II. Pension Reform in the Slovak Republic1
1. Like many other OECD countries, Slovakia faces substantial fiscal risks—in particular to the pension system—from the aging of the population in coming decades. Demographic projections show low birth rates and significant increases in life expectancy resulting in a doubling in dependency ratios by 2040. For a pay-as-you-go (PAYG) system, this implies a need for some combination of lower replacement rates, higher retirement ages and higher contributions to restore sustainability—and contribution rates in Slovakia are already among the highest in the OECD.
2. Pension reform is in its early stages in Slovakia. In 1999 the government announced plans to introduce a three-pillar reformed pension system. Two of the three pillars—the public PAYG system (first pillar), and a voluntary, fully funded supplementary pension system (third pillar)—are in place, and Parliament approved significant changes to the first pillar in May 2002. However, the mandatory, fully funded (second) pillar remains at the planning stage.
3. This paper argues that the present reform program goes some of the way towards placing the pension system on a sustainable footing, but further measures will be needed. The paper is organized as follows. Section B reviews the current pension system and its performance in the 1990s, following its separation from the Czechoslovak system, and early reforms. Section C outlines the authorities’ reform plans as envisaged in their Pre-Accession Economic Program, and the recently approved reform of the first pillar, which reflects a blend of reforms in other advanced transition economies. Section D discusses the implications of the current round of reforms for the long-run sustainability of the pension system, the medium-term implications of the introduction of the second pillar, and the next steps for reform. Section E concludes.
B. The Current Pension System and the Need for Reform
4. The public PAYG pension system has generated small but increasing deficits over the past decade. The pension fund was balanced or ran modest surpluses in the mid-1990s, but by 1999 was running a deficit of 0.7 percent of GDP (see Table 1). This deterioration has come on the contribution side: pension expenditures have been stable at around 7½ percent of GDP over the past decade, but revenues have fallen from around 8 percent of GDP in the mid-1990s to around 7 percent of GDP now (see Table 2). The deterioration in contributions has reflected mainly weaker employment conditions in recent years, as well as subdued wage growth. A further factor underlying the deterioration in contributions was a period of nonpayment of contributions by some key large public enterprises.2
|Number of old-age pensioners (thousands)||553||556||558||561||567||573||585||587||…|
|Number of contributors (thousands)||2,060||1,999||2,048||2,149||2,135||2,200||2,102||2,057||…|
|Old-age dependency ratio (percent)||26.8||27.8||27.2||26.1||26.6||26.0||27.8||28.5||…|
|Number of other pensioners (thousands)||619||622||615||612||610||612||614||616||…|
|Total number of pensioners (thousands)||1,172||1,178||1,173||1,173||1,177||1,185||1,199||1,203||…|
|Systemic dependency ratio - all pensioners (percent)||56.9||58.9||57.3||54.6||55.1||53.9||57.0||58.5||…|
|Life expectancy (years)|
|Monthly pension benefits (average, Sk)|
|Monthly wage (average, Sk)||5,379||6,294||7,195||8,154||9,226||10,003||10,728||11,430||12,365|
|Replacement rates (percent)|
|EXPENDITURE (Sk billion)|
|Pension revenue in Sk billion:|
|Contribution rate, old age (percent)||26.5||26.5||26.5||27.5||27.5||27.5||27.5||27.5||28.0|
|Contribution rate, disability (percent)||5.5||5.5||5.8||4.8||4.8||4.8||4.8||4.8||4.8|
|Maximum assessment base (end-year)||19,600||21,600||21,600||24,000||24,000||32,000||32,000|
|Memo: minimum wage (end-year)||2,450||2,450||2,450||2,700||2,700||3,000||3,600||4,400||4,920|
|Collection rate (percent)||98.2||…|
|BALANCE (Sk billion)|
|Old age pensions, in percent of GDP:|
|Pension system balance||0.1||0.0||0.7||0.0||-0.1||-0.7||-0.2||-0.4|
|Nominal GDP (Sk billion)||406.6||486.1||568.9||628.6||708.6||775.0||835.7||908.8||989.3|
|CPI inflation (year-average)||23.0||13.5||9.8||5.8||6.1||6.7||10.6||12.0||7.3|
|Nominal wage growth (percent)||18.4||17.0||14.3||13.3||13.1||8.4||7.2||6.5||8.2|
|Index of real wage (1993=100)||100.0||103.1||107.3||115.0||122.6||124.5||120.8||114.9||115.9|
|Index of real pensions (1993=100)||100.0||98.3||103.7||108.6||112.4||116.4||113.0||110.2||…|
|Labor force participation rate||47.9||46.9||47.3||47.9||48.3||48.3||49.3||49.9||50.1|
|(In millions of koruny)||(In percent of GDP)|
|Revenues of sickness insurance, total||7,693||7,310||9,576||9,817||10,001||10,994||11,970||1.4||1.2||1.4||1.3||1.2||1.2||1.2|
|Of which:||Insurance premia paid by|
|National Labor Office||12||93||122||162||204||186||137||0.0||0.0||0.0||0.0||0.0||0.0||0.0|
|Revenues of pension scheme, total||44,603||50,932||51,503||56,299||56,546||66,625||67,800||7.8||8.1||7.3||7.3||6.8||7.3||6.9|
|Of which:||Insurance premia paid by|
|National Labor Office||424||696||709||930||1,168||1,056||802||0.1||0.1||0.1||0.1||0.1||0.1||0.1|
|Memo:||Total excluding government, other||39,839||46,740||49,589||53,255||53,479||57,485||64,118||7.0||7.4||7.0||6.9||6.4||6.3||6.5|
|Systematic sickness insurance benefits||5,708||7,388||8,115||8,974||9,496||9,144||8,879||1.0||1.2||1.1||1.2||1.1||1.0||0.9|
|Systematic pension scheme benefits||41,438||46,089||51,000||56,883||61,870||68,314||73,379||7.3||7.3||7.2||7.3||7.4||7.5||7.4|
|Social insurance benefits paid by government||14,884||15,487||16,100||17,378||16,437||15,570||…||2.6||2.5||2.3||2.2||2.0||1.7||…|
5. The state budget has covered an increasing share of the recent deficits of the Social Insurance Agency (Sociálna Poist’ovňa, or SIA). In the years up to 2000, the state budget provided around Sk 1 billion in annual transfers to the SIA. But in 2001, state budget support for old-age pensions was increased to Sk 3.7 billion, and the 2002 budget provides for a corresponding transfer of Sk 6.7 billion. The state budget has provided a further Sk 1 billion in both 2001 and in the 2002 budget for sickness benefits. Aside from the state budget, public money from the National Property Fund has also been used in 2001 and 2002 to finance repayment of arrears of public enterprises to the SIA.
6. Even allowing for some improvement in the financial condition of the pension system over the next few years, the long-term outlook is worrying. The current pickup in wages and employment suggests there is scope for some near-term recovery in contributions. But the aging of the population poses serious longer-term challenges. Demographic projections prepared by the Ministry of Labor, Family and Social Affairs (MLSAF) show a steady increase in life expectancy and a gradual decline in the Slovak population from 2010 (see Table 3), implying a doubling in the old-age dependency ratio by 2040. Figure 1 shows MLSAF projections of the long-term evolution of the demographic support ratio (the inverse of the dependency ratio) under different retirement age assumptions.
Figure 1.Demographic Support Ratio
Macroeconomic Scenario, for the period 2010–2085
|Average rate of inflation||2.1||2.6|
|Average unemployment rate||6.2||10.2|
|Labor productivity growth (real)||2.23||1.73|
|Average wage growth (real)||1.62||1.12|
|Average GDP growth (real)||1.68||1.17|
Demographic Scenario, for the period 2010–2080
|Number of live births||55 152||52 691||59 011||64 287||56 635||56 151||56 474||57 257||54 990|
|Population (in millions)||5.41||5.42||5.35||5.34||5.2||4.99||4.81||4.65||4.56|
|Average population age:|
7. The MLSAF estimates that, if unchanged, the current system would run steadily increasing deficits reaching around 6 percent of GDP by 2050. The deficits would arise both from lower contributions (as a share of GDP) because of the shrinking labor force, as well as from higher pension expenditures (Figure 2). For a PAYG system, restoring sustainability implies a need for some combination of lower replacement rates, higher retirement ages and higher contributions.
Figure 2.Long-Run Balance of the Pension System
8. Pension contributions are currently 28 percent of gross income, among the highest in the OECD. The total comprises employer contributions of 21.6 percent, and employee contributions of 6.4 percent. Contributions are normally limited to a ceiling set at eight times the minimum wage, though this ceiling is not always adjusted to reflect increases in the minimum wage. Since January 1, 2000, the maximum monthly assessment base has been set at Sk 32,000.3 Contribution rates in Slovakia are among the highest in the OECD, especially after including other social insurance taxes, although comparable to those in other advanced transition economies (Box 1; see Table 4).
|Pensions||All Social Insurance|
|Memo: EU average (unweighted)||13.2||9.4||22.5||36.7|Box 1:Payroll Taxes in Slovakia
Old-age pensions account for just over half of total social security contributions. Including health and unemployment insurance contributions, payroll taxes in Slovakia currently total 50.8 percent of gross wages, as follows:
Slovakia’s social insurance contribution rates are high by OECD standards, though comparable with those in other advanced transition economies, including the Czech Republic (48.5 percent), Hungary (44.3 percent), and Poland (48.0 percent). By comparison, the average contribution rate in EU countries is 37 percent, of which 23 percent is for pensions (see Table 4).
9. The current benefit formula is extremely redistributive and distorts labor market incentives. Assuming no special privileges, all contributors with income above Sk 10,000—and economy-wide average monthly wages are now above Sk 12,000—retire on the same monthly pension of Sk 6,395 (Box 2). For an employee earning the economy-wide average, this implies a replacement rate of just over 50 percent; for an employee earning up to the contribution ceiling of Sk 32,000, the replacement rate is around 20 percent. However, an employee insured for 42 years and earning the minimum wage would retire on a monthly pension exceeding previous earnings—strengthening incentives for early retirement.
10. Retirement ages are low, especially for women. The statutory retirement age for men is 60; for women, the retirement age is 57, or one year earlier for each child up to the fourth child, hence as early as 53. These retirement ages are quite low by OECD standards; 65 is the usual statutory retirement age (see Table 5). However, according to MLSAF staff, actual retirement ages in Slovakia much closer to the statutory ages than in most other OECD countries, where actual average retirement ages may be several years less than the statutory retirement ages.
|Memo: EU average (unweighted)||64.7||63.3|
Data refer to 2000.
11. Following separation from the Czech Republic, early reforms gave priority to clarifying the institutional framework for social security provision. The Social Insurance Act of 1994 established the SIA as the provider of pension and disability insurance, carving it out of the former National Insurance Agency which had also administered health benefits.
Box 2:Current Benefits
Pension benefits under the existing system are based on “reduced” late-career earnings. A full pension is paid after 25 years insurance, based on 50 percent of reduced average earnings during the highest five of the last ten years in employment. A further 1 percent of reduced average earnings is paid for each year of employment between 26 and 42 years. The following marginal reduction scale (unchanged since 1988) applies to average monthly earnings:
|Sk 0 - Sk 2,500||no reduction|
|Sk 2,501-Sk 5,999||two-thirds reduction|
|Sk 6,000-Sk 9,999||90 percent reduction|
|Sk 10,000 and above||100 percent reduction|
The scale implies maximum reduced monthly earnings of Sk 4,067. The unadjusted monthly benefit for a worker employed for 42 years would be 67 percent of this amount, i.e. Sk 2,725. Parliament has increased pensions annually, such that in 2002, the actual monthly pension is given by the unadjusted benefit multiplied by 1.905, plus Sk 1,204. This implies a standard maximum monthly pension of Sk 6,395. The current law also provides for pension privileges for numerous groups, in some cases implying (unadjusted) pensions of 100 percent of the reduced wage. The current maximum monthly pension for all pensioners is Sk 8,282.
12. Voluntary pension insurance was introduced in 1996.4 There are now four supplementary insurance companies with Sk 5-6 billion in assets covering around 300,000 participants. The third pillar was initially open to employees only; in 1999 self-employed workers were given the right to participate as well. Participation in the third pillar has been less than originally expected, perhaps reflecting a lack of transparency in the operations of the supplementary insurance companies and their managed funds.
13. Employee contribution rates were increased in 2001. As a step towards addressing the growing imbalances of the social security funds, the government increased the employee contribution rate for pensions from 5.9 percent to 6.4 percent, effective January 2001; health insurance contributions were also raised from 3.7 percent to 4 percent.
C. The Government’s Current Pension Reform Strategy
14. Early in its term, the current government adopted the objective of introducing a three-pillar pension system. The MLSAF (1999) outlined a three-pillar system with two compulsory pillars and one voluntary pillar:
First pillar: a compulsory PAYG pillar financed as in the existing system, providing basic pensions;
Second pillar: a compulsory, fully funded pillar with individual accounts, operated by the SIA and with principal (adjusted for inflation) guaranteed by the state, but asset management subcontracted to managers of the insured person’s choice; and
Third pillar: a voluntary, private, fully funded pillar.
The concept assumed the first two pillars would replace about 50-55 percent of gross real income, though with a ceiling of about three times the economy-wide average wage; the third pillar would replace 20-25 percent of net earnings without a ceiling.
15. The government’s 2001-2002 Staff-Monitored Program with the Fund included two structural benchmarks on pension reform:
The first benchmark, for June 2001, envisaged government approval of a gradual increase of the retirement age. The benchmark was eventually met in October 2001 with government approval of a timetable for increasing the retirement age to 60 for all women by 2018; the government later agreed to extend the transition period to 2026.
The second benchmark, for March 2002, envisaged government approval of a comprehensive model for the introduction of a three-pillar pension system. This benchmark has been delayed owing to the need for technical assistance to help design and implement the second pillar.
16. Reform of the first pillar has moved ahead but important decisions remain pending on the design of the second pillar. Parliament recently approved the new Social Insurance Act, which reforms the PAYG system. Turning to the second pillar, the government has yet to decide on its size or institutional framework, but has already set aside privatization receipts to finance the transition costs arising from the diversion of contributions from the first pillar.
17. A World Bank-assisted project to support these reforms, and other improvements in the social protection system, is also pending. The Bank approved the Social Benefits Reform Administration project (SBRA) in February 2002. The project will provide technical assistance to help design the second pillar; improve the collection and administration of social security contributions, with a single agency collecting for the SIA and National Labor Office (NLO); and support capacity building at the MLSAF, SIA and NLO. The government has yet to complete the steps needed for disbursement of the SBRA loan.
18. The move to a three-pillar pension system in Slovakia follows the recent introduction of multi-pillar systems in several other advanced transition economies. The reform of the first pillar in Slovakia reflects a blend of reforms in several other countries, including Latvia, Poland and Hungary. Progress to date in introducing the second pillar in the advanced transition countries has been more mixed (Box 3).
Box 3:Lessons from Other EU Accession Countries
The move to a three-pillar pension system in Slovakia follows the recent introduction of multi-pillar systems in several other advanced transition economies. The recent experiences of Latvia, Poland and Hungary are particularly instructive.1 Latvia in 1995 was the first transition economy to introduce a notional defined-contribution (NDC) system to reform the PAYG first pillar, though progress since in introducing its second pillar has been slower. Hungary launched a multi-pillar pension system in 1997, though it has recently backed away from some of the second-pillar reforms. Poland introduced a multi-pillar system in 1999, like Latvia reforming its PAYG pillar along NDC lines.
The reform of the first pillar in Slovakia reflects a blend of reforms elsewhere. The approved Slovak personal wage point system, though still defined-benefit PAYG, shares a number of features with the NDC systems in Latvia and Poland (Box 4). The continuing broad role for the first pillar in providing basic incomes, rather than a more specific focus on poverty relief, also remains in reformed first pillars in other transition economies. Although Latvia introduced the NDC system to tighten the benefit-contribution link, it also retains a minimum pension guarantee within the first pillar. On the other hand, Hungary and Poland—now followed by Slovakia—have abolished minimum pensions and provide a minimum income from outside the pension system.
Comprehensive approaches linking reform of the first pillar to the introduction of the second seem to have advantages over more piecemeal reforms—on balance. Comprehensive reforms allow for changes to first-pillar benefits to mesh with the second pillar, improving their design. Also, given the continuing role of reformed first pillars in providing basic pensions, some policymakers may come to perceive the introduction of the second pillar as less urgent. Fox and Palmer (1999) argue that the introduction of the second pillar in Latvia suffered from being left until after the first pillar; the Latvian second pillar started operations in 2001, six years after the reform of the first pillar. In Poland, the reformed first pillar and the new second pillar were implemented simultaneously, allowing for their close integration (Góra, 2001). On the other hand, the comprehensive reform in Hungary has gone less smoothly. Following the 1997 reform that both modified the PAYG pillar and launched the second pillar, in late 2001 the government reversed several features of the initial second-pillar reform, including keeping the contribution rate below envisaged levels, and eliminating its mandatory coverage and minimum guarantee.
The transition costs of introducing funded pillars have complicated implementation in some countries. Latvia is spreading the transition costs over a longer time frame by phasing in its second pillar. The fully funded pillar was introduced in 2001 with an initial contribution rate of 2 percent; the contribution rate will rise to at least 4 percent in 2007, and reach 10 percent from 2010 (Schiff et al., 2000). In Hungary, concerns that the transition costs would increase the difficulty of meeting the Maastricht fiscal deficit criterion may have contributed to the recent stalling of the second-pillar reform (Wagner, 2002).
But other countries have moved more ambitiously than Slovakia to increase retirement ages, despite considerable opposition. In Latvia, retirement ages will be equalized at 62, up from 60 for men and 57½ (with the option of early retirement at 55) for women. The originally approved timetable for retirement age increases was slowed after opposition forced a national referendum; the revised timetable now provides for the retirement age to increase annually by six months until it reaches 62 for both men and women. In Poland, the original reform proposal to increase retirement ages to 62 for both men and women also proved controversial; the eventual law increased the retirement age for women from 55 to 60, and for men from 59 to 65. In Hungary, a retirement age of 62 for both men and women will be fully phased in by 2009. The Czech Republic, which formerly had the same retirement ages as Slovakia, is phasing in higher retirement ages for men from 60 to 62 years, and for women from 53-57 to 57-61 years, to be completed by 2007 (Laursen, 2000).1Fox and Palmer (1999) and Schiff et al. (2000) review the pension reforms in Latvia. Rocha and Vittas (2001) assess the implementation of the pension reform in Hungary, and Chlon et al. (1999) outline the pension reform strategy in Poland.
Reforming the first pillar: the 2002 Social Insurance Act
19. Parliament approved the new Social Insurance Act in May 2002. The Act reforms the first pillar by raising the retirement age for women and tightening the link between benefits and contributions. The Act becomes effective in July 2003. Its main features are:
The Act equalizes the retirement age at 60 for both men and women. In response to opposition to a (previously approved) shorter transition period, the approved transition period will last until 2017 for childless women and until 2026 for women with four children.
Social insurance will cover only those benefits that replace previous earned income. Accordingly, the SIA will pay sickness/disability benefits, accident benefits and pension benefits. It will not pay the “non-systemic” benefits from the previous system, including the minimum pension and wife’s pension; these would instead be covered by state benefits (paid by the state budget). A range of special pension privileges are also cancelled.
The Act introduces “personal wage development points” to tighten the link between benefits and contributions. Points will be earned in each year of employment, so that benefits will reflect lifetime earnings and not just earnings in the final years of employment. Although the new points scheme will remain defined-benefit (DB), it has much in common with the notional defined-contribution (NDC) systems introduced in some other European countries (Box 4).
Regarding the personal wage point system, for which a “wage point” is the ratio of the employee’s wage to the economy-wide average wage in a given year of employment, the new system will continue to apply a reduction scheme. For wages less than 1.25 times the economy-wide average, the employee will receive full wage point credit; between 1.25 and 2 times the economy-wide average, one-third wage credit; and zero for more than 2 times the economy-wide average. For example, an employee earning 1.55 times the economy-wide average would be credited with 1.35 wage points.
Workers with 40 points, equivalent to 40 years earning an economy-wide average wage, will receive a pension upon retirement equal to 50 percent of the economy-wide average wage. Pensions will be increased every July 1 by the lower of the increase in the CPI and the increase in the economy-wide average wage over the previous twelve months.5
Unemployed workers, whose premia are paid by the NLO, earn wage points at an annual rate of 0.3 points. Women receive 0.5 points for a childcare period.
Box 4:The Slovak Reform and Notional Defined-Contribution Schemes
The reformed Slovak PAYG pension system will remain defined-benefit (DB), but with several key characteristics of the notional defined-contribution (NDC) or “notional accounts” systems introduced in several European countries, including Latvia, Sweden, Italy and Poland.
The reformed Slovak PAYG system and NDC systems record wage points or pension contributions over an employee’s working life in individual accounts. These accounts then represent individual claims on future public resources.
The resulting benefit-contribution link is generally less than one-to-one—the systems typically include some redistributive element, such as a minimum pension (Latvia) or a reduced-earnings scheme (Slovakia). But the link is tighter than in the unreformed PAYG systems that base benefits on income in the final year or years of employment.
The reformed Slovak system will record “personal wage points” for conversion to pension income upon retirement, based on the prevailing economy-wide average wage. Hence, the system remains DB.
NDC systems record individual contributions, which earn “notional” rates of return. This rate of return is typically set at the growth rate in die covered wage bill. Hence, although these systems are unfunded, they still operate on defined-contribution (DC) principles.
Despite the apparent DB-(N)DC distinction, there is an underlying equivalence to these systems. In the Slovak system, basing pensions on economy-wide average wages at the time of retirement implies an implicit rate of return equal to wage growth. This is equivalent to an explicit but notional rate of return based on wage growth in an NDC system.
NDC systems remain somewhat controversial. Fox and Palmer (1999) argue that NDC systems are more transparent because the actuarial indexation and benefit calculations of the PAYG system are made explicit. Disney (1999) argues that notional accounts systems are effectively identical to well-designed PAYG systems, but lack transparency because of the complexity of the actuarial calculations. On this criterion, the reformed Slovak PAYG system may be more easily understood by the public, to the extent that the role of wage growth in determining pensions is clearer than in systems that use a notional rate of return as an extra intermediate step in the pension calculation.
The corollary of the tighter link between benefits and contributions is a reduction in the degree of redistribution. The proposed system, while still redistributive, does envisage significantly reduced pensions for newly retiring lower-income earners. The Act includes a three-year transitional period under which new benefits would be no lower than under current law.
The reformed first pillar relies on the state budget to provide a safety net for lowest-income retirees. The new benefit formula would yield some pensions significantly below the subsistence level, particularly for women on low incomes who retire early. In these instances, the state budget will top up pensions to cover subsistence.
The new benefit formula results in higher new pensions, at least during the three-year transition period, but the additional costs should be modest. The MLSAF estimates higher spending on new pensions by 20 percent in the first year of the new system, by 18 percent in the second year, and by 16 percent in the third. With annual spending on new pensions of around Sk 2.5 billion, the cost of new benefits in the first year would be Sk 500 million; the additional cost in 2003 would be only Sk 250 million because the law would apply only in the second half of the year.
Social insurance contributions remain unchanged, both in terms of the rate (28 percent) and the maximum monthly assessment base (Sk 32,000). Parliament rejected a provision in the draft act to raise the maximum assessment base to 3.25 times the average wage.6
Parliament also approved an amendment reducing employees’ contributions by 0.5 percent per child, with an annual cost to the SIA of Sk 750 million. However, there is also a deduction of 0.5 personal wage points when the benefit is received.
A further amendment reallocates the costs of spa care from the health system to the pension system. The annual cost to the SIA is estimated at Sk 650-700 million.
Introducing the second pillar: plans and provisions
20. The Board of Economic Ministers is now considering two alternatives for the institutional framework for the second pillar. Under the first alternative, an independent unit of the SIA would be responsible for the collection, payout, and management of accounts, while private asset managers will manage the funds. Under the second, the SIA would be responsible for collecting contributions, but private managers would undertake all remaining functions. In either case, the Financial Market Authority would be responsible for licensing and supervision. With parliamentary elections scheduled for September 2002, it is likely that the final decision on the institutional framework will fall to the next government. Nevertheless, MLSAF staff are optimistic that preparatory work could be completed in the first half of 2003 and that the second pillar could be functioning by January 2004.
21. Privatization receipts have been set aside to finance the transition costs associated with the introduction of the second pillar. In early 2002, the government allocated Sk 65 billion (over 6 percent of GDP) for the pension reform following the successful sale of a 49 percent stake in the gas company SPP. These funds have been deposited with the National Bank of Slovakia (NBS) and interest will accrue in this account. With an average rate of return of 6 percent, these funds should be sufficient to finance the estimated Sk 75 billion transition costs7 of the second pillar.
22. The government has yet to decide on the size of the second pillar, but it is likely to start with a contribution rate of 3.7 percent of gross income. The government plans to expand the second pillar gradually. MLSAF staff expect the second pillar to start with a contribution rate of 3.7 percent, and envisage an eventual rise to 6 percent. Participation is expected to be mandatory for employees below 40 years of age, and optional for employees aged between 40 and 50; the second pillar would be closed to employees aged over 50.
23. The initial annual transition costs could be around 1 percentage point of GDP. MLSAF estimates of the transition costs are quite high—Sk 15 billion in 2004, falling to Sk 10-11 billion over 2005-07 and starting to taper off thereafter. However, these costs include not only the cost of diverting 3.7 percent of contributions to the second pillar, but also the projected deficit of the first pillar under no reform (assumed to be Sk 4 billion in 2004), plus administrative costs, currently equivalent to 3.5 percent of the revenues of the first pillar. The funds at the NBS will be used exclusively to compensate for the diversion of contributions to the second pillar, and will not fund second-pillar administrative costs.8
D. Assessing the Reform So Far
24. This section discusses the implications of the reforms for the long-run sustainability of the pension system, the medium-term implications of the introduction of the second pillar, and the next steps in the reform process.
The long term: sustainability
25. The first pillar reform significantly reduces the long-term imbalance of the SIA, but does not in itself ensure financial sustainability (Figure 3). The MLSAF estimates that the approved reform results in the deficit of the SIA reaching 2 percent of GDP by 2050, compared with a deficit of 6 percent of GDP under the current system. The lower deficit is achieved by containing expenditures, through increases in the retirement age and (effectively) limiting growth in pensions to inflation. Figure 3 also shows scenarios with varying retirement ages: assuming the present reform without changes in the retirement age, the deficit would reach 3 percent of GDP by 2050; with the retirement age at 63, the deficit would reach 1-1½ percent of GDP by 2050.
Figure 3.Long-Run Balance of the Pension System
26. The indexation mechanism is a major factor constraining future expenditures, and should also reduce incentives for early retirement. As noted earlier, pensions will be adjusted annually by the lower of the increase in the CPI and in economy-wide average wages. Assuming positive real wage growth in most years implies, for practical purposes, CPI-indexation. This is the cheapest form of indexation from the fiscal point of view; the MLSAF estimates that indexing pensions by the increase in wages would imply an additional deficit of 1.3 percentage points of GDP by 2085. Moreover, at the time of retirement, points are converted to a pension based on prevailing economy-wide wages. Hence, assuming positive real wage growth in the long run, then the implicit pre-retirement rate of return on accumulating pensions exceeds the post-retirement growth of pensions, creating incentives to delay retirement.
27. The tighter link between benefits and contributions should mitigate labor market distortions. The shift to basing pensions on lifetime contributions, instead of salaries in the final ten years, tightens the link between benefits and contributions, albeit partially because of the continuing redistributive function of the reduction scheme. The tighter benefit-contribution link should promote participation in the formal sector to the extent that employees see social security contributions more as insurance premia and less as taxes.
28. Although the MLSAF analysis projects a balanced system in the long run after the inclusion of the second pillar, risks to these projections appear skewed to the downside:
The imbalance in the system may be greater if life expectancy increases more rapidly than projected by the MLSAF. The United Nations projects that by 2050, life expectancy for Slovak men will reach 76.6 years, compared with 71.6 years in the MLSAF projections; and for women to reach 82.4 years, compared with 80.3 years in the MLSAF projections (see Table 6). Without taking a view on which of the two sets of demographic projections is better, risks to dependency ratios may be to the upside. More pertinently, the lack of a mechanism to cope with increases in life expectancy leaves the system still open to demographic risk.
The new indexation mechanism may be subject to political risk. Because pensions will be maintained in real terms but not increased, political pressures could emerge in the medium term for indexation reflecting wage growth as well as inflation, for example so-called “Swiss” indexation that gives 50 percent weight to wage growth, and 50 percent to inflation. This could increase the deficit of the SIA by ½-¾ percentage points of GDP compared with the approved reform.
More work is needed to quantify the impact of the reform on the state budget. The MLSAF quantification of the reform provides little information on the projected impact on the state budget beyond the medium term. Because of the abolition of the minimum pension and other nonsystemic pensions (such as wife pensions), some pensions will be below subsistence levels—particularly for women on low incomes who retire early—and will need to be topped up by the state budget. According to MLSAF staff, the effect on the state budget will be small, and the state budget already provides some top-up income support; but more information is needed to analyze the impact of the reform on the state budget as well as on the pension system.
The medium term: transition costs, and next steps
29. In general, the transition costs associated with the introduction of a funded second pillar need not imply a loosening of the fiscal stance, depending on the implications for private savings. Mackenzie et al. (2001) observe that, under certain conditions, a shift to a second pillar can leave the fiscal stance unchanged, because the payroll taxes formerly collected by the public social insurance scheme are transformed into the surpluses of the private sector pension funds. In other words, the higher fiscal deficit is offset by higher private saving, and national saving is unchanged. However, Mackenzie et al. further note that an individual accounts reform that addresses the aging problem without new incentives to increase private saving would loosen the fiscal stance and hence require offsetting fiscal measures.
30. But in the case of Slovakia, the government would still need to take fiscal measures to offset the transition costs in order to meet its other medium-term objectives, including adoption of the euro. The government recently announced its intention to meet the Maastricht fiscal deficit criterion, that is, reduce the fiscal deficit to below 3 percent of GDP on an ESA95 basis, by 2006. Annual transition costs of around 1 percent of GDP will clearly add to the challenges in meeting the deficit criterion—these costs still increase the fiscal deficit9, even though the funds have been allocated to finance them. It is therefore critical that the staffs of the MLSAF and the Ministry of Finance collaborate closely in designing the next stages of the pension reform, and that the medium-term costs of pension reform are fully incorporated into Slovakia’s medium-term fiscal framework.
31. The shift to funded pensions is not in itself a panacea for the demographic problem but can facilitate adjustment, as well as promoting old-age income security by diversifying risks and assisting the development of Slovak capital markets. Rates of return on funded pensions are still subject to demographic pressures; declines in labor force growth could be associated with falling returns to capital.10Disney (2000) argues that funding exchanges the potential political risk of unfunded systems for potential investment risk. To the extent that these risks are uncorrelated, however, old-age income security is enhanced by diversifying the sources of retirement income to include both PAYG and funded pensions. Funded pension systems may also have political economy advantages, being self-adjusting to shocks and avoiding locking in the unsustainable benefit entitlements that can arise in PAYG systems. Moreover, through their role as institutional investors, private pension funds help to deepen capital markets. The second pillar will make an important contribution to this process, especially in view of the limited growth to date in the third pillar.
32. The envisaged private management of second-pillar assets should protect against political interference. Private management of second-pillar assets is still controversial—as indeed are individual accounts—in particular because of higher administrative costs (see, for example, Orzsag and Stiglitz, 2001). But public pension funds may be subject to more restrictions, or pressures to direct investments according to social and political objectives, that lower rates of return compared with private funds (Holzmann and Palacios, 2001).
33. Moving forward with the introduction of the second pillar in Slovakia requires considerable further analytical and logistical work, with World Bank assistance. As noted above, the government is yet to approve the institutional framework for the second pillar, but has decided on private management of second pillar funds. At a more technical level, the authorities still need to conduct a variety of simulations with regard to alternative second-pillar contribution rates, retirement ages, phase-in periods, and choice mechanisms. Technical assistance with the simulations is available through the World Bank SBRA project. The SBRA will also support the administrative reforms—improving collections, and building capacity, including for management of individual accounts—needed for the successful introduction of the second pillar.
34. The authorities will also have to revisit measures in the medium term to ensure the longer-run sustainability of the first pillar. Barr (2000) observes that for countries with large, unsustainable PAYG systems, the only choice is to make the PAYG system sustainable, despite any partial shift towards private, funded arrangements. In turn, sustainability of the PAYG system can only be achieved by some combination of higher contributions and lower benefits.
35. There is limited room to move on the contribution side. As discussed earlier, contribution rates are already very high in Slovakia and there is no apparent room to increase them. In this light, the SBRA reforms to improve the collections system are critical. As also noted, the parliament has rejected the recent proposal to increase (significantly) the maximum assessment base for contributions to 3.25 times the average wage. The revenue effect of this measure would have been modest, because only 2-3 percent of employees would have been affected (though the cost to these employees would have been large). However, there is a clear case for indexing the maximum assessment base—preferably by wage growth, in line with the implicit rate of return on contributions.
36. Measures will have to come mostly on the benefit side. The most feasible option may be to make benefits less generous through further increases in retirement ages.11 Unlike the NDC systems in some other transition economies, the reformed Slovak system does not include a mechanism to adjust benefits according to unexpected increases in life expectancy. Moreover, the recently approved increases in retirement ages for women will roughly cover the projected increases in female life expectancy over the next three decades; but the projected increase in male life expectancy is greater, and not yet matched by any increase in the retirement age for men.
37. Sustainability can best be achieved if pension reform is supporting economic growth. Output is the key variable in a context of demographic problems; lower replacement rates are consistent with old-age income security if the economy continues to grow. Thus, one of the questions for pension policy is how the reformed system can best support economic growth—highlighting the importance of reforms that not only address sustainability directly, but also reduce labor market distortions, deepen capital markets, or promote national saving.12
38. The recently approved Social Insurance Act represents a first step towards putting the pension system on a sustainable footing. The Act should help reduce both the long-run deficit of the PAYG system, and labor market distortions: the new indexation arrangements should help contain expenditures in the long run, and the tighter link between benefits and contributions should encourage participation in the formal sector.
39. The introduction of the second pillar will come with a significant medium-term price tag—but the price is worth paying. Although the necessary funds have already been set aside to finance the transition costs from diverting contributions to the second pillar, these transition costs could still increase the fiscal deficit by up to 1 percent of GDP annually in the medium term. Nevertheless, the introduction of the second pillar—while not a panacea for the demographic problem—will allow a partial shift away from the still-unsustainable first pillar, diversify retirement incomes, and contribute to capital market development. It is therefore critical that the staffs of the Ministry of Labor and the Ministry of Finance collaborate closely in designing the next stages of the pension reform, and that the medium-term costs of pension reform are fully incorporated into Slovakia’s medium-term fiscal framework. Also, the recently approved World Bank project offers technical assistance with the design of the second pillar, and will support the administrative reforms needed for its successful introduction.
40. Further steps will still be needed to ensure the sustainability of the pension system. Some measures should be relatively painless, such as indexing the maximum assessment base for contributions. Others measures, such as further increases in the retirement age, would clearly be more difficult. But even after the approved increases in the retirement age for women, retirement ages will still be well below those in most OECD countries—which are facing serious sustainability problems in their own pension systems.
BarrNicholas2000 “Reforming Pensions: Myths, Truths and Policy Choices,” IMF Working Paper WP/00/139 (Washington: International Monetary Fund).
BütlerMonika2001 “The Political Feasibility of Increasing Retirement Age: Lessons from a Ballot on Female Retirement Age,” CEPR Discussion Paper No. 2780 (London: Centre for Economic Policy Research).
ChlonAgnieszkaMarekGóra and MichalRutkowski1999 “Shaping Pension Reform in Poland: Security Through Diversity,” World Bank Social Protection Discussion Paper No. 9923 (Washington: World Bank).
DisneyRichard1999 “Notional Accounts as a Pension Reform Strategy: An Evaluation,” World Bank Social Protection Discussion Paper No. 9928 (Washington: World Bank).
DisneyRichard2000 “Crises in Public Pension Programmes in OECD: What are the Reform Options?,” The Economic Journal 110 (February) pp. F1–F23.
FoxLouise and EdwardFox1999 “Latvian Pension Reform,” World Bank Social Protection Discussion Paper No. 9922 (Washington: World Bank).
GóraMarek2001 “Polish Approach to Pension Reform,” in OECD 2000 Private Pensions Conference No. 3 (Paris: Organisation for Economic Cooperation and Development).
HellerPeter S.1998 “Rethinking Public Pension Reform Initiatives,” IMF Working Paper WP/98/61 (Washington: International Monetary Fund).
HolzmannRobert and RobertPalacios2001 “Individual Accounts as Social Insurance: A World Bank Perspective,” World Bank Social Protection Discussion Paper No. 0114 (Washington: World Bank).
LaursenThomas2000 “Pension System Viability and Reform Alternatives in the Czech Republic,” IMF Working Paper WP/00/16 (Washington: International Monetary Fund).
MackenzieG.A.PhilipGersonAlfredoCuevas and Peter S.Heller2001 “Pension Reform and the Fiscal Policy Stance,” IMF Working Paper WP/01/214 (Washington: International Monetary Fund).
Ministry of Labor Social Affairs and Family1999Concept of the Social Insurance Reform in the Slovak Republic (Bratislava).
OrzsagPeter R. and Joseph E.Stiglitz2001 “Rethinking Pension Reform: Ten Myths about Social Security Systems,” in New Ideas about Old Age Securityed. byRobertHolzmann and Joseph E.Stiglitz (Washington: World Bank).
PalaciosRobert and MontserratPallarès-Miralles2000 “International Patterns of Pension Provision,” World Bank Social Protection Discussion Paper No. 0009 (Washington: World Bank).
RochaRoberto and DimitriVittas2001 “The Hungarian Pension Reform: A Preliminary Assessment of the First Years of Implementation,” World Bank Policy Research Working Paper No. 2631 (Washington: World Bank).
SchiffJeraldNikoHobdariAxelSchimmelpfennig and RomanZytek2000“Pension Reform in the Baltics: Issues and Prospects” IMF Occasional Paper 200 (Washington: International Monetary Fund).
Prepared by David Moore.
The railways were a leading nonpayer to the social security funds in the mid- to late 1990s. Agreement was later reached for the railways to reimburse the social security funds (using privatization receipts transferred from the National Property Fund to repay old debts); in 2000 these repayments accounted for Sk 6 billion of contributions to the Social Insurance Agency—and for practically all of the improvement in reported revenues that year.
This ceiling was based on a minimum wage of Sk 4,000. However, although the minimum wage was increased to Sk 4,920 in October 2001, the maximum assessment base has remained unchanged.
The accompanying Financial System Stability Assessment document reviews the third-pillar supplementary pension insurance companies in Slovakia.
The provision for automatic indexation contrasts with the current system in which the parliament must approve pension increases. However, this provision will not apply until 2004; Parliament will still have to approve the indexation increase in 2003.
This would have implied a maximum assessment base of some Sk 45,000 per month, representing a large tax increase for upper and especially upper-middle income earners, and their employers. The MLSAF estimates the revenue impact of the parliament’s decision to be quite small in the near term, because only about 2-3 percent of employees earn more than the maximum assessment base. In the longer term, the MLSAF’s baseline projections are annual losses equivalent to around 2 percent of contributions.
This refers to the sum of current-price flows, rather than net present value.
These costs are estimated at about 0.3 percent of assets and include, for example, setting up client accounts.
If the second pillar is public, diverted contributions remain within the consolidated general government, that is, the diversion of contributions does not increase the fiscal deficit. Other EU accession countries are currently discussing with Eurostat the possibility of including the private second-pillar pension insurance companies as part of general government, for the purposes of the Maastricht fiscal deficit and debt criteria.
See for example Heller (1998).
Explaining the case for increases in retirement ages is a difficult task, but has been done elsewhere. Butler (2001) reviews the political economy of a single-issue Swiss referendum in 1998 to block an approved increase in the retirement age for women from 62 to 64; the referendum was defeated by a 60 percent majority.
See for example Barr (2000). Barr notes that the impact of funding on economic growth remains highly controversial; there is empirical evidence that funding contributes to higher savings in the United States but evidence for other countries is not robust.