Chapter 7: Public Expenditure Rationalization

Alfred Schipke, Aliona Cebotari, and Nita Thacker
Published Date:
April 2013
  • ShareShare
Show Summary Details
Shamsuddin Tareq, Alejandro Simone, Abdoul Wane, Kenichiro Kashiwase and Koffie Nassar 

Current fiscal trends in the Eastern Caribbean Economic and Monetary Union (OECS/ECCU) are not sustainable.1 After good fiscal performance in the 1990s, the fiscal position of the union’s member states weakened considerably during the 2000s (Table 7.1). Fiscal deficits and public debt worsened as the region responded to a series of shocks that reduced growth. The fiscal deficit doubled to an annual average of 5.5 percent of GDP during the 2000s and public debt levels soared in all countries—from an average debt-to-GDP ratio of 56 percent during the early 1990s to 90 percent in 2009. The resulting accumulation of debt was financed mostly by captive financial institutions, including indigenous banks and social security schemes (see Chapter 5 on Public Debt).

Table 7.1OECS/ECCU: Selected Economic and Financial Indicators, 1990–2009
Indicator and country1990–992000–091990–941995–992000–042005–0920082009
Real GDP growth (percent)
Antigua and Barbuda3.–10.3
St. Kitts and Nevis4.–5.6
Saint Lucia3.–1.3
St. Vincent and the Grenadines3.–0.6–2.3
Simple average3.–4.3
Overall fiscal balance (percentage of GDP)
Antigua and Barbuda–2.9–13.2–2.3–3.5–15.8–10.5–5.7–18.1
St. Kitts and Nevis–2.6–6.7–0.8–4.4–9.8–3.5–3.9–2.9
Saint Lucia0.3––4.7–3.7–0.9–3.2
St. Vincent and the Grenadines–1.6–2.6–1.6–1.6–2.0–3.1–1.4–3.2
Simple average–2.2–5.5–1.8–2.6–7.1–3.9–2.5–5.4
Public sector debt (percentage of GDP)
Antigua and Barbuda92.4103.991.893.0119.388.576.9101.8
St. Kitts and Nevis62.8134.249.368.2124.6143.8131.0148.5
Saint Lucia30.858.127.334.454.961.458.863.2
St. Vincent and the Grenadines50.960.055.346.558.561.657.065.5
Simple average55.886.156.056.785.986.378.790.1
Sources: Country authorities; and IMF staff estimates.Note: n.a. = not available.
Sources: Country authorities; and IMF staff estimates.Note: n.a. = not available.

Significant fiscal consolidation is inevitable. Increasing debt levels are putting pressures on the interest bill and are likely suppressing growth, resulting in a vicious cycle of high debt and low growth and further contributing to an eventual need for fiscal adjustment (Kumar and Woo, 2010; and IMF, 2011).2 Large fiscal deficits, in turn, may be increasingly difficult to finance if concerns about debt sustainability intensify.

The size of the required consolidation for each member depends on the country’s initial level of debt and debt tolerance. The OECS/ECCU has adopted a debt target of 60 percent of GDP, which the authorities deem to be consistent with the region’s capacity to service its debt while maintaining a comfortable level of reserves. For some countries the adjustment needed to meet the debt target will be large. For example, St. Kitts and Nevis would have to cut its ratio by almost 90 percentage points of GDP to meet this medium-term target. Antigua and Barbuda would need to reduce debt by about 40 percent of GDP.

What policies will deliver the needed fiscal adjustment in the OECS/ECCU economies? First, the fiscal imbalances in the member countries reflect high growth in spending rather than low revenue collection (Figure 7.1). Significant tax reforms, including implementation of a value added tax by all countries (except Saint Lucia) have increased revenue from 23 percent of GDP in the 1990s to 26 percent of GDP in 2009 (see Chapter 6 on Revenue). On the expenditure front, central government spending jumped from 25 percent of GDP on average during the 1990s to 33 percent of GDP in 2009.3 This increase in government spending reflects mainly escalating current transfers (Antigua and Barbuda, Dominica, and Grenada) and capital expenditure (Antigua and Barbuda, Dominica, Grenada, and Saint Lucia).4 With certain social assistance programs provided to all regardless of income, the costs of delivery swelled, in part due to inefficiencies. The increasing weight of nondiscretionary recurrent spending (including civil service pensions and subsidies) and the relatively high wage bill have reduced the flexibility of budgets to respond to shocks. The resulting deficits have been financed by borrowing, which has led to swelling interest bills. Given the high levels to which taxes have risen in some countries, reducing government spending offers the best means of reducing fiscal imbalances.

Second, the literature finds that expenditure-based adjustments are more successful. Analyses of large and successful fiscal consolidation efforts emphasize the importance of reducing public spending, especially wages and transfers, rather than increasing revenues (Alesina and Perotti, 1995; McDermott and Wescott, 1996; and Alesina and Ardagna, 1998, 2009). Several of the more institutionally advanced economies in the world established medium-term expenditure frameworks to help governments set and meet multiyear priorities and build credibility (Tsibouris and others, 2006).

A broadly based expenditure rationalization strategy could reduce public spending on a cumulative basis by 3 to 5 percent of GDP over the medium term in OECS/ECCU countries. This chapter reviews the factors underpinning the large increases in public spending and discusses measures for expenditure rationalization. It is based on country-specific analyses covering the main expenditure items in the central government budgets. The analysis shows that expenditure rationalization measures could reduce spending without diminishing protections for the poor. The proposed measures would contribute significantly to the fiscal adjustment needed to meet the region’s debt target of 60 percent of GDP.

Figure 7.1OECS/ECCU: Revenue and Expenditure, 1990–2009

Sources: Country authorities; and IMF staff estimates.

The next section of this chapter reviews trends in the government wage bill and employment and possible reform measures, and is followed by a section on issues related to social security. Subsequent sections provide analyses of education, health expenditures, and social assistance spending. Capital spending and issues related to public investment programs are the focus of the penultimate section, and the final section looks at expenditure policy issues related to parastatal entities (PEs).5 See Table 7.2 for a summary of the short-term and medium-term measures reviewed in the remainder of this chapter.

Table 7.2OECS/ECCU: Selected Expenditure Rationalization Measures
Civil service employment and wages
Short-term optionsMedium-term options
Temporary freeze on wages

Streamline allowances

Selective hiring freeze

Eliminate known duplication and overlaps
Rationalize compensation on the basis of the compensation and classification review results

Review the role of government
Social security
National pension schemesCivil service pension schemes
Increase the contribution rate

Set ceiling on insurable earnings appropriately and index to the growth of insurable earnings

Increase retirement age to 65 years and thereafter link it to life expectancy at retirement

Correct the front-loaded nature of accrual rates

Introduce automatic indexation of pensions

Diversify investment portfolios and strengthen investment policies
Adjust the retirement age and the calculation of pensionable earnings

Reduce the high replacement rates arising from combined pensions

Freeze civil service pensions at current levels

Make the civil service pension a contributory scheme

Undertake regular actuarial reviews

In the medium term, phase out civil service pension schemes where applicable
Short-term optionsMedium-term options
Increase the student-teacher ratio in primary and secondary schools

Consider greater cost recovery in publicly provided tertiary education
Reform the organization of education to better match resources with demographic changes

Establish training qualifications for new hires

Evaluate efficiency of bonding arrangements in leave programs and scholarship provision

Achieve economies of scale through greater regional cooperation
Health care
Short-term optionsMedium-term options
Revise user fees system to better reflect the cost of health services

Ensure adequate financing for proper maintenance of hospital infrastructure

Set a nominal freeze on compensation of health personnel

Strengthen incentives for the collection and payment of user fees

Undertake cost audits in health facilities and improve cost tracking

Expand the implementation of treatment protocols for different diagnoses
Review public benefits package to ensure consistency with financial constraints

Base some portion of hospital reimbursement on prospective methods

Rationalize the use of hospital infrastructure, where occupancy rates are low

Achieve economies of scale through greater regional cooperation
Social assistance
Short-term optionsMedium-term options
Introduce objective and transparent targeting mechanisms and strengthen accountability

Compile a complete inventory of all social assistance programs

Establish a central beneficiary registry for all programs

Consolidate social assistance programs under a single institution
Rationalize social assistance programs on the basis of a coherent social protection strategy

Link social assistance programs to actions that promote human capital
Capital spending
Short-term optionsMedium-term options
Improve coverage of public sector investment programs and the reporting of public investment

Define and strictly enforce simple selection criteria for screening investment projects

Review the existing portfolio of capital projects with a view to rationalizing it

Tighten procurement procedures and establish commitment controls

Strengthen and enforce monitoring procedures
Develop capacity for cost-benefit analysis at the regional level

Carry out systematic ex post cost-benefit analyses of large investment projects

Strengthen the legal framework for public-private partnerships and the government’s capacity to undertake them
Parastatal entities (PEs)
Short-term optionsMedium-term options
Establish and enforce existing legislation on governance and reporting requirements

Ensure that consolidated PE operations reports are prepared on a regular basis

Strengthen the oversight of PEs by ministries of finance

Scrutinize proposals for PE creation and periodically assess their cost-effectiveness
Review the functions of and rationales for the existing PEs in the context of the review of the role of government
Source: IMF Technical Assistance Reports.
Source: IMF Technical Assistance Reports.

Civil Service Employment and Wages

Government wage bills represent the biggest expenditure items in the budgets of OECS/ECCU countries. Expenditure on personal emoluments and wages averaged more than 9 percent of GDP during 2006–09 (close to one-third of total government expenditures) (Table 7.3). When other elements of the compensation package are included (i.e., other benefits and allowances, employer contributions, and civil service pensions), the total compensation costs of civil servants is even higher. Moreover, these figures do not include PEs, which tend to have compensation arrangements at least as generous as those in the civil service sector.6 St. Kitts and Nevis, Saint Lucia, and St. Vincent and the Grenadines have the highest wage-bill-to-GDP ratios among the OECS/ECCU countries (near to or greater than 10 percent of GDP in 2008), while Antigua and Barbuda has the lowest (8.3 percent).

Table 7.3Wage-Bill-to-GDP Ratio(Percentage of GDP)
Antigua and Barbuda10.210.710.910.910.
St. Kitts and Nevis12.211.411.211.611.911.310.110.010.2
Saint Lucia11.011.511.010.810.
St. Vincent and the Grenadines11.911.411.410.710.410.710.410.211.0
OECS/ECCU weighted average10.811.110.810.510.
Source: Country authorities.
Source: Country authorities.

Figure 7.2Wage Bill as a Share of Government Revenues, 2008

Sources: Country authorities; and IMF staff estimates.

Note: OECD = Organization for Economic Cooperation and Development. The small island economies group includes The Bahamas, Barbados, Belize, Brunei, Trinidad and Tobago, Fiji, Jamaica, Kiribati, Maldives, Marshall Islands, Palau, Papua New Guinea, Samoa, Solomon Islands, Sri Lanka, Timor-Leste, Tonga, and Vanuatu.

A number of countries in the region implemented reforms to reduce high wage bills. These reforms led to a decline in the ratio of the wage bill to GDP during 2000–08. Short-term measures included wage and hiring freezes, although wage definitions and degrees of coverage of the central government varied by country. Dominica actually cut nominal wages. Medium-term measures included voluntary separation packages (Antigua and Barbuda) and outsourcing of selected functions (Dominica and Grenada). However, these measures often were not underpinned by an overall employment and compensation reform strategy, and the wage-bill-to-GDP ratio, after declining somewhat, trended upward in 2008 in some countries.

Main Issues

The wage bill and total compensation of civil service employees in OECS/ECCU countries consume too large a share of budget resources.7 On average, the wage bill accounted for 36 percent of government revenues in 2008, compared with 28 percent in Latin American countries and 24 percent in member countries of the Organization for Economic Cooperation and Development (OECD). The union average was higher than that for most small islands (Figure 7.2). This ratio varies considerably among the countries: Dominica had the lowest ratio at 28 percent, while both Antigua and Barbuda and St. Vincent and the Grenadines had ratios greater than 40 percent. High compensation costs raise questions about affordability and sustainability.8 In addition, compensation expenditures tend to be difficult to adjust in the short term, thus limiting the ability of the government to respond to exogenous shocks—a particularly important issue for OECS/ECCU countries, which historically have suffered from recurrent natural disasters.

Figure 7.3Civil Service Employment as a Share of Population

Source: Country authorities.

Several general factors have contributed to high compensation costs.

  • High government employment. Small island economies tend to have relatively large governments, reflecting economies of scale in the provision of public services (Remmer, 2010). However, high government wage bills in these countries also are a consequence of high unemployment in the region—governments have sometimes acted as an employer of last resort.9 This situation has led to a high ratio of government employment as a share of the population (Figure 7.3), and in some countries, to duplications in functions across government entities. Dominica and Grenada have the lowest ratios of government employment to population, while Antigua and Barbuda has the highest.

  • Large wage increases. Collective bargaining agreements have led, in some cases, to large across-the-board real increases in salaries bearing little relation to changes in productivity or performance of employees. For example, in Saint Lucia, civil service wages increased by 8.5 percent in real terms during 2001/02–2009/10. Based on a 2009 labor survey, the civil sector wage premium was estimated at 35 percent in Saint Lucia in 2009.10 In St. Kitts and Nevis, government wages increased 18 percent more than did cumulative inflation during 2006–09.

  • Other benefits. Collective bargaining agreements have also led to the proliferation of allowances and other benefits.11 Between 16 and 20 types of different allowances exist in each member country. Spending on such allowances has been about 1.0 percent of GDP in 2008–10. Rules governing these allowances are typically unclear, permitting a large degree of discretion in their use. The large number of small allowances suggests that it would be useful to review their purposes and streamline them, which could also save administrative resources.

Some sectors show signs of overstaffing. Low and declining student-teacher ratios for primary and secondary schools are the result of the number of teachers and associated supporting staff not being adjusted in line with demographic developments. This lack of adjustment is a consequence of budgeting on the basis of historical levels of input use (see the section on education, later in this chapter).

Functions among ministries, departments, and PEs are overlapping and duplicative. For example, three public sector institutions procure pharmaceuticals in Antigua and Barbuda. In Saint Lucia, functional reviews of the Ministry of Communications and Works and the Ministry of Social Transformation revealed considerable overlap between the two ministries, their autonomous agencies, and among departments within each ministry. Similarly, the survey of public officials carried out for the OECS Institutional and Organizational Capacity Review in 2001 also pointed out duplication of functions in Dominica. More generally, social assistance benefits are administered by different government agencies with a variety of targeting and implementation mechanisms, multiplying administrative costs in all OECS/ECCU countries (see the section later in this chapter on social assistance).

Efficiently managing “established” civil service employment is a challenge because of the difficulties identified below:12

  • Adjusting employment in response to changing needs.13 Redundancy provisions in public service regulations make it difficult to adjust government employment levels in OECS/ECCU countries. In some cases, the provisions are unclear, resulting in uncertain costs (Antigua and Barbuda); in others, the cost of laying off redundant workers is high (Grenada),14 or the state requires hard-to-establish “obvious and compelling” reasons for employees to be declared redundant (St. Kitts and Nevis). In addition, generous benefits from the civil service pension plans have acted as a strong disincentive for mobility from civil service employment to the private sector, reducing attrition rates and thus limiting possibilities to adjust employment (see the section on social assistance). Legal provisions for terminating underperforming employees involve numerous stages of appeal that can take several years, making it difficult to lay off employees based on performance. Nevertheless, Dominica was able to reduce the size of established workers in the government as part of its stabilization program in the early 2000s through a combination of wage freezes, voluntary resignations, and retrenchment.

  • Rewarding performance and specialized skills. Collective bargaining agreements mandate across-the-board compensation increases irrespective of performance. Formal performance evaluations, when pursued, are either highly subjective (St. Kitts and Nevis) or give similar performance ratings to a majority of workers (Dominica and Grenada).15 In addition, the wage scale is very compressed. The compression ratio—the ratio of the highest to the lowest wage in the government—is 4.7 in Saint Lucia, 4.5 in Antigua and Barbuda, and 4.4 in Grenada, the three OECS/ECCU countries for which data were available. These ratios are low by international standards, suggesting that highly skilled workers are underpaid while unskilled workers are overpaid.16 Wage compression makes it difficult to attract and retain highly qualified workers in the government. Low compression ratios are found to be associated with corruption (Van Rijckeghem and Weder, 2001; and Abed and Gupta, 2002). These factors have led to rigid civil service employment practices that have inhibited the hiring and retention of qualified staff in key areas and reduced the flexibility needed to formulate and implement policies effectively.

These human resources management challenges for established workers have led OECS/ECCU countries to expand nonestablished employment. Reliable information on the total number of nonestablished workers is not available in most member countries.17 In one country where data are available—Antigua and Barbuda—such workers constitute more than half of total central government employment. The fact that conditions of employment in these positions are determined in separately negotiated contracts has led to (1) important inequities in pay and benefits between established and nonestablished workers, (2) increased politicization in the appointment of civil servants, and (3) difficulties in keeping track of the number of nonestablished employees and their compensation arrangements.

In some cases, these difficulties have led to the creation of PEs. The number of PEs has grown over time in several member countries. Employment in these entities is governed by legislation different from civil service legislation and regulations (see the section in this chapter on parastatal entities). Available information suggests that employment in PEs is significant in some of the countries, but PEs are neither part of the statistics on government employment presented here nor of the wage bill. The total government wage bill is therefore much higher than the numbers presented here suggest and decisions to rationalize the government wage bill should take into account the larger base.

Information systems and analysis related to basic human resources management are weak. Personnel records, including data on leave, allowances, and other benefits that staff receive, are produced for established workers. In some cases, this record keeping is required by law (Antigua and Barbuda). However, the information in the records is not used for analysis. The widespread prevalence of manual procedures has also complicated rapid access to human resources information.

Policy Reforms

Policies for an overall employment and compensation reform strategy combine measures with short- and medium-term impacts. Short-term measures focus on attaining essential short-term savings and on preparing for reforms with a lasting impact. Medium-term measures are designed to increase or make permanent the savings secured with short-term actions by reforming the role and structure of government. Theoretically, in both the short and the medium terms, the role and scope of the government must be constrained to be consistent with society’s ability and willingness to cover the cost of government services.

Several fundamental building blocks for sound fiscal policy management need to be put in place for any wage-bill rationalization strategy to succeed. First, a clear overall budget constraint should apply to compensation and should be based on comprehensive definitions of “government” and “compensation” to guide the process. Any package of measures should be adequately appraised to ensure it is in line with this budget constraint. Second, a specific minimum level of information on employment and compensation should be available for adequate quantification and evaluation of reforms. Third, adequate public financial management procedures should be in place to monitor and enforce the execution of any desired target wage bill.

Short-Term Actions

  • Temporary freeze on wages. Nominal wage levels can be frozen broadly for the entire civil service or selectively for particular groups of employees for a limited period. Assuming unchanged employment levels, a freeze should result in a reduction in the wage bill relative to GDP as the economy expands in nominal terms. A freeze will also provide time for the governments to undertake necessary background studies for designing and implementing medium-term civil service reform strategies.

  • Streamlining of allowances. Tightening eligibility for and reducing the number and size of allowances can achieve fiscal savings as well as increase the transparency of such remuneration. Similarly, steps could be taken to control overtime payments, for example, by replacing monetary compensation with leave time or by suspending the overtime premium on the hourly wage.

  • Selective hiring freeze.18 Combined with natural attrition, a selective or partial hiring freeze in nonpriority areas can help to reduce compensation spending. A reduction in employment levels can be achieved without the costly severance payments associated with lay-offs. This gives the government some time to prepare for a more comprehensive civil service reform while containing the growth in the wage bill in the interim. Critical replacements could be made by redeploying staff from other jobs, to the extent they have the appropriate skills. Some OECS/ECCU countries have used a freeze as a temporary measure to contain growth in the wage bill in times of budgetary pressures but not yet as part of a comprehensive effort to consolidate the wage bill or government employment.

  • Legislative changes to increase flexibility in the management of human resources. The revisions should address issues related to severance and procedures for termination of underperforming workers. International good practices that protect workers but avoid the perpetuation of inefficiencies should be implemented. In particular, where applicable, the current lack of clarity regarding the costs associated with redundancies should be clarified.

  • Elimination of known duplication and overlaps. Functional reviews and other available information can be used to identify superfluous positions. For example, reviews in Saint Lucia recommended merging some agencies to reduce excessive administrative costs.

  • Reform of civil service pension schemes. Generous pension benefits that exceed 100 percent of salary in some countries should be reduced to increase labor mobility to the private sector and reduce labor costs in the economy. Diminished benefits could eventually help increase the attrition rate, facilitating a reduction in civil service employment and an improvement in competitiveness (see the section on social security).

  • Preparatory steps for comprehensive civil service reform. The time and fiscal space provided by the above measures should be used to prepare for more comprehensive civil service reform by undertaking the following:

    • A payroll audit. An audit of the payment of civil service salaries, allowances, and pensions could help identify and correct any existing pay anomalies or fraud, especially for payroll systems that are not automated. Antigua and Barbuda, Grenada, and Saint Lucia are already planning to undertake such audits with assistance from the World Bank.

    • An employment census. A census will help develop a database on employment and compensation, which would be a critical input for preparing a civil service reform strategy. Given the importance of PEs, it is important that they be included in the census. The census may also help identify the presence of ghost workers in public institutions. To work within resource constraints, the focus should be on the central government and parastatal institutions with the largest wage bills. Antigua and Barbuda, Grenada, and Saint Lucia have initiated employment censuses of the central government with the assistance of the World Bank.

    • Functional reviews. Functional reviews help with rationalizing the structure of government by identifying duplication and overlap in functions across public institutions. Two pilot reviews were completed in Saint Lucia, as noted above, and similar reviews have begun in Antigua and Barbuda and in Grenada with World Bank assistance.

    • A compensation and job classification review. Such a review would identify distortions and inequities related to pay and assess the degree to which civil service compensation is competitive with compensation in the private sector. Well-defined job descriptions and a comprehensive definition of compensation are needed so that the compensation comparison ideally incorporates all forms of payment in the civil service and private sectors for a given job description and set of qualifications.19 The authorities in Antigua and Barbuda are working with the World Bank under the Public Sector Transformation Plan to update job descriptions.

Medium-Term Actions

  • Review the role of government. The assessment should help determine whether all the functions currently performed by government are needed or justified from a public finance point of view. The review would identify overlaps; prioritize functions and specify institutions that could be closed, divested, or merged; and highlight functions that could be discontinued or outsourced to the private sector.20 The envisaged role of the government should be consistent with the availability of resources, that is, the roles and responsibilities of the government should be defined within the feasible budget. Antigua and Barbuda is planning to undertake a review of the role of the government in the context of its Public Sector Transformation Plan.

  • Eliminate “unfunded” positions in the budget. Although there is some basis for retaining a few unfilled vacancies to ensure flexibility in meeting the government’s personnel needs and emerging priorities, in some countries (e.g., Saint Lucia) the number of “unfunded” positions is large. Moreover, the practice of approving civil service positions without corresponding financing is contrary to standard good practice relating to budget preparation (Potter and Diamond, 1999).

  • Rationalize compensation on the basis of the compensation and job classification review results. Pay levels should be adjusted to remove distortions and inequities and to ensure that civil service compensation is competitive. However, reducing compensation levels that are too high with respect to comparable positions in the private sector is likely to be difficult and may need to be gradually phased in.

  • Gradually integrate nonestablished employees that fit in the new organization of the civil service into the establishment. As reforms advance and basic distortions in the grading and classification of established personnel are resolved, it will be necessary to have one personnel system for all civil servants to ensure equal pay for equal work. This will prevent the inequities and adverse consequences created by the two existing parallel systems.

Social Security

National pension funds in OECS/ECCU countries administer pay-as-you-go pension schemes and provide a variety of other benefits. The funds provide old-age, disability, and survivor benefits in addition to other age-related, sickness, maternity, and employment injury benefits.21 In some countries, the national pension fund also pays out social assistance pensions for the poor (Grenada, St. Kitts and Nevis, and St. Vincent and the Grenadines) and certain medical payments to enrollees (Dominica and Saint Lucia).22 Member countries also have separate civil service pension schemes, although they are being phased out in some countries. Some groups of civil servants are subject to special pension rules.23

Private occupational pension schemes complete the spectrum of social security benefits in the OECS/ECCU. Occupational pension plans are usually governed by regulations in the Income Tax Act. In Saint Lucia and St. Vincent and the Grenadines, additional legislation within the Insurance Act governs these schemes. Coverage under occupational schemes has been limited to the formal sector labor force, particularly larger enterprises including multinational companies, PEs, and local private companies. Occupational schemes include both defined-benefit and defined-contribution plans.

Main Issues

Many of the issues discussed in this section are also covered in the report of the Commission on the Pension and Pension Administration Reform in the OECS/ECCU.24 This commission was set up by the Monetary Council of the Eastern Caribbean Central Bank in early 2006 “to review and make recommendations to achieve the goals of a stable, predictable, and adequate income security throughout retirement.” The issues raised here and the associated policy reforms were discussed with the chairman of the commission at a meeting in Castries, Saint Lucia.25

National Pension Schemes: Sustainability, Structure, and Portability

Sustainability issues—Pension schemes in OECS/ECCU countries are unsustainable over the medium term under current system parameters. Most national pension funds have built up substantial reserves and are currently operating in surplus. However, system dependency ratios are expected to worsen significantly as a result of system maturation, aging of the population, and emigration trends.26 The latest available actuarial reviews indicate that under broadly unchanged policies, schemes already have or will start running cash deficits in the near future as benefit payments exceed contribution revenues.27 These deficits would be financed by a drawdown of reserves. Eventually, all schemes will run down their reserves, although the dates for depletion vary by country. IMF staff baseline projections broadly confirm the results of the actuarial reviews (Figure 7.4).

Figure 7.4Net Cash Flow and Reserves of Pension Funds

Source: IMF staff estimates.

One way to assess the sustainability of a pension system is to estimate the pay-as-you-go (PAYGO) contribution rate.28 These rates would have to increase to levels ranging from almost 22 percent of the salary in Saint Lucia (more than four times the 2010 rate) to 32.3 percent in Grenada (more than seven times the 2010 rate) by 2050 for contributions to cover promised benefits (Figure 7.5). IMF staff projections of dates of reserve depletion were found to be similar to those in the actuarial reviews for those OECS/ECCU countries with no significant arrears problems. However, for Antigua and Barbuda, staff projections indicated that if pension parameters remained unchanged and the government policy of nonpayment of contributions and interest continued, cash reserves would be depleted the year after the analysis was undertaken (i.e., 2011).

Figure 7.5Estimated PAYGO Rates for Pension Systems


Source: IMF staff estimates.

The schemes’ sustainability problems arise primarily from low retirement ages and other pension parameters that vary across OECS/ECCU countries. Table 7.4 shows key parameters of pension systems across the Caribbean. Retirement ages in member countries are low compared with other countries in the region and significantly lower than those in OECD countries, where the retirement ages for both men and women tend to be converging to 65 years.29 Among OECS/ECCU countries, Saint Lucia is gradually increasing the retirement age to reach 65 years by 2015. Dominica has also planned similar increases but not yet implemented them. Among the other parameters, contribution rates are also low in some countries (Antigua and Barbuda, St. Vincent and the Grenadines) while some other countries have high accrual rates (Saint Lucia).

Table 7.4Comparison of Key Parameters of Pension Systems across the Caribbean
CountryRetirement age (years)Contributions (% of earnings or payroll)Total contributions (% of earnings)Replacement ratesAdministrative costs (% of contributions)
Antigua and Barbuda602.0 to to 8.01025355017.01
Belize651.0 to 3.53.5 to 6.04.5 to 9.51031356052.2
JamaicaM 65, F 6022.52.55.010125010.1
St. Kitts and Nevis625.06.011.01030356016.01
Saint Lucia6325.05.010.014340326011.31
St. Vincent and the603.54.58.01030406015.3
Trinidad and Tobago603.57.010.51030n.a.607.1
Sources: Pettinato and Cassou, 2005; Mitchell and Osborne, 2005; Brunton and Masci, 2005; and World Bank, 2010.Note: — = Not applicable.

Average for 2006/07–2009/10 for Antigua and Barbuda; 2006/07–2008/09 for Saint Lucia; and 2006–09 for Grenada and St. Kitts and Nevis.

In these cases, the retirement age is gradually being increased to 65 years. In Jamaica, the retirement ages of men and women will be equalized by 2015 at 65 years.

Increasing to 15 years after 2012.

Sources: Pettinato and Cassou, 2005; Mitchell and Osborne, 2005; Brunton and Masci, 2005; and World Bank, 2010.Note: — = Not applicable.

Average for 2006/07–2009/10 for Antigua and Barbuda; 2006/07–2008/09 for Saint Lucia; and 2006–09 for Grenada and St. Kitts and Nevis.

In these cases, the retirement age is gradually being increased to 65 years. In Jamaica, the retirement ages of men and women will be equalized by 2015 at 65 years.

Increasing to 15 years after 2012.

Administrative expenses associated with pension funds are high in OECS/ECCU countries. Table 7.4 shows administrative costs as a share of contributions. Among the countries covered in this chapter, Saint Lucia has the lowest administrative costs at 11.3 percent of contributions. In OECD countries, these costs tend to be much lower, and in many cases are in the low single digits. The lower administrative costs in the OECD likely reflect economies of scale available in larger systems. Nevertheless, significant real growth in the compensation of civil service workers (in St. Kitts and Nevis and in Saint Lucia) and the administrative costs associated with activities not directly related to social security (in St. Kitts and Nevis) have also contributed to higher administrative costs.30 The high level of administrative expenses is a significant drag on the effective return that national pension funds can provide to their contributors.

Figure 7.6Internal Rates of Return in OECS/ECCU Pension Systems

Source: IMF staff estimates.

Note: Excludes Antigua and Barbuda because the data for calculations were not available. The internal rate of return calculations are for a male contributor and assume the current parameters for pension schemes summarized in Table 7.4; a life expectancy at retirement of 17 years for Grenada, St. Kitts and Nevis, and Saint Lucia, and 16 years for Dominica and St. Vincent and the Grenadines; and real growth for salaries equal to 1 percent in line with similar assumptions in the latest actuarial review. Inflation rate assumptions vary by country: 2.75 percent in Grenada, 2.5 percent in St. Kitts and Nevis and Saint Lucia, and 2 percent for Dominica and St. Vincent and the Grenadines.

Structural issues—The current approach to computing an individual’s average earnings results in significant inequities and reduces the insurance value of old-age pensions. The average earnings on which an individual’s pension is based are calculated on a relatively short number of years and using only the highest salaries. The system thus provides a distortionary incentive to the worker to minimize the earnings on which contributions are paid early in the worker’s career and maximize earnings as retirement approaches. In addition, it creates inequities by discriminating in favor of individuals whose earnings rise rapidly during their careers (typically high-income workers) and against those whose earnings are flatter over their careers (typically low-income workers). This methodology also reduces the insurance value of old-age pensions, because an individual’s pension is highly susceptible to fluctuations in earnings in the final years before retirement.

Accrual rates in benefit formulas are front-loaded, strongly favoring pensioners with short wage histories. Pension benefits are front-loaded with workers accruing a significant replacement rate after only 10 years of contributions, and much lower rates thereafter (Figure 7.6). This generates inequities in rates of return among contributors with contribution periods of different lengths. Figure 7.6 shows the real internal rates of return on contributions for individuals who differ only in the lengths of their contribution periods.31 Real returns exceed 3 percent for all workers and can be many times higher for workers with short work histories. Few investments of comparable risk could produce such high real rates of return over a long period.

Pension payments are typically adjusted on an ad hoc fashion, which has several disadvantages. First, pensioners face a degree of uncertainty about the future real value of their pensions. Second, if inflation is relatively high, pensioners can also lose a substantial amount of real income between adjustments. For instance, if adjustments were to happen every three years and inflation were 10 percent annually, pensions would lose about one-third of their real value between adjustments. Finally, cohorts may be treated unequally, depending on when they retire, because the real value of their pensions will depend on these ad hoc adjustments and developments in inflation.

Some parameters of the pension systems also tend not to be adjusted systematically. In some countries (Antigua and Barbuda, Saint Lucia), the ceiling on insurable earnings has eroded significantly relative to average earnings because it was not adjusted for a long time.32 The minimum pension is also typically not adjusted for inflation. To the extent that the minimum pension is set at a level that is consistent with the sustainability of the pension system, the lack of adjustment erodes its purchasing power and thus its social protection value.

The finances of pension funds in the region are vulnerable to country-specific shocks, low returns on social investments, and exposure to the public sector. International portfolio diversification is limited. The share of domestic assets in the reserve portfolios of pension funds in all cases exceeds 80 percent and in some cases is close to 100 percent (Antigua and Barbuda). Pension funds are thus exposed to fluctuations in the local economies, which are vulnerable to shocks, including from natural disasters. This exposure was demonstrated vividly in Grenada and Montserrat, where natural disasters destroyed much of these systems’ assets. Even where investment guidelines exist, they typically allow a significant portion of the reserves to be invested in social investments, government securities, or PEs, demonstrating the important role pension funds play in most OECS/ECCU economies in financing the public sector.33

Portability issues34—The Caribbean Agreement on Social Security (CASS) provides mobility incentives that can give rise to duplication of benefits. The CASS is the most important agreement related to national pension scheme benefit portability in the economic union, but no provisions in the agreement prevent the overlap of front-loaded benefits. Some pensioners can end up with replacement rates greater than the maximum of any single country if they work in several countries, and this can be true even ignoring differences in the generosity of pensions across countries.35 The problem arises because (1) the agreement does not apply when a worker has made sufficient contributions to qualify for a pension in more than one country, and (2) the accrual rates in all national schemes in the union are front-loaded.

There is no common framework for portability of occupational pensions, either within most countries or between countries. In particular, there have been two critical constraints: (1) the absence of a common approach to the regulation and supervision of occupational schemes, and in some cases very limited regulation and supervision of such schemes within countries; and (2) diverse and conflicting economic regulations for occupational schemes, including conflicting tax treatment and foreign exchange controls. As a result, workers changing jobs often cash out funds accumulated in the plans, and might spend such funds instead of saving them for retirement.

Civil Service Pension Schemes: Sustainability and Structure

Sustainability issues—For civil service pension schemes sustainability issues are especially pressing in countries that are not phasing them out. In these countries, expenditure on government pensions in 2009 ranged between 1 and 2 percent of GDP. Expenditures will likely rise, assuming the aging process is similar to that faced by national pension plans. Expenditure levels could exceed 3 percent of GDP by 2020 in some cases (Antigua and Barbuda, St. Vincent and the Grenadines). Even in countries where civil service pension schemes are being phased out, pension expenditures have been a significant burden but will gradually decline as pensioner mortality exceeds retirements.

Benefit provisions for civil service pensions are quite generous, inequitable, and not well integrated with national pension fund benefits in countries where civil service pensions are not being phased out. Civil service pension schemes are generally more generous than national schemes because of their noncontributory nature, higher accrual rates, and lower retirement ages. In countries where civil service schemes are not being phased out, government employees receive pensions that, when combined with their national pensions, can exceed 100 percent of their preretirement salaries.36

The generous civil service pension schemes generate adverse labor market incentives and complicate public sector reform. The problems include (1) reduced labor mobility between the civil service and private sectors through long vesting periods and a “golden handcuff effect,” both of which severely penalize early retirement, and limited pension portability; (2) reduced transparency of the compensation package, because it is so concentrated in future benefits; (3) costly separation of inefficient workers, which reduces the net fiscal gains from privatization, divestment, or downsizing initiatives; and (4) increased administrative costs associated with administering multiple pension schemes (Palacios and Whitehouse, 2006). The latter issue is especially important in the region, given the high administrative costs prevalent in relatively small systems.

Structural issues—Civil service pension schemes suffer from many of the same structural weaknesses that national pension schemes do. These weaknesses include low contribution rates; low retirement ages, especially for special groups (i.e., teachers, parliamentarians, and police); a faulty approach to computing earnings—in particular, the use of short wage histories; and the practice of adjusting pension payments in an ad hoc fashion. Unlike the national schemes, however, information on beneficiaries and pension levels is not readily available and the schemes have typically not been subject to actuarial reviews.

Policy Reforms for National and Civil Service Pension Schemes

National Pension Schemes

Five broad reforms can be considered for improving the finances of a PAYGO system.37 The reforms comprise (1) raising the contribution rate, (2) improving collection of contributions, (3) reducing the replacement rate, (4) reducing the system dependency ratio, and (5) lowering administrative costs. Increasing the contribution rate and reducing benefits (or their growth) for current and soon-to-be retirees can generate short-term savings that can improve the finances of the pension schemes, while the other reforms will typically do so only over the medium term.

The optimal mix of reforms will depend on the short-term financial situations of the schemes. Pension schemes that do not have a short-term financing problem (Grenada, St. Kitts and Nevis, Saint Lucia, and St. Vincent and the Grenadines) should ideally start adjusting their parameters as soon as feasible so that long-term sustainability can be restored gradually. Gradual adjustments have the advantage of making sustainability-oriented reform more equitable because the burden of adjustment can be spread across many generations. Pension schemes that have short-term financial pressures (Antigua and Barbuda and Dominica) will need to rely on less-gradual parametric adjustments. Different reform measures are discussed below for OECS/ECCU pension systems.

The revenues of national pension schemes could be improved by raising contribution rates and setting contribution ceilings at appropriate levels.

  • Increase the contribution rate. All national pension schemes in the member countries would need to consider raising their contribution rates to achieve medium-term financial sustainability. However, schemes under short-term financial pressure, as in Antigua and Barbuda and Dominica, may need to implement a large upfront increase. For this reason, the ABSSB has proposed an upfront 2 percentage point increase in the contribution rate and an increase of 0.4 percentage points per year until reaching 13.6 percent in 2020. In contrast, the pension system in Saint Lucia, which has the most comfortable financial situation among the countries of the OECS/ECCU, could consider increasing the contribution rate starting in 2012/13 by 0.1 percentage point annually for about 40 years for a total increase of 4 percentage points by 2050/51. However, other reasonable schedules are possible and should be informed by ongoing and future actuarial assessments.

  • Ensure that the ceiling on insurable earnings is set at an appropriate level and then index it to the growth of average insurable earnings. The ceiling should be updated using information on the distribution of wages to ensure that the contribution base is not unduly restricted. Then the ceiling should be adjusted regularly so that it does not erode relative to average wages over time. The ceiling could be adjusted annually in line with changes in the previous year’s average insurable earnings. These adjustments would not significantly improve the long-run finances of the pension schemes because benefits for high-income earners would eventually increase in line with their higher contributions. In St. Kitts and Nevis, this would require severing the link between the threshold for the levy on salaries and the ceiling on monthly insurable earnings to avoid affecting government revenue every time the ceiling is adjusted.

Changes to the policies governing benefits will further improve the viability of the system:

  • Increase the retirement age to 65 years and thereafter link it to life expectancy at retirement. The speed of adjustment in the retirement age will depend on the financial condition of the schemes. For example, in Antigua and Barbuda, the national pension fund has approved a proposal to increase the retirement age to 62 years and 4 months in 2011, then further to 63 years and 6 months in 2012, and by 6 months annually thereafter until it reaches 65 years. Saint Lucia is already increasing the retirement age gradually to reach 65 years by 2015. Once the target of 65 years is reached, the retirement age should be automatically adjusted based on life expectancy at retirement. Incorporating automatic adjustment provisions in the pension laws would not only depoliticize the process, but would also provide some modest additional revenue for the pension schemes.

  • Modify the calculations for average pensionable earnings. Common international practice is to calculate average earnings over 25 to 40 years, with wages from past years typically converted into current dollars using an index of the average economy-wide wage. The goal is to gradually extend the averaging to the full careers of workers, with missing years counted as zeros. If information on pensionable earnings for a full career is not available, then the full period for which information is available should be used. As additional wage history becomes available, the period could be increased by one year each year.

  • Introduce automatic indexation of pensions. Once retirement pensions (including the minimum pension) are set and sustainable replacement rates have been achieved, pension benefits should be adjusted annually in line with inflation to preserve the purchasing power of pensions. Indexing pensions (including the minimum pension) to the consumer price index will improve equity, reduce uncertainty, and depoliticize pension adjustments. Ad hoc increases should be prohibited by law. In countries where pension spending is traditionally indexed to wage growth, changing indexation to consumer prices could also generate savings. If necessary, a cap on the adjustment could be introduced (e.g., 5 percent), with adjustments for inflation above the cap dependent on government approval based on an assessment of pension fund finances.

  • Correct the front-loaded nature of accrual rates by introducing a proportional accrual rate system. This change would entail making accrual rates the same for every year worked until the maximum replacement rate is reached. For example, a maximum replacement rate of 50 percent could be targeted after 40 years of work, yielding an accrual rate of 1.25 percentage points for every year of service.38 Such an adjustment would not only address the inequities of the current front-loaded accrual rate system but would also reduce real internal rates of return to more realistic levels (Figure 7.7). To avoid an abrupt change in benefit formulas, the adjustment could be phased in. The combination of wage averaging discussed in a previous paragraph, the introduction of a proportional accrual rate system, and the phased reduction of accrual rates would contribute to partially offsetting the impact of the maturing of the system and generate savings. Moreover, correcting the front-loaded nature of accrual rates would help eliminate distortionary mobility incentives in the OECS/ECCU’s current pension portability framework.

Figure 7.7Post-Reform Internal Rates of Return in OECS/ECCU Pension Systems

Source: IMF staff estimates.

Note: The internal rate of return calculations are for a male contributor and assume the current parameters for pension schemes summarized in Table 7.4; a life expectancy at retirement of 17 years for Grenada, St. Kitts and Nevis, and Saint Lucia, and 16 years for Dominica and St. Vincent and the Grenadines; and real growth for salaries equal to 1 percent in line with similar assumptions in the latest actuarial review. Inflation rate assumptions vary by country: 2.75 percent in Grenada, 2.5 percent in St. Kitts and Nevis and Saint Lucia, and 2 percent for Dominica and St. Vincent and the Grenadines.

The net impact of these measures would be to improve the medium-term sustainability of the pension funds. Figure 7.8 demonstrates this effect by showing the degree to which the projected dates of reserve depletion are pushed back as a result of the reforms.

Figure 7.8Pension Reform Options: Stretching Net Cash Flow

Source: IMF staff estimates.

Note: Years in parentheses indicate when net cash flow turns negative. Reform 1: Increase the contribution rate. Reform 2: Increase the retirement age. Reform 3: Automatically index pensions. Reform 4: Introduce proportional accrual rate systems. The magnitudes of changes in reform options are not identical for the six countries because the starting point varies by country.

1 The baseline for Antigua and Barbuda includes the increase in the contribution rate.

2 Saint Lucia is implementing a gradual increase in the retirement age; consequently, this impact is included in the baseline projections.

Administrative expenses could also be reduced through greater regional cooperation. Combining some administrative responsibilities (such as actuarial services and investment management) with other countries in the OECS/ECCU could help produce economies of scale and reduce costs. Some countries will need to contain growth of national pension fund employees’ compensation to reduce administrative costs—compensation expenditures account for a large share of those costs.

Investment portfolios should be diversified and investment policies strengthened. To reduce country-specific risks and potential politicization of their investment portfolios, pension funds should begin gradually diversifying away from domestic assets. In particular, investment guidelines can be adopted or strengthened to facilitate the investment of newly available funds into an internationally diversified portfolio of assets. The guidelines should include such items as restrictions on the proportion of assets that can be held in any one company or country;39 a requirement that assets be managed by competent professionals selected through a transparent and competitive bidding process; restrictions on the costs that can be incurred in managing the assets (e.g., less than 0.5 percent of assets); restrictions on the types of investments that are permissible (e.g., highly leveraged and other risky investments should be prohibited); and requirements that investments and investment returns be publicly disclosed, be benchmarked against an appropriate market portfolio, and be independently audited.

Reforms to pension portability can relieve some of the pressure on finances of the pension schemes:

  • Improve the CASS by eliminating artificial incentives to migrate within the union. In principle, workers should not profit from decisions to migrate for the purpose of accruing multiple entitlements. This issue can be addressed by simply applying aggregation and apportionment rules to all contributions made by each worker, irrespective of whether the worker has enough contributions to qualify for a pension in one or more countries.

  • Enhance the portability of occupational pensions across countries and within countries. Across countries, portability could be enhanced through the harmonization of regulatory measures, ideally through a unified regulatory framework, and through coordinated supervision. Within countries, the following policies could help improve portability of occupational schemes: (1) limiting the length of the vesting period; (2) mandating that the sponsor of the scheme reimburse the contributions made by a worker who leaves the job before completing the vesting period; (3) requiring the sponsor to transfer the accrued pension rights and the necessary funds to the pension scheme of the new job, assuming the worker has completed the vesting period in the old job and the new job offers an occupational pension (and reimbursing otherwise); and (4) mandating that pension schemes have sufficient funding to make portability payments.

Civil Service Pension Schemes

Some of the reforms discussed above apply equally to civil service pension schemes, especially practices relating to retirement age and the calculation of average pensionable earnings. In addition, the following reforms could be considered.

  • Reduce the high replacement rates that arise from national pensions and civil service pensions in those countries where both can be collected. The preferable option would be to prohibit beneficiaries from drawing both pensions simultaneously. If this prohibition is not feasible, replacement rates can be reduced in a number of ways: (1) reducing civil service pensions by some percentage of the national pension that an individual receives, (2) reducing the accrual rate, and (3) directly capping the replacement rate of the combined pension benefits an individual could receive.

  • Freeze civil service pensions at current levels. In countries where the civil service pension schemes are being phased out, freezing those pensions would not only achieve savings but would also be equity improving. Workers who joined civil service later are not eligible for civil service pensions, but those who were grandfathered in and are currently collecting pensions are benefiting from the more generous retirement conditions and the noncontributory nature of the discontinued civil service pension schemes.

  • Make the civil service pension a contributory scheme. Requiring that employees contribute would level the playing field with the private sector and reduce the net fiscal cost of the scheme. This measure would also provide relatively immediate and easily collectable revenues because they would be directly withheld from salary payments.

  • Undertake actuarial reviews of the civil service pension schemes. As occurs for the national schemes, an actuarial review should be carried out every three years to determine the PAYGO cost the budget is expected to face. Such reviews would be a natural way to gather information that would enable more in-depth analysis of the sustainability of the pensions fund’s finances, highlight the magnitude of the problem, and help address the insufficient information and lack of transparency that seem to affect the schemes currently.

  • Reform pension arrangements for special civil service groups (teachers, legislators, police) so that they gradually converge to those of the national scheme. In many cases these reforms are unlikely to generate large savings, but reducing the generosity of special pensions can help build political capital to push for the difficult but unavoidable reforms required to put public finances on a sustainable path.

  • In the medium term, gradually phase out civil service pension schemes where they still exist. Dominica, Grenada, and Saint Lucia have begun the phasing out by making new workers ineligible for civil service pensions. If a supplementary pension scheme for civil servants is still viewed as necessary, one alternative would be to create a new, voluntary, defined-contribution scheme to top up civil servants’ national pensions. The defined-contribution nature of the scheme would ensure that benefits and contributions are appropriately aligned, and that benefits are portable. Such a system would be best created at the regional level to reduce administrative costs and ensure strong governance.


The public sector is the main provider and financer of education services in most OECS/ECCU countries. At about 5 percent of GDP, average government spending on education in member countries is lower than that in small island countries but relatively high compared with the Latin American region and similar to that in OECD countries (Table 7.5). In per capita terms, public education spending in the region is higher than that in either small island economies or in Latin America. The high level of spending reflects the strong commitment to education by governments in these countries. The policy agenda in the post-independence period emphasized access to primary and secondary education. As a result, by 2000 all countries in the union had achieved the Millennium Development Goal of universal primary education. However, there is significant disparity in spending across countries. As a share of GDP, public education spending in Grenada, Saint Lucia, and St. Vincent and the Grenadines is much higher than the average for the OECS/ECCU group. In per capita terms, spending is highest in St. Kitts and Nevis (Figure 7.9).

Table 7.5Education Expenditures
CountryPublic education spending (Percent of GDP)
Antigua and Barbuda2.6
St. Kitts and Nevis4.1
Saint Lucia5.5
St. Vincent and the Grenadines8.0
OECS/ECCU weighted average4.9
Small island economies6.2
Latin America3.9
Source: Country authoritiesNote: The small island economies group includes Barbados, Belize, Fiji, Jamaica, Maldives, Samoa, Sri Lanka, Timor-Leste and Vanuatu. Figures are for 2009 or latest year available. OECD = Organization for Economic Cooperation and Development.
Source: Country authoritiesNote: The small island economies group includes Barbados, Belize, Fiji, Jamaica, Maldives, Samoa, Sri Lanka, Timor-Leste and Vanuatu. Figures are for 2009 or latest year available. OECD = Organization for Economic Cooperation and Development.

Figure 7.9Public Expenditure on Education versus GDP per Capita in Island Economies

Source: IMF staff estimates.

Despite the relatively high expenditures, educational outcomes have not been strong as shown by available outcome indicators, such as the number of students who pass the Caribbean Secondary Education Certificate (CSEC) examination (Figure 7.10). Dominica, with the lowest spending on education among union members on a per capita basis, has the highest percentage of students passing the CSEC exam. Conversely, St. Vincent and the Grenadines, with one of the highest public education spending levels as a share of GDP and on a per capita basis, has one of the lowest percentages of students passing the exam.

Figure 7.10Percentage of Students Passing the Caribbean Secondary Education Certificate Examination

Source: Saint Lucia, Ministry of Education and Culture, 2009.

Note: Average of scores in mathematics, English A, biology, physics, and chemistry.

Main Issues

Student-teacher ratios have been declining in all OECS/ECCU countries because education systems have yet to adjust to demographic changes. The student-teacher ratios for primary education and secondary education have declined to about 20 or below in most OECS/ECCU countries (Figure 7.11). The ratios are now lower than in other Caribbean and Latin American countries and below international standards. The World Bank recommends student-teacher ratios of 25 and 30 for primary and secondary education, respectively, based on international evidence that increasing student-teacher ratios to those levels would not harm outcomes (World Bank, 2005a, 2005b, 2005d; and Hanushek, 1998). The number of enrolled students has been falling steadily as a result of declining birth rates. However, the total number of teachers has increased considerably in several OECS/ECCU countries, reflecting, in part, an inertial, input-based approach to budgeting based on historical spending rather than on the number of students. But in some countries, the increase in the number of teachers reflects a conscious policy. For example, St. Kitts and Nevis expanded the range of programs offered by schools despite lacking sufficient numbers of students for each of the programs. In St. Vincent and the Grenadines, additional secondary school teachers were hired in anticipation of higher enrollment resulting from the policy on universal access to secondary education; however, enrollment did not increase as expected, leading to a decline in the ratio. Demographic trends point to a further reduction in the student population in both primary and secondary education in the region. Unless the number of teachers drops significantly, the student-teacher ratio will decline even further.

Figure 7.11OECS/ECCU: Number of Students and Teachers in Primary and Secondary Education

Sources: Country authorities; and IMF staff estimates.

The resulting high spending on wages and salaries is crowding out other, more productive spending. In some countries (Dominica, St. Vincent and the Grenadines) salaries account for more than 90 percent of current education outlays and in others (St. Kitts and Nevis) it is close to 90 percent. This leaves few resources for supplies and materials, operating and maintenance services, and teacher training. Underprovision of these inputs can have a detrimental impact on the quality of education.

Figure 7.12Ratio of Trained Teachers to Total Number of Teachers

Source: UNESCO Institute for Statistics (http://www/

A high proportion of teachers in public schools have not received professional training; in some countries, this deficiency has been addressed through costly government training programs.40 On average, fewer than half the teachers at the secondary level and about two-thirds at the primary level have received preservice training (Figure 7.12).41 The ratio of trained teachers to total teachers varies widely across the countries, but in some the ratio is less than 40 percent at the secondary level. International evidence shows that low teacher qualification leads to weaker educational outcomes (Darling-Hammond, 1999; and Rowe, 2003). Furthermore, lack of adequate training leads to lower efficiency because repetition rates increase. Governments have sought to improve training through leave benefits to pursue higher education degrees. Depending on length of service, teachers may take study or duty leave and still receive part or all of their salary while they study full time. The study-leave program attempts to improve the quality of teachers already in service, rather than ensuring that newly hired teachers are properly trained. Thus, the leave benefits merely treat the symptoms rather than the underlying causes of an undertrained teaching force, and therefore are possibly a more expensive way of dealing with the problem. In addition, it is unclear whether the bonding conditions attached to the leave benefits allow the government to reap the benefits of its investment.42

Cost recovery in tertiary education is low in most countries. Typically, tuition and fees cover only a small portion of total costs of tertiary education. In fact, cost recovery of tertiary education spending is considerably higher in Latin America and North America than in the OECS/ECCU.43 Similar to what is observed internationally, the benefits of subsidized fees in tertiary education accrue mostly to the non-poor in OECS/ECCU countries. For example, household survey data indicate that in Saint Lucia, only 11.4 percent of students in tertiary education come from poor households. Greater effort could also be placed on means-testing scholarships that would allow more needy students to attend universities while retaining some funding for top-performing students regardless of need.

Administrative expenses are relatively high because of the small size of the countries, and some subsidized programs are not well targeted. Larger countries spend roughly half of the amount OECS/ECCU countries do on education administration as a share of recurrent spending in education. The fixed costs of developing curricula, information systems, and guidelines for the education sector drive this difference. Inefficient targeting of some programs, such as school feeding and subsidies for tuition and textbooks, also raise equity concerns (see the section on social assistance).44

Short-Term Policy Reforms

One focus of policy reform is to increase the student-teacher ratio in primary and secondary schools. The aim should be to achieve a student-teacher ratio closer to 25 or 30 as recommended by the World Bank. The student-teacher ratio could be increased by a combination of strategies, such as freezing the hiring of new teachers and allowing the number of existing teachers to decline by attrition. In addition, in schools where student-teacher ratios are the lowest but enrollment is sufficiently high, classes can be consolidated within the same school; in schools where enrollment levels are relatively low, multigrade teaching could be implemented. Mandating a minimum class size may be a complementary measure to help guide consolidation.

Second, consideration should be given to greater cost recovery in publicly provided tertiary education, by reducing budget financing to universities and by means testing for scholarships. Increasing the share of financing for colleges from student tuition has advantages in addition to reducing the cost to the budget. First, it will strengthen the response of tertiary institutions to new labor market conditions because they will need to attract students to finance their activities. Second, it will give students appropriate incentives to finish their studies on a timely basis. Savings from this measure will vary depending on the extent to which cost recovery is increased and the size of the budgets of public tertiary institutions. More emphasis could also be placed on means testing for scholarships, which would allow more needy individuals to attend universities, in line with other efforts to improve the targeting of social assistance benefits. At the same time, scholarship funding should be retained for top-performing students.

Medium-Term Policy Reforms

Medium-term reforms include the financing and organization of primary and secondary education to better match resources with demographic changes. The reforms could include the following:

  • Establish a system that bases funding on the number of students to automatically adjust financing so that spending on teachers and schools declines as fewer are needed. At some point, closing or merging schools may be required to achieve higher student-teacher ratios.45 Consolidating schools would likely require increased public spending on transportation as well as higher costs for students (such as additional money needed to travel to schools that may now be further away from home). Therefore, governments should carefully monitor developments to ensure that educational outcomes do not suffer.

  • Consider establishing regulations on training qualifications for new hires. Hiring qualified teachers may be less costly than current programs that train teachers after they have entered the classroom; however, establishing a pool of qualified, younger teachers will take time. The authorities should consider allowing only individuals with certified teaching credentials to apply for teaching positions, as recommended by the World Bank in some countries in the region (World Bank, 2005a). If the high cost of preservice training relative to current compensation levels dissuades too many talented individuals from pursuing teaching careers, a review of the compensation package may be useful. Once enough trained teachers are in place, the study- and duty-leave benefits could be phased out. In the meantime, adjustments to teacher salaries could be based on a distinction between those with certified training and those without, so that trained teachers receive favorable treatment.

  • Evaluate the efficiency of the bonding arrangements in leave programs and scholarships. Leave programs and scholarships are a sizable component of government spending, so evaluating the efficiency of these provisions would be useful. Similar arrangements in the private sector should be part of the assessment. For example, in Antigua and Barbuda, private sector arrangements for study and duty leave often require the individual to pay the education and living expenses up front, and only after the bonding conditions are met is reimbursement provided. This structure is more likely to ensure that bonding conditions are met and that the employer reaps the benefits of its investment, compared with current practices in public sector contracts.

  • Deepen regional cooperation to achieve economies of scale. Improved coordination among OECS/ECCU countries could help reduce administrative costs without sacrificing the quality of education. At the primary and secondary levels, economies of scale could be achieved in areas such as curriculum development. Cooperative efforts could be especially relevant at the tertiary level, so that different countries specialize in particular fields.

Health Care

Public financing and delivery dominate the health care systems in OECS/ECCU countries. Antigua and Barbuda is the only country in the region that finances public health care through a payroll tax.46 As a share of GDP, public spending on health has averaged between 3 and 4 percentage points in member countries during 2007–10. Government expenditure on health accounts for more than half of total health spending in OECS/ECCU countries; in some it is as high as two-thirds.47 External financing from donors tends to play a large role in the financing of capital expenditures related to health care.48 Private financing is most prevalent in Antigua and Barbuda and Saint Lucia. In Saint Lucia, private health insurance covers individuals employed in the formal private sector and a majority of government employees. User fees for public health facilities (mainly for hospital services) apply in all countries except Antigua and Barbuda; but they tend to finance, on average, only about 10 percent of hospital spending and thus contribute only a small fraction of health financing needs.

Health care indicators compare favorably with relevant comparator groups. According to World Bank public expenditure reviews, health indicators are generally good with respect to the countries’ income levels (World Bank 2003, 2005a, 2005b, 2005d, 2005e). In addition, they tend to compare favorably with other small island economies (Table 7.6).

Table 7.6Health Outcomes and Spending across the OECS/ECCU and in Small Island Economies
CountryLife expectancy at birthInfant mortality per 1,000 live birthsAge-standardized mortality rate per 100,000 (Noncommunicable diseases)Age-standardized mortality rate per 100,000 (Cardiovascular)Age-standardized mortality rate per 100,000 (Cancer)Age-standardized mortality rate per 100,000 (Injuries)Per capita total expenditure on health (Purchasing power parity international $; average exchange rate)
Saint Lucia751352220512867677
Antigua and Barbuda741167429616045964
St. Vincent and the Grenadines711267428915264457
St. Kitts and Nevis731469142410843959
Small island economies702670134811992523
Source: World Health Organization.Note: Data are for latest year available. For most countries this is 2007 for life expectancy, 2004 for standardized death rates, and 2006 for remaining variables. The small island economies group comprises The Bahamas, Barbados, Belize, Fiji, Jamaica, Maldives, Palau, Papua New Guinea, Samoa, Solomon Islands, Sri Lanka, Timor-Leste, Tonga, Trinidad and Tobago, and Vanuatu.
Source: World Health Organization.Note: Data are for latest year available. For most countries this is 2007 for life expectancy, 2004 for standardized death rates, and 2006 for remaining variables. The small island economies group comprises The Bahamas, Barbados, Belize, Fiji, Jamaica, Maldives, Palau, Papua New Guinea, Samoa, Solomon Islands, Sri Lanka, Timor-Leste, Tonga, Trinidad and Tobago, and Vanuatu.

Main Issues

Primary care facilities are underutilized, increasing the cost of providing primary care. Many patients seek primary care at more costly outpatient and emergency wards in hospitals for two reasons—user fee policies and concerns about the quality of care at primary health care facilities.49

  • Low user fees for hospital services. Access to primary care is free at public primary care facilities in all countries in the OECS/ECCU. All but one country have user fees for hospital services, but these fees are very low and have not been adjusted for many years in most countries.50 Low hospital fees do not provide adequate incentives to patients to use primary care facilities even if use of the facilities is free. As a result, emergency departments in secondary care hospitals are often clogged by walk-in patients with nonemergency conditions whose ailments range from coughs and colds to chronic illnesses.51 This problem is further exacerbated by widespread exemptions from fees and problems with their collection.

  • Concern about quality of care in primary care facilities. Patients consider the quality of primary care provided at hospitals to be superior to that provided at primary care facilities. This attitude is partly due to the fact that primary care facilities are open for limited periods during the day (mainly during working hours) while hospitals are open around the clock. Another reason is that in some countries (Antigua and Barbuda) physician absenteeism is reportedly high at primary care facilities. Quality of care concerns thus arise when patients are discharged from the hospital back to the community health center, but physicians are then unavailable to ensure continuity of care.

Expenditures on pharmaceuticals and medical supplies have been rising in the region. The demographic and epidemiological transitions occurring in the region have led to rapidly increasing spending on medical supplies and pharmaceuticals. These items, mainly procured by secondary care hospitals, are among the leading cost drivers for budgetary recurrent health expenditure in several OECS/ECCU countries (Saint Lucia, St. Kitts and Nevis, and Grenada). Moreover, user fees for drugs are not only low but bear no relation to actual costs, thereby contributing further to increased pharmaceutical expenditures.52

Table 7.7Saint Lucia: Allowances as a Share of Personal Emoluments for Established Health Sector Workers(Percent)
Type of allowance2006/072007/082008/09Revised budget 2009/10Budget 2010/11
Total allowances for health sector workers8.57.97.610.715.2
Call out2.
On call1.
In lieu of private practice0.
Anesthetist fees0.
Night differential00.
Source: Government of Saint Lucia, Ministry of Finance.
Source: Government of Saint Lucia, Ministry of Finance.

Another important component of costs has been compensation of health care providers. As noted earlier, collective bargaining agreements have increased compensation spending, including for health care workers. For example, in Saint Lucia, the collective bargaining agreement for health workers resulted in a real 19 percent increase in average personal emoluments per established health sector worker for 2006/07–2010/11. In St. Vincent and the Grenadines, three-fourths of the increase in government health spending during 2006–10 was attributable to higher compensation of employees. In Saint Lucia and Antigua and Barbuda, spending on allowances has been increasing and has accounted for a growing share of compensation of established health care workers in recent years.53Table 7.7 illustrates this issue for Saint Lucia.

Health facilities face significant difficulties in collecting fees. Patients are required by law to pay user fees before service is provided, with the exception of medical emergencies. In practice, no health facility can deny care if user fees are not paid in advance. Therefore, health facilities face difficulties collecting fees from patients after services are provided. Moreover, broad, untargeted exemptions from user fees result in additional revenue losses.54 To illustrate the extent of revenue losses, in Saint Lucia, Victoria and St. Jude hospitals collect on average only about 45 percent and 55 percent of their patient bills, respectively, as a result of the combined impact of exemptions and collection problems.

The use and maintenance of hospital infrastructure are suboptimal in some countries. In Grenada, St. Kitts and Nevis, and Saint Lucia, hospital occupancy rates are low by international standards, leading to unnecessarily high operational costs.55 Health facilities are also insufficiently maintained, which will eventually lead to costly rebuilding or repairs. Spending on operations and maintenance in Grenada, St. Kitts and Nevis, Saint Lucia, and St. Vincent and the Grenadines has been on average about 1 percent or less of total health spending in 2007–10. Increasing compensation costs and spending on medical supplies and drugs are crowding out these expenditures in most OECS/ECCU countries.56

Public health programs for preventing disease are not sufficiently exploited. Health professionals in the public and private sectors share the perception that certain diseases that lead to high drug consumption could be most cost-effectively combated by public health campaigns to inform the population of the risk factors and how to prevent the conditions. Two widely cited illnesses that would benefit from this approach are diabetes and hypertension.

Cost controls are weak. The costing of health services and drugs provided in the public benefits package, and formal processes to examine their cost-effectiveness, are deficient. No information is typically available from public health facilities about the full cost of providing specific types of treatment. Health information systems are also not sufficiently developed to allow all expenditures related to providing treatment to patients to be recorded. Adequate health information systems would also enable the monitoring of whether proper treatment protocols are followed.57 Lack of pertinent information has led to decisions to expand the benefits package without a clear understanding of the cost implications, thus contributing to expenditure overruns and arrears.58

Public spending on health care is likely to rise at an even faster rate in the future, underscoring the need for effective cost control. Noncommunicable diseases—often more expensive to treat than infectious diseases—have become a greater share of the disease burden in the region. The aging of the population will also lead to higher health care costs. These trends are likely to continue. Figure 7.13 presents projections for health care spending in the region during the period 2010–50 under alternative scenarios. The projected average increase in public health expenditures ranges from 0.3 percent to 8.9 percent of GDP over the period. Projections for individual countries show large variations, ranging from 0.3 percent of GDP by 2020 in Dominica under the low cost growth assumption to 10 percent of GDP by 2050 under the high cost growth assumption. The projection methodology is discussed in Appendix 7A.

Figure 7.13Projected Central Government Health Expenditure in the OECS/ECCU Region

Source: IMF staff estimates.

Note: Low-growth scenario: Public health spending per capita grows at the same rate as GDP per capita growth. High-growth scenario: Public health spending per capita grows 2 percentage points faster than GDP per capita growth.

Short-Term Policy Reforms

  • Adjust the system of user fees to better reflect the cost of providing health services and encourage better use of primary care facilities. Potential adjustments include:

    • Charging cost recovery user fees for secondary and tertiary hospital services. Such fees would not only help increase patients’ awareness of actual health care costs, but would also encourage them to make better decisions about their own care, would shift part of the costs from the government to the private sector, and would encourage greater use of primary health care facilities. User fees for medicines should also be designed as a function of the actual cost of the drugs to encourage both greater use of generic substitutes and more efficient consumption.

    • Redesigning the user fee system so that it (1) exempts only the poor and those with chronic conditions who cannot afford to pay;59 (2) exempts services that treat infectious diseases; and (3) does not allow private insurance to cover user fees, thereby blunting incentives for efficient use. The number of exempted groups should be reduced to exclude those who could pay.

    • Implementing complementary reforms. For the adjustments to the user fee system to be most effective, the quality of service at community health centers needs to be improved by extending the hours of operation and reducing physician absenteeism in countries where it is a problem.

  • Reduce absenteeism through the increased use of fixed-term contracts for physicians and through enhanced monitoring and enforcement. Physicians and other health workers should be monitored to verify that they are present at primary care facilities and carry out their work as expected. Penalties should be based on the number of infractions, should be large enough to deter such behavior, and should be enforced to instill credibility. In particular, random monitoring, as is being done in Saint Lucia, and increased flexibility to hire new physicians on a fixed-term contract basis, may improve performance.

  • Implement a nominal freeze on compensation, including allowances, of health care personnel. Given the large real increases awarded to health care personnel in 2008–10 and the governments’ tight financial constraints caused by the global financial crisis, a case could be made for no further compensation increases, especially in allowances, for some time.

  • Strengthen the incentives for the collection and payment of user fees. In compliance with the relevant health legislation, health facilities should be allowed to deny treatment in nonemergency situations if the required user fees are not paid in advance. In addition, the discretionary budget allocations and transfers received by health facilities should be reduced, and in return the facilities should be allowed to keep the fees they collect from patients.

  • Ensure adequate financing for proper maintenance of hospital infrastructure and for additional public health programs. Part of the resources freed by the rationalization of the wage bill could be used for infrastructure maintenance and public health programs. Public health education campaigns for prevention, which could start at the school level, will increase people’s awareness of major risk factors for certain diseases (e.g., AIDS, diabetes, hypertension), and may contribute to a reduction in their incidence. Thus, they can help contain treatment costs over the medium term.

  • Undertake cost audits in health facilities and enhance the ability of health care information systems to track costs. Audits are critical for evaluating existing benefits packages, analyzing the consistency of the costs of providing those services with resource constraints, and avoiding unfunded expansion of new medical services, which could undermine the financial sustainability of the system. Audits will also provide critical information for redesigning the schedule of user fees and for laying the groundwork for medium-term reforms. Improvements to health information technology to support the reforms are also needed.

  • Expand the implementation of treatment protocols for different diagnoses. Physicians should be explicitly asked to justify the lab tests, procedures, and prescriptions for any given diagnosis on the basis of the treatment protocol for that diagnosis. This measure could be complemented by random reviews by a medical board of justifications submitted for the procedure. If the justifications are deemed to be invalid, the physicians’ salaries could be reduced in line with the cost of the excess medical care provided.

Medium-Term Policy Reforms

  • Review the public benefits package to ensure its consistency with financial constraints. Given the strong medium-term financial pressures resulting from the demographic and epidemiological transitions, current benefits need to be reviewed to ensure that appropriate prioritization and cost-control decisions are made to maintain their affordability.

  • Rationalize the use of hospital infrastructure. Because hospitals have relatively low utilization rates, all possible means of rationalizing the number of health facilities and the distribution of hospital beds should be pursued.

  • Consider basing some portion of hospital reimbursement on prospective methods, such as diagnosis-related groups (DRGs). Under such methods, hospitals are paid a fixed amount to treat a patient based on diagnosis at admission. Just as copayments can be seen as cost sharing on the demand side, prospective payment methods amount to cost sharing on the supply side. These payment methods put part of the provider’s income at risk, so the provider will enhance efficiency and minimize costs. However, DRGs may also encourage providers to increase admissions and to reduce the quality of service. It is thus important to monitor treatment patterns and adherence to standard protocols to ensure appropriate delivery of care financed within a given budget envelope.

  • Deepen regional cooperation in health care to reap benefits from economies of scale. OECS/ECCU countries are already benefiting from regional procurement of pharmaceuticals. Regional action could be expanded further to include other medical supplies and the establishment of a mechanism to evaluate the cost-effectiveness of the drugs and other technologies in the public benefits packages. Because establishing a formal cost-effectiveness agency is a challenging and expensive task, an alternative strategy would be to base reimbursement decisions on the evaluations of agencies from other countries, such as the National Institute of Clinical Excellence in England and Wales. Pharmaceutical costs could also be reduced further by eliminating multiple procurement systems in those countries that still have them, thus increasing the scale economies provided by the OECS Pharmaceutical Procurement Service.

Social Assistance

OECS/ECCU countries implement a wide array of social assistance programs to provide income support and access to basic services to the poor and the vulnerable.60 This assistance includes transfers (both cash and in-kind), labor-market programs (including public works and training programs), welfare programs based on education or health, programs for the disabled, and housing programs. The programs are typically implemented by different ministries and in some countries also by PEs (Antigua and Barbuda).61

Spending on social assistance programs has increased steadily in almost all countries in the region since 2008. Among other reasons for the increase, governments sought to enhance protection for the most vulnerable from the adverse impacts of the recent financial and food and fuel crises. Spending on such programs varies across OECS/ECCU countries—ranging from about 1 percent of GDP in St. Kitts and Nevis to slightly more than 2 percent of GDP in St. Vincent and the Grenadines (Figure 7.14).62 These levels are comparable to average spending of about 1.5 percent of GDP in Latin America and the Caribbean.

Figure 7.14Social Assistance Spending

Main Issues

The numerous social assistance programs do not appear to be underpinned by coordinated social protection strategies (Blank, 2009a, 2009b, 2009c, 2010a, 2010b). As a result, the way in which the different programs fit together in achieving an overall objective is not clear. Social assistance programs in OECS/ECCU countries have some specific shortcomings:

  • Programs lack flexibility to respond to shocks. Financial crises, price shocks, and natural disasters can all have adverse impacts on the poor and the vulnerable. Protecting the population from shocks requires that programs can be expanded or contracted as appropriate, both in beneficiaries and in benefit levels. Weaknesses in targeting and financing constraints have contributed to program inflexibility. Moreover, because most programs are based on categorical targeting, they are not designed to protect the poor and vulnerable from these shocks.

  • Programs emphasize protection rather than human capital building. Almost all programs provide some sort of income transfer to meet the immediate needs of beneficiaries. With the exception of the school feeding programs, none of the benefits is conditioned on behaviors that help recipients get out of poverty.

The lack of coordinated social protection strategies is also reflected at the budgeting and execution stages, compounding inefficiencies. In the absence of clearly identified priorities, no mechanisms for fiscal prioritization of programs are in place. This lack of priorities is reflected in the budgeting process, during which budgets are determined by the previous year’s allocations and not in response to changing needs and priorities. Social assistance programs are implemented by different government ministries and related agencies, each with its own administrative system and procedures, thus precluding synergies, leading to duplication of benefits and beneficiaries, and stretching limited capacity. High start-up and administrative costs also accrue.

Programs are not effectively evaluated and monitored, and no procedures are in place to ensure accountability. Policies and procedures for many programs are not fully documented. As a result, implementation arrangements are open to the discretion of government officials. In some cases, eligibility requirements are stated in terms of unverifiable income, subjective criteria, or both. These weaknesses create opportunities for political interference in the selection of beneficiaries. Moreover, no systematic monitoring takes place to ensure that program objectives are being met. Deficiencies in record keeping and centralization of beneficiary information, compounded by limited computerization, greatly reduce the ability to effectively monitor delivery of social assistance.

Programs suffer from significant errors of exclusion and inclusion as highlighted by poverty assessments. For example, public assistance programs covered fewer than 20 percent of the poor and indigent in Dominica, 25 percent of the elderly in Grenada, and 9 percent of the poor in Saint Lucia. Although these low coverage levels may be partly due to financing constraints, targeting issues are also a likely contributing factor as evidenced by considerable leakage to non-poor groups. For example, children from both the highest and lowest quintiles appear to benefit almost equally from school feeding programs in Saint Lucia, while close to half of the benefits of the welfare assistance program accrues to the non-poor. Similar issues are observed in St. Kitts and Nevis and Grenada.63,64

Figure 7.15Grenada: Spending on Social Assistance and Poverty Incidence

Sources: Country authorities; and IMF staff estimates.

These weaknesses limit the effectiveness of social assistance spending in reducing poverty or preventing additional increases in poverty during crises. For example, in Grenada, despite large and increasing social assistance spending and a rise in real income levels, poverty has increased by 5.6 percentage points to 37.7 percent between 1998 and 2008 (Figure 7.15). Although progress has been made in reducing extreme poverty, Grenada’s headcount poverty incidence is still one of the highest in the region.

Short-Term Policy Measures

  • Compile a complete inventory of all social assistance programs and evaluate them to judge how well they meet their objectives and are targeted to the poor. This review should include information such as the number of programs, number of beneficiaries, total spending on benefits, eligibility conditions, targeting mechanisms, and administrative costs to deliver the benefits.65 This information would help the governments identify which programs could be scaled up and which should be phased out. Undertaking this exercise would set the stage for the development of a coherent social assistance strategy over the medium term.

  • Integrate social assistance programs under a single institution to create opportunities for synergies among programs, limit duplication of benefits and beneficiaries, make the most of limited capacity, save on administrative and personnel costs, and pave the way for more efficient provision of services.

  • Establish a central beneficiary registry for all programs. Such a system would enable program managers to cross-check information about beneficiaries and prevent abuse of the programs and duplication of services. It would also make possible more regular and frequent performance assessments and would provide information that could inform program design. A central beneficiary registry could reduce administrative costs by streamlining procedures for application, targeting, and approval. A registry has been set up in Dominica with assistance from the World Bank, currently covering three programs—public assistance, the Education Trust Fund, and the school feeding program (see Box 7.1).

  • Introduce objective and transparent targeting mechanisms and strengthen accountability.66 All nontransparent eligibility criteria should be replaced with explicit and verifiable criteria, building on the results of the evaluation of existing programs. Household survey data provide important information for construction of poverty maps and collection of other relevant information for targeting. Policy and procedures should be fully documented in operations manuals and strictly enforced to reduce the discretion of government officials in the selection of beneficiaries. Dominica recently set up a Beneficiary Identification System under which beneficiaries are selected using a proxy means test (see Box 7.1).

Box 7.1The Beneficiary Identification System in Dominica

As part of its efforts to enhance its capacity to better target, plan, and implement social assistance programs, the Government of Dominica, with assistance from the European Union and the World Bank, has set up a Beneficiary Identification System (BIS). The system is meant to create a central registry to be used for the selection and registration of beneficiaries for key social assistance programs—public assistance (PA), school feeding, and the Education Trust Fund (ETF). The BIS is expected to ensure transparency and consistency in the selection of beneficiaries, reduce leakage by focusing on the poorest, and enhance the cost-effectiveness and efficiency of the system.

Under the BIS, all applications are received using a newly designed Benefits Application form that is available at village council offices and at the Welfare Division. A welfare officer is assigned to evaluate an applicant’s condition using a Household Assessment Form. The form includes 22 questions relating to the applicant’s social conditions, such as family composition, education of household members, occupation, housing conditions, access to basic services, and ownership of assets. These questions are designed so that a proxy means test (PMT) can be performed to determine the eligibility of the applicant for social assistance. Information from the assessment form is then fed into a data bank that stores the information and provides an overall score for the applicant.

Applicants are classified into three groups according to the PMT score: (1) indigent—those who receive all benefits covered under the three programs; (2) poor—those who receive PA plus one benefit under the ETF; and (3) vulnerable—those who are not currently eligible for benefits under any of the three programs but will be considered for assistance if an emergency occurs.

A number of complementary reforms have also been implemented for the three social assistance programs covered by the BIS. Policies and procedures have been documented, operations manuals have been prepared, and relevant staff have been trained. A monitoring plan has also been approved. All three programs now have functioning program databases to track beneficiaries and expenditures.

All systems are in place for implementation of the BIS based on PMT. Information on all existing PA beneficiaries has been collected using the Household Assessment Forms. In the first instance, existing PA beneficiaries will be evaluated using the PMT. Once the system is fully operational, all future applicants for any of the three programs will be evaluated using the PMT system. In late 2010, the Ministry of Education decided to award free school meals and ETF benefits to PA clients. This will bring greater coherence and transparency to the social protection system.

Medium-Term Policy Measures

  • Rationalize social assistance programs on the basis of a coherent social protection strategy. This strategy should specify objectives, fiscal allocations for core programs that match these objectives, implementation arrangements, accountability and control mechanisms, and monitoring and evaluation plans. One set of programs could address key dimensions of extreme poverty and intergenerational transmission of poverty, and cover the elderly and the disabled who are not covered by the social security scheme. Small, labor-intensive public works programs could form a second group that could be scaled up or down in response to economic shocks. A third group could concentrate on skills training and development. Saint Lucia is already developing such a strategy following a Cabinet decision in 2010.

  • Link social assistance programs to actions that promote human capital development. Conditioning social assistance on regular visits to health clinics, and attendance and progression of children in schools, for example, can be an effective method of breaking the intergenerational transmission of poverty (see Box 7.2). However, these conditional programs require sophisticated administrative and institutional frameworks. A possible approach could be to phase in these programs gradually, consistent with the development of administrative capacity.

  • Strengthen information management systems. Computerization would enhance the ability of the governments to monitor and evaluate programs and would also reduce administrative costs. Programs and beneficiaries can be tracked over time and evaluated with respect to inputs and outcomes. A regional approach to automation might be explored with the goals of improving cost-effectiveness and overcoming capacity constraints.

Capital Spending

Public investment in the OECS/ECCU is relatively high.67 During the period 1970–2010, annual public investment spending averaged about 9.3 percent of GDP in these countries, 3 percentage points of GDP higher than in other Latin American and Caribbean countries. For example, public investment is on average less than 4 percent of GDP in Argentina, Bolivia, Brazil, Chile, Costa Rica, Paraguay, and Peru. As illustrated in Figure 7.16, capital spending hovered around 5 percent of GDP during most of the 1990s. It grew considerably in the early 2000s, and reached 9 percent of GDP in 2002 and 2006.68 Since then capital spending levels have started to decline, and further adjustment is estimated to have occurred in 2010 as the region dealt with the fiscal impact of the global financial crisis and stepped up efforts to reduce fiscal imbalances and debt levels.

Figure 7.16Public Investment Spending in the OECS/ECCU, 1990–2010

(Weighted average)

Source: Country authorities.

High public investment spending did not translate into faster growth in OECS/ECCU countries (Figure 7.17). St. Kitts and Nevis is the only ECCU member country with both high public investment and relatively fast GDP growth. Fast-growing countries (Antigua and Barbuda and St. Kitts and Nevis) improved productivity mainly by finding niches for their high-end tourism services (IMF, 2011). Also, GDP growth in non-OECS/ECCU Latin American countries with public-investment-to-GDP ratios of about 4 percent of GDP (5 percentage points below the average of the OECS/ECCU) was comparable to GDP growth in the OECS/ECCU. See Box 7.3 for an analysis of public investment productivity in the region.

Figure 7.17Public Investment and Growth in Selected Latin American Countries, Average, 1970–2009

Sources: IMF and country authorities’ data; and IMF staff calculations.

Note: Values above (below) average plus one-half of one standard deviation of the relevant variable are labeled high (low). The vertical and horizontal lines in the figure are the averages plus one-half of one standard deviation.

Box 7.2Conditional Cash Transfer Programs in Selected Latin American Countries

Conditional cash transfer programs provide poor households with money on the condition that they make specified investments in the human capital of their children. Beneficiary households are asked to comply with a series of conditions—generally school enrollment and attendance of children, and regular visits to health centers. These programs have often provided entry points for reforming badly targeted subsidies and upgrading the quality of safety nets. Impact evaluation studies have concluded that these programs have contributed to increased consumption levels among the poor and have reduced poverty—by a substantial amount in some countries. This box briefly reviews three such programs in Latin America.

Solidario, Chile

Introduced in 2002, Solidario provides cash transfers to extremely poor households. In the first instance, households identified through a proxy means test are provided with psychosocial support through a local social worker for two years. During this period, households work with the social worker to assess their needs and design a strategy to exit extreme poverty. The strategy includes actions that households commit to undertake, which then become conditions under the program. During these two years, households receive direct cash transfers with the amounts declining over time, and preferential access to a number of other social programs. After the two-year period, households continue to receive smaller cash transfers for three years, along with preferential access to social services but the social worker services are terminated. Impact evaluations revealed significant gains in a number of different dimensions after two years of program implementation. School enrollment and adult literacy showed significant improvement. Enrollment in public health programs and preventive health visits for children also showed significant increases. A strong take up of employment programs also occurred.

Oportunidades, Mexico

Originally named Progresa, this program was developed as a replacement for a number of poorly targeted food subsidies. Launched in 1997, the program aims to eradicate extreme poverty by promoting human capital investment by the poor. Beneficiary households are provided with cash transfers, school supplies, and nutritional supplements, conditional on regular school attendance by children and preventive health care visits. In 2002, the program was renamed Oportunidades and expanded nationally with several changes to its objectives and operational features. By 2007, the program had reached 5 million households (approximately a quarter of the population) at a total cost of 0.4 percent of GDP. Impact evaluations have indicated that the program succeeded in reaching the hard-core poor, with poverty being reduced by 17 percent in communities served by the program. Per capita consumption, school enrollment, and visits to health centers by infants and pregnant women increased significantly among beneficiary households.

PATH, Jamaica

Jamaica introduced the Program of Advancement through Health and Education (PATH) in 2001, replacing three previous programs. PATH has two components: (1) child assistance grants that provide education grants to children conditional on regular school attendance (for ages 6 through 17) and health grants (through age 6) conditional on periodic visits to health clinics; and (2) social assistance grants to adults. Initially, the latter component was conditional on regular visits to health clinics, but this was later withdrawn. Another important benefit was the waiver of certain education or health fees for PATH households. Evaluation of the program indicates that 58 percent of the benefits accrue to the poorest quintile. The program has been effective in increasing school enrollment and preventive health visits by children.

Sources: This box draws from Grosh and others (2008); Fiszbein and Schady (2009); and Levy and Ohls (2007).

Box 7.3Productivity of Public Investment in OECS/ECCU Countries

Most of the member countries pursued expansionary fiscal policies before the 2008 recession, mainly by increasing public investment. Countries aimed to sustain growth in the short run and crowd in private investment by boosting public infrastructure assets. However, the impact of these policies has generally not been as expected. Notwithstanding the other factors contributing to the slowdown in growth, the effectiveness of these public investments has been a key issue. Table 7.3.1 provides the results of the estimation of a standard growth model in which capital accumulation is split between the public and private sectors. The basic model is estimated for OECS/ECCU and non-OECS/ECCU countries using a panel regression technique. The results suggest that the accumulation of public capital assets did not have a significant effect on GDP growth in OECS/ECCU countries, while public investment was found to spur growth on average in non-OECS/ECCU countries. This result is consistent with findings in IMF (2011) and Sahay (2005).

Table 7.3.1Public Capital and Growth in OECS/ECCU and Non-OECS/ECCU Countries
Non-OECS/ECCU countriesOECS/ECCU countries
Public capital0.07**0.0381
Private capital0.253***–0.33
Labor force0.551***0.51
Source: World Bank, 2005e.***p< 0.01, **p< 0.05, *p < 0.1
Source: World Bank, 2005e.***p< 0.01, **p< 0.05, *p < 0.1

Natural disasters and the quality of public investment could explain the apparent disconnect between public investment and growth in the OECS/ECCU, beyond the debt overhang. A significant fraction of capital budgets has gone toward reconstruction following natural disasters, which have increased in frequency since the 1990s. However, a closer examination of post-hurricane spending patterns in some OECS/ECCU countries shows that there is a tendency for subsequent increases in capital spending to become permanent rather than temporary (World Bank, 2005e). Natural disasters that forced high reconstruction costs and a reduction in the capital stock could partly explain this disconnect between public investment and growth. At the same time, it could also be that public investment has not been allocated to the right sectors. An analysis of the distribution of public investments by the World Bank found that the share of expenditure going to economic infrastructure has declined steadily, while spending on transfers increased and the wage bill remained high.

Main Issues

Weaknesses in managing public investment projects appear to have constrained the efficiency of public investment spending. World Bank public expenditure reviews for several countries in the region have flagged weaknesses in the planning, execution, and monitoring of public sector investment programs (PSIPs), which have contributed to high and inefficient investment spending. Projects in a country’s PSIP are generally selected after negotiations between ministries and cabinet members, and often announced to the public before technical discussions, if any, of their feasibility have taken place. Some countries have formalized the process by introducing explicit procedures for project review and selection, but this practice is not yet common across the region. Many reportedly “strategic” public investments—such as sports stadiums, marinas, airports, and fishing complexes—were not supported by overall strategic plans embodied in sector analyses that indicated how these investments were to impact growth. Financial and technical requirements for operating and maintaining these investments are discussed only toward the end of construction, and consultations with stakeholders are rare during both project preparation and implementation. The main issues that have adversely affected the productivity of public investment are discussed below.

The coverage and reporting of public investment in governments’ fiscal accounts are inadequate. On one hand, projects carried out by PEs (statutory bodies and state corporations) or financed using special arrangements (such as public-private partnerships or guaranteed loans to private companies) are not included in the PSIPs. As reported in World Bank public expenditure reviews, creative accounting practices have also contributed to reducing the coverage of the PSIP.69 On the other hand, the figures presented in the capital budget and the PSIPs of member countries do not accurately reflect public investment in fixed capital. Some projects included in PSIPs are recurrent expenditures. For example, OECS/ECCU countries by and large include in their PSIPs all projects for which they will seek external financing (loans and grants) irrespective of the nature of the project. In addition, the recurrent costs of capital projects are not reclassified as such in the budget documents. Although the project summary forms contain information on the economic classification of the projects’ expenditures, these expenses are included wholly in the capital budget.

Despite recent progress, several weaknesses strain the formulation of capital budgets and PSIPs. Most countries in the region have now set up medium-term rolling frameworks for investment planning, but the environment for planning public investment remains weak:

  • First, strategic guidance for public investment decisions is limited. Line ministries derive their investment programs from the country’s medium-term development strategy that sets economy-wide priorities, but the sector- and subsector-level strategies have not been translated into detailed plans consistent with the overall objective in the medium-term strategy. Projects from priority sectors are given perfunctory approval unless major problems arise that require reconsideration.

  • Second, dual budgeting70 is still prevalent in most OECS/ECCU countries (Dominica, Grenada, and St. Vincent and the Grenadines). As a result, budget appropriations for current and capital spending are not well balanced and do not consistently support the countries’ medium-term development strategies. Examples of such unsatisfactory balance can be found in the health sector, for which large new hospitals are being built but with inadequate funding for maintenance and utilities (see the section on health care). Serious inefficiencies then arise because under-budgeting of maintenance and operations spending can significantly reduce the rate of return on capital projects and lead to expensive reconstruction needs in the future.

  • Finally, the plethora of tiny projects, some with questionable priority or impact, also overstretches planning ministries’ limited capacity to manage these programs and their resources for the careful preparation of large and potentially high-return projects.

Investment projects are not clearly prioritized. The project summary forms submitted with proposals for inclusion in the budget are not based on rigorous project analyses, because no formal requirements for full-fledged feasibility studies are in effect. This shortcoming is partly due to capacity constraints.71 Also, no formal requirements are in place for cost-benefit analyses, even for large projects, that would oblige line ministries to build such capacity. Finally, no forum exists in which the results of any such analyses are considered by decision makers in establishing capital spending priorities. As a consequence, decisions to include specific projects in the budgets are often based on the criterion of the availability of finance rather than intrinsic quality and rate of return. The risk of approving unaffordable projects with low or negative rates of return and the politicization of investment decisions thus both increase.72

Lack of clear budget constraints for capital spending appropriations results in low execution rates of the government investment program, especially in Dominica, Grenada, St. Kitts and Nevis, and St. Vincent and the Grenadines. Entities involved in investment planning are not given clear budget constraints. Consequently, a long and unfinanced list of projects evolves, which is not conducive to prioritization at the budget formulation stage.73 Low implementation capacity in line ministries is also a significant constraint. Budgeting is thus unrealistic and based on consistently overly optimistic projections of capital spending, which forces inefficient prioritization at the execution stage. Projects are arbitrarily matched with available resources, contributing to low execution ratios of overall planned capital spending. The low execution rates are mostly in priority sectors. For example, less than half of investment appropriations in education and health are executed in St. Vincent and the Grenadines (Figure 7.18).

Figure 7.18St. Vincent and the Grenadines: Average Public Investment Execution Rate, 2000–05

Source: Government of St. Vincent and the Grenadines, Ministry of Finance.

Note: The execution rate is the ratio of actual investment spending to the corresponding budget appropriations in any given year.

Weaknesses in procurement and commitment controls further undermine cost-effective project implementation.

  • Public Expenditure and Financial Accountability assessments identified deficiencies in the legal framework for procurement and limited implementation and enforcement of existing procurement legislation. These assessments have also noted the lack of open competition for contracts as a critical issue. Moreover, relevant procurement data, such as records of awarded purchases and other financial documents, were generally not available to allow assessment of the methods used to award public contracts.74

  • Expenditure commitment controls apply to only a few categories of spending,75 which has led to an automatic ex post regularization of spending overruns on investment projects through special warrants or supplementary estimates. This practice creates difficulties in ensuring that correct procurement procedures are followed before a commitment can be registered.

These weaknesses translate into low value for money (VfM) in public investment. A World Bank study (World Bank, 2005c) reveals wide disparity in unit costs among projects in different OECS/ECCU countries. The high costs of public investment in some countries result from (1) a lack of standardized procurement rules; (2) over-design by consultants without consideration for budget constraints and actual needs; (3) poor contract management as evidenced by the large number of contracts in which the actual expenditures varied from what was contracted, but standard procedures were not followed to clarify or correct them; and (4) a lack of oversight of ongoing works by ministries due to a shortage of qualified staff. Inadequate monitoring and ex post project evaluation (Dominica, Grenada, St. Kitts and Nevis, and St. Vincent and the Grenadines) limits the ability to take corrective action. Detailed project execution report requirements are not complied with on a regular basis. Ex post project evaluation is only carried out for externally funded projects when required by the financing institution.

No clear legal and institutional framework governs public-private partnerships (PPPs). Member countries do not have integrated national legal frameworks, requiring project regulations to be established separately in the contract for each project through low-level decrees. The absence of a law specifically dedicated to PPPs leaves the environment for PPPs subject to continuous change, and makes the environment more conditional on current socioeconomic contexts, greatly reducing the likelihood of successful long-term outcomes. This weak legal and institutional environment explains the limited number of PPP operations undertaken in the region,76 with greenfield operations focusing on the telecom sector, and limited involvement of private investors in the construction of economic infrastructure such as roads and dams. Given the limited number of such operations, the ministries of finance did not set up dedicated units to monitor PPPs and their associated fiscal risks. There are also no units to promote the use of PPPs in line ministries, mainly because of capacity constraints and the complexity of these operations. Limited private participation in infrastructure has prevented the development of public sector comparators to assess VfM of PPP projects. However, interest in PPPs has been growing in several countries.

Policy Reforms

OECS/ECCU countries can implement a number of measures to rationalize capital spending and enhance its efficiency in the short and medium terms.

In the short term, there may be scope for

  • Improving the coverage of PSIPs and the reporting of public investment. First, PSIPs should include all public investment projects that satisfy certain selection criteria (discussed below) irrespective of their financing arrangements and whether the institution executing the project is within the budget or is a PE. Second, only those projects involving the creation of nonfinancial assets of the government (e.g., public roads) should be classified as capital expenditure. Recurrent spending should be excluded from public investment data and reclassified as current spending in fiscal reports.

  • Giving the ministries of finance (MOFs) a greater gateway role. The MOFs can become more involved by requiring that all investment projects be carried out by public sector institutions, or that those that may create contingent liabilities for the budget above a certain threshold be screened by the MOF.

  • Defining and strictly enforcing simple selection criteria for screening investment projects. For example, investment projects considered for inclusion in the PSIP and ultimately the budget should (1) be submitted with all the information needed to evaluate them (including information on their recurrent spending implications); (2) have gone through appropriate feasibility studies and cost-benefit analyses, especially large projects; and (3) be consistent with sectoral priorities and spending envelopes provided by the MOF to the requesting entity for the current budget year and a certain number of future years.77 The criteria should apply to all public investment projects independently of their financing. All project submissions—including those in priority sectors—not satisfying these basic criteria should be rejected.

  • Applying the simple criteria above to the existing portfolio of capital projects, with the objective of rationalizing the portfolio. The criteria can readily be applied to new projects given that they have not been started, but the criteria should also be applied to ongoing projects. Consideration should also be given to downsizing, postponing, or cancelling existing projects altogether. The net benefits of these measures should be compared with the costs of changing or cancelling the projects embedded in the relevant contracts.

  • Tightening procurement procedures and establishing commitment controls. The tightening of procurement procedures should ensure open competition, especially for large projects, including through greater public disclosure of information on awarded contracts and through complaint mechanisms. Commitment controls should be established to prevent the automatic ex post regularization of overruns and to ensure that procurement procedures are followed. Appropriate sanctions and enforcement mechanisms should also be implemented to ensure compliance with procedures and controls.

  • Strengthening and enforcing monitoring procedures. In particular, monthly monitoring of capital spending execution should be required and credible penalties should be applied to noncompliant institutions.

In the medium term, consideration could be given to

  • Strengthening the multiyear perspective of fiscal planning, expenditure policy, and budgeting. The medium-term plans for public services provision need to be better integrated into the projections of fiscal aggregates. To that end, line ministries should develop credible, multiyear budget estimates. Governments that have developed three-year and program-based budgeting frameworks should complement these frameworks with strategic goals and should prioritize expenditures and link resources to policy objectives anchored into the countries’ poverty reduction strategies.

  • Developing capacity for cost-benefit analysis at the regional level. In the short term, feasibility studies and cost-benefit analyses can be contracted out to get around capacity constraints, but steps should be taken to develop capacity internally over the medium term. Standard manuals on ex ante and ex post evaluation of projects could be developed to guide this process. A regional approach could be considered to take advantage of scale economies, not only in the development of the manuals but also in the training of government officials.

  • Carrying out systematic, ex post cost-benefit analyses of large investment projects. Information on what went right and what went wrong with projects is invaluable for refining project selection criteria to limit the losses from projects with low or negative rates of return.

  • Over the longer term, strengthening the legal framework for PPPs and the government’s capacity to undertake them. This measure would require the introduction of specific PPP laws for countries where there are none. The laws should regulate the initiation and implementation of public projects financed by private funds and specify the broad structure for risk sharing and for addressing issues that could arise in implementing PPP contracts. The laws should also provide a framework for the institutional setup for supporting the development of PPPs in member countries. The countries could also consider developing capacity for conducting VfM assessments (see Box 7.4). The development of this capacity should take place after the more fundamental aspects of capital budgeting have been strengthened. Until that time, new PPP contracts should be avoided, given the substantial fiscal risks they can generate if administrative capacity is weak.

Parastatal Entities

Parastatal entities (PEs) are government units that operate separately from the budget. In OECS/ECCU countries, a variety of organizations operate as PEs, including public financial and nonfinancial corporations, regulatory commissions, extra-budgetary funds (e.g., pension funds), and government agencies.78 Many of these PEs perform noncommercial functions and in the past operated within the confines of the budget. The motivation to separate their operations from the budget was in part due to difficulties in hiring and retaining specific types of workers resulting from pay scale limitations and other rigidities in human resources management within civil service.

Budgetary transfers to these entities have been increasing steadily over the years. Transfers have ranged from 0.2 percent to about 3.1 percent of GDP in recent years (Table 7.8). The increase in the number of PEs partly explains the increase in transfers. In some countries, the number of PEs has increased over the years either to manage new initiatives (e.g., a new authority for the Cricket World Cup or for building a new airport), or because some government departments were transformed into PEs. In Antigua and Barbuda, more than half of the transfers to PEs in 2010 were allocated to a newly created medical center. In St. Vincent and the Grenadines, the sharp increase in transfers to PEs in 2010 reflects allocations to two entities that accounted for more than three-quarters of the total.

Table 7.8Budgetary Transfers to Parastatal Entities(Percent of GDP)
Antigua and Barbuda0.
St. Kitts and Nevis10.
Saint Lucia1.
St. Vincent and the Grenadines0.
Source: Country authorities.

Refers only to transfers to parastatal entities (PEs) in St. Kitts. No information was available for transfers to PEs in Nevis.

Source: Country authorities.

Refers only to transfers to parastatal entities (PEs) in St. Kitts. No information was available for transfers to PEs in Nevis.

PEs account for a significant share of civil service employment and spending in some countries. A paucity of information about PEs makes any assessment of their operations difficult. Data from the social security board indicate that in Antigua and Barbuda, employment in PEs amounted to more than 45 percent of budgetary employment. The corresponding figures for Dominica and St. Vincent and the Grenadines were 27 percent and 35 percent, respectively.79 In Grenada, employment in 9 out of 29 PEs was estimated to be 20 percent of central government employment in 2008. Similar information was not available for other countries. With respect to spending, the most comprehensive data were available for St. Vincent and the Grenadines. In 22 of its 25 PEs, spending was almost 16 percent of GDP in 2007 and the overall deficit of these entities was 3.6 percent of GDP (Table 7.9). These figures indicate that a significant portion of public spending takes place outside the budget and is thus a source of significant fiscal risk.

Table 7.9St. Vincent and the Grenadines: Operations of Parastatal Entities
(Millions of EC$)
Total receipts213.2251.1390.1232.8
Current receipts206.5243.8384.9229.6
Total expenditure332.9324.3449.4288.0
Current expenditure211.2249.8333.2238.8
Capital expenditure121.774.5116.349.2
Overall balance–119.7–73.3–59.3–55.2
(Percentage of GDP)
Total receipts10.511.016.910.4
Current receipts10.210.716.710.4
Total expenditure16.515.922.314.9
Current expenditure10.412.216.612.3
Capital expenditure6.
Overall balance–5.9–3.6–3.0–2.8
Source: Government of St. Vincent, Ministry of Finance.Note: The number of entities for which information is available varies across years. Of the 25 entities identified, information was available for 12 in 2006, 22 in 2007, 16 in 2008, and 14 in 2009. The increase in spending in 2008 was partly due to increased activity at the international airport project, which is being implemented by one of the entities.
Source: Government of St. Vincent, Ministry of Finance.Note: The number of entities for which information is available varies across years. Of the 25 entities identified, information was available for 12 in 2006, 22 in 2007, 16 in 2008, and 14 in 2009. The increase in spending in 2008 was partly due to increased activity at the international airport project, which is being implemented by one of the entities.

Box 7.4Assessing Value for Money in Practice

Value for money (VfM) is essentially an outcome based on the most cost-effective method for public service provision. According to the U.K. Office of Government Commerce, VfM is defined as “the optimum combination of whole-life cost and quality (or fitness for purpose) to meet the user’s requirement.” VfM is rarely synonymous with lowest price.

The term is also sometimes used to refer to the appraisal and assessment processes that led to this outcome. The United Kingdom’s experience is particularly useful for understanding how the appraisal and assessment are conducted in practice to ensure VfM.

  • The first step (investment planning phase) is to choose the sectors for which a PPP would be an appropriate procurement route. PPP should be chosen when services are contractible and outputs can be clearly defined, and should be targeted to areas in which private sector participation can bring important efficiency gains.

  • The second stage (project level phase) is to prepare the business case for a particular project. At this point, the project has been identified as belonging to a sector having “PPP potential,” but the decision to opt for either a PPP or traditional public procurement has not been made yet. The decision rule can take two different forms: (1) the use of a public sector comparator (PSC), or (2) a more qualitative assessment. The PSC is a hypothetical risk-adjusted estimate of the cost of service or output delivery if the public sector were the supplier. It is expressed in net present value terms and is based on existing data on costs and benefits associated with the provision of the specific output under consideration. An alternative methodology is to use qualitative methods to make the decision. This alternative responds to the criticism that PSCs are based on uncertain assumptions and prone to significant error and manipulation. The qualitative assessment would consider the advantages and disadvantages of being in a long-term contract relationship, including the credibility of potential bids to maintain standards of quality, the risks of litigation and renegotiation of contracts, and similar factors.

  • The third stage is to correct “optimism bias.” At this stage, if a PSC has been used, the public sector cost is increased to take account of what is often referred to “as the systematic tendency for appraisers to be over-optimistic about key project parameters.” Optimism bias is corrected by using estimates of upward cost drift from similar public sector projects. These adjustments are sometimes very large. For example, the National Health Service in the United Kingdom recently estimated that for projects with a capital value of between £10 million and £25 million, the cost drift averaged about 40 percent, while for larger projects it was about 30 percent.

  • The fourth and final stage is the procurement test. If, at the end of the third stage, a particular project is still perceived as having the potential to provide VfM in a PPP, tender documents are prepared and the project moves to the final procurement phase. At this stage, a variety of considerations are assessed, including the quality of competition for the contract, the success achieved in transferring risk to the private sector, the credibility of the cost estimates provided by the bidders, and other considerations (for example, whether the VfM will be achieved by reducing workers’ benefits or wages). If a particular project fares poorly in some of these dimensions, consideration is given to delaying it or reconsidering traditional public procurement methods.

Sources: This box draws on Grout (2005); Leahy (2005); and Monteiro (2005).

PEs’ debt and debt-service obligations also pose significant fiscal risks. Available information suggests that parastatal debt could be significant (Table 7.10). In Saint Lucia, parastatal debt amounted to about 5 percent of GDP in 2010. However, in St. Kitts and Nevis and St. Vincent and the Grenadines, government-guaranteed debt of PEs was reported to be 29.5 percent and 15.2 percent of GDP, respectively, at end-2009. Moreover, in both Grenada and St. Vincent and the Grenadines, the governments recently had to assume debt-service obligations of delinquent PEs. In Grenada, such operations in 2005 cost the budget about 2 percent of GDP, while in St. Vincent and the Grenadines, the costs were about 3 percent of GDP in 2010.

Table 7.10Government-Guaranteed Debt of Parastatals(Percent of GDP)
Antigua and Barbudan.a.n.a.n.a.n.a.
St. Kitts and Nevis27.826.729.5n.a.
Saint Lucia5.
St. Vincent and the Grenadines12.313.815.2n.a.
Source: Country authorities.Note: n.a. = Not available.
Source: Country authorities.Note: n.a. = Not available.

Main Issues

Compliance by PEs with the legal governance and reporting requirements is weak. The level of compliance varies across countries, but the MOFs have audited financial statement information for only a small subset of PEs, if any at all. In Antigua and Barbuda and Grenada, information on fewer than half of all PEs is available at the ministry; in some other countries, even less. A notable exception is St. Vincent and the Grenadines, where the MOF recently launched a largely successful initiative to collect such information, even though the PEs are not legally required to provide it to the ministry. However, even when financial information is available, it is often dated, and in some cases, as in Antigua and Barbuda, the data are incomplete and insufficiently reliable to evaluate the financial condition of the PEs.80

Incomplete financial information and limited resources prevent adequate monitoring of performance and fiscal risks arising from PEs. In most countries no dedicated unit or body monitors the operations of PEs. In Grenada and St. Vincent and the Grenadines, only one staff member in the MOFs was allocated the task of collecting and monitoring the reports for all PEs. In Dominica, this task is the responsibility of the debt unit in the MOF, which has only two employees. Even in countries with formal monitoring units or oversight bodies, as in St. Kitts and Nevis and Saint Lucia, the effectiveness of monitoring is limited because of the absence of higher frequency reports from the PEs, capacity constraints (Saint Lucia), or lack of sufficient legal authority (St. Kitts and Nevis and St. Vincent and the Grenadines). As a result, consolidated financial statements for PEs are not available except in St. Vincent and the Grenadines. The growing importance of PEs and the significant spending they undertake make this level of monitoring insufficient for guarding against the fiscal risks they pose.

The proliferation of PEs, combined with insufficient oversight and limited information on their financial operations, is a source of concern. Country experiences have shown that the absence, or lack of enforcement, of appropriate governance arrangements and reporting requirements can create inefficiencies and loss of fiscal control. In particular, this can lead to

  • A multiplicity of institutions with related responsibilities but no mechanism to prioritize and avoid duplication, thus generating high overhead costs and wasting limited public resources.

  • A reduction in budget discipline. PEs may be used to circumvent regular budget procedures and their checks and balances, resulting in reduced transparency, a loss of control over spending, and the potential creation of significant contingent liabilities.

  • Reduced credibility for meeting deficit and debt targets. When the financial situations of the PEs are not well known, the government may suddenly face a realization of contingent liabilities, forcing it to intervene in ways that generate deviations from the deficit and debt targets.

Short-Term Policy Reforms

  • Establish and enforce legislation on governance and reporting requirements for PEs and strengthen sanctions for noncompliance. In particular, deadlines for submitting audited financial statements need to be clarified and penalties for noncompliance strengthened and enforced. Penalties should include institutional and personnel sanctions if the reporting obligations are not met in line with the legislation. Institutional sanctions could include the withholding of transfers and loan approvals. The legislation should also clarify that PEs should submit, on a timely basis, audited financial statements that adhere to international accounting and auditing standards.

  • Ensure that consolidated reports on the operations of the PEs are regularly prepared. Compiling the information in line with international government finance statistics standards, such as in the Government Finance Statistics Manual (GFSM 2001), could help significantly in achieving consolidated reporting. Bridge tables could be prepared from the audited financial statements of the PEs into the GFSM 2001 methodology to prepare the information for analysis. The authorities could seek technical assistance on these issues, if needed, including from the IMF’s Statistics Department.

  • Strengthen the oversight of PEs by MOFs. Dedicated monitoring units within the MOFs could be created or existing ones buttressed with the necessary resources and legal powers to allow them to carry out their functions effectively. A possible approach could be to focus initially on the largest PEs and those that are likely to have the largest impact on the budget. Setting up a monitoring framework for PE oversight would also be helpful. The Caribbean Regional Technical Assistance Center (CARTAC) is providing technical assistance to St. Vincent and the Grenadines for this purpose. This effort could possibly be expanded to other member countries. Appendix 7C presents a short summary of the main elements of a possible monitoring framework for PEs in OECS/ECCU countries. The monitoring process should lead to the identification of measures to restore the financial sustainability of troubled PEs and minimize their reliance on budget transfers or debt takeovers in the short term. The MOFs should also play a prominent role in approving PEs’ investment projects. New initiatives should be subject to careful feasibility studies to ensure that they provide VfM. Savings from these measures are likely to more than compensate for any costs arising from the creation and strengthening of monitoring units.

  • Increase the scrutiny required to justify the creation of new PEs and periodically assess whether their objectives remain relevant and are being met in a cost-effective way. The first recommendation could be implemented by requiring an independent assessment of the business case for the establishment of a PE. The primary purpose of the business case—and its assessment—should be to show that a PE would meet the objectives at a lower cost than would other alternatives, including existing government departments. The second recommendation could be met by introducing automatic sunset clauses for some institutions in the relevant legislation and requiring periodic independent reviews of PEs’ performance with respect to the objectives for which they were created and of their cost effectiveness.

Medium-Term Policy Reforms

The functions and rationale of existing PEs should be reviewed in the context of the review of the role of government discussed in the earlier section on civil service employment and wages. The review of the role of government will likely identify areas of duplication, which could create rationalization opportunities and savings for the budget. PEs that no longer perform functions considered necessary or that can be performed adequately by the private sector should be liquidated or divested. Many of the parastatals carry out government functions, so careful consideration should also be given to whether autonomy from the budget is really necessary for efficient execution of their activities. If autonomy were found not to be needed, PEs could be merged into an appropriate government department and included in the budget.

Over the long run, PEs must be financially sustainable. They should be able to cover their costs from their own resources. Any subsidy has to be related to the government’s policy goals and included explicitly in the government budget as a transfer. Such expenditure would have to be evaluated in light of other priorities and demand on government resources.


Expenditure rationalization will be a critical element in the fiscal consolidation efforts in the OECS/ECCU region. The analysis in this chapter shows that a broadly based expenditure rationalization strategy could reduce the expenditure-to-GDP ratio by 3 to 5 percentage points on a cumulative basis over the medium term. The main findings and recommendations of the chapter are as follows:

  • Addressing the high cost of government should be a top priority in the region’s fiscal consolidation efforts. The government wage bill makes up the single largest expenditure item in the budgets of OECS/ECCU countries. Its large size relative to total expenditures makes reducing the wage bill an essential element of the expenditure consolidation effort. Short-term measures to control wage spending include temporarily freezing wages and employment, streamlining benefits and allowances, and eliminating known duplications and overlaps. Medium-term measures include reviewing the role of the government to identify government units that could be streamlined, closed, or divested. Some countries in the region are already moving in this direction.

  • Social security systems are a key source of fiscal vulnerability in OECS/ECCU countries. Unfavorable demographic trends and relatively generous pension benefits threaten the sustainability of the region’s national pension systems. Although most schemes are currently operating in surplus, under current policies most will soon begin to incur cash flow deficits. Even though some reforms are being implemented in some countries, additional efforts are needed in all countries to increase the retirement age and contribution rates, change pension calculation formulas, and reduce front-loaded accrual rates. These proposed reforms can increase pension system reserves in these countries between 4 and 15 percentage points of GDP by 2050. In addition, investment policies for system reserves should be strengthened to increase portfolio diversification and the rate of return.

  • Health systems in the OECS/ECCU face a number of challenges. Noncommunicable diseases—which are more expensive to treat than infectious diseases—now account for a greater share of the disease burden in the region. This, along with the aging population, will increase future health care costs significantly. Member countries can take a number of measures to enhance the efficiency of health spending over the short and medium terms. Improving the targeting of user fees could increase revenue while still ensuring that the poor have access to basic health care services. Steps need to be taken to better control costs through auditing, enhancing health information systems, and restraining growth in compensation. Rationalizing the configuration of health care facilities could improve service quality and efficiency. Finally, enhanced regional cooperation in health care may generate substantial cost savings as has already occurred in pharmaceutical procurement.

  • Government spending on education in OECS/ECCU countries is relatively high compared with the rest of the Caribbean, reflecting the strong commitment to education by governments in member countries. Most countries in the union have achieved the Millennium Development Goal of universal primary education. However, the efficiency of public spending in education has been low. Inefficient targeting of some programs, such as school feeding and subsidies for tuition and textbooks, also raise equity concerns. Enhancing the efficiency of spending could lead to better education outcomes while also generating cost savings for the budget. Measures could include increasing student-teacher ratios, consolidating the number of primary schools consistent with the decline in birthrates, better targeting subsidy programs, and introducing or raising user fees at the tertiary level. Greater regional cooperation could also reduce administrative expenses and improve economies of scale.

  • OECS/ECCU countries implement a wide array of social assistance programs to provide income support and access to basic services for the poor and the vulnerable. The numerous programs are implemented by a variety of government agencies, leading to high start-up and administrative costs. Weaknesses in targeting have resulted in substantial leakage to non-poor groups and have limited the effectiveness of these programs in fighting poverty. Introducing objective, transparent targeting mechanisms and consolidating the institutions involved in delivering social assistance can enhance the efficiency of these programs. Linking social assistance to actions that promote human capital growth can help break the intergenerational transmission of poverty.

  • Public sector investment programs (PSIPs) need to be rationalized to better link public investment and economic growth. Capital spending is relatively high in OECS/ECCU countries. Nevertheless, as a result of weaknesses in planning, execution, and monitoring of PSIPs, this high spending has not translated into faster growth. Strategic guidance for public investment decisions is limited, and appropriations for current and capital spending do not consistently support the countries’ medium-term objectives. The execution rate for government investment projects is low, while weaknesses in procurement and commitment controls result in low value for money (VfM) of public investment. Steps to rationalize capital spending and enhance its efficiency include improving the coverage and reporting of public investment in fiscal reports, making procurement procedures transparent, tightening commitment controls, defining and strictly enforcing project selection criteria, and strengthening and enforcing monitoring procedures. Medium-term measures could focus on strengthening capacity for project appraisal (perhaps through regional cooperation) and the legal framework, and increasing the capacity for implementing capital projects under public-private partnerships (PPPs).

  • Parastatal entities (PEs) constitute a significant source of fiscal risks as demonstrated in Grenada and St. Vincent and the Grenadines, when the governments had to take over the debt obligations of some PEs. PEs account for a significant share of civil service employment and spending, and their debt obligations are also large in some countries. Compliance with legal governance and reporting requirements by PEs is weak, and incomplete financial information prevents adequate monitoring. The oversight of PEs by the ministries of finance should be enhanced by strengthening and enforcing existing legislation on governance and reporting requirements. Regular consolidated reports on the operations of the PEs need to be prepared. In the medium term, the authorities should undertake a review of the functions and rationale of existing PEs with a view to decreasing their number.

Although not discussed in this chapter, successful expenditure rationalization will also require strengthening public financial management systems. OECS/ECCU countries are at various stages of formulating and implementing reform measures for strengthening these systems. Continued technical assistance from donors, CARTAC, and the IMF would be critical in this regard.

Appendix 7A. Health Spending Projections

A simplified model was developed to project health expenditure. The model focuses on demographics and all other factors combined. It serves to illustrate a range of possible spending trajectories under different assumptions about spending growth relative to income growth and is based on projected demographic changes.

  • The central element for the projections is a profile of public health spending per capita for five-year age cohorts from the OECD.

  • The shape of the average OECS/ECCU public health spending per capita profile is assumed to be the same as in OECD countries.

  • The profile of absolute spending in local currency units for each age cohort is calculated using data on public health spending, the number of people in each age cohort, and the relative spending weight of each cohort.

  • The shape of this expenditure profile remains constant over the projection period.

  • Changes in the number of people within each five-year age cohort, based on U.S. Census Bureau projections for each of the OECS/ECCU countries, yield overall spending changes resulting from demographics.

  • An increase in the spending level at a given age (i.e., an upward shift of the expenditure profile) represents changes in spending caused by technology, income, insurance, and any other factors excluding demographics. Following convention, this increase is referred to as “excess cost growth.”

  • The baseline scenario is that public health spending per capita grows 1 percentage point faster than projected GDP per capita growth (exclusive of demographic changes) for each age cohort.

  • Given the large degree of uncertainty in nondemographic factors, two alternative scenarios are simulated to demonstrate the following possible spending trajectories:

    • Low-growth scenario—public health spending per capita grows at the same rate as GDP per capita growth.

    • High-growth scenario—public health spending per capita grows 2 percentage points faster than GDP per capita growth.

  • Data on health spending measured in local currency units is taken from the WHO National Health Accounts (WHO, 2010).

Appendix 7B. Alternative Approaches to Targeting

Experience has shown that cost-effective targeting can be achieved through combining various targeting methods (Coady, Grosh, and Hoddinott, 2004):

  • Geographic targeting. Household survey and national census data can be used to allocate the overall transfer budget across regions throughout the country. Typically, data from the census and household surveys are combined to construct a socioeconomic index for, say, districts. Poverty maps are often used to estimate the share of low-income households in each region. Similar maps can be constructed for other socioeconomic outcomes such as education, health, or access to physical infrastructure. These maps then provide a basis for targeting disadvantaged areas and allocating budgets.

  • Self-selection targeting. Designing programs so that lower-income households self-select in and higher-income households self-select out can substantially reduce the costs of targeting associated with processing claims by ineligible households and the costs of leakage of benefits to higher-income groups. Low transfer levels relative to, say, median income, or requiring households to apply at regular intervals to program offices with a clear statement that only those in need are eligible, can help to achieve this goal. However, care needs to be taken to ensure that low-income households are aware of the program and its objectives and have easy access to program offices, and that the information they are required to provide is accessible to them and can be verified at a low cost.

  • Categorical targeting. A strong correlation between individual or household socioeconomic characteristics and low incomes, if it exists, can be used when determining eligibility. These characteristics can be used to include or exclude applicants. Such an approach needs to be strongly founded on prior and convincing empirical evidence, such as information available from recent national household surveys.

  • Community targeting. Information available at the community level can often be useful when identifying households in need, especially if current economic status is not easily captured by a more statistical and deterministic approach to targeting. This information could come from program officers living in the community who have “local knowledge” not available in any survey. Prominent local persons (e.g., teachers, doctors, local elders, and religious groups) are often involved, usually in the form of a committee. However, care needs to be taken to ensure that these persons have objectives consistent with those of the program.

Appendix 7C. Main Elements of an Oversight System for Parastatal Entities in OECS/ECCU Countries

The operations of PEs can sometimes pose significant risk for the central government budget.81 In more than one country in the region the central government has had to take over the debt and debt-servicing obligations of a PE. Improved governance and enhanced monitoring could help contain fiscal risks. The elements of such an oversight system can be grouped into the broad categories discussed below.

Legislative Framework

The oversight system should be underpinned by standardized legislation, policies, and regulations governing the activities of PEs in each country. The government’s overarching financial responsibility, accountability, and administration legislation should contain a section that sets out in general terms the operating, accountability, and reporting provisions to be applied. These general terms should deal with such matters as (1) how the mandates of PEs are established; (2) appointment of personnel, including the board of directors; (3) good governance practices; (4) initiatives and transactions requiring approval by ministers; (5) authority and responsibility of ministers, particularly the minister of finance, and other officials, directors, and managers; (6) preparation of annual plans, operating and capital budgets, and reviews; (7) borrowing authority; (8) reporting requirements; and (9) audit requirements. The legislation should also clarify the powers of the government to intervene in the affairs of PEs. Legislation should also specify sanctions for noncompliance. In addition, each PE would have its own specific legislation that would define its mandate(s) and would set out any elements unique to that corporation (e.g., specific appointments to the boards of directors, pricing and service considerations, and so forth).

Central Oversight Committee

At the center of the oversight system should be a cabinet committee chaired by the prime minister or the minister of finance (they are often the same person in the Caribbean). The committee should comprise a small number of other ministers and senior officials, and it should have the responsibility to (1) establish regulations, guidelines, and directives for the PEs, including in areas such as compensation policies and borrowing authority; (2) in cooperation with line ministers responsible for each entity, establish the mandate for each PE; (3) approve the PEs’ annual plans (including operating and capital budgets); (4) monitor the performance of the PEs against approved plan targets and guidelines; (5) appoint the members of the boards of directors of the entities, including the chief operating officers and auditors; and (6) call for strategic and operational reviews of the PEs as necessary.

Role of the Ministry of Finance

Responsibility for supporting the oversight committee with secretarial and analytical functions should be centralized within the MOF. The MOF should collect all the financial statements, reports, and plans. It should analyze the plans and reports and prepare summary briefings—highlighting emerging issues and recommendations—for the oversight committee based on the information prepared by the PEs. With respect to the financial statements, the MOF should regularly prepare an analysis of the consolidated balance sheets for each PE with comments and recommendations as required.

Annual Plans

Each PE should prepare an annual plan that (1) identifies key initiatives for the coming and future years; (2) sets quantitative targets (performance measures) for its objectives; (3) identifies significant threats that could adversely affect achievement of the targets; (4) provides operating and capital budgets; and (5) provides actual and pro forma financial statements (i.e., financial projections for the five-year period covered by the plan) both with and without the proposed changes in strategy or capital expenditure. These plans should be discussed by the boards of directors and submitted to the oversight committee for approval before the start of the next financial year.

Strategic Plans and Operational Reviews

A mandatory review of each PE should be undertaken periodically, focusing on both fundamentals and operations. Such reviews will study environmental factors that may have changed significantly over time and in ways that affect the PE’s business prospects and possibly even its raison d’être. The reviews would focus on factors that do not typically change much over a short period (e.g., one year), but may do so over an extended time. Particular events (such as a major capital project proposal) may also prompt the need for a strategic review. The review will examine the basic assumptions about the PE’s current business, the implications of fundamental changes in the competitive landscape, and other key variables that may affect the business of the PE. This review will then form the basis for the development of a strategic plan that will define the strategy or direction of the entity, specify the objectives and goals of the organization for the coming years, and guide the allocation of its resources (human, physical, and financial) to pursue those objectives and goals. This strategic plan will then serve as the basic framework for subsequent annual plans.

Like strategic reviews, operational reviews should also be undertaken periodically. However, unlike the former, the latter do not involve structural changes in business strategies with respect to external factors. Rather, operational reviews, during which internal operating procedures and processes are assessed, have the objective of improving efficiency.

Performance Reports

Each PE should prepare an annual report, approved by its board, then submitted to the oversight committee generally within three months after the termination of its financial year. The annual report should include both a management report (or a commentary) and the latest financial statements. The management report should include information on financial and policy performance, any noteworthy issues, accomplishments, performance measures compared with targets, and key financial ratios. The annual financial statements should be prepared in line with international accounting standards and should be audited.

In addition to these annual reports, PEs should also prepare quarterly management reports for the line ministries to whom they report and provide these to the oversight committee. Quarterly reports should include pertinent current financial data and ratios. They should also track progress in the achievement of plans and set out the main issues and concerns. In particular, the financial and operational risks, their time horizon and potential impact, together with the actions taken and the likelihood of success, should be covered in these reports. These in-year reports would form the basis of an early warning system by providing information for the MOF to evaluate fiscal risks. The MOF, in turn, would alert the oversight committee of any potential fiscal risk and put forward recommendations for corrective action as required.


    AbedG. and S.Gupta2002Governance Corruption and Economic Performance (Washington: International Monetary Fund).

    AlesinaAlberto F. and SilviaArdagna1998Fiscal Adjustments: Why They Can Be ExpansionaryEconomic Policy: A European Forum Vol. 13 No. 27 pp. 487517.

    AlesinaAlberto F. and SilviaArdagna2009Large Changes in Fiscal Policy: Taxes Versus SpendingNBER Working Paper No. 15438 (Cambridge, MA: National Bureau of Economic Research).

    AlesinaAlberto F. and R.Perotti1995Reducing Budget DeficitsDepartment of Economics Discussion Papers No. 759Columbia University.

    BlankL.2009aGrenada Social Safety Net Assessment (Washington: World Bank, United Nations Children’s Fund, and United Nations Development Fund for Women).

    BlankL.2009bSt. Kitts and Nevis Social Safety Net Assessment (Washington: World Bank, United Nations Children’s Fund, and United Nations Development Fund for Women).

    BlankL.2009cSaint Lucia Social Safety Net Assessment (Washington: World Bank, United Nations Children’s Fund, and United Nations Development Fund for Women).

    BlankL.2010aAntigua and Barbuda Social Safety Net Assessment (Washington: World Bank, United Nations Children’s Fund, and United Nations Development Fund for Women).

    BlankL.2010bSt. Vincent and the Grenadines Social Safety Net Assessment (Washington: World Bank, United Nations Children’s Fund, and United Nations Development Fund for Women).

    BruntonP.D. and P.Masci2005Workable Pension Systems: Reforms in the Caribbean (Washington: Inter-American Development Bank and Caribbean Development Bank).

    CoadyD.M.Grosh and J.Hoddinott2004Targeting of Transfers in Developing Countries: Review of Lessons and Experience (Washington: World Bank).

    Darling-HammondL.1999Teacher Quality and Student Achievement: A Review of State Policy EvidenceCenter for the Study of Teaching and Policy (Palo Alto, CA: Stanford University).

    de WitH.I.Jaramillo J. Gacel-Ávila and J.Knight (eds.)2005Higher Education in Latin America: The International Dimension (Washington: World Bank).

    Dominica Ministry of Health2010Strategic Plan for Health: Investing in Health—Building a Safer Future Vol.1Government of the Commonwealth of Dominica.

    FiszbeinA. and N.Schady2009Conditional Cash Transfers: Reducing Present and Future Poverty (Washington: World Bank).

    GroshM.E.C.del Ninno E. Tesliuc and A.Oeughi2008For Protection and Promotion: The Design and Implementation of Effective Safety Nets (Washington: World Bank).

    GroutP.2005Value-for-Money Measurement in Public-Private PartnershipsEIB Papers Vol. 10 No. 2 pp. 3358.

    HanushekE.1998The Evidence on Class SizeW. Allan Wallis Institute of Political Economy Occasional Paper No. 98–1 (Rochester, NY: University of Rochester).

    International Monetary Fund (IMF)2011Eastern Caribbean Currency Union: Selected Issues IMF Country Report No. 11/32 (Washington: International Monetary Fund).

    IzvorskiI. and S.Kahkonen2008Public Expenditure Policies in Southeast EuropeWorld Bank Working Paper No. 134 (Washington: World Bank).

    KumarManmohan S. and J.Woo2010Public Debt and GrowthIMF Working Paper 10/174 (Washington: International Monetary Fund).

    LeahyP.2005Lessons from the Private Finance Initiative in the United KingdomEIB Papers Vol. 10 No. 2 pp. 5971.

    LevyD. and J.Ohls2007Evaluation of Jamaica’s PATH Program: Final Report (Washington: Mathematica Policy Research).

    McDermottJ. and R.Wescott1996An Empirical Analysis of Fiscal AdjustmentsIMF Working Paper 96/59 (Washington: International Monetary Fund).

    McRobertsS.2009Improving Governance and Oversight of State-Owned Enterprises in St. Vincent and the Grenadines (Barbados: CARTAC).

    McRobertsS.2010An Analysis of Financial Risk in State-Owned Enterprises of St. Vincent and the Grenadines (Barbados: CARTAC).

    MitchellO.S. and D.M.Osborne2005Old Age Pension Reform in the English Speaking Caribbeanin Workable Pension Systems: Reforms in the Caribbeaned. by P.D.Brunton and P.Masci (Washington: Inter-American Development Bank and Caribbean Development Bank).

    MonroeH.2009Can the Eastern Caribbean Currency Union Afford to Grow Old?IMF Working Paper 09/38 (Washington: International Monetary Fund).

    MonteiroR.2005Public-Private Partnerships: Some Lessons from PortugalEIB Papers Vol. 10 No. 2 pp. 7381.

    PalaciosR. and E.Whitehouse2006Civil Service Pension Schemes Around the WorldSocial Protection Working Paper No. 602 (Washington: World Bank).

    PettinatoS. and J.D.Cassou2005A Needs Assessment of Pension Systems in the English Speaking Caribbeanin Workable Pension Systems: Reforms in the Caribbeaned. by P.D.Brunton and P.Masci (Washington: Inter-American Development Bank and Caribbean Development Bank).

    PotterBarry H. and J.Diamond1999Guidelines for Public Expenditure Management (Washington: International Monetary Fund).

    RemmerK.L.2010Why Do Small States Have Big Governments?European Political Science Review Vol. 2 No. 1 pp. 4971.

    RoacheS. and T.N.Rasmussen2007Social Security in the Eastern Caribbean Currency Unionin Eastern Caribbean Currency Union: Selected Issues IMF Country Report No. 07/97 (Washington: International Monetary Fund).

    RoweK.2003The Importance of Teacher Quality as a Key Determinant of Students’ Experiences and Outcomes of Schooling” (Camber, VIC, Australia: Australian Council for Educational Research).

    SahayRatna2005Stabilization, Debt, and Fiscal Policy in the CaribbeanIMF Working Paper 05/26 (Washington: International Monetary Fund).

    Saint Lucia Ministry of Education and Culture2009Education and Culture Digest: Past Trends, Present Position and Projections up to 2009/10” (Castries, Saint Lucia: Ministry of Education and Culture).

    TsibourisG.M.A.Horton M.J. Flanagan and W.S.Maliszewski2006Experience with Large Fiscal Adjustment IMF Occasional Paper No. 246 (Washington: International Monetary Fund).

    Van RijckeghemC. and B.Weder2001Bureaucratic Corruption and the Rate of Temptation: Do Wages in the Civil Service Affect Corruption, and by How Much?Journal of Development Economics Vol. 65 No. 2 pp. 30731.

    WeigandC. and M.Grosh2008Levels and Patterns of Safety Net Spending in Developing and Transition CountriesSocial Protection Discussion Paper No. 0817 (Washington: World Bank).

    World Bank2003Saint Kitts and Nevis: OECS Fiscal IssuesReport No. 25185-LCR (Washington: World Bank).

    World Bank2005aDominica: OECS Fiscal IssuesReport No. 27467-DOM (Washington: World Bank).

    World Bank2005bGrenada: OECS Fiscal IssuesReport No. 28344-GRD (Washington: World Bank).

    World Bank2005cOrganization of Eastern Caribbean States: Towards a New Agenda for GrowthReport No. 31863-LAC (Washington: World Bank).

    World Bank2005dSaint Vincent and the Grenadines: OECS Fiscal IssuesReport No. 30855-SVG (Washington: World Bank).

    World Bank2005eA Time to Choose: Caribbean Development in the 21st Century” (Washington: World Bank).

    World Bank2010Strengthening Caribbean Pensions: Improving Equity and SustainabilityReport No. 47673-LAC (Washington: World Bank).

    World Health Organization2009Global Health Observatory (Geneva).

    World Health Organization2010. World Health Statistics 2010 (Geneva)

Countries covered in this chapter are Antigua and Barbuda, Dominica, Grenada, St. Kitts and Nevis, Saint Lucia, and St. Vincent and the Grenadines.

The adverse impact of debt on growth has been shown to be higher for low-income and emerging market economies than for advanced economies (Kumar and Woo, 2010).

These figures are a lower bound because they exclude both spending by parastatal entities and investment financed through special financing arrangements. Insufficient cross-country information is available to include these types of spending.

Excludes spending by statutory bodies.

Public spending on natural disaster insurance is not examined in this chapter because all the countries covered are members of the Caribbean Catastrophic Risk Insurance Facility, which provides a range of insurance products aimed at reducing the economic impact of natural disasters. More information on the facility is available at

In this section, the term “civil service” is used rather than “public service,” although in some countries there is a difference between the two.

The wage bill is defined as personal emoluments and wages paid to all budgetary central government workers.

The room to further increase government revenue as a share of GDP is assumed to be limited, given that several reforms have already been implemented.

For example, labor survey information available for Saint Lucia suggests that unemployment has exceeded 14 percent during most years since the mid-1990s.

This analysis is based on an econometric estimate of civil service and private sector wage differentials for workers with comparable levels of human capital. It uses the Blinder-Oaxaca decomposition of hourly income, using age and education level as its measure of human capital endowments. If total income is used instead of wage income, the estimated premium would be 26 percent. A previous World Bank study estimated the premium at 12 percent in 2005 (World Bank, 2005e). In most countries, the expectation is for the premium to be negative—wages in the civil service sector tend to be lower because employees typically have greater job security and enjoy a number of other pecuniary and nonpecuniary benefits that may not be available in the private sector.

Civil service pensions are typically adjusted in line with changes in civil service wages.

In most OECS/ECCU countries, there are two types of civil servants—established and nonestablished. Although there are slight variations in the definitions, established workers are generally those who are appointed through the public service commissions. Nonestablished workers are generally contractual employees.

This problem may force inefficient increases in civil service employment. For example, the Grenada health authorities pointed out that when a new need arises they have to request additional positions and budget without being able to eliminate existing positions that are no longer needed.

The constitution of Grenada establishes that if a worker is made redundant, a full pension must be paid to him or her for life irrespective of age or length of service.

In Grenada, large salary overlaps between successive positions provide insufficient incentives to civil servants to upgrade their skills and move to a new grade.

For example, the average compression ratio for member countries of the OECD is 8 (Izvorski and Kahkonen, 2008). One reason for the low ratio in Saint Lucia is that senior management wages have not been adjusted for 16 years.

In Grenada, wages paid to nonestablished workers are reflected under budget code 340, which is intended to record spending for professional/consultancy services. Workers hired under this code are not reflected in the government’s budgetary workforce statistics. In other countries in the union, budget documents record the salary expenditures for such employees under “wages” but do not provide information on their total number.

The term “selective” means allowing critical personnel to be hired when the function cannot be fulfilled by redeployment of staff. Departure of critical personnel is typically a small percentage of attrition. Care would need to be taken to avoid defining all personnel in any particular sector as critical because doing so would reduce the scope of achievable savings. On a number of occasions in OECS/ECCU countries all security, education, and health workers were defined as critical, which reduced significantly the effectiveness of hiring freezes because these sectors comprise the large majority of civil service employment.

For example, and to illustrate the importance of using a comprehensive compensation definition, a simple comparison of salaries that does not take into account the impact of pension benefits (which in some OECS/ECCU countries exceed 100 percent of salaries) would be highly misleading.

Some of the OECS/ECCU countries have tried outsourcing with varied results (Dominica, Grenada, and Saint Lucia). Savings from these measures were limited and the quality of service varied. The experience highlights the importance of truly competitive bidding and strict monitoring to ensure that a high-quality product is delivered.

With regard to other age-related benefits, Antigua and Barbuda and Saint Lucia, for example, provide partial pensions and age/retirement grants for those who did not accumulate sufficient contributions to be eligible for full pensions. There are also transitional pensions for those who could not have accumulated sufficient contributions to qualify for a pension because they retired soon after the creation of the pension fund. Not all benefits are provided in every country. For example, Antigua and Barbuda does not offer employment injury benefits.

Social assistance pensions are discussed in a later section of the chapter. Medical payments do not refer to comprehensive health insurance but to coverage for specific risks, such as hospitalization.

In Antigua and Barbuda, for example, teachers become eligible for pensions after 25 years of service, regardless of age. The income replacement rate for police officers also starts out higher but then grows more slowly with years of service than do regular government pensions. Parliamentarians also have their own pensions. To qualify, they must be elected to two terms in office, at which point they receive 100 percent of their salary for life upon their exit from parliament.

More information on the work of the commission can be found at

Two of the chapter authors, Shamsuddin Tareq and Alejandro Simone, met with the chairman in June 2010.

See Monroe (2009) for a discussion of the impact of emigration on population dependency ratios.

Antigua and Barbuda’s pension fund, unlike the rest, has been experiencing cash deficits since 2007, and limited cash reserves are available to pay pensions in spite of a still relatively favorable dependency ratio of more than five contributors per old-age pensioner. For several decades, the government made virtually no contributions for government workers and paid only a limited amount of interest on government debt held by the Antigua and Barbuda Social Security Board (ABSSB), which constitutes a large portion of the fund’s portfolio. In addition, during the period up to March 2004, the government made repeated demands on the ABSSB to provide funding for various budget-related purposes. In 2009, the ABSSB’s projected cash deficit was 0.7 percent of GDP, which was financed out of its limited liquid reserves, which stood at EC$35 million at the end of 2009. Arrears to the ABSSB amounted to approximately US$500 million at the time of the IMF mission visit in February 2010.

PAYGO rates are the contribution rates needed each period to equate benefit payments with contribution revenues.

For references on OECD retirement ages and other parameters, see IMF (2010).

According to the actuarial review, compensation of pension fund personnel in Saint Lucia grew by 7.5 percent per year in real terms between fiscal years 2003/04 and 2007/08. In St. Kitts and Nevis, the real growth of staff costs was 6.6 percent between 2006 and 2009. In addition, the pension fund in St. Kitts and Nevis carries out other activities not normally carried out by social security administrations, which leads to extra administrative costs. These activities include the collection of a levy on salaries that is transferred to the government as revenue and the provision of insurance to private sector firms for severance payments.

The internal rates of return are the rates of return that equalize the present value of a person’s contributions and benefits. The real internal rate of return could be thought of as the real interest rate obtained from investing an amount equal to the contributions during a certain number of working years in the pension scheme.

In St. Kitts and Nevis, difficulties in adjusting the ceiling on insurable earnings arise because the ceiling is also used as a threshold above which a levy on salaries is collected in behalf of the government. Thus, increasing the ceiling on pensionable earnings reduces revenue collections from the levy for the budget.

For a more detailed discussion of this point, see Roache and Rasmussen (2007).

See World Bank (2010) for a more detailed discussion.

A simple example illustrates this result. Assume that all countries have the same scheme, under which the vesting period is 10 years, the accrual rate is 4 percent during the first 10 years and 1 percent thereafter, and the pensioner has a 30-year working career. If a worker worked in only one country, he or she would get a replacement rate of 60 percent of salary [(10 × 4) + (20 × 1)]. However, by working 10 years in three different countries, the replacement rate would double to 120 percent [(4 × 10) + (4 × 10) + (4 × 10)].

For example, in both Antigua and Barbuda and St. Kitts and Nevis, civil service employees can draw several pensions, in addition to pensions from the national scheme. In such cases, the combined pension benefits could exceed 130 percent of preretirement salary.

These reforms emerge from the formula below, which is derived by assuming that the PAYGO pension scheme is financially self-sustaining, that is, current contributions are meant to cover current pension payments and administrative costs in every period:


where C is the contribution rate, R is the replacement rate (average pension/average wage), D is the dependency ratio (number of pensioners/number of contributors), and A represents administrative costs (administrative cost/divided by the covered wage bill).

The maximum replacement rate should be chosen to keep costs in line with contributions, given the other parameters in the PAYGO system.

In particular, the guidelines should seek to minimize investment in social projects and government or government-related entities performing government functions (such as statutory bodies).

Currently, the minimum requirements for employment as a teacher are four passes at the Ordinary level of the General Certificate of Education (GCE) or Caribbean Examination Council exams.

Trained teachers are those who have successfully completed a program in teacher education methods and teaching techniques (usually a two-year program).

For example, in Antigua and Barbuda, discussions with public officials suggest that exemptions from bonding conditions are often granted.

Cost recovery in many public tertiary education systems, such as those in Chile, Colombia, the Republic of Korea, and the United States, is in the range of 20 to 30 percent. See de Wit and others (2005).

For example, in St. Kitts and Nevis, at least 65 percent of the well-off, or half of the student population, receive free meals. A similar pattern is observed in the free textbook program. In St. Vincent and the Grenadines, larger numbers of students from better-off households benefit from free meals than do those from less well-off households.

In some countries (e.g., Dominica), some schools were closed because of the very low number of students.

In Antigua and Barbuda, formal sector workers are required to contribute 7 percent of their salaries to the medical benefits scheme, which covers treatment of nine diseases. Expenses covered include inpatient and outpatient care, medicines, and some overseas care not available in the country (up to a threshold).

World Health Organization, Global Health Observatory Data Repository, 2009,

For example, in Saint Lucia, external grant financing accounted for 72 percent of all budgeted capital expenditures, on average, during 2007/08–2009/10.

In Saint Lucia, individuals in the poorest quintile of the income distribution more often use hospital facilities than community health centers for their care, according to the 2005/06 household survey. The use of hospital facilities is even higher in the wealthiest quintile, which almost exclusively uses private doctors and hospitals. The 2007/08 poverty assessment for St. Kitts and Nevis shows qualitatively similar results. In Antigua and Barbuda, more than 70 percent of public health spending occurs at the hospital level, suggesting that a significant portion of primary care is delivered by hospitals.

For example, in Saint Lucia and Grenada, user fees in the hospital legislation have not been adjusted for more than 25 years. In St. Kitts and Nevis, adjustments were more recent but the fees are still low. In Dominica, the Cabinet recently approved a revised fee schedule that incorporates fees for new services, which are expected to reflect better the cost of providing health services.

In Dominica, only about 25 percent of such visits are classified as genuine emergencies (Dominica, Ministry of Health, 2010).

Fees for drugs tend to be fixed irrespective of the number of different medicines per prescription and do not depend on the cost of drugs. In Antigua and Barbuda, three different systems are used to procure pharmaceuticals, which reduces the possibility of realizing economies of scale. The government procures medicines through the OECS Pharmaceutical Procurement Service. The Medical Benefits Scheme purchases pharmaceuticals through the Antigua and Barbuda Tender Board, and Mount St. John’s Medical Center has its own procurement system.

Information on the number of nonestablished workers was not available for computing average compensation. However, similar trends would most likely be observed. In Antigua and Barbuda, allowances ranged between 23 percent and 38 percent of established workers’ salaries during 2008–10.

In Saint Lucia, exempt groups include patients with diabetes and hypertension, contributors to the national pension scheme, several categories of government workers, and individuals below a specified income threshold. In St. Kitts and Nevis, children under age 16, adults over age 62, and all persons suffering from chronic diseases are exempt from user fees. In Antigua and Barbuda, those over age 60 or under age 16 as well as individuals diagnosed with one of the covered diseases are exempt irrespective of whether they contribute to the Medical Benefits Scheme.

Occupancy rates for hospitals range from 60 to 90 percent internationally. These rates have typically been below 60 percent at Victoria Hospital and less than 50 percent for St. Jude Hospital in Saint Lucia in recent years. Similarly, in St. Kitts and Nevis, occupancy rates at the Joseph France Hospital where between 40 and 45 percent in 2006–09. In Dominica and Grenada, the problem is most acute at the level of district hospitals. Occupancy rates at district hospitals in Dominica reportedly range between 20 and 30 percent. World Bank (2005b) reported occupancy rates for the Princess Royal Hospital in Grenada to be between 12 and 15 percent.

Antigua and Barbuda is the exception because it earmarks and collects payroll contributions for drugs and medical supplies while salaries of health care workers are paid out of general revenues.

General hospital authorities in Grenada have described a variety of situations in which treatment protocols are not followed, resulting in an excessive number of medications, lab tests, and x-rays.

For example, in Saint Lucia, decisions made to include in the benefits package certain health services or drugs without full costing are frequent and have led to repeated cost overruns at Victoria Hospital during 2007/08, 2008/09, and 2009/10. One example was the decision to include drugs for diabetes and hypertension in the free benefits package because of increased demand. Some arrears to suppliers resulted and were carried over for several years. These arrears were subsequently cleared.

For example, the availability of free diabetes and hypertension drugs has induced patients with private insurance, who could afford to pay for treatment, to shift their demand for treatment of these diseases to the public sector.

This section discusses publicly financed, noncontributory transfers and subsidies targeted in some fashion to the poor and vulnerable. It draws on social safety net assessments for Antigua and Barbuda, Grenada, St. Kitts and Nevis, Saint Lucia, and St. Vincent and the Grenadines, prepared by Lorraine Blank with support from UNICEF, the World Bank, and the UN Development Fund for Women

In Antigua and Barbuda, PEs financed a significant proportion of social assistance spending in 2009.

Comparison of spending figures across countries must be done with caution in light of different definitions of social assistance and the coverage of such programs.

The food voucher program reaches only 4 percent of the poor in St. Kitts. About one-third of book-loan participants come from the poorest quintile, while 21 percent come from the wealthiest quintile. A larger share of students from the top income quintile (24 percent) than from the poorest quintile (17 percent) benefit from the free textbook facility. The school feeding program is universal and 43 percent of students from the top income quintile receive free meals.

In Grenada, eligibility for the main social assistance program is determined through a set of subjective and difficult-to-verify criteria that exclude the working poor. Beneficiaries tend to stay on the public assistance list for life, regardless of improvements in their social conditions, because reviews are not conducted systematically. This limits the opportunities for new poor to enter the list and for expanding the amount of assistance per capita for indigent people. The majority of other programs are based on categorical targeting and cover very few beneficiaries.

To facilitate this work, several countries in the region have undertaken social safety net assessments. The database of programs collected in this context could be completed with detailed administrative data from all programs.

Appendix 7B presents an overview of the variety of targeting methods that can be used to target the poor cost-effectively.

This section draws on various public expenditure reviews undertaken by the World Bank and on Public Expenditure and Financial Accountability assessments or draft assessments shared by country authorities.

These figures do not include (1) capital spending by PEs; (2) capital spending financed under special arrangements (e.g., public-private partnerships such as build, operate, and transfer); and (3) capital spending recorded in nonstandard ways. Capital spending is thus understated by these amounts. Conversely, these figures do not correct for the fact that all foreign-financed capital projects are automatically included as capital expenditure in most, if not all, countries in the region—even though these include recurrent expenditures. This practice results in an overestimation of capital spending.

For example, “below the line” recording of capital investment projects is reported for St. Kitts and Nevis.

“Dual budgeting” refers to the practice of budgeting recurrent and capital expenditures separately. This practice has led typically to underestimation of the recurrent cost implications of capital projects, because recurrent costs are normally not taken into account when developing PSIPs.

In particular, detailed manuals covering basic project appraisal techniques and staff trained to implement the techniques effectively are absent in most, if not all, OECS/ECCU countries. The situation is different for donor-financed projects, which typically require some form of project appraisal.

World Bank public expenditure reviews mention several examples, including national stadiums. National stadiums have often turned out to be more expensive than anticipated and have not generated the projected revenues.

Some projects are also included in the budget in the hopes that donors might fund them during the fiscal year.

Consistent with this, Public Expenditure and Financial Accountability reviews note the absence of a procurement appeals system or any case of procurement complaints being heard by the courts.

St. Kitts and Nevis has commitment controls for debt-service and payroll obligations.

Information available in the World Bank’s Public-Private Infrastructure Advisory Facility database suggests that since 2000, no country in the region has implemented more than one PPP operation.

The more advanced medium-term budgeting becomes, the more it will be possible to check compliance with envelopes in future budgets.

The discussion in this section excludes pension funds because their operations are dealt with in detail in a previous section.

These figures are not strictly comparable to those for Antigua and Barbuda because information on nonestablished government workers in Dominica and St. Vincent and the Grenadines was not available.

Information relating to the budgets of PEs makes up most of what is available. Detailed execution information for previous years on revenues, expenditures, and reconciliation with financing is absent for most PEs. Other issues with the information include nonstandardized reporting across institutions with unclear statistical definitions and standards; and concerns about quality, given the absence of audited financial statements.

This appendix is adapted from McRoberts (2009 and 2010). It has benefited from comments from Michel Marion of the Caribbean Regional Technical Assistance Center.

    Other Resources Citing This Publication