Welfare State Retrenchment in Central and Eastern Europe: the Case of Pension Reforms in Poland and Slovenia Igor Guardiancich This paper endeavours to shed some light on the mechanisms that led to the divergence of welfare state arrangements across Central and Eastern Europe (cee). In particular, pension system reforms displayed a great deal of variance, which surprised both institutionalists and convergence theorists. The Polish and Slovenian cases are thus presented and compared in a political economy perspective. Theories of retrenchment, recent studies on the dynamics of cee pension reforms and consultations with some of the relevant actors, were employed in order to account for the divergence of reform outcomes in the two countries. The study focused on three main explanations: partisan competition, the interaction between relevant external (World Bank) and internal actors (Minister of Labour and Minister of Finance) and the trade-off between power concentration and accountability concentration. The latter yielded the best explanation. While Polish reformers managed to internalise most veto actors’ reservations, Slovenian politicians excluded from consultation the country’s main trade union. Its opposition determined the rejection of radical reforms recommended by the World Bank Introduction After the collapse of Communism in 1989, most cee countries embraced the neo-liberal strategy of stabilisation, liberalisation and privatisation. Although the inherited socialist premature welfare state was intellectually bankrupt, mismatched with the objectives and principles of a market economy and illegitimate in the eyes of the public (Rueschemeyer !999, !5), its reform was not seen as a priority, and it was postponed until the second wave of reforms. Local politicians were afraid of social unrest, were still convinced that the socialist system could cope with the fiscal pressures, and they used social security to mitigate the worst strains Igor Guardiancich is graduate alumnus of the European Institute, London School of Economics and Political Science, United Kingdom. Managing Global Transitions 2 (1): 41–64 42 Igor Guardiancich engendered by the transition. Price liberalisation (ending the implicit subsidies), privatisation of firms (establishing hard budget constraints and shifting the responsibility for welfare provision from the firm to the state), rising unemployment, increasing absolute poverty and the output decline, which entailed a collapse in state revenues, made social security reforms unavoidable from both the social and fiscal standpoints (Wa-gener 2002,156–7). Most institutionalists expected from cee countries the creation of a corporatist-conservative (Esping-Andersen 1990, 26–9) welfare arrangement, because the socialist and pre-war systems were employment-based and because continental Western Europe was seen as s powerful influencing factor. However, they overlooked the persisting instability of cee institutions and the fact that both the European Agreements and the acquis communautaire say precious little in the welfare state realm, since the details of social policy are delegated to single Member States (Wagener 2002, 153–4). In contrast, neoclassical convergence theorists predicted that cee countries would heed the World Bank’s (wb) recommendations and end up with a liberal welfare state, forgetting that path-dependence plays a crucial role where policy feedback and high fixed costs, determined by large, inherited Pay-As-You-Go Defined Benefit (payg-db) schemes, prevent radical policy shifts (Myles and Pierson 2000, 312). The comparison of pension reforms in Poland and Slovenia proves that both camps were wrong.* If fundamental reform of the pension system means its (partial) privatisation, which is what the wb recommends, then Poland introduced radical changes, while Slovenia ended up with patching-up measures (Ch³on´ et al. 1999, 6; Müller 1999,37). While this essay will employ the literature on retrenchment in post-communist cee, it will not directly address its dynamics.2 Instead it will try to explain, through a comparison of the political economy of reforms in Poland and Slovenia, the factors accounting for the divergence of their pension systems and to discuss the problems associated with passing a politically and economically unattractive reform, such as a radical departure from the Bismarckian-Beveridgean paradigm (Müller 1999,107) in old age security. This study will be structured as follows. In the first part, Pension reforms in premature welfare states, the problems associated with the privatisation of cee pension systems will be addressed. In the second, Pension reforms in Poland and Slovenia, a brief political economy account on the evolution of the socialist pension system up to the present schemes will be presented. The last section, Hypotheses on divergence, will Managing Global Transitions Welfare State Retrenchment in Central and Eastern Europe 43 analyse the institutional and political factors accounting for the lack of convergence between the two countries’ reform paths. Pension Reform in Premature Welfare States THE SOCIALIST PENSION SYSTEM The socialist welfare state roughly matched the continental European Bismarckian system. In fact, benefits were a worker’s privilege and not a social citizenship right. The difference was not straightforward since socialism maximised labour participation, and unemployment was practically absent (Wagener 2002,154). Most socialist countries had a unified pension scheme, which was included into the state budget, thereby cross-subsidising other expenditure items. Employees’ contributions were largely abolished, leaving employers’ contributions as the only source of financing. The coverage was expanded to the majority of the population. Therefore the system guaranteed horizontal equity, but could not prevent vertical efficiency problems, i. e. leakages in targeting, and rising expenditures (Barr 2001, 245–52). The low pensionable age was seen as a victory for socialism. The scheme was managed by trade unions. The contribution-benefit link was weak, as benefits were usually based on best-earnings formulae (cfr. Myles and Pierson 2000, 324). Benefits were low and hardly differentiated, but universalism was breached by the privileges granted to employees who held risky or unhealthy jobs, classified as important for the advancement of socialism. Unsystematic indexation produced trans-generational inequity, since older pensions were continuously losing purchasing power (cfr. Müller 1999, 94–6). During the early transition these problems degenerated. The access to benefits became even easier after the relaxation of early retirement and disability criteria, intended to alleviate the pressure on the overstrained labour markets by pushing the unemployed out of the labour force altogether, and to buy social peace by granting privileges to the most endangered groups (miners, heavy industry workers etc.; Gomulka 2000,1–2). The falling revenues associated with the output decline, tax evasion and the informalisation of the economy rendered the pension systems fiscally unaffordable and in need of deep reforms. In Table 1 the main indicators of the Polish and Slovenian pension systems are presented. Barr (2001, 252–3) uses ebrd data to calculate the ratio between system dependency and age dependency. The results (175 for Poland, 186 for Slovenia) show that the leakages (people below the retirement age receiving a pension, compared to those above the age) are respectively 75% Volume 2 · Number 1 · Spring 2004 44 Igor Guardiancich Table 1: Pension system selected indicators Poland Slovenia System dependency ratio (1996) 61.2% 60.6% Age dependency ratio* 22.33%–23.77% l6.l%–20.3% Replacement rate (1995–1996) 65% 75.8% Social contributions 19.52% 15.85%- 24.35% 13.65% (old-age and non old-age) 23.57% 1990 1996 1991 1996 Public pension spending! 8.6% 15.2% 10.9% 14.5% State budget transfersf 0.0% 4.1% –1.8% 3.3% Implicit debtf 220–251% 200–260% No reform scenario 2050 2044 Public pension spending! 27.1–34.5% 27.3% Contribution rate needed in 2050 54–93 n.a. * Poland: old-age population/working population in 1990–2000 (Security 1997 24); Slovenia: 64+/15–64 in 1991–2001 (Stanovnik 2002, 30). f As percentage of gdp. Sources: Gomulka 2000; Hausner 2001; Lindeman et al. 2000; Orenstein 2000; Security 1997; Stanovnik 2002. and 86%, among highest in the world. This similarity, and the fact that after the failed early attempts at stabilisation both countries increased payroll taxes to cover the soaring expenditures, makes the two systems well suited for a comparison focused on institutional and political, rather than on structural or economic factors.3 THE WORLD BANK’S PRIVATISATION RECIPE The wb became a major external actor in post-communist social policy reorganisation (Orenstein 2000,13) after it included specific pension reforms into the neoliberal ‘Washington Consensus’ package, following the positive developments in Chile and Argentina (Müller 1999,25). wb’s 1994 publication ‘Averting the Old Age Crisis’ supported the introduction of private elements into existing payg schemes. According to the Bank, these not only render the schemes fiscally sustainable when confronted with mounting demographic problems, but also promote growth through the expansion of capital markets, savings and investments. In this framework the state retains a residual role of poverty alleviation and pension fund supervision, mainly because state-run payg systems had Managing Global Transitions Welfare State Retrenchment in Central and Eastern Europe 45 Table 2: The three wb pillars4 Redistributive plus Savings plus coinsurance coinsurance Means-tested, minimum pension guarantee, or flat Tax-financed Mandatory publicly managed (ist) pillar Personal savings plan or occupational plan Regulated fully funded Mandatory privately managed (2nd) pillar Savings plus Objectives coinsurance Personal savings plan or occupational plan Fully funded Voluntary (3rd) pillar Form Financing Source: World Bank 1994,15 financially spun out of control in middle- and high-income countries, becoming neither efficient, nor equitable, nor sustainable (wb 1994, 1, 13–4). ‘Averting the Old Age Crisis’ became a catalyst for intellectual confrontation, which generated a global social policy debate between various international organisations (cfr. Deacon 1997, 91–152), and it even triggered a wide diversification of views within the wb (Stanovnik 2002, 64). wb’s approach was termed by Lo Vuolo the ‘new pension orthodoxy’ (Müller 1999, 26) and it is schematically summarised in Table 2. This paper’s main assumption is that the introduction of a privately managed mandatory fully funded second pillar might be politically feasible, but it is economically hardly desirable in cee. In fact the institutional prerequisites of private pension funds are more demanding than those of public systems, because they call for developed financial markets, regulatory capacity and political stability to function effectively. These are often missing in cee. In addition, the only available financial assets might be high risk and low yield (Barr 2001,103). Finally, and more convincingly, the transition costs are gigantic, due to the implicit debts inherited from the socialist era.5 The set up of even a modest wb inspired second pillar renders part of the implicit debt explicit, which causes the double payment problem. Myles and Pierson (2000,312–20) make a clear distinction between mature payg systems and latecomers, claiming that the former cannot heed the wb’s recommendation and should instead opt for a rationalisation (i. e. parametric reform) of the existing payg arrangement. cee schemes can be assimilated with these. Fixing a payg is probably less costly than resorting to funding. Following this wave of disapproval, the wb conceded that an uncritical acceptance of its recommendations should be avoided. In fact Von Gers- Volume 2 · Number 1 · Spring 2004 46 Igor Guardiancich dorff, a wb official, recently admitted that too heavy a load had been placed on cee pension systems, in terms of using them to simultaneously promote savings, capital markets, economic growth and address social problems (Report 2003). Pension reforms in Poland and Slovenia THE EXISTING LITERATURE In contrast to the political economy of retrenchment, authors who explicitly deal with pension reforms in cee are relatively few. The account of the political economy of pension reforms in Poland and Slovenia will be based upon the research by Müller (1999) and Orenstein (2000) for Poland and by Stanovnik (2002) for Slovenia. Orenstein’s framework will be used. It is based on Weir and Skocpol’s (cfr. Orenstein 2000, 9–10) seminal work, which laid the foundations for the new institutionalism, stressing the importance of both the state as an autonomous actor and policy feedback leading to path-dependence. Orenstein adds the international organisations (especially the wb and imp), which influence decision making through: • contributions to the global social policy discourse, conferences and publications which constrain national policy, and • direct intervention, which includes all forms of technical assistance (sending of missions, financial assistance, the secondment of wb employees; Orenstein 2000,11–4). wb conditionalities are used as blame avoidance mechanisms for unpopular policy measures by radical reformers, whose commitment to reforms is a signalling mechanism, which is highly appreciated by financial markets (Müller 2002,130). Specifically for pension reforms, the wb exerted its influence through the agenda setting role in local debates (Müller 1999, 54) and by spreading intellectual innovations within the global social policy discourse. Despite this focus on institutions and international actors, both Müller and Orenstein (Stanovnik follows his logic) emphasise the interaction between domestic actors, thereby taking on a more pluralistic approach. Since negotiated reform of the welfare state is the norm, as compared to governmental unilateralism (cfr. Rhodes 2000,165–94), these interactions are present in both actor-centred (Müller 1999,49–52) and rational choice institutionalism. In the former the constellation of available actors is given by the institutional setting. The actors display determined pref- Managing Global Transitions Welfare State Retrenchment in Central and Eastern Europe 47 Volume 2 · Number i · Spring 2004 48 Igor Guardiancich erences and their interaction decides the final policy outcomes (Müller !999, 49–52). The latter interprets the institutional setting as a constraint for the available policy options (especially the size of the implicit debt, i. e. the maturity of a pension system), but it also focuses on the power concentration of the executive (veto actors and veto points, cfr. Tsebelis 1995), which enhances its ability to pass unpopular legislation. The description of the four phases of pension reforms in Poland and Slovenia will follow Orenstein’s (2000,16–9) timeline. From socialism to transition economics describes the inherited system and the initial reforms. Commitment-building starts when political actors become aware that reform is necessary (usually following the advice of external actors, cfr. Cashu 2003, 5) and ends with the government’s adoption of a single reform programme. During coalition-building final legislation is passed and reformers try to prevent anti-reform mobilisation. Implementation instead poses new problems and introduces new actors (private pension funds). The outcomes can be eventually evaluated. Since relevant actors were not excluded from earlier reform phases in Poland and Slovenia, probably only further fine-tuning measures are to be expected (Oren-stein 2000,19–20; Stanovnik 2002, 66–7). POLAND From Socialism to Transition Economics During the interwar period two unified Bismarckian pension schemes covered white- and blue-collar workers. In 1954 a payg-db system with Soviet features replaced the fully funded schemes. It continued to be managed by the Social Insurance Institute (zus) (Müller 1999, 93–4). Pensions were financed by employers’ contributions and benefits were relatively high, but the ‘old pension portfolio’ problem, stemming from unsystematic indexation, implied that older pensions were losing purchasing power. Therefore 40% of those who reached their pensionable age continued to be employed (Müller 1999,94–6). After 1989 high unemployment and relaxation of retirement and disability requirements brought the system into open crisis.7 Stabilisation was attempted through ad hoc measures (Orenstein 2000,46) which, despite their excessive support of benefit adequacy, were fiercely opposed by the ‘grey lobby’ (Müller 1999,103). The latter was eventually responsible for p ost-Solidarity’s electoral defeat against the Democratic Left Alliance (sld) and Polish Peasant Party coalition in 1993 (Ch³on´ et al. 1999,11). Managing Global Transitions Welfare State Retrenchment in Central and Eastern Europe 49 Commitment-Building Even before the sld-psl coalition stepped into office the debate around pension reforms was heated. There were three camps: ‘rationalisers’, reformers and non-reformers (Ch³on´ et al. 1999, 11). The non-reformers enacted the early ad hoc measures, which were stopped by the Constitutional Tribunal in 1995. The reformers proposed different radical plans. The Ministry of Finance (mof) under Ko³odko published ‘Strategy for Poland’ in June 1994 and suggested the introduction of a mandatory second pillar. Solidarity advocated a similar arrangement. The Minister of Labour and Social Affairs (mol) Miller was a ‘rationaliser’. He supported voluntary savings and a parametric reform of the first pillar, but the public found his plan too timid (Hausner 2001, 216). The stalemate which ensued between Ko³odko and Miller was broken only by the appointment of a new mol, Ba²czkowski, who was an independent and a former Solidarity activist. From 1994 he was chairman of the Tripartite Commission for Socio-Economic Affairs. He managed to create a link with the opposing Solidarity Electoral Action (aws) and established the Office of the Plenipotentiary for Social Security Reform (ops) in April 1996 (Orenstein 2000, 47–8). The wb reinforced its position by financing the creation of the ops and by letting the Polish wb official Rutkowski advise it (Müller 1999,107). Coalition-Building mol Ba²czkowski died in November 1996, putting the whole reform at risk. Hausner became the plenipotentiary and he won independence from the new mol Zielin´ski, who opposed radical reforms. (Orenstein 2000, 49). ‘Security through Diversity’ was approved in April 1997 and advocated a multi-pillar system, including a Notional Defined Contribution (ndc) payg first, a mandatory funded private second8 and a private voluntary funded third pillar. The Tripartite Commission approved the proposal in April 1997 without opposition from the social partners, and the government supported the proposal with a massive information campaign (Ch³on´ 2002, 160). The Sejm approved the easy laws (especially the regulation of the second and third pillars) swiftly and with high cross-parliamentary support (Orenstein 2000,51) in September 1997, because these introduced new elements into the system, leaving unchanged the existing ones (Ch³on´ et al. 1999,14). The aws-uw coalition won the 1997 general elections. The second phase of coalition-building started. Although rare in Polish politics, Volume 2 · Number 1 · Spring 2004 50 Igor Guardiancich a cross-coalition consensus was reached (Müller 1999, 112). The newly appointed plenipotentiary Lewicka, a senior Solidarity pension expert, wholeheartedly embraced the reforms (Ch³on´ et al. 1999,14). The Freedom Union (uw) did not oppose ‘Security through Diversity’ as it was involved in its preparation, but aws was sceptical. After the government’s formal approval of reforms in April 1998, aws and uw entered a stalemate (resolved with the support of sld) in the all-parliament Extraordinary Commission. Solidarity and opzz raised a number of issues, but they preferred to reach an agreement, rather than mobilise against reforms (Orenstein 2000, 53). The whole debate took much longer, because the laws (the reform of the first pillar and the reorganisation of zus) remodelled the existing schemes (Ch³on´ et al. 1999,16). In December 1998 both laws were adopted. Implementation The hasty approval was achieved at the expense of many details, whose determination was postponed (the law on annuities) (Ch³on´ 2002,120). The krus (the farmers’ fund) and disability pensions were left untouched. The Polish reform achieved a double individualisation through the mandatory second pillar and the first payg-ndc scheme, thereby granting more transparency, fiscal stability and less redistribution. Lower indexation will eventually lead to a surplus after 2025. These savings, along with explicit debt and privatisation revenues will be used to finance the transition. According to Ch³on´ (2002, 122–4), ops’s forecasts are overoptimistic (cfr. Security 1997,9–10,85). Falling replacement rates will give rise to benefit adequacy and poverty alleviation concerns (Geroldi and Merano 2001,14). Implementation was extremely problematic: the participation rate in the mandatory scheme exceeded all expectations (Ch³on´ 2000, 34–6; Fultz 2002, 18–9), the zus was technically unprepared and fraud coupled with high administration costs, determined negative rates of return for all the funds. Not surprisingly, the public became increasingly disillusioned with the enacted reforms (Ch³on´ 2002, 183–91). SLOVENIA From Socialism to Transition Economics The Bismarckian welfare state inherited from the Austro-Hungarian empire (Müller 2002, 114) was perpetuated during the Yugoslav kingdom Managing Global Transitions Welfare State Retrenchment in Central and Eastern Europe 51 (Stanovnik 2002, 19–20). After wwn the pension funds were nationalised and a Soviet payg system was introduced. The Yugoslav-Soviet rift put the system in flux (Müller 2002, 114), and not until Kardelj’s 1974 Constitution did federalisation and contribution financing become definitive. Thereafter guidelines were set at federal level, while at lower levels ‘self-managed communities of interest’ had extensive regulatory rights (Stanovnik 2002, 21). The pension system preserved some integrity, since the organisation of the Institute for Pension and Disability Insurance (ipdi) was never broken down to sub-republic levels (Böhm 2003). The 1983 Slovene Pension and Disability Insurance Act (pdia) included farmers and the self-employed into the scheme. Pensioning criteria were generous and early retirement was only temporarily penalised (Stanovnik 2002,22–4). Due to insufficient indexation, benefits were anchored to wage growth in 1986, instantly worsening the looming crisis. The Yugoslav recession hit Slovenia, but the pension system resisted and the population was sufficiently satisfied with it (Böhm 2003). After independence the system became unsustainable due to the relaxation of eligibility criteria, unfavourable demographic prospects (Geroldi and Marano 2001, 2) and structural unemployment. The 1992 pdia was a timid attempt to tighten eligibility criteria and a simple adaptation of the 1983 pdia. Stabilisation failed and the ipdi became dependent on state budget transfers (Stanovnik 2002,25–9). Reforms became unavoidable. Commitment-Building The IMF and wb were active in Slovenia and suggested a two-stage multi-pillar reform plan. The 1995 imf-wb report influenced all subsequent reform drafts (Stanovnik 2002, 41–3). The new Minister of Labour, Family and Social Affairs (mol) Rop was appointed after zlsd (United League of Social Democrats) exited the LDs-led (Liberal Democrats) government coalition. Rop invested his entire political career in the pension reform (Böhm 2003), and (almost) unilaterally published the ‘Starting Points for the Reform of the Pension and Disability System of Slovenia’, which advocated the introduction of a wb inspired second pillar. A phare team criticised the ‘Starting Points’, asserting that the transition costs would be unbearable. In 1996 the general elections produced a new coalition between lds, Desus (Pensioners’ Party) and sls (Slovenian People’s Party). Desus and sls were included into the team preparing the ‘White Paper on the Reform of the Pension and Disability Insurance in Slovenia’ (wp) and subsequently neutralised. The wp was published in Volume 2 · Number 1 · Spring 2004 52 Igor Guardiancich November 1997. It envisaged many parametric reforms to the first pillar, the introduction of a mandatory funded privately managed second pillar and the expansion of the (nonexistent) voluntary private third pillar (cfr. Stanovnik 2002, 39–46). The wp was the only official reform proposal, because all potential opponents were deliberately excluded, especially the Association of Free Trade Unions of Slovenia (ftus). Coalition-Building Due to the concertation character of Slovenian decision-making (Geroldi and Marano 2001,15) this phase started with the appointment of the tripartite negotiating working group for pension reform, which included the m ol, the trade unions (and thus the previously excluded ftus) and the business associations. The ftus opposed the wp and was backed up by the opposition party zlsd. It protracted negotiations from January 1998 until April 1999 (Stanovnik 2002, 47, 64) and started a counter-information campaign in favour of rationalisation (Böhm 2003). In March 1998 it organised the largest protest march ever in independent Slovenia, claiming that the second pillar would undermine the inter-generational solidarity and create poverty among pensioners. During commitment-building mof Gaspari delegated the design of reforms to Rop, but now he opposed the mandatory funded pillar on fiscal grounds. The combined action of the ftus, a critical evaluation of the wp by the economist Bole in a paper commissioned by the wb and the negative opinion of influential economists determined the quiet demise of the second pillar in April 1998 and the final wb’s forfeit. The government acknowledged ftus’s veto role, and gave Rop a full mandate in order to overcome the persisting stalemate (Stanovnik 2002, 49). The 1999 pdia contained deep amendments to the first pillar and very generous tax treatments for the voluntary funded privately-managed third pillar. The new pdia was approved in Parliament in June 1999. Implementation The ILO praised the Slovenian parametric reform for cutting 25% of future benefits, without introducing a second pillar (Fultz and Ruck 2001, 40), thereby enabling the use of public pillar savings for other social security purposes. The wb instead calculated that high replacement rates and (quasi) wage indexation would still engender fiscal unsustainability (Geroldi and Marano 2001, 8–11), forgetting to include the newly introduced trans-generational solidarity indexation element, which reduces Managing Global Transitions Welfare State Retrenchment in Central and Eastern Europe 53 the yearly adjustments of the stock of pensions in function of the eligibility and accrual criteria of new pensioners (Kidric¡ 2002, 4–6). This index yielded encouraging results: both the replacement rate and overall pension spending fell. In addition, the recent agreement on public employee collective schemes is transforming the third pillar into a quasi-mandatory private pillar of considerable proportions (Stanovnik 2003), and there is still some leeway for further parametric reforms. Hypotheses on Divergence Summing up, both Slovenia and Poland achieved significant retrenchment in their pension schemes through parametric reforms, but they chose a fundamentally different stance towards redistribution and risk. The structural factors which could account for this divergence are the inherited socialist pension system, the economic crisis in early transition and the demographic prospects. The latter two hinge on the ‘benefit of crisis’ argument, which holds that radical reforms are feasible under critical conditions, when the opposing interest groups can easily be broken (Müller 1999,46). In contrast Pierson (1996,156) contends that while retrenchment might be feasible, radical reforms would still meet with considerable resistance. Other authors (Orenstein 2000, 11; Brooks and James 2001, 140–6) suggest that faced with high implicit debts reformers tend to be more moderate and retain larger public payg pillars. Neither fiscal nor demographic crises can alone explain the final outcomes. In fact the large implicit debts, the early measures cushioning the labour market and the use of higher contributions to match the expenditures were common to both countries. The lags in tax collection (the Olivera-Tanzi effect) also played a significant role (Kidric¡ 2003). The demographic prospects were worse in Slovenia (cfr. Stanovnik 2002, 30; Geroldi and Marano 2001,3–4; Ch³on´ 2002,96–100), but its reforms were milder. With regard to the inherited systems, it is renowned that Bismarck-ian schemes are resistant to changes, due to the powerful networks of social support generated by welfare state expansion (Müller 1999, 41–2; Pierson 1996, 150–1). In fact early reforms were a clear reflection of the existing systems in both countries. The subsequent developments can be (partially) attributed to the fact that Poland had just earnings-related benefits, while in Slovenia both contributions and benefits were individualised already during the 1970s (employers’ contributions reflected the Volume 2 · Number 1 · Spring 2004 54 Igor Guardiancich wages of individual employees). Therefore, the Slovenian system was not perceived as unfair by the population. Fiscal sustainability became the major concern. In Poland pensions were seen as opaque and discriminatory, thus both fiscal sustainability and individualisation were required. Therefore the introduction of a new instrument, such as a mandatory private funded scheme, would have proved more difficult in Slovenia and necessitated a better-designed governmental strategy (Kidric¡ 2003). Since structural factors are not sufficient to explain the divergence, the paper will try to apply three (overlapping) political economy theories of retrenchment. PARTISAN COMPETITION According to Kitschelt (2000, 265–6) the dynamics of party competition directly affect retrenchment policies. If the median voter opposes unpopular policies, the rational ruling party will not enact them. Thus even large but weakly organised groups, such as pensioners, gain much electoral power. Nonetheless multiparty politics offer other options, among which the policy appeal of an alternative government coalition is worth emphasizing. Some authors (Müller 1999, 41; Pierson 1996, 150–1) include social-democratic parties among the potential opponents to radical pension reforms, due to their commitment to existing welfare states. Kitschelt (2000,281–2) instead argues that it is not the strength of labour or social-democratic parties per se, which impedes retrenchment, but a partisan alignment of government and opposition parties, where competition is based on socio-cultural issues and all parties claim to be defenders of the welfare state. Similarly, Ross (1997,188–9) emphasises the electoral strategy of the left as a means of covering the policy ground of the right: the ‘Nixon in China’ argument permits retrenchment, the refusal of neolib-eralism retains the state in charge, i. e. pension cutbacks without privatisation. Ross’ and Kitschelt’s approaches are maybe only partially suited for post-communist politics. Various authors (cfr. Innes 2001) stress the instability of cee party systems, especially the lack of programmatic alternatives on the economic front (dictated by external constraints). Notwithstanding, former national-accommodative communist regimes (cfr. Kitschelt et al. 1999,1–92) such as Slovenia and Poland have the best chances to develop party competition not exclusively focused on valence Managing Global Transitions Welfare State Retrenchment in Central and Eastern Europe 55 issues. Until then the locus of alternative economic views should maybe be sought outside the party system. In Poland the sdRP (Social-democracy of the Republic of Poland), a communist successor party with a social-liberal, secular inclination, represented the sld-psl coalition’s core and had to compete against aws (national, religious and anti-communist) and uw (radical, pro free-market, secular). Party competition developed over radical reforms (treated as valence issues, due to the conformity of views) with cultural values (religion) at the core and was probably conducive to a pre-emptive move by sld in old age security, because no alternative or credible party was defending the welfare state’s status quo (cfr. Kitschelt 2000). Similarly, in Slovenia the parties identified themselves through cultural values and agreed on the direction of economic reforms, transforming them into valence issues (Ionit¸a¢ 1997, 36, 56). The lds (civic-liberal) almost never lost power during the country’s brief history, opting for various coalition arrangements. The reformed communists, zlsd (mildly social-democratic), exited the government in January 1996, because they opposed outright retrenchment, and remained until 2000 in quasi-opposition. When lds proposed a radical reform no alternative party could have blocked its policies, which were staunchly opposed only by the ftus, probably the only credible supporter of social-democracy in Slovenia.9 Thus, while eliticisation of politics attributable to the weakness of non-political actors is representative of Polish decision-making (Ion-it¸a¢ 1997, 28–9), Slovenia displays a stronger civil society - party system interaction, which contrasts with the ‘associational wasteland’ scenario supported by authors like Offe and Bönker. INTERACTION AND PREFERENCES OF EXTERNAL AND INTERNAL ACTORS The mol, mof and wb play crucial roles in cee pension reforms. During the second wave of reforms ‘extraordinary politics’ had already given way to ‘politics as usual’ (cfr. Balcerowicz 1994,176–7). Müller (1999,52– 3) goes very far in that she claims that the mol and mof display diverging preferences, in order to benefit their constituencies. The former favours a parametric reform within the Beveridge-Bismarckian paradigm and the latter advocates radical reforms for fiscal stability. Thus the mof and wb usually agree, but again, according to Müller (1999,53–4) their very presence is contingent on structural factors: fiscal imbalances of the pension Volume 2 · Number 1 · Spring 2004 56 Igor Guardiancich system for the mof, external indebtedness for the wb. Their mutual involvement should guarantee privatisation. Orenstein (2000,14) is more cautious and only argues that greater exposure to the wb translates itself into more fundamental pension reforms. All of these conjectures are either theoretically or empirically (within this comparison) at least partially unfounded. Although mol Miller was staunchly opposed to radical reforms, the opposite must be said about mol Rop in Slovenia. While Miller was substituted by Ba²czkowski, in order to diminish the distance between veto actors (Orenstein 2000,12), Rop was given, right after the departure of zlsd from government, an exclusive mandate in order to enact the needed reforms after years of immobility and stalemate. This same immobility probably dictated his preference for radical reforms. The other governmental actor, the mof, may also have an ambiguous stance towards radical reforms, since he faces various trade-offs. The implicit debt puts reforms on the agenda, but may render the mof ambivalent towards privatisation. With high explicit debt, the mof prefers reforms, which reduce future fiscal obligations, but the implicit debt conversion may become impossible. Underdeveloped capital markets make reforms which increase savings very palatable, but they also give rise to concerns about the funds’ performance (Brooks and James 2001, 138–42; Müller 2002, 137). In Poland mof Ko³odko was consistently pro reform. In Slovenia mof Gaspari was not. At first he was absent, due to the decision-making procedures in Slovenia (the exclusive delegation of reforms to the relevant Ministry); subsequently he became passive, as he did not participate in the drafting of the wp and he de facto delegated the responsibility of transition financing to Rop. Gaspari intervened only when the stalemate between Rop and ftus was harshest, and opposed the second pillar only after Bole’s publication (Stanovnik 2002, 49, 58–60). Thus the standard roles of the mof and mol, as they are envisaged by Müller, were in Slovenia inverted. Since the wb was not operating in a political vacuum it had to use both direct and indirect interventions to overcome opposition (Orenstein 2000, 73). In both countries the wb was active and it influenced the political debate from the beginning. Interestingly, its proposals were at first met with greater reservations in Poland than in Slovenia (Goli-nowska 1999, 181), and both countries homogenised their views on the funded element with the establishment of a single governmental special office (the ops in Poland) or working group (for the wp in Slovenia). Managing Global Transitions Welfare State Retrenchment in Central and Eastern Europe 57 However, the outcomes were different: ‘Security through Diversity’ was better designed than the White Paper. Nonetheless, it cannot be a priori said that the wb had less leverage in Slovenia than in Poland. While direct assistance was greater in the latter, the wb sponsored trips and conferences, allocated funds and relied on Milan Vodopivec, a former wb official, for advising mol Rop (Stanovnik 2002, 42). The reaction to Bole’s paper (which claimed that transition costs would be unsustainable) demonstrates the wb’s consistent pursuance of a specific plan for Slovenia. The Bank’s ‘top brass’ was aghast and ceased sponsoring the country’s pension reforms from the very moment the paper was presented, refusing even to publish it (Stanovnik 2002, 62–4; Stanovnik 2003). POWER CONCENTRATION AND ACCOUNTABILITY CONCENTRATION Retrenchment policies are essentially unpopular, resulting in a political exercise in blame avoidance (Müller 1999, 40–1), since their advocates face a trade-off between their policy preferences and electoral ambitions (Pierson 1996, 146). Therefore pension reforms must minimise anti-reform mobilisation from concentrated interest groups (Orenstein 2000, 60) using obfuscation, division and compensation tactics (Pierson 1996,147; Brooks and James 2001,159–61). In order to spread the blame, as many veto actors as possible have to be included in the reforms’ design, otherwise their exclusion from an early phase may result in harsher opposition at a later stage (Orenstein 2000, 72–5). In this respect some authors (Bonoli 2000, 239; Orenstein 2000, 12–3) contend that fewer veto points are equalled with more radical reforms and retrenchment. The number of veto points depends on institutional factors, electoral results and the strength of the pro-welfare coalition (Bonoli 2000, 240–1). Thus minority governments and grand coalitions will have to accommodate more demands than bare majority governments, and the strength of trade unions will diminish the executive’s room for manoeuvre. On the other hand power concentration poses accountability problems: competitive political systems, electoral rules and the political cycle all play a role (Bonoli 2000, 244–5). In fact Ross (1997, 192) demonstrates that grand coalitions, where blame is spread among the coalition partners, were in the 1980s more conducive to retrenchment than less diffuse executives. The task is now to explain why Slovenia did reject radical reforms despite the fact that its institutions and politics constantly produced fewer Volume 2 · Number 1 · Spring 2004 58 Igor Guardiancich veto points than the Polish counterparts. In addition to institutional factors, such as a quasi-unicameral parliamentary system in Slovenia and a semi-presidential bicameral one in Poland, Slovenia had even greater power concentration, because Rop was given a full mandate for the reform of pensions (which is a distinctive feature of Slovenian decisionmaking). On the other hand the strength of the pro-welfare coalition was much underestimated in Slovenia. The analysis of cee pro-welfare coalitions shows that pensioners were considered unanimous voters, while trade unions appeared too fragmented to be threatening (Böhm 2003; Ch³on´ 2002, no). Radical reforms have therefore an advantage over rationalisation, in that they can be designed to minimise the resistance of pensioners by excluding them from structural changes (Nelson 2001,246). In Poland the new system was precluded to the population older than 50 and in both countries pensioner parties were effectively neutralised. However, exclusionary compensation is not sufficient to win trade unions over. Rop made a double faux pas in that he excluded the ftus from both a stage of reforms (commitment-building) and a deliberative forum (the working group for the White Paper). In addition, he probably did not believe the ftus could reverse the large public support built around the wp. In any case Poland had an easier task than Slovenia. Polish unions are weak and divided along ideological lines, while union density and collective agreements’ coverage is in Slovenia the highest among cee countries. In addition ftus was constantly affiliated to the opposition party zlsd and has considerable institutional power over legislation on labour relations in the Economic and Social Council of Slovenia (Stanovnik 2002, 64). Therefore the ftus could effectively act as a veto actor by threatening the government with social conflict. ftus’s uncompromising opposition against the mandatory second pillar can be attributed to its conviction that the government pursued a ‘hidden agenda’ (Müller 1999,28), which implied the slashing of first pillar benefits, when the transition costs would become unsustainable. Rop understood that further pursuance of partial privatisation would have put his whole political career at risk and therefore the proposal was quietly withdrawn. This confirms that concentration of accountability may outweigh concentration of power effects. After presenting the three approaches, it is now clear that the most important actors in both reforms were the wb, mof, mol and the trade unions. Their systematic comparison shows evidently the veto Managing Global Transitions Welfare State Retrenchment in Central and Eastern Europe 59 Table 3: Comparison of Polish and Slovenian actors and policy legacies10 Poland Slovenia Policy Only benefits were individualised. Both benefits and contributions were legacies Need for further individualisation individualised under socialism. Fiscal and for fiscal stability. stability was paramount. World Great leverage due to external General political and public Bank indebtedness. Both direct and acceptance of the‘new pension indirect intervention. orthodoxy’. More indirect than direct intervention. mol Miller - rationaliser. Baczkowski - Rop-consistently pro-reform. appointed to diminish the distance between veto actors. ops -pro-reform, independent. mof Ko³odko and his successors - Gaspari - first absent due to the consistently pro-reform. delegation of responsibility to Rop. Not participating in or opposing reforms until Bole’s paper. Trade Solidarity and opzz- numerically ftus-numerically strong, unions weak, ideologically divided, little ideologically committed to zlsd, institutional power. Either pro- great institutional power. reform or accepted side-payments. Consistently against reforms and did not accept side-payments. role played by ftus, as compared to Solidarity or opzz. Additionally ftus’s role was crucially reinforced by the policy legacies, which prevented Slovenian decision-makers from acting unilaterally. Conclusions This paper demonstrates how politics and institutions affected pension reforms in Poland and Slovenia and clearly points out that the divergence of CEE welfare states is contingent on a multitude of factors, among which the trade-off between power concentration and accountability concentration is paramount. However it is worth stressing that, after substantial accommodation of special interest groups’ demands, high cross-parliamentary consensus was reached on the reform programmes in both Poland and Slovenia (Böhm 2003; Orenstein 2000, 51). This confirms that while there was overall agreement on the need for retrenchment the inclusion of neoliberal prescriptions was animatedly debated. Therefore the actor-centred institutionalist framework is very suitable for studying cee welfare state retrenchment. Both reforms that have been analysed display path-dependent elements, while the political ac- Volume 2 · Number 1 · Spring 2004 6o Igor Guardiancich tors’ leeway is testified by the fact that both radical and incremental reform possibilities have been carefully taken into account. Despite this indisputable general framework, the rest of Müller’s early theories cannot be systematically applied to Slovenia. Both the wb’s approach and the ambiguous stance of mof Gaspari are telling of the ‘small n’ problem which affects Müller’s research. On the other hand Orenstein’s skilful adaptation of Tsebelis’ veto actors methodology is fully applicable to the Polish and Slovenian cases. Nonetheless its limitations are straightforward. The dependent variable in Tsebelis’ work is the capacity for policy change (in this study the adoption of a mandatory fully-funded second pillar). Due to the dichotomous nature of this variable, its use is somehow restricted to the comparisons of cases when the executive struggles for the adoption or rejection of a specific policy. In this respect the approach does not explain much of the complex exchange and interaction between the involved actors, such as the government and the social partners, which characterised all cee welfare state reforms. In order to fill in these lacunae, a whole new theory should be developed. A good basis to start from could be the recent research on the reform of Bismarckian pension systems carried out by Natali and Rhodes (2003). Poland and Slovenia show enough similarities in order not to exclude its applicability. However, even the most sophisticated approaches developed for Western Europe have to be adjusted in order to account for the patronage, which has been exerted by international organisations on cee politics. Notes 1. Pension reforms in the Czech Republic and Hungary confirm the lack of convergence. The former retained its payg scheme, the latter introduced a fairly large privately managed fully funded second pillar. 2. Pierson (2001, 422–55) mentions three types of retrenchment. Privatisation falls under recommodification, rationalisation under re-calibration and other cutbacks under cost containment. He associates each type of retrenchment with one of Esping-Andresen’s three worlds of welfare capitalism (liberal, corporatist-conservative and social-democratic). In contrast, Natali and Rhodes (2003, 5–6) contend that a mix of the three is used in any reform of Bismarckian welfare states. 3. The trilateral comparison featured in Orenstein (2000) cannot count on such striking similarities. 4. The wb terminology will be used throughout the paper. Managing Global Transitions Welfare State Retrenchment in Central and Eastern Europe 61 5. See Table 1. Poland and Slovenia display greater implicit debts than most OECD countries, excepted Italy. 6. Weir and Skocpol’s approach differs from the previous ones in that it stresses the two dimensions generating the state’s autonomy and influence: a) the autonomy of its bureaucrats; b) its organisational configuration, which influences all other societal aspects - the interaction of political actors, the demands of interest groups and the inclusion of intellectual innovations in national policies. 7. See Table 1. 8. The second pillar was gradually phased-in for implicit debt and tactical sequencing concerns. The population over 50 was excluded to avoid pensioners’ opposition (Müller 1999,115). 9. Orenstein (2000, 65) rightly predicted that a right government supporting neoliberal arrangements would meet with greater resistance by ‘old’ trade unions. Nevertheless this idealistic role should not be overemphasized. Natali and Rhodes (2003, 5, 7–8) contend that retrenchment politics are not only about blame avoidance, but also involve credit claiming. Political parties and trade unions alike act as vote-, office- and policy-seekers. In fact even in Slovenia the ftus approved some reforms, which were detrimental to the outsiders. 10. For details refer to Guardiancich 2003. References Balcerowicz, L. 1994. Poland. In The political economy of policy reform, ed. John Williamson, 153–77. Washington, dc: Institute for International Economics. Barr, N. 2001. The welfare state as piggybank: Information risk, uncertainty and the role of the state. Oxford and New York: Oxford University Press. Bonoli, G. 2000. Political institutions, veto points, and the process of welfare state adaptation. In The new politics of the welfare state, ed. P. Pier-son, 238–64. Oxford: Oxford University Press. Böhm, L. (Executive Secretary, Association of Free Trade Unions of Slovenia). 2003. Interview by I. Guardiancich, 18 July. Brooks, S., and E. James. 2001. The political economy of structural pension reform. In New ideas about old age security, ed. R. Holzman and J. Stiglitz. Washington, dc: World Bank. Cashu, I. 2003. The politics of pension reform: An east-west comparison. Interim Policy Research Paper 2001–2003 (continuing). International Policy Fellow, Open Society Institute. Budapest, March. Ch³on, A. 2000. Pension reform and public information in Poland. Social Protection Discussion Paper Series 0019, The World Bank. Volume 2 · Number 1 · Spring 2004 62 Igor Guardiancich Ch³on´, A. 2002. The Polish pension reform of 1999. In Pension reform in Central and Eastern Europe. Vol. 1, Restructuring with privatization: Case studies of Hungary and Poland, ed. E. Fultz, 95–205. Budapest: International Labour Office. Ch³on´, A., M. Góra and M. Rutkowski. 1999. Shaping pension reform in Poland: Security through diversity. Social Protection Discussion Paper Series 9923, The World Bank. Deacon, B. 1997. Global social policy: International organizations and the future of welfare. London: Sage. Esping-Andersen, G. 1990. The three worlds of welfare capitalism. Cambridge: Polity. Fultz, E. 2002. Pension reform in Hungary and Poland: A comparative overview. In Pension Reform in Central and Eastern Europe. Vol. 1, Restructuring with privatization: Case studies of Hungary and Poland, ed. E. Fultz, 11–23. Budapest: International Labour Office. Fultz, E., and M. Ruck. 2001. Pension reform in central and Eastern Europe: Emerging issues and patterns. International Labour Review 140 (1): 19–43. Geroldi, G., and A. Marano. 2001. The pension system and the pension reform in Slovenia. A paper presented at the World Bank - iiasa Conference, Learning From Partners, Vienna, 6–7 April. Golinowska, S. 1999. Political actors and reform paradigms in old-age security in Poland. In Transformation of social security: Pensions in Central and Eastern Europe, ed. K. Müller, A. Ryll, and H. J. Wagener. Heidelberg: Physica-Verlag. Gomulka, S. 2000. Pension problems and reforms in the Czech Republic, Hungary, Poland and Romania. London: Centre for Economic Performance, LSE. Guardiancich, I. 2003. Why did pension system reforms in Poland and Slovenia lead to different outcomes? Est–Ovest 34 (6): 41–79. Hausner, J. 2001. Security through diversity: Conditions for successful reform of the pension system in Poland. In Reforming the state: Fiscal and welfare reform in post-socialist countries, ed. J. Kornai, S. Haggard, and R. Kaufman. Cambridge and New York: Cambridge University Press. Innes, A. 2001. Party competition in post-communist Europe: The great electoral lottery. Working Paper Series 54, Centre for European Studies Central and Eastern Europe. Ionit¸a¢, S. 1997. Party systems in Eastern Europe. A comparative study on seven countries - Poland, Czech Republic, Slovakia, Hungary, Romania, Bulgaria, Slovenia. Working Papers Series 2, Romanian Academic Society. James, E. 1996. Providing better protection and promoting growth: A de- Managing Global Transitions Welfare State Retrenchment in Central and Eastern Europe 63 fence of averting the old age crisis. International Social Security Review 49 (3): 3–20. James, E 1998. The political economy of social security reform: A crosscountry review. Annals of Public and Comparative Economics 69 (4): 451–82. Kidric¡, D. 2002. Pension reform in Slovenia. Paper presented at the conference Practical Lessons in Pension Reform: Sharing the Experiences of Transition and oecd Countries, Warsaw, 27–28 May. Kidric¡, D. (Head of Department for Social Development, Institute for Macroeconomic Analysis and Development, and Advisor, Insurance Supervisory Agency). 2003. Interview by I. Guardiancich, 13 August. Kitschelt, H. 2000. Partisan competition and welfare state retrenchment. When do politicians choose unpopular policies? In The new politics of the welfare state, ed. P. Pierson, 265–302. Oxford: Oxford University Press. Kitshcelt, H., Z. Mansfeldova, R. Markowski and G. Toka. 1999. Post-communist party systems, competition, representation, and inter-party cooperation. Cambridge: Cambridge University Press. Lindeman, D., M. Rutkowski and O. Sluchynskyy. 2000. The evolution of pension systems in Eastern Europe and Central Asia: Opportunities, constraints, dilemmas and emerging best practices. Washington, dc: World Bank. Myles, J., and P. Pierson. 2000. The comparative political economy of pension reform. In The new politics of the welfare state, ed. P. Pierson, 307– 33. Oxford: Oxford University Press. Müller, K. 1999. The political economy of pension reform in Central-Eastern Europe. phd diss., Europa-Universität Viadrina. Müller, K. 2002. Between state and market: Czech and Slovene pension reform in comparison. In Pension reform in Central and Eastern Europe. Vol. 2, Restructuring of public pension schemes: Case studies of the Czech Republic and Slovenia, ed. E. Fultz, 113–46. Budapest: International Labour Office. Natali, D., and M. Rhodes. 2003. The new politics of the ‘Bismarckian welfare state: Pension reforms in continental Europe. Florence: European University Institute. Nelson, J. 2001. The politics of pension and health care reforms in Hungary and Poland. In Reforming the state: Fiscal and welfare reform in post-socialist countries, ed. J. Kornai, S. Haggard, and R. Kaufman. Cambridge: Cambridge University Press. Orenstein, M. 2000. How politics and institutions affect pension reform in three postcommunist countries. World Bank Policy Research Working Paper. Volume 2 · Number 1 · Spring 2004 64 Igor Guardiancich Pierson, P. 1996. The new politics of the welfare state. World Politics 48 (2): 143–79. Report on World Bank - Trade Union meeting on pension reform, draft. 2003. Washington, dc, 21–22 May. Rhodes, M. 2000. The political economy of social pacts. In The new politics of the welfare state, ed. P. Pierson, 165–94. Oxford: Oxford University Press. Ross, F. 1997. Cutting public expenditure in advanced industrial democracies: The importance of avoiding blame. Governance 10 (2): 175–200. Rueschemeyer, D. 1999. Left parties and policies in Eastern Europe after communism: An introduction. In Left parties ad social policy in post-communist Europe, ed. L. Cook, M. Orenstein, and M. Rueschemeyer. Boulder and Oxford: Westview Press. Security through diversity: Reform of the pension system in Poland. 1997. Warsaw: Office of the Government Plenipotentiary for Social Security Reform. Stanovnik, T. 2002. The political economy of pension reform in Slovenia. In Pension reform in Central and Eastern Europe. Vol. 2, Restructuring of public pension schemes: Case studies of the Czech Republic and Slovenia, ed. E. Fultz, 19–73. Budapest: International Labour Office. Stanovnik, T. 2003. Rezultati pokojninske reforme. Finance, 11 June. Stanovnik, T. (Professor of Public Finance, Faculty of Economics, University of Ljubljana). 2003. Interview by I. Guardiancich, 8 July. Tsebelis, G. 1995. Decision making in political systems: Veto players in presidentialism, parliamentarism, multicameralism and multipar-tism. British Journal of Political Science 25 (3): 289–325. Wagener, H.-J . 2002. The welfare state in transition economies and accession to the EU. West European Politics 25:152–74. World Bank. 1994. Averting the old age crisis: Policies to protect the old and promote growth. Washington, dc: World Bank. Managing Global Transitions