Kingdom of the Netherlands—Netherlands Antilles: Selected Issues and Statistical Appendix

This Selected Issues paper and Statistical Appendix explores four policy issues—fiscal policy, public sector pension reforms, monetary management, and labor market performance—which are crucial for understanding the recent performance of the economy of the Netherlands Antilles and which will need to be addressed to restore the prospect of durable economic growth. The paper reviews experience with fiscal adjustment in the Netherlands Antilles, focusing in particular on the 1996–97 adjustment program. The paper also analyzes the sustainability of the public pension system of the country.

Abstract

This Selected Issues paper and Statistical Appendix explores four policy issues—fiscal policy, public sector pension reforms, monetary management, and labor market performance—which are crucial for understanding the recent performance of the economy of the Netherlands Antilles and which will need to be addressed to restore the prospect of durable economic growth. The paper reviews experience with fiscal adjustment in the Netherlands Antilles, focusing in particular on the 1996–97 adjustment program. The paper also analyzes the sustainability of the public pension system of the country.

III. Sustainability of the Public Sector Pension System14

A. Introduction

57. During the 1990s, the pension scheme for civil servants underwent several changes affecting its degree of funding and sustainability. The system first moved towards a fully funded arrangement, and then, given the financial repercussions of this change, returned to a partially pay-as-you-go (PAYG) system. In spite of a reduction in entitlements since 1996, benefits have remained generous and the system still lacks financial sustainability, implying a strong case for further reforms and fiscal consolidation. Particular problems are the excessive share of claims on the government in the public sector pension fund’s asset portfolio, the accumulation of premium arrears by the government, and the projected rise in the costs of the general old age pension scheme as a result of population aging. The lingering problems illustrate the importance of governance considerations: there is a need for strengthening policy-makers’ accountability for safeguarding pension claims, including through the application of strict prudential standards.

58. The pension system of the Netherlands Antilles consists of a general state-provided old age pension (AOV), a range of private sector pension schemes, and the public sector pension system.

  • All residents of 60 years of age or older are entitled to general old age insurance. This AOV (“Algemene Ouderdoms Verzekering”) scheme is administered by the Social Insurance Bank (SVB) and financed on a PAYG basis. The premium amounts to 10 percent of income up to a ceiling. The monthly AOV pension was NA f. 466 per person in 1998. Benefits are not linked to previous income and, since 1996, also do not depend on marital status, A deduction is applied for those who have not been residents continuously between the ages of 15 and 60.

  • About one-third of the workforce is covered by supplementary pension schemes.15 Private sector pension schemes are administered by firm-specific pension funds, insurance companies, and—in some cases—by the firm itself Generally, these pensions supplement the AOV benefit to 70 percent of the average wage over the final (e.g., ten) years of employment, and also cover disability. Insurance premiums—of which, typically, one third is paid by the employee—are tax deductible; as are pension or life-insurance premiums (up to a ceiling) for nonemployees.

  • Civil servants are covered by the public sector pension scheme. This scheme is administered by APNA.

B. Recent Reforms of the Public Sector Pension Scheme

59. During 1996–98, the pension system for civil servants was adjusted to reduce the government’s premium costs. In the early 1990s there was a major shift from a mixture of a PAYG and a fully funded financing system to an almost completely fully funded system. However, the shift to fully funding, together with a substantial wage increase in 1992, swelled the stock of acquired pension rights, and at end-1994 APNA’s actuarial deficit amounted to NA f. 579 million (14.1 percent of GDP),16 Accordingly, further changes to the pension system were introduced in 1996 and 1998.

60. At the core of the recent reforms was a reduction in the entitlements for new participants. Until 1996, civil servants could retire at age 55 and—if they had worked for at least 30 years—receive a pension equal to 70 percent of their final salary, while also receiving the AOV. Changes were made in phases:17

  • Since the beginning of 1996, employees can only retire at age 60. However, for those with at least five years of service and already employed at end-1995, the early retirement option was maintained.

  • Starting 1998, total retirement income for new employees was reduced to 70 percent of the final wage. Total retirement income is the sum of a civil servant’s pension and AOV benefit. To achieve this, the pension was no longer based on the wage, but on the wage minus a deduction, which is equal to (10/7) times the AOV.18

  • Finally, also since 1998, new employees can only build up a full pension if they have worked for 35 (instead of 30) years. Accordingly, pension rights are not built up as quickly as in the old system; for each year of employment, civil servants are entitled to a pension equal to 2 percent of their final wage.

61. The reforms also included adjustments in the financing arrangement, returning to a partial PAYG system for current retirees. The government’s capital funding premiums were reduced to cover pension rights only up to the reduced level granted to new employees.19 For current participants, the government is to compensate APNA for the difference between the obligations under the old and the new system through PAYG premiums. In addition, pensions of newly retiring civil servants aged between 55 and 60, are financed entirely on a PAYG basis, through the “VUT-uitkering” (early retirement benefit).

62. The 1996 reforms resulted in a sharp decrease in the government’s premium obligations; however, the pay-as-you go premiums were projected to grow rapidly after a few years. The changes can be divided up as follows:

  • The reduction in entitlements (for new participants) and in the degree of funding (for current participants) reduced pension funding obligations, allowing a reduction of the governments’ capital-funding premium (in percent of total wages) from 26 percent to 20 percent in 1996, and then further to 17 percent in 1998.

  • As a result of the higher retirement age, APNA’s 1994 actuarial deficit was largely closed. Accordingly, the regular annual premium payments for covering the 1994 actuarial deficit, were waived, except for relatively small charges in 1997.20

  • Partially offsetting the above, there are additional expenses for the “VUT-uitkering” and higher PAYG premiums.

In sum, these changes lowered the 1996 contributions to NA f. 79 million for the central government and NA f. 91 million for Curaçao, compared to NA f. 105 million and NA f. 129 million, respectively, under the old system (see Table 3).

63. The government has failed to make contributions to a sinking fund which was established in 1997 to provide for the projected rapid increase in VUT and PAYG payments. The amounts the government would need to pay for the VUT and the PAYG premiums were projected to escalate, from NA f. 5 million in 1996, to NA f. 15 million in 1998, and would already amount to NA f. 36 million in the year 2000. To partially provide for these costs, a sinking fund was included in the 1997 budget, as part of the government’s adjustment program. However, as a result of financing difficulties no contributions were made in 1997 and 1998.

C. Evaluating the Current Situation

64. In view of APNA’s recent actuarial surpluses, public sector pension premiums were reduced for 1999 and 2000. An actuarial analysis indicated that at end-1997 the fund’s assets available for covering future pension liabilities exceeded the required provision by NA f. 108 million.21 Likely, a further surplus was recorded in 1998. On this basis, it was decided to waive the final premium payment for covering the 1989 actuarial deficit, scheduled for 1999, and to refund the 1998 payments. In addition, the capital funding premium was lowered temporarily, for 1999 and 2000, from 17 percent to 12 percent.

65. APNA’s actuarial balance, however, does not indicate the financial viability of the public sector pension system. Properly assessing the system’s sustainability requires a more comprehensive analysis of all public sector pension obligations and the broader effects of population aging.

Table 3.

Netherlands Antilles: Pension Premiums of the Central Government and the Island Government of Curaçao, 1993–99 1/

(In millions of NA guilders)

article image
Source: APNA.

Pension premiums are based on APNA’s final settlements, rather than on the premiums charged by APNA at the beginning of each year, except for 1998 and 1999.

Includes fines and interest.

66. APNA’s recent actuarial surpluses to a large extent reflected high returns on investment in 1997 and 1998—which could be temporary. Since 1997, APNA’s assets are recorded at market value instead of at historical value. As a rule, half of asset price increases is reserved for cushioning possible future asset price declines (the ‘investment reserve’), while the other half helps improve the actuarial balance. In 1997, this ‘other half accounted for just over half of an annual surplus of NA f. 114 million—which brought about the above mentioned actuarial surplus by year-end.

Table 4.

Netherlands Antilles: APNA’s Preliminary Balance Sheet at end-1997

(In millions of NA guilders)

article image
Source: APNA.

67. APNA’s assets include sizable premium arrears, indicating a potential future liquidity problem and the existence of doubtful claims. At end-1997, total arrears on scheduled payments to APNA amounted to NA f. 308 million, comprising 14 percent of ‘available’ assets.22 The largest amounts were owed by the island governments of Curaçao (68 percent) and Bonaire (15 percent). Further growth in arrears was recorded in 1998, as the island government of Curaçao was largely unable to pay. This accumulation of premium arrears could eventually undermine APNA’s ability to meet its obligations, as its cash flow is projected to become negative over the next decade.23

68. More generally, arguably APNA’s claims on the government, which comprise almost half of APNA’s assets, should be excluded from a broader assessment of the pension system’s sustainability. APNA has financed about 25 percent of total general government debt, and the option of delaying payments to the pension fund has served as a major shock absorber for the budget. At end-1997, government loans constituted 36 percent of APNA’s total investments, below the statutory ceiling of 40 percent. However, a further 3 percent consisted of other direct loans to the public sector and government guaranteed loans. In addition to the government loans, the share of the island government of Curaçao already exceeded its limit within the overall ceiling. Furthermore, premium arrears are not included in investments; if they had been regularized, this would have brought the share of government loans to 45 percent. On a net basis, claims on the government provide no funding for future pension obligations. To honor these claims, ultimately, fiscal consolidation is required. Even if the loans could be refinanced, the government would end up paying debt service costs instead of pension benefits—i.e., the annual costs would be similar to those in a formally nonfunded system. At a practical level, however, the scope for refinancing outside the pension system is limited, given a fiscal projection in which the government already exhausts the room for domestic financing without crowding out private investment, and the limits to external financing.

69. In the absence of a sinking fund, the projected surge in the government’s PAYG payments could be unfeasible. APNA’s actuarial analysis assumes full payment by the government of the early retirement scheme and of the increase in the PAYG contributions for supplementing the funded part of the pensions of current retirees. The PAYG contributions are expected to rise from NA f. 3 million in 2000 to NA f.11 million in 2005, and peak at NA f. 38 million by 2020. Regarding the early retirement scheme, actual costs have so far been very limited compared to initial projections of the maximum costs. In 1998, the actual VUT costs amounted to NA f. 0.5 million for the central government and NA f. 2.4 million for the island government of Curaçao, compared to maximum amounts of NA f. 5.1 million and NA f. 9.4 million, respectively. In practice, mainly teachers (employed by the island government) make use of the scheme, as they are allowed to maintain a paid job without losing the early retirement benefit. No projections are available of the likely future use of the scheme, but it appears very unlikely that the projected build up to annual costs of about NA f. 33 million by 2002 will be approached, as this scenario assumed full use of the facility.

70. Population aging will limit the tax base, while requiring higher AOV and health insurance contributions—putting pressure on the room for taxation. The proportion of the population of over 60 years of age is projected to expand from 10.3 percent in 1997 to 13.2 percent in 2005 and 15.1 percent in 2010.24 For the current AOV arrangement, this was calculated to imply a rise in the premium from 7.8 percent of total income in 1997 to 11.1 percent in 2005 and 12.8 percent in 2010. In addition, under unchanged policies there will be an upward effect on the costs of health care, which is also largely financed through social insurance premiums. However, the combined level of premiums and taxes is constrained by tax competition with other islands and the need to ensure proper functioning of the labor market.

71. Although data limitations preclude a comprehensive analysis, the public sector pension system does not appear to be viable. Even a cursory assessment of the above-mentioned considerations underscores the emphasis that was put on improving the pension system’s financial outlook within the 1996–97 adjustment program. The then scheduled sinking fund was largely targeted at creating a provision for the future costs of early retirement. While this particular component of the pension system appears to be less costly than envisaged, the other actuarial liabilities of the public sector indicate an urgent need for higher public saving.

D. Avenues for Further Reform

72. The need to prepare for rising future pensions costs for civil servants calls for an improvement in the consolidated net asset position of the government and the public sector pension fund. The lack of sustainability of the current pension arrangement for civil servants implies a need for reform, that could comprise both entitlements and the financing arrangement. Concerning the latter, it is crucial to see through the division between the government and APNA, and focus on improving their consolidated position.

73. The future cost increase, argues for a reduction in entitlements of current employees and/or retirees and raising the premium for employees. There is not much scope for boosting the government’s pension contributions through higher taxes and reduced expenditure outside the pension system. In addition, reducing pension entitlements for new participants only would not have a strong immediate impact. These constraints leave a reduction in pension entitlements for existing retirees, and shifting pension premiums from the government to its employees as the primary remaining avenues for lowering public sector pension liabilities.

74. Phasing in a pension deduction for the AOV benefit and terminating the early retirement scheme would provide immediate savings and eliminate a contingent liability. Pension entitlements of public sector employees covered by the old pension system (i.e., who were employed before January 1, 1998) are overly generous: in the absence of an AOV franchise, benefits are often higher than previous income. Introducing a franchise for future retirees would eliminate the scheduled escalation of PAYG payments. Phasing in such a deduction for the about 4,700 current retirees also, would provide large immediate savings. A further option would be to phase out the early retirement scheme for current employees. Given the limited use of the early retirement scheme, the immediate financial impact of such measure would be small, but a large potential liability would be eliminated.

75. Additional measures could be adopted to lower pensions. Despite the recent adjustments for new employees, pension entitlements are still generous in comparison with those outside the government. This is illustrated by the required pension premium for civil servants of 20 percent, which is well above what is customary in the private sector.

  • Pensions could be based on the average wage instead of on the wage at the end of a civil servant’s career.

  • The replacement ratio could be reduced, e.g., from 70 percent to 60 percent.

  • The period required to build up a full pension could be extended further, from 35 years to, for example, 40 years.

76. For the longer term, an important choice relates to the degree to which the system is funded. On efficiency grounds, a fully funded system is superior if the risk-adjusted real interest rate exceeds the rate of growth of wages and salaries (Box 2). For the Netherlands Antilles, with a stagnant labor force, this condition appears to be satisfied—as is the case for most countries. It is, however, important to recognize that financial consolidation of the public sector can be implemented regardless of the degree of funding. Within the funded part of the system, APNA could reduce its provision of loans to the government; given the government’s financing constraint, this would compel higher public saving. To the degree there is no funding, the delayed establishment of a sinking fund would still be the appropriate step.

Efficiency of Pension Schemes

The choice between a funded and a pay-as-you-go (PAYG) system can be judged by the rate of return on pension contributions, as shown by Aaron (1966). Under a PAYG system, the implicit rate of return on pension contributions corresponds to the growth rate of these contributions, as this translates into increased pension income. On the assumption that pension contributions are a constant fraction of labor income, this rate is given by the real rate of growth of wages and salaries. Accordingly, if the (risk-adjusted) real interest rate earned on the pension fund’s assets exceeds the sum of the growth rates of the labor force and labor productivity, a fully funded system yields a higher rate of return, making it superior from the point of view of each generation, as it can provide the same benefits at lower contribution rates. This advantage of a fully funded system reflects the associated higher capital stock, which yields this higher rate of return.

The above analysis ignores the transitional effects of changing the financing arrangement. Although changes to public pension systems are often of a piecemeal nature, it is illuminating to consider the effects of an overall change in the financing arrangement

Moving to a PAYG public pension scheme from a situation without any public scheme would reduce domestic savings (as the need to save for one’s future would be reduced), while introducing a funded scheme would have no such effect (as savings made through toe scheme would (more than) compensate for reduced private savings). However, it does not follow that replacing a PAYG scheme by a fully funded one necessarily has a positive effect on savings (See Mackenzie, Gerson, and Cuevas, 1997). The crucial issue is the financing of the pension of the current generation of retirees. Since no funding has been built up for these obligations, the financing requires either fiscal consolidation or government borrowing. In the latter case, public sector dissaving may increase the stock of government debt by the amount of the newly created pension fund, leaving overall savings unchanged. This reflects the more general point that to the extent that pension funds assets are invested in government securities which, in turn, allow for an increase in government spending, there is no real counterpart to the system’s reserves.

The analysis of replacing a fully funded scheme by a PAYG system largely mirrors the above. The transition creates a windfall gain, as the existing fund would no longer be needed and could be disbursed. To the extent the proceeds are spent, there is an immediate temporary decline in savings. In addition, there would be a smaller permanent reduction in savings, as a smaller stock of assets is maintained.

77. Stronger institutional guarantees should be introduced to ensure the integrity of the pension system. The undermining of the sustainability of the pension system during the 1990s, points to inadequate governance. To some extent this development was facilitated by the reintroduction of PAYG financing. Politicians may be held accountable more strongly for exercising the budgetary discipline needed to prepare for future pension outlays if the required contributions are clearly visible. In this respect, a fully funded system makes explicit what would be the nonfunded liabilities of a PAYG system. In the case of the Netherlands Antilles, indeed, the planned creation of a sinking fund linked to the new PAYG element of the scheme has been discarded, whereas the nonpayment of obligations to APNA are recorded as arrears and, as such, have remained on the policy agenda. As a further guarantee, a pension fund can be made subject to strict external prudential supervision.

E. Bibliography

  • Commissie Pensioenen, 1997, “Op Weg naar een Pensioenbeleid voor de Nederlandse Antillen,” Willemstad, December.

  • G.A. MacKenzie, P. Gerson, and A. Cuevas, 1997, “Pension Regimes and Saving,” IMF Occasional Paper, No. 153, August.

  • H. Aaron, 1966, “The Social Insurance Paradox,” Canadian Journal of Economics and Science, Vol. 32, pp. 371374.

  • S.K. Chand and A. Jaeger, 1996, “Aging Populations and Public Pension Schemes,” IMF Occasional Paper, No. 147, December.

14

Prepared by Jan Kees Martijn.

15

A 1995 study by Ernst, Moret, and Young indicated that of the total workforce of 78,589, 22,327 workers had supplementary pension insurance. Of the remaining workers, 7,994 had an income below 10/7 times the AOV benefit, implying that the AOV could be considered to provide full—i.e., at least 70 percent—coverage (see Commissie Pensioenen, 1997).

16

APNA’s actuarial balance, which is determined once every five years, compares the pension fund’s assets to the pension liabilities for current employees and retirees. In case of a deficit, an additional premium is established for the government, to cover the difference within a ten-year period.

17

In addition, starting in 1998, widowers’ pensions were reduced from (5/7) to (5/8) of the regular pension.

18

To clarify: total retirement income = 0.7*(final wage-10/7 AOV) + AOV = 0.7*final wage.

19

In addition, employees are to pay a premium equal to 8 percent of their wage minus the deduction for the AOV (i.e., 0.08*(wage-deduction)).

20

These charges amounted to NA f. 7 million for the island government of Curaçao and NA f. 5 million for the central government; compared to annual premium of NA f. 40 million and NA f. 30 million, respectively, in case the reforms had not been introduced.

21

This surplus equals capital excluding the investment reserve for covering adverse future asset price developments (see Table 4). The calculation of the 1997 actuarial balance did not incorporate the reforms introduced in 1998.

22

Available assets comprise total assets minus debts, the investment reserve, and the provision for maintenance (see Table 4).

23

According to a 1996 projection by Towers Perrin, for a constant number of civil servants, the number of retirees would increase sharply, by 30 percent between 1996 and 2004, as a result of the age structure of current employees and retirees, and increasing life expectancy.

24

Commissie Pensioenen (1997).