Poland, Hungary, retirement, Czech Republic, family benefits, women and men, gender equality, child care, income, equal treatment, retirement age, CEE countries, equal, gender inequality, household income, industrial employment, child care benefits, women, working, family members, retirement ages, social security benefits, child allowance, Child raising, cicka, national benefits, family benefit, gender dimensions, Central European women, social security, pension credit, gender issues, life expectancy tables, private pension funds, social security system, gender issue, Central European countries, maternity benefits, EU member states, social context, pension schemes, redistribution, gender pay gap
Chapter 1 The Gender Dimensions of Social Security Reform in the Czech Republic, Hungary, and Poland Elaine Fultz and Silke Steinhilber 1. Introduction The social security reforms that the new governments in the Czech Republic, Hungary, and Poland adopted in the 1990s affect women and men in quite different ways. The differential impact depends in large part on the particular benefits that were altered, and thus the domains of social and professional life touched by the reforms: work, unemployment, child bearing, parenthood, sickness, disability, or retirement. It is quite possible that different observers concerned with gender equality will assess these reforms differently, depending on their own views of what constitutes equal treatment, their priorities for addressing unequal conditions, and their notions of the shared rights and obligations of members of society. Without abandoning our own preferences, we have tried in what follows to provide an objective account of the gender dimensions of reform in the three countries. The study analyzes two broad categories of benefits that address distinct areas of experience and need: on the one hand, a set of family benefits that supports parents with children (family allowances, child care benefits, and maternity benefits) and, on the other, one that replaces lost income as a result of old age or death (retirement and survivors' pensions). Although support in each case facilitates a retreat from the work force, the first may anticipate return while the second typically does not. More relevantly for this study, the first category includes a mix of benefits, some of which are shaped by biological differences (i.e., maternity benefits) and others that are provided without regard to these (family allowances and child care benefits), while both benefits in the second category address contingencies experienced by women and men. Since the economic and Social Context
shapes the need for benefits, the profiles of those who receive them, and their costs, this context is the starting point of analysis in each of the three countries.1 The country studies 1 See sections 2.1 in Chapters 2, 3, and 4.
2 show that women achieved a high level of work force participation under socialism, but the contexts in which they worked were characterized by significant gender inequality. This inequality persisted into the 1990s, with some indicators worsening or becoming more obvious, but on the whole falling within the range observed in Western Europe
.2 unemployment rates
were persistently higher for women than for men in two of the three countries, the Czech Republic and Poland, throughout the 1990s, a disparity that exists in all EU Member States
.3 The gender pay gap was, and remains, in the range of 20-25 percent in all three countries, with the Czech Republic showing the greatest gap. These figures are comparable to those in the EU member states with the greatest pay inequality.4 Gender segregation in employment is strong, with women being concentrated in services. Compared to the EU member states in 2000, however, the share of women in industrial employment is higher in all three countries.5 Women are underrepresented in managerial jobs in all three countries, although in Hungary and Poland they hold larger fractions of such positions than women in all EU member states with the exception of France.6 At the beginning of the 1990s, all three countries provided the benefits examined in this study on an unequal basis to women and men. Retirement age preferences existed for women, family benefits provided women with greater support for balancing their lives as workers and mothers, and, in two of the three countries, women had greater survivors' benefits (the Czech Republic and Hungary). While gender equality was part of the communist ideology, benefits were not aimed at achieving equal treatment of women and men. Rather, in an environment of substantial gender inequality, the benefits rewarded women for motherhood and eased their dual role of worker and principal family care giver. 2 See EU DG Employment and Social Affairs, Unit Equality for Women and Men, Thematical statistical sheets (based on Eurostat, Key Employment Indicators), http://europa.eu.int/comm/employment_social/ equ_opp/statistics_en.html. 3 In Hungary, women's lower unemployment rate is explained by a larger exodus of women from the work force in the early 1990s. See Lukбcs and Frey, section 2.1, this volume. 4 These are Austria, the Netherlands, United Kingdom, and Ireland. See Eurostat "Gender pay gap in unadjusted form - Average gross hourly earnings of females as a percentage of average gross hourly earnings of males" 2003, http://europa.eu.int/comm/eurostat/Public/datashop/printcatalogue/EN?catalogue=Eurostat&product=1-em030-EN. 5 The highest share of women in industry is found in the Czech Republic, 27.3 %. In Hungary, it is 24.8%, and in Poland 18.9 %. In the EU, the average share of women in industrial employment is 14%. Source, see footnote 2. 6 See UN/ Economic Commission for Europe, "Women and Men in Europe and North America
", New York and Geneva 2000.
3 In the period of rapid change following 1989, eliminating these Gender differences
was not among the highest priorities for reform. In these early years, gender equality was eclipsed by other concerns that were seen as more pressing, namely, the need for new benefits to protect workers and families against high inflation, job loss, and poverty.7 The new governments' capacity to deliver such relief was an immediate test of their legitimacy, and they moved quickly, adopting measures to reconfigure existing social security schemes to meet these needs. Criteria for early retirement were liberalized for both women and men, pension benefits were improved in various ways, and new family benefits were established in an effort to compensate for inflation and the removal of subsidies on basic commodities. As their economies gained a measure of stability in the mid- to late 1990s, all three governments set out to restructure social security more fundamentally. Even in this less extraordinary context, however, gender issues were not a major driving force in reform. This is attributable in part to the absence of a well-organized gender lobby in any country, to the divergent goals of the women's organizations that did exist, to the complexity of social security issues, and to the continuing existence of problems in labour market
s that captured greater attention.8 The reforms of the later 1990s rather aimed at reshaping earlier emergency measures, containing scheme costs arising from them, and reflecting the new relationship between citizen and state. Family benefits were redirected toward the largest families and those in greatest need. Some family benefits were cut, either directly or by failure to provide regular cost-of-living adjustments. Pensions were made more individualized and earnings-related, although to a different extent and through different means within each country. In all three countries, retirement ages were increased.9 Although gender received little explicit attention, the new laws themselves raise major questions with respect to equal treatment. Among 7 See sections 2.1 and 2.2 in Chapters 2, 3, and 4. 8 Steinhilber identifies three distinctive sets of goals adhered to by women interviewed for this study: formal gender equality in social security, affirmative action to compensate for disadvantages in labour markets, and unequal treatment which recognizes women as partners in a family unit. Section 5.3, this volume. 9 In Poland, early retirement was eliminated prospectively (2006). This will lead to an increase in the actual retirement age.
4 them, five stand out. Formulated below, these will be dealt with in the subsequent analysis. · The channel through which family benefits for child care is made available is of importance in relation to women's attachment to the labour force
. In national contexts where child care is still overwhelmingly provided by women, making such benefits available through the work place or contingent on employment creates incentives and rewards for labour force participation
. Conversely, if child care benefits are restricted to those with low income or limited means, mothers who stay at home to care for young children
may become isolated from the world of work and find their integration or subsequent reintegration into the labour market in later years more difficult.10 · Second, a basic issue arises as to what policies permit or promote more equal sharing of family caring responsibilities
between women and men. In the course of the 1990s, all three countries made child care benefits available to men and women on an equal basis. As yet, equal treatment in this regard has had no discernible effect on the allocation of child care responsibilities within the family, which continue to rest overwhelmingly with women. As populations age, elderly family members will pose increased requirements for care, while the need for family child care will be ongoing. It seems clear that women as a group cannot, as a matter of simple arithmetic, gain greater freedom to seek more diverse forms of income and fulfillment until men and women share the tasks of caring for family members more equally. · Third, women have entered the new political era with privileges from the socialist years with respect to retirement. In some cases, the retirement age has been set by law to give all women the possibility to retire at an earlier age than men. In others, it is determined by the number of children a woman has raised. Clearly these privileges afford women and men unequal treatment in an area that does not relate to their biological differences. Is there a principled rationale for continuing them in the current era? 10 In addition, some women will be ineligible due to income or resources that exceed the limit.
5 · Fourth, the 1990s witnessed a major regional trend in the direction of linking the magnitude of an individual's pension benefit more closely to his or her earnings and work history; in other words, in the direction of eliminating income redistribution toward low-income workers. This trend appears in both the public pension schemes and the new privatized ones. It is advantageous to women and men with higher incomes, and hurts all workers, women or men, with lower ones. Given that women earn significantly less than men during their professional lives and tend to work fewer years (both as a result of more time taken for bearing and raising children and more likely unemployment throughout much of Central and Eastern Europe
), this trend affects them more negatively. · Fifth, the partial privatization of pension schemes, such as took place in Hungary and Poland, raises a major question concerning the size of men's and women's pensions as a result of their different average life expectancies. This issue arises in the new systems of commercially managed individual savings accounts mandated by privatization laws. When the amounts accumulated in these accounts are converted to annuities at retirement, it will be necessary to consult one of two types of life expectancy tables: either one that treats women and men separately or a second that gives a single, joint life expectancy projection for both. The use of the latter will result in the payment of an equal monthly benefit for a man and a woman who retire at the same age with the same history of contributions. However, since women on average live longer than men, they will, on average, accumulate higher total lifetime benefits than comparable men. On the other hand, the use of gender-specific tables will give a man and a woman retiring at the same age with the same histories of contributions the same total lifetime accumulations of retirement benefits but will give the woman a lower monthly benefit because her savings must, on average, be stretched to cover a longer lifetime. What should we take as equal treatment for men and women with identical histories of earnings retiring at the same age? Identical total average accumulation over the entire period of retirement? Or identical average monthly benefits?
6 2. Family benefits: Overview of the Reforms Family benefits are payments made in cash or in kind to support parents in bearing and raising children.11 They include general allowances, paid to supplement wages according to the number of children in a family, as well as two benefits that are contingent on specific actions: giving birth (maternity benefits) and withdrawing temporarily from the work force to care for a young child (child care benefits).12 Depending on program rules and the environments in which they operate, these benefits can have a different impact on men and women collectively across society. During the socialist period, all three countries developed an extensive array of family benefits. The country studies examine the most relevant among these in shaping the life and professional choices of women and men.13 In contrast with pension reform, where major changes were adopted within short time periods in many CEE countries, the reform of family benefits has been more incremental and continuous. The sequence is simplest in the Czech Republic, where major reforms were implemented in 1995 and 1996.14 In Hungary and Poland, by contrast, changes were made throughout the 1990s and continue today. Amidst this complexity, several broad trends stand out. First, in the early 1990s, there was a decoupling of family benefits from employment status.15 The new governments in all three countries used various family benefits, along with other forms of social security, to provide 11 In Hungary, family benefits are also paid through the tax system. 12 ILO, Introduction to Social Security, Chapter 11, "Family Benefit," Geneva, Switzerland, 1984. The actual names of the benefits that fit the above definition of child care vary in the three countries. In the Czech Republic, it is the parental allowance; in Hungary, the child care fee, child care allowance and child raising allowance; and in Poland, the child care benefit and child raising allowance. See Table 1. 13 The Czech study examines four such benefits (the maternity benefit, child allowance, social allowance and parental allowance); the Hungarian study examines seven (the maternity benefit, family allowance, child protection benefits, tax credit
s, the child care fee, the child care allowance, and child raising benefit); and the Polish study, seven (the maternity benefit, child care benefit, the child-raising allowance, the family allowance, the alimony fund, the guaranteed periodic benefit, and benefits for pregnant women
and women raising children). 14 Other less comprehensive reforms were enacted in 1990 and 1993 as well. 15 Only maternity benefits were exempted from this process. Today they continue to be provided as employment-related insurance benefits in all three countries.
7 relief from the effects of inflation, job loss, and poverty.16 This was accomplished by converting some employment-related benefits to universal ones while targeting others to families with low incomes or limited resources. Some new benefits were also created. Later in the 1990s, fiscal pressures provided a major impetus for change. In Hungary and Poland, the governments adopted austerity programs that called for major cuts in overall social spending, including family benefits.17 These were achieved through greater income testing of benefits, cuts in their level and duration, and omitting adjustments for inflation. The shift to income testing was most marked in Poland, where today eligibility for nearly all benefits is based on the income or means of the family.18 In Hungary, income testing (of the examined benefits) was adopted in the mid-1990s and then abolished later in the decade.19 Income testing was less marked in the Czech Republic, where it was applied to some but not all benefits and in a manner that was not highly restrictive.20 A third pattern across the countries is provision of greater support for large families. In Hungary and Poland, these changes were part of broader efforts to redress stagnant or declining population growth
. They were supported by certain religious groups
and political parties favoring the restoration of traditional family roles or national values. In Hungary, a new child-raising benefit was created for families with three or more children; and progressively larger tax credits were provided for the second, third, and subsequent children.21 In Poland, larger families received a higher childraising allowance for their third and each subsequent child.22 In the Czech Republic, although population growth was less ardently promoted, there too larger families received larger child allowances.23 16 Sections 2.2.1 of Chapters 2, 3, and 4, this volume. 17 Wуycicka et al., sections 2.2.1 and 2.2.2(a), and Lukбcs and Frey, section 2.2.1, this volume. 18 The child care benefit, a short-term payment for those who leave work to care for a sick child, remains employment related. Wуycicka et al., section 2.2.5, this volume. 19 Even when income testing was applied, it was rather loose, excluding only nine percent of families and seven percent of children from the family benefit. Lukбcs and Frey, Table 14, this volume. 20 94 percent of two-parent families with one economically active parent received the child allowance before it was income-tested (1996), and 93 percent received it afterward (1999). Kucharovб et al., Table 2, this volume. 21 Lukбcs and Frey, Boxes 2 and 3, this volume. 22 Wуycicka et al, Box 1, this volume. 23 Kucharovб et al., section 2.2.3, this volume.
Finally, all of the countries equalized benefits for men and women providing child care, thus eliminating provisions of preexisting law that had barred fathers from using these benefits or imposed stricter rules on them. In one country, Poland, equalization of one child care benefit was coupled with a cut in the rate of the benefit.24 Given the gender wage gap, this was judged necessary to avoid an increase in spending outlays, since the benefit level was based on the worker's pay. See Table 1.
Table 1. Equal treatment of men and women with respect to
various national benefits for child care, benefit and year equalized
Right to return to same job
Child care fee
Child care allowance
Child raising benefit
Child care benefit
Child raising allowance 1996
Source: Kucharovб et al., Lukбcs and Frey, and Wуycicka et al.,
sections 2.2, this volume.
These varied changes in family benefits contrast with the limited change in maternity benefits. All three countries maintained these benefits as employment-related social insurance. Hungary, making a single change as part of a 1996 fiscal austerity plan, reduced the replacement rate of wages from 100 to 70 percent.25 In 2000, the Polish government initiated a series of rapid changes in the duration of the benefit but ultimately restored its original duration of 16 weeks.26 Throughout, Poland maintained an income replacement rate at 100 percent of the worker's current wage. Poland extended maternity benefits to fathers, allowing them to take two weeks of the total period available. The Czech Republic made a single small change, shifting the basis for computation of benefits from gross to net wages.27 See Table 2. 24 This is the benefit for care of a sick child. Wуycicka et al, section 2.2.2(b), this volume. 25 Lukбcs and Frey, Box 1, this volume. 26 Poland first introduced an increase in the duration of benefits from 16 to 20 and then 26 weeks, followed by a cutback to 16 weeks. Wуycicka et al., section 2.2.2(a), this volume. 27 In the Czech Republic, the percentage is calculated from the so-called Daily Assessment Base which is based on wages but entails a complex formula that can cause some variations from actual wage amounts.
Table 2. Reform of maternity benefits
Reform and year
Social insurance benefit Unchanged
28 weeks' duration
90% of previous net wage 1993, changed to 69% of
previous gross wage
Hungary Social insurance benefit Unchanged
28 weeks' duration
100% of previous wage
1996, replacement rate
reduced to 70 %
Social insurance benefit Unchanged
16 weeks' duration
2000, increased to 20 weeks
2001, increased to 26 weeks
2002, reduced to 16 weeks for
first child (2 available to
father), 18 weeks for
100% of previous wage
Source: Kucharovб et al., Lukбcs and Frey, and Wуycicka et al., sections
2.2, this volume.
3. Impact of the Reforms
At the end of the 1990s we can discern four broad effects of the modifications previously outlined. First is a sharp decline in spending on family benefits as a portion of GDP in Hungary and Poland. While the definitions of benefits and years for which data are available differ, a trend is nevertheless clear: In Hungary, in 1990-2000, spending on the benefits under discussion fell in relation to GDP by nearly one-half, from 3.8 percent to 2.0 percent, and in Poland in 1990-98, from 1.7 percent of GDP to 1.1 percent.28 See Table 3 and Figure 1. As a result, family benefits as a component of household income
also fell sharply: in Poland (1993-99), for
28 Lukбcs and Frey, Table 10 and Wуycicka et al., Figure 2, this volume.
all households, from 3.8 percent to 1.4 percent and in Hungary (1993-99), for a family with two children, from 17.0 percent to 10.8 percent.29
Table 3. GDP per capita, family benefits total and as a
percentage of GDP, Hungary, 1990-2000 (1990 HUF)
GDP per capita Total expenditure Family benefits
as a % of GDP
Source: CSO Statistical Yearbooks and Statistical Yearbooks of the
National Health Insurance Fund Administration, in Lukбcs and Frey,
Table 11, this volume.
It is noteworthy that, in Poland and Hungary during most of the 1990s, the share of family benefits in household income declined significantly while the share of social security benefits overall remained relatively stable.30 This suggests that as a category families were less well protected than other beneficiary groups, for example, old-age and disability pensioners.
29 Lukбcs and Frey, Table 17, and Wуycicka et al., Table 1, this volume The figures for Poland represent only the family allowance, child care fee, and child-raising allowance, as these are the only benefits for which household income statistics are available. 30 In Poland, total social benefits ranged between 31 and 33 percent of household income during 1992-99, whereas in Hungary, social benefits dropped temporarily during the mid 1990s and were then restored. For a Hungarian family with one child, for example, social benefits were 21 percent of household income in 1993, 16 percent in 1997 and 20 percent in 1999, whereas the three family benefits dropped from 9.2 to 5.7 percent over this period. Wуycicka et al., Table 1, and Lukбcs and Frey, Table 17, this volume.
11 Figure 1. Family benefits as percentage of GDP. Poland 1990-98
Source: Hagemejer, K. Liwiski, J; and Wуycicka, I., Poland: Social Protection in Transition (ILO:2002) in Wуycicka et al., Table 3, this volume.
These losses contrast with the small gains posted in the Czech Republic. Here, in the postreform period (1996-2000) family benefits as a percentage of GDP increased modestly from 1.78 percent of GDP to 1.85. See Table 4. This increase resulted primarily from higher unemployment beginning in 1997 and a consequent increase in demand for the newly income-tested family benefits. In the Czech Republic, different levels of aggregation of family income data by household size complicate comparisons with the other two countries. What is clear, however, is that after the reforms, Czech benefits increased slightly as a fraction of family income in households of all sizes but somewhat more in larger ones.31 This upward drift contrasts with the sharp decline in Hungary and Poland.
Table 4. Three family benefits, total spending and spending as a
percentage of GDP,
the Czech Republic, 1996-2000 (CZK, thousands)
1996 1997 1998 1999 2000
12,194 12,495 11,493 12,474 12,748
6,244 6,224 6,273 6,251 6,199
7,357 7,612 7,780 7,718 7,691
31 Kucharovб et al., Table 6, this volume.
Total of three benefits
25,795 26,331 25,546 26,443 26,638
Gross domestic product 1,447,70 1,432,80 1,401,30 1,390,60 1,433,80
GDP accounted for by the 1.78% 1.83% 1.82% 1.90% 1.85%
Source: Kucharovб et al., Appendix Table 9, this volume.
In the second place, the reforms concentrated family benefits on families with low incomes. After Czech reforms (1995-96), the fraction of families receiving all three of the benefits examined in this study fell by onehalf, while at the same time, those that remained eligible for all three received increases equal to three to ten percent of net family income.32 This concentration probably accounts in part for the fact that, in Poland at the end of the decade (1998), family benefits constituted 12 percent of income for families in the lowest income decile but just 0.15 percent for those in the top decile.33 In Hungary, as previously mentioned, income testing was adopted and then discontinued.
The authors of all three studies express concern that the increased scope of income testing of child care benefits may reduce the incentives for women to engage in economic activity, especially women with low skills.34 In Hungary and Poland, they raise similar concern about the relatively small gap between child care benefits and the minimum wage
.35 At the moment, however, we cannot quantify the strength of these incentives in the professional and life choices made by family caregivers.
Third, as previously described, all three countries gave equal access to child care benefits to women and men. Access alone, of course, cannot be expected to change entrenched modes of distributing child care between parents. While the data on use of these benefits is not complete, what does exist suggests that there has been no pent-up demand for access to child care 32 Kucharovб et al., Table 6, this volume. 33 Wуycicka et al., Table 3, this volume. 34 The statement actually refers to the "child raising benefit" in Poland. There "child care benefits" refer to short-term benefits for the care of a sick child, while the "child raising benefit" is paid for caring for young children at home for extended periods. In Poland, means testing was applied to the child raising benefit but not the child care benefit. Wуycicka et al., footnote 13. 35 In Poland, the guaranteed benefit is 88 percent of the net minimum wage, whereas in Hungary the combination of the child care allowance and child-raising benefit equaled 84 percent of the minimum wage until 2001, when it was increased by 25 percent (from HUF 40,000 to HUF 50,000). Wуycicka et al., Section 2.2.2(f)(i), and Lukбcs and Frey, section 2.2.4; this volume.
13 benefits on the part of fathers. In the Czech Republic during 1995-98, fathers constituted less than one percent of child care beneficiaries, roughly the same as in Hungary.36 In Poland, statistics on the use of child care benefits by men and women are not available. However, our authors were unable to find any evidence, even anecdotal, that men are using these benefits to stay home and care for their children. Clearly, the main barriers to greater sharing of family responsibilities lie in inequalities in the labour market (the gender wage gap) and in cultural values.37 Nevertheless, the equal treatment by these countries removes barriers to possible future shifts toward more equal sharing of responsibilities for child care between parents. A final pattern worthy noting is anecdotal evidence reported in all three studies that women are increasingly reluctant to use employment-based rights to family benefits for fear of reprisals by employers in the form of job loss and unfavorable re-assignment.38 There are, however, no estimates available on the scope of this phenomenon in any of the countries. From the perspective of gender equality in employment, these observations are a serious cause for concern. To sum up, the reforms of family benefits have achieved mixed results. On the positive side, income-testing of some family benefits has succeeded in targeting scarce resources to those most in need. This has softened some of the financial shocks associated with the shift toward a market economy. Given that women have lower average earnings and higher rates of poverty, this targeting has no doubt benefited them accordingly. On the other hand, this change has shifted the nature of the support from wage replacement to poverty alleviation
; and it has shifted the status of beneficiaries, mostly women, from holders of personal rights to petitioners of the state. At the same time, income testing provides a disincentive to economic activity. This threatens to capture some women in a trap of dependency on these benefits. While all three countries have extended equal treatment to men and women with respect to child care benefits, these reforms have not as yet been associated with any behavioral changes. Rather, the care of young children continues to be provided overwhelmingly by women, as before the reforms. In two of the three countries (Hungary and 36 Kucharovб et al., section 2.2.3(b), and Lukбcs and Frey, Table 18, this volume. 37 Steinhilber, section 5.2, this volume. 38 Lukбcs and Frey, Section 2.2.4; Kucharovб et al., sections 2.2.3b & 2.2.5; and Wуycicka et al., section 2.2.4; this volume.
14 Poland), large cuts in family benefits leave working parents, mostly women, with considerably less support for efforts to balance family and professional responsibilities than they received before transition. 4. Pensions: Overview of the Reforms In 1989, the national pension schemes of Czechoslovakia, Hungary, and Poland had several features in common. With no private supplemental arrangements in place, all served as the main providers of retirement income in their respective countries. All were financed on a pay-as-you-go basis through transfers from state-owned firms to a social insurance account within the state budget. There was little public accounting of the collection or allocation of these resources. Benefits were computed on the basis of a worker's final earnings, thus penalizing workers with steady earnings compared to those whose earnings rose as their careers advanced.39 Retirement ages were relatively low: in Czechoslovakia, 60 for men and 5357 for women; in Hungary, 60 for men and 55 for women; and in Poland, 59 for men and 55 for women.40 Two countries, Czechoslovakia and Poland, provided pension privileges i.e., higher benefit amounts and/or lower retirement ages for certain occupations, such as miners, pilots, and certain high-level members of the Communist Party. In the later socialist years, adjustments of pensions did not keep pace with increases in prices or nominal wages, to the economic detriment of older pensioners in comparison with those more recently retired. In the early 1990s, rising unemployment caused a drop in contribution income in all three schemes. This was largest in Poland and Hungary, where the number of contributors declined by 15 and 25 percent respectively. In the Czech Republic, the number of contributors dropped by only eight percent.41 The Hungarian and Polish governments liberalized early retirement in order to absorb excess unemployment, causing rapid increases in the number of 39 In Hungary, the best three out of five years before retirement were counted, with a rising replacement rate for those with longer careers. In Poland, the best two consecutive years out of the last ten were counted. In the Czech Republic, the benefit was based on the average wage earned in the five years immediately prior to retirement. Given the relatively compressed wage structures of these countries under socialism, the use of final earnings did not produce great differences in pensions. 40 For Poland, these are actual ages, based on early retirement options. The regular retirement age was 65 for men and 60 for women. Wуycicka et al., section 2.3.2, this volume. 41 Palacios, R., M. Rutkowski, and X. Yu, Pension Reform in Transition Countries, World Bank
, Washington, D.C.
, June 1999.
15 pensioners in the range of 20 percent in Hungary and 50 percent in Poland. In the Czech Republic, the increase was about 10 percent
.42 In addition, all three governments took early action to improve pension benefits. Czechoslovakia and Poland took aim at the inequalities between old and new pensioners with a step-up in benefits. This adjustment was greatest in Poland, where it contributed to a sharp rise in pension spending.43 In addition, all three governments regularized cost-of-living adjustments. Both Hungary and the Czech Republic established schemes of voluntary supplemental pension savings (1993-94). In the early period, Czechoslovakia moved decisively to eliminate pension privileges (1991). As the decade progressed, the three countries enacted major pension reforms.44 In general, these reforms aimed at containing scheme costs and at individualizing benefits. The Hungarian and Polish governments revised pension formulas so that benefit levels more closely reflected each individual's record of earnings, thus eliminating redistribution toward lowincome workers. The Czech Republic, by contrast, adopted a new formula which continued to include redistributive terms. All three countries adopted laws that would increase the retirement age gradually. They also increased the number of years of earnings counted in computing pensions. This reduced the penalties for those with sustained flat wages and increased the rewards for contributions paid early in a career. In an effort to control costs, Hungary and Poland restricted the indexing of wages counted for pension purposes and reduced pension cost-of-living adjustments.45 These same countries revised the pension credits for periods spent out of employment caring for young children, with the effect in both cases of reducing their value. The Czech Republic and Hungary made survivors' benefits for men and women equal.46 In a major policy shift in the late 1990s, Hungary and Poland introduced mandatory systems of commercially managed individual savings 42 Palacios, Rutkowski and Yu, as previously cited. 43 During 1990-96, pension spending rose from 9.6 to 16.1 percent of GDP. Chlon, Agnieszka, "The Polish Pension Reform of 1999," Table 4, p. 106, in Elaine Fultz, editor, Pension Reform in Central and Eastern Europe, Volume 1, Restructuring with Privatization: case studies
of Hungary and Poland, Budapest: ILO-CEET, 2002. 44 Lukбcs and Frey, Wуycicka et al., and Kucharovб et al., sections 2.3.1, this volume. 45 Hungary adopted the so-called Swiss method (benefit adjustments reflecting wage and price increases in equal proportions), while Poland moved to heavier reliance on price indexing (the ratio for adjustments was 85 percent reliance on prices and 15 percent, on wages). 46 No such action was needed in Poland since these were equal prior to 1989.
16 accounts.47 These accounts replaced a portion of the countries' public, payas-you-go schemes and put part of workers' contributions (savings) with private investors. When these schemes are phased in, retiring workers will use their savings to buy an annuity, that is, a pension payable monthly until death. Proponents portrayed these reforms as needed to develop capital markets, boost economic growth, and increase the security of pensions through risk diversification. The Czech Republic debate
d this type of reform, but in view of the high transition costs has so far rejected it. Unlike changes in maternity and family benefits that may be implemented quickly after enactment, pension reforms are usually implemented gradually over a period of years. The greatest time horizon belongs to the so-called radical reforms that replace one type of system with another, as is the case with pension privatization. This time lag makes assessing the reforms far more difficult in the short run. We can, however, make some educated guesses about long-run effects using existing data and macroeconomic simulations. Using these approaches, the following three chapters identify the major gender dimensions of the reforms just described. A comparison of these analyses reveals several broad patterns across the countries. These relate to: 1) the retreat from income redistribution in pension benefit formulas; 2) changes in the retirement age; 3) crediting periods devoted to child care for pension purposes; and 4) the conversion of individual savings to annuities in the new private pension schemes.48 5. Impact of the Reforms a) The retreat from redistribution. All three countries began the 1990s with pension policies that effected substantial redistribution of income from workers with higher earnings to those with lower ones. These schemes provided defined benefits computed on the basis of workers' wages in their final years of employment. The benefits were constrained by upper and lower limits that enhanced their redistributive character. In Hungary, a progressive scale within the pension formula provided further redistribution 47 These reforms are discussed in detail in another volume in this series. Elaine Fultz (ed.), Pension Reform in Central and Eastern Europe: Volume 1, Restructuring with Privatization: Case Studies of Hungary and Poland, Budapest: ILO CEET, 2002. 48 Survivors' benefits are not included here. In two of the three countries (Czech Republic and Hungary) these were equalized for men and women during the 1990s, but these changes were of a secondary importance compared to those discussed here.
17 by weighting low wages more heavily. In the course of the 1990s, Hungary and Poland eliminated these redistributive elements entirely, while the Czech Republic revised its pension formula but retained elements of redistribution. The governments achieved these changes in rather different ways. In 1998, the Hungarian government deleted the progressive term in the formula, thereby benefiting middle and upper income workers and disadvantaging those with low incomes.49 This change is being phased in gradually and will become fully effective in 2013. The Polish government revised the formula twice in the 1990s. In 1992, it adopted a two-part formula that increased the weighting of individual workers' earnings, and at the same time provided some redistribution toward those with low incomes.50 Seven years later, under the weight of large current deficits and unsustainable future
ones resulting from the earlier adjustments, the government reformulated the public scheme to reduce both benefits and redistribution. This was accomplished by the introduction of a new Notional Defined Contribution (NDC) formula that makes an individual worker's pension directly proportional to total lifetime contributions.51 In 1995, the Czech government adopted a new formula with two terms, both of which benefited low-income workers.52 The first consists of a flat amount that is the same for all workers. The second uses the individual worker's earnings and gives greater weight to low wages. At the same time that Hungary and Poland individualized their formulas (1998-99), they adopted new systems of private individual savings. This major policy shift redirected approximately 20 percent of contributions to individual accounts managed by commercial firms. By their nature, these accounts lack any redistribution toward low-income workers; and they offer no benefit promise at retirement. Instead, each worker will receive an annuity based on the investment performance of his or her own savings, 49 Lukбcs and Frey, section 2.3.2, this volume. 50 The latter was achieved through a constant term in the formula equal to 24 percent of the average wage. 51 In the NDC scheme, benefits will be based on each worker's own contributions. The amount of the pension will be calculated by dividing the total lifetime contributions that he/she pays by the average statistical life expectancy of the worker's age cohort at the normal retirement age (gender neutral life tables will be used in this calculation). Thus, benefits will decline automatically in response to increased life expectancy (unless the individual keeps working and delays retirement). Individual accounts will be established to record each worker's contributions. Past contributions will be adjusted at the rate of 75 percent of the real growth of wages which are subject to contributions. This reform applies to all those who were 49 or younger on the date the reform came into force. Others will continue to be covered by the preexisting defined benefit system. Wуycicka et al., Box 7 this volume. 52 Kucharovб et al., section 2.3.1, this volume.
18 minus administrative charges.53 The method for converting savings to an annuity is of great potential importance for equal treatment and is discussed later. The new formulas in the public schemes make no direct reference to sex and are therefore formally neutral with respect to equal treatment: A man and a woman retiring at the same age with identical earnings and contributions will receive the same monthly pension. The Hungarian and Polish reforms will of course leave workers with the lowest lifetime earnings with lower pensions. Since women figure prominently in this group, they will be disadvantaged in a way disproportionate to their numbers in the populations at large. The private component lacks any redistributive effect and will contribute further to their disadvantage. The Polish study illustrates the likely future effect. Its simulation shows that the average pension paid to a woman who retires at age 60 under the old rules is about 82 percent of that paid to a man retiring at the same age.54 After the new NDC and privatized schemes are fully phased in, the average woman retiring at 60 will receive just 74 percent of the average pension paid to a man with the same retirement age.55 The declining female/male benefit ratio is largely the result of eliminating redistribution toward low-income workers and those with shorter work histories. Not all women are losers under the Polish reform, of course; nor are women the only disadvantaged group: men with low earnings will be similarly affected. The greater impact on women results from their lower average earnings and shorter periods of contribution which the reformed pension system will no longer offset. b) Retirement age. In all three countries, socialist retirement policies gave women an advantage. The possibility of retiring earlier than men was one of a number of advantages that encouraged and rewarded motherhood. In Czechoslovakia, women's preference ranged from three to seven years, depending on the number of children they had raised. In Hungary and 53 In Poland, if payments from both the public and private schemes fall below a certain level, the individual can receive a supplement from the state budget that brings his/her pension up to a specified minimum. 54 Wуycicka et al., section 2.3.4, this volume. 55 Specifically, the typical retired woman retiring at age sixty will draw a pension equal to 22.4 percent of the average wage in the economy while her male counterpart will draw one equal to 30.4 percent. Should they both retire at age 65, the woman's pension would equal 29.2 percent of the average wage while the man's would equal 39.6 percent. In both cases the average woman's pension will be just 74 percent of the average man's. Wуycicka et al., section 2.3.4, this volume.
19 Poland, women were afforded a five-year advantage (in Hungary, retirement at age 55 versus 60 for men; in Poland, 60 versus 65).56 In the current environment of pension financing difficulties, shrinking or stagnant fertility rates, and a projected drop in the ratio of active workers to retired persons, achieving equality with respect to retirement age by revising men's retirement age to agree with women's is not a plausible alternative. Practically, equality can be achieved only by eliminating women's preferences. In some sense, movement in the direction of equal treatment is a losing proposition for Central European
women (we will argue below, however, that it should be seriously considered nonetheless). The difficulties associated with this change will probably fall most heavily on those women who, closest to retirement, planned to retire under the old rules but find themselves nearing retirement under the new ones. But it may also fall on the young and the aging as women, working to more advanced ages, are restricted from performing traditional roles as care-givers of aging parents and grandchildren. Reflecting these considerations, only Hungary has so far established equality of retirement ages. The Czech Republic has narrowed the gap between women and men, and Poland has chosen to maintain the preexisting five-year differential.57 See Table 5. 56 Kucharovб et al., Lukбcs and Frey, and Wуycicka et al., sections 2.3.2, this volume. 57 Poland did, however, eliminate early retirement in the 1999 reform, effective in 2007. See Wуycicka et al., section 2.3.2, this volume.
Table 5. Retirement ages in the Czech Republic, Hungary, and
60 for men, 53-57 for 62 for men, 57-61 for
women, being phased
in by 1 January 2007
60 for men, 55 for 62 for both sexes,
being phased in by 1
65 for men, 60 for 65 for men, 60 for
women, with early women, with early
retirement options retirement eliminated
beginning in 2007,
with the exception of
a narrow list of
occupations, to be
specified in future
Source: Kucharovб et al., Lukбcs and Frey, and Wуycicka et al., sections
2.3.2, this volume.
However, it must be recognized that the value of the retirement age preferences that continue to exist in two countries is being eroded through the reforms described in the previous subsection. With the elimination of redistribution toward workers with low lifetime contributions, most women who exercise the option to retire early will receive substantially lower benefits than they would have before these reforms. This effect, too, is illustrated in the Polish simulations. See Table 6. Once the new mixed system is fully implemented, a woman retiring at age 60 with an average female's pension will receive an amount equal to only 57 percent of a man retiring at age 65 with an average male's pension. By delaying retirement until age 65, she would receive a pension equal to 74 percent of his. Given this large difference, the surest way to avoid further erosion of women's retirement protection is to bring the statutory retirement age for women into equality with men's. Without this equalization, the retreat from redistribution will leave women at substantially greater risk of poverty as the
21 reforms are phased in. This risk provides strong justification for equal treatment.
Table 6. Average pension for a woman (as a percentage of the average pension for a man) Poland, simulation for 2050
earlier retirement retirement at age
for women (60,65)
Source: Wуycicka et al., section 2.3.4, this volume.
c) Caring credits. During the socialist era, all three schemes provided credits toward a pension for years that workers (mostly mothers) spent out of employment caring for young children at home. While the rules for counting such periods varied across schemes, a year spent in such status was generally treated as equal to a year of employment, even though no contributions were paid.58 This meant that periods of child care did not reduce the pension that a parent would receive. These credits were financed within the pension system by a cross subsidy from other contributors. During the 1990s, Hungary and Poland revised the rules for crediting such periods, in each case diminishing their value. Hungary retained the old rules in the public component but applied the new rules in the new mandatory private component adopted in 1998. In this case, participants must contribute six percent of their child care benefit to a commercially managed individual savings account. As explained previously, their future pension benefits will be calculated as a simple return on this contribution i.e., investment performance minus management fees.59 As with all other contributions to the privatized component of the pension system, there is no employer matching contribution. Six percent of the child care benefit is a tiny amount, equal to less than US$4.00 per month. This policy is especially
58 The treatment also varied depending on which child care benefit was received by the parent. See Lukбcs and Frey, section 2.3.2, this volume. 59 This private benefit will supplement the individual's public pension. The amount of the public pension will be reduced due to the diversion of a part of the contribution to the private tier.
22 disadvantageous for middle- and upper-income workers, since the pension entitlements that they earn while working are substantially higher than those based on the child care benefit. The Polish government took a different approach by providing a transfer from the state budget to finance pension credits for periods of child care (1999). This improves the transparency of pension financing and shifts the burden of financing pension credits for which no contributions are paid to the public at large. However, the subsidy is based on the minimum wage which makes the benefit much less generous than it was before this reform. As a result, most individuals who take leave from work to provide child care will receive lower pensions. As it is almost exclusively women who take leave and receive child care benefits, it is their earnings history, and consequently their pensions, that will be reduced. This reform also creates disincentives for men, who typically have higher earnings, to take child care leave. d) Life expectancy and private pensions. Under the new mixed pension schemes in Hungary and Poland, part of each worker's monthly pension contribution is redirected to a commercially managed individual savings account. At retirement, the savings accumulated in this account will be converted to an annuity that will pay a monthly pension benefit until the worker's death. The annuity provider will set the level of this monthly benefit based on how long the worker is likely to live, that is, how many years the savings must be stretched to cover. Since it is not possible to know this in advance, the provider will use a statistical estimate of the average life expectancy of all those in the worker's age group. Two quite different methods for making this estimate are possible. The first will give a retired woman the same monthly benefit as it gives a man who retires at the same age with identical savings and investment earnings. This results from the use a joint life expectancy table for both sexes. The second approach will give the woman in the range of 20 percent less, depending on the specifics of the situation. This results from the use of two distinct life expectancy tables, one for women that reflects the fact that, on average, they live longer than men. The simulations in the Polish study show that, using a joint table for both pension pillars, the average woman would receive a pension equal to 74 percent of that of the average pension
23 for a man. Switching to gender-specific tables would reduce her benefit from 74 percent of his to 59 percent.60 What constitutes equal treatment in this situation? Should women and men with comparable histories of earnings accumulate equal benefits over the duration of their retirement? Or should they receive an equal monthly benefit? From a public policy
perspective, the latter approach is preferable for three reasons. First, the use of gender-specific averages would mask the substantial overlap that exists in the actual mortality of individual men and women. In fact, substantial numbers of men live longer than the average female life expectancy; substantial numbers of women die before they reach it; and substantial numbers of men and women live to be nearly the same age. Using group averages to set the pension benefits of individuals whose actual longevity does not match the averages would have a capricious effect, creating many unjustified winners and losers i.e. men who outlive the female average but receive a higher pension nonetheless based on their own sex's shorter average longevity (winners), and women whose longevity falls short of the male average but who receive lower pensions anyway because other women live longer (losers). Second, though the new individual savings schemes of Hungary and Poland are privately managed, they are still part of the pension system and as such have public purposes. The most basic of these is to pool risks across the population so as to provide everyone a minimal level of protection against poverty arising from uncertain longevity. Paying lower benefits to those who live longer would defeat this objective, subjecting them to greater risk of poverty at every stage of their retired lives. Third, women are by no means the only, or even most prominent, group in society with greater average longevity. If we apply group treatment to them, should we not also give smaller monthly benefits to nonsmokers who, on average, outlive smokers; to the more affluent members of society who, on average, outlive the less affluent; to members of racial and ethnic majorities who, on average, outlive members of minority groups; and to those free from genetic vulnerability to life threatening diseases such as cancer, hemophilia, or heart disease
who, on average, outlive less lucky members of society? 60 Wуycicka et al., Tables 13 and 14, this volume.
24 For all three reasons, the preferred principle is that individuals with the same earnings history and the same actual retirement tenure should, in practice, receive the same monthly benefits. This is precisely what the Hungarian law provides. It requires that annuity providers use the same joint life expectancy table in calculating monthly benefits for both women and men. In Poland, by contrast, the issue remains undecided; as yet there has been no legislation that stipulates how mandatory individual savings will be converted to pensions. In 1999, the Government presented a proposal to Parliament allowing the use of genderspecific calculations, and then in the face of heavy criticism there, withdrew it.61 The successor government has not yet addressed the issue. Beyond laws requiring equal treatment, there is need for regulatory structures that support compliance and enforcement of these laws. This need arises from the simple fact that the required use of joint life expectancy tables creates rational incentives for annuity providers to devise ways to attract men and discourage women. After all, the annuity for a woman will, on average, cost the provider more than an annuity for a man, for the simple reason that, on average, women will collect their benefits for longer periods. For providers, averages matter. They might, for example, target men in advertising, dispatch agents to locations where men congregate, or offer gifts that men value. There are many other possibilities for subtle discrimination.62 Thus, both the method chosen for calculating the new annuities and the arrangements put in place for enforcing this method are of great importance for the future well-being of women. To sum up, the first decade of transformation brought greater losses of pension protection for women compared to men in all three countries. These losses reflect two broad trends across Central Europe. First, the strong appeal of individualism in all areas of life shaped the politics of pension reform, leading to benefits in both the public and private schemes that more 61 Wуycicka et al., sections 2.3.2 and 2.3.4, this volume. 62 This potential is greatest under a system such as that in Hungary where the law allows for multiple, competing annuity providers. Were there instead one single provider for the country, there would be no opportunity for that firm to discriminate in selecting its customers from among a larger pool of private scheme members. This option is among three arrangements being considered in Poland. Agnieszka Chlon, "The Polish Pension Reform of 1999", Framework 1, "Options for providing annuities in Poland", in Fultz (2002), p. 135.
25 closely reflect contributions paid, that is, to a curtailment of redistribution. Second, tight fiscal limitations in all three countries constrained the options for achieving equal treatment in pension schemes, leading to reforms that imposed greater disadvantages on women in some regards. The losses are most severe in Poland, where the particular combination of changes adopted the elimination of redistribution in both the new public scheme and in the new commercially managed savings accounts in conjunction with continuing earlier retirement age for women poses major disadvantages for women. They are least severe in the Czech Republic, where elements of redistribution so far retained in the pension system benefit women. In Hungary, the new equal treatment of women with respect to retirement age has helped to offset some of the economic losses that would otherwise be associated with reduced redistribution. The equal treatment implied in the adoption of gender-neutral life expectancy tables is an important step to avoid further disadvantages. The four issues examined in this section call for attention and action by those with an interest in advancing women's interests. First, in those countries that have privatized their pension schemes, the use of genderspecific life expectancy tables (as opposed to unisex tables) would put women at considerable risk. In several CEE countries, it remains an OPEN QUESTION
whether this issue will be settled on the basis of social insurance principles or on the operating norms of private pension funds. Advocates of gender equality should weigh in. Second, there is great need for adherence to minimum standards that ensure that all those with low earnings will receive decent levels of protection in retirement. During the 1990s, the pendulum of public policy swung far in the other direction in many Central European countries; and corrections are needed. Useful benchmarks are provided by ILO Convention 102, Minimum Standards of Social Security, and the European Social Security Code. While this is not exclusively a gender issue, advocates of gender equality should lend their support. Third, the value of women's retirement age preferences is being eroded indirectly, as reduced redistribution in pension schemes makes payment levels increasingly inadequate for those who retire early. In this circumstance, equalizing the retirement age for women and men becomes a practical necessity. Finally, child care has been devalued for pension purposes in two of the three countries through credits that penalize parents, overwhelmingly women, who leave work to care for young children. Advocates of gender equality
26 should join forces with all those who recognize the social value of child care to ensure that pension systems should not penalize such periods.63 Beyond these issues, the analysis raises a number of problems whose solution lies outside the social security system: i.e., the persistent gender wage gap in CEE countries, the higher unemployment rates that women experience in most of them, and the skewed division of responsibilities for child care between women and men that continues to exist across the region. As noted previously, social security alone is a weak instrument for reshaping the entrenched beliefs and practices that sustain unequal treatment of men and women in the larger environment. Rather, these problems require an integrated response that combines social security reforms with broader changes ie, changes in labour law, affirmative action, public education
, and stronger legal protection
s against discrimination. Together these studies provide ample illustration of the need for broader action on gender issues that shape the impact of social security schemes. 63 Germany, for example, takes several approaches to ensuring this. For every child born after 1 January, 1992 a parent receives one pension credit point per year (calculated by dividing the insured wage for a calendar year by the average wage of all insured for the same year) during three years, regardless if she/he is employed or not (Before the 2001 reform, credits were given for one year). If the parent is employed, the caring credits are added to the obligatory pension contributions from the wage. In addition, if a parent has a low income (e.g. because of part-time work) while the child is between three and ten years old, her/his pension contributions are boosted by 50% up to a maximum of the pension contributions from the average wage of all insured for the calendar year. Only one parent, typically the mother, can benefit from this rule. Parents without income from employment who have two or more children below the age of 10 are credited an additional 0.33 pension credit points per year. The total entitlement must not exceed average contributions from all insured, and eligibility for the uprating of pension contributions between the 3rd and 10th birthday of the child is based on 25 years of contribution payments.