Information about Western Hemisphere Hemisferio Occidental
Journal Issue
Share
Article

Eastern Caribbean Currency Union: Selected Issues

Author(s):
International Monetary Fund
Published Date:
March 2003
Share
  • ShareShare
Information about Western Hemisphere Hemisferio Occidental
Show Summary Details

VII. Social Security Issues in The Caribbean73

A. Introduction

156. Population aging is a global phenomenon. A confluence of increasing life spans and declining fertility rates is expected to swell elderly dependency ratios worldwide. Most developed countries have already started to experience the first signs of aging, while the developing countries are still enjoying the benefits of a relatively young population and workforce. However, in a few decades from now, the situation of the developing nations will dramatically change, due to a much more rapid aging process than in the developed world. The countries of the Caribbean are no exception. As shown in Figure 1, the elderly dependency ratio in the countries of the Eastern Caribbean Currency Union (ECCU),74 defined as the number of individuals aged 65 and older divided by the total number of working-age individuals (16-64 years old), is projected to quadruple from about 0.09 to about 0.36 in only thirty five years starting in 2015.75 In contrast, the same ratio for the countries of Western Europe, Japan, and North America will have only doubled over a period of fifty years between 2000-2050 (when projections end).

Figure 1.Elderly Dependency Ratios in the ECCU, and Western Europe, U.S. and Japan

Source: U.S. Census Bureau, International Data Base

157. Increasing dependency rates are expected to put pressures on national saving rates, government budgets, and public pension systems, as fewer workers will have to support a larger number of retirees. For the ECCU countries, the projected rapid aging calls for immediate action, although there are still almost two decades before the so-called age-quake starts to unfold. A quick policy response would avoid disruptively large last-minute policy changes that would be needed to shore up the public social security systems in the ECCU. Despite having accumulated significant surpluses, unless they are reformed, the systems will quickly run out of resources soon after the currently large labor force reaches retirement. Future prospects look even bleaker when one takes into account a fiscal situation that in some countries is already unsustainable or is heading in that direction.

158. The purposes of this chapter are to shed light on the challenges facing the social security systems of the ECCU countries and to recommend policy actions that may help to alleviate the aging crisis projected to start around 2020. The main theses of the chapter are: (1) the projected population dynamics in the ECCU are expected to render the current social security arrangements insolvent and to exacerbate their already critical fiscal position; (2) besides its direct negative effect on pensions and government debt, aging will have spillover effects into the financial sector, due to the close link between public pensions, the government, and commercial banks; and (3) immediate action is required to prevent the potentially disastrous consequences of the aging problem, including: structural reform of the national insurance schemes, a revamping of the investment strategy of the pension funds, a policy of continuing accumulation of reserves, and restructuring government pensions.

159. This chapter is structured as follows: Section B describes the social security systems in the ECCU, delineating the common elements and main differences across countries. The main issues and problems with the current arrangements stemming from population aging are analyzed in Section C. In Section D, we outline a few policy recommendations. Section E concludes the discussion.

B. Characteristics of Public Social Security Systems in the ECCU

160. The public social security systems in the ECCU are composed of a National Insurance System (henceforth named the NIS) and a system of government pensions.76 The NIS is the major component of public social security and is structured as a partially funded, defined-benefit system financed by payroll taxes. In the majority of ECCU countries, public sector employees also receive government pensions in addition to NIS benefits. Government pensions, however, are unfunded and non-contributory and constitute a contingent liability for the government.

161. Given that the ECCU countries are bound to face a severe demographic challenge in the next decades, a careful analysis of the structure of their national insurance systems is warranted. Identification of structural risks and weaknesses early enough could permit these countries to find adequate solutions and adopt them in a more gradual and hence less socially disruptive manner. The rest of this section will describe, in some detail, the structure of pension systems in the ECCU. Based on this discussion, the next sections will isolate the main problems with the systems and will propose solutions for reform.

C. The Basic Structure of the NIS

162. The National Insurance Systems (NIS) in the ECCU were initially set up in the mid 1960’s, following the British social insurance model. As such, they share many common elements. All started up as Provident Funds, which were later expanded into public social insurance systems. At present, they operate as partially funded defined-benefit plans financed by payroll taxes. This arrangement implies that the social security administration collects payroll taxes from employees and employers and uses the proceeds to provide a given level of benefits to those covered under the system. Since the ECCU economies are predominantly young, the taxes collected exceed the current claims on benefits. The remaining contributions are saved in the systems’ investment funds.77 These funds constitute a cushion against possible shocks to the systems, but as this chapter argues, they will prove insufficient to guarantee the solvency of the NIS when the population starts to age.

163. The NIS is financed by payroll taxes paid by both employees and employers. Total contributions to the systems range from 6.1 percent in St. Vincent and the Grenadines to 11 percent in St. Kitts and Nevis, with the majority of contributions being around 8-11 percent, as Table 1 shows. Furthermore, all countries have ceilings on covered earnings. These ceilings facilitate intra-generational redistribution from the higher to lower income groups, as the benefits of low earners represent a higher fraction of total earnings compared to the benefits of the high earners. Essentially, the lower the ceiling, the higher the degree of redistribution under the system78. As shown in Table 1, there are significant differences among the NIS of the ECCU, with countries such as St. Vincent and the Grenadines, St. Lucia and Grenada being relatively more progressive compared to Dominica and St. Kitts and Nevis.79

Table 1.NIS Contribution Rates in Percentage of Earnings/Payroll
CountryEmployee’s contribution and max./min. taxable earningsEmployer’s contributionTotalSelf Employed
Long-term benefitsShort-term benefitsLong-term benefitsShort-term benefits
In percent
Anguillan.a.n.a.n.a.n.a.n.a.n.a.
Antigua and Barbuda3.0

2.0 public sector

Max. EC$ 54,000/year
0.05.00.08.08.0
Dominica3.0

Max. EC$ 60,000/year
0.07.00.010.07.0
Grenada4.0

Max. EC$ 36,000/year
0.05.00.09.06.8
Montserratn.a.n.a.n.a.n.a.n.a.n.a.
St. Kitts and Nevis5.0

Max. EC$ 6500/month
0.05.01 work

injury
11.010.0
St. Lucia5.0

Max. EC$ 36,000/year
0.05.00.010.0n.a.
St. Vincent and the Grenadines2.5

Max. EC$ 2600/month
0.03.50.16.1n.a.
Source: U.S. Social Security Administration; Social Security Systems Throughout the World, 1999.
Source: U.S. Social Security Administration; Social Security Systems Throughout the World, 1999.

164. Social insurance taxes are used to provide benefits, which are divided into three broad categories: long-term benefits, including old-age pensions, disability and survivorship benefits, short-term benefits, such as sickness and maternity, and work injury benefits. The benefit structure of the NIS, summarized in Tables 2 and 3, is widely similar across the ECCU. Two important common elements, low legal retirement ages and low required contributions, facilitate the redistribution of resources inter-generationally80. Despite the uniformity across countries with respect to benefit levels, there is large variation in the maximum old-age benefit-to-contribution rate, which measures the relative generosity of pension systems. According to this indicator, the NIS systems in Grenada and St. Kitts and Nevis are less generous than those in the other ECCU countries, with the systems in St. Vincent and the Grenadines and Dominica emerging as the most generous (Figure 2).

Table 2.NIS Long Term Benefits
CountryMin. AgeQualifying ContributionsMinimum/Maximum Old-Age PensionSurvivorship BenefitsDisability Benefits
Anguilla*n.a.500 credited (250 paid)

disability: 50 paid
EC $120/wk - 60 percent of earningsEC $27/wk - 60 percent of pensionEC $105-30 percent
Antigua and Barbuda60500 paid/credited

disability: 156

paid/credited old-age

grant: 26 paid/credited
25 percent - 50 percent of earnings Old age grant

Old -age assistance
50 percent of pension to widow 20-40 percent of pension to children

Funeral grant: EC$2000
same as old-age pension
Dominica60500 credited (150 paid)

disability: 150

paid/credited
30 percent - 70 percent of earnings Old age grant

Old -age assistance
50 percent of pension to widow 25-30 percent of pension to children

Funeral Grant: EC$ 1800
same as old-age pension
Grenada60500 credited(150 paid)

disability: 50

paid/credited old-age

grant: 150 paid/credited
30 percent of earnings

min.: EC$40/wk Old age grant
75 percent of pension to widow 25-50 percent of pension to children

Funeral Grant: EC$ 2000, 1500 (spouse), 750 (child)
same as old-age pension
Montserrat*n.a.150 paid disability: 150 paid/creditedEC$173/mt-60 percent of earnings50 percent of pension to widow

30 percent of pension to children
same as old-age pension
St. Kitts and Nevis62500 credited (150 paid)

disability: 150

paid/credited
30-60 percent of earnings min.

EC$200/month Old age grant
50 percent of pension to widow 15-30 percent of pension to children

Funeral Grant: EC$ 2500, 2500 (spouse), 1600 (child)
same as old-age pension
St. Lucia6010 years of contributions

disability: 5 years of contributions
40 percent of earnings, no max. Old age grant old age grant75 percent of pension to widow 25-50 percent of pension to children

Funeral Grant: EC$ 1500
same as old-age pension
St. Vincent and the Grenadines60500 credited(150 paid)

disability: 150

paid/credited old-age

grant: 50 paid/credited
30-60 percent of earnings min. EC$

40/week Old age grant
75 percent of pension to widow 25-50 percent of pension to children

Funeral Grant: EC$ 2000
same as old-age pension
Source: U.S. Social Security Administration: Social Security Systems Throughout the World, 1999.
Source: U.S. Social Security Administration: Social Security Systems Throughout the World, 1999.
Table 3.NIS Short Term and Employment Injury Benefits
CountryQualifying contributionsSickness benefits (amount, duration)Maternity benefits (amount, duration)Employment injury benefits (EI)
Anguilla*sickness: 26 wks

maternity: 26 wks
60 percent of avg. wage

26 weeks
60 percent of avg. wage

13 weeks

grant: EC$250
n.a.
Antigua and Barbudasickness: 26 wks

maternity: 26 wks
60 percent of avg. wage60 percent of avg. wage

13 weeks

grant: EC$400
none
Dominicasickness: 13 wks

maternity: 30 wks
60 percent of avg. wage

26 weeks (max. 52)
60 percent of avg. wage

12 weeks

grant: EC$500
60 percent of avg. wage

medical benefits
Grenadasickness: 13 wks

maternity: 30 wks

EI: max. wages: EC$36000/yr
65 percent of avg. wage

26 weeks (max. 52)
65 percent of avg. wage

12 weeks

min. EC$450
70 percent of avg. wages

26 weeks

medical benefits
Montserrat*sickness: 26 wks

maternity: 30 wks
60 percent of avg. wage

26 weeks
60 percent of avg. wage

12 weeks

grant: EC$400
n.a.
St. Kitts and Nevissickness: 26 wks

maternity: 39 wks

EI: mas. wages: ECS

6500/mth
65 percent of avg. wage

26 weeks
65 percent of avg. wage

13 weeks

grant: EC$450/child
75 percent of avg. wages

26 weeks

medical benefits
St. Luciasickness: 6 mts

maternity: 7 mts

EI: max. wages: EC$36000/yr
65 percent of avg. wage

26 weeks (max. 52)
65 percent of avg. wage

3 months

grant: EC$450
EC $ 30 for medical costs
St. Vincent and the Grenadinessickness: 26 wks

maternity: 30 wks
65 percent of avg. wage

13-26 weeks
65 percent of avg. wage

13 weeks

grant: EC$300
70 percent of avg. wages

26 weeks

medical benefits
Source: U.S. Social Security Administration: Social Security Systems Throughout the World, 1999.
Source: U.S. Social Security Administration: Social Security Systems Throughout the World, 1999.

Figure 2.Maximum Old-Age Benefit Rates Relative to Individual Contribution Rates

Source: ECCB

165. Rather than operating on a strict pay-as-you-go basis, the NIS is partially capitalized. This arrangement allows the systems to save part of the resources that are not used to pay out benefits. Table 4 summarizes the historical evolution of the revenues, expenditures, balances and reserves of the NIS from 1980 to 2001. As shown, NIS balances as percentage of GDP have been roughly stable over time at around 2 percent of GDP, with two notable exceptions. The first exception is Dominica, where despite relatively similar increases in both contributions and benefit expenditures, NIS balances doubled in recent years. They now stand at above 4 percent of GDP, mainly due to increases in interest and other revenues, which more than compensated for the increase in administrative expenses. The second exception is St. Kitts and Nevis, where higher than average contributions relative to GDP led to higher balances, in the range of 4 to 6 percent of GDP.

Table 4.NIS Receipts, Payments, and Reserves in Percent of GDP
Country198019851990199520002001
Antigua and BarbudaNIS Balances2.381.912.070.481.81n.a.
Total Revenues3.203.033.292.373.86n.a.
Total Expenses0.821.121.221.811.99n.a.
Total Reserves12.7916.0016.2622.4925.82n.a.
DominicaNIS Balances2.332.912.412.104.054.33
Total Revenue3.134.314.484.9810.8611.17
Total Expenses0.681.011.351.548.308.47
Total Reserves12.2017.6820.9424.9629.0530.16
GrenadaNIS Balancesn.a.0.091.782.22n.a.2.57
Total Revenuen.a.0.112.413.34n.a.4.22
Total Expensesn.a.0.030.631.12n.a.1.65
Total Reservesn.a.6.0111.5919.80n.a.24.84
St. Kitts and NevisNIS Balances4.545.124.886.506.106.02
Total Revenue5.777.705.106.707.277.79
Total Expenses0.721.281.651.892.862.58
Total Reserves23.7442.1731.6041.4746.5848.63
St. LuciaNIS Balances2.041.711.752.443.03n.a.
Total Revenue2.322.242.373.694.72n.a.
Total Expenses0.280.360.530.991.53n.a.
Total Reserves8.3413.2415.2321.8629.71n.a.
St. Vincent and the GrenadinesNIS Balances2.301.601.712.572.00n.a.
Total Revenue2.712.052.313.133.23n.a.
Total Expenses0.410.460.600.561.08n.a.
Total Reserves12.8913.7415.7719.7923.92n.a.
Source: ECCB and WEO
Source: ECCB and WEO

166. Since the NIS systems have consistently run positive balances over time, they have accumulated substantial investment funds. As illustrated in Table 4, reserves as a percentage of GDP have doubled from 1980 to 2000 in most countries except Grenada and St Lucia, where they have actually quadrupled. In recent years, reserves have stood at around 20 to 30 percent of GDP in all countries except in St. Kitts and Nevis, where they have been as high as almost 50 percent of GDP. The reserve-expenditure ratio measures how many years of payments the funds could make if no future contributions or interest were received and no additional benefits were awarded. As shown in Figure 4, this ratio is in the range of 10 to 20. St. Vincent and the Grenadines has the highest reserves in the region, which could cover about 22 years of benefit payments with no additional revenues. At the other extreme, Dominica could only pay for 3.5 years if no other income were received.

Figure 3.Administrative Expenses as Percentage of Total Contributions

Source: ECCB

Figure 4.Reserve to Expenditure Ratio

Source: ECCB

167. Despite their centralized structure, the NIS systems in the ECCU entail very high administrative costs, ranging from 0.3 to 0.7 percent of GDP, except in Dominica, where they are at 1.5 percent of GDP. An alternative, and perhaps more useful, measure of administrative expenses is their ratio to total contributions. This ratio, as shown in Figure 3, ranges from about 10 to 20 percent across the ECCU region. These unusually high numbers can be partially justified by the large scale of NIS operations, which include administration of income and other individual taxes not related to social security. However, major inefficiencies in the system exist and need to be corrected.

168. Since capital markets in the ECCU countries are relatively shallow, the investment funds of the NIS are concentrated on a narrow range of investments, including mainly fixed deposits, treasury bills, debentures, and cash. Figure 5 shows the investment portfolios of each country’s NIS for 2001, broken down into their main components. The average share of investments in deposits at banks and other financial institutions is the largest component, at around 42 percent of total assets. It ranges from 15 percent in Dominica to 68 percent in St. Vincent and the Grenadines. The next most prominent average share is comprised of other assets (including other local and non-local assets), at about 32 percent. Average regional foreign investments are at 11.6 percent, but excluding Montserrat, whose share of investments in foreign assets is at 53.3 percent, the average for the region falls to 5.7 percent. As will be discussed later in the chapter, the NIS could benefit from more diversified investment portfolios.

Figure 5.NIS Investment Portfolio (2001)

Source: ECCB

D. Government Pensions

169. In addition to NIS benefits, public employees in many ECCU countries also receive government pensions.81 These non-contributory, defined-benefit pension schemes are restricted to established government workers, who are entitled to pension at retirement or after a specified number of years of service. At retirement (usually 55 but with an option to take early retirement at 50), government employees can opt for a gratuity (lump sum) based on years of service, or a reduced gratuity along with a pension for the rest of their lives. Most employees opt for the reduced gratuity for which different combinations of gratuity and pensions can be chosen.

170. Pension payments are a small but growing part of government expenditure, but they have the potential to mushroom as the population ages. Pension payments are paid out of the government-consolidated fund on a pay-as-go basis, since they are non-contributory. In 2001, the total pension payments by the ECCU countries amounted to around 2 percent of GDP. Some governments have tried to modify the pension arrangements in order to reduce their impact on general resources. The Government of Grenada has integrated the government pensions into the National Insurance Scheme by mandating that all workers who joined the service after 1983 would only be covered by the NIS, while existing workers at that time would continue under the old pension arrangements. A similar attempt was made by St. Vincent and the Grenadines in 1996, but the government reverted to the old arrangements in 2000. Government non-contributory pensions in Dominica only cover the period between the government retirement age (55) and the age at which they begin to receive social security pensions (60).

E. Issues with the Current Public Social Security Arrangements in the ECCU

171. The current public social security arrangements in the ECCU face significant challenges stemming from population aging. Although demographic trends are similar among the ECCU countries, the levels and rates of aging vary across the region. As Table 5 shows, by 2050, elderly dependency ratios are projected to range from 26.9 percent in St. Kitts and Nevis, to 56.5 percent in St. Vincent and the Grenadines. The most rapid population aging is projected to occur in Antigua and Barbuda and St. Vincent and the Grenadines, which will experience a six-fold increase in their elderly dependency ratios from around 2010 to their peak in 2050. At the other extreme, St. Kitts and Nevis is expecting a more ‘modest’ increase of about 2.5 fold in its dependency ratio during 2015-2050.

Table 5.ECCU Dependency Ratios
200020102020203020402050
OECS10.058.7110.6118.2427.9635.79
Anguilla10.5910.1514.7625.0435.6442.14
Antigua and Barbuda7.564.979.2324.4234.0432.29
Dominica12.4011.2911.8619.3932.5535.22
Montserrat17.9215.3613.3118.1228.1042.78
St. Kitts and Nevis14.6811.3013.4220.4523.5126.89
St. Lucia8.597.618.6614.0622.7831.05
St. Vincent and the Grenadines10.049.5112.1419.9133.1056.46
Source: U.S. Census Bureau, International Data Base
Source: U.S. Census Bureau, International Data Base

172. Unless they are reformed, the NIS will be threatened by bankruptcy, as pension liabilities will exceed contributions. Increases of 200 to 600 percent in the number of elderly individuals relative to the working–age population will swell expenditures relative to the revenues of public social security systems. Although the schemes have accumulated substantial investment funds, these will prove insufficient when the currently large labor force starts to draw the promised pension benefits. The countries with larger current reserves or expecting a more gradual increase in their elderly dependency ratios are at a slight advantage to the rest, in that they have more time until their National Insurance Systems will become bankrupt.

173. According to official projections, under current arrangements, the NIS of both St. Kitts and Nevis and St. Vincent and the Grenadines are expected to be able to meet expenditure requirements using contribution and interest revenues until about 2020. From 2020 onwards, they will need to draw on their accumulated reserves to pay off pension benefits that come due. In approximately 10 years, reserves are projected to be exhausted, and the long-term branches of the NIS will be bankrupt. Since the LTB concentrates more than 70 percent of the resources of the NIS, bankruptcy of the LTB essentially implies the likely bankruptcy of the whole system.82

174. It is interesting to note that while the end outcome is projected to be roughly the same in St. Kitts and Nevis and in St. Vincent and the Grenadines, the reasons behind it are quite different. On one hand, St. Kitts and Nevis is able to stay afloat for another three decades due to the more favorable demographics it will experience. St. Vincent, on the other hand, will manage to function under the current pension arrangements for another thirty years due to the returns generated by its high level of accumulated reserves.

175. In contrast to the two cases described above, a much more worrisome situation is projected for Dominica. The reserve to expenditure ratio of the Dominican NIS is a meager 3.5, currently the lowest in the region. By 2011, its LTB is expected to have to dip into reserves to finance benefits. Consequently, reserves will be quickly exhausted by 2023. A lower growth rate of the economy and non-recovery of contributions in arrears are shown to lead to an even faster pace of reserve depletion and an earlier bankruptcy of the system.

176. Actuarial data is not available for the rest of the ECCU countries. However, similar to the three case studies presented above, the NIS of these countries are likely be bankrupt around 2020-2030.83 Unless their national insurance systems are reformed soon, drastic measures will need to be taken to bail out the systems later on, as will be explained in the next section.

177. The portfolios of the ECCU NIS schemes have not generated a large enough return to ensure the viability of the schemes (sometimes with long periods of negative real returns), as a result of insufficient diversification and ineffective portfolio management. Diversifying into foreign assets and strengthening their investment profile can improve their performance.

178. The ECCU NIS have a total portfolio of EC$ 1.5 billion, or approximately 20 percent of GDP. Such a significant amount of resources needs to be managed appropriately so as to generate returns to pay pensions when surpluses dry out. However, over the period 1980 to 1998, the NIS achieved average real yields of around 1 percent, which are substantially lower than the average real growth rate for the period (5 percent), suggesting significant under-performance (Table 6 from Nicholls 2002).84 St. Lucia achieved average real yields of around 3 percent over the same period, while Antigua and Barbuda experienced negative real yields.85 Furthermore, in all countries except St. Lucia, the variability of the yield was greater than the mean, which could indicate excessive risk in the portfolios.

Table 6.NIS Investment Yields, Inflation, and Economic Growth
Real YieldsInflationEconomic Growth
1980-19981980-19851985-19901990-19951995-19981980-19981980-1998
MeanVariabilityMeanMeanMeanMeanMeanVariabilityMeanVariability
Anguilla0.14.3-0.8-4.23.5363.83.75.6
Antigua and Barbuda-1.72.3-1.8-2.80-1.15.43.15.24.9
Dominica1.82.30.80.723.74.92.34.13.7
Grenada2.23.5-11.63.74.13.62.54.33
Montserrat178.9NA18.715.2NA4.73.42.26.5
St. Kitts and Nevis0.13.6-0.2-11.11.55.73.85.22.9
St. Lucia3.12.622.82.63.54.12.53.14.7
St. Vincent and The Grenadines2.23.20.62.52.73.553.73.84.3

179. Some of the reasons for the low yields on the NIS investment portfolios are: the skewness of the portfolio to deposits in commercial banks and government paper, concentration of the portfolio in domestic assets, and social lending. A shortage of high-grade domestic securities has caused the portfolios of the NIS to be highly skewed towards deposits in commercial banks and domestic government paper. These instruments have generally been remunerated at less than market rates.

180. In addition, among the category other local investments (14 percent in 2001) are land and buildings and other social lending programs, which are not remunerated at market rates. It can be argued that as custodians of such large amounts of resources, the NIS should have some social responsibilities. However, the lower returns on the social activities need to be offset by higher earnings in the other portfolio categories.

181. While there is no formal legislation restricting investment in foreign securities, the unwritten policy is that the savings of the NIS should be invested at home.86Some of the reasons for investing domestically include the purported higher returns in capital scarce islands, the need to develop capital markets at home, exchange rate risk in foreign investment, and lower returns in developed countries due to their age structure. While, theoretically, returns should be higher in capital scarce countries, the vulnerability of the ECCU economies to natural disasters and external shocks implies that in practice, the portfolios would be riskier. Furthermore, the link between pension assets and capital market development is not well established. Although the investments of pension funds helped to develop the stock market in some Latin American countries, Reisen (1997) found a negative relationship between stock market development and pension assets in a survey of OECD and non-OECD countries. Finally, exchange rate risks can be hedged and the effect of population structures on investment yields can be offset by productivity increases.

182. The nexus among the NIS—government—state-owned banks is increasingly fragile. The relationship between the central government and public sector financial institutions masks some of the risks in the financial system and the sustainability of the public sector finances. A large part of social security funds is deposited into government-owned commercial banks, and at the same time, the commercial banks have large exposure to government debt. There are several risks to this arrangement.

183. First, the NIS and the domestically owned commercial banks have excessive mutual exposure to each other, which could result in instability of the financial system. By policy, the NIS keep the majority of their deposits in domestically owned institutions, averaging around 42 percent of their total portfolios. This practice was initially adopted to give the newly created national banks a slight advantage over their well-established foreign competitors, but the domestic banks have not yet been able to wean themselves from this largely captive source of funds. Domestic commercial banks, which depend on the NIS for an average of around 12 percent of their deposits, can be easily destabilized by modest shifts in the portfolio of the NIS. At the same time, the social security schemes have excessive exposure to the credit risk of these usually weak institutions. Thus even though the NIS may want to shift resources to optimize their return and reduce risk, they are held captive by the weakness of the national financial institutions.

184. Second, in some cases, the social security funds are deposited in the commercial banks at below market rates to provide loans to the government, lowering the overall return on their portfolio. In addition, a part of their other local investments are loans to the development banks at below market rates to fund social projects such as student loans and low-income housing schemes. In this way, the NIS are required to subsidize governmental social programs, when in fact, they should command higher rates in the market because of their size. The lower rate of return lessens the long-run viability of the NIS and increases the possibility of a government bailout in the future.

185. Third, the commercial bank lending to the government masks the exposure of the NIS to the government. In many cases the NIS are over-exposed to the government. The ECCU average proportion of the social security portfolio invested in government paper in 2001 was approximately 10 percent but was as high as 30 percent in Grenada and 23 percent in Dominica. An arrangement that practically permits the commercial banks to lend social security deposits to the government indirectly increases even further the exposure of the NIS to the government, as well as raising borrowing costs and lowering rates of return to the NIS.

186. Finally, this system of intra-public sector resource flows could easily spill over onto the payments system if fiscal pressures cause the central government to build up arrears on their liabilities to the weak domestically owned commercial banks.87 In this situation, there is a risk that the failure of one of the commercial banks would put a significant portion of the NIS funds at risk.

187. Government pensions will add to the contingent liabilities of the government once population aging starts to unfold. While the government pension payments are not a huge burden now, they have the potential to grow as the population ages and the size of the government workforce increases, as has been the case in recent years. The promised benefits under these schemes will have to be paid either from the resources of the NIS or from general revenues, which will exacerbate the problems of the NIS or will add to the fiscal deficit and government debt.

Policy Recommendations

188. The rapid pace of population aging projected for the ECCU countries is expected to lead to significant rises in public pension expenditures over the next five decades. Options for pension reform in the next few years need to be explored in order to avoid last-minute large changes of the social security systems, which would be economically disruptive and socially undesirable. With debt to GDP ratios already on an unsustainable path and with high budget deficits, the scope for using general fiscal measures to address the problems of the social security systems is very limited. Hence, reforms of the pension systems must be considered now. This section discusses some guidelines for reforming the social security systems in the ECCU.

189. Reforming the NIS, which is the largest component of pensions, should have top priority. There are two main approaches to reform public pension systems:88 the “parametric approach,” which maintains the pay-as-you-go deflned-benefit structure of the system while modifying its parameters, and the “the systemic approach,” which involves a structural change of the system towards a fully funded, defined-contribution, privatized scheme. Parametric reform is potentially easier to implement than changing the entire structure of the system, but it is not immune to future adverse demographic shocks. Consequently, additional changes could be required if the population aging is more severe than expected, or if other unforeseen shocks render the system insolvent. Systemic reform has the advantage of linking individual contributions to benefits and of not requiring additional changes, but usually entails high administrative costs and very large implicit transition costs89.

190. Parametric reform of the NIS seems more appropriate for the ECCU, since systemic reform would imply large transition costs that would add to the already unsustainable debt-to-GDP ratios in the region. The ECCU governments favor the parametric approach, and have put forth a series of recommendations, including increasing the minimum retirement age, raising contribution rates, increasing the period over which insurable earnings are averaged to calculate benefits, reducing the maximum benefit level, and indexing the parameters of the schemes to price or wage indices. In addition, since the LTB branch of the NIS is bound to be hit hardest by population aging, redirecting part of the funds from other branches to the LTB may help alleviate some of the pressures stemming from aging. While the proposed modifications are likely to improve the long run viability of the NIS, their effectiveness depends on several factors, which need to be considered.

191. First, the proposed changes under the parametric approach may meet with political resistance and could be delayed. The currently large labor force, which constitutes the majority of voters in the ECCU countries, will not easily be persuaded to accept higher rates now and lower benefits in the future. Furthermore, since the countries of the ECCU already face high debt levels and large budget deficits, they may need to raise income or other individual taxes to restore sustainability of their fiscal balances. Having to increase social security rates at the same time is likely to be difficult politically, yet delaying reform may have potentially very serious future social welfare consequences.

192. The NIS in the ECCU need to adopt a gradual approach to raising pension rates with immediate implementation. Delaying pension reform implies either sharp increases in contribution rates and reductions in benefits later on, or significant intervention by the government to bail out the insolvent pension systems at the risk of running even larger budget deficits in the future. According to projections in its actuarial review, St. Kitts and Nevis would need to triple its LTB contribution rate from 8 percent to 25 percent in 2030, when its reserves will be exhausted, and to continue to raise rates to more than 30 percent over time in order to keep its system afloat. Similarly, Dominica would need to double its rate around 2023 from 6 to 12 percent and to continue to raise it to 22 percent by 2050. Hence, the sooner reforms are implemented, the lower will be the required changes in contribution rates.90

193. Second, raising the retirement age may not necessarily be as effective as intended. The usual argument for increasing the retirement age is that it would not only lower pension spending, but it would also potentially increase the contribution base. However, as Bakker et. al. (1996) point out, if participation rates of the elderly into the labor force are low due to the availability and generosity of early retirement or disability,91 then this reform may turn out to be counterproductive, as individuals would simply substitute old-age pensions for early retirement or disability benefits. In addition, increasing the retirement age may crowd out the employment chances of younger persons. This could add to the already significant burden of high unemployment rates in the ECCU countries.

194. Third, the parametric modifications put forth may not be sufficient to ensure the viability of the NIS. As discussed in the previous sections of the chapter, the social security systems of the ECCU suffer from high administrative costs, poor investment performance and exposure to the risks in the domestic financial markets. Changing the parameters of the system does not help solve these important problems. Thus, in addition to parametric reform, the NIS must undertake other congruent measures, such as reducing administrative costs, modifying the investment strategy of the pension funds so as to increase their real return and to reduce their exposure to domestic banks and the government.

195. The investment portfolios of the NIS need to be improved through greater diversification and strengthening of the investment decision-making process, while ensuring that portfolio adjustments are not too abrupt to destabilize the financial sector. Diversification of the portfolios to include more regional and foreign securities could result in higher returns and lower riskiness of the portfolios. It would also reduce some of the inherent financial sector instability generated by the mutual exposure of the financial sector and the NIS. However, restructuring of the portfolios must be undertaken gradually.

196. The optimal level of diversification depends on the social objectives of the institutions, the size of the portfolios, the macroeconomic effects and the impact on the domestic financial system. Reisen (1997) suggests that pension funds optimally hold world portfolios with domestic assets proportional to their country’s financial market capitalization as a percentage of global capitalization. For the ECCU, this would imply that the bulk of the resources should be invested abroad. However, the opportunity cost of foreign investment is the cost to the economy in terms of forgone real sector development and through lower interest rates. The optimal portfolio needs to balance the benefits of risk diversification against these costs. Hence, the outcome would be greater investment in foreign assets but less than that predicted by the Reisen model.

197. Adjusting investment portfolios too rapidly to reach the optimal level of foreign investment can be dangerous. Fast adjustment of the NIS portfolios could destabilize the banking system given the dependence on the domestically owned banks on NIS deposits. Moreover, sudden portfolio shifts could result in a rapid depletion in the level of official reserves, even though the total foreign assets of the region remain the same. A gradual approach to the adjustment of the NIS portfolios could minimize the risks to the financial system while steps are taken to strengthen the position of the domestically owned commercial banks.

198. Strengthening the investment decision-making process could also enhance the returns of the NIS. Nicholls (2002) notes that while the NIS legislations require the establishment of investment committees, they do not appear to function effectively. He recommends strengthening these committees by broadening the membership to include persons with investment and negotiating skills or the function could be contracted out to investment professionals.

199. Government pensions need to be either funded or eliminated before the population aging crisis hits. This can be accomplished by governments making specific funding arrangements or incorporating them into NIS. Several attempts that have been made to reform government pensions failed because they did not adequately deal with the accumulated pension rights. Thus, attempts at reform would need to manage these obligations in a fair manner.

200. One solution would be to fully fund the government pension scheme and separate it from the consolidated fund. This would ensure that the rising obligations do not put pressure on government financing. The major difficulty with this approach would be to find a way to provide the resources needed to fund the accumulated rights given the fragile budgetary position of most governments of the ECCU. One possibility would be to make specific budgetary allocations on an annual basis to achieve actuarial viability of the fund by a specific time. Another possibility would be to move the government pension funds to private sector organizations and finance the transition with a combination of cash and government bonds. In this way, the contingent liabilities would be recognized explicitly as government debt, but the transitional costs would add to the already unsustainably high debt burden in the ECCU.

201. A second solution, favored by most governments, is to integrate the government pensions into the National Insurance Schemes. This is fairly straightforward when government workers are not covered by the NIS, as in the case of St. Lucia. The government would provide the resources to cover the pension rights as under the privatization option (discussed above) and workers will receive the NIS benefits. In cases where NIS benefits are inferior to the expected government pensions, the difference could be made up in cash payments or government securities. Where the government pension benefits are additional to the NIS, privatization may be the better option.

F. Conclusions

202. The ECCU faces a major population-aging challenge. A quadrupling of elderly dependency ratios is expected to swell pension expenditures relative to contributions.

203. Although still a few decades away, this demographic prospects warrants attention and action now, before potentially devastating fiscal consequences become too difficult to manage. Early identification of risks and weaknesses associated with the social security arrangements in the ECCU could allow the governments to find adequate solutions and adopt them in a more gradual and hence less disruptive manner. The present chapter describes the social security system in place in the ECCU, flags the main issues stemming from population aging, and makes several policy recommendations.

204. Due to their pay-as-you-go structure, the social security systems in the ECCU show increasing signs of vulnerability to population aging. We have identified several salient issues with the social insurance arrangements in the region that will need to be addressed before the aging challenge. First, unless they are reformed, the National Insurance Schemes of the ECCU will be threatened by bankruptcy as their liabilities will exceed their assets. Although the schemes have accumulated substantial investment funds, these will prove insufficient when the current large labor force starts to draw the promised pension benefits. Second, the nexus among the NIS—government—state-owned banks is an increasing source of financial sector vulnerability, with the government borrowing from the NIS, and the NIS depositing its funds in weak domestic banks. Third, the current investment portfolio of the NIS is risky and insufficiently diversified. Fourth, government pensions will add to the contingent liabilities of the government.

205. Options for pension reform in the next few years need to be explored in order to avoid last-minute large changes of the social security systems, which would be both economically and socially undesirable. With debt to GDP ratios already on an unsustainable path in several of the ECCU countries, the scope for using general fiscal measures to address the problems of the social security systems is very limited. Hence, reforms of the pension systems must be considered. This chapter proposes some guidelines for reforming the social security systems in the ECCU, including a gradual approach to parametric reform of the NIS with immediate implementation, diversification of the investment portfolio of the NIS so as to improve rates of return, prolong the viability of the system, and lessen the spillover effects into the financial system, and funding or incorporating government pensions into the NIS.

References

    BakkerBasIoannisHalikiasFrankLakwijk and JoaquimLevy(1996). “Kingdom of the Netherlands, The Netherlands—Selected Issues.IMF Staff Country Report No. 96/80.

    • Search Google Scholar
    • Export Citation

    ChandSheetal and AlbertJaeger(1996). “Aging Populations and Public Pension Schemes.IMF Occasional Paper No. 147.

    DiamondPeter(1993). “Issues in Social Insurance.Nancy L. Schwartz Lecture, Kellog School of Management Northwestern UniversityApril14.

    • Search Google Scholar
    • Export Citation

    DiamondPeter (1998). “The Economics of Social Security Reform.NBER Working Paper No. 6719.

    JamesEstelle(2001). “Coverage under Old Age Security Programs and Protection for the Uninsured: What are the Issues?” in Nora Lustig(ed.)Shielding the Poor Ch. 7. Brookings Institution Press Inter-American Development Bank: Washington DC pp. 149174.

    • Search Google Scholar
    • Export Citation

    Mesa-LagoCarmelo(2001)Social Assistance on pensions and Health Care for the Poor in Latin America and the Caribbean. “in Nora Lustig(ed.)Shielding the Poor Ch. 8. Brookings Institution Press Inter-American Development Bank: Washington D.C. pp. 175215.

    • Search Google Scholar
    • Export Citation

    NichollsG.(2002). “Is there a case for foreign Investment by the Eastern Caribbean Currency Union’s (ECCU) Social Security Schemes (SSS).Savings and Development Vol. 26 No. 3.

    • Search Google Scholar
    • Export Citation

    NichollsG.(1997)Implications for financial stability: Social Security DepositsMoney Affairs Vol. 10 No. 1 pp. 6586.

    ReisenH.(1997). “Liberalizing foreign investments by pension funds: Positive and normative aspects.OECD Development Centre Technical Paper no. 120.

    • Search Google Scholar
    • Export Citation

    SamuelWendell(2002)Economic Impact of the Elderly,unpublished manuscript.

    SamuelWendell(2000)The Harmonization of Social Security in the Caribbean.” inCurrent Issues in Social Security: Pensions Poverty and Health Inter-American Conference on Social Security Studies Series 53eds: Karl Theodore and Althea La Foucade pp. 4573.

    • Search Google Scholar
    • Export Citation

    SamuelWendell(1990)Social Security, Saving and Investment in the ECCU Countries.Bulletin for Fiscal Documentation. pp 220225.

    • Search Google Scholar
    • Export Citation

    ShillerRobert(1998)Social Security and Institutions for Intergenerational, Intragenerational, and International Risk Sharing,NBER Working Paper No. 6641

    • Search Google Scholar
    • Export Citation

    ThomasC(2000)The social protection framework in the OECS countries,Mimeo

73This section is prepared by Wendell Samuel and Delia Velculescu (both WHD) and is part of a larger analytical paper to be completed later in the year.
74These are: Antigua and Barbuda, Dominica, Grenada, St. Kitts and Nevis, St. Lucia and St. Vincent and the Grenadines; and the UK territories of Anguilla and Montserrat. These countries are also members of the Organization of Eastern Caribbean States (OECS) and the Caribbean Community (CARICOM).
75The data is from the U.S. Census Bureau, International Data Base, and can be found at http://jolis.worldbankimflib.org/Pathfinders/Sectors/HNP/hnpstats.htm
76Social safety nets are also available in the ECCU. Thy are mean tested and age-independent, and constitute only a small fraction of total public benefits.
77In contrast, a pure unfunded pay-as-you-go system would simply adjust taxes to match claims on benefits at each point in time. This system is rarely seen in practice, as frequent adjustments in tax rates are socially undesirable.
78Assuming no limit or earnings or contributions.
79However, the systems may be less progressive than this suggests, because high earners tend to live longer and thus collect more benefits, as suggested by a study with U.S. data conducted by Panis and Lillard (1995).
80At the same time, low retirement ages, together with ceilings on old age benefits discourage extended participation in the labor force by elderly individuals.
81Grenada, Dominica, and St. Lucia offer either NIS benefits or government pensions to public employees.
82Varying the assumptions that underlie the projections could worsen the projections presented above, A lower real rate of return and higher administrative expenses would precipitate the bankruptcy of the NIS.
83We are currently working on an analytical model, which would allow us to forecast the evolution of NIS balances and reserves for all ECCU countries. This analysis will be available in the final form of this paper later this year.
84An analysis of the sub-periods shows gradual improvement in performance of most of the NIS with the exception of Antigua and Barbuda. The average yield for the ECCU was close to zero during the period 1980-1985. However, Anguilla, Grenada, and St. Kitts and Nevis moved from around 1 to 3 percent.
85Real yields for Montserrat are much higher, but the data is incomparable due to data unavailability over the entire period.
86Less than 6 percent of the NIS portfolio was invested in foreign securities in 2001, excluding Montserrat.
87The situation could be exacerbated if government arrears to the private sector cause a default on obligations to the commercial banks.
88Chand and Jaeger (1996), Diamond (1988, 1993), etc.
89The costs associated with the transition from a pay-as-you-go system to a fully funded one stem from the need to pay the accumulated benefits of current or future retirees who have contributed to the pay-as-you-go system throughout their careers. These costs could be either paid by the present generations, by increasing their contribution rates to provide both for their own retirement and for the benefits of current retirees, or spread among present and future generations, by issuing recognition bonds (as in the case of Chile) While raising contribution rates may be difficult politically, issuing new bonds may add to the fiscal burden of the government, and may not be advisable for the ECCU governments, given their already alarmingly large debt-to-GDP ratios.
90We plan to generate simulation results for alternative scenarios of parametric reform using our analytical model. This would enable us to determine the optimal timing and magnitude of required tax increases in each country case.
91See Samuel (2002) for a discussion of alternative employment opportunities for the elderly.

Other Resources Citing This Publication