POLITICAL STUDIES: 2000 VOL 48, 706-723
Productivist Welfare Capitalism: Social Policy in East Asia
Ian Holliday
City University of Hong Kong
The article engages with the literature on the 'East Asian welfare model' by using Esping-Andersen's 'worlds of welfare capitalism' approach to analyze social policy in the region. It describes the main features of a productivist world of welfare capitalism that stands alongside Esping-Andersen's conservative, liberal and social democratic worlds. It then shows that Japan, Hong Kong, Singapore, South Korea and Taiwan are all part of this world, though they divide into sub-groups within it. To account for productivist welfare capitalism in East Asia, the article focuses particularly on bureaucratic politics at the unit level, and on a range of key shaping factors at the system level. It closes by considering the implications of East Asian experience for comparative social policy analysis.
In the decade since Esping-Andersen (1990) published The Three Worlds of Welfare Capitalism, one strand of comparative analysis has focused on the conservative-liberal-social democratic typology of welfare state regimes constructed in that book. In the process, some authors have noted that welfare arrangements in capitalist East Asia - usually said to consist of Japan plus the four little 'tigers' or 'dragons' of Hong Kong, Singapore, South Korea and Taiwan (Vogel, 1991) - cannot readily be fitted into this tripartite framework. Indeed, in the course of the 1990s something of an academic consensus emerged in support of the contention that 'the East Asian experience is distinctive, differing decisively from the Euro-American models current in social policy discourse' (Kwon, 1998a: 27).
There is, however, no agreement about the precise nature of East Asian excep-tionalism. Some argue for a regional cluster, floating such concepts as 'oikonomic welfare states' (Jones, 1990), 'Confucian welfare states' (Jones, 1993), 'Japan-focused East Asian welfare regimes' (Goodman and Peng, 1996), and an 'East Asian welfare model' (Kwon, 1997).1 But even these authors differ about the extent of clustering, with some holding that the significant similarities do not extend beyond Japan, South Korea and (possibly) Taiwan (Goodman and Peng, 1996; Kwon, 1997). The leading survey maintains that 'it is misleading to think in terms of one homogenous, overarching "East Asian welfare model"' (White and Goodman, 1998: 14). Other studies reach much the same conclusion (Midgley, 1986). One found enough difference in social security arrangements in Singapore and South Korea alone to 'challenge the widespread perception that social policies in the four Asian "Tigers" or "Dragons" are broadly similar' (Ramesh, 1995: 229).
Social policy analysts also take different positions on the causes of East Asian exceptionalism. Jones's early focus on culture soon came in for criticism (Jones, 1990; 1993). White and Goodman (1998: 12, 15) claim that cultural explanations are 'a more or less prominent theme' in the literature, but dismiss them as 'unhelpful' in accounting for the evolution of East Asian welfare systems. Competing
© Political Studies Association, 2000.
Published by Blackwell Publishers, 108 Cowley Road, Oxford 0X4 UP, UK and 350 Main Street, Ualden, MA 02148, USA
PRODUCTIVIST WELFARE CAPITALISM
707
studies focus mainly on the role of the state. Deyo (1992: 289-90) holds that with few exceptions 'East Asian social policy has been driven primarily by the requirements and outcomes of economic development policy'. Ramesh (1995: 232) argues that 'social security in [South] Korea and Singapore should be conceptualized as being shaped by the objectives of the two states and the international and societal constraints they face'. Kwon (1997: 479) and White and Goodman (1998: 15) also develop state-centric accounts. Statist rather than cultural explanations are in fact the most prominent theme in the current literature.
This article picks up these two main issues. It frames the first in terms not of an East Asian welfare model, which is both loose and geographically deterministic, but of a world of welfare capitalism. In so doing it draws on Esping-Andersen's work, seeking first to specify the main features of what it calls a productivist world of welfare capitalism, and secondly to demonstrate the ways in which the five East Asian welfare regimes commonly brought within this kind of analysis conform to the type. The former exercise results in some extension of Esping-Andersen's approach. The latter acknowledges that social policy differences within the region are striking, but nevertheless holds that there is a single world of welfare capitalism here within which distinct sub-sets may be identified. The sub-sets are facilitative (Hong Kong), developmental-universalist (Japan, South Korea, Taiwan) and developmental-particularist (Singapore). On the second issue, the article argues that, with very few exceptions, existing explanations of East Asian welfare development are too narrow. Although the state has emerged as the central focus of analysis, still not enough is said about bureaucratic politics. In addition, very little attention is paid to the external context in which domestic factors are placed and by which they are often shaped. Towards the end the article considers the implications of East Asian experience for comparative social policy analysis.
The Universe of Welfare Capitalism
The Three Worlds of Welfare Capitalism is an obvious reference point, not only because others have turned to it in examining East Asian social policy, but also because it provides a powerful framework for comparative analysis. If we are to use it here, however, we need to be clear about its approach and argument. The task Esping-Andersen sets himself is to examine 'a novel phenomenon in the history of capitalist societies' (p. 1), their transformation into welfare states. The focus of his study is, then, not all those capitalist states that happen to engage in a bit of social policy on the margins of everything else they do (or do not do), but only those that are so deeply affected by their social policy that they are best defined as welfare states. For Esping-Andersen, 'the welfare state [is] a principal institution in the construction of different models of post-war capitalism' (p. 5). It is 'qualitatively different welfare-state logics' (p. 4) that divide the 18 states he looks at into conservative, liberal and social democratic regimes marked by distinct developmental trajectories.
When we examine the criteria for inclusion in what might be called the universe of welfare capitalism, we find they are multiple. The notion of a 'welfare-state regime' (p. 2) is set up as the organizing concept of the book in order to 'denote the fact that in the relation between state and economy a complex of legal and
708
IAN HOLLIDAY
organizational features are systematically interwoven' (p. 2). But two criteria are critical: a capitalist order, and a welfare state that extends social rights. On the latter, Esping-Andersen holds that the key component of social rights is 'the degree to which they permit people to make their living standards independent of pure market forces' (p. 3). Here Polanyi's concept of 'de-commodification' plays a central role. It occurs 'when a service is rendered as a matter of right, and when a person can maintain a livelihood without reliance on the market' (p. 22). These criteria, then, mark out the universe of welfare capitalism.
Three further criteria are employed to distinguish worlds within that universe. They are 'the quality of social rights, social stratification, and the relationship between state, market, and family' (p. 29). On this basis Esping-Andersen identifies three clusters of regime types: liberal, conservative and social democratic. A liberal welfare state engages in social provision reluctantly, preferring to leave things to the market wherever possible. The result is a stratification system 'that is a blend of relative equality of poverty among state-welfare recipients, market-differentiated welfare among the majorities, and a class-political dualism between the two' (p. 27). In a conservative welfare state the liberal fixation on markets is not present, the church is an important force, and the granting of social rights is largely uncontested. The key issue here is 'preservation of status differentials; rights, therefore, [are] attached to class and status' (p. 27). One result is that redistribution is negligible. Another, prompted by the power of the church, is that the family is promoted. In a social democratic welfare state 'an equality of the highest standards, not an equality of minimal needs' (p. 27) is pursued. However, social programmes are not uniform, instead being 'tailored to differentiated expectations' (p. 28) with benefits 'graduated according to accustomed earnings' (p. 28). The market is crowded out, and the costs of raising a family are socialized.
The argument developed here is that a fourth criterion needs to be added to the three Esping-Andersen uses. Moreover, once this is done a fourth world of welfare capitalism becomes apparent. The criterion emerges from analysis of Esping-Andersen's claim that the sphere of welfare capitalism should be restricted to those capitalist states so strongly affected by their social policy as to be identifiable as welfare states. This essentially arbitrary restriction rules out examination of capitalist states that do engage in social policy, while also subordinating it to other policy objectives. There is no good reason, however, why such states should be excluded. Rather, the aspect of their social policy that marks them out from Esping-Andersen's three worlds - its subordination to other policy objectives - should be used as a fourth criterion for identifying worlds within the universe of welfare capitalism. When we look inside the state in this way, we can say that social policy is not subordinate to other policy objectives in any of Esping-Andersen's three worlds (as he himself makes clear). In the liberal and conservative worlds social policy is not especially privileged either. By contrast, in the social democratic world it does have a privileged place. In the fourth, productivist world of welfare capitalism, the further contrast is that the reverse is the case. Here, social policy is strictly subordinate to the overriding policy objective of economic growth. Everything else flows from this: minimal social rights with extensions linked to productive activity, reinforcement of the position of productive elements in society, and state-market-family relationships directed towards growth.
PRODUCTIVIST WELFARE CAPITALISM
709
Table 1: Four Worlds of Welfare Capitalism
	Social	Social	Stratification	State-market-
	policy	rights	effects	family relationship
Liberal	Neither	Minimal	Equality of poverty	Market
	privileged		for minority;	provision
	nor		market-differentiated	encouraged
	subordinate		welfare for majority	
Conservative	Neither	Quite	Existing status	Family
	privileged	extensive	differentials	protected
	nor		preserved	
	subordinate			
Social	Privileged	Extensive	Universal benefits	Market
democratic			graduated according to accustomed earnings	crowded out; family socialized
Productivist	Subordinate	Minimal;	Reinforcement of	Premised on
	to economic	extensions	productive elements	overriding
	policy	linked to productive		growth objectives
activity
Feeding all this into Esping-Andersen's analysis, we get a liberal world prioritizing the market, a conservative world defined by status, a social democratic world focused on welfare, and a productivist world premised on growth. Each is an ideal type to which no real-world example conforms exactly. Each is shown in Table 1, with its dominant feature highlighted.
The Productivist World of Welfare Capitalism
The two central aspects of the productivist world of welfare capitalism are a growth-oriented state and subordination of all aspects of state policy, including social policy, to economic/industrial objectives. Alongside these defining features a series of additional elements may be found. Policy makers might seek to pursue economic growth by facilitative means. Then social policy will be geared towards facilitation of economic growth too. Alternatively they might adopt a developmental stance, said by Johnson (1982) in his classic study of MITI and the Japanese Miracle to comprise elite policy makers setting economic growth as the fundamental goal and working out a concrete strategy to attain it. Then social policy can be either universalistic, with the state stepping in to create universal welfare programmes (at least for the productive elements in society), or it can be particularistic, with the state directing individual welfare provision.
710
IAN HOLLIDAY
Within the productivist world, it is therefore possible to identify distinct clusters, each of which addresses the central issues of growth orientation and subordination of non-economic policy in different ways. The three clusters identified here are facilitative, developmental-universalist and developmental-particularist. The facili-tative regime has something in common with Esping-Andersen's liberal type, except that it espouses an overarching commitment to growth, which puts a different spin on things. Social rights are minimal, stratification effects are very limited, and the market is prioritized. The developmental-universalist regime extends some social rights to productive elements of the population and thereby creates a kind of aristocracy of labour. The state, which is an important player in economic policy, is therefore of some social policy significance alongside the market and families. The developmental-particularist regime has almost no social rights as such. Instead, satisfactory levels of individual welfare provision are promoted among productive elements in society. In this way their position is again advanced, and the state plays a directive social policy role alongside the market and families.
If we look within the productivist world of welfare capitalism we find, then, some variations, though as is only to be expected they are not extensive. They are shown in Table 2, which opens up the productivist line in Table 1. On the defining first dimension, all three strands of productivism subordinate social policy to economic objectives. Again, this is a set of ideal types to which no real-world example will conform exactly.
Table 2: The Productivist World of Welfare Capitalism
Social Social       Stratification State-market-
policy rights effects family
relationship
Facilitative
Developmental-universalist
Developmental-particularist
Subordinate to economic policy
Subordinate to economic policy
linimal Limited
Subordinate to economic policy
Limited; extensions linked to productive activity
Minimal;
forced
individual
provision
linked to
productive
activity
Reinforcement of the position of productive elements
Reinforcement of the position of productive elements
Market prioritized
State underpins market and families with some universal programmes State directs social welfare activities of families
PRODUCTIVIST WELFARE CAPITALISM
711
Social Policy in East Asia
The five capitalist states of East Asia on which this article focuses are impossible to place in Esping-Andersen's framework.2 They do engage in social policy, but only once they have first attended to their overriding policy objectives in the economic sphere. It is certainly not their 'welfare-state logics' that define their developmental trajectories. Nevertheless, they are brought within an Esping-Andersen-style analysis here because by any reasonable yardstick they are welfare capitalist.3 The argument is that each state approximates to one of the regime sub-types found in the productivist world of welfare capitalism. A general point to be made at the outset is that East Asian societies retain a strong conception of the family, which therefore looms large in welfare provision in all five states. This is an entirely contingent feature of productivist welfare capitalism in contemporary East Asia. It will not necessarily be found elsewhere.
Japan
Japan is a developmental-universalist state within the productivist world, and has always subordinated social policy to other policy objectives. Its social rights are not entirely limited, for the Allied (mainly American) Occupation of 1945-52 ensured that Article 25 of the 1947 Constitution gave all Japanese the right to 'a minimum standard of healthy civilized life'. It also prompted reconstruction of the health insurance scheme (from 1948) and of the pension system (from 1954). Together, these schemes extended social rights across the whole population. Nevertheless, that extension was very basic, as is revealed by the fact that throughout the high-growth period of 1955-75, government expenditure on social welfare was stable at around 2 percent of GNP (Goodman and Peng, 1996: 201). Indeed, in these years economic growth was very much the overriding policy objective pursued by the key part of the political elite, technocrats working in the Ministry of International Trade and Industry (MITI) (Johnson, 1982). They established financial control of the economy, 'soft authoritarian' control of labour relations, very close links with domestic conglomerates (known, pejoratively, as zaibatsu) and tight control of foreign capital (Johnson, 1987). A major extension of social rights seemed possible when the government proclaimed fiscal year 1973 'Year One of the Welfare Era', and improvements in health care and indexation of pensions did follow (Collick, 1988: 214). Otherwise, however, this initiative was undermined economically by the oil crisis, and politically by the fiscal crisis of those Western states that had made major extensions of social rights in the post-war years (O'Connor, 1973), and were now seen as negative rather than positive models. Japan duly reverted to its own way, though the proportion of public expenditure taken by welfare did start to rise in the 1980s and 1990s, partly as a result of the economic crisis that dominated the latter decade. Today the Japanese state is active in the spheres of medical care, income maintenance, social services and housing (Goodman et ah, 1997: 371). The major elements of its welfare regime remain a series of compulsory health insurance and national pension schemes covering the entire population. In these systems of funded social insurance the primary orientation of the state is regulatory (Kwon, 1997: 468-9). In Japan social rights to basic health care and pensions are very nearly universal. In other spheres they are much more limited,
712
IAN HOLLIDAY
and the market and families play important roles. This is not a pure form of developmental-universalism, for some social rights extend beyond the productive elements in society. Those extensions are, however, limited.
Hong Kong
Hong Kong is mainly a facilitative state within the productivist world, though it has elements of developmental-universalism. Its strategy for economic growth has been to present itself as a citadel of free-market capitalism, though in the post-war period the generalist officials who staffed the 'secluded bureaucracy' of the British colonial administration (Lau, 1982) took responsibility for making land available for production, building large amounts of public housing, renting it at sub-market rates, engaging in industrial planning and controlling the prices of many goods and services (Castells, 1992: 45-9). Since the 1997 handover this kind of activism has continued. In the social policy sphere the dominant stance was initially highly reactive: to the influx of close to one million refugees from the Chinese Revolution of 1949; and to the general crisis of legitimacy for the British colonial regime signaled by 'Cultural Revolution' riots in 1966-67. However, the construction programme started in the 1950s has resulted in almost half the population living in public rented housing. In health care and education social rights are basic but universal. Hong Kong has a substantial publicly-funded hospital sector sitting at the heart of what is otherwise a private-sector health care system. In education, large parts of the primary and secondary sectors are funded by the state but provided by non-state bodies. The tertiary sector, which grew substantially in the 1980s and 1990s, is a state system. The territory has a basic system of social security (MacPherson, 1993). Personal social services are supplied mainly by the voluntary sector, but funded almost wholly by government. The most important gaps in Hong Kong's social rights are various forms of transfer payment. There are no publicly-provided pensions, unemployment benefit or child benefit (Wilding et ah, 1997; Chow, 1998). The resultant package of social rights, which today consumes nearly 20 percent of GDP, has a rather odd profile (Wilding and Mok, 2000). Whilst Hong Kong is certainly closest to the facilitative strand of productivist welfare capitalism, with the market and families performing key welfare functions (Chan, 1998), the state's long-standing intervention in the housing sphere, and its more recent initiatives in education and health, give it elements of developmental-universalism.
Singapore
Singapore is different again, being a developmental-particularist state within the productivist world. The two central elements of its social policy are the Central Provident Fund (CPF) and extensive public housing. The CPF is a compulsory savings scheme, created by the British colonial administration in 1953 (and implemented in 1955), to which employers and employees contribute largely similar, and sometimes identical, amounts. Currently they both put in 20 percent of wage income (up to a fixed ceiling). Crucially, individual entitlements, which can now be used to finance housing, health care and pensions, cannot exceed individual contributions, plus interest. This is, then, a system of forced saving to which both employees and employers contribute part of what is actually the individual's wage.
PRODUCTIVIST WELFARE CAPITALISM
713
It is not redistributive. A massive public housing programme has also been a feature of Singapore's growth years, with the proportion of publicly-constructed homes rising from 9 percent in 1960 to more than 86 percent in 1996. However, this too is a far less 'social' programme than it might appear, for individuals through the CPF buy almost all homes. More than 90 percent of Singaporeans own their flats (Tremewan, 1998: 83-8). In the early 1990s, Singapore's public expenditure consumed about 16 percent of GDP, though the semi-autonomy of the CPF meant that its spending on social protection was greater than this figure suggests (Goodman et al, 1997: 362-4). The unusual aspect of the Singaporean system, in which the state plays a highly interventionist and directive role, is that it nevertheless has a highly restricted social dimension. The CPF is not a social insurance system, as in Japan, but a mandatory provident fund based strictly on individual contributions: individuals can take out up to but not exceeding what they have put in. Crucially, 'the Singapore welfare system grants no entitlement or access to benefits to those outside the CPF system' (Goodman etal, 1997: 366). In Singapore, then, social rights are very limited and something close to the developmental-particularist form of productivist welfare capitalism is found.
South Korea
South Korea is in many respects a pale, and conscious, imitation of Japan (Johnson, 1987). It is, however, a purer form of developmental-universalism. From the beginning, President Park Chung-hee pursued a social policy motivated almost entirely by economic objectives, the influence of which it has never really thrown off (Kwon, 1999). In most cases social rights were extended first to industrial workers, and many have still not been fully universalized. In the case of the National Pension Programme, introduced in 1988, a central aim was capital mobilization during the phase of surging economic growth (Kwon, 1998b; 1999). Singapore's CPF also had this objective. South Korea's other main social policy programmes are Industrial Accident Insurance (introduced in 1964), a Public Assistance Programme for the poor (1965), National Health Insurance (1977), and an Employment Insurance Programme (1995). With the exception of the Public Assistance Programme, South Korea's social policy is based on the social insurance principle also seen in Japan, and the state plays a largely regulatory role (Goodman et al, 1997: 371). In 1995, the public and private sectors in South Korea respectively spent 10 and 1 percent of GDP on social protection. This is a rather pure form of developmental-universalism because the Public Assistance Programme provides no more than a very basic universal safety net, and all other programmes cover only those who have paid into them (though not on the strictly individualistic basis found in Singapore).
Taiwan
Taiwan is also a developmental-universalist state, though here the contingent fact of immediate secession from the People's Republic of China means that the military has always featured prominently in social programmes. Indeed, the tone of Taiwan's social policy was set by its first piece of welfare legislation, the Military Servicemen's Insurance Law of 1953, and social rights, based on the social insurance
714
IAN HOLLIDAY
principle, remained very limited until the emergence of democracy in 1987. Even since then, 'those groups which are seen as most closely related to the stability of the Taiwanese state' (the military, state bureaucrats and teachers) have received the lion's share of central government welfare expenditure: 75 percent in 1991 for a total of 9 percent of the population. Meanwhile, 'the disadvantaged' (the poor, handicapped, young, aged and women) have received next to nothing: 3 percent in 1991. The instrumental productivism of the political elite is further demonstrated by the fact that '[significant periods of growth in social expenditure invariably correspond with the emergence of political crises facing the government'. Taiwan has appropriately been characterized as only 'reluctantly' a welfare state with a restricted set of social rights (Ku, 1995: 355-9). However, in 1995, National Health Insurance - 'the most significant welfare effort by the Taiwanese state in the post-war era' - was introduced as a means of integrating and extending existing programmes (Ku, 1998: 119). A national pension scheme should soon come on stream, though the economic crisis of the late 1990s may delay it. There are, then, 'some signs that a Taiwanese welfare state may be emerging' (Ku, 1998: 137). Nevertheless, Taiwan remains very much a developmental-universalist state with a strong tilt towards the military.
Productivist Welfare Capitalism in East Asia
These brief sketches reveal obvious differences. The states within which social rights have emerged have different profiles. Those rights themselves also exhibit clear variation. In terms of basic orientation, the social insurance principle that dominates social policy in Japan, South Korea and Taiwan contrasts with forced individual savings in Singapore and a kind of residual universalism in Hong Kong. Social rights differ both in consequence and for additional reasons: even within the category of social insurance regimes there are distinct levels of coverage. At the level of detail, the large public housing sectors in Hong Kong and Singapore set them apart from Japan, South Korea and Taiwan. But in Hong Kong public housing has traditionally been rented, whereas in Singapore it is privately owned. The lack of public pension schemes in Hong Kong and, currently, Taiwan distinguishes them from Japan, Singapore and South Korea. Again, Singapore's arrangements are unusual, differing very clearly from those in Japan and South Korea. And so on. It is on these sorts of grounds that critiques of the 'East Asian welfare model' are constructed. Thus, whilst Goodman and Peng (1996: 194) are prepared to 'believe that there is a good case for talking about the development of what might be called "Japan-focused East Asian social welfare regimes" as seen in Taiwan and [South] Korea', they will go no further. Similarly, White and Goodman (1998: 14) build on this analysis to argue that whilst Japan, South Korea and Taiwan exhibit 'systematic similarities', Hong Kong is a 'clear out-rider' and Singapore is 'also distinctive'. It is for this reason that The East Asian Welfare Model, despite its title, becomes a detailed refutation of the concept (Wilding, 2000).
All these points taken, there are nevertheless pan-regional similarities. Most obviously, these are high rates of economic growth into the 1990s, low rates of taxation, public expenditure and social provision, and considerable social stability even during the 'Asian crisis' of the late 1990s. But analysis need not stay at this
PRODUCTIVIST WELFARE CAPITALISM
715
superficial level. Wilding (2000) notes that at least nine common features of East Asian social policy can be drawn from the literature. They are: low public expenditure on social welfare; a productivist social policy focused on economic growth; hostility to the idea of the welfare state; strong residualist elements; a central role for the family; a regulatory and enabling role for the state; piecemeal, pragmatic and ad hoc welfare development; use of welfare to build legitimacy, stability and support for the state; and limited commitment to the notion of welfare as a right of citizenship. We can add in all the qualifications we like, and Wilding himself is certainly very careful to do so, but the notion that there is a core of shared experience which sets East Asian social policy apart from those variants known in the West is hard to resist. Wilding's (2000) further contention is that 'whether or not one discerns an East Asian welfare model through the mists of difference depends on how tightly one conceptualizes the notion of "model"'. The approach taken here is to identify a single world of welfare capitalism within which sub-worlds may be found. To validate this approach, it is necessary to show first, and most importantly, that the overarching categorization is appropriate, and secondly that the sub-divisions work.
The main categorization is a clear fit in all cases but one. In Japan, Singapore, South Korea and Taiwan, post-war social policy was initially developed by either an authoritarian or an undemocratic and unaccountable state. None changed particularly quickly, for even Japan, formally democratic since the Allied Occupation, was under single-party control until the 1990s. Within these polities, technocrats took charge of policy development, and in each case gave economic policy clear priority over social. The social stratification objectives of each of these welfare regimes were also similar, comprising reinforcement of the position of capital, whether large or small in scale, incorporation of the productive working and middle classes, and exclusion of very nearly everyone else. The strategy for social integration was uniformly stability through growth. The political class standing behind each of these welfare regimes was always a largely unaccountable conservative elite that imposed its vision of social policy on the rest of society (White and Goodman, 1998: 15). The exception is Hong Kong, where the largely hands-off, laissez-faire approach of the British colonial elite does not look especially productivist, and few relevant changes have been made post-handover. However, appearances can deceive. Even here, the political elite consciously developed a set of policies to secure maximum economic gain, with part of that set being whatever package of social rights was needed to underpin growth. The result does not look productivist partly because it was never meant to. The appearance of laissez-faire policies, though by no means in full conformity with reality, is an important part of Hong Kong's economic strategy. Behind the facade, the extensive land clearance, public housing and price control schemes in which colonial and post-colonial administrators have engaged make Hong Kong something other than a liberal form of welfare capitalism (Wilding, 1997).
Taking the sub-divisions within the productivist world, the clearest differences are between the mainly facilitative approach of Hong Kong and the thoroughgoing developmentalism of the other four states. These all engage in substantial economic intervention, are clearly identifiable as 'governed markets' (Wade, 1990), and systematically exploit social policy for direct economic policy gain. In Hong Kong the approach is different, but the goal is much the same. Among developmental
716
IAN HOLLIDAY
states, a distinction has to be drawn between Singapore and the rest. As we have seen, Singapore does not have the social insurance principle that characterizes Japan, South Korea and Taiwan. It does have mandatory individual savings plans and a large state role in housing construction. In fact, Singapore's developmentalism is marked by a strong element of particularism, whereas in the other three states universal solutions are more prominent. It is of course necessary to know something about Singapore's peculiar ways of forcing people to be free to make sense of the tension between individual and state evident in its developmental-particularist label. On the other side, it is important to note that although Japan, South Korea and Taiwan are all social insurance systems, in South Korea and Taiwan in particular it is quite possible for many people to fall through the welfare net, thereby limiting the extent of their developmental-universalism.
East Asia does not, then, have a single 'welfare model'. Those who say this are right. But five of its states can be placed in the productivist world of welfare capitalism. Those who identify developmental welfare states in the region are also right, though they usually fail to see that Singapore can be placed in this group while at the same time being distinguished from the 'Japan-focused' regimes. Beyond that, there is a tendency in the literature to confuse the entire issue by holding that economic advance has been subordinate either to 'the prime goal of nation-building' (Goodman and Peng, 1996: 210), or to regime legitimation. This may be true, but it is not what is distinctive about East Asia's productivist world of welfare capitalism. Indeed, in this world states tend to share with states from Esping-Andersen's three worlds an ultimate objective of social solidarity and regime legitimation. It is the ways in which they pursue that objective that set them apart.
Accounting for Productivist Welfare Capitalism in East Asia
Getting in place a categorization of the five East Asian welfare regimes examined here is only the first stage in analysis. The second task is to account for this regional clustering of welfare capitalist states. Two methodological remarks must be made at the outset.
One is that we are not seeking a monocausal explanation. 'The hope of finding one single powerful causal factor must be abandoned,' writes Esping-Andersen (1990: 29); 'the task is to identify salient interaction-effects.' For this reason the approach that sets up a contest between explanations is less helpful than one that focuses on factors lying behind them: not functionalism, structuralism, societal and statist explanations (Ramesh, 1995), but economic base, external factors, social base and political superstructure. In essence what we are attempting to do is establish the correct balance within a matrix of factors, many of which have already been broadly debated in the literature and all of which differ by time and place. Getting the right balance is key, not listing the factors. The second methodological remark is that we need not restrict analysis to internal matters. At the very least we need to focus on two levels of analysis: unit and system. By unit is meant state and all that goes on within its borders. By system is meant the ways in which the various parts of the international 'order' (loosely understood) interact. Existing analyses of East Asian welfare regimes say a great deal at the unit level, but very little at the
PRODUCTIVIST WELFARE CAPITALISM
111
system level. There is no need to be so unbalanced; external factors can take their places alongside internal ones.
Looking first at the unit level, the existing literature places its main emphasis on social base and political superstructure. The Confucian welfare state argument is the best case of a social base explanation, though Jones (1993) also makes clear that the political superstructure performs an important mediating role. Beyond this, any account must say something about the demographics of East Asian states. An important part of the reason why, for example, they tend to have little or no unemployment benefit is that for many years they tended to have little or no unemployment. Similar reasons explain their patchy development of pension systems. The developmental arguments advanced by Deyo (1992), Ramesh (1995) and Kwon (1997), and the developmental nation-building case made by Goodman and Peng (1996), are leading examples of political superstructure arguments. Some accounts also focus on factors that cut across the base-superstructure divide. One instance is labour relations, where state and capital often come together to promote company-level corporatism that undermines trade union mobilization and diminishes the need for state intervention (Gould, 1993). Here, as elsewhere, Japan is the model and both Hong Kong and Singapore are non-conformist.
Missing from the evidently very full analyses of internal factors that have shaped East Asian social policy development is one key dimension, bureaucratic politics. Developed most famously by Allison (1971) in his examination of the Cuban missile crisis, this approach focuses attention on the structure of the state, and on linkages and divisions between its component parts. The existing literature points tentatively in this direction (Kwon, 1997: 479-80; White and Goodman, 1998: 15), but a great deal remains to be done if the unit-level dynamics of East Asian social policy development are to be fully understood. We know a lot about the technocrats and elite policy makers who staffed key East Asian economic agencies like the fabled Japanese Ministry of Finance and MITI, their direct South Korean equivalents, and the Taiwanese Production and Economic Stabilization Boards (Henderson and Appelbaum, 1992: 14). We also know something about the administration of economic policy in Hong Kong and Singapore. Moreover, in the East Asian cases such individuals and agencies are central to social policy analysis, for the argument made here and elsewhere is that social policy has in many ways been a sub-category of economic policy. That said, they still need to be brought within analyses of welfare dynamics, and their social policy counterparts also need to be incorporated. This is a research agenda for future studies rather than something that can be undertaken here.
For the purposes of this analysis, the system level has two key headline aspects -economic and political - and a series of subsidiary elements. The economic dimension chiefly comprises international markets and international business. The political dimension is mainly made up of regional actors and dominant international powers or agencies. In existing accounts these dimensions feature, if at all, as no more than shadowy background factors. Jones (1990; 1993) and Ramesh (1995) make some reference to them, but in neither case is it systematic. Other authors are largely silent about them. There is, however, a case for saying they have been critical to the development of East Asian welfare regimes.
718
IAN HOLLIDAY
It would clearly be hard to exaggerate the extent to which each and every one of the five states examined here is dependent on the world market and world business. Each has, moreover, chosen to make a virtue out of this necessity. Singapore is perhaps the extreme case, single-mindedly pursuing under Lee Kuan Yew a strategy of openness to inward investment. But it differs from its regional neighbours in degree, not kind. Even Japan, traditionally extremely hostile to foreign capital, has always been heavily dependent on the world economy. Similarly, South Korea and Taiwan, which have always had stringent trade protection policies, have been just as dependent as all other East Asian states on trade. Indeed, Vogel (1991: 1) notes that in 1988 'Japan and the four little dragons constitute [d] five of the world's seventeen top trading nations.' In the mid-1990s Japan, Hong Kong and Singapore were all in the world's top ten. The economic crisis of the late 1990s is unlikely fundamentally to change this. The integration of these states into the world economy remains substantial.
Equally, political pressures are highly significant. Within the region all East Asian states face a threatening or potentially threatening neighbour of the kind that might in the West be called a 'hostile Other'. Japan is both threatening and threatened. Its long-standing insecurity was reinforced by defeat in the Second World War, Allied occupation for the best part of a decade thereafter, and revolution in China in the meantime. But it also posed a major twentieth-century threat to the other four territories analyzed here, occupying all of them in the Second World War, and Taiwan and Korea from 1895 and 1910 respectively. These states are also threatened in differing ways by mainland China, and have all been destabilized to greater or lesser extents by the course revolutionary and post-revolutionary politics have taken there since 1949. The greatest impact has been felt by Hong Kong and Taiwan, which were major recipients of refugees in the late 1940s and early 1950s and also experienced knock-on effects from the Cultural Revolution. But mainland China has challenged all these states in one way or another. Internationally, the course taken by Cold War politics had a major impact on East Asia. Henderson and Appelbaum (1992: 14) argue that '[t]he economies of Japan, Taiwan, and South Korea - and the militaries of the latter two countries - were deliberately built with US aid and technology transfer as bulwarks against communism'. East Asian industry also benefited from the demand generated by the Korean and Vietnam Wars. Beyond that, the direct role played by America and American-dominated global institutions, such as the United Nations, clearly cannot be ignored. Among many other things, America led the 1945-52 Allied occupation of Japan, participated (under UN auspices) in the 1950-53 Korean War, and chose from 1949 to 1971 to recognize the Republic of China (Taiwan) instead of the People's Republic (China). It has also been a major post-war aid donor and investor in the region, and remains either the most important or the second most important trading partner for all these East Asian economies.
The argument is not that to explain East Asian social policy development we need to adopt a structurally deterministic approach of the kind associated in international relations theory with neo-realism (Waltz, 1979). We do need to recognize, however, that each of these factors has helped to condition the perceptions of all -not just all significant, but pretty much all - East Asian actors in the post-war period of economic and social policy development. Vulnerability has been key
PRODUCTIVIST WELFARE CAPITALISM
719
(Jones, 1990; 1993), and has persuaded elites and masses to support a development strategy embracing social as well as economic policy. For many and varied, but nevertheless critical, reasons, all sides have agreed not to rock the boat. Indeed, we can draw an analogy here in which places like Hong Kong, Singapore and Taiwan in particular feature as a lifeboat, and mainland China is the forbidding sea. The economic crisis of the 1990s has merely reinforced pre-existing fears.
The value of this systemic dimension is that it helps to explain what has gone on at the unit level: limited, though not entirely non-existent, popular mobilization and extensive elite control. If we look at existing explanations for, say, working-class quiescence, we find that they focus on factors like rapid industrialization, a transient population, Confucianism and benevolent authoritarianism. Clearly none of these aspects of East Asian experience can be overlooked, but on their own they do not constitute a complete account. After all, rapid industrialization is not necessarily unpromising ground for formation of trade unions, and some significant working-class mobilization has taken place (Deyo, 1989). Many people in fact came to East Asian states to stay, knowing from the outset that they were unlikely ever to be able to leave. There is no clear reason why Confucianism need be damaging to working-class mobilization, though it may of course shape it in ways that are unfamiliar to Western observers. Finally, the fact that benevolent authoritarianism was for many years accepted rather than challenged by workers is something that needs itself to be explained. The dimension that must be added to make sense of each of these observed aspects of East Asian development is the systemic, which has acted as an important conditioning (and disciplining) agent in all five states.
Analyzing Worlds of Welfare Capitalism
Bureaucratic politics at the unit level and the range of factors that constitute the system level must, it is argued here, be incorporated into an analysis of East Asia's world of welfare capitalism. Moreover, if they are important in East Asia they are presumably at least worth considering elsewhere. What, then, can we learn from East Asia about social policy development in the West?
The focus of The Three Worlds of Welfare Capitalism is strictly unit-level. In seeking to move beyond 'a simple class-mobilization theory of welfare-state development', Esping-Andersen (1990: 32) adopts 'an interactive approach with regard to both welfare states and their causes'. In so doing, he shifts the focus of analysis to the nature of class mobilization, the nature of class coalitions, and historical patterns of regime institutionalization. A large part of the resultant thesis is that 'politics not only matters, it is decisive' (Esping-Andersen, 1990: 4). This is a critically important argument, but reflection on the East Asian experience suggests that it does not go far enough in developing a political account. The movement to 'bring the state back in' (Skocpol, 1985) that is a feature of contemporary comparative analysis must encompass social policy too. This is an obvious point, and one that has already been taken up in a number of studies. It is used here to identify a fourth world of welfare capitalism.
A point to which less attention is paid is that much more can be said at the system level to explain welfare dynamics in the West. If we look at the main post-war
720
IAN HOLLIDAY
phase of welfare state development in, say, Western Europe, we find significant differences from East Asian experience. In Western Europe, the economic and political dimensions of individual states' external environments have never been wholly threatening, and at times have been supportive. In the main years of welfare state development, West European states created the political institutions of the EEC, EC and EU, and came under the strategic protection of NATO and the American nuclear umbrella. Economically, they benefited from the Bretton Woods system of managed exchange rates, and when that fell apart in 1971 they began to put in its place the Deutschmark zone that eventually gave birth to the euro. None of these structural responses made Western Europe an entirely secure place, but they certainly made it far less vulnerable to external shocks than East Asia.
A fairly crude counter-factual reinforces the point. If post-war West European states had faced, say, a real Soviet threat on their very borders and had, in some cases, contained a large population of Soviet exiles, would they have had anything like the 'maximalist' social policy development they in fact experienced? By the same token, if late-twentieth-century East Asian territories had faced no more threat from, say, mainland China than West European states faced from the Soviet Union for most of the post-war period and had, moreover, been much less prone to international market pressures, would they have experienced 'minimalist' social policy development? Without denying the critical importance of unit-level factors, system-level developments must be brought within any satisfactory analysis of welfare capitalism. Interactions between states can be as important as interactions within them. Sometimes they may be more important.
One of the leading analyses that recognizes this needs, finally, to be considered. Katzenstein (1985) contends that welfare states develop more readily in small, open economies that are particularly vulnerable to international markets, the reason being that both capital and labour are subject to forces beyond domestic control. From all that has been said to this point, it is immediately clear that the East Asian and West European experiences are diametrically opposed. In East Asia's small, open and vulnerable economies, welfare states have developed far from readily. How are we to account for this discrepancy? The answer lies in a modification of Katzenstein's thesis. Small states in world markets constitute promising territory for welfare state development if, and only if, the systemic vulnerability perceived by most actors is of a middle-range sort. When it moves beyond this they will take the East Asian path and prove quite hostile to welfare state development. This is, of course, in line with all forms of socialist analysis that state, roughly, that workers will tend to get a bad deal from capitalists unless they can effectively mobilize against them. It is chiefly system-level factors that have made such mobilization so difficult in East Asia.
Conclusion
A productivist world of welfare capitalism needs to be added to Esping-Andersen's conservative, liberal and social democratic worlds. Japan, Hong Kong, Singapore, South Korea and Taiwan can all be placed in that world, though they separate into additional clusters within it.
PRODUCTIVIST WELFARE CAPITALISM
721
At the start of the twenty-first century, challenges to the productivist welfare capitalist states of East Asia are visible in the continuing economic crisis, resultant unemployment, popular mobilization that has brought real democracy to some, though not all, parts of the region, and demographic changes now threatening most welfare states. The thrust of this analysis is that the continuing vulnerability of all these states to system-level forces makes it highly unlikely that they will move beyond productivist welfare capitalism in the foreseeable future. This is not to argue that the brute facts of economic recession and an ageing population will have no more than a negligible impact on East Asian welfare regimes. Clearly all of them will be forced to extend their reach. However, the important point is that they will almost certainly do so within a productivist perspective. Hong Kong may remain more facilitative than its developmental counterparts, though now that the handover has taken place that is uncertain. But, one way or another, productivism will remain the name of the game.
Perhaps the more interesting question concerns the West. One of the major issues raised by economic globalization is whether almost all of the world's population will soon live in what are effectively small states in world markets. There are, of course, countervailing trends, notably in the continued development of the EU. However, the extent to which this and other defensive structures will succeed in erecting security barriers against the dynamic forces of global capitalism remains unknown. The East Asian experience strongly reinforces the argument that globalization must generate real pressure to roll back the frontier of social policy. It could help make productivist welfare capitalism something of an international standard in the twenty-first century.
(Accepted: 7 October 1999) About the Author
Ian HoIIiday, Department of Public and Social Administration, City University of Hong Kong, Tai Chee Avenue, Kowloon, Hong Kong; email: saian@cityu.edu.hk
Notes
I thank Mok Ka-ho, Adrian Sinfield and Paul Wilding for very helpful comments on a draft of this article. I also thank Paul Kelly and the Political Studies reviewers for sharp and constructive criticism of the version originally submitted to the journal. The usual disclaimers apply.
1 In classical Greek, oikos means 'household economy'. Jones argues that the four small East Asian polities conform to Aristotle's 'household management' style of government.
2 Technically, Hong Kong is not a state. Until the end of June 1997 it was a British Crown colony. Ever since it has been a Special Administrative Region of the People's Republic of China.
3 One East Asian state, Japan, did make it into Esping-Andersen's 1990 study, but received almost no analysis and is a rather awkward fit. In response to some critics' contention that it is distinctive, Esping-Andersen returned to this case in 1997. Acknowledging that Japan 'gives the appearance of a hybrid system', he nevertheless argued first that 'there is little to indicate a distinct ... model' and secondly that 'any attempt at labelling the Japanese welfare state is premature since it has not yet sunk its roots, institutionally speaking' (Epsing-Andersen, 1997: 179).
References
Allison, Graham T. (1971) Essence of Decision: Explaining the Cuban Missile Crisis. New York: Harper Collins.
722
IAN HOLLID AY
Castells, Manuel (1992) Tour Asian Tigers with a Dragon Head: a Comparative Analysis of the State, Economy, and Society in the Asian Pacific Rim', in Richard P. Appelbaum and Jeffrey Henderson (eds), States and Development in the Asian Pacific Rim. Newbury Park CA: Sage, pp. 33-70.
Chan, Chak-kwan (1998) 'Welfare policies and the construction of welfare relations in a residual welfare state: the case of Hong Kong', Social Policy and Administration 32, 278-91.
Chow, Nelson (1998) 'The Making of Social Policy in Hong Kong: Social Welfare Development in the 1980s and 1990s', in Roger Goodman, Gordon White and Huck-ju Kwon (eds), The East Asian Welfare Model: Welfare Orientalism and the State. London: Routledge, pp. 159-74.
Collick, Martin (1988) 'Social Policy: Pressures and Responses', in J. A. A. Stockwin (ed.), Dynamic and Immobilist Politics in Japan. Honolulu: University of Hawaii Press, pp. 205-36.
Deyo, Frederic C. (1989) Beneath the Miracle: Labor Subordination in the New Asian Industrialism. Berkeley, CA: University of California Press.
Deyo, Frederic C. (1992) 'The Political Economy of Social Policy Formation: East Asia's Newly Industrialized Countries', in Richard P. Appelbaum and Jeffrey Henderson (eds), States and Development in the Asian Pacific Rim. Newbury Park CA: Sage, pp. 289-306.
Esping-Andersen, Gosta (1990) The Three Worlds of Welfare Capitalism. Cambridge: Polity.
Esping-Andersen, Gosta (1997) 'Hybrid or unique? The Japanese welfare state between Europe and America', Journal of European Social Policy 7, 179-89.
Goodman, Roger and Peng, Ito (1996) 'The East Asian Welfare States: Peripatetic Learning, Adaptive Change, and Nation-Building', in Gosta Esping-Andersen (ed.), Welfare States in Transition: National Adaptations in Global Economies. London: Sage, pp. 192-224.
Goodman, Roger, White, Gordon and Kwon, Huck-ju (1997) 'East Asian Social Policy: A Model to Emulate?', in Margaret May, Edward Brunsdon and Gary Craig (eds), Social Policy Review 9. London: Social Policy Association, pp. 359-80.
Gould, Arthur (1993) Capitalist Welfare Systems: a Comparison of Japan, Britain and Sweden. London: Longman.
Henderson, Jeffrey and Appelbaum, Richard P. (1992) 'Situating the State in the East Asian Development Process', in Richard P. Appelbaum and Jeffrey Henderson (eds), States and Development in the Asian Pacific Rim. Newbury Park CA: Sage, pp. 1-26.
Johnson, Chalmers (1982) MITI and the Japanese Miracle: The Growth of Industrial Policy, 1925-1975. Stanford CA: Stanford University Press.
Johnson, Chalmers (1987) 'Political Institutions and Economic Performance: The Government-Business Relationship in Japan, South Korea, and Taiwan', in Frederic C. Deyo (ed.), The Political Economy of the New Asian Industrialism. Ithaca NY: Cornell University Press, pp. 136-64.
Jones, Catherine (1990) 'Hong Kong, Singapore, South Korea and Taiwan: Oikonomic Welfare States', Government and Opposition 25, 446-62.
Jones, Catherine (1993) 'The Pacific Challenge: Confucian Welfare States', in Catherine Jones (ed.), New Perspectives on the Welfare State in Europe. London: Routledge, pp. 198-217.
Katzenstein, Peter (1985) Small States in World Markets. Ithaca NY: Cornell University Press.
Ku, Yuen-wen (1995) 'The development of state welfare in the Asian NICs with special reference to Taiwan', Social Policy and Administration 29, 345-64.
Ku, Yuen-wen (1998) 'Can we afford it? The development of national health insurance in Taiwan', in Roger Goodman, Gordon White and Huck-ju Kwon (eds), The East Asian Welfare Model: Welfare Orientalism and the State. London: Routledge, pp. 119-38.
Kwon, Huck-ju (1997) 'Beyond European welfare regimes: comparative perspectives on East Asian welfare systems', Journal of Social Policy 26, 467-84.
Kwon, Huck-ju (1998a) 'Democracy and the Politics of Social Welfare: A Comparative Analysis of Welfare Systems in East Asia', in Roger Goodman, Gordon White and Huck-ju Kwon (eds), The East Asian Welfare Model: Welfare Orientalism and the State. London: Routledge, pp. 27-74.
Kwon, Huck-ju (1998b) 'The South Korean National Pension Programme: Fulfilling its Promise?', in Roger Goodman, Gordon White and Huck-ju Kwon (eds), The East Asian Welfare Model: Welfare Orientalism and the State. London: Routledge, pp. 106-18.
Kwon, Huck-ju (1999) 'Inadequate policy or operational failure? The potential crisis of the Korean National Pension Programme', Social Policy and Administration 33, 20-38.
Lau, Siu-kai (1982) Society and Politics in Hong Kong. Hong Kong: Chinese University Press.
MacPherson, Stewart (1993) 'Social security in Hong Kong', Social Policy and Administration 27, 50-7.
PRODUCTIVIST WELFARE CAPITALISM
723
Midgley, James (1986) 'Industrialization and welfare: the case of the four little tigers', Social Policy and Administration 20, 225-38.
O'Connor, James (1973) The Fiscal Crisis of the State. New York: St Martin's Press.
Ramesh, M. (1995) 'Social security in South Korea and Singapore: explaining the differences', Social Policy and Administration 29, 228-40.
Skocpol, Theda (1985) 'Bringing the State Back in: Strategies of Analysis in Current Research', in Peter B. Evans, Dietrich R. Rueschemeyer and Theda Skocpol (eds), Bringing the State Back In. Cambridge: Cambridge University Press, pp. 3-43.
Tremewan, Christopher (1998) 'Welfare and Governance: Public Housing under Singapore's Party-State', in Roger Goodman, Gordon White and Huck-ju Kwon (eds), The East Asian Welfare Model: Welfare Orientalism and the State. London: Routledge, pp. 77-105.
Vogel, Ezra F. (1991) The Four Little Dragons: The Spread of Industrialization in East Asia. Cambridge MA: Harvard University Press.
Wade, Robert (1990) Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization. Princeton NJ: Princeton University Press.
Waltz, Kenneth N. (1979) Theory of International Politics. New York: McGraw-Hill.
White, Gordon and Goodman, Roger (1998) 'Welfare Orientalism and the Search for an East Asian Welfare Model', in Roger Goodman, Gordon White and Huck-ju Kwon (eds), The East Asian Welfare Model: Welfare Orientalism and the State. London: Routledge, pp. 3-24.
Wilding, Paul (1997) 'Social policy and social development in Hong Kong', Asian Journal of Public Administration 19, 244-75.
Wilding, Paul (2000) review of Roger Goodman, Gordon White and Huck-ju Kwon (eds), The East Asian Welfare Model: Welfare Orientalism and the State, Public Administration and Policy 7, forthcoming.
Wilding, Paul, Huque, Ahmed Shafiqul and Tao Lai Po-wah, Julia (eds) (1997) Social Policy in Hong Kong. Cheltenham: Edward Elgar.
Wilding, Paul and Mok, Ka-ho (2000) 'Hong Kong', in Pete Alcock and Gary Craig (eds), International Social Policy. Basingstoke: Macmillan, forthcoming.