Paper prepared for the 2nd EASP Conference
University of Kent, 30th June to 7th July 2005
Theoretical review on pension politics and structural accounts
: Explaining pension development and reforms
in Japan, Korea, and Taiwan after 1990s
<Please quote as a draft>
Contents of paper
Introduction
Pension politics: theory and limits
Explaining pension development and reforms with ‘pension politics’
Bringing structure back in: Socio-economic structure and business structure
Socio-economic structure: industrialism, globalisation, and post-industrialism
Business structure: Instrumental and structural power of capital
Implications of structural accounts: towards an integrated framework
Conclusion
Contact:
Young-Jun Choi
Department of Social and Policy Science,
University of Bath, BA2 7AY
44-1225-384518
Introduction
Since 1990s pension schemes in Japan, Korea and Taiwan have been markedly expanded and reformed under the influences of considerable socio-economic changes. In Japan, there were three major reforms in 1994, 1999, and 2004, mainly focusing on enhancing the financial sustainability of the overall pension schemes by increasing the pensionable age and contribution rate, changing the benefit indexation, and reducing the benefit level. The Korean government has extended the coverage of the National Pension (NP) to firms with 5 or more employees (1992), farmers and rural self-employed people (1995), and urban self-employed and firms with less than 5 employees (1999). In the 1999 reform, there was a substantial effort to curb future pension spending by increasing the contribution rate and lowering the benefit level. In addition, the government is proposing another pension reform for the financial sustainability in 2005. The Taiwan government has also implemented several pension reforms by introducing an old-age allowance for farmers (1995) and making it eligible to all citizens (2002). Upgrading the retirement benefit to the Labour Pension (LP) in 2004 is another significant move for old-age security in Taiwan.
In spite of increasing interests and importance of public pensions, there are very few scholarly studies of the development and changes of pension systems in East Asia in a comparative perspective. In addition, newly emerging studies on pension development in single countries tend to heavily rely on political institutional approaches, which have also dominated in recent pension reform literature in the western world. However, it seems that they have limitations to explain the similarities and differences between the experiences of pension development in each country. The research aims to provide a theoretical analysis on changes of the pension system in globalising and post-industrialising Japan, Taiwan, and Korea. In order to fully understand pension development in this region, I will draw on a variety of recent theoretical writings on East Asian welfare regimes, pension politics, and structural accounts. In the first part, I will theoretically examine the arguments of pension politics, and subsequently review them with three countries’ experiences. Then, in the second part, the usefulness and validity of structural accounts will be theoretically examined and subsequently their empirical implications will be suggested in explaining cross-national differences of the pension developments. Consequently, the study will emphasise the integrated framework, taking into account policy legacies, political institutions, socio-economic factors, and business structure and power.
Pension politics: theory and limits
There seems very little specific effort to unify and define pension politics, but it is commonly and broadly used as a term representing political activities in the process of public pension decision making. In fact, pension politics means not just manifestations of political activities but implies that politics is crucial in the understanding of pension development and transformation in an academic usage. Furthermore, this approach based on new institutionalism emphasises policy dynamics of pension schemes. The ‘new politics’ approach[1] defined and explained by Pierson (1994, 1996, 2000, 2001a) is the most systemic representative for it. One of the strengths of this approach, which could also be a critical weakness, is its large domain and scope to embrace the various political explanations used in welfare state development so far, such as electoral politics, interest-group politics or even the role of labour unions or corporatism with the injection of the following concepts; veto points (Bonoli 2000), blame avoidance (Weaver 1986, 2003) and credit claiming (Natali and Rhodes 2004). The proliferating welfare-state regime research after Esping-Andersen (1990, 1996, 1999) commonly more or less agree or accept the view that institutions structure actors’ behaviour and enable its form to maintain stability (North 1990). As a result, this view has been often used for counter-arguments against structural-convergence thesis, emphasising domestic political institution and welfare state structure.
In this account, polity or policy institution is a significant factor in connection with other factors (Pierson 1994). Once an institution is set up, it generates increasing returns (North 1990) or self-enforcing mechanism (Pierson 2000) and they made institutional change stable and predictable. Furthermore, it goes beyond just existing, but actively affects actors’ values, norms and behaviours (Swank 2002) and structures political interactions (Thelen and Steinmo 1992). Again, deeply embedded interests in the institution generate institutional inertia resistant to changes. However, they are careful not to fall into the danger of an overly static view of the social world derived from path-dependent changes (Pierson 2000; Pierson and Myles 2001). Contingency or unintended outcome from a relatively small event or a critical role of sequencing and timing provide changes within a certain limit. Also, inside the logic of path dependence, there are ceaseless political activities enabling or disabling institutional changes by various political actors (Orloff 1993, Mueller 2002, 2003). Since pension reforms, probably pension development as well, involve the alteration of existing distributional equilibria, “those who are going to lose out are likely to exploit veto points where available, in order to affect the course of policy-making or simply to try to block the adoption of legislation perceived as disadvantageous” (Bonoli 2000:43).
Blame avoidance is another factor, frequently mentioned in pension politics (Weaver 1986, 2003, Pierson and Weaver 1993, Shinkawa 2003) in that policy-makers are overwhelmingly concerned with blame avoidance to reduce the associated social and political risks. For example, the strategies of division and obfuscation, i.e. playing off one group of beneficiaries against another and lowering the visibility of cutbacks, e.g. by increasing the complexity of reforms, making a scapegoat or diffusion of blame (Pierson 1994, Pierson and Myles 2001, Shinkawa 2003, Mueller 2003) are commonly used in the process of pension retrenchment and are alleged to be significant factors shaping pension institutions. In addition or in a similar vein, Natali and Rhodes (2004) maintain that blame avoidance behaviours are often adopted with the combination of credit claiming exercises such as the introduction of supplementary fully-funded schemes (also Schludi 2001) because of politicians’ strategic moves in competition with one another. As a result, pension politics causes stable and continuous changes within the institutional background.
Nevertheless, it seems that there have been several reasonable criticisms on this approach in different respects. Firstly and most fundamentally, some academics challenge the broadness or vagueness of its approach. Institutionalism and path dependence arguments have been discussed by state-centred theorists with the accounts of state capacity and bureaucratic influence, mainly by Skocpol (1985) and Orloff (1993). However, during the 1990s it seems to expand its scope to a polity-centred one, including parties, electoral rules, and legislatures (Weaver and Rockman 1993). Furthermore, bringing polity-centred institutionalism into welfare retrenchment discourse, Pierson (1996, 2000) and other historical institutionalists extend its view from polity to policy centred institutionalism. Contrary to the view that regards a policy as a passive object, this approach defines a policy as an active and dynamic entity constraining actors’ choice and value and, consequently, producing path dependence by the logic of increasing return. The enlarging scope of this theory results in the blurred boundary of institutionalism, and it is criticised that this approach is suitable for ‘guiding perspectives’ or ‘framework’ rather than ‘theory’ (Gorges 2001).
The second issue is the question on the path dependence of policy legacy. In most new institutionalist accounts, established institutions are usually fairly stable and ‘change’ is unlikely because of path dependence generated by institutional legacy. However, shifting our sight to Eastern Europe and Latin America, it is obvious that there have been abrupt pension reforms and transformations during the last two decades, triggered in Chile in 1981 (Mueller 2002, 2003, Holzmann 2003). Also, to a lesser extent than developing countries, there have been pension reforms in the western countries (Overbye 1992, Holzmann et. al 2003). Some countries, such as the UK, Sweden, and Italy, have implemented pension reforms by introducing private factors or notional defined contribution pension and others also keep planning and conducting parametric reforms to different extents (Nay 2000:35). They appear to aim at securing the financial sustainability and minimising the adverse effects of the public pension on the national economy. In this sense, the institutional approach only provides a partial explanation, but there are still other phenomena unsolved by this approach.
Thirdly, there is a danger of ‘institution-determinism’ in explaining policy changes. For example, Andersen and Larsen (2002) argue that Sweden has ended up with earnings-related social insurance system towards conservative model whereas Demark has ended up with multi-pillar system towards liberal model by path breaking reforms according to path dependence. Also, Korpi (2001) observes that the state corporatist model shows the strongest tendency of path dependence, whereas the basic security institutional form turns out to be the most vulnerable one to pressures for change not only because it has the lowest transaction costs, but also because it has more possibility of political conflicts derived from its low-level benefits. In the study of pension retrenchment in Japan, Shinkawa (2005:157) also argues that “the early take-off pension retrenchment in Japan was due to institutional vulnerability embedded in its pension scheme”. They tend to find out core sources of changes inside policy-institutions, but institutions seem to be used as a panacea to justify every changes occurring in different pension models, e.g. Swedish, British, and Japanese model.
Finally, pension politics arguments seem to pay less heed to newly emerging actors and non-decision makers influencing pension reforms. The new international actors such as international NGOs, international organisations or multi-national corporations are relatively free from blame avoidance or credit claiming politics since they are not directly constrained by electoral politics. Even though many studies still underscore the role of domestic politics, e.g. Mueller (2003), it is also undeniable that their influence has been rapidly expanded and will be. In particular, there has been a ‘take-for-granted’ approach on the role of business in welfare state literature (Farnsworth 2004). However, it is also apparent that business is a strong non-decision maker that constrains the agenda of pension development and reform, particularly in the era of the globalised competition economy when the structural power of capital is increasing (Gough and Farnsworth 2001). In summary, even though institutions matter, there is a limit to explain pension development and reform, particularly in a long-term perspective. Heuberger and Ney (2004) argue that the price of emphasising continuity and path dependency, at least in terms of analytical focus, has been to underestimate and in some cases completely miss fundamental change. Consequently, it is necessary to establish a new framework embracing other factors rather than solely relying on a pension politics discourse based on the new institutionalism.
Explaining pension development and reforms with ‘pension politics’
Political-institutional approaches have been employed in many studies on East Asian welfare regimes. As is widely acknowledged, a dictatorship or one-party dominated system in this region had been exercised until recently and state had monopolised all the decisions regarding social and economic issues without much intervention from labour unions and civil society. While the weakness of progressive political forces, as a result of successful oppression by the state, is a decisive role in retarding welfare development, it is argued that the state has also taken advantage of social policy legislation in order for political legitimisation. Also, after democratisation, it has been frequently noted that electoral competition is a major reason behind the legislations of new welfare programmes. Many studies, mostly with qualitative historical methodology, such as Kwon (1998a, 1999), Lee (1999), Tang (2000a, 2000b), Ramesh (2000), Yang (2000, 2003), Lin (2002), Aspalter (2002a, 2002b), Shinkawa (2003), Estevez-Abe (2002), and Shin (2004), put political-institutional reasons on the primary level in their explanation.
Table 1. History of pension development and reforms after 1990s
Year / Japan1994 / Pensionable age of the first basic tier phased out from 60 to 65, from 2001 to 2013 for male and from 2006 to 2018 for female.
Benefit indexation changed in line with net wages
1999 / Reduce aggregate pension benefits by 20 percent by 2025 by means of
1) reductions in the earnings-related benefits by 5 percent
2) shift from wage indexation to CPI indexation
3) new earnings-test introduced for those aged 65 to 69
4) pensionable age for the second earnings-related tier increased from 60 to 65, from 2013 to 2025 for male, from 2018 to 2029 for female.
2001 / Defined Benefit/Contribution Corporate Pension Law passed
2004 / Modified ‘Automatic Balancing Mechanism’ introduced
- Lower benefits (40 years working model workers: 50% replacement rate)
- Higher contribution rate to 18.30 by 2017
- Increase state subsidy to the benefit payment of the BP from 1/3 to 1/2
Korea
1992 / National Pension (NP) compulsory coverage extended to firms with 5 or more employees
1995 / NP extended to farmers, fishermen, and the self-employed in rural areas with 3% contribution rate
1997 / Amendment of the LSA allowed employers to convert their retirement allowance schemes into Corporate Pension Scheme (CPS)
Elderly Welfare Act enacted, including new Old-age Allowance scheme including near-poor elderly people, implemented in 1998 (approx. £13 to £25 according to economic status)
1999 / NP reforms
- extended to the self-employed in urban areas with 3% contribution rate (increase to 9% by 2005)
- Reduced replacement level to 60% to 45% (30 years contribution)
- Increasing pensionable age to 65 by 2033
2003 / firms with 1 or more employees started to include as workplace-based insured workers from individual-based insured (three stages extension to 2006)
Taiwan
1993 / Old-age allowance implemented in local areas where the Democratic Progress Party (DPP) controlled
National Pension plan started to discuss
1995 / Old-age Peasants’ Welfare Allowance (OPWA) introduced
1998 / Fishermen covered by the OPWA
1999 / National Pension implementation postponed to 2000 due to earthquake and further delay due to changing the government in 2000
2002 / Old-age Citizens’ Welfare Allowance (OCWA) and Old-age Indigenes’ Welfare Allowance (OIWA) introduced
2004 / OPWA increased the benefit level to NT$3000 to $4000
Retirement benefit, lump-sum program, in the Labour Standard Act revised to the transferable annuity program, the Labour Pension, defined contribution scheme and financed by employers (mandatory, less than 6%) and employees (voluntary)
This trend has been also found in case studies explaining pension development and reforms in three countries, even though the main point of their explanation is slightly different from one another. From the previous literature, this explanation with political competition and credit-claiming behaviours has been widely used for analysing the Taiwanese development (Ku and Chen 2001, Aspalter 2002, Chen and Chen 2002, Hill and Hwang 2005:156-157, Lin 2005). According to this introduction of democratic election system in Taiwan prompted the steep competition between political parties and subsequently led to increasing populist measures by politicians. In fact, there have been increasing political campaign promises for social policy reforms (Lin 2005:185) and one of the most important debates has been old-age pension. The opposition party, the Democratic Progressive Party (DPP), brought welfare issues into the centre of election by emphasising the role of state in welfare provision[2]. In the course of the local authority elections of 1993, the DPP launched the agenda of non-contributory old-age allowance, even though the Kuomintang (KMT) criticised heavily it as ‘a collective bribe of elderly voters’ (Ku and Chen 2001:102). Against the swift movement by the DPP, the KMT started to establish the policy for old-age security in two ways: national pension plan and old-age allowance for lower-income elderly. After that, on the one hand, the KMT has made repeated promises about establishing national pension system, particularly during election campaigns. By setting up principles of the national pension and developing sophisticated pension plans (Chen and Chen 2004), the KMT tried to attain an advantage over the DPP’s non-contributory old-age allowance plan that KMT regarded as ‘a political show’.