Journal Issue

Republic of Poland: Selected Issues

International Monetary Fund
Published Date:
August 2005
  • ShareShare
Show Summary Details

III. The Polish Pension Reforms After Six Years16

45. This paper examines the impact of labor market and fiscal trends on the sustainability of the pension system since the 1999 pension reform. The 1999 pension reform—which phased out the old pay-as-you-go (PAYG) system and created a new mixed private-public pension system—was a major effort to restore the solvency of the public pension system. The calculations at the time of the reform suggested that the previously large implicit debt in the pension system had been eliminated. Recent official projections, however, show a different picture. As a result of a sizable decline in employment and a faster-than-expected increase in pension expenditure, the state-managed part of the old-age pension system has a much larger than expected deficit. Moreover, it is now projected to turn into a surplus that is both smaller and later than envisaged in 1999, and the accumulation of the pension savings in the privately-managed part remains moderate.

A. The Pension System Prior to the Reforms17

46. Before 1999, Poland had one of the most generous pension systems in Europe—an unsustainable situation. Despite a relatively young population,18 old-age pension spending reached 6.5 percent of GDP. Though the system was in a surplus, long-term projections showed that without reforms, the system dependency ratio19 would have reached 1 by 2050 (with a demographic dependency ratio of close to 0.4), and the annual balance of the old-age pension system would have moved from a small surplus to a deficit of about 3¾percent of GDP by 2050. Besides the unfavorable demographic trends, the main reasons for the projected rapid deterioration were a low average effective retirement age (57 years) and generous replacement ratios.20

B. The 1999 Pension Reforms

47. The 1999 pension reform phased out the previous PAYG system and created a new mixed private-public pension system. The old system was terminated for those born after 1948. The new system created two mandatory accounts for each individual: a notional defined-contribution account in the state-managed part and a funded defined-contribution account in the privately-managed part. Participation in the privately-managed part was optional for those born between 1948 and 1968. In addition to the mandatory parts, there is a voluntary part, which is funded and privately managed. The mandatory retirement age (60 for women and 65 for men) was left unchanged. The combined contribution rate for employers and employees to the mandatory old-age pension system was also left unchanged at 19.52 percent of the gross wage, but for those who participate in both parts, 7.3 percentage points of this is paid to the privately-managed mandatory pension funds (OFEs).

48. The state-managed part of the mandatory system is a notional defined-contribution scheme. Contributions are recorded in private accounts, where accumulated notional capital earns a return equal to the increase in contributions due to the state-managed part in the preceding year.21 Reflecting their participation in the old PAYG system, people were given initial notional capital in the new state-managed part. Upon retirement, the yearly benefit is calculated by dividing the accumulated notional capital by the average remaining unisex life expectancy at retirement age in the calendar year of the retirement. After retirement, benefits are indexed to the CPI.22 The benefits are paid from current contributions and the system is not funded in general, but a small demographic reserve fund was created to serve as a partial funding for this part.23

49. The privately-managed part of the mandatory system is a funded defined-contribution scheme. Participants can freely choose among licensed OFEs. Accumulated funds are invested by the fund managers, subject to regulation. Upon retirement, beneficiaries will be obliged to purchase an annuity from a specialized annuity company.24 Most risks involved in this part (return and longevity) are assumed by the participants, but there is a minimum pension guarantee which was carried over from the old system.25 Out of the eligible 11.5 million members, 9.7 million, close to 85 percent, opted to participate in both parts.

50. The third part is a voluntary, privately-managed, funded component. It can take the form of life insurance, investment fund, mutual insurance, and employee pension funds. Part of the contribution to an employee pension scheme in the third part, up to 7 percent of the gross wage, can be deducted from the social security contribution base. As this allowance existed also in the old system, the creation of the third part was not expected to result in any significant revenue loss.

51. The reform has reduced future replacement ratios and is expected to increase average effective retirement age. The replacement ratio from the state-managed part for men is projected to decline from 76 percent before the reform to 60 percent by 2035 (while the combined replacement ratio from the two parts is projected to decline to around 66 percent.26 The replacement ratio from the state-managed part for women is projected to decline from around 58 percent before the reform to 42 percent by 2030, while the combined replacement ratio drops to 48 percent.27 By making the conditions for early retirement (before 60/65) actuarially neutral, the reform is expected to increase the effective retirement age. Though the lower retirement age in fact results in correspondingly lower pensions (old-age income) for women, attempts to have the same the retirement age for men and women have failed.

52. Longevity risk, which is substantial, was also shifted to beneficiaries in the state-managed part. Available estimates (Chłon, Góra, and Rutkowski, 1999) suggest that an increase of 2 years in the remaining life expectancy upon retirement (for the entire population, which is used to calculate the pension benefit) would result in a drop of about 3 percentage points in the replacement ratio.

C. The Long-Term Sustainability of the State-Managed Part of the Old-Age Pension System

53. Projections made at the time of the pension reform suggested that the deficit following the introduction of the reform would be relatively small and transitory and that the long-term sustainability of the state-managed part was secured. Projections—based on rather conservative growth assumptions and assuming a gradual reduction in contribution rates, but assuming significantly increased labor force participation—suggested several major improvements (Chłon, Góra, and Rutkowski, 1999):28

  • Old-age pension expenditure in the state-managed old-age pension system was projected to decline to slightly above 2 percent of GDP by 2050, from about 6 percent in 1999.

  • The deficit during the transition was projected to peak at around 1.2 percent of GDP in 2004–05 and turn into a surplus by 2012, staying in the range of 0.8 to 1 percent of GDP throughout 2050. This compares to a small surplus until 2010 and afterwards a rapidly widening deficit, reaching 4 percent of GDP a year by 2050 in the old system (Figure 1).

  • Based on these projections, the implicit debt in the old PAYG system until 2050, amounting to about 140 percent of GDP, was turned into an implicit asset in the new system of about 40 percent of GDP. 29

Figure 1.Poland: The Fiscal Impact of the 1999 Pension Reform:

(The balance of the state-managed part of the pension system relative to GDP)

Source: Chłon, Góra, and Rutkowski (1999).

Note: Ratios to GDP have been recalculated to reflect revisions to the national accounts since 1999.

54. Unfavorable labor market developments, changes in the parameters, and generous pension indexation since 1999, however, have significantly worsened the position of the state-managed old-age pension system.30 A faster-than-expected decline in employment, a rapid switch to self-employment, and a faster-than-assumed (ex-post) increase31 in real pensions in 2000–03 have significantly increased the deficit following the introduction of the reform (Figures 24). Given the current labor market trends, the assumption of a rapid increase in employment in the long term made in 1999 also seems too optimistic.

Figure 2.Poland: Increase in the Number of Old-Age Pension Beneficiaries

(In percent)

Sources: IMF Staff calculations; Polish authorities.

Figure 3.Poland: Average Old-Age Pension Relative to Average Wage

(In percent)

Sources: IMF Staff calculations; Polish authorities.

Figure 4.The first-pillar pension system, 1999-2004

(in percent of GDP)

Sources: IMF Staff calculations; Polish authorities.

55. Reflecting these changes, current official projections show a smaller and later improvement in the long-term position of the state-managed old-age pension system. The projections made by ZUS in 2003 already showed a major deterioration (ZUS, 2003). Even under the most optimistic scenario,32 the state-managed old-age pension system would be in deficit until 2036 and would have an implicit debt of about 46 percent of the 1999 GDP until 2050 (Figure 5). The ZUS projections for 2006–10 published a year later (ZUS, 2004) show a further deterioration even under the most optimistic scenario, in fact a larger one than the latest estimates by staff show.

Figure 5.Poland: Projected Balance of the State-Managed Part of the Old-Age Pension System Relative to GDP

Sources: IMF staff projections and calculations, Polish authorities, and Chłon, A., M. Góra, and M. Rutkowski, “Shaping pension reform in Poland: Security through diversity”, Social Protection Discussion Paper Series No. 9923, The World Bank, August 1999.

D. Pension Savings in the Privately-Managed Part of the Mandatory Pension System

56. As new generations entered the labor market, the number of members has increased significantly since the introduction of the new system in 1999. The number of participants reached 11.3 million in January 2005. Reflecting the low level of employment and the exclusion of agricultural labor from the pension reform, this is only about 44 percent of the potential contributor base.33

57. Although the investment performance of the OFEs has been impressive since 1999, accumulated pension savings in this part of the pension system remained moderate. The average annual real rate of return on OFEs assets in 2000–04 was over 8 percent. Nonetheless, after 6 years, the average accumulated funds per member reached only about Z1 5,500, or about 2¼ times the monthly average gross wage in the enterprise sector in January 2005. Owing to administrative inefficiencies, ZUS, the collecting agency accumulated large arrears to the OFEs in the early years of the new system. ZUS now has an agreement to pay these arrears, which with interest amounted to some Zl 10 billion (1.2 percent of GDP) in mid-2003.34 Taking this into account, the accumulated assets per member would be about 2.6 months of salary. Reflecting a continuous decline since 2001, the average contribution base per member dropped from about 77 percent of the average economy-wide wage in 2000–01 to 50 percent in 2004. (Figure 6). Declining economy–wide employment and an increasing share of young contributors—who have lower-than-average income and a higher-than-average unemployment rate—are likely to be the main factors explaining this development.

Figure 6.Poland: Average Contribution Base of OFEs Relative to Economy-Wide Average Wage

Sources: IMF Staff calculations; Polish authorities.

58. High costs of running the privately-managed part of the mandatory pension system lowered asset accumulation. Distribution fees charged by fund managers reached 8.5 percent of contributions in the first two years, and though the rate has been declining since then, it was still above 6 percent in 2003. In addition, the management fee was 0.6 percent of total assets in 2003. With fees included, the average annual real internal rate of return of the OFEs in 1999–2003 was about 3 percent (KNUiFE, 2003).35

59. The savings accumulated in the OFEs are financing mostly the public debt. Close to ⅔ of the OFEs portfolio is invested into Polish government securities, leaving about ⅓ to the private sector, mostly in form of equity capital (Figure 7). The foreign investment of the OFEs was marginal, some 2.2 percent of the total in January 2005, considerably below the regulatory limit of 5 percent of total assets.

Figure 7.Poland: The Portfolio Structure of Second-Pillar Private Pension Funds (end-January 2005, in percent)

Sources: IMF Staff calculations; Polish authorities.

E. The Fiscal Implications of the 1999 Pension Reform

60. The asset accumulation in the OFEs resulted in a matching increase in public debt. The cyclically adjusted balance of the general government including the contributions to the OFEs has deteriorated since 1999. Moreover, the non-pension structural balance of the general government—which excludes the impact of the deterioration of the (old) state-managed old-age pension system since 1999—has also deteriorated since 1999 (Table 1). This suggests that the cost of the pension reform (the diversion of part of the pension contributions to the OFEs) has not been provided for, although a firm conclusion on this matter would require a clearly-specified counter factual to the actual outcome.

Table 1.Poland: The Fiscal Impact of the 1999 Pension Reform(In percent of GDP)
Headline fiscal balances
Overall balance (excluding contributions to OFEs)-2.4-3.0-3.4-5.3-6.3-5.9-6.1
Contributions transferred to the OFEs0.
Balance including contributions to OFEs1/-2.7-2.5-4.2-5.1-4.7-4.9
Balance excluding receipts to and payments from the the state-managed part of the pension system-1.2-1.6-3.2-3.7-3.3-3.5
Structural fiscal balances
Overall balance-4.5-4.4-4.2-4.9-4.9-4.7-5.4
Balance including contributions to OFEs1/-3.9-3.1-3.9-4.0-3.7-4.3
Non-pension balance-1.8-2.0-2.9-2.9-2.6-3.1
Privately-managed pension funds (OFEs)
Accumulated assets0.
Financial balance0.
Sources: The authorities and staff estimates.

F. Conclusions

61. The 1999 pension reform was a major effort toward restoring the long-term stability of public finances, but less-favorable-than-envisaged developments since then resulted in less progress than targeted.

  • Owing to a decline in employment, the coverage of the pension system has declined significantly since 1999. This has also lowered the asset accumulation in the privately-managed part.

  • Declining employment and a faster-than-previously envisaged increase in the real value of pensions have resulted in a weakening in the long-term financial position of the (old) state-managed old-age pension system since 1999. Recent official projections show that this part will be in deficit for a considerably longer period than expected at the time of the pension reform, and the deficits during this period will be significantly larger.

62. The analysis point to several fiscal risks implicit in the present pension arrangement.

  • Because the coverage of the mandatory pension system has declined, future claims on budget to provide support to those who will have low effective replacement ratios (or no pensions at all) are a risk. The implicit contribution base (calculated based on the collected contributions and the statutory contribution rates) was only about 26 percent of GDP in 2004, and it has been declining since 2001. As the longevity risk is now fully assumed by beneficiaries, a faster-than-assumed increase in remaining life expectancy, which will reduce the effective replacement ratio, may aggravate this problem.

  • The reform of the pension system for farmers (KRUS) has been postponed. KRUS, which is about 90 percent subsidized, provides an average pension of Zl 650, some 78 percent of the minimum wage. While a major burden for the budget because of the very small contributions participants pay, it provides only a modest income for its beneficiaries.

63. These developments suggest a need for reviewing the pension system with a view to identifying measures to strengthen its long-term financial position and reduce the risk of old-age poverty. The most effective way of addressing these problems would be to promote employment, but there are other measures to consider, too, including:

  • Broadening the base for social security contributions, in particular for self-employed as envisaged in the Hausner plan; and

  • promoting higher voluntary private pension savings.

The authorities’ plan to set up a National Actuary Office to continuously monitor the pension system would also be an important step to increase public awareness and promote early actions in these areas.


    Chłon, A., M.Góra, and M.Rutkowski (1999), Shaping pension reform in Poland: Security through diversity, Social Protection Discussion Paper Series No. 9923, (Washington: The World Bank, August1999).

    Góra, M. (2003), Reintroducing Intergenerational Equilibrium: Key Concepts behind the New Polish Pension System, William Davidson Institute WP No. 574, June2003.

    Góra, M. and M.Rutkowski (1998), The Quest for Pension reform: Poland’s Security Through Diversity, Office of the Government Plenipotentiary for Social Security reform, Warsaw, October1998.

    KNUiFE (2003), Insurance and Pension Funds Yearbook 2003, Insurance and Pension Funds Supervisory Commission, Warsaw 2003.

    World Bank PEIR (2002), Towards a Fiscal Framework for Growth, A Public Expenditure and Institutional Review, January21, 2003 (Washington: The World Bank).

    ZUS (2003), Prognoza wpłwów I wydatków Funduszu Emerytalnego do 2050 roku, ZUS Departament Statystyki, Warszawa, Listopad2003.

    ZUS (2004), Prognoza wpłwów I wydatków Funduszu Ubezpieczeń Społecznych na Lata 2006–10, ZUS Departament Statystyki, Warszawa, Grudzień2004.

Prepared by István P. Székely.

Sections A and B draw on Góra and Rutkowski (1998), Chłon, Góra, and Rutkowski (1999).

The demographic dependency ratio was below 0.2 in 1999.

The system dependency ratio is the ratio the ratio of pension beneficiaries to contributors. In 1999 it was about 0.5, 2½ times the demographic dependency ratio.

The replacement ratio is the ratio between the last salary and the first pension.

There is a floor to the real notional return set at zero percent.

As part of the Hausner Plan, the indexation rule was changed in 2004. Previously, pensions were indexed to the change in the CPI index plus at least one fifth of the real wage growth in the economy. There is, however, a new law submitted to parliament that would change the indexation rule again, making it more generous.

Originally, the pension reform law envisaged an annual transfer of one percent of the contribution base to this fund. The size of the transfer, however, was reduced in 2001 (before the demographic reserve fund started to function): to 0.1 percent of the contribution base in 2002-03; 0.15 percent in 2004; 0.2 percent in 2005; 0.25 percent in 2006; 0.3 percent in 2007; 0.35 percent in 2008, and to nil afterwards.

This part is not yet fully legislated. The first payment from the funds will be made in 2009.

The guarantee is for the sum of the pension rights from both parts combined, topping up the accumulated rights to a guaranteed minimum. The cost of this guarantee is covered by the budget.

The actual replacement ratio depends to a large extent on market returns and the conditions of the annuity, which, especially the latter, are almost impossible to predict at this stage.

The difference in replacement ratios between male and female participants is due to the difference between mandatory retirement ages.

Assumed average growth was 3.5 percent, slightly below staff’s current medium-term growth assumption in the no-reform scenario. Pension rules were assumed to remain unchanged but the contribution rate to the state-managed part was assumed to be reduced from 12.22 (19.52 less the contribution to the privately-managed part of 7.3) to 7.3 percent between 2013 and 2030, as was envisaged in the reform. The labor force participation rate was assumed to increase by some 20 percent.

Using in both calculations the current 10-year rate of 5 percent as a discount rate.

Until there are participants and beneficiaries of the old system, the state (ZUS) will manages two systems, the old one and the new one. The calculations in this paper refer to the combined costs of the two systems. The negative impact of the unfavorable factors discussed in this part concern the old system. The transparency of the pension system would be enhanced by producing and releasing separate accounts for the two systems.

Projections made in 1999 assumed that pensions would be increased annually by the rate of CPI inflation plus one fifth of the real wage growth.

Assuming a 2.5 percent annual real wage growth until 2030 (declining to 2 percent afterwards), a large decline in the rate of unemployment, from over 20 percent in 2003 to 4.5 percent by 2050, and a one-percentage point improvement in the collection ratio, from 97.5 percent in 2003 to 98.5 percent by 2004.

Women of 16-60 years and man of 16-65 years.

Based on a law passed on July 23, 2003.

This compares to an average annual real interest rate on newly issued government debt of above 6½ percent (on 2Y T-Bonds) during the same period, which was the real cost to the tax payers to finance the additional deficit the creation of the privately-managed mandatory part of the pensions system resulted in.

Other Resources Citing This Publication