Haiti: Selected Issues
Author:
International Monetary Fund. Western Hemisphere Dept.
Search for other papers by International Monetary Fund. Western Hemisphere Dept. in
Current site
Google ScholarClose
Publication Date:
20 Apr 2020
eISBN:
9781513541501
Language:
English
Keywords:
ISCR; CR; Haiti; trade; income; export; value; per-capita GDP; inequality in Haiti; Gini coefficient; income-generating activity; income share; fuel price; customs data; government social spending; unrecorded imports; social policy; supply cost variation; price transmission effect; Electricity; Income inequality; Fuel prices; Imports; Caribbean; authorities' priority; social protection; financing plan; disaggregated data spending; PNPPS objective; Social assistance spending; Social protection spending
Download PDF (2.1 MB)

Selected Issues

Abstract

Selected Issues

Social Protection Spending1

Haiti is one of the poorest and most unequal societies globally. This paper provides an overview of the current social context in Haiti and coverage and effectiveness of social safety nets. Existing social programs suffer from volatile financing and are fragmented, poorly targeted, and undermined by weak governance. As a new national policy on social protection and development is being prepared by the government, this paper builds on the literature and lessons from recent experiences to examine policy options to effectively reduce poverty and inequality. It argues in favor of a comprehensive strategy focused on a few quasi-universal cash transfer programs with simple demographic and/or geographic targeting. The programs must include systematic annual reports and be supported by a communications strategy and sustainable financing plans.

A. Stock-taking of Existing Social Policies2

1. Social policy has historically been underdeveloped in Haiti. With few resources and low capacity, the engagement of the Haitian government on social policies has been limited. The first modern public social security organizations created in the late 1960s have remained weak and underdeveloped (World Bank, 2016). Government spending in the three main social sectors—social protection, health, and education—is complemented by: i) programs funded and managed by development partners; ii) semi-independent contributory insurance schemes; and iii) domestic private spending.

2. Total government spending on social sectors peaked at 5.7 percent of GDP in 2013. After the catastrophic 2010 earthquake, a surge in foreign financial and technical support contributed to a proliferation of projects and interventions. Most public social spending at that time was allocated to education (3.5 percent of GDP), reflecting the authorities’ priorities. The remainder was divided between health (0.8 percent of GDP) and other general social protection programs (1.4 percent of GDP). These levels are comparable to the averages for low-income countries but well below those of other Latin American and Caribbean (LAC) countries (Figure 2).

3. Government-financed social protection spending in Haiti has declined sharply since 2010. At its peak, the national social assistance strategy called Ede Pèp consisted of a wide array of programs aimed at protecting vulnerable populations. At the time, these government programs were funded in roughly equal part by the Petrocaribe agreement, by the public treasury, and by special taxes on remittances and international phone calls (Figure 2).3 While Petrocaribe ended in April 2018, revenues from the special taxes decreased to 0.3 percent of GDP from 0.5 percent of GDP per year during FY2013-FY2017. Other budget resources allocated to social protection have also dwindled, largely because social protection spending has been displaced by rising spending on energy subsidies (see SIP on energy sector). Consequently, many of the Ede Pèp programs have stopped operating and assessing the level of activity in those remaining is difficult because of their fragmentation and a general lack of oversight.

Terminology and Methodology Note on Social Spending

article image

  • This note distinguishes between different forms of social programs―also called social protection and social safety nets. All social spending and programs have in common the objective of addressing vulnerabilities by preventing individuals from falling into poverty and by helping the poor to exit poverty. There are two types of programs that differ based on their financing sources: i) programs that are financed by workers and employers’ contributions, possibly along with government subsidies, are defined as social insurance programs and include health insurance programs and pension schemes; ii) other programs that are financed by the government and/or development partners and that do not rely on private contributions are defined as social assistance. This note uses the term “social protection” to cover the full spectrum of programs included in social insurance and assistance (i and ii above), public and private.

  • Social assistance and insurance programs operate across different social sectors, such as health, education, labor, gender and agriculture. Because of the lack of data on social protection, social spending sometimes corresponds to a broader category of expenditure that includes total spending in social sectors, even if some of the beneficiaries are not vulnerable. The definition of social protection used in this note is thus a subset of “social spending”.

  • Data. Data quality and availability are challenging because of the multiplicity of actors, different definitions of social protection, and the lack of systematic reporting. A comprehensive overview of spending on social protection would require disaggregated data spending. This note relies heavily on the comprehensive World Bank 2016 report entitled “Better Spending Better Services”, the most recent and comprehensive attempt at collecting data consistently across social programs. The study is based on data from FY2013 (FY ending September 30)—the year the last household survey was completed―and is complemented whenever possible with more recent data. All available time series later than FY2013 show that social spending has decreased, suggesting that social conditions are likely to have deteriorated since then.

4. Total spending on social protection, education and health has largely been funded and complemented by development partners, although that support has declined. Some development partners have financed the government’s social programs (Petrocaribe, World Bank and InterAmerican Development Bank-IDB) including with a technical assistance component, while others have favored fully independent programs funded and operated by NGOs (e.g., USAID programs are operated by CARE). Independent social programs financed by development partners were estimated at about 0.5 percent of GDP in 2013. In addition, development partners have also financed independent programs amounting to 1.9 percent of GDP for health programs and 0.5 percent of GDP for education in 2013 (World Bank, 2016). The downward trend in overall support from development partners since 2013 suggests that the authorities should not count on higher levels going forward.

5. Beyond government and development partner assistance, social protection is also provided under three contributory insurance programs (see Table 1). The Office of Work Accidents, Illness and Maternity (OFATMA) under the Ministry of Social Affairs and Labor (MAST) provides work injury and disability insurance to both private and public workers. Pension schemes are run by the National Office for Old Age Insurance (ONA) for the private sector and by the Directorate for Civil Pensions (DPC) in the Ministry of Finance (MEF) for the public sector. Social insurance contributions (from employers and employees) represented about 1.2 percent of GDP in 2013, over 90 percent of which came from private sector worker contributions. Total spending under these three programs remained slightly below 2 percent of GDP from 2013 to 2018. These programs have typically run surpluses because, with current Haitian demographics, there are many more contributors than beneficiaries for now.

Table 1.

Haiti: Main Existing Social Protection and Promotion Programs

article image
Source : « Protection et promotion sociales en Haïti, document-cadre pour une politique nationale » (2018), WBG 2016

B. The Limited Effectiveness of Social Spending

6. Coverage of all existing public social assistance programs is low and fragmented, and transfers are poorly targeted. Despite the large number of programs, institutions, and partners, coverage of the population is relatively low (Table 1). Only about 8 percent of the population received non-contributory social assistance benefits such as scholarships, food aid or transfers in 2012 (World Bank, 2016). Coverage is progressive as the population of the two poorest quintiles are twice as likely to receive assistance relative to the top two quintiles. However, coverage is uneven and young children (pre-school-age) are under-represented among beneficiaries, which is a concern given their vulnerability. Child labor is a significant issue in Haiti. In contrast with the progressivity of coverage, social assistance transfers are regressive as their amount increases with income (Figure 1).

Figure 1.
Figure 1.

Haiti: Government Social Spending and Technical and Financing Support

Citation: IMF Staff Country Reports 2020, 122; 10.5089/9781513541501.002.A002

7. Few Haitians are covered by existing contributory social insurance programs and their sustainability has recently been undermined by weak governance. Access to contributory programs is out of reach for most Haitians because of the high level of informality. It was estimated that only 400,000 workers, or around 10 percent of wage workers and less than 4 percent of the population, were insured in 2012. Most households with social insurance belong to the highest consumption quintile and live in the Port-au-Prince area (World Bank, 2016). The operating surplus at ONA―the largest contributory program―have recently turned into deficits. The 2019 report from the audit court Cour Supérieure des Comptes et du Contentieux Administratif (CSCCA) raised concerns about ONA’s sustainability due to weak governance, poor portfolio choices, and misuse of funds.

8. The impact of social spending has been limited due to low capacity, fragmented programs, and a lack of continuity due to volatile financing. Haiti would benefit from a cohesive and coordinated national strategy. Reports from the Commission Économique pour l’Amérique Latine et les Caraïbes (CEPALC), World Bank and IDB highlight the fragmentation of social policies as a major source of inefficiency. At over 20, there are too many different programs, and many are too small, with narrow target groups. These programs are supervised by nine ministries and executed by 11 public agencies and ministries. Development partner support is not well coordinated and future financing is always uncertain. To some extent, fragmentation reflects the absence of a national strategy and low capacity, with the latter compounding the former. Furthermore, monitoring, analysis and evaluation of results are generally missing, preventing the application of lessons from experience to inform current policies. Estimates from the last household survey (2012) suggest that social assistance and insurance programs contributed to reducing the poverty headcount by only one percent (World Bank, 2016).

9. Despite gradual improvements and total spending levels comparable to other low-income countries (LICs), health outcomes remain relatively low, implying that public spending has been comparatively inefficient. Health indicators have improved, with life expectancy rising by 9 years and child mortality declining by 8 percentage points from 1990–2016. While noteworthy, this is a slower pace than the average improvement in all LICs. The poorest continue to benefit the least from health services (World Bank 2016). Health service utilization in percent of the population is about 20 percent lower on average than LIC peers, because the density of medical staff to hospital beds is below average, despite similar levels of public spending. Health facilities are judged to have low efficiency because of low consultation-per-medical personnel ratios and excessive share of administrative staff.

10. Education outcomes and government spending on education are at comparable levels to LIC peers, though affected by fraud and misuse. Education attainments and learning indicators have improved but, based on existing data up to 2013, are still relatively low. The literacy rate of adults in 2013 was close to 80 percent while 45 percent of the population had completed secondary school (high school). This exceeded by a small margin LIC country averages of 60 and 40 percent, respectively (World Bank 2016). Gains are due mostly to increased access to private schools—less than 15 percent of schools are public—and to increased public financing through school canteen programs and tuition waivers under the Programme de Scolarisaton Gratuite et Universelle (PSUGO) and topped up by development partners’ tuition waiver program (Table 1).4 In 2013, the government reported that over 0.5 percent of GDP was spent on waivers supporting approximately 1.4 million students. However, implementation challenges have plagued PSUGO and independent reports (ULCC 2013, CCSCA 2019) have repeatedly pointed to fraud, such as “ghost” schools and nonexistent student beneficiaries. While 70 percent of the education budget is allocated to salaries, on par with other LAC countries, it is heavily skewed towards administrative or non-teaching staff. The quality of service delivery in education is poor and disparities across regions are wide (World Bank, 2016).

C. The New National Policy

11. The new Politique Nationale de Protection et de Promotion Sociales (PNPPS) is a very welcome national initiative led by the ministry of social affairs and labor (MAST) and the ministry of planning and external cooperation (MPCE).5 The ministries of health, education, and women’s conditions and four public organizations participate in the relevant drafting sub-commissions.6 Sub-commissions have collaborated with international organizations and civil society organizations (CSOs).7, 8 A first draft of the policy was submitted for national consultation in June 2019. A revised draft was expected to be presented to the government for approval by the Council of Ministers by the end of 2019. This inclusive approach is an excellent starting point to prepare a homegrown and more effective and efficient social safety net. Ideally, the implementation plan should quickly articulate some short-term actions to address urgent needs as well as to maintain momentum and traction.

12. The PNPPS assigns three objectives to social protection: (1) breaking the inter-generational transmission of poverty by supporting children’s health, development, and education; (2) creating conditions so that all individuals can get out of poverty; and (3) creating conditions of equality of people and eliminating discrimination related to gender, age, and disability.

13. While aiming to provide universal coverage, the PNPPS focuses on four strategic priorities: (1) childhood; (2) labor, employment and employability; (3) health and social security, and (4) resilience to shocks. The corresponding social protection schemes will prioritize rural populations living in extreme poverty and the most vulnerable.

D. Policy Considerations

Key goals in the design of the PNPPS should be to reduce fragmentation and overlap and boost ownership. In this context, and drawing from the literature and international experience, the following recommendations present important features that would contribute to attaining those goals.

14. Given the conditions in Haiti, cash transfers to households would be preferred over other forms of assistance that are more complicated to administer. The literature generally finds a positive effect of cash transfer programs on poverty, education, health, and nutrition indicators.9 In-kind support such as Panye Solidarité and Kantin Mobile require the government to buy food products and redistribute them. This assumes that the government has this administrative and logistical capacity and at the lowest cost. It also presents opportunities for leakages. The May 2019 report of the CSCCA could not confirm the data provided by the administrative unit set up within the Ministry of Finance, the Fonds d’Assistance Economique et Sociale (FAES), on the distribution of food baskets or free meals.

15. International experience and the recent literature suggest that the best strategy for Haiti would aim for quasi universal coverage of basic needs. Given the limited resources available for social protection, this would involve broad demographic or geographic targeting of beneficiary groups. Recent research has found more refined targeting, based on income for example, less effective in countries not able to identify beneficiaries or lacking capacity to implement (Banerjee et al. 2019). While much of the population is poor in Haiti and could benefit from social assistance, some social spending would aim for simple targeting of groups in ways that are easy to implement, for instance all children under some age cutoff. For example, providing average annual transfers of US$150 to all households yields un upper bound estimate for total transfers of US$345 million (4 percent of GDP).10, 11 Blunt targeting, for example by demographic or region, would lower total costs to more feasible levels. The goal is to strike a balance between more universal, simple-to-administer programs and cheaper but more complicated conditional and/or targeted programs that involve greater costs and more capacity to administer, implement, and monitor.

16. Staff recommend that implementation of the PNPPS could focus on three unconditional cash transfer schemes: (1) to households with children under a certain age and/or with pregnant women, (2) to people with disabilities, and (3) to old people aged 65 and above. Focusing on those three categories of beneficiaries is consistent with the strategic priorities set by the PNPPS, addresses Haitians’ strong sensitivity toward and against social assistance which is perceived as a charity handout, and factors in the limited availability of resources. The lack of conditionality is motivated by the likely difficulties for the most vulnerable households to comply with conditions and by lack of administrative capacity to monitor compliance. 12

17. Effective social programs have to be complemented with improvements in the delivery of health and education services. In view of the large gaps in health and education, supply side reforms in health and education are needed. Conditions for cash transfers, such as school attendance or vaccinations, could only be considered after improvements in education and health infrastructure actually improve the availability and access to services. Policy makers should work with development partners to support public investment and improvements in these areas. Tavneet Suri (2019), “Universal Basic Income in the Developing World,” NBER Working Paper, 25598; Abramo, Laís, Simone Cecchini and Beatriz Morales (2019), Social programmes, poverty eradication and labour inclusion. Lessons from Latin America and the Caribbean, ECLAC.

18. Existing social programs need to be rationalized. Small programs that do not effectively contribute to the objectives defined in the PNPPS should be phased out. Social insurance programs and large or relevant social assistance programs that meet PNPPS objectives would need to be reformed and transitioned to the new simplified and more efficient cash transfer schemes. Rationalizing existing programs should be complemented by an overhaul of current agencies whose credibility has been undermined by allegations of mismanagement and misuse of public and donors’ funds.

19. The PNPPS would need to involve a sustainable financing strategy based on domestic resources. In the context of declining development partner support and given the questionable impact of foreign-financed programs in terms of fragmentation and on program ownership, an appropriate financing strategy has to rely on domestic resources. To finance social programs, the government should boost customs revenue mobilization and reduce unproductive public spending, particularly on goods, services and energy subsidies. Any reform to the energy sector or subsidies must be preceded by effective mitigating measures addressing the adverse impact of such a reform on key groups affected, particularly the most vulnerable (see SIP on Energy Sector). Short and medium-term budget formulation should include a coherent and comprehensive framework that specifies explicitly the financing sources for social spending at different horizons.

20. Delivering unconditional cash transfers to broad groups requires the proper identification of potential beneficiary groups.13 The development of the MAST Information System (SIMAST), a database containing detailed socio-economic information will facilitate such identification.14 SIMAST is being developed with technical assistance from development partners and currently covers 1.2 million individuals (14 percent of the population). The marginal cost of increasing coverage by one percent is estimated at US$350,000 implying a total one-off cost of US$28.4 million (0.4 percent of GDP) to achieve full coverage of the population. Potential shortcomings of SIMAST include that it may not be expanded to cover dangerous urban areas, it will need regular updates, and its financing could end. Over the long term, national coverage based on a modernized national identification system could be envisaged and possibly coordinated down the road with other government functions like taxpayer IDs.

21. The national policy should designate one sole institution to coordinate and develop a robust and secure system to deliver social assistance benefits. Building capacity and institutions is as important as the design of social programs. The delivery of social services in Haiti has involved fraud and corruption and a lack of accountability, and the credibility of the agencies involved has been undermined by allegations of mismanagement of public and donors’ funds. To avoid the risk of fragmentation and program interruption, the government should give the primary coordination role to the MAST since its central role in designing and executing social protection is established by the organic law 24/11/1983 and reinforce its capacity.

22. Developing digital financial services to support financial inclusion would be a critical for implementing large-scale cash-transfer schemes. The ex-post assessment of the response to Hurricane Matthew emphasized the operational challenges related to cash-based initiatives, including cash-transfer delivery through mobile phones: Catholic Relief Services and others “met challenges with beneficiaries not having access to phones, not being comfortable with the system, not having SIM cards that worked with mobile money, not having coverage at disbursement sites, and provider agents not ‘seeing’ transfer amounts on beneficiary phones…”.15 The 2015 Financial Inclusion Strategy of the Banque de la République d’Haiti (BRH) aims to develop new mobile financial services which may provide reliable cash distribution channels.

23. All existing and new programs should report systematically and annually on their activities, coverage, resources received and transferred, beneficiaries, and administration costs. In the medium term, this could be delegated to an independent organization, like the Observatoire National de la Pauvreté et de l´Exclusion Sociale (ONPES) mentioned in the last national plan for poverty reduction in 2014. Involving development partners and CSOs in the oversight of the implementation of the PNPPS would be crucial to limit leakages and mismanagement risks.

24. Proactive communications will be essential for the success of the PNPPS. Communications should work towards building broad public support for comprehensive reforms, with a focus on social protection. The government should aim to advance understanding of the links between unpopular revenue mobilization reforms and the sustainable development of social protection programs. Close monitoring of benchmarks and targets is needed and should be appropriately communicated to build trust in public institutions.

Table 2.

Haiti: Illustrative Template of Reform Implementation

article image

References

  • World Bank. 2016. Better Spending, better services – a review of public finances in Haiti (Vol. 2) (English). Public expenditure Review (PER). Washington, D.C.: World Bank Group (link).

    • Search Google Scholar
    • Export Citation

  • Cour Supérieure des Comptes Et du Contentieux Administratif (CSCCA). 2019. Rapport Sur La Situation Financière De L’état Et L’efficacité Des Dépenses Publiques Pour L’exercice 2017–18 (RSFEEDP V).

    • Search Google Scholar
    • Export Citation

  • Ministère des Affaires Sociales et du Travail (MAST). 2018. Protection et promotion sociales en Haïti. Document-cadre pour une politique nationale.

    • Search Google Scholar
    • Export Citation

  • Unité de Lutte Contre la Corruption (ULCC). 2013. Rapport annuel de l’exercice 2012–13.

  • Lamaute-Brisson, Nathalie. 2015. Protection et promotion sociales en Haïti La stratégie nationale d´assistance sociale (SNAS/EDE PEP), enjeux stratégiques et institutionnels. Rapport de la Commission Économique pour l’Amérique Latine et les Caraïbes (CEPALC).

    • Search Google Scholar
    • Export Citation

  • InterAmerican Development Bank. 2018. Temporary social safety net and skills for youth. Washington, D.C.

  • Banerjee, Abhijit, Paul Niehaus, and Tavneet Suri. 2019. Universal basic income in the developing world. Annual Review of Economics 11.

1

Prepared by Matthieu Bellon (FAD) and Frederic Lambert (WHD).

2

See SIP on Inequality for data on poverty and inequality. Recent data on social indicators in Haiti is limited. Based on developments over the past five years, most social indicators are not likely to have improved (see Box 1).

3

Under the Petrocaribe agreement between Venezuela and Haiti started in 2007, Venezuela supplied fuel to Haiti with concessional financing for a portion of the imports (SIP on Energy Sector Reform, Annex II). The Haitian government could use the proceeds from domestic fuel sales to finance investments and social programs.

4

PSUGO is financed from the government budget and from special taxes on international phone calls collected by the telecom company CONATEL and on international money transfers collected by the Central Bank, both of which are accumulated in the National Education Fund (FNE).

5

The last three-year plan for poverty reduction in 2014 “Thinking and Fighting for a Haiti without Poverty: Action Plan for Accelerating the Reduction of Extreme Poverty” (PAARP) is now outdated.

6

The Bureau of the Secretary of State for Integration of the Handicapped (BSEIPH), the Directorate of Civil Protection (DPC), the Office of Citizen Protection (OPC), and the Socio-Economic Assistance Fund (FAES).

7

ECLAC, ILO, IADB, PAHO/WHO, UNICEF, UN WOMEN, WFP, WBG.

8

The American Chamber of Commerce, the Alternative insurance Company, the Alliance for Disaster Risk Management, the Center for the Promotion of the Women Workers, the Chamber of Commerce and industry, the Economic Forum of the Private Sector and Haitian Unions.

9

See, for example, Parker, Susan and Petra Todd (2017) “Conditional Cash Transfers: The Case of progresa/Oportunidades,” Journal of Economic Literature, 55(3), 866–915; Banerjee, Abhijit, Paul Niehaus and

10

The program for mothers of school-age children that was supported by the WB provided cash transfers of US$100 per year for mothers with one child, US$150 for mothers with two children, and US$200 for mothers with three children or more. The program ended in 2018 because of financing issues.

11

Total costs would need to additionally account for administration costs.

12

For instance, if access to schools is limited in rural areas, conditionality on school attendance would not be feasible.

13

Individual identification with biometric details could reduce duplication and fraud risks and may strengthen the security of cash payments if the individual ID number is linked to a mobile account. However, India’s experience with the “Aadhaar” identification system offers a cautionary tale, especially in regions where internet access is unreliable. Smart cards with a microchip storing information accessible without internet access, may be a better alternative.

14

There is a consensus among technical and financial partners to unify efforts and focus on developing the SIMAST. The SIMAST has been used in projects by USAID, the WFP, CARE, Action Contre la Faim, the EU, the Red Cross, Concern Worldwide and others. SIMAST is made available by MAST employees whose work is funded by the WFP. Updates and extensions of the database have so far been financed under one-off projects by development partners.

15

CRS Haiti, ’The Vendor Effect’: Hurricane Matthew Response. A cash transfer programming learning study (2018).

Same Series

Other IMF Content

Other Publishers

  • Collapse
  • Expand
Cover IMF Staff Country Reports
Haiti: Selected Issues
Author:
International Monetary Fund. Western Hemisphere Dept.
Volume/Issue:
Volume 2020: Issue 122
Publisher:
International Monetary Fund
ISBN:
9781513541501
ISSN:
1934-7685
Pages:
57
DOI:
https://doi.org/10.5089/9781513541501.002
  • <sup><xref ref-type="fig" rid="RA02fig01">Figure 1.</xref></sup>
    View in gallery
    Figure 1.

    Haiti: Government Social Spending and Technical and Financing Support

Figure 1.

Haiti: Government Social Spending and Technical and Financing Support