in urban settings. In order to attract skilled workers, firms must compensate for their travel costs by offering higher wages. While this will benefit households, it can prevent firms from reaping productivity gains and entering international markets because the average wage needed to compensate workers is higher than competitive international standards. 206 to jobs in Mexican cities. It can also lead to lengthy commutes for those that have found jobs in the main job centers primarily located downtown. In Mexico City, low-income households living in the peri-urban areas can spend an additional four hours commuting per week compared to low-income families residing in more central areas (Kim and Zangerling 2016). This Annex provides a detailed description of distribution of economic opportunities in Haiti with a focus on Port-au-Prince and Cap-Haïtien. THE DISTRIBUTION OF ECONOMIC OPPORTUNITIES AND PEOPLE The overall job structure of Port-au-Prince is very similar to its residential structure, when jobs are identified from areas with strong daytime activity as described in Box 2 (Chapter 3). Apart from significantly higher daytime population densities in the above identified sub-centers of Port-au-Prince, namely Carrefour, National Palace, Pétionville, and Croix-des-Bouquets, the two maps of Figure 1 look very similar.1 SPATIAL DISTRIBUTION OF ECONOMIC OPPORTUNITIES (LEFT) AND Figure 1. RESIDENCES (RIGHT) IN PORT-AU-PRINCE Source: Authors’ elaboration using Digicel data. 1 The fact that observations show only small differences in daytime and nighttime populations can be explained by a number of factors: the sample of callers captures elderly or young people unlikely in both cases to move beyond the threshold distance of a cluster; the sample captures unemployed active people who would not have a strong motivation to travel far; and the sample captures people who either work from home or work close to home (i.e. within the 1 km radius from home that was used to define home and work clusters) so that their home and job locations would be considered the same cluster. 207 Figure 2. DISTRIBUTION OF COMMUTERS DURING DAYTIME (LEFT) AND DURING THE EVENING (RIGHT) IN PORT-AU-PRINCE Source: Authors’ elaboration using Digicel data. Notes: This figure shows that commuters flow to downtown Port-au-Prince for work and related purposes (left) and come from everywhere in the urban area. Figure 2 shows that in this case, most daytime population densities tend to cluster around the National Palace with up to 90,000 people/km2, i.e. around 1.5 times its residential population. Pétion- ville also sees a net increase in population densities during daytime indicating job-related attraction. Mean population density in Pétionville increases by 27 percent to just above 50,000 persons/km2 from just 40,000 persons/km2 during the evening. Likewise, some other areas to the northeast of the city center also experience net increases in population densities during daytime, for example Saint Martin. Farther north, the center of Croix-des-Bouquets sees a net increase of 39 percent during daytime. On the other hand, other areas are notably residential. Carrefour and Canaan are two of those areas that see net decreases of population densities during daytime of 8 and 30 percent, respectively. The business district right in the center of Cap-Haïtien strongly dominates any other daytime desti- nation, with daytime population densities reaching up to 80,000 people/km2. In comparison, the other high-density location within Cap-Haïtien, Petite Anse, sees significantly fewer commuters than the center of the city, with daytime population densities actually dropping by 24 percent compared to evening. Driving south of the Mapou River, daytime densities decrease at a faster rate than during nighttime, indicating again that commuters tend to leave these places for work. 208 DISTRIBUTION OF COMMUTERS DURING DAYTIME (LEFT) AND DURING Figure 3. THE EVENING (RIGHT) IN CAP-HAÏTIEN Source: Authors’ elaboration using Digicel data. COMMUTING FLOWS It is worth analyzing the flow of people from each of the buffers in order to understand usual commuting behavior. Figure 4 depicts the flow of commuters from each specified buffer. The column on the left depicts the daytime (work) location of the individuals who live in the selected buffer. The column on the right depicts the evening time (home) location of the individuals who work in the selected buffer. For instance, for the first row, the column on the left shows where those people living in the center of Port-au-Prince go to work, whereas the column on the right shows where those people working in the center of Port-au-Prince live. The first row of Figure 5 depicts commuters from the center of town. It is possible to see that most commuters who live in this area travel to Delmas and Pétionville during the day. The share of commuters in the center of Port-au-Prince is rather small, which is likely because most people who live in the center of Port-au-Prince also work there. The center of Port-au-Prince sees a huge influx of people from different areas during the day including Carrefour and Martissant on the west, Pétionville on the southeast and Delmas on the northeast. The second row of Figure 5 reiterates that most people within a buffer of at least one and at most five kilometers from the city center will commute to the city center to work, and a significant share will also commute to Pétionville. There is a considerable number of people who work and live in the same buffer, as depicted in the left pane of Figure 12 in chapter 3. The same figure also shows that most commuters will not commute a much greater distance than those living in the center. In fact, the trend seems to go the other way. This buffer is a net importer of people during daytime, with about a five percent more people coming in during the day than leaving. People working in this buffer will travel from two to five kilometers to work and seem to flow from the same directions as those coming to work in the center of town. 209 People who live in the most distant buffer see similar movements of large number of people going to work in the center of Port-au-Prince and Pétionville. The region has almost twenty percent more commuters commuting to work in the morning than commuting from work during the evening. Nevertheless, there is more variation, with people living in this buffer going to work as far as Croix-des-Bouquets and Canaan. Figure 12 in chapter 3 shows that there are a number of people who travel relatively great distances to work. The graph shows that those who tend to travel longer distances to work also tend to go out of the buffer during the day, suggesting that those are the ones who travel to the center of town. A large proportion of those people likely travel from Carrefour. On the other hand, those who live and work in this buffer tend to travel smaller distances and they are likely spending their daytime in regions such as Croix-des-Bouquets and Pétionville. The pane on the right of Figure 5 confirms this, as the many well-lit places all over the buffer indicate that people who work in this buffer also tend to live in the same region and travel smaller distances. The flow of commuters from each buffer during the day and evening time in Cap-Haïtien is depicted in Figure below. The first row of Figure 5 (left pane) shows that a number of commuters who live in the center of town tend to go to Petite Anse during the day and some others tend to travel farther south along the Mapou River. The center of Cap-Haïtien sees about six people commuting into the area during the day for every commuter who travels outside of it. The histogram in Figure 13 (Chapter 3) shows that about seventy percent of the trips to the center are less than five kilometers. Most of those trips are from commuters who live in the second buffer, either farther south along the Mapou River or in Petite Anse. The second row (left pane) of Figure shows that the vast majority of commuters in the second buffer takes that direction. About 95 percent of the commuters (Figure 13 in chapter 3, second row, left column) in the second buffer travel less than five kilometers to work. 210 COMMUTING FLOWS FOR EACH DISTANCE TO THE CITY CENTER Figure 4. BUFFER IN PORT-AU-PRINCE Source: Authors’ elaboration using Digicel data. Notes: The left column shows the work destination of people living in each buffer. The right column shows the home location of people working in the selected buffer. The first row shows results for locations within 1km from city center, the second row for locations between 1km and 5km from the center, and the last row for locations beyond 5km and within 25km from the center. 211 Figure 5. COMMUTING FLOWS FOR EACH DISTANCE TO THE CITY CENTER BUFFER IN CAP-HAÏTIEN Source: Authors’ elaboration using Digicel data. Notes: The left column shows the work destination of people living in each buffer. The right column shows the home location of people working in the selected buffer. The first row shows results for locations within 1km from city center, the second row for locations between 1km and 5km from the center, and the last row for locations beyond 5km and within 25km from the center. 212 CRITICALITY ANALYSIS The criticality analysis undertaken for this study deserves a few explanations. Road links were removed individually, and the travel times between each potential Origin and Destination were recomputed using a network analysis toolbox following a shortest path algorithm. In total, the removal of 268 road links was tested. The most critical road links were identified by calculating how the average accessibility to opportunities in the urban area of Port-au-Prince was affected. The lower the average accessibility compared to the baseline, the more critical the road link. The average accessibility is moderately impacted by the removal of the road links. The maximum impact is to reduce the average accessibility by five percent for a sixty-minute threshold. There are two main explanations for this low impact. First, when even localized natural hazards hit, they are likely to take out more than one road link, making some areas much more difficult to bypass. We do not capture this impact. Secondly, the calculations we used in this exercise do not account for congestion. In the model, when one road link becomes unavailable all traffic will use a lengthier route, but the travel speed on the alternate route will remain identical and unaffected by extra traffic. In reality, when traffic is rerouted to an alternate link, the capacity of that link is likely to become overused leading to congestion that would further reduce the average accessibility in the urban area. With these limits in mind, Figure 6 shows how accessibility is modified at the local level by the removal of the fifteen most critical links in the urban area. 213 MAPS OF LOSS OF ACCESSIBILITY TO OPPORTUNITIES RELATIVE TO THE BASELINE (IN PERCENTAGE POINTS) FOR DISRUPTION OF THE Figure 6. FIFTEEN MOST IMPORTANT ROAD LINKS IN THE URBAN AREA OF PORT-AU-PRINCE 214 215 Source: Authors’ elaboration using CNIGS data on road and Tap-Tap networks. 216 REFERENCES Aslund, O., J. Osth, and Y. Zenou. 2010. “How Important Is Access to Jobs? Old Question--Im- proved Answer.” Journal of Economic Geography 10 (3): 389–422. doi:10.1093/jeg/lbp040. Ciccone, Antonio, and Robert E. Hall. 1996. “Productivity and the Density of Economic Activity.” The American Economic Review 86 (1): 54–70. Duranton, Gilles, and Diego Puga. 2004. “Micro-Foundations of Urban Agglomeration Economies.” Handbook of Regional and Urban Economics 4: 2063–2117. Glaeser, Edward L., and Matthew E. Kahn. 2004. “Sprawl and Urban Growth.” In Handbook of Regional and Urban Economics, edited by J. Vernon Henderson and Jacques-François Thisse, Volume 4:2481–2527. Cities and Geography. Elsevier. Melo, Patricia C., Daniel J. Graham, David Levinson, and Sarah Aarabi. 2013. “Agglomeration, Accessibility, and Productivity: Evidence for Urbanized Areas in the US.” In Paper Submitted for the Transportation Research Board 92nd Annual Meeting. Rosenthal, Stuart S., and William C. Strange. 2004. “Chapter 49 Evidence on the Nature and Sources of Agglomeration Economies.” In Handbook of Regional and Urban Economics, 4:2119–71. Elsevier. http://linkinghub.elsevier.com/retrieve/pii/S1574008004800063. Venables, Anthony J. 2017. “Breaking into Tradables: Urban Form and Urban Function in a Devel- oping City.” Journal of Urban Economics 98 (March): 88–97. doi:10.1016/j.jue.2017.01.002. 217 ANNEX 6 – METHODOLOGY AND DATA LIMITATIONS ON MUNICIPAL FINANCES IN HAITI The objective of this study consists in analyzing the main issues linked to municipal finance at the national level. More specifically, it is a review of the different funding sources, which looked at the global and local tax system, the territorial aspect, the distribution of resources and expen- ditures, and finally, outlining the problem of municipal finance in Haiti. The methodological approach adopted included understanding the tasks, literature review, data collection, interviews with key decentralization stakeholders, and report generation. The document review was to identify and analyze different legal documents relevant to local taxation, research reports, and books and articles on fiscal decentralization and local finance. The purpose of the data collection was to obtain from the main sources concerned information on local finance stemming from the MICT, DGI, the six targeted communes by the Municipal Development and Urban Resilience (MDUR) project, and the official government gazette le Moniteur the legal texts on decentralization and taxation. The main difficulties encountered were the inconsistencies between the data coming from different sources over the same research period. Often, the data are incomplete for the last three years. Therefore, they can only be subject to limited use in the analysis. Indeed, beyond the values reflecting the weight of all Haitian municipalities in the global tax values (both expenditures and revenues), the universe of municipalities must be disaggregated according to their size. For that reason, we worked from three different groups, according to the classification used for the tax definition on business licenses (Article 28 of the Act on the business license).1 NO. OF % OF TOTAL AVERAGE POPULATION MODEL AVERAGE / TYPE COMMUNES POPULATION BY COMMUNE TOTAL AVERAGE Type 1 5 2,401,032 22% 480.206 6.2 Type 2 12 2,069,621 19% 172.468 2.2 Type 3 123 6,441,166 59% 51.945 0.7 TOTAL 140 10,911,819 100% 77.389 1.0 Source: Authors’ elaboration using Digicel data. 1 Including Tabarre, which was also part of the former municipality of Delmas. 218 a) First group: Port-au-Prince, Pétionville, Carrefour, Delmas, Tabarre: 2,401,032 inhabitants b) Second group: Aquin, Cap-Haitian, Cayes, Fort-Liberté, Gonaïves, Hinche, Jacmel, Jérémie, Miragoâne, Petit-Goâve, Port-de-Paix, Saint-Marc: 2,069,621 inhabitants c) Third group: the 123 other communes: 6,441,166 inhabitants Financial autonomy must be taken into consideration as an important question in carrying out the fiscal analysis of the communes; this is the element that could give an effective role to local governments in the management of their territories, whether it be in urban or rural settings. In this case, percentage of own resources must be estimated against total resources of the communes. In the Haitian context, we will consider as the municipalities own resources those collected directly by them, and as tax resources those collected on their behalf by DGI. The difference between the costs incurred by the communes and the revenues is paid by the transfers made by the MICT through the FGDCT. To calculate the effects and be able to make comparisons between municipalities, we estimated the resources and expenditures per capita. It is also important to classify the communes based on the department to which they belong, so as to identify the existence of specific features amongst the different regions. In relation to the information sources, we have worked with the data provided by the MICT and by the DGI. We also used specific information derived from the communes of the Cap- Haïtien agglomerate. Note that the data is incomplete and there are inconsistencies and serious shortcom- ings. In some cases, we made estimates based on other data. For the spending of the FGCDT, we have had to work with the programming data at that time, since information pertaining to imple- mentation was unavailable. Finally, it is necessary to point out that the issue of communal sections, institutionally defined as basic territorial collectivity (section 2.1, Article 15, Chapter II, decree defining the decentralization framework) is not directly relevant to the context of this work. While the institutional framework currently in force considers the communal section as an important entity, in practice, its devel- opment status is embryonic and the provision of paragraph 1 contained in Article 140 from Part II of the decentralization decree that specifies that 25 percent of the collection of the property tax (CFPB) received by the communes should be transferred to the communal sections has not even come into force. 219