Chapter 1 made the case for the continued relevance of the universal social democratic welfare state based on de-commodification, of which the Swedish welfare state is closest approximation to the ideal type (Esping-Andersen 1985a). The argument was not that the social democratic welfare state is without its problems with regards to pluralism and democracy. But, invoking particularly Himmelstrand et.al (1981) and Rothstein (1998), the argument was that the institutional form of the social democratic welfare state provided a propitious framework and context for a socio-political transformation that could address the legitimation-problems of contemporary advanced capitalist society. Universalism with high levels of entitlement are required to ensure that policy implementation remain simple (and hence avoiding bureaucratic 'governability crisis'); to ensure that state intervention does not become intrusive and infringes on diverse lifestyles (distinction between entitlement levels and monistic implementation); and to sustain solidarity between the middle class and the most vulnerable in society. In short, contra Giddens, chapter one constitutes an advocacy of the project of a humanisation of capitalism without apologies.

But chapter 1 did not address the question of the economic viability of this welfare state in the era of 'global competition'. In other words, Giddens' neo-liberal argument against the economic rationality of 'traditional social democracy' has yet to be addressed. That is the objective of this chapter. The chapter will argue that the 'traditional' social democratic project of a humanisation of capitalism is still a viable project in the 'late-modern', 'post-industrial', or more to the point, 'post-Fordist' world. What is more, an analysis of Swedish political economy in the 1980s and the first part of the 1990s that goes beyond the superficialities of the mainstream, demonstrates this.

'Post-Industrialism', Models of welfare capitalism and global competition

Whilst globalisation is an important theme for Giddens in The Third Way, his argument against the economic rationality of traditional social democracy was in essence a microeconomic one. His critique focuses on the level of the individual and her performance on the labour market. Welfare benefits, based on the principles of social citizenship and risk-minimisation generate, according to Giddens, economically dysfunctional behaviour since these schemes encourage individuals to consume social services excessively whilst they perform sub-optimally and dysfunctionally as factors of production. Individuals tend to reduce job-search efforts; they prolong claims to unemployment insurance; they call in sick too often or remain on sick-leave for too long a time and hence they prolong claims to health-insurance; and, they have incentives to claim pensions prematurely. Such microeconomic irrationalities then undermine the competitiveness of corporations in the global marketplace. As a result, the material basis of the welfare state is undermined. 'Globalisation' understood in terms of reduction of trade barriers and barriers to capital movements have reduced the possibility to tolerate such economically dysfunctional behaviour. (Giddens 1998: cf. Lindbeck 1995; see also Lindbeck et.al. 1994).

I will return to Giddens argument later in this chapter and seek to refute it with reference to the Swedish case. The argument cannot be addressed immediately because it does not reflect the best possible case against the type of welfare state in question. Giddens account does not analyse in sufficient detail the socio-economic pressures that the welfare state faces. Furthermore his account of the welfare state is too stylised and does not analyse in detail important variations in different types of welfare states. There are, however, more nuanced and detailed analyses of the welfare state that are in basic agreement with Giddens. These must be addressed before I can provide a more convincing defence of the socio-economic rationality of the social democratic welfare state. Gøsta Esping-Andersen's (1996; 1999) recent work is in basic agreement with Giddens, and this work should perhaps be taken especially seriously since Esping-Andersen is an erstwhile avid defender of the decommodifying universal social democratic welfare state. Indeed, the concept itself can be attributed primarily to him in present-day research and debate.

According to Esping-Andersen, welfare states face a dramatically increased trade-off between social protection, equality and economic rationality, because of changing demographics, fundamental shifts in the capitalist socio-economic structure and increased global competition. Esping-Andersen argues that advanced welfare states face a basic dilemma: How should they cope with ageing populations, lower fertility rates, and differentiated ‘life-cycles’ on the one hand, and with a slower rate of growth that he (invoking Baumol 1969) suggests is inherent in the shift from a manufacturing-centred ‘industrial’ economy, to a service-centred ‘post-industrial’ economy on the other, whilst they at the same time face the increased constraints of global competition. For Esping-Andersen, the essence of the problem is one of demographics. With a projected 50 percent increase in the age-dependency ratio between 1996 and 2020, given present-day entitlements and rates of economic growth, welfare state capacities will be over-stretched. To meet current commitments, he estimates that an increase of social expenditure equivalent to an extra 5-7 percent share of GDP would be required. This indicates the severe fiscal strain that secular demographic trend exert on welfare states (Esping-Andersen 1996a: 7 cf. European Community 1993: 24 and OECD 1988). Given the imperatives of global competition, increased taxation hardly constitutes a feasible route towards raising the required additional resources, according to Esping-Andersen. Increased rates of productivity growth could, however, provide the basis for the required additional expenditure. Productivity growth would need to underwrite an additional increase of 0.5-1.2 percent of real taxable earnings on top what current rates of productivity growth allows. This would be sufficient to finance additional expenditure (Esping-Andersen, 1996a: 7).

But Esping-Andersen suggests that such additional productivity and real social wage growth is difficult to achieve in the post-industrial economy. In post-industrialism there is an increased trade-off between employment, equality and high rates of productivity growth. This is because high value added jobs in manufacturing are becoming increasingly scarce as a result of increased capital intensity, generated by the introduction of automation-technology. The new jobs are increasingly to be found in the low productivity service sector, where wages consequently have to be lower. The alternative to the creation of a low-wage economy is mass-unemployment, which further exacerbates the tendencies towards fiscal crisis, because of increased claims on unemployment insurance funds. Given his conclusion that the aforementioned tradeoffs have become more severe, he advocates welfare and labour market reforms that are similar to those suggested by Giddens.

The 'Post-Industrial' Dilemma and Esping-Andersen's Three 'Policy-Regimes' of Welfare Capitalism

Esping-Andersen concretises his argument with reference to different types of advanced welfare capitalism. For this purpose, he uses the famous typology of three ‘policy regimes’ that he has developed elsewhere (Esping-Andersen 1985a; 1990: esp. 26-32, 38-54). In that work, Esping-Andersen showed that the distinct institutional characteristics of different types of welfare state regimes were determined by, and tended to reproduce, distinct socio-political political power structures. One of Esping-Andersen's ideal types is the 'conservative', or 'Christian democratic' welfare-policy regime, which finds its approximate concrete expression in the continental European states (such as Germany, Austria and the Netherlands). In this model, conservative social forces, connected with the state and the church, and with their roots in pre-capitalist society, took the lead in welfare state development in the late 19^th century. Following the lead of the 'Bismarck reforms' of 1880s, these forces mediated between the forces of capitalist development and the emerging socialist working class, in order to ensure that the disintegrative effects of the former did not lead to a socialist revolution. The essential social purpose in this context was to counteract what they considered to be the morally degrading and anomic effects of capitalist commodification, and to ensure that the proletariat was organically integrated into national society as a loyal subject of the state. This type of welfare policy regime, then, keeps commodification in check in order to maintain 'traditional values' such as the family, church, and the respect towards preordained status and hierarchy between social groups. Here workers are granted corporative status and protection in the national community, in exchange for 'responsible behaviour'. The main components of social policy in this regime are corporatist interest intermediation in collective wage-bargaining and, and occupation-based (status-determined) social insurance.

The liberal welfare policy regime is another of Esping-Andersen's ideal types, and it approximates developments in, for example, the USA, Britain and Canada. Here the raison d'etre of welfare policy is to maintain and reproduce the ethos of possessive individualism. To be sure, also the liberal welfare state was designed in terms of 'public goods' to protect people against contingencies such as illness, accidents and unemployment. But the protection - simulating individual insurance-contracts - was designed to maintain rather than mitigate the cash-nexus (between individual contribution and benefits) and the individual wage labourer as a privileged reference for socio-economic identity. This type of welfare state tolerates trade unions under the guise of 'freedom of association', but is reluctant to give them official status or acknowledge the wage relation as an 'unfree' relation that requires a collective mobilisation of workers to counteract the power of capitalists. Welfare policy is organised along voluntary, individualist-actuarialist lines and public intervention is kept to a minimum. The state only contravenes the market logic through a residual means-tested 'safety net' - social assistance, intended for those 'problem cases' that are not capable to provide for their own protection through market-performance. This corresponds to what, following Titmuss (1971), was referred to as the residual welfare state in the end of chapter 1.

Esping-Andersen reserves the label 'social democratic' welfare policy regime to the ideal type that approximates welfare state design in the Scandinavian countries. It was only in these countries that the labour movement of the second international become hegemonic and succeeded in making welfare policy reflect the concrete strategy of the politically conscious reformist working class. The essential raison d'etre of the social democratic policy regime is to provide a social income or wage that is independent of wage labour, and that is granted as a universal entitlement of 'social citizenship'. Such 'decommodification' was intended to generate working class unity and loyalty to the social democratic project and to extend the appeal of this project beyond the working class to 'the people' in general. Like the conservative regime, this regime provides decommonification, but rather than allocating particularist entitlements and assigning particularist 'places' for social groups in society, this decommodification-strategy attempts to project universalism. The policies of this decommodification include transfer payment and services that provide universal entitlements or income maintenance, a collective bargaining regime supporting solidaristic wage policy, flanked by a macroeconomic policy commitment that blunts the disciplinary stick of the threat of unemployment by providing an unconditional commitment to full employment.

These welfare state regimes were developed in the course of development of industrial capitalism and the institutionalisation of distinct 'Fordist' models of economic growth. However, Esping-Andersen argues that these institutional arrangements have also set the framework for distinct, path-dependent, trajectories of 'post-industrial' capitalist development, some of which Esping-Andersen considers to be more viable than others (Ibid: 191-217; 221-29). He argues that the Christian democratic welfare policy regime of continental Europe faces some fundamental contradictions. These result from the fact that this institutional form has not responded to the decrease of labour demand in the industrial manufacturing sector, generated by increased capital intensity and automation, by an offsetting increase of labour-demand in an expanding post-industrial service sector. Rather there has been a reduction of the labour supply and the volume of employment.[i].

Two institutional determinants, both intimately related to the shape of welfare state arrangements, are held to be central in this context: the 'rigid' nature of wage-determination in the corporatist labour market; and, social policy encouraging the reproduction of the traditional family and the male breadwinner model (Esping-Andersen 1996b). Apart from failing to address poverty generated by the increased incidence of breakdown of the traditional family[ii], these determinants have generated an unsustainable trajectory of ‘jobless growth’, characterised by mass unemployment and fiscal crisis. Comparatively high wages and 'rigid' negotiated wages, asserted by the corporatist collective bargaining regime on the economy as a whole, have led to capital intensive corporate restructuring in the export-oriented manufacturing sector. At the same time, these high wage-levels have retarded the development of the service sector, given the limits to productivity growth in services. Furthermore, this type of welfare state deliberately encourages the maintenance of the traditional family, and has not pursued heavy public investments in, for example, public childcare that might have boosted the service sector. Whilst this development results in high productivity growth and high wages for those who remain employed (the ‘insiders’ of the labour market), labour is to a significant extent shedded. Christian democratic welfare states have responded to this excess labour-supply, through schemes of early retirement, reduction of working hours, and high levels of unemployment insurance payments. The problem is, however, that these benefits are funded through increases of payroll-taxes (employers' and employees' contributions), which increase wage costs further, and compel companies to pursue labour shedding further. According to Esping-Andersen, the continental model is engaged in a hopeless battle to remain fiscally solvent, as growth rates no longer can be maintained to generate sufficient revenue for its transfer payment programmes. This is especially so, because increased capital intensity means that the payroll-tax revenue/growth rate ratio tends to decrease whilst demographic changes increase demands for expenditure and reduce revenue.[iii] Additionally, the unemployment problem becomes increasingly acute, which generates further fiscal pressure. The scope to switch to alternative forms of taxation, such as the taxation of capital, is held to be very limited in an era of global competition and transnational capital mobility (Esping-Andersen, 1996b; see also Streeck, 1995).

According to Esping-Andersen, the neo-liberal route of deregulation of the Anglo-Saxon residual welfare state as practised in the United States and Britain in the 1980s and the 1990s is a more viable response to post-industrialism. Here, curtailment of union-rights, labour market deregulation, and a reduction of social benefits have resulted in an increase of employment-growth and a reduction of unemployment. This has created the conditions for the development of a labour-intensive service sector, which depends on low wage flexible labour but that also offers high skill and high (albeit often flexible and insecure) wages at the apex of its occupational stratum.[iv] Esping-Andersen argues that fiscal pressures and unemployment problems can be resolved in this model. The emerging service sector has proved to be capable of absorbing surplus labour, which, in turn, off-loads fiscal burdens from the state. Furthermore, pensions become increasingly a private matter, generated through private savings and managed by mutual funds. Indeed, pension-fund management plays an important role in the emerging service economy. In this context, the difference between the normative and institutional relationship to the traditional family in the liberal and the conservative regimes is of critical importance. In the liberal regime, there has been a dramatic increase of the labour force participation rate and the employment-rate of women - in large part in low-end service jobs, but also in some high wage professional jobs (Esping-Andersen 1990: 201-02)[v]. This increases the tax-base, which significantly contributes to the relief of the aforementioned demographic pressures on social expenditure (Esping-Andersen 1996: ).

On the basis of this comparison Esping-Andersen argues that re-commodification, retrenchment, and a certain convergence towards the liberal welfare-policy regime will be necessary if welfare capitalism is to be stabilised. Compared to Giddens however, he is refreshingly blunt about what the distributive implications are of this type restructuring will be. In a pessimistic tone, he predicts that the result of such neo-liberalisation will be increased polarisation and segmentation between core and periphery in the labour market, (leaving little room for Giddens’ ‘social cohesion’ on the Third Way). Drawing on the American and British experience he points out that the trend towards ‘flexible’ and reduced wages has produced ‘unprecedented levels of poverty’ (Esping-Andersen, 1996: 8). He (rightly) does not even entertain the prospect that this type of welfare state might generate more participatory forms of democracy.

Esping-Andersen's recent work is also strikingly pessimistic about the post-industrial prospects of the social democratic welfare-policy regime. Here there has been a dramatic shift in his argument between the publication of The Three Worlds of Welfare Capitalism in 1990 and Welfare States in Transition in 1996. Indeed, in the former work he argued that the social democratic policy regime did provide the basis of a viable post-industrialism, which maintained and perhaps even enforced and extended the norms or de-commodification and social citizenship. He argued that mass unemployment could be avoided in the social democratic model through tax-funded public investments in labour-intensive social welfare services. He pointed especially towards the expansion in ‘reproductive’ services and female employment in Sweden in the 1970s and the 1980s, ensured by public child-care programmes, education, and in the expansion of public care and services for the elderly. Employment expansion proceeds without significant wage-segmentation in this policy-regime, because public service workers are unionised, and wages are determined through inter-sectoral, co-ordinated, and solidaristic wage bargaining. Resources are thereby transferred to the low productivity, service sector, oriented towards the national economy, from the export oriented high-productivity manufacturing sector. This transfer is ensured in two ways: First, high taxes (income taxes, corporate taxes, sales taxes and payroll taxes) are used to finance public service sector production. Second coordinatied and solidaristic wage-policy implies that workers in the export-oriented manufacturing sector forfeit some of the wage increases that productivity increases in their sectors would grant, in favour of wage increases above productivity rates in the public service sector.

Esping-Andersen supported his argument with some striking empirical evidence that set Swedish developments since 1960 apart from the developments in the USA and Germany. In contrast to Germany, in Sweden, high and negotiated wage growth in the manufacturing sector did not correlate with reduced participation rates and employment. Rather 'post-industrial' employment grew at the same rate in Sweden as in the USA, with a dramatic increase of female employment growth in 'health, education and welfare' making up for low rates, comparable to the German rates, in the low end consumer service sector. The total employment rate also increased Sweden, although at a lower rate than in the USA. Moreover, a significant part of employment growth was growth in part-time employment. On the other hand, participation rates in Sweden were much higher even than in the USA, and the part-time ratio was significantly lower than in Germany[vi].

In The Three Worlds Esping-Andersen argued that this social democratic form of post-industrial welfare capitalism also was fiscally viable. Tolerance of higher tax-rates obviously helped and full employment obviously contained unemployment insurance expenditure. But furthermore the welfare policy regime ensured a more sustainable transition from the traditional family structure. High female participation rates and exceptionally and high fertility rates would ensure an adequate expansion of the labour supply, a higher tax base and a lower dependency ratio, which in turn would address the problem of demographic pressures. These developments were facilitated by public childcare and extensive provisions for parental leave that would allow women to 'harmonise careers with fertility' (Esping-Andersen 1996: 13, see also 76)

It is evident, however, that Esping-Andersen was impressed by the rapid developments towards severe fiscal imbalances, high levels of unemployment and years of negative GDP-growth that emerged in Sweden almost immediately after the publication of Three Worlds (Figure 2.1). In his subsequent dismissal of the socio-economic viability of this policy-regime, the same microeconomic interpretation appears as we know from the work of Giddens, relying on the same school of neo-classical economists for the diagnosis. Invoking the moral hazard argument we encountered with Giddens and Lindbeck, he states that 'high wages and taxes are widely believed [sic.] to spur negative work incentives and hidden employment' (Ibid 13 n 15), and he concludes:

…[the] achilles heel of the system….is the growing tax burden that a huge public labour market incurs. With high rates of productivity growth the system can be sustained: when productivity of private investments are sluggish, severe cost problems emerge. This is exactly the situation that especially Sweden faces today: declining fiscal capacity combined with rising pressures on public job creation/or income maintenance. Swedish policy makers and unionists can no longer avoid wage flexibility and major social benefit cuts. (Ibid: 13)

I will return to and critique this 'belief', for which, Esping-Andersen admits in a footnote, that it hard to 'come by hard evidence' (Ibid: 13 n 15) in later in this chapter. But before this can be done, it is necessary to discuss in further detail Esping-Andersen's prescriptions for the continental 'Christian Democratic' welfare states. Such a discussion is important, because European social democratic 'modernisers', in charge of welfare state restructuring in Europe at the time of publication of this book, depart from conclusions like those generated by Esping-Andersen's analysis. They have ruled out a politics of convergence towards the Scandinavian social democratic welfare policy regimes. Instead, they seek the resolution in a partial convergence with the Anglo-Saxon type of liberal/residual welfare state. In this context, the practical experiences of Dutch restructuring (the 'Polder Model') have generated particular interest among social democratic modernisers, not the least in Europe's economically dominant state, Germany.

Neo-Liberalism with a Human Face?

Esping-Andersen concludes that the realities of post-industrialism requires a convergence towards the Anglo-American policy regime despite its costs in terms of poverty and inequality. However, he does hold up the hope that if this policy is modified with one component of the social democratic model, an active labour market policy of retraining of the workforce, then the worst of the inequalities can be checked. Indeed, insofar as people are likely to change jobs more often over the life-cycle, 'lifelong learning' might result in equality over the life-cycle as people move from low-skill to high-skill employment (Esping-Andersen 1996: 261-65.) This would amount to a kind of 'neo-liberalism with a human face'. Here Esping-Andersen's prescriptions dovetail with those of the social democratic 'modernisers', who provide the current intellectual input for welfare state reform in continental Europe. A prominent representative of these intellectuals is Wolfgang Streeck, a Max Planck Institut professor and advisor of the German Social Democratic government, who is involved in the tripartite negotiations 'Alliance for Jobs' (Bündnis für Arbeit). Streeck would merely add that 'responsible' corporatist bargaining also has a role to play, if unions accept wage-segmentation and 'competitive corporatism' in exchange for a reduction of unemployment and an effort to avoid the starkest inequalities (eg. Riester and Streeck 1997; Streeck 1998). For these analysts, the Dutch 'Polder Model' of restructuring in the 1990s has been held up as a possible case of ‘best practice’ to emulate. More broadly, the Dutch Model has become a favourite case in social democratic modernisation discourse in Germany and elsewhere. For example, its merits have been encanted over the last few years in fora such as Die Zeit , the left-liberal/social democratic 'quality' weekly broad-sheet newspaper, published in Hamburg. Here, the 'Polder model' is seen as an affirmation of the possibility to implement market oriented, 'supply-side' reforms whilst maintain a somewhat retrenched, but nevertheless well developed welfare state (eg. Perner 1999).

While explicitly avoiding the term 'Polder Model' and the hubris surrounding the aforementioned discussion in the German press, Visser and Hemerijck's (1997) detailed study - in part inspired by German fascination - provides a benevolent account of the Dutch welfare state restructuring the in the 1980s and the 1990s. Theirs is a story about how a Christian democratic continental, welfare state has managed to generate employment growth and resolve fiscal crisis. (For some basic economic indicators, see Figure 2.2) A crucial reason behind this 'relative success' has been wage moderation, negotiated between the corporatist ‘social partners’ (a process started by the Wassenaar Accord in 1982.) Wage moderation resulted, in part, from an exchange between unions and business. But it was also in part the result of an exchange between unions and the government, where the latter reduced taxes and the level of non-wage charges for social insurance as a reward for wage-moderation. This proved to be possible due to an increase in the tax-base generated by an increase of part-time work in the service sector. This, reduction of non-wage costs, in turn, increased demand for social services which contributed to further service employment. Fundamentally, this development included negotiated wage segmentation, where wages of new jobs were set at low increments on the negotiated wage-scale. Here the government used the threat of not accepting the negotiated settlements as the wage norm for the non-unionised sector to enforce the usage of the lower increments. According to Visser and Hemerijck, the increase of part-time work has exactly generated female employment, and made it possible to combine child rearing with paid employment.

Another important aspect of Dutch reform has been a tightening of the eligibility criteria for long term health insurance and early retirement, to shift social policy from income maintenance of the long term unemployed towards employability. Finally, thirdly, the Dutch state has started to pursue an active labour market policy and retraining, through public and private labour bureaus. These also contribute towards wage segmentation since they are used as temp-agencies by employers, whose lower wages and benefits are regulated through the contracts of the bureaus, and not directly by the businesses for whom the worker in question perform their work (Ibid).

No doubt, the Netherlands remains a more egalitarian and welfarist society than, for example, the United States and therefore Dutch fiscal and macroeconomic re-stabilisation should not be rubbished. Also, the combination of part-time work with child-rearing is not without interest when one considers question of work-sharing, demands for civic participation and reproduction in a post-traditional society. Therefore, Visser and Hemerijck's modest pronouncement of 'relative success' is more convincing than the hubris generally associated with the 'Polder Model'.

The achievements of the Dutch Model ought to be relativised, however. Employment rates may have increased. But this is from one of the absolutely lowest levels in Europe. Figure 2.2, showing the rate of employment of the total adult population shows that it is dubious to suggest that the 'flexible' Netherlands was better than the 'rigid' Sweden in generating work, even at the depths of the unemployment crisis of the latter and the height of success of the former. Second, the higher employment levels in the Netherlands come at the price of increased income polarisation. Visser and Hemerijck may be correct to argue that Dutch wage-inequality still lies somewhere 'between' Sweden and Germany on the one hand, and the USA and the UK on the other. But the percentage increase of wage dispersion in the Netherlands (as measured by ‘90-10 ratios’) was just slightly below that of the UK and much higher than that of the USA, and these countries stand out as the 'income-polarisers' of the OECD countries (Figure 2.3).[vii] Considering the P90/P10 ratios is calculated in terms of full-time employment and the extent that Dutch employment relies on part time employment, this probably understates wage polarisation in the Netherlands. Third, the Dutch record of production and productivity growth is actually not particularly impressive. If one of the criteria of social democratic modernization is higher productivity growth, then the ‘Polder Model’ is a failure.[viii] Finally, claims that the Dutch Model is gender-progressive should be treated with great scepticism. The policies are not designed to relieve women from reproductive work. Rather, they seek to make waged work compatible with women’s (assumed) child rearing duties, and hence make it possible for them to take on a ‘double burden’. This is also a double burden they are increasingly compelled to take if their partners also are employed in precarious work. The main point here is, that even if progressive males agrees to share this burden, the work time a household requires for generating sufficient wages and reproductive labour has increased for the family as a whole. Capital and the state, on the other hand, incur no extra labour force reproduction costs, and average wage costs decrease. In other words, the rate of exploitation of women and men has increased.

In addition to this relativisation of Dutch welfare-state achievements, it is important to ascertain whether the Dutch Model could be emulated elsewhere. To their credit Visser and Hemerijck eschew making any claims to that effect (1997: 184). In the Dutch case, a facilitating condition seems to have been its export structure. The Netherlands has managed to reduce its energy import-costs as a result of gas-reserves in the North Sea. This has allowed the Netherlands to tie its currency to the Mark, and to maintain a low cost of imports. Consequently, wage moderation has been easier to achieve. The Netherlands also specialises in special niches of international service trade with a high value-added. In particular, the Netherlands functions as a service centre and a transportation-hub for the German economy. Finally, insofar as wage-competition with Germany has been important to the Dutch Model, this is of course not an option for Germany. More generally, undercutting the costs of others is necessarily a particularist strategy for someone in relation to another entity and it can therefore not be a general answer to the problem of post-industrialism.

The fallacies of the comparative political economy of post-industrialism and beyond: Regulation Theory and 'Post-Fordist' restructuring

The final point of the last section points to an important blind-spot of the one-sidedly comparativist focus of the literature reviewed so far in this chapter. This literature commits a fallacy of composition and fails to account for the problem of competitive austerity in contemporary capitalist restructuring. That is when each country reduces domestic demand and adopts an export oriented strategy of dumping its surplus production [by keeping wage increases below productivity growth and pushing down domestic costs], for which there are fewer consumers in its national economy given the decrease in workers’ living standards and productivity gains all going to the capitalists, in the world market. This has created a global demand crisis and the growth of surplus capacity across the business cycle.” (Albo 1994: 147).

Competitive austerity is central to the diagnosis of the problems of welfare capitalism of the so-called 'regulation school'. According to the regulation school, economic policy that accepts global competition uncritically as the adjudicator of economic rationality and that is based on one-sided supply-side strategies pre-empt more welfarist approaches to technological change. As a result its call for welfare state retrenchment becomes a self-fulfilling prophesy (Lipietz 1989: ).

It is also possible to invoke the regulation school to critique the notion of 'post-industrialsim' as a concept intended to make sense of capitalist restructuring. The problem with this notion is not so much that 'manufacturing that matters'; clearly, both the manufacturing and services are important in contemporary capitalism. The problem is rather that the notion of a 'service' is too heterogenuous to capture the dynamics of capitalism. It refers to too many different instances of economic activity to be useful as an abstract concept. No doubt, service employment has become more important in the last twenty years for advanced capitalism. But there is no single self-referential logic that has driven this development. Rather, the growth of service employment is linked to several logics that are contingently related. At the most, these logics are related to one another only through a set of highly mediated processes that are not identified by the post-industrialist literature (Sayer 1992). Some increases in service employment is due to the disembedding of the international monetary order, which has led to a sharp rise in the need for risk-management in financial services (Sassen 1991). Another increase of the demand for labour in services is indeed, as Esping-Andersen suggests, due to the breakdown of traditional family patterns, which generates demand for child-care, the care of the elderly, cleaning services etc. Some increase in the demand for services is due to income polarisation generated by neo-liberalism itself. For example, the rise of an intensively consuming middle class has led to the expansion of the entertainment industry and the provision of 'spectacles' (eg Davis 1984). Neo-liberalism has also generated an increased demand for services due to the increasingly punative methods which are associated with its methods of regulation of the marginalised population. This includes the increase of demand for prisons (Wacquant 1997). More generally a problem with the notion of post-industrialism is that it falseley assumes a determinist logic of socio-economic development.

A Theory of Capitalist Regulation

The regulation school provides a more fruitful approach to the analysis of capitalist restructuring. Regulation theory relates technological and industrial change to the particular social form of capitalism with its attendant contradictory and crisis-prone circuit of expanded reproduction. It explains how such change may or may not help sustain the coherence of this circuit and hence maintain order in capitalist society.

Michel Aglietta (1998) has succinctly explained the premises and key-concepts of this approach in the postscript of the new English edition of his seminal A Theory of Capitalist Regulation (1979). The theory starts from the premise that money is 'the primordial social link' in capitalism. Following, the analysis of Marx, this is because the link between actors takes place through the exchange of commodities, whose value is expressed through the 'universal equivalent' of money (Marx 1867: ). This makes capitalism growth dependent as well as inherently contradictory.

Capitalism is growth-dependent because it fundamentally depends on specialised financial institutions to provide credit for investments of economic activity, and the motivation of these institutions is to make more money out of money. This, in turn is a raison d'etre that is passed on to productive enterprises as an imperative, because they need to make profits, if for no other reason, to settle their debts to their creditors. This is expressed through Marx's famous formula M-C-M'.

This growth dependence tends to generate contradictions, and a rather basic contradiction is inherent in the system. The making of more money out of money is only materially possible on an aggregate (or 'social') level through the exploitation of wage labour, understood in the specific sense that labour as a subordinate class produces more than it receives in return in wages. This is possible in capitalism, because the private property regime has separated the workers from the ownership and control of the means of production that is necessary for their survival. As a result they sell themselves on the labour market for a price below the value that they actually produce in the labour process (because their 'supply' of labour-power is not determined by this value, but by their opportunity cost for not working, determined by the imperative to provide for the necessities of life). This is the root of the basic social contradiction in capitalism. This contradiction has a distinctly economic dimension, which constitutes the analytical focus of regulation theory: The imperative of making more money out of money requires a continuous expansion of aggregate consumption. However such expansion tends to be frustrated by the fact that individual capitalists necessarily minimise wage costs to sustain profits (the rate of labour exploitation). Hence, the prospects of the expansion of consumption required to realise (or 'validate') expanded reproduction tends to be undermined by the exploitation of labour, which also constitute the mass of consumers.

On the basis of this Marxist ontological position, regulation theory has developed its key concepts, regime of accumulation, mode of regulation and hegemonic project in order to explain how such a mode of production, which is so prone to instability, can sustain itself.

The economic dimension of the aforementioned basic contradiction of capitalism necessarily compels capitalists to constant innovation (a revolutionising of the means of production (pace Marx) or creative destruction (pace Schumpeter) to maintain profit-rates. Indeed without such innovation, expanded reproduction cannot take place and the system enters a state of crisis. Following Schumpeter, regulation theorists argue that such innovation tends to be defined by particular technological paradigms that prevail in particular epochs, with their attendant core-products and core production-processes (eg. the textile industry and the putting-out system in the mid-19th century, the automobile and the conveyor belt system since the 1930s). It is the nature of these norms of production that determine whether the basic contradiction of capitalism can or cannot be successfully counteracted, together with the norms of consumption, and together they constitute a particular régime of accumulation.[ix] Because of the contradictory role that people play in the system as consumers and wage labourers, there is no guarantee that paradigm that defines the norms of production automatically produces a compatible paradigm for the norms of consumption. It may be that competition compels firms to cut costs or increase productivity so as to generate surplus capacity. Indeed, regulation theory interprets the depression of the 1930s in these terms: a highly productive norm of (mass) production, based on industrial mechanisation had then been implemented, but the consumer demand was not sufficient to realise this production.

This tension between norms of production and consumption implies that the capitalist market economy needs to be encased in a broader stabilising institutional framework, which also needs to ensure basic requisites for a market economy such as the honouring of contracts and the respect for property. It is this institutional framework that regulation theorists call the mode of regulation. In part, it is of course possible that capitalists themselves solve aspects of these regulatory problems themselves through collective action, but for the most part the mode of regulation is constituted, and depends upon, collectively binding decisions within the territory of operation which can only be ensured through the sovereignty associated with the state. The exact object of regulation is contingent on politics and the prevailing ideology (Jessop 1990b: ). However, at the very least the capitalist mode of production demands that the mode of regulation deals with the regulation of the wage relation, the regulation of finance, the reproduction of the management of the currency, property and contracts. (Lipietz 1985: )

Regimes of accumulation can either be extensive or intensive. Intensive accumulation means that the making of more money out of money, that is expanded reproduction, is ensured through increased productivity (that is, through the increased value of output per unit of labour power). Extensive accumulation is based on an increase of the labour time required to ensure the reproduction of labour power, which manifests itself through a downward pressure on wages. The distinction is important from the point of view of class relations, because intensive accumulation regimes provide the potential for positive sum games between capital and labour, where wage- and profit-rates can increase simultaneously. Extensive accumulation implies a zero-sum game, where profitability only can be maintained through the suppression of aggregate wage-rate increases (Ibid. )

The meaning of the term regime of accumulation is often misunderstood in empiricist social science. It is important to point out that it is not a description of immediate empirical phenomena. Rather, it guides empirical investigations by reflecting on the more abstract generative structures and structural functional requisites inherent in capitalism and the relation of these to the technology of the forces of production. This is to provide more empirical concepts with sufficient analytical depth to reflect on conditions of existence that are not immediately observable (see Appendix pp. for further elaboration).

The analysis of empirical-historical manifestations of capitalism can only take place through the concepts of 'accumulation strategy', 'hegemonic project' and 'growth model' (Jessop 1990a; 1990b). Modes of regulation do not emerge automatically as functional effects of the imperatives of régimes of accumulation. They are rather the outcome of prevailing socio-political strategies advanced by concrete political subjects in competition, conflict and co-operation with other concrete political subjects over the nature of the collectively binding decisions of society. These subjects are understood in a Gramscian sense as conscious subjects, whose outlook, to be sure, depends on their particular social position and interests, but who reflect on the problems as they emanate from these positions in an intellectual and immanently creative way. Hence the subjective dimension is to be understood as having an independent effective reality, which interacts, and is not wholly determined by objectivity. 'Accumulation strategy' refers to the particular regulation strategy of the many possible strategies that is chosen to deal with problems of capital accumulation, and these depend on the successful articulation and advancement of a particular economic ideology (or 'hegemonic paradigm'). Since regulation depends on collectively binding decisions, accumulation strategies also need to be embedded in the state-institutions, whose legitimacy in society depends on the particular articulation of the 'general will' and the 'general interest' of society, that is, the particular articulation of a 'hegemonic paradigm' (Jessop 1990: 193-218; 311-16). The particular mode of regulation that thus is configured by the accumulation strategy and the hegemonic project regulates the 'growth model' of a society. The growth model is a more concrete expression of a particular economy of a particular society than the régime of accumulation, because it reflects on contingencies such as particular product specialisms, natural endowements and other individual peculiarities that are not captured by considerations of the capitalist mode of production in general, even considered in a particular phase of historical development.

Together the aforementioned concepts provide an analytical refinement of the Gramscian concept of 'historic bloc' as discussed in the introduction (see also Appendix ), which allows for a consideration of the terms for material reproduction in society.

Fordism and Post-Fordism

On the basis of this theoretical introduction, we can discuss the alternative conception of regulation theory to the question of the welfare state and 'post-industrialsim'. Rather than understanding contemporary restructuring in terms of 'post-industrialism', it understands it in terms of a transition from Fordist to post-Fordist regimes of accumulation.

According to regulation theory, the Fordist regime of accumulation constituted the socio-economic basis of the 'policy-regimes' of welfare capitalism identified by Esping-Andersen as well as their crisis. This regulationist account is rather well known: In the Fordist phase of capitalism, the norms of production was organised according to Taylorist principles, applying product specific machinery on a large scale. Adequate market outlets were in this context ensured through ex ante integrated regulation in the form of collective bargaining and Keynesian economic policy ensured by welfare national states, which constituted the mode of regulation. Accumulation was in this context intensive, and it provided the basis of a positive-sum game and 'historic compromises' between organised labour and capital. Mass consumption could thereby validate mass production, which in turn underwrote mass consumption through productivity increases (repressing inflationary tendencies in Keynesian regulation and more fundamentally the real costs of investment goods - in Marxian terms the 'organic composition of capital') (Aglietta 1979; Lipietz 1985: 35-41).

Stagflation signalled the breakdown of Fordist circuits of valorisation in the 1970s, While many factors intervened in the economic crisis (including, for example, the oil crises the breakdown of the Bretton Woods system, inflationary Keynesianism and mounting government debt), Lipietz argues that in the last instance the crisis was induced by Taylorism reaching its organisational frontier of production process innovation. Significant economies could no longer be yielded through the fragmentation of work-moments and the separation of conception from execution. Especially important in this context was the legitimation and motivation-constraints manifested through labour force resistance. (Workers resisted change through absenteeism, sabotage, or wildcat strikes) As a result, returns to capital decreased and continued demand expansion could not be underwritten through productivity increases. This expressed itself through stagflation and fiscal crisis (Ibid: 41-46).

Regulation theory understands the new phase of capitalism, not in terms of a single unfolding 'post-industrialism' but in a more open ended way, in terms of a set of potential 'post-Fordist' trajectories of capitalist accumulation. The prospects of these competing post-Fordist trajectories depends on supporting modes of regulation, accumulation strategies and hegemonic projects. The conclusion of regulation-theoretical research has been that there is nothing inherently neo-liberal implied in the logic of this contemporary dynamic of socio-economic restructuring. Regulation theoretical research has reached these conclusions in their search for emerging norms of production, through theoretically informed empirical research of different experimental corporate responses in the OECD region to the 'stagflation crisis' of the 1970s and the attendant profits-squeeze. These responses entailed different mixes of cost-cutting strategies and production-technological innovation, and they have set the conditions for a new phase of capitalist development based on new core products and processes.

At the most abstract level, 'Post-Fordist' production is based cybernetic automation of industrial processes, and a breakdown of information bottlenecks. This allows for increased capital intensity, a breakdown of information bottlenecks which allows for corporate organisation over a larger geographical space, a reconnection of conception and execution without productivity losses (especially through computer assisted integration of design, management and manufacturing ['CAD/CAM']), as well as more flexible adjustment to demand (Kaplinsky, 1984). One of the most important contribution of regulation theory has been their ideal-typical sketch of different possible régimes of post-Fordist capital accumulation. Here, macro-economic implications are inferred from the observation of microeconomic experimentation in the 1980s (Leborgne & Lipietz, 1988).

Certainly, flexible neo-liberalism provides the basis of one possible post-Fordist trajectory. Here the economies of CAD/CAM and 'general purpose machines' are facilitated through the elimination of collective bargaining and the usage of individual contracts, incentives and threats, and numerically flexible wages, all used to create an 'enterprise-corporate' culture. Flexible liberalism implies labour market polarisation, with a shrinking core work-force with enterprise-corporate contracts and a growing periphery of precariously employed. However, on the basis of corporate experiments in Sweden, Germany and Northern Italy, Leborgne and Lipietz argue that another trajectory is more compatible with trade unionism. In sharp contrast to what Esping-Andersen, drawing on neo-classical economics, assumes, the premise of the negotiated involvement model is that numerical flexibility of wages is not essential for post-Fordism, but that they can continue to be set through bi- or tripartite collective bargaining. Rather, co-determination provides an organisational form for functional flexibility - or, 'networking and skill to adjust volume to demand without prodictivity-losses' (Amin, 1994: p. 20-21) - compatible with the application of CAD/CAM technology and general purpose machines. In this model, active labour market policy facilitates workforce training and mobility, and collective bargaining provides for a solidaristic distribution of work, wages and leisure, in exchange for public goods such as the stable supply of a skilled workforce and public provisions cutting social overheads (such as health- and childcare). The final ideal-type is the neo-Taylorist model where new technology merely is used in a Fordist way to increase capital intensity and to extend further the separation between conception and execution.

Considered from the point of view of the functionalist requisites for expanded reproduction, the different models have different macroeconomic strengths and vulnerabilities. One central strength of the negotiated involvement model is that collectively bargained wages and social entitlements (operating as “automatic stabilisers”) could still integrate expanding production with demand and consumption ex ante. This would mean that flexible adjustment and economies of scope could be combined with other economies, like capacity utilisation and returns to scale, adequate investment levels, and a stable environment and time-horizons for “learning by doing. [x] At the same time, however, this model seems to presuppose stable and expanding aggregate demand. Apart from the role that demand-expansion in and of itself plays for productivity increases, this is because of the limited scope for cost cutting during contracting demand in the negotiated involvement scenario. Hence it requires a Keynesian dimension. This is further reinforced by the fact that it is associated with tight labour markets that increase the capacity of labour to sustain solidaristic wage policy and assert demands for meaningful co-determination in the production process (Sandberg, 1992; Pontusson, 1992). Furthermore, wage rigidity tendentially shifts the burden of adjustment to capital as it requires capital accumulation at lower profit/value added ratios, and/or as it sets higher demands for innovation - a “Schumpeterian dimension”. This, in turn sets a higher demand for stable and predictable access to sources of finance, and other public goods such as vocational training.

At the same time, non-market relations associated with the regions that have developed production based on negotiated involvement, may resolve “collective action” problems and provide positive externalities that facilitate exactly such economies. Corporatist arrangements, whether under the auspices of the Swedish Labour Market Board (AMS), or Germany’s apprenticeship system do facilitate the supply of skilled labour (Standing, 1988; Streeck, 1989). Positive externalities of a long term commitment to finance of production, and coordination and diffusion of innovation, (as well as the propensity to organized interest intermediation) are also generally associated with intimate, sustained, organisational “voice” linkages between financial institutions and firms (such as house-banks) (Zysman, 1983; Pontusson, 1992: pp. 152-53; Streeck, 1995). Such finance linkages are also generally associated with lower requisite levels interest payments to finance investments (Zysman, 1983), which counteracts the pressure on profits exerted by negotiated wages. Finally welfare state services, more developed in Scandinavia than in the continent, provide the foundation for labour intensive growth despite automation, operate as a stabiliser of aggregate demand, and can provide user-producer linkages in social systems of innovation (Edqvist & Lundvall, 1993).

A weakness of the flexible-liberal model is that (especially downward) wage flexibility set ex post may mean that it encounters difficulties in generating sufficient demand for a sustained new growth phase, and it may be associated with wide business-cycle fluctuations.[xi] It may also have difficulties in generating adequate public goods such as vocational training (Albo, 1994: pp. 150-53). But the former problem might possibly be compensated for by intensified consumption by the upper middle class, and especially a reduced turnover time in consumption in, especially, the service sector. Moreover, the financial sector has also become very flexible in terms of managing slumps and booms, through risk-management and credit extension (Harvey, 1990, Chs. 10 & 11).[xii] It should be noted, however, that these are services for which it exerts rents, although the financial sector itself is responsible for much of the risk (Strange, 1989).[xiii] One should also not forget territorial expansion as a possible mechanism to sustain accumulation.[xiv] Moreover of course, numerical wage flexibility allows for wage reductions as an instrument to sustain profit rates. Together these measures may prove to be sufficient to displace contradictions in capital accumulation.

The nature of the financial sector is a critical factor upon which the direction of post-Fordist capitalist development hinges. The presently predominant trend towards more market-based and globalised finance operates tendentially both as an obstacle to negotiated involvement and as and as an impetus for flexible liberalism. The balance of payments constraints and the shortening of time horizons in investment associated with globalisation and marketisation of finance, constitute severe obstacles to the welfarist, Keynesian as well as the Schumpeterian requisites of negotiated involvement. With regards to the latter requisite, increased markets for risk-management, and increased competition in corporate finance spurred on by informatics, provide incentives for a breakdown of voice-based links between banks and firms[xv], as well as a relative focus on financial transaction as such (including mergers and acquisition), as opposed to production process innovation.[xvi] With regards to the former requisites, rents exert inhibiting costs on the prospects for public sector expansion as well as fiscal and monetary stimuli, especially in the present context of high debt ratios. Whether this takes the form of higher deficits, increases in inflation (in relation to other currencies), or taxes, currency markets will punish such moves with higher interest rates and “risk-premiums” on currencies. Moreover, in a context when all states try to run surpluses to reduce their debt burden, the result is a reduction of demand-growth in the world economy and “competitive austerity”. At the same time, the increased importance of risk management as an economic activity, and the rent it exerts, provides the basis for an expanding financial services industry, organized along flexible liberal lines (Sassen, 1991). This may provide neo-liberal accumulation with a “core product”. The consequence of all these trends is that global finance operates as a formidable bias for flexible liberalism and against negotiated involvement. Hence, the question of regulation of financial markets is central to the question as to whether flexible liberalism or negotiated involvement prevails as the hegemonic régime of accumulation. In this context it is significant to note, as even Giddens grants, that the globalisation of finance was not unleashed by a predetermined endogenous economic logic, but rather by contingent political decisions and non-decisions (Strange 1989; Helleiner 1994) Hence, unlike social democratic modernisers we have good reasons to treat 'globalisation' not as an objective economic logic, but as a socially constructed phenomenon that can be questioned, and whose re-regulation ought to be a high priority on the political agenda.

Regarding the neo-Taylorist model, a regime of accumulation based “purely” is unlikely. Due to expensive overheads it requires stable markets outlets, but undermine these through increased capital intensity, unemployment, and a breakdown of the Fordist ex ante wage-relation. However, the model remains relevant because a “fallacy of composition” is possible (what is “irrational” on macro-level, might be rational for the individual components), engendering an “organic crisis of production”. But perhaps more importantly, there is evidence that suggests that no concrete case of economic restructuring follows the ideal types "purely", but articulates elements from all models.[xvii] Taylorist practices might survive as a subordinate mode in the dual economy or "sunset" sectors (notably in the consumer service sector). Moreover, corporate strategy may employ elements of the different models in attempts to combine economies of scope and scale. For example, corporations may combine flexible processes with scale and mergers designed to retain price-setting privileges in their markets (Amin & Malmberg, 1994). The mixing of Fordist/Taylorist elements makes scale more relevant, and decreases the capacity of market adaptation of flexible liberalism to fluctuating and sluggish demand. However, insofar as strategies of scale have become transnational, it also generates further balance of payments constraints for the reflationary aspects of a negotiated involvement strategy, as well as constraints on industrial policy

The crisis of the Swedish model reconsidered

We are now in a position where we can return and consider Esping-Andersen's rejection of the social democratic welfare policy regime as a viable response to post-industrial restructuring. Like Giddens critique of traditional social democracy, Esping-Andersen's argument is that the socio-economic crisis of social democracy can be reduced to 'moral hazard' that has generated dysfunctions on the labour market. To support this argument, he draws on empirical references to Sweden's economic performance in the 1990s. Drawing on the regulation-theoretical conception of post-Fordism, which was presented in the previous section, I will challenge this interpretation of the crisis of the Swedish model. There is no evidence that regulations of Swedish labour markets cause dysfunctional behaviour that is incompatible with technological change and global competition. Rather, I suggest that Esping-Andersen is to quick to give up on the social democratic model of 'post-industrialism' that he so eloquently formulated in the Three Worlds of Welfare Capitalism. My argument is that evidence suggests that Sweden provided a fertile ground for a post-Fordist regime of accumulation based on the norms of negotiated involvement in the 1980s. The subsequent crisis of the Swedish model was not due to moral hazard, but due to the failure to institutionalise a mode of regulation that could stabilise and support such a growth model. We cannot point to an intransitive logic to explain this failure. There is nothing inherently anti-social democratic in post-Fordism. Rather, the failure to institutionalise such a mode of regulation must be expalined with ref

© Routledge 2002. All rights reserved.

Read also Chapter 1: Recent Discourse on the 'Third Way'

• Dossier Globalisation