Journal Issue

Haiti: Selected Issues

International Monetary Fund. Western Hemisphere Dept.
Published Date:
June 2015
  • ShareShare
Show Summary Details

Public Expenditure in Haiti: Balancing Human Capital and Infrastructure Formation1

A. Introduction

1. Capital spending increases underpinned Haiti’s public expenditure trends in the last decade. Despite relatively high and sustained levels in public investment, Haiti’s performance compares negatively with that of the region and low-income countries (LICs) both in terms of translating resources into better infrastructure and with respect to management practices.

2. Spending on education increased recently and progress was made in achieving higher enrollment rates. The strong focus on increasing primary enrollment rates helped to bring Haiti closer to universal primary school enrollment (88 percent). However challenges remain in closing the gap with the region in terms of increasing average years of schooling, increasing enrollment rates for secondary and tertiary education and enhancing the quality of the education system.

3. Rebalancing public expenditure to support both human capital formation and better infrastructure will improve the growth environment. The combination of high, but inefficient, capital spending calls for enhancing capital efficiency, rather than increasing spending. The declining school-age population relative to the working population provides room to raise enrollment rates, but additional fiscal space is needed to increase spending per student. Reallocating resources towards education while enhancing its quality and accessibility would support human capital formation, but would also increase the productivity of the capital stock. This should foster long term growth.

B. Public Expenditure Trends, 2004–2014

4. Haiti’s central government spending increased by 14 percentage points of GDP in the last decade. (Chart 1). It rose from 11 percent of GDP in 2004 to 25 percent of GDP in 2014, while revenues (including grants) increased from 10 percent to 19 percent of GDP.2 Current revenue increased by 4 percentage points while budget support increased by 1 percentage points. Project grants peaked in the post-earth quake period but returned to lower levels in 2013–2014. The total increase in project grants amounted for 4 percentage points in 2004–2014. Public expenditure was highest (at 28 percent of GDP) in 2012–2013, as post-earthquake reconstruction accelerated. The decrease in public spending in 2014 suggests a return to more sustainable levels in the years ahead.

Chart 1.Public Expenditures and Revenues

(Percent of GDP)

Sources: Haitian Authorities; and IMF staff calculations

5. Central government spending (as a share of GDP) is similar to other countries in the region but it is expected to decrease in the coming years. At 27 percent of GDP in 2012–2014 central government spending was slightly lower than the average for Latin America and the Caribbean (LAC) (30 percent of GDP), and LICs (28 percent of GDP) (Chart 2). As the post-earthquake effort winds down, government spending (as a share of GDP) is projected to decrease to 23 percent in 2015–2017 which is lower than the regional average and the LIC’s but in line with lower revenue-to-GDP ratio in Haiti. Government revenues were 20 percent in 2014 compared to 26 percent of GDP in the region and the LICs.

Chart 2.Public Expenditure (2012-14)

(Percent of GDP)

Sources: World Economic Outlook; and IMF staff calculations.

6. Central government expenditure growth was mainly driven by capital spending. Capital spending increased by about 9 percentage points of GDP between 2004 and 2014.3 This significant increase was mainly financed by project grants (4 percent of GDP) and Petrocaribe financing (3–4 percent of GDP). Current spending (in percent of GDP) increased by about 5 percentage points, largely driven by increases in the wage bill (about 3 percentage points). The other main contributors were transfers and subsidies (about 2 percentage points) (Chart 3). The increase in capital spending started before the 2010 earthquake, during 2006–08. Although from a lower level than capital spending, the wage bill and transfers and subsidies also increased. Interest payments are relatively minor given a low and concessional debt (Chart 4).

Chart 3.Growth of Expenditures (2003-14)

(Percent of GDP)

Sources: World Economic Outlook; and IMF staff calculations.

Chart 4.Increase in expenditures 2003-2014

(Percentage increases in three-year averages)

Sources: Haitian Authorities; and IMF staff calculations.

7. As of 2014 capital spending in Haiti is higher than the average for both the LAC region and LICs. In 2012–14, capital spending was 15 percent of GDP, much higher than the average for the LAC region (6 percent of GDP) and the LIC average (9 percent of GDP) (Chart 5). Among LAC countries, only Ecuador shows a similar high level of capital spending with 14 percent of GDP. A number of LICs, however, also have relatively large public investment ratios, as e.g., Republic of Congo (24 percent), Djibouti (16 percent), Mozambique (16 percent) and Niger (14 percent). Although low development levels may justify high public investment ratios, these may also respond to inefficiencies, weak governance and a lack of planning.

Chart 5.Capital Spending (2012-2014)

(Percent of GDP)

Sources: World Economic Outlook; and IMF staff calculations.

8. In turn, spending on other categories are lower than in the LAC region. Haiti’s central government wage bill (5 percent of GDP), is relatively low with respect to LAC (8 percent of GDP), and LICs (7 percent of GDP) (Chart 6). This difference can be explained by the different coverage of employee compensation across countries. Spending on social benefits and security is negligible given the low level of coverage and benefits.4

Chart 6.Decomposition of Public Expenditure (2012-2014)

(Percent of GDP)

Sources: World Economic Outlook; and IMF staff calculations.

9. However, spending in education increased in the last few years, due mainly to PSUGO, and is similar to the LAC average. Spending on education was 4.2 percent of GDP in 2014 and is projected to increase to 4.8 percent of GDP in 2015 (MENFP 2015). Central government spending was comparable with the LAC average in 2013 (4.7 percent of GDP), slightly lower than the average for Sub-Saharan Africa (5 percent of GDP) and higher than that in MENA (3.8 percent of GDP) (World Bank 2013). Half of the education spending in Haiti is domestically financed and half foreign financed. The government launched in 2011–12 the PSUGO (Programme de Scolarisation Universelle, Gratuite et Obligatoire). The program aims providing free basic education to children from poor households. PSUGO spending amounted for 0.5 percent of GDP during 2012–2014 and is budgeted to increase to 1.1 percent of GDP in 2015. PSUGO is financed by the fees collected on international calls and remittances, and to a lower extent through the budget.5

C. Efficiency of Capital and Education Spending: Some Indicators

10. The efficiency of Haiti’s public investment spending appears low, according to a number of metrics. An efficiency frontier analysis is used to evaluate capital spending efficiency (IMF 2014). This method assesses efficiency by the extent to which monetary inputs are translated into infrastructure outputs. The methodology reflects how far a country is from the production possibility frontier which is determined by the best performers worldwide. The used inputs are the real public capital stock per capita in purchasing power parity (PPP). The infrastructure output is estimated using the infrastructure component of the Global Competitiveness Indicator (GCI), which considers, inter alia, the overall quality of infrastructure, of roads, of (air)ports, of electricity supply, and of (mobile) telephone lines. The efficiency frontier analysis uses two metrics to evaluate capital spending efficiency: the “Partial Free Disposal Hull” (PFDH) and the “Data envelopment Analysis” (DEA). The main difference between the PFDH and DEA metrics is that the first measure benchmarks against a group of peers and re-samples to improve robustness and reduce sensitivity to outliers while the second is more sensitive to the presence of outliers and could overestimate inefficiencies by benchmarking a country relative to only a few best performers in the sample. According to both measures capital spending efficiency in Haiti is significantly lower than the average for the LAC region and LICs. In particular, Haiti’s PFDH score (at 0.5 on 0.5–1.2 range) is among the lowest in the world (Table 1).

Table 1.Efficiency Scores
Source: IMF (2014)
Source: IMF (2014)

11. Low capital spending efficiency in Haiti may be rooted in a weak public investment framework. In this regard, the Public Investment Management Index (PIMI) (Dabla-Norris, et al., 2012) assesses the public investment framework by focusing on the main stages of the project cycle (strategic guidance and project appraisal; project selection; project implementation; and, project evaluation and audit). Haiti lags behind the LAC region and LICs in all stages of the project cycle, and thus its overall PIMI-index is significantly lower than that for both country aggregates (Table 2). Haiti performs particularly poorly in the appraisal and selection of projects.

Table 2.Public Investment Management Index
OverallBy stage

12. Despite increased primary education enrollment rates and average years of education, these indicators lag the averages for the region. Efforts in the last few years resulted in an increase of the enrollment rate for primary education from 78 percent to 88 percent in 2011, slightly lower than the LAC average (92 percent in 2012). However secondary enrollment rate is still in sharp contrast with the region. Only 25 percent enrolled in 2012 to secondary education compared to 73 percent in the LAC (World Development Indicators (2012). Available data up to 2010 shows that the average years of schooling (for people 15 and older) increased significantly (from a low base) in the past 60 years but remain lower than those in the Caribbean and Latin America (Table 3). Although the recent increase in enrollment rates will result in new adult cohorts with higher years of education, more progress is needed to reach regional averages given the low base level. Collaboration with private schools is important since the Haitian education system is heavily dominated by private schools. More than 80 percent of all enrollments are in private schools. These schools are managed by international institutions, NGOs, or religious institutions, and their quality varies significantly.

Table 3.Average Years of Total Schooling, Population > 15, 1950-2010
Dominican Republic2.
Latin America2.
Sources: UNESCO (2011).
Sources: UNESCO (2011).

13. PSUGO provided subsidies to private schools and eliminated school fees. PSUGO has 11 sub-programs with different objectives. Table 4 summarizes the execution status of the PSUGO program during 2011–2014 with regard to the objectives under the 11 sub-programs and the resources provided by the PSUGO Fund and the budget. The emphasis of PSUGO has been on a few areas, namely subsidies to participating private schools, the elimination of school fees due by parents to participating public schools, and payments to public school teachers and administrators participating in the program. Subsidies are provided for 6–12 year old children in about 8,400 private and public schools. The program objectives for 2011–2016 (hiring teachers and school directors, and subsidies to children in and outside public schools) were achieved at end 2014 with execution rates of 79 percent or more. At the same time, only 55 percent of the planned resources for 2011–2016 (G 12.2 billion) had been used and there was an underestimation of the program cost. As a result, cost overruns led to off-budget spending estimated at G 2 billion. Going forward, the authorities intend to contain costs by progressively incorporating students into the program, and by more selectively granting subsidies to private schools, mainly in areas that are not adequately covered by public schools.

Table 4.Haiti: Achievements under the PSUGO Program, 2011-2014
Performance per sub-programCosts per sub-program (G million)
Sub-programsActivitiesProg.Est.Execution rate (%)Prog.Est.Execution rate (%)
Stengthening human resources1Recruited teachers and education professionals11,0008,725791,44397167
Identification of needy kidsNeedy kids identified in 2 departments100,55194,513942800
Strengthening the data base351440
Strengthening the information system25310
Subsidizing kids outside public schoolsNumber of kids covered860,250785,674919,4945,15354
Eliminating school fees in public schoolsNumber of kids covered639,750680,30010653841978
Auditing schools covered by the programNumber of schools audited10,0009,552961017473
Improving coordination and communicationNumber of vehicles10023231224738
Monitoring and evaluationNumber of physical and financial reports15533974951
Training of teachers and Directors, supervisionNumber of teaching professionals10,0007,495754049
Caring for street kidsNumber of kids covered2,0001,25063501019
Sources: Ministry of Economy and Finance; Ministry of Education (PSUGO Program); and Fund Staff estimates.

Out of 4,207 teachers hired in FY2014, only 3,388 are reported to have signed contracts.

Sources: Ministry of Economy and Finance; Ministry of Education (PSUGO Program); and Fund Staff estimates.

Out of 4,207 teachers hired in FY2014, only 3,388 are reported to have signed contracts.

D. Assessing Future Demand for Public Education: Fiscal Implications

14. Haiti’s fertility rate has decreased in recent years, and it is projected to continue declining through the medium term (Figure 1). According to the U.N. population projections, the fertility rate (average number of children per woman) in Haiti has been declining over the past 35 years to 2.9 in 2015 (2.1 percent in LAC) and is expected to decrease further to 2.2 in 2035 (Figure 1).6 Accordingly, the share of population below 19 (which underpins the demand for schooling) is projected to decrease from 45 percent in 2015, to 36 percent in 2035 (27 percent in LAC). In turn, life expectancy is projected to increase from 64 years in 2015 to 69 years in 2035 (76 years in LAC). The increase in life expectancy combined with declining fertility rates will push up the dependency ratio from 7.5 percent in 2015 to 9.9 percent in 2035 (20.8 percent in LAC).

Figure 1.Demographic Trends and Baseline Projections

Sources: UN World Population Projection(2013).

15. Education cost per student, school enrollment rates, and demographics determine education expenditure (as a share of GDP). Expression (1) decomposes public spending on education (as a share of GDP, ED/GDP) in: (i) benefit generosity (average education spending per student to GDP per worker); (ii) the school enrollment (or coverage) rate (i.e., the number of students to the school age children); and, (iii) demographics, defined as the ratio of school-age children to population 20–64.7

16. Projected demographic trends provide scope to increase enrollment rates. A projection assuming unchanged benefit and coverage ratios, and the U.N. “medium” fertility rates, is consistent with education spending (as a share of GDP) decreasing to 3.5 percent in 2035. This result would be driven by the projected decrease in the ratio of school-aged population to the labor force. A gradual increase in enrollment rates, in particular secondary enrollment rate, to catch up with the regional average (in 2015) would imply an increase in education spending to 4.9 percent of GDP in 2035. The increase in education spending needed to increase enrollment rates is sensitive to fertility rates. Using the U.N. “low” fertility rates would require a level of education spending (as a share of GDP) of 4.1 percent of GDP while U.N.’s “high” fertility rate would result in a significantly higher education spending ratio, of 5.7 percent of GDP.

Chart 7.Education Projections

(percent of GDP)

Sources: UN World Population Prospects (2013); Haitian Authorities; and IMF staff calculations.

17. Increasing benefits would require additional fiscal space. The analysis above isolates the implications of demographics and coverage on education spending (as percent of GDP). The fiscal space to finance increases in education will be larger, if in addition to increases in enrollment rates, there is an increase per-student benefits. Increasing per-student benefits by, e.g. 20 percent, would require additional fiscal space of about 1 percent of GDP.

E. Concluding Thoughts

18. Haiti should enhance the efficiency of capital spending rather than in increase its level. While high infrastructure-related spending was needed in the aftermath of the earthquake, capital spending appears to be crowding out other, equally needed, spending. In addition, a number of metrics show that the efficiency of public investment in Haiti remains low when compared to the region and LICs. Thus, increasing the efficiency of capital spending by strengthening public investment institutions should now be prioritized. Transparency of budget execution, openness of the procurement process and an efficient cash management system are critical to ensure stability of public investment and reduce opportunities in rent seeking. More transparency and accountability in project management, monitoring and evaluation will strengthen incentives to deliver on time and on budget and increase the integrity in the use of public resources (IMF 2015).

19. Investing in human capital through better education is essential to foster sustainable economic growth and social cohesion. A strong commitment to education is necessary. Pursuing higher enrollment rates, including gradually increasing secondary school attendance, should be accompanied with increases in the quality of the education system to ensure learning. Recent research has shown that economic growth is not only responsive to average years of schooling, but on measures of the quality of education (Hanushek and Woessman 2012).8 Qualified teachers are an important determinant of cognitive achievement (Wold Bank, 2014)9. PSUGO is investing in training of teachers and directors. On the institutional front, the government should have a strong role in regulating and monitoring the quality of the education system, including the private schools. Spending needs to be allocated transparently and public and private schools should coordinate policies and approaches. Improving education statistics is vital for monitoring developments and evaluating outcomes. Together with the implementation of other institutional reforms, the accumulation of human capital should make Haiti more attractive for investors and promote employment and social cohesion.


Prepared by Emine Hanedar (FAD)

Figures for Haiti refer to fiscal years (October – September).

Capital spending excludes PSUGO (Programme de Scolarisation Universelle, Gratuite et Obligatoire). PSUGO-related expenditures are classified as public investment in the budget.

Social benefits (GGES) are defined as transfers in cash or in kind to protect the entire population or specific segments of it against certain social risks. They are classified according to the type of scheme governing their payment, and consist of social security benefits, social assistance benefits, and employer social benefits (GFSM 2001, paragraphs 6.67-6.72). Note: The payment of pensions and other retirement benefits through employer social insurance schemes are not expense; they are treated as reductions in liabilities.

See the Selected Issues Paper “Haiti’s Public Sector: Explaining the ECF’s Fiscal Target” for a comprehensive discussion of the PSUGO fund and its resources.

The U.N. periodically publishes data on demographic trends and performs long-term projections. The projections are based on a probabilistic method using the medium fertility assumption. The probabilistic method for projecting total fertility consists of three separate processes: (i) a high-fertility pre-transition phase, (ii) the fertility transition itself and (iii) a low-fertility post-transition phase. The method incorporates country-specific assumptions and empirical experience of all low-fertility countries having already experienced a recovery (U.N., 2013).

Fiscal Monitor, “Public Expenditure reform, Making difficult choices”, (2014).

The authors use cross-country growth regressions to assess the relationship between educational achievement and GDP growth. The results show a close relationship between educational achievement and GDP growth that is remarkably stable across countries, time period and sensitivity analyses of specification.

Other important factors are nutrition, health, cognitive and social-emotional development in the early years of life.

Other Resources Citing This Publication