Edited by Peter Taylor-Gooby MittuniversitetetN Biblioteket J 831 25 Östersund^/ OXFORD ERSITY PRESS Contents List of Figures and Tables x List of Abbreviations xi Notes on the Contributors xii r i^:New Risks and Social Change i ^ Peter Taybr-Gooby 2. New Social Risks in a Conservative Welfare State: The Case of Germany 29 Andreas Aust and Frank Bonker 3. The UK—A Test Case for the Liberal Welfare State? 55 Peter Taylor-Gooby and Trine P. Larsen 4. New Risks—Are They Still New for the Nordic Welfare States? 83 Virpi Timonen 5. France: A New World of Welfare for New Social Risks? 111 Bruno Palier and Christelle Mandin 6. Spain's Transition to New Risks: A Farewell to 'Superwomen' 133 Luis Moreno 7. Switzerland: Negotiating a New Welfare State in a Fragmented Political System 157 Giuliano Bonoli ',sT^ew Risks at the EU Level; A Spillover from '"•-'Open Market Policies? 1S1 Trine P. Larsen and Peter Taylor-Goaty { 9}. New Social Risks and Welfare States: New Paradigm and v"" New Politics? 209 Peter Taylcr-Gooby Index 239 I Peter Taylor-Gooby introduction: The Post-Industrial Welfare State This book presents a new perspective on the question of what to cio about the welfare state. Changes in population, family structure, labour markets, and in the coming to maturity of expensive welfare systems present formidable challenges to the current welfare settlement. For the most part, discussion of these developments focuses on the pressures they impose on existing benefits and services (see, for example, Pierson, 20010, b; Scharpf and Schmidt, 2000; Ferrera and Rhodes, 2000a, b; Kuhnle, 2000). Key questions in the literature concern the maintenance of standards and the politics of retrenchment, realignment, and recalibration. In this book, we consider different and parallel changes that require new policies and offer the possibility of more positive directions in welfare reform—the emerging policy agenda of new social risks. The development of European welfare states in the 1950s, 1960s, and SVs-lOj 1970s took place under highly favourable circumstances, aided by four -——- key factors: a /golden age' of relatively continuous growth in economies characterised by large stable manufacturing sectors which provided high levels of family-wage employment for the mass of the population;%table nuclear family structures which supplied care for young children, frail older people, and other dependent groupsr^overnments able to manage their national economies through broadly neo-Keynesian policies which achieved continuing low unemployment and secure wages;^and political^ systems in which coalitions'^ able toj>ress effectively for the provision of benefits and services to meet their needs and in whichjheJ^_cj^nse^ This chapter has benefited from debates in the WRAMSOC group, and particularly from PTftfnrfffl Hic/*iic-;™<- —i-K ---d---1:---j c---i~i------1.-----............ ... |-"L > LR IA f LuR-GOOB Y H i W K i'. K NU Mh" I.a.I ( 7- legitimated. These circumstances favoured the development of a family of European welfare states characterised by a specific division between the appropriate spheres of public and private action. As an ideal type, the main business_of the welfare statgJnJndu^Mal society was to provide for needs which wgrj^nm_gdequately met through thj^market— interrupjion_ofjnc20^ (retirement, unemployment, sickness, or disability) and mismatch between incomeandjm^ cycle (e.g. child endowment)—or for needs^where^statep yndelY_recognised as desirable (e.g. highly-valued services in areas where the costs of privately checking professional expertise are high, such as health care or education). Social care, however, was mostly provided through the family system. Interventions in the family were limited and the corresponding services weakly developed in most countries. The hut *u5 -Jwrtf outcome was the Keynes-Beveridge or Keynes-Bismarck welfare state: gov-' ^ernments managedeconomies to promote full employment and organised socialprOTisiqn for needs whichjnarket and famijydid not meet. Things have changed. In an ideal typical post-industrial society, • economic growth rates are lower and more uncertain,. Technological changes mean that stable employment irrthg manufacturing sector is no longerayailabl£_ojn_a mass scale, with implications for the job security of semi- and unskilled workers and for class structure and the political interests i associated with it. Stricter competition promoted by economic globalisa-i tion has advanced labour market flexibility. The fact that women have ; succeeded in gaining greater advancement in education and in employ-■ ment and are continuing Co press for more equal opportunities means that j traditional unwagedsocial care based on a gender dmsion_ofJUihQux. ! imposesstrainson the family (EU20000, b; Daly, 2000: 490). | Thesechanges create new social risks and a new reform agenda for the ' welfare state. This book analyses the emergence of these risks in a range of I Europeanlocieties and examines political responses to them by govern-1 ments and at the European Union level. In this chapter, we define new social risks, consider how they relate to old social risks, identify the implications for citizens and government, discuss how new risks are to be analysed, review the emergence of new risks and policy response to them at the national and EU level and consider the contribution of reforms in this area to innovation in welfare state policy-making. New Social Risks: A Definition New social risks arethe__risks_that people now face in the course of their lives as a result of the economic and social changes associated A 2 with the transition to a post-industrial society. Four processes are of importance: First, women have moved into paid work in large numbers, while the proportion of men who are economically active is falling. Men's labour force participation in EU countries fell from 89 per cent in 1970 to 7S per cent by 2001, while women's participation rose from 45 to 61 per cent (OECD 2001b, 2002a). One powerful driver is the importance of two earners to maintain a satisfactory family income. Another is the increasing demand from women for greater equality in access to education and to independent employment. Analysis of Luxembourg Income Study data shows that new sociaI.risks emerg£jnos£_a£utejyfor lower skilled women who find most difficuktv_jnj^ and M^dTtemm^aiiccMjnjhks (Cantillonet al.,2001: p.447). Second, the increase in the absolute and relative numbers of elderly people has implications for social care as well as for the cost of traditional welfare state pensions and health services. The ratio of those over sixty-five to the population of working age in Europe Is projected to rise by 73 per cent between 2000 and 2030 (OECD, 20orfe: 27). Most domestic care is still provided by women. Just over twice as many women as men spend time on care tasks for children and older dependants; women of family age (20-49 years old) who have care responsibilities for children spend about 46 hours a week on them compared to 22 hours for comparable men. Corresponding figures for those aged 50-64 who care for older dependants are 22 and r6 hours, respectively (Eurostat, 20020: tables A.17 and A.19). Traditional patterns of care hnpose stresses on women seekjTig_p_aid work and generate .adgmand for provision from alternative sources—men, the private sector, and the state. Care responsibilities also impact on employment and on incomes. Data from the 1998 ECHP show that, for couple households with dependent children, 90 per cent of men of prime working age (20-49) are in employment compared with only 57 per cent of women. When we turn to older couple households with care responsibilities for dependent older people, employment rates fall to 47 per cent for men and 29 per cent for women—the same two-thirds ratio of women to men (Eurostat 20020: table A.21). The impact of care responsibilities pn, wpmen^ejco^ turn affectsthe risk of family poverty. The Luxembourg Income Study data show that poverty rates for couple households in the European Union where only one partner is in paid work are between three and six times higher than those where both work and here the effect is most marked in liberal countries with their weaker benefit systems (Esping-Andersen, 2002: table 2.5). n 11 2\ Third, labour market changes (to do primarily with technical develop- .' raents in production, which have reduced the proportion of unskilled manualjqbs in industry, and secondarily with th^growth in scale and. intensity of cross-national competition, which allows countrtojmth lower pay levels to use their comjjaj^twg^^ work) have tightened the link between education and employment. This in turn affects therisk of social exclusion among those with poor education. Those with a minimum level of education are about two and a half times more likely to be unemployed and nearly five times more likely to be in long-term poverty compared with those who have attended university (Eurostat 2000, tables 2 and 3; OECD, 2002^: table D). Education and skill levels are also linked to progress in work and in quality of working life. The Employment in Europe survey in 1996 showed that 47 per cent of employees had experienced a significant increase in the skill level of their job in the Last five years, and virtually none a decrease. Skill increases are much more likely higher up the occupational ladder and the lower skilled more likely to anticipate insecurity and unemployment (Gallie, 2002: H3-18). The fourth change lies in the expansion of private services resulting / primarily from attempts to constrain state spending to meet the pressures on the old risk welfare state listed in the first paragraph. While privatisation is not in itself a risk, it can generate new risks when citizen-consumers commit themselves to unsatisfactory ^°j£g_s' and when refflilajionj^ standards in private provision is ineffective. Thejshift towards the private sector has been most marked in responses to the pressures on state pensions. A number of countries are also providing benefits which widen access to private provision as part of their care strategy for children and frail older people. ■/'■'J- The United Kingdom, which already has the most extensive private t filar pensions system, has gone furthest, by radically weakening the state second-w^e) pillar pension and developing private alternatives. Non-state pensions, which currently provide 60 per cent of income for the top 30 per cent of retired people are intended to provide 60 per cent for all by 2050 (OECD, 2000: figure 4.1; DSS, rp98). The Netherlands also has extensive second-pillar private provision. Other European countries are pursuing private pensions as supplements to state pensions. Germany provides subsidies and strong encouragement, Sweden requires workers to invest in complementary private pensions and Switzerland has well-established compulsory occupational pensions. In France, long-term tax-subsidised savings schemes are in process of implementation. The OECD concludes that the strengthening qfgrivate pensions is the most im porta nUrendjii thecurrent reform of pension systems (OECD, 2000: 46). Promoting the growth of occupational and private pensions is a major element in the EU's strategy for modernising social protection^ EU, 2002a: 38). The UK private pension scandals are well known (chapter 3). While problems on a similar scale have not emerged elsewhere, successful regulation of new private services is essential to avoid escalating risks for more vulnerable groups (Esping-Andersen, 2002:16-17). This brief review of the way in which changes associated with the post-industrial transition affect people's lives indicates that more vulnerable groups are likely to experience new needs in three areas: In relation to changes in the family and gender roles: • balancing paid work and family responsibilities, especially child care « being called on for care for a frail elderly relative, or becoming frail and lacking family support. In relation to labour market_cjianges^ 'l • lacking the skills necessary to gain access to an adequately paid and ;! secure job 1« having skills and training that become obsolete and being unable to upgrade them through life long learning. In relation to welfare state change^ Ij • using private provision that supplies an insecure or inadequate pension ^ or unsatisfactory services. We go on to consider the context in which new social risks have emerged on the policy agenda. New Social Risks, Old Social Risks, and Pressures on Government At the same time as the new risks deriwd_imm_ihe transition to_p_ost-industrialism are confronting the citizens of European welfare states^ the established structures designed to cope with the social needs generated by industrialism and centred on interruptions to the family wage are facing pressures from different directions. The crisis of the welfare state has been extensively analysed (e.g. Ferrera and Rhodes, 2000; Kuhnle, 2000; Pierson, 2001a, b; Scharpf and Schmidt, 2001; Taylor-Gooby, 2001. 2002). It involves three factors, operating simultaneously: rising demand, restricted resources, and constraints on the capacity of government to reconcile the two. The most important pressure derives from population ageing as a result of rising life expectancy and lower birth rates. The proportion of the EU population aged sixty five and over rose from 13 to 17 per cent between 1980 and 2000 and is expected to rise to 27 per cent by 2040. The proportion of working age (15-64) increased slightly from 65 to 67 per cent in the earlier period but is expected to fall to 5S per cent in the latter (EU, 2002a; 30). Since spending on old age and survivors' benefits is far and away the largest proportion of welfare state spending (44 per cent for the European Union as a whole) and that on health care (which is mainly consumed by older people) is the second largest, at 22 per cent (EU, 2002a: table 2), and since provision for older people is financed mainly by the taxes and social contributions of those of working age, these changes in age structure imply substantial financial pressures. Demands increase just as resources come under constraint. Drawing on the work of Baumol (1967), Rowthorne and Ramaswamy (1997). and Iversen and Wren (1998a, b), Pierson (2001a: 84) argues that the shift in the labour force from the manujactujingto the service sectoMn the post-industrial transition also involves_aJojxg^teimJxencUx^ tivity. This has implicationsjpr economic growth. Productivity rates in the major economies fell through the 1970s, 1980s, and 1990s. Growth rates, which were between 2.7 and 4.6 per cent for the four largest European economies (France, Germany, Italy, and the United Kingdom) for the decade to 1973, fell to between r.i and 2.7 for the following decade, then to 0.4-1.7 per cent between 19S9 and 1995. A rally in the late 1990s collapsed after the bursting of the dot.com bubble so that rates for 1996-2002 were between 1.4 and 2.6 per cent. The current outlook is one of weak growth (Calderon, 2001; OECD, zoosd, 2003: table A.t). ( At the same time, governments' capacity to manage key aspects of their economies is weakened by the implications for tax and subsidy regimes of increasing capital mobility and thejmj^;tj}fjT^^ lation on exchange rates, so that djmand^icle^^ ment have become less attractive to poiicy-makers (Scharpf and Schmidt, 2000: 51-68). For EU member states these effects are reinforced by the 'open market' policies of the 1986 Single European Act, the 1994 Maastricht Treaty, and the constraints on state borrowing of the 1997 Growth and Stability Pact. The fact that the working class is jio^onger^ orgajiised through mass ej^loym^^ the effect of greater income inequality (Esping-Andersen, 2002: table 2.2) and greater diversity in working lives, and the emergence of new political pressures in societies that have been enriched through continuing growth (involving age cleavages, gender interests, divisions over environmental issues, and other concerns) tend to erode the solidarity that supported and legitimated the traditional welfare state (Beck, 1992: ch. 2). Pierson argues that welfare states, faced with growing needs and mount-ing pressures on resources, face a future of permanent aUstenty. The loss of some of the levers by which national governments managed their economies and the]miits_to solidarity make the task of funding expensive state services more difficult. The services developed by the industrial welfare state to meet the needs associated with lack of access to market income_(pensions, sickness, disability, and unempioymenFbeliefitJan^'so on) are everywhere undergoing reform. Measures to constrain spending by curtailing entitlements, expanding the role of the private sector or setting more stringent eligibility conditions are widespread (Myles and Pierson, 200r: 312). Market-based systems are introduced to control spending on health services through competitive pressures (Rico, Saltman, and Boerma, 2003: 599-600; Rathwell, 1998; ch. 17). An indication of the impact of these policies is that average benefits for each person of pensionable age in the EU-15 fell by 2 per cent between 1994 and 1999; unemployment compensation per capita fell by one per cent a year between 1990 and 1994, accelerating to 2.5 per cent between 1994 and 1999 (£U, 20020: charts 9,10, and n). This approach stresses austerity as the overriding consideration in welfare reform. However, analysis of welfarestatedevelopment must also include the societal changes of post-industrialisation that are creating a second set ofdemands_^ a second reform agenda in the context of the transition to post-industrialism. From the individual perspective these changes emerge as 'new sociaTnsks/, concerned with access to employment and opportunities in work, and with managing the conflicting pressures of ^HiMzJi^^~i2£^L£?£ej_Pi!^ work, and career. From_the welfare state per-sp_etagjM3^^ rather than with health care needs or retirement pensions. New risks in relation to long-term care and pension reform may become more significant in the future. 3- For those groups who successfully manage the transition ijyjo^d^epL^loj^^. ment or develop strategies to meet care needs, new social risks cease to'be so pressmg. "01e!Y-J£EI£^^ without access to adequate training or education, or who are unable to draw on child or elder care from the family or the state. Old social risks, to do with retirement and ill health, were more likely to form part of the continuing life-experience of the mass working class of industrial society. This has implications for the politics of new social risks. 4- They_jnyojvgJxrth labour market and family life, and thus extend demand for state intervention into areas of life that had been seen as private from an old risks perspective, redrawing the public/private boundary, and raising normative issues of where in the family the responsibilities for generating income and providing care lie. New risks are significant but transitory and particular. They open up new issues about the moral assumptions behind welfare state policy-making. The Perspective of Government The new risks generate new constellations of interests, which cross-cut old social risk constituencies in complex ways. They bear particularly on womenJ_j3n_ younger workers, and on those without relevant skills. One hypothesis is that newcross-class alliances will emerge to pufj>uejwjfju^ state restructuring in a more diffuse life-politics' {Giddens, I994: 48-9)-Those most affectejlaj£_ty4iicj^^ linked in families with partners and_parents who do not face thejpressures directly, which may explain why such groupings are only weakly developed. The reform process is likely to involve groups other than the immediate -risk-bearers, and the interests of groups such as employers in relation to t child care may lead policy in particular directions. Policy-jTiakmgJn^ welfare state essentially involved the designand implementation of programmes to meet needs that giarket incomes did not satisfy. The welfare states that developed operate mainly through horizontal transfer over the life cycle, with somex^rtical transfer to poor minorities. Such arrangements depend on intergenerationa I and (some) social solidarity^During the 'golden age' redistribution was relatively painless because the costs of transfers could be mitigated for the immediate losers by growth: higher incomes outweighed rising tax, an option unavailable in an age of 'permanent austerity'. Under current circumstances, the old risk politics of welfare is in large part about the gradual erosion of commitments to continue the expansion of state spending through processes of recommodification, retrenchment, recalibration, and rationalisation (Pierson, 2001: 455}. The most important example is the ^"T^' adjustment of pension entitlements as schemes mature through such ' tactics as the raising of pension ages, increased stringency of payment calculations, a reduction in early retirement, and a greater role for the private sector (Hinrichs, 2oor; Myles, 2002; Bonoli, 2003I. In this context, the zero-sum game of net tax:p_ayer and net benefit tVvny\S recipient (or rather, those who believe themselves to be net payers and i recipients) is particularly intense. This brings factors which inhibit reform to the fore—governments wish to avoid blame for unpopular reforms and interested parties seek to veto or delay change. Differences in institutional structures influence the extent to which these factors operate. An approach which places stress on new risks alongside the old risks of established welfare states must consider a modified agenda of political divisions and social values. Fro_rnjhe_perspective oi_goy£inment the main differ-gr^es^betweennewand old risks a_reas follows. 1. Old risk policies tend to involve financial commitments requiring &t+wew*-> substantial tax and social insurance contributions. For this reason, and ^ aA*^ because they involve mainly horizontal redistribution, they tend to require a politics of solidarity, for example through the notion of a 'generational contract', 'risk-pooling', or a state that provides 'cradle-to-grave' care or offers a 'people's home'. Since the overwhelming majority of the population of industrial societies thought they might need the services supplied, such solidarity could be mobilised. Virtually all attitude surveys (see, for example, Ferrera 1993; Kaase and Newton 1996; 'nevysiL. K., .'Hi K lAfJJP, ^uUli , Svallfors and Taylor-Gooby 1999; ch. 1; van Oorschot 2000) concur that pensions, health care and provision for disabled people—the major programmes to meet needs not met through the market—enjoy high legitimacy. Government in the transition to post-industrialism faces the problem of how to justify cut^acks_and constraint in these areas while developing new risk services that tend to go to minorities. This involves the tactics of blame avoidance and cumulative change charted by Pierson (1994) and others, and is expressed in the enormous resilience of these services against change. 2. New risks affect particular subgroups at particular life stagey most keenly. They involve political divisions that do not map easily onto the traditional class and party structures and are likely to involve alliances with other social actors interested in the expansion of the workforce and in enhancing national economic competitiveness. If old social risk welfare was often seen as the °JJtc£]^gQ^g_^en}ocrat|5 Ji^^pI^SS^.' (Korpi, 1983), new risk welfare "prcigrarnrne^^ the intergstsjm^nched by the outcome of that_struggle. 3. Oldjjskjw^^ primarily to support people atstages in their lives whenn^edswe^no^metthrough the wage relationship. They thus involve substantial transfer expenditure and may be seen as an economic burden at a time of stringency. New risk policies are often concerned to help more people support themselves through paid work. They may form part of a national strategy to mobilise a greater proportion of the population and to enhance economic competitiveness in a globalised market, and open up particular agendas for business and unions. To the extent that policy focuses on these issues, labour market reform predominates and child care becomes a more important issue than care for older people. About 42 per cent of women of prime working age (between twenty and forty-nine) are involved in child care as against 6 per cent in elder care {Eurostat 2002«: table A.17). 4. New risk policies meet needs mainly by encouraging and enabling dif-ferent choices and behaviour patterns rather than providing benefits. They are concerned with the engagement of the citizen in paid work and with changes in the pattern of family life. They involve issues of responsibility for providing income and for domestic care that cut across the boundary between public and private spheres. New risk politics directs attention to issues of legitimation and moral values. 5. Because new risks are less likely to involve the entrenched interests, major expenditures, and neo-Keynesian apparatus of interventionism that concern national governments and more likely to involve equal access to employment, the balance of family and work and the issues of training and education that concern an open market in labour, the European Union is likely tojggk a strongerj;oJg_[n this area thanir has in relation to old risks. New social risk activity will also enable the European Union to intervene directly in citizen's lives and may help to repair the 'democratic deficit' (Richardson, 2001: p. xv). The politics of old social risk policy-making concern the extent to which ^od'Iccs welfare states are able to resolve the tensions that emerge between different of 0 ^ groups when governments seek to retrench or contain spending on highly popular policies. Interest focuses on the extent to which it is possible to construct agreements which allow the interests of labour, business, and welfare state service users to be reconciled and to contain the burden of financing provision (Pochet 1999; Rhodes 2001; Hemerijck 2002), and on examples of successful accommodation such as the 'Dutch miracle' (Hemerijk and Visser, 2000). -prtckes New risk welfare politics is concerned primarily with mobilising the ofntW/ population to enhance competitiveness and with expanding opportunities < SKs and changing behaviour and assumptions about responsibilities. While I the mass services of the traditional welfare state generate their own con- * stituency, new social risk cleavages are much more likely to cross-cut existing social divisions. At the same time, new possibilities for employers and for those trade unions which represent workers most_ajfj^d. jyjiew risks toform alliances in response to_the shifts in the labour market _emerge. An important theoretical concern is with changing modes of economic i regulation and social roles in the family. This raises questions of how public policy innovations are legitimated and of how shifts in the approach of policy-makers and of business and unions are to be understood. Paradigms and the Legitimation of Policy Change A convenient approach to the policy stance of a government is provided by the notion of policy paradigm. Borrowed from Kuhn's influential work on the development of core ideas shared by communities of natural scientists (r97o)» the notion of policy paradigm is used to sum up the shared core beliefs of a policy community. It includes ideas about the goals of policy; the identification of issues as problematic in relation to those goals; explanations of why problems arise; solutions to identified problems; explanations of why they will meet the problem; and definitions of the appropriate role for government and other actors (Hall, 1993: 279). ' I ! L K i A I L 0 I-; UJL'C The Keynesian paradigm was dominant in industrial countries for much of the post-war period. This approach included an account of the workings of political economy and of why a market system might be subject to unevenness and shortfalls in growth through mismatch in the availability of capita] and investment opportunities at particular times (leading to depression or inflation), together with a recipe for economic intervention that explained how government could intervene through manipulation of interest and exchange rates in order to stimulate or restrain the economy. In this context, the traditional welfare state was legitimated as intervention which redistributed to groupswith a high propensity to consume, especially at times of demand deficit, and which enabled_goyernment to regulate demand through social programmes and social [nf^a^rucjure^ projects. Meeting political pressures from the working class and other interest groups could thus be reconciled with policies which secured the conditions for continued economic growth. After the oil crisis of the 1970s, policy-makers in many countries lost confidence in this paradigm as traditional approaches to economic management became more difficult to pursue. The intemationaiisation of financial markets reduced the capacity toinanipulate interest rates and_ imdemuned governments' ability to dictate exchange rates; stronger international competition, reinforced by the growing importance of the European Union as a free trade zone, influenced the extent to which' the return to labour in cash and social wage could be determined in response to internal political demands without regard to developments elsewhere (Scharpf and Schmidt, 2000: ch. r). Policy-makers sought for new paradigms. The chief response among economists was monetarism, which str&isesjie^^ poiicies_ rather than employment or^growth^ objectives. The fact that 'Germany's success with a pragmatic version of monetary policy that emphasized a strong and stable currency provided policy-makers with a powerful example to emulate' at the time, ensured that broadly monetarist approaches became the dominant paradigm in policy-making (McNamara 199S: 6). Monetarism enjoins governments to pursue the welfare of citizens byreduLing economic intervention and containing inflationary pressures by; manir>ulatmj^he.money supply as the primary tool of economic management, rather than working directly on the level of demand. Throughout Europe, political economic paradigms were increasingly influenced by the assumption that the objectives of the welfare state are best advanced by ensuring that the market works efficiently, rather than expanding the provision of benefits and services. Such an approach is implicit in the priority given to low inflation, low budgetary deficits and low public debt in the f{ Maastricht treaty criteria for membership of EMU. As a recent President of the European Central Bank put it: Greater flexibility in labour, product and financial markets together with sound fiscal positions and wage moderation will support the objective of maintaining price stability and will create stable_conditioris to foster employment creation. Such an interaction of policies... is the best possible way to enhance the long-term welfare of the citizens of the Euro area. (Duisenberg, 2002) It is also reflected in the original guidelines for the European Employment i Strategy, in which the four sections deal with increasing employ-ability, . developing entrepreneurship, encouraging aaVa^ta^fit^ and strengthen- [ ing equal opportunities policies (EU 1998, Annex), rather than reflation, job-creation, and passive benefits. From this perspective new social risks, which direct policy-makers' attention to labour market change and issues surrounding women's greater involvement in paid work, offer opportunities which may fii. the developing paradigms of social policy-makers more closely than the high tax/ high service spending approach of the industrial society welfare state. The paradjgm-shift m ecQnomic_golicy also plays an important role in the legitimation of new developments in social policy^ New risk policy-ma king is bound up with new approaches__to economic policy. We move on to consider new risks in the context of theoretical' accounts of the welfare state. Analysing New Social Risks: Welfare State Regimes The regime categorisation of welfare states has proved remarkably durable in providing a framework for the analysis of social policy. It distinguishes Nordic social democratic welfare states, where entitlement is based on citizenship principles and where the objective is a high level of universal protection against social risks; Continental corporatist welfare states, based on social insurance systems, where levels of provision are generally high, but the social hierarchies of work are reflected in entitlement; Mediterranean welfare states where pensions, health care, and education are highly developed and other services are now catching up with average ..European Union levels and where the family plays a prominent role; and the liberal regime, where state provision is more limited and targeted and private market systems are encouraged to a greater extent. The regime categorisation '4 14 DLVV KI'jKi AND SOCIAL CliANui also corresponds to distinctions in levels of welfare. Nordic regimes tend to have lower rates of poverty and inequality and better protection for vulnerable groups than corporatist or Mediterranean regimes, with libera! regimes faring worst (Esping-Andersen, 2002: table 2.3). The hierarchy of social opportunities follows a similar pattern (Korpi, 2000). Thus the regime framework applies both to policy mechanisms and to the outcomes achieved in welfare states. This approach has provided the backbone for work focused primarily on old risks, whether globalisation and labourjnarket shifts are seen as the chief problem (Scharpf and Schmidt, 2000), P^£H^£^Lf!^£i!y (Pierson, 20ora, b) or population ageing and the rigidity of existingsys-tems (Ferrera and Rhodes, 2000). It requires some modification to analyse the emergence of new risks and the way in which welfare states address them. The regime perspective focuses on the way different social groups, mainly class-based, contest the allocation of the resources generatedby economic growth, drawing on 'industrial society' and 'power resources' theories. The key concepts deployed are 'decommodification' (how far welfare provision reduces the domination of citizens' lives by market forces) and 'stratification' (how far class inequalities are modified by or simply reflected in the redistribution of resources by the welfare system— Esping-Andersen, 1990: 3). - Family and gender issues were relegated to a subsidiary role in the traditional welfare settlement, on the assumption thac family provision for breadwinner husbands also met the needs of their wives (see, for example, Beveridge, 1942: 9-11). Policies to manage new social risks address_ the opportunities and burdensfbrwojmer^ had previously been tackled via Keynesian economic policies and welfare had been largely concerned to reallocate market resources to those whose needs were not met through the market wage system. New risk policies are concerned strongly with meetingjieedsjhrough the mobilisation of labour, with direct targeted interventions to support wages, where these arp inadpqTiatppvnrrw^ services .In^area^jormedy addressed via the state. 'Recommodification' and 'flexibility' emerge along-■ s'ide decommodification and stratification as key concepts for analysing welfare reform. We now examine patterns of welfare state policy inputs and outcomes in relation to family and employment in more detail, to see whether regime categorisation needs modification to provide an adequate framework for understanding the emergence of policies to meet new social risks. New Social Risks in Balancing Work and Family Life State social policies differ in the assumptions implicit in them about women's activities and in the division of responsibility for waged work in the labour market and unwaged domestic work necessary to provide care for children and older people and maintain a home. Initial approaches to analysing this issue reflected the gender cleavages between the male productive and female reproductive and domestic spheres of ideal typical industrial society. Thus Lewis categorised European welfare states in the 19S0S according to the extent to which policy assumes a gender division between male breadwinner and female homemaker (Lewis, 1993: ch.i). Similar approaches have been elaborated by Orloff (1993), Bussemaker and Kersbergen (1994), Rubery (1999)- Knijn and Kremer (1997) and others. More recent analyses point to the increasing recognition of women as workers in a number of welfare states, but indicate that social provision is in most cases ill-adapted to support mothers who also participate actively in paid work. 'Broadly speaking, states have tended Lo recognise women for the purpose of social entitlements either as mothers or as workers, in a rare case, like France, both ... have been recognised. In Sweden in the late twentieth century, recognition as mothers has been grafted onto recognition as workers. Where the male breadwinner model still has major purchase, then women find that their position as paid workers is at best a matter of secondary concern (Britain and Germany) and at worst actively discouraged (Italy)' (Lewis, 1998:15). The chief input in this area is the level of statespending on services which_ relieve the care burden on the family_and^a^icuj;arly on_hquse-mves_fr^m^bdjdjind^elder care. Outcomes are indicated by the extent to which women are able to participate in paid work alongside men. Table 1.1 shows that the inclusive Nordic citizenship welfare states committed substantially more than any other group of European countries to care services for children and older people between 1980 and r999, spending roughly three times the EU average on elder care and twice as much on family services. Care spending in the corporatist heartland of Europe has roughly doubled. The liberal-oriented states continue to rely . heavily on market and family provision. Spending in the Mediterranean states has expanded rapidly from a low base and now exceeds the liberal group in family care. When the statistics are analysed on a national basis, France stands out through the level of spending on child care support, 10 iJL I CK l.-.t'LüK tjÜüD < Table i.i Trends in state spending on new social risks (% GOP 1980-99) 19S0 1999 Services Services Active Services Services Active for for labour for for labour elderly families market elderly families market and support and support disabled disabled Nordic 1.77 1.60 0.8S 2-73 1.78 1.67 Corpora tist 0.46 0.38 0.13 0.75 0.74 1.14 Liberal 0.53 0.29 0.28 0.59 0.33 0-74 Mediterranean 0.08 0.04 0.02 0.25 0.37 0.47 EU-15 0.65 0-55 0.25 0.98 0.S3 1.00 Notes: Nordic: Sweden, Denmark, Finland; Corpoiatist: Austria, France, Germany, Belgium, Netherlands; Liberal: UK, Ireland; Mediterranean: Greece, Italy, Portugal, Spain. EU average is based on all fifteen members. Services for elderly and disabled include: residential care, home-help services, day care, and other services In kind. Services for families Include: formal day care, personal services, household services, other family benefits in kind. Active Labour Market Policies include: training, youth measures, subsidised employment, employment measures for disabled people, employment services. Source: Calculated from OECD (aoora) which at 1.23 per cent of GDP in 1999 is midway between the corporatist and Nordic groups. This pattern of spending on care is broadly reflected in the extent to which women with family responsibilities participate in paid employment. Couples where both members are engagedinpaid work may pursue one of four employment strategies—the 'dual earner' model, in which both women and men particTpate more equally in paid employment, the 'mod-ified industrial' model, which casts the male as a full-time and the woman as a part-time worker, the 'dual carer', in which both work part time and share domestic care work equally, and the 'r^rsejndustnal', in which the woman works full time and the man part time. The more flexible work and family roles of ideal typical post-industrial society are reflected in progress towards the dual breadwinner or dualcarerjrtodels and in the extent to wTuTh^ociaTpolicy seeks to sustain this (Lewis, 2001; see also Daly, 2001). The European Labour Force Surviy^xamines the employment patterns of dual earner households with and without dependent children. Nowhere N' VV KIM '-, ANI> V_>< . MANi.if do more than two per cent of working couple households follow the dual carer model with both partners in part-time work. Similarly, fewer than 2 per cent of working couples follow the reversed traditional pattern of female full-time, male part-time employment {Eurostat 2002&). For the vast majority, the division is between the modified industrial and the dual bread-winner pattern. The proportion of dual earner households is rising across Europe, as part of the shift towards a post-industrial society, and accounts for at least 60 per cent of all couple households supported through work everywhere except in Greece, Italy, and Ireland, where it exceeds 45 per cent. In all cases, among working couples without children at least two-and-a-half times as many follow the dual earner as follow the modified industrial pattern. However, among those with children, there is a rapid shift to the modified industrial model, which predominates for this group in the United Kingdom, Germany, and the Netherlands, and is widespread in the other corporatist and liberal countries with the exception of France (Eurostat 2002b). Table 1.2 shows how motherhood reduces women's engagement with work ro a very limited extent in Nordic countries, but has a much greater impact in the corporatist, liberal, and particularly Mediterranean context. In corporatist and liberal countries, mothers who remain in work are much more likely to do so part time. More detailed analysis of trends between 1984 and 1999 in employment patterns in couple households with a young dependent child (under six) points to a wider impact of child care on women's paid work across the corporatist countries and the United Kingdom. Everywhere there is a trend away from the traditional model of man in full-time work, woman as full-time unwaged mother, which was dominant in 1984. However, in the Mediterranean countries the majority of those mothers who work do so full time, while in corporatist continental Europe and the United Kingdom, mothers tend to work part time. France with its extensive child support policies is again the exception (OECD 2001c: table 4.2). In relation to long-term care, a recent European study shows that midlife informal care is associated with 'reductions in work hours or exit from the labour force' which is not recovered after care responsibilities end (Spiess and Schneider (2002:33). Regime theory broadly corresponds to patterns of both welfare state inputs and of outcomes in relation to the balance of care and paid work for women and men. However, these outcomes are produced through differen t means as the following chapters show. In Nordic countries, state support iS ŕ'l ' i R TA 'ŕ L C' R (.,( U.',, i Table 1.2 Women's employment and child care % Women aged 25-55 in Share of part-time employment employment, 2000 for women, aged 25-55, 2001 No One Two All women Those with a children child children child under six Denmark 78.S 88.1 77-2 20.8 6.1 Finland 79-2 73.5 73-5 n.a. n.a. Sweden 81.9 80.Ó 81.S n.a. n.a. Austria 76.0 65.7 24.8 50.4 Belgium 65.6 71.S 69-3 33-4 45-0 France 73-5 74-1 ss.s 23-S 3