xii Divisions of Weife SA social assistance SACO white collar union federation in Sweden SAP Swed.sh Social Democratic Party JMP Statutory Maternity Pay SPn Slckness Insurance *™ Social Democratic Party (FRG) TEc Supplemental Security Income (US) ^ Training and Enterprise Council ™ unemployment insurance benefit 1 ,;) YOUth Traini«« CI____ • j-------■■( liiauidllĽt Youth Training Scheme 1 Annlysine and Comparing Welfare States Ki... .i.'-.i^p1" a critical appraisal of modern social policy in four w' ■ o. ^-.!'. •■ ! i the chapters on each welfare state there is a strong ■n -i* »■■■ i »■ íl ttata about social policies, largely from official ■i. ľi..: -".i"■ -"-a-js. but the approach is very far from being just an ■ „„■ i , .-ii :'i.-jniiition of data. In the selection and presentation of ..i-ľ-.i- is "'•■"■es have been exercised in order, wherever possible, to , ,-i ,.!. *. i'i; i. .'i. al themes about the origins, purposes and outcomes of v .t ■ si p- -n. •■ u- ^ - "tern societies, particularly in relation to the key social i!i\i-i.\i-* •■" «..•■^■- i ice' and gender. Needless to say, it is easy to ask such r,, i' .iii.,-ii,'i.- !l it frequently extremely difficult to answer them on the :-lľ.i- .'I -\.ii-i k d 'ta. In many ways all we have are pinpricks of light in a ...i.ir» ■ .i n. *> * hich are bathed in gratefully and exploited as fully as j ,.«,■■' 1 ■■». :*■1.1, r will hopefully be wary of instances where the data is ,i\.-: i'--i *i».:.*: ■ ' -;ven massaged. This chapter is devoted to examining •....in i I l.j i-i-.'1 ■ iological, conceptual and theoretical issues which such l'i . *■ .'i ::>■*." i-, i, in order to explain the approach taken in the rest of I li*. .-n-i» - .i . policy' and 'the welfare state' are virtually synonym-.i!i^ i if*.^ .■:.■ <.'■ i-entionally used to describe government action in the ik-i.i* "i p.:- m ' id family income, health care, housing, education and i ..nil!'. ■ i:.- \w nal 'care' services. 'Government action' embraces not ii:-l\ .li'.-ii pi.n .i.i of benefits and services, but also the regulation and v-u-Mi!'. i iii'u-l n., iiscal reliefs) of the various private forms of welfare. Mu -r I ľ u i .ink... occupational welfare provided by employers, welfare (■■ 'viii.v I», i ■■ jM.-fit, charitable, trade union, community, religious and " !m \ i ■" _ 11. i. s - ■ ■>!. inizations, as well as that provided informally by family IIKII- v> iik-ii "n ,i id neighbours. Clearly the boundaries of social policy 1 *il '• -I i " ■■'■ hich are conventionally ascribed to 'economic policy' " -. iMipi.iM m i industrial, monetary and fiscal policy) and other areas ■I "i i: 'I. ,» mi \ i. g. immigration, law enforcement, industrial relations ■ s 'I | .■!.■ pniu' i phe concept of 'policy' is also usefully extended in at .*■ k iv.- ..-V i .in. otions. First, it must cover the activities of agencies to ,'11'- ' "' ■»líni", i " and legislation frequently delegate responsibility for ",,-i '"'^ q. ingos, regional and local government, and so on. •■ ■ '.\». p'-ifessional and local government discretion often create us complexity in defining what policy actually amounts to. Second, 2 Divisions of Welfare the term 'social policy' should cover areas of inactivity by government an< its agencies in relation to social issues. Thus 'government inaction, o nondecision, becomes a policy when it is pursued over time in a fairl consistent way against pressures to the contrary' (Heidenheimer et al. 1990: 5). This is particularly apparent in areas such as family policy, when governments uphold the privacy of 'the family', while inevitably, at th same time, intervening in many ways into family life. From whatever value perspective it is pursued and however its boundaries are defined, the study of social policy involves analysis of three basic elements - the origins, the substance and the impact of policy, or as Heidenheimer et al. (1990: 3) put it, 'how, why and to what effect . . . governments pursue particular courses of action or inaction'. Pursued from a radical and critical perspective, policy analysis should attempt to keep all three elements in play, although this is a difficult juggling act. To examine the origins of policy is to understand the processes of policy change with a view to understanding how to shift policy in a more radical direction. In examining the substance of policy it is necessary to look critically behind the ostensible aims and functions of policy. Above all it is essential to examine the impact or outcomes of policy in terms of critical parameters such as class, 'race' and gender inequality without taking for granted benign intentions or successful implementation of policy. Analysis of the impact of social policy is therefore discussed first, followed by brief discussions on the origins and the substance of social policy. The Impact of Social Policy Both supporters and detractors of the welfare state would agree that one of its purposes is to heal fundamental social divisions or at least to mitigate social inequalities. This is essential to the idea or ideology of the welfare state, but such outcomes cannot be taken for granted. There is considerable evidence that, in fact, the welfare state institutionalizes class, gender and racial divisions and inequalities. Yet it is also true that without the welfare state, the extent of class, 'race' and gender inequalities and divisions would in most instances be even more substantial. This contradictory nature of the welfare state in both mitigating and furthering social inequalities and divisions is a key theme of this book/We start from the premise that in the United States (US), the FederafRepublic of Germany (FRG - known as West Germany until 1990), Sweden and Britain the welfare state operates in the context of a 'patriarchal and racially structured capitalism' (Williams, 1989: xiv). Of course there are many other important social divisions upon which welfare states have a critical impact and around whose welfare needs social movements have mobilized. These would include divisions by age, physical impairment, intellectual impairment, sexual orientation, religious affiliation, and national identity. Such divisions in industrial societies are just as universal as those of class, 'race' and gender. The choice of class, 'race' and gender is largely dictated Analysing and Comparing Welfare States 3 by the contemporary politics of welfare states and the social science data which reflects that politics, in which these divisions have played a preeminent role. It is therefore assumed here that three of the key social divisions of welfare are those of class, 'race' and gender - concepts upon which some brief reflection is required. Class and Class Divisions of Welfare In a capitalist society the great majority of the population is primarily dependent for its present and future welfare on the selling of labour power, that is the potential to work in exchange for wages. This dependence may be direct in the case of a worker or former worker, or indirect in the case of a worker's dependants. Wages are used to purchase, amongst other things, the welfare necessities of life. Labour power is purchased by employers and employing organizations (capital), over whom the workers collectively (the working class) have limited effective control. Through party political and trade union organization and many other forms of popular pressure (the class struggle) the working class has frequently been able to exert a degree of informal and formal influence over capital, both at the workplace and through the state by taxation and regulation of capital. Capital's sole reason for existence is to make profits and to survive the harsh pressures of capitalist competition. Hence legal and bureaucratic regulation of capital by the state is required to achieve a measure of 'fair' competition. Modern capital may also require the state to ensure that the education, health and welfare of the working class is sufficient to meet the demands of the labour process. Above all capital requires the working class to be available for disciplined, paid employment. The welfare state is therefore also the result of capital's requirements for the reproduction of labour power. The combination of the working class struggle for welfare and capital's requirements for the reproduction of labour power have produced the welfare state. The term 'welfare state' covers both the direct provision of welfare benefits and services by public agencies, and the subsidy and regulation of occupational, for-profit, voluntary, charitable, informal and other forms of private welfare. Capitalist economies inevitably experience booms and slumps - periods of profits growth and expansion, and periods of profits decline and recession. The welfare state is obviously more secure in periods of growth and expansion, though capital is always seeking out new areas for profit in the area of welfare services itself. In some circumstances this can develop into pressure for the privatization of parts of the welfare state. Whatever the state of the economy, the welfare state under capitalism is never entirely secure or stable. Particularly during recession and mass unemploy- -í f ment, capital may seek retrenchment of the welfare state, when the welfare ■'■"■"//'-lr1r,-y needs of the working class are most pressing. On the other hand in such..': -i'-*-. ^ v periods, capital may support the use of social welfare to counter working class anti-capitalist sentiment and agitation. č he A 4 Divisions of Welfare Such a Marxian 'class analysis' assumes therefore that 'capital' struggling for profits and fair competition, and 'the working class' struggling for a decent standard of living, have played a major role in shaping the welfare state. It is readily apparent however that such an analysis unravels with some difficulty and complexity when applied, for example, to the actual sociological and political divisions of the working class by income, skill. type of employment, status, 'race', religion, gender, locality and so on. In pursuing a class analysis of the impact of the welfare state more precisely, it is customary to use more neutral Weberian concepts of occupational social class and to use data on income distribution and poverty which can be interpreted in terms of a social class analysis. Official statisticians and sociologists in Britain, Sweden and the FRG have deployed classifications of occupational divisions for many years-, but unfortunately they are constructed on different bases. There is no widely accepted definition of occupational social class for the US, though 'race1 is widely used as a key parameter of social stratification. Occupational social classifications use criteria such as skill, status, and source of income. Frequently the working class in a Weberian sense is equated with manual and lower-status white collar workers, while the middle class(es) are identified as those in managerial and professional occupations. Of course when households are classified in this way, the occupation of the so-called 'head of the household', usually a man, is used, which gives such classifications an inevitable sexist bias. The aim of the modern welfare state, at least within liberal democratic ideology, is to mitigate welfare inequalities between the middle class and the working class thus defined. However, welfare states differ quite significantly in the degree to which this is achieved. It is also the case that the middle class have everywhere benefited extensively from the welfare state, and are politically well equipped to defend those benefits if and when they are threatened. It can be plausibly argued that the overall impact of the welfare state under capitalism on the social class distribution of welfare has been to stabilize and thus reinforce the class structure, at least in the post-war boom decades. Nevertheless the dismantling of the welfare state would, almost inevitably, deepen social class inequalities catastrophically. 'Race' and the Racial Divisions of Welfare Capitalism was built upon colonialism and slavery and is sustained by modern imperialism in the form of the continued exploitation and dependency of the peoples of the so-called Third World. Migrant workers, their descendants and the descendants of slaves are to a varying extent colonized internally in Western societies by processes of racialization. Such processes are of course much older than capitalism, and certainly cannot be understood or explained fully by reference to slavery, colonialism and the power of modern capital, as the history of anti-semitism, for example, shows. Nevertheless the domestic interests of capital have coincided with Analysing and Comparing Welfare States 5 these processes, particularly in the recruitment of cheap, migrant labour from poor, formerly colonized societies and from formerly enslaved «roups. Ideologies of racial superiority, enhanced by colonialism and modern imperialism, have also fostered racial division of the working class, which saps its potential for collective strength in pursuing the reform and transformation of the capitalist system. The concept 'race' is used here as a political and sociological category, whose meaning is established in concrete historical and political circumstances, and is therefore subject to change. Racial differences are not a constant or objective phenomenon. Real biological differences frequently inform the construction of racial difference, but they are not necessary to the process, in other words, 'race' has to be socially and politically constructed and elaborate ideological work is done to secure and maintain the different forms of 'racialization' which have characterized capitalist development ... It is struggle that determines which definition of 'race' will prevail and the conditions under which they will endure or wither away. (Gilroy, 1987: 38-9) Hence, in the contemporary welfare states examined here, different groups are significantly racialized - African American and Hispanic people in the US, Asian and Afro-Caribbean people in Britain, Turks and other Southern European people in the FRG. The US is the only state where data on 'race' and welfare needs and provisions is extensively collected, but there is some useful data for Britain and the FRG. Liberal democratic welfare states in the modern era generally claim to be undermining racial inequalities and racialized processes of social control. Yet serious policies which follow up such intentions have only been implemented in a few states (such as the US and Britain) where anti-racist movements have been able to exert some real pressure. The lived experience of ethnic minorities and the sociological evidence both suggest that racialized processes are embedded in modern welfare states. These processes cannot by any means be explained solely in terms of explicit racism in legislation (e.g. immigration laws), and overt subjective racism in bureaucratic and professional practice. Nor can they just be explained satisfactorily in terms of the structural, socioeconomic requirements of capital as outlined above. The term 'institutional racism' is useful to identify policy and administrative processes in the welfare state, which result in relatively adverse treatment of ethnic minorities. Institutional racism takes many forms such as stereotyping ethnic minority clients, failing to recognize their particular welfare needs or to consider them legitimate, and accommodating to overt racist pressures in the wider community. The impact of contemporary social policy on racial inequalities is quite different in each welfare state, but in all of them social welfare has both mitigated and reinforced racial inequalities and institutional racism. 6 Divisions of Weifare Patriarchy and Gender Division Patriarchy means simply male domination, or the subordination of women. Here a universal concept of patriarchy is assumed, for as Dahlerup (1987: 95) explains, the fact that male dominance seems to be somewhat universal speaks in favour of constructing a universal concept. A universal concept of patriarchy implies that male dominance is not just an effect of, for instance, capitalist society, or a leftover from feudalism, but an independent structure of its own. Patriarchal gender division is certainly a fundamental characteristic of individual welfare in industrial societies. Welfare states have played an important role in both reinforcing male supremacy, and, under pressure from women's movements, in transforming it to some extent. Conventionally in capitalist societies since the advent of the modern poor law, social policy has been built around the concept of the family wage and the patriarchal division of labour in the family. In this model, the married male breadwinner earns a family wage sufficient to meet the subsistence and other welfare needs of his dependants, including his wife who works at home unpaid to care full-time for children and other family members in need. Poor law and means-tested assistance systems have given deterrent, stigmatized relief to families which deviate drastically from the norm. Contributory social insurance was established under pressure from the male-dominated labour movement, to provide a semblance of the family wage to the sick, the old, the unemployed and widows to protect them from the degradation of poor relief. The family-wage model is patriarchal because it puts women financially in a wholly dependent position, which reinforces men's overwhelming private economic power in the family, bolstered by the wider patriarchal culture. In reality men's wages have usually been insufficient to support their families adequately, which has fuelled the pressure from working class men and women for the welfare state. Paradoxically the very elusiveness of the family wage 'may have deepened popular commitment to the family-wage ideology, which came to represent respectability and stability for working and middle class alike' (L. Gordon, 1988: 619). Thus patriarchal social insurance based on the family-wage model was a cornerstone of the New Deal welfare reforms in the US and the Social Democratic welfare reforms in Sweden in the 1930s, and the Beveridge and Adenauer welfare reforms in Britain and the FRG in the 1940s and 1950s. The material inadequacy of the family wage and the impoverishment of lone mothers have always helped to encourage working class women to find paid employment, as well as putting women into the forefront of the struggle for the welfare state. Demands for maternity rights, safe birth control, decent benefits for mothers, particularly lone mothers, and so on were just as central to the first phase of feminism around the turn of the century as they have been in the recent past. The development of the Analysing and Comparing Welfare States 7 welfare state since the Second World War has generally been accompanied by substantial growth in women's paid employment, particularly employment within the welfare state itself, as well as by increasing numbers of lone mother families. To a limited extent this has shifted the locus of women's economic dependence away from male breadwinners and onto the state, both as employer and as provider of benefits and services. patriarchy has been transformed in the sense that, under the welfare state, more women shoulder the dual burden of unpaid welfare work at home and lower-paid, lower-status paid employment in the labour market. The development of the welfare state has therefore made a central contribution to the modern reconstruction of patriarchy. A 'feminization of poverty' is sometimes said to have occurred in recent decades with increasing numbers of ione mothers and older women in poverty, that is outside the patriarchal family. This perhaps reflects a process in which women's poverty has become less hidden within the patriarchal family than it was in earlier periods. The increasingly visible poverty of lone mothers and older women outside the patriarchal family reveals the inadequacies of welfare systems based on the patriarchal family model. Of course patriarchal gender division is not just enforced through ideology and policy concerning women's incomes; it is also maintained, for example, by obstructing their control of their fertility, by condoning male violence and by the reluctance of welfare states to relieve some of the burdens of caring, particularly for young children and the growing numbers of infirm elderly people. Women have struggled hard to resist and challenge both traditional and modern patriarchal processes in social policy. Hence the struggles for women's right to an independent income, for abortion on demand and against sterilization abuse, and for day care for young children are highlighted in this book as centra! examples of policy conflict over gender division and patriarchy in the welfare state. The available data on gender division and inequalities in welfare is clearly limited by the widespread use by official statisticians of the family household as a unit and the concept of the head of household. Data on fundamentally important areas like day care and abortion is not really gathered on a comparative basis. Nevertheless the pressures created by the modern women's movements have helped to generate more and more useful sources of information for critical policy analysis. The Origins of Social Policy Analysing policy origins requires an examination of the political and social pressures, the forces of agency, which push governments into policy formation. Mainstream policy analysis offers several schools of thought on how the welfare state came into being and how social policy is shaped. These give primacy respectively to 8 Divisions of Welfare 1 parliamentary party politics, both of the right and the left, in response to pressure and interest groups as well as direct electoral pressures {party pluralism) 2 corporatist negotiation between government, capital, organized labour and other corporate bodies such as the medical profession {corporatist pluralism) 3 the institutional capacity of the state itself in the shape of the civil service, and other public administrative structures and traditions {state capacity) 4 political class struggle pursued by trade union movements and parties of the left in advancing the interests of the working class {neo-Marxismi socialism). In the policy analysis in subsequent chapters, we draw on and refer to these mainstream tradmons where appropriate to the particular national and policy context. With the possible exception of the neo-Marxist model, such mainstream approaches tend to portray policy as being delivered 'from above by establishment organizations and their leaders. In certain interpretations, one establishment figure such as Bismarck, Beveridge, Per Albin Hansson, Roosevelt, Kennedy, or Thatcher may be portrayed as a key initiator of policy change. Radical approaches naturally tend to emphasize policy change as a response to pressures 'from below'. As well as conventional lobbying, pressure groups and interest groups, these pressures are also represented by more radical social movements and self-activity, and occasionally by agitation and uprising. Women's movements, civil rights movements, welfare rights movements and anti-racist movements in recent decades have been particularly significant in relation to the policy areas covered here. For the US perhaps the most celebrated example of such an analysis ot policy change is the work of Piven and Cloward (1971; 1977). They argue that popular resistance and organization by poor people in the US, particularly the unemployed and the industrial workers in the 1930s and the civil and welfare rights movements in the 1960s, pushed governments into major programmes of social policy reform. Piven and Cloward have been criticized for not recognizing that 'the welfare-rights movement was a womens liberation movement' (L. Gordon, 1988: 623) too, and that women s demands for abortion rights and affirmative-action programmes were a complementary part of the radical challenge to the policy establishment. In Western Europe too, such popular struggles and social movements are undervalued in many accounts of the origins of social policy. Unquestionably what sociologists have called the 'new social movements' since the 1960s have been in the forefront of the pressures for welfare state expansion (Olofsson, 1988). Besides the women's movement, in Western Europe such new social movements have included the squatters, the student movements, the welfare rights movements and occasionally cataclysmic urban uprisings or riots. In some senses there is nothing particu- Analysing and Comparing Welfare States 9 larlv new about these movements, which are just the latest wave of radical and popular movements which have shaped social policy throughout modern history. The relationship of such movements to organized labour has been close in some respects and at some moments in history, particularly in Western Europe. However, frequently labour movements have distanced themselves from poor people's movements and radical, new social movements. Trade unions have advocated and defended elements of the welfare state when it is perceived to be in their interests to do so, but at other junctures such intentions have been undermined by divisions over workers' perceived political and economic interests. The relationship between popular struggles and movements on the one hand, and political and trade union organizations on the other hand, is raised by the continuing debate about welfare corporatism. Welfare corporatism envisages a contract between the citizenry and the welfare state to ensure that welfare and employment needs are adequately catered for in exchange for industrial peace and reasonable freedom of manoeuvre for capital. This contract is regularly negotiated between trade unions (representing the citizenry), employers and government. Such a system has apparently flourished in Sweden and Austria, and has been advocated strongly for Britain by writers such as Mishra (1984) and Lee and Raban (1988). Socialist theorists such as Korpi (1983) and Stephens (1979) see Swedish welfare corporatism as the essence of the democratic class struggle and the gradual transition to socialism. Some feminists, such as Adams and Winston (1980), advocate social democratic, quasi-corporatist feminist social policies on the Swedish model for the US. Some black activists in the US such as Wilson (1987) have put forward a programme for social democratic corporatism, as the means of achieving effective implementation of equal rights and employment rights legislation in order to redress the widening racial inequalities in the US. There are however several weaknesses in the corporatist vision of social policy. The first problem concerns representation; sometimes corporatist regimes, notably Weimar Germany, rest on too narrow a social base, so that for example the organized, male working class benefits, leaving the less powerful elements of the working class out of the social contract. Second, the corporatist welfare state is; likely to inhibit relatively autonomous activity around welfare needs, the popular struggles which have always pushed the welfare state forward. It will almost inevitably try to stifle, control and divert such pressures into bureaucratic avenues. Radical social movements of many kinds tend to be 'hidden from history' partly by their very nature and partly because of the establishment bias of historians and social scientists. This book attempts to keep the influence of pressures 'from below' in focus, though it is equally important not to overinflate their power. Another key theme is therefore that social policy emerges out of a continual conflict involving pressures 'from above' emanating from established economic and political forces in power and 10 Divisions of Welfare pressures 'from below' arising out of unmet welfare needs and movements advocating them. The Substance of Social Policy The substance of social policy covers the structure and functions of policy, or, if you prefer, its nature and purposes. Obviously policy substance is analysed empirically in terms of public finance, legislation, and administration of welfare (public and private), or, in other words, 'provision, providers and payment' as Higgins (1986: 226) puts it. Governments and policy analysts are adept at producing factual information on these matters, but this often gives relatively little to bite on in terms of the definition and the incidence of the welfare needs to which policy is responding, a question which rapidly leads back into the discussion of the impact of policy dealt with above. However, behind empirical questions about the substance of policy lie fundamental issues about its functions and purposes, that is what 'pulls' governments into social policy action or inaction. There are at least three ways of looking at this, elements of all three of which are taken up in later chapters. First, the idealist approach, frequently adopted by political scientists, suggests that each welfare state has been shaped by a clash of ideologies out of which has emerged an institutionalized consensus or dominant ideology. A social policy consensus or welfare ideology reflects a semipermanent, hegemonic resolution of the active political processes already discussed. Such an approach frequently produces a typology of welfare states and/or welfare ideologies, such as that adopted in this book to analyse the particular character of social policy ideology in each state. Conventionally such typologies are constructed on the basis of empirical data on welfare expenditures, rights and outcomes analysed in the context of the prevailing political hegemony. Looked at from critical perspectives on 'race', class and gender, the function of social policy is the interpretation and application of the political consensus in order to contain or restructure class, gender and 'race' relations. The functions of particular policies can then be assessed for the extent to which they maintain or transform the prevailing consensus. In this book the modern welfare consensuses in the four states under examination are characterized as Social Democratic (Sweden), the Social Market Economy (Federal Republic of Germany), Voluntarism and Liberalism (the United States) and Liberal Collectivism (Britain). Second, more sociological approaches essentially theorize social policy as functioning to deal with the problems of order, social integration and social discipline in industrial societies. For example at moments of crisis, states have sometimes seized upon social policy reform in the attempt to engineer national solidarity - Bismarck in the 1880s, the Swedish Social Democrats in the 1930s, the Beveridge Report, Johnson's Great Society Analysing and Comparing Welfare States 11 urosrammes of the 1960s. Even in the less solidaristic context of the oresent day, the more universalistic elements of the welfare state such as the British National Health Service continue to evoke a popular sense of national pride. The view that the substance of social policy represents the social rights of citizenship is a Fabian variant of the nationalist approach, extending the social rights of citizenship, that is the coverage of needs and individuals, functions to promote social solidarity. Such views of the welfare state continue to have strong support in capitalist societies, particularly in Western Europe, but they have been heavily condemned by the New Right and the new social movements. The libertarian New Right see citizenship rights as imposing economic dependence of individuals on the state and a bureaucratically defined solidarity, thereby undermining the social disciplines imposed by markets, as well as individual and voluntary initiative. The new social movements suspect the concept of citizenship in practice as implicitly excluding many people, including migrant workers, lone mothers, dependent wives, long-term unemployed people and so on. From a critical perspective a sociological function of the welfare state is the maintenance and legitimation of social divisions, notably class, patriarchal and racialized social relations. Substantively, social policy functions to impose labour discipline on the working class, to reproduce labour power on behalf of capital, to reinforce the patriarchal family and women's dual burden, and to reproduce the interna! colonization of racialized groups. Such a critical, functionalist approach to policy analysis offers enormous insights and is frequently adopted in this book, but it also has severe limitations, as critics of neo-Marxist analysis of the welfare state in particular have pointed out, for example Lee and Raban (1988: 125-8). Essentially the limitation boils down to the emphasis on the continuity and inevitability of these functions, which underestimates the state's capacity for policy change and the pressures which social movements and class struggle bring to bear on social policy making and implementation. When used in an undiluted form, critical functionalism implies an unyielding logic of capitalism, patriarchy and racism. The critical analysis of policy substance should reveal that such 'logics' are constantly modified and shaped by the immediate political and economic context. These problems are aired further in the section on methodological issues below. The third form of substantive policy analysis can be labelled economic approaches. Here social policy is conceived as a subdivision of economic policy, and the purposes of social policy are therefore largely subservient to the dictates of economic forces and economic policy making. In the real world of most Western governments, this is certainly the dominant view of social policy. Conventional economic analysis of social policy has been dominated by two schools of thought since the 1930s, commonly described as neo-classical microeconomics and Keynesian macroeconomics. Keynes-ianism was ascendant in the post-war boom years, while neo-classicism has seen a great revival since the stagflationary recession of the mid 1970s in 12 Divisions of Weifare such modern forms as monetarism and supply-side theory. Both traditii continue to exert enormous influence over social policy making. N> classicism starts from the premise that unrestrained markets tend maximize welfare, except under special circumstances where 'externaliti (self-destructive effects of free-market competition) intervene or wh 'the public good' is of paramount importance (e.g. national defenc Substantive policy is therefore analysed by neo-classical economists terms of the extent to which it either undermines or contributes to economic efficiency of private enterprise by limiting externalities and promoting the public good. Keynesianism is more intrinsically favourable towards the welfare state. Social policy is considered as an element of fiscal policy (government taxation, expenditure and borrowing) which is a key tool for regulating the whole national economy in order to counter inflation and unemployment, and to generate economic growth. Hence the welfare state is manipulated, expanded and contracted in order to stabilize the economy. Substantive social policy is therefore analysed by Keynesians in terms of whether it undermines or enhances macroeconomic equilibrium. Increased welfare expenditure may be used to generate economic demand and hence employment and economic growth. A reduction in welfare expenditure may be used to damp down inflationary economic demand and economic overactivity. Keynesians would argue that the growth of the welfare state from the 1940s to the 1970s generally reflected successful, expansionary macroeconomic management, and that the restraint of welfare state growth since then has contributed to the lowering of inflation. Neo-classicals would generally be more sceptical of the former claim. There is, of course, a lot of common ground between the two schools of thought, particularly as applied in practice by governments. Neo-Marxist, anti-racist and feminist analyses of social policy tend to avoid pure economic approaches because of their functionalism and abstraction from human, social relations. Modern Marxian economics, however, in its critique of the neo-classical and Keynesian traditions, suggests that the post-1930s expansion of the welfare state has been part of the institutionalization of the 'Fordist regime of accumulation'. Under this regime improved real wages and increased welfare expenditure have 'both reconciled workers to the intensification of labour associated with Fordist methods of production [assembly lines etc.] and provided the rising mass consumption' (Clarke, 1988: 65) which has absorbed the growing volume of consumer goods pouring off the assembly lines. The limits of the Fordist regime have emerged since the mid 1970s, as capita! has become unable to raise productivity sufficiently to counter the inflationary costs of ever rising consumer demand and the vastly increased costs of technological change. Cuts in real wages and in social expenditure can only provide a respite, as capital struggles to establish a new, post-Fordist regime of accumulation. In this analysis, the substantive development of social policy is intimately bound up with the fundamental disequilibria of capitalist economies, and the attempts of capital to overcome them. Analysing and Comparing Welfare States 13 iher central theme of this book is therefore that social policy and ; expenditure have been crucially shaped by the crises, slumps and of capitalist economies. In the West, there have been three key s in which the ideology of welfare has fundamentally shifted, all of i have been moments of international economic crisis - the 1870s, the ind the 1970s. Hence the development of social policy is intimately ;ted with attempts, often deeply conservative, to resolve capitalist ecullulTiic and consequent social crisis. The crisis of the 1870s eventually ushered in Bismarckian social insurance in Germany, New Liberalism in Britain and Progressivism in the US, a period of reform which peaked in the first decade of the twentieth century. The crisis of the 1930s generated the New Deal in the US, the fascist welfare state in Germany, the Social Democratic welfare transformation in Sweden and somewhat belatedly the Beveridge reforms in Britain. These ideological shifts towards various forms of welfare collectivism begun in the 1930s were consolidated in the post-war decades sustained by the economic boom. The expansion of the welfare state came to an end, and the welfare states have been reshaped to some extent in order to conform with more stringent and disciplinary capitalist requirements. The welfare state under capitalism is not irreversible, contrary to the intimations of Block (1987) and Therborn and Roebroek (1986). it is a movable feast, though even the neo-conservative project requires an authoritarian, regulatory welfare state to underpin the privatization of welfare services and benefits. The recessions of the mid 1970s and early 1980s saw considerable restructuring of industrial relations and social policy in the Western welfare states, on the basis of a fragmentation of the working class in the attempt to confine the aspirations of the working class within the limits of capital by confining wages and social expenditure within the limits of profitability. (Clarke, 1988: 86) The fragmentation of the working class, enhanced by neo-Hberal economic and social policies in some states, may take different forms, but racial and gender divisions are common to almost all industrial societies. The exploitation of such divisions has been fiercely resisted and it is by no means obvious that racial and gender inequalities have actually worsened over the past two decades in the states examined here. However, in those states such as the US and Britain where neo-liberalism has made its greatest impact, the fragmentation of the working class by income, status and occupation has become more pronounced, instanced by the wide use of the term 'underclass' to describe people living in poverty, whether benefiting from a welfare safety net or not. The recession of the early 1990s is likely to further these processes. Methodological Issues So far we have sketched out three areas of social policy analysis, within which a variety of analytical approaches, both mainstream and critical, have been indicated. The term 'mainstream' is used here to denote the 14 Divisions of Welfare work of liberal {in the American sense) and social democratic writers wh< are hegemonic in the West in the disciplines of sociology, political sciena and economics. The term 'critical' is used here to denote Marxist, feminis and anti-racist analysis, bearing in mind that there is enormous variation o perspective on social policy within these three, as discussed in detail fo example in Williams (1989). In the mainstream, questions of 'race', gende and class (in the Marxian sense) are, at best, of peripheral concern. Needless to say, the literature on social policy analysis is dominated by debates within and between the mainstream disciplines, and also between them and Marxism. While feminist work and anti-racist work have established themselves to some extent as subdivisions within the mainstream disciplines, as yet they appear to be distinctly ghettoized from mainstream social policy analysis. The analytical discussion of the origins and substance of social policy offered above is essentially derived from debates in the mainstream and in Marxism, and it is far from clear how adequately they can be applied to feminist and anti-racist policy analyses. Clearly it is insufficient simply to tack on the latter to established conventions. The distinction between idealist, sociological and economic forms of substantive policy analysis, as drawn above, is not meant to convey these as necessarily competing or mutually exclusive approaches. This book draws on key themes from each as already indicated. It almost goes without saying that an idealist welfare consensus or ideology of welfare is a fundamental element in forging sociological integration or social reproduction, and also embraces assumptions about economic policy and ideology. Ideally a political economy of the welfare state has to keep all three forms of analysis in play. However the separation of the analysis of policy origins and policy substance implied from the start of the chapter is artificial and problematic, simply because policy analysis in practice inevitably carries assumptions about both origins and substance, whether this is made explicit or not. Agency and Structure In the discussion of policy origins the focus is on 'agency', the concrete institutions and human pressures shaping social policy. This implies that the substance of policy emerges from the resolution of power conflicts between the 'actors' or 'players'. Such a resolution will reflect the interests of and the balance of power between those forces of agency, whether the emphasis is on political parties, pressure groups, the class struggle, radical social movements or other significant players. In reading off the actual purposes and implementation of policies directly from the process of, say, 'political choice' (e.g. Heidenheimer et «/., 1990) or 'working class mobilisation' through social democratic politics (e.g. Stephens, 1979), the undertowing effect of the deep structures of economic development or ideological power is inevitably missing. Analysing and Comparing Welfare States 15 in the discussion of policy substance the focus is unquestionably on ,-'structure' (i.e. functionalist), that is suggesting that social policies fulfil ' functions essential for the maintenance of the political consensus, social integration and/or economic equilibrium and growth. Social policies therefore emerge of necessity to meet essential social and economic needs which increasingly cannot be met by unaided voluntary and private means. In critical analyses, the implication is that the welfare state functions to reproduce capital and capitalist social relations, patriarchal economic and social relations, and racialized economic and social divisions. Functional-ism is open to the criticism of being ahistorical, deterministic and denying the possibility of radical change. At a theoretical level, many Marxists and social theorists such as Giddens resolve the agency-structure problem by proposing that 'the notions of action [agency] and structure presuppose one another [emphasis in original] . . . [and have a] dependence which is a dialectical relation' (Giddens, 1979: 53). Nevertheless, in going about practical policy analysis, the methodological questions of agency and structure pose particular problems for critical approaches. On the one hand, the origins of policy and welfare reform must be sought in the activity and struggle of working class movements, women's movements and anti-racist pressures. On the other hand, it is difficult to concede that the patriarchal, capitalist and racist imperatives structurally embedded in the Western welfare states can be shed, short of a radical transformation to a quite different political economy. Critical analysis is therefore open to the accusation of 'wanting its cake, and eating it' - celebrating social policy reform and defending the welfare state as positive gains of pressure from below, while in the same breath portraying the functions of the welfare state as fundamentally oppressive. For socialists in relation to social policy, this dilemma was perhaps first confronted in the 1880s when Bismarck promulgated his social insurance schemes for organized workers, while at the same time criminalizing working class political organization. The measures aimed to stem the rising influence of Marxism within the workers' movement by conceding that an authoritarian capitalist state was capable of social reform in favour of the working class, or in this case its most politically and economically organized elements. Bismarck exploited the division among socialists between fundamentalists who denied the possibility of progressive reform and pragmatists pressing for piecemeal progressive reform of the state. The pragmatists were significantly strengthened by Bismarck's reforms, not least because they gained considerable bureaucratic power in the administration of the insurance schemes, while in reality only an elite of the working class benefited. Welfare reform, ever since, has often raised such dilemmas for socialists. For feminists in the Western welfare states, comparable dilemmas have been raised by the sex discrimination and equal opportunities measures of recent decades. On the one hand it can be argued that patriarchal power has not been significantly undermined except possibly for some middle class women. The emphasis on 'equal 16 Divisions of Welfare rights' feminism, particularly in the US, has not fundamentally challenge^ institutionalized sexism in the labour market, the home and the welfare state. On the other hand it can be argued that the women's movement has generated a gathering momentum behind such policies, which has at least opened up the possibility of the withering of patriarchy. In the US, and to a much lesser extent in Western Europe, similar dilemmas have also been raised by the civil rights and 'race relations' reforms since the 1960s. On the one hand it can be argued that such measures have failed to achieve significant positive changes in racial inequalities and racism institutionalized in the public and private sectors, while benefiting a small ethnic minority elite. On the other hand it can be argued that, subject to further reform and more effective implementation of present policies, irreversible changes in breaking down racial inequalities and racialized processes have been achieved. Here we will examine briefly some examples of critical policy analysis which have attempted to resolve the agency-structure questions. In relation to a class analysis of social policy, attempts to resolve the agency-structure problem have been usefully described as a 'reconciled" (Jones, 1985: 47) or 'compromise' (Lee and Raban, 1988: 133) Marxist viewpoint, which attempts to reconcile the apparent contradictions of class struggle and capital logic. Hence Gough (1979: 65) suggests that periods of progressive welfare reform are explained by 'a coincidence of interests' between the welfare functions required by capital and the working class political pressures for the welfare state, but 'the respective importance of each varies over different policy issues'/As Gough implies, the question has lo be resolved differently for particular policy examples and historical moments in different welfare states. Piven and Cloward (1971) is an outstanding example of a reconciled analysis which applied Marxian functionalism to the patriarchal and racist administration of welfare benefits in the US. They showed how the welfare benefits system was used in the southern states to push poor people, particularly black women, into low-paid employment. They also document how this system was, to some extent, overturned by the civil and welfare rights movements of the 1960s. As these struggles have continued the welfare system has become more liberal and less punitive, but the enforcement of low-paid employment in a gendered and racially structured labour market remains at root a function of the system. Feminist analysts of social policy have also sought to reconcile agency and function. For example, according to L. Gordon (1988: 628-9) 'promoting the family-wage system is a better overall explanation of the social-control functions of the welfare state than has been previously offered . . . [but] the welfare state has, on balance, increased women's power'. Gordon emphasizes that the welfare state cannot successfully implement traditional palriarchal functions in the face of the reality of labour markets and of women's resistance. In her analysis of British social policy, Pascali (1986) puts a predominant emphasis on the functions of the Analysing and Comparing Welfare States 17 welfare state in reproducing women's dependence and the social control of women's lives. However she also notes that "one difficulty with such arguments is the danger of portraying women as victims of outside structures - whether of capitalism or of patriarchy' (Pascali, 1986: 238). Hence the vital significance of women's campaigning and protest over social policy issuesr West European feminists such as Hemes (1987) and Borchorst and Sum (1987) have reconciled agency and structure by ■suggesting that the women's movements of recent decades have succeeded to some extent in transforming gender relations into a new form of patriarchy. According to Sassoon (1987: 180), these transformations in the nature of women's subordination and in our relative freedom, and the new contradictions which are arising, are missed by any functionalist view, with or without a Marxist label, which defines the relationship between women and the state in terms of how state policy maintains the nuclear family or acts to keep women in the home . . . Women are at one and the same time freer and enmeshed in a new web of dependence. In the new patriarchy most women experience the dual burden of paid employment and unpaid domestic caring, less financially dependent on men perhaps but more dependent on state services. Anti-racist analysis of contemporary social policy in Britain and the US has hardly dwelt on theoretical questions about agency and function. In the US attention has focused on the urgent debate about the real impact of the reforms achieved in the Second Reconstruction in the 1960s. Even the most critical analysts accept that African Americans made 'significant strides during the civil rights movement of the 1960s' (Pinkney, 1984: 177) and that the Second Reconstruction was by no means a 'splendid failure' unlike the First after the abolition of slavery (Marable, 1984b: 208). Yet radical writers like Pinkney and Marable certainly imply a structural or functionalist perspective in debunking the liberal myths of significant and continuing black economic and welfare progress. Hence the demand for racial parity within a state apparatus and economy which is based on institutional racism and capital accumulation at the expense of blacks and labour is fatally flawed from the outset. Racism and capitalist exploitation are logical and consistent by-products of the American system. (Marable, 1984b: 211) Marable (1983) and Pinkney (1984) describe how the reforms of the 1960s helped to create a privileged black elite, brought very modest benefits to the mass of the black working class and contributed to the expansion and the disciplining of the urban underclass. Marable's analysis has understandably been described as 'structuralist' (Hochschild, 1988: 172), yet at the same time he puts enormous emphasis on the activity and resistance of African Americans in creating the Second Reconstruction and developing the possibility of the Third. In Britain anti-racist policy analysis has largely concentrated on the important task of elucidating and challenging racialized processes in particular policy fields, such as immigration, policing and housing. Resistance and pressure from the black communities and anti- 18 Divisions of Welfare racist movements have unquestionably had some positive impact on ! welfare state. Yet racism continues to be structurally embedded in | labour market, the national 'culture' and the welfare state. As Williai (1989) has shown, several anti-racist schools of thought have emerged British social science which reconcile the agency-structure questions. . contrasting ways. If racism is essentially structurally functional, then raci*1 processes have to be understood and challenged at the broadest cultui : and macroeconomic levels (e.g. Gilroy, 1987; Miles, 1989). If racism .* located essentially in institutionalized processes and in the actions n-individuals and communities, it must be analysed and challenged in ten-.* of local politics and community resistance (e.g. Ben-Tovim et al, 19S:-Ball and Solomos, 1990), Clearly these are differences of emphasis !■ relation to action and structure; they are by no means necessarily mutua . exclusive. The critical policy analysis adopted here attempts to reconcile some nl the methodological and political dilemmas raised by the questions •■! agency and structure. There is no logical inconsistency in emphasizi'i. agency in analysing the origins of policy, and structural functions • analysing the implementation and impact of policy. The welfare state in ■ patriarchal and racially structured capitalism is obviously capable of significant reform in favour of workers, women and racialized minorities, particularly for their most organized or most privileged elements. However such reforms only come about as a result of serious and sustained social and political pressures from those groups. Reforms, such as those mentioned above, restructured and transformed class, 'race' and gender relations in a long-term, positive direction, but such transformations are not irreversible nor do they necessarily open the way to the end of capitalism, patriarchy or racism. On the contrary it is manifestly clear that, particularly in periods of capitalist economic crisis, such progressive welfare reform is reversible, and that the welfare state can contribute to an adverse reinstatement of class, 'race' and gender divisions. Comparing Welfare States The very suggestion of 'comparative' analysis of social policy is problematic because it conjures up the hope that social scientists have developed rigorous methods and established schools of thought for comparing welfare states. Nothing could be further from the truth. Twenty years ago the field of study barely existed, though writing and research have expanded considerably since then. Attention has been particularly focused on the historical and political origins of benefits systems and on comparing aggregate public expenditures on social needs. In areas such as family policies and on questions such as the implementation and impact of policies, we only have a few pinpricks of light. In particular on questions of 'race', racial inequalities and racialized processes of social control, there is Analysing and Comparing Welfare States 19 arkably little critical, cross-national material. The outstanding excepts the work of Stephen Castles and his collaborators on migrant leers in Western Europe (Castles, 1984; Castles and Kosack, 1973). j work reviews the migration process, the citizenship status and the . uour market position of racialized minorities in Western Europe. It shows that governments and racist popular pressures have denied full citizenship to racialized minorities, implying a strong convergence amongst all the states. Yet the data also suggests enormous differences in the welfare status of racialized groups, for example between Germany, Sweden and Britain. However, as yet, there has been no attempt to theorize these cross-national differences and the experiences of racialized groups in the modern welfare states. Students of cross-national social policy under capitalism are faced with several kinds of literature. First, there is an increasing amount of useful data from international agencies and databases, such as that used for the statistical appendix in this book. The most important and accessible sources are perhaps the Organization for Economic Cooperation and Development (OECD), the European Commission (EC) and the Luxembourg Income Study (LIS). Such organizations have reworked national government statistics to put them on a comparable basis. The student is of course left to analyse and theorize the data, though obviously it contains its own biases shaped by the kinds of statistics which governments collect. Also agencies like OECD and the EC are not primarily concerned with comparative social policy. Their interest in the field is in examining the relationship between social policy and the regulation of public expenditure and labour markets. Such data tends to concentrate on direct public services and benefits, so that fiscal, occupational and private welfare forms are sometimes neglected, not least because governments often do not collect good data on these. Many critical parameters in terms of policy outcome like class, 'race' and gender feature marginally at best in most of this empirical data. Systematic Comparison of Social Security Systems Another form of cross-national material attempts to develop quantitative indices and concepts for comparing the origins and performance of welfare states. It is therefore perhaps the most deserving of the title 'comparative' in the eyes of social scientists. Here analysts use published and unpublished data from national government and international agencies such as those discussed above to develop comparative hypotheses about the development and impact of welfare states. Hence such analyses are very much subject to the critical limitations already mentioned regarding definitions of social welfare and of social divisions. We will discuss, briefly, three prominent examples of the genre. Wilensky (1975) compared social security spending as a proportion of GNP in sixty-four states for 1966. He concluded that 'economic level', that is GNP per head of population, 20 Divisions of Welfare 'overwhelms regime type as a predictor of social security effort' over t long term (Wilensky et a!., 1985: 9). In other words differing welf; ideologies and political regimes amongst nation states are much 1 significant than different levels of economic development, and relai differences in demographic structure, in explaining differences in soc security spending. Wilensky thus takes a structural functionalist posiii and supports the thesis of convergence, which suggests that industri ization and economic growth encourage convergent welfare state forr despite differences in political ideology. Structural functionalist co parison such as this has become unfashionable in the wake of ■ sociological critiques of functionalism and of 'end of ideology' thes However the inability of Western states to reverse fundamentally growth of the welfare state, despite lower levels of economic growth, may lend some support to the thesis. In direct contrast to structural functionalists, analysts who emphasize that 'politics matters' have attempted to derive a quantitative relationship between welfare spending and the political composition of governments in the West. The thesis here is that political forces of agency, particularly party politics, have had a predominant influence over welfare slate development, and that therefore comparative political differences account in large measure for differences in welfare expenditure. Hence, for example, Alber (1983) compared social security expenditure and the political complexion of governments in thirteen Western states over the years 1949 to 1977 using quantitative indices. He found that left cabinets tended to increase the social expenditure ratio markedly stronger than cabinets which excluded socialist parties or centre-left coalitions' (Alber. 1983: 166). Such a view is sometimes described as the 'class mobilization' thesis, suggesting that the development of the welfare state is closely linked to the mobilization of the working class for parties of the left. Casilcs (1982) related social expenditures on education, income maintenance and health care in eighteen OECD states to a number of political variables. He found that 'partisan control of government' was a key determinant of patterns of social expenditure in the 1960s and 1970s, 'with strong parties of the Right acting as an impediment to expansion and social democralic and other parties, jointly or separately, serving as a stimulus' (Castles, 1982: 85). The restraint of welfare state growth in the 1980s has been more marked in states with predominantly conservative governments, so that recent experience lends some support also to this thesis. Both the structural functionalist and the class mobilization theses in these quantitative forms have naturally been subject to criticism for their insensitivity to the social and political histories of individual welfare states, limitations which are of course recognized by their authors (e.g. Castles, 1982: 88; Wilensky et ai, 1985: 12). Taking account of this kind of criticism but upholding a 'politics matters' approach, Esping-Andersen (1990) uses seven quantitative indices for measuring the accessibility, coverage and redistribuíive impact of benefits systems and four quantitative indices of Analysing and Comparing Welfare States 21 re regime', that is characteristics of the political economy of welfare Applying these to the eighteen OECD states, reasonably close ations are established between the nature of the benefits systems and ilitical regime types. The performance indices for pensions, sickness nemployment benefits cover such critical questions as the balance en private and public pensions systems, the degree of universal to benefits, and the extent of differential benefits for different social s. One of these indices is the scale of 'de-commodification', that is the to which pensions, sickness and unemployment benefit schemes aged, sick and unemployed workers to survive economically outside ?our market. The de-commodification index therefore quantifies the to which social security benefits exempt such workers from offering Iheii labour power as a commodity (Esping-Andersen, 1990: 54). Table 1.1 gives the de-commodification scores for 1980 for eighteen OECD welfare stales, which Esping-Andersen divides into three clusters - the Anglo-Saxon and Scandinavian states at the extremes and the continental West European states in the middle range. Table 1.1 The rank order of welfare states in terms of combined de-commodification in old age pensions, sickness benefits and unemployment insurance, 1980 State De-commodification score Austraiia 13.0 United States 13.8 New Zealand 17.1 Canada 22.0 Ireland 23.3 United Kingdom 23.4 Italy 24.1 Japan 27.1 France 27.5 FRG 27.7 Finland 29.2 Switzerland 29.8 Austria 31.1 Belgium 32.4 Netherlands 32.4 Denmark 38.1 Norway 38.3 Sweden 39.1 Mean ^__________________ 27.2________________ Source: Esping-Andersen, 1990: Table 2.2 The indices used to assess regime type take account of both economic development and political power, including indices very similar to those used by Wilensky and Castles. Esping-Andersen's model offers little support for the structural functionalist view, because 'economic development is negatively correlated with de-commodification' (Esping-Andersen, 22 Divisions of Welfare 1990: 52) and has little or no explanatory power in relation to the othe: indices of benefits systems' performance. The three political factors which all offer significant correlations with benefits systems' performance are the degree of left power' (working class mobilization) in government, electoral support for Catholic conservatism, and the extent of absolutism (authoritarian rule, limited franchise) in the history of each state. Aggregating these political indices produces three clusters of political regime types, which Esping-Andersen describes as 'liberal', 'conservative' and 'socialist', as in Table 1.2. Table 1.2 Clustering of welfare states with strong degrees of conservative, liberal and socialist regime attributes Strong libera! Strong conservative Strong socialist Australia. Austria Denmark Canada Belgium Finland Japan France Netherlands Switzerland FRG Norway United States Italy Sweden Source: Esping-Andersen, 1990: Table 3.3 The liberal welfare states tend to have benefits systems in which stigmatized means-tested assistance has a central role, social insurance benefits are modest and private benefits are promoted by the state. The liberal political regimes are characterized by a relative absence of both working class mobilization for the left and of Catholic and absolutist mobilization for the right. Comparing Tables 1.1 and 1.2, the US, Canada and Australia fit well with the typology. For other states, such as the UK and Japan, low de-commodification does not appear to correlate well with liberal regime type. The conservative welfare states emphasize social insurance over both means-tested and private benefits, but in forms which tend to maintain firm class and status distinctions, so that income redistribution through benefits systems is low. Such regimes tend to be characterized by a Catholic conservative tradition and absolutist tendencies in their political histories. Comparing Tables 1.1 and 1.2, Italy, France, FRG and to a lesser extent Austria fit reasonably well with this model. The socialist (or social democratic) states have achieved the most universal and most class redistributive benefits systems, modifying the class structure but by no means removing it. Labour market measures rather than benefits are deployed to discipline and support the unemployed and those threatened with unemployment. The political regime is characterized not only by strong working class mobilization in government, but also by the ability of social democratic parties to form 'class coalitions with other groups such as farmers. Clearly the correlation between de-commodification and socialist regime is strongest in the Netherlands, Denmark, Sweden and Norway, but the model fits less well for states such as Finland and Belgium. Analysing and Comparing Welfare States 23 It almost goes without saying that this typology is based on correlations which merely suggest tendencies rather than direct relationships. Clearly too Esping-Andersen's comparative model has a limited application to the issues with which this book is concerned. 'Race' and gender play no part in his quantitative analysis. Questions, for example, around gender discrimination and assumptions of women's dependence built into benefits systems are not included in the performance indices, nor is women's political mobilization included amongst the regime type indices. Nevertheless the model carries strong implications about the class divisions of welfare, or what Esping-Andersen describes as 'the welfare state as a system of stratification'. The indices of benefits systems' performance suggest strong division along social status and occupational social class lines, although these concepts are not explicitly developed on a quantitative basis. The index of working class mobilization carries with it the implication that class (in a Marxian sense) shapes welfare politics very directly. The Esping-Andersen model is only applied to pensions, sickness and unemployment benefits and to labour market policies. It has not been applied to policy areas such as health care, family benefits and services, education, housing and so on. Also, while the model does take into account private and occupational benefits, fiscal welfare, such as tax reliefs on pensions contributions, cannot be accounted for. Of course most of these limitations are imposed by the absence of relevant data from national governments. Despite the limitations, Esping-Andersen's model is the most successful attempt thus far to develop a quantitative approach to a class analysis of comparative social policy. The following chapters of this book take an example of each of the three types of welfare state generated by the model, with the addition of Britain, a welfare state which reflects an odd mixture of the 'socialist' and 'liberal' types. Structured Diversity: Comparing Social Security Systems Beyond the quantitative literature discussed so far, qualitative comparative analysis of social policy is dominated by a 'structured diversity' approach, which emphasizes the diversity or even the uniqueness of each welfare state in its national social and historical context. The diversity may be structured by domestic political processes, cultural values, economic forces, demographic factors or whatever. The last two decades have seen a growing number of such texts, sometimes confining themselves to single policy areas and/or a few states. Notable examples, which all offer their own structural emphases, are Rimlinger (1971), Rose and Shiratori (1986), Ashford (1986), Friedman et al. (1987), Morris (1988), Castles (1989), Heidenheimer et al. (1990), Baldwin (1990) and Mishra (1990). Whether this kind of literature deserves the accolade 'comparative' is a matter of debate, since methodical and/or theoretical comparison is often either underdeveloped or absent. Many of these texts devote a chapter to each welfare state, which in itself tends to emphasize diversity and uniqueness, 24 Divisions of Welfare especially if the chapters are written by different people! Ail of them ar written broadly within the mainstream pluralist or social democrati perspectives of political science. This book too adopts a 'structure diversity', state by state method which has been eloquently advocated b Castles (1989: 12-13), who argues that the investigation of particular cases is not a return to the particularistic ove determination of single nation histories . . . the logic of comparative explanatic does not suddenly disappear when we are treating intentional, institutional an historical variables. Learning from a particular national experience will alwa; take particular forms, but patterns of human action and purposes, especially ; moulded by the fact of living in societies constrained by common structur parameters, are likely to manifest certain intrinsic similarities as well as residu differences. Here as in most other structured diversity approaches, it is sometimes le to readers to assess these intrinsic similarities and residual differences i the light of their own knowledge and experiences of welfare states. It is suggested in particular that the diversity of welfare states is structured by the common features of 'race1, class and gender divisions and of capitalist economic development already outlined. Comparing Health Care Systems Most of the cross-national research on social policy discussed so far concentrates on social insurance and social assistance benefits. Cross-national research on health care policy is for the most part empirical and descriptive in nature, being primarily 'addressed to policy planners, who pay more attention to pragmatic goals' (Wilensky et al., 1985: 48). However the diversity of public interventions in health care, of power struggles within health care systems and of health status outcomes is striking. There has been relatively little analysis or theoretical discussion about the extent to which cross-national factors have shaped this diversity. The key political dimensions for comparative analysis which suggest themselves from policy analysis of national health care systems are, on the one hand, the influence of the political context (i.e. the welfare consensus) and, on the other hand, the influence of the medical profession. Indeed most accounts of the origins and administration of health care systems focus on the conflicts and the establishment of consensus between national governments (or health bureaucracies regulated by the state) and medical professional organizations. There are sharp differences of interpretation as to the relative importance of such exogenous and endogenous players (see Wilensky, 1975: 51-2). In the state by state analysis, here, it is suggested that both the medical profession and national politics have played a critical role, the relative influence of each being shaped by the national context. Comparative analysis of the increasingly significant movements of health care consumers and workers barely exists, but the women's health movement is widely recognized as having exerted an international influence on health care provision. The gay movement must surely take the Analysing and Comparing Welfare States 25 credit for pushing the health care issues surrounding AIDS onto policy ■idciidas. Although mainstream policy analysis of health care seems dorn mated by pluralist, political choice perspectives, there are a number of socioeconomic dimensions which structure the diversity of health care «ivstems. First, all the Western welfare states witnessed a rapid rise in public expenditure on health care as a proportion of GNP in the post-war boom decades, followed by varying degrees of restraint since the mid 1970s (Tables A.20, A.21, A.22). This is linked in part to another 'structural' dimension, namely the cross-national influence of scientific medical knowledge and technology, allied with the interests of increasingly transnational pharmaceutical and medical technology corporations. Health care can be usefully conceived as a major industry, the hospital as a factory, fundamentally shaped by the development of industrialism, the macroeconomy and the industrial class struggle. The functionalism and reductionism of such an analysis, which obviously underemphasizes national differences in health care, is unfashionable. Yet the evidence from the states examined here suggests that health care is being increasingly commodified as techniques of management and cost control, for example, are imported from industry. Such structural convergences in health care policy have to be kept in mind in pursuing critical comparison. Comparing Family Policies Family policy is a much less clearly demarcated area of welfare state intervention than health care and social security. It is also only quite recently and rather reluctantly that governments have been forced to consider the impact of social policies on the achievement (or lack of it) of equal opportunities and equal rights for women. 'The family' continues to be considered as a private area of civil society, into which the state should only intervene at moments of crisis. The Nazi and Stalinist experiences with explicit family policy continue to give the very idea a bad name. This ideology of non-intervention in the family coexists very awkwardly with the modern array of social policies, almost all of which carry direct implications about, for example, the rearing of children and the care of other 'dependants' such as the infirm elderly. This book confines itself to examining selected aspects of family policy, namely income maintenance and labour market policies in relation to women, child benefits, abortion, pre-schooi day care and parental leave. Cross-national study of family policies and the impact of social policies on women has been particularly circumscribed by the inadequacy of national government statistics and cross-national incompatibility of parameters (e.g. definitions of day care and part-time employment). In some critical areas such as access to abortion services, sterilization abuse and the demand for day care, governments may be reluctant to collect good data for political reasons. The cross-national diversity of family policies, at least in the areas specified above, has been structured by a number of common functions and political 26 Divisions of Welfare concerns of the modern capitalist state, some of which have a long history. These include: 1 care and control of poor families deviating from or unable to sustain tl patriarchal norm, mostly lone parent families 2 fertility control measures, more or less explicitly targeted at po people and racialized groups 3 pronatalism, that is the use of family policy measures to attempt increase the birth rate, sometimes targeted away from poor people a racialized groups 4 interventionist family policy as a means of promoting state investme in child development or children as 'human capital' 5 the deployment of women's paid and unpaid labour, in and out of t labour market. The relative prominence of these functions or concerns varies considerat from state to state and in different historical periods. A wide range social forces have shaped such policies including the churches, charitii women's organizations, the medical profession and educationalist. Although party politics has generally not featured a great deal in family policy making, the wider political context and the welfare consensus have inevitably shaped the form and content of family policies. In recent years, several women writers have developed comparative analyses of family policies, predominantly dealing with the relation between women's paid employment and family policies and taking up the methodological questions about agency and structure discussed above. Adams and Winston (1980) compared a range of policies affecting the paid employment of women in Sweden, the US and the People's Republic of China. They conclude that neither the level of economic development nor direct pressure from social movements and political parties was very significant in explaining governments' family policies. They develop what might be described as a 'state capacity' explanation, suggesting that the key difference between the US and Sweden is the differing ideological and administrative capacities of their governments to adopt 'fuil employment' policies, and interventionist and centrally coordinated approaches to family policy. Ruggie (1984) compared policies on women's paid employment and on day care in Britain and Sweden, reaching broadly similar conclusions to Adams and Winston. According to Ruggie (1984: 19) 'the explanation ... for the differences in British and Swedish policies for women is grounded in a theory of state/society relations . . . the nature of these relations varies according to the particular state formation.' Ruggie applies a conventional dual typology of welfare states to these areas of family policy. British policy is analysed as an example of the 'liberal welfare model' in which, at the policy making level, traditional ideas about women's role remain pre-eminent and labour market policy for women is market-led. Sweden's policies are analysed as an example of 'the corporat-ist welfare model' in which active intervention in the labour market to Analysing and Comparing Welfare States 27 omote WOmen's paid employment and even gender equality is legitimate, ^pgie's 'corporatisť model corresponds to Esping-Andersen's 'social |,,m0Cratic' regime type. Although he does not analyse family policies, p Spíng-Andersen (1990: 27) notes that the 'conservative' welfare states (not considered by Ruggie) are 'strongly committed to the preservation of traditional family-hood. Social insurance typically excludes non-working wives, and family benefits encourage motherhood. Day care and similar family services are conspicuously underdeveloped,' The analysis of family X policies in the US, Britain and the FRG presented below suggests that r differences in family policies between conservative and liberal regimes are in fact relatively insignificant in comparison to the differences with social democratic regimes. The application of comparative typologies derived from analysis of benefits systems does not provide a very good fit for the analysis of family policies. Kamerman and Kahn (1978: 3) have made a useful distinction between welfare states with an implicit family policy such as Britain, the US and perhaps the FRG, and those with an explicit family policy such as France and Sweden. In contrast with the analyses above, Norris (1987) offers evidence in support of the view that party politics matters, at least in relation to some aspects of family policy. Norris sought correlations among the OECD states between indices of the economic and social positions of women on the one hand, and macroeconomic, party political and social variables (e.g. the strength of Catholicism) on the other hand. She found that GNP per capita correlated with the proportion of women in paid employment, while party political complexion of governments did not. Social democratic governments, however, had a significant positive impact on women's earnings as a proportion of men's and also on breaking down occupational segregation by gender in the labour market. Right-wing governments (rather than Catholicism) correlated with restrictive legal rights to abortion and limited day care provision. Access to abortion services was, however, clearly correlated with the strength of Catholicism as well as GNP per capita. Using a large sample of states and a range of indices thus suggests a more complex structured diversity in the comparative positions of women and in comparative family policies than indicated by state capacity analyses. Rubery (1988) indicates some other aspects of this diversity. Comparing Britain, France, Italy and the US she found no association between public expenditure on benefits and services for families, and either the level of lone motherhood or the proportion of women in paid employment. This points to the possible ineffectiveness of family policies in some of its aims, or at least the contradictory functions which such policies frequently embrace. The treatment of lone mothers is a vivid illustration. They may be encouraged to find paid employment and not to depend on benefits, while at the same time the virtues of full-time parenthood are extolled by welfare agencies and politicians. In comparing women's labour market position in the four states, Rubery concludes that it is impossible to say which is better or worse for women. Thus 28 Divisions of Welfare societies which offer more flexible employment forms, providing opportunit to combine domestic with wage work, may also offer lower pay and stat associated with these types of employment (for example part-time work Britain; informal sector work in Italy). Opportunities to maintain employms continuity and status over the family formation stage (for example in France a Italy), have to be set against the disadvantages of taking on full dual burdens domestic and wage labour, or may only be taken up because of the availability family assistance based on unpaid female labour. Reliance on state wclfj provision may provide a more equitable basis for women to enter the labi market, but where this provision is linked, as in France, to strong state supp for the ideology of the family and motherhood, it has contradictory impacts women's labour market and social roles. (Rubery, 1988: 280-1) Clearly the labour market strategies of women and employers, as well governments, vary enormously according to aspects of the socioeconom demographic and cultural context, apart from the state's capacity a direct political factors. The state by state analysis here can only throw soi shafts of light on this diversity. Summary and Organization of the Book Cross-national policy analysis, as attempted here, adopts a critic structured diversity approach. Welfare states are uniquely shaped by th political, cultural, social and economic context within a nation sta However, social policy is also structured by common elements shared wealthy Western states, notably the context of patriarchal and racia structured capitalism. All the welfare states have had a contradict impact on class, 'race' and gender divisions. They owe their origins t< combination of pressures from those advantaged and those disadvantage by the social structure prominently divided by class, 'race' and gender. The substance of social policy is structured by a number of ideological, sociological and macroeconomic requirements necessary to the continued survival of modern capitalism, which are historically redefined under changing economic and social circumstances. Most importantly, welfare states have been significantly restructured in periods of international economic crisis, the inter-war years and the period since 1974. The main chapters of this book are case studies of four welfare states, Sweden, the Federal Republic of Germany, the United States and Britain. The discussion covers the decades since the inter-war restructuring, but concentrates as far as possible on the contemporary period. The US, the FRG and Sweden are often held up as clear examples of three contrasting forms of welfare regime, Esping-Andersen's (1990) 'three worlds of welfare capitalism'. Britain is included as a fourth case study, being something of a hybrid and illustrating the significance of diversity and uniqueness amongst welfare states. Each case study is structured around five headings: ideology and welfare expenditure; income maintenance policies and outcomes; 'race' and racial inequalities; women and family policies; the health care system. This structure evolved as a pragmatic Analysing and Comparing Welfare States 29 ilution to the inevitable problems of integrating critical perspectives into rhc empirical material of conventional policy analysis without losing a cn^c of structural, cross-national commonalities in the diversity. Three o0licv areas are focused on: income maintenance, family and health care, income maintenance or social security policy clearly lies at the heart of the y,elhire state. Considerations of the class, 'race' and gender dimensions of income inequality, poverty and labour market processes are essential to critical policy analysis. Family policies clearly reflect the patriarchal structure of welfare states, and the examples of abortion, lone motherhood and day care have been particular foci of struggle by contemporary ^omen's movements. Health care systems exhibit enormous cross-national diversity in their financial and administrative forms. Yet there are comparable structural conflicts around resources, management and health status inequalities- The data used here has been trawled from international agencies, government statistics and social science research sources, as discussed above. The limitations of the available data are colossal, such as the unknown extent of unmet needs, incompatible definitions for comparative purposes, absence of critical data on class, 'race' and gender divisions, and many other forms of bias shaped by established social science and government norms. The reader must decide what has the ring ol truth. Sweden: the Social Democratic Welfare State Most enthusiastic advocates of the welfare state under capitalism m. Sweden as having achieved the closest to the ideal. For most of the pos war period Sweden has achieved healthy economic growth, 'full' empliv ment, and one of the highest standards of living in the world, combine with the largest, most expensive and possibly most egalitarian state welfai system in the West. It is almost as if the Swedes have obstinately defic gravity - or in this case the 'logic' of capitalism. Detractors on the he market right suggest that this has been achieved by creating a particuhi form of authoritarianism at the expense of individual, family and entrepreneurial freedoms, a system whose rigidity may eventually be sell-destructive. On the left, many Marxists and feminists would point out ih: limitations of the achievement in view of the enduring structures of cht^ and patriarchy in modern Swedish society, suggesting that the Swedi>., welfare state only modestly blunts the oppressions and inequalities u hie. capital and patriarchy exploit. With the growth of the gjobai economy, irk, emergence of the united capitalist states of Europe and the faltering of \ he Swedish economy in the 1980s, the survival of the present Swedish welfare state is increasingly coming into question. There is no dramatic turning point from which the modern welfare stale in Sweden can be said to have begun. Policy history in Sweden is a story of gradualism, a slow but fairly inexorable development of social welfare. As many analysts point out, this reflects a political tradition of compromise and coalition in an ethnically and confessionaily homogeneous society, which has not been engaged in imperialist excursions or wars since the industrial revolution. The clearest moment of ideological shift towards the welfare state occurred in the 1930s when the Social Democratic Party (SAP) for the most part abandoned both Marxist and neo-classical or deflationary economic policies in favour of Keynesianism. Initially on ihe basis of counter-cyclical economic policies involving public works and lood price subsidies, the SAP was able to dominate Swedish coalition go\ern-ments from 1932 onwards, with the exception of a period of centre-right coalition government from 1976 to 1982. During the 1930s what has been described by Korpi (1978; 1983) as a historic compromise between capital and labour was forged: on the one hand, private capital was guaranteed economic freedom, particularly from the threat of nationalization; on the Sweden: Social Democratic Welfare State 31 nd, beyond the economic sphere, the Social Democrats, in the the people rather than the working class, had overwhelming ■ over state policy. In the 1940s and 1950s there was in fact only jrowth of the welfare state, albeit largely on a universalist and ive class redistribution basis. The 1960s and 1970s saw rapid n of welfare benefits and services on a more neutral distributional lidifying middle class commitment to the welfare state. Policies o shaped by the pressure to recruit more women into paid lent. The gathering problems of stagflation, unemployment and union militancy from the late 1960s prompted governments to ;tout Keynesian defence of the welfare state until the early 1980s, unemployment relatively low. The Swedish response to the fiscal the state and recession in the 1970s was therefore quite different st other capitalist welfare states. Since the recession of the early iwever, the growth of the welfare state has been all but halted with ents gradually moving towards more stringent anti-inflationary i public expenditure restraint policies. Ideology and Welfare Expenditure The Social Democratic Hegemony It is fruitless to search for a clear theoretical or programmatic statement of the Swedish Social Democratic conception of the welfare state. It has to be reconstructed from the pragmatic, piecemeal development of social policy over the last century. Policy formation by the Social Democrats since the 1930s has been shaped by two broad influences - populism and socialism. _.._ c Populism entails seeking a popular mandate, a broad appeal within the /^-^ electorate, and thus coalition and compromise with social groups and * ^Vv interests structurally hostile to socialism. Socialism involves the development of economic democracy, political accountability, social equality and the advancement of working class interests. Having dominated government for so long, the Swedish Social Democrats have continually wrestled with the contradictions between populism and socialism. Commentators often convey the ideology of Swedish Social Democracy either as being essentially socialist, for example Korpi (1983) or Stephens (1979), or alternatively as having abandoned socialism in favour of a progressive liberal collectivism, for example Tingsten (1973). Neither is the case; socialist ideology and rhetoric remain the rationale for the party's existence and a benchmark for ideological renewal, but populist pragmatism and political realism shape much of the party's policies in government. In the 1930s, three elements of this welfare ideology came to prominence. First, 'the People's Home', to describe the future welfare society, was characterized by 'equality, concern, cooperation and helpfulness' (quoted by Korpi, 1978: 84). This famous phrase was coined by the party 32 Divisions of Welfare leader Per Albin Hansson in 1928. The emphasis on 'the people1 signalled consciously the party's intention to move beyond an emphasis on the needs of the working class. The party was soon to form a successful coalition with the Farmers' Party. Later in the 1950s the SAP dropped its alliance wiih the farmers, turning to the growing ranks of white collar workers and professionals as allies of the blue collar working class. The word 'home" implied the re-creation of a 'sharing, just community on a national basis ... a person would not take more than his or her share, but neither would anyone stand outside the system of community provision' (Heclo and Madsen, 1987: 158). The 'People's Home' connects directly with a second idea, that of 'solidarity' as a limit to competition and inequality. This translated into a keen emphasis on universal, flat-rate benefits as the underpinning of the welfare state in the 1940s, and also trade union commitment to soiidaristic wages policy, adopted in the late 1930s. This emphasizes that wages should be related to workers' performance rather than capital's profitability, and also that pay differentials should be minimized. A third element of Social Democratic ideology from the 1930s is the commitment to Keynesian counter-cyclical economic policies, particularly the use of infrastructural public employment and labour market policies to limit unemployment. In practice in the 1940s and 1950s the principles of the Social Democratic ideology of the welfare state were essentially the commitment to full employment and the provision oť universa! flat-rate benefits as protection against fundamental risks of sickness, low income etc. This had of course much in common wilh Beveridge's Liberal Coilectivist conception of the welfare state as incorporated by Labourist social democracy in Britain. During the 1960s pressure built up on the leadership of the SAP to renew socialist commitment to equality of outcome in terms of both class and gender. This was inspired by the student movement, the rediscovery of poverty within the welfare state, the rebirth of feminism and other radical, new social movements. A turning point was reached in 1969 with the party's adoption of a policy document calling for increased social equality, arguing that this was the primary issue faced by the labour movement. It said that an equalization of living conditions is a means for attaining altered human relations, a better social climate . . . Those who are left behind, with insufficient resources to contribute to the common good, represent an obstacle to both efficiency and desirable social change ... In the Social Democratic conception, there is no reason that extreme differences in endowments, in health, in intellect, or in work capacity should lead to an assignment of standards and life chances that differentiate some from others, (quoted by Heclo and Madsen, 1987: 174-5) The report highlighted wage and salary differentials, poverty, unemployment and gender inequalities as issues which should be tackled more energetically, arguing that the welfare state should build upon Beveridgean universalism 'through "targeted" reforms for groups with special difficulties' (Heclo and Madsen, 1987: 178), that is those who had been left behind ■^ Sweden: Social Democratic Welfare State 33 in the post-war economic boom. This 'equality movement' had a considerable impact on SAP ideology and policy making in the 1970s, above all perhaps with the introduction of sharply progressive income tax in 1971. It led to a great expansion of services and staffing within the welfare state with some modest success in improving equality of outcome. During the 1970s the energy of the rejuvenated trade union wing of the labour movement was focused on the question of 'economic democracy', the achievement of more popular direct control of private industry and commerce without nationalization or socialization. The proposal was to establish trade union controlled 'wage-earner funds' out of a substantial profits tax; the funds would build up stock holdings in leading companies, thereby gaining substantial leverage over their management and investment policies. The pressure for 'economic democracy' in part reflected > concern about the growth of structural unemployment and workers' -inability to influence decision making in increasingly transnational firms, which resulted in long-term deterioration in the welfare of working class communities. In 1983 with the labour movement in retreat in the wake of the recession, the government implemented an extremely watered down version of the wage-earner funds, largely funded by workers' contributions and unlikely to have much influence over capital's decision making (Pontusson, 1987). In the 1980s the movements for equality of outcome and for economic democracy have declined in their influence, as the SAP leadership has increasingly adopted a 'new realist' posture under ideological pressure from the New Right and the fiscal crisis of the state. The attack on anti-capitalist ideology in the SAP was led by Kjell-Olof Feldt who was the Minister of Finance from 1982 to 1990. As far back as 1984 he complained that 'anti-capitalist agitation still plays a role in the Labour movement' (Linton, 1984). On April 17th 1989, interviewed by The Financial Times, he said that the market economy 'is a dynamic and efficient economic system, which produces the best results in this imperfect world ... as long as it is influenced by the democratic forces operating in a free society'. Reading behind his public statements and his policies, it is clear that he was 'trying to drag Sweden forward through a radical programme of "free-market" reforms' (Woodall, 1990: 15), including tax and benefit cuts, higher unemployment and the encouragement of private for-profit health and welfare services. Despite Feldťs resignation in early 1990 in response to the SAP leadership's failure to support his policies, the influence of such new realist, pro-capitalist or market socialist views is gaining ground in the SAP. In the 1990s the party therefore faces something of a serious ideological crisis over future social and economic policy. Welfare Expenditure Not surprisingly Sweden has the highest level of public expenditure on social welfare amongst the capitalist states. In 1981, the last year for which 34 Divisions of Welfare comparable statistics are available, real social expenditure (Table A accounted for 33.5 per cent of GDP, compared to 15.9 per cent in 15 which was not far above the OECD average in that year. Until the 19( the actual levels of welfare expenditure in Sweden lagged far behind wl one might have expected from the Social Democratic hegemony. Therb< (1989: 220-2) explains what he calls this 'social lag' in terms of the traditi of low taxation due to non-beiligerence, fiscal conservatism and workc resistance to indirect taxation. The OECD figures also jeveal that 1 growth of welfare spending as a proportion of GDP growth in Sweden the years 1960-75 was not far above the OECD average, but in the years 1975-81 real social welfare spending increased four times as fast as the economy as a whole, twice as fast as the OECD average in that period (Table A.l). In other words, Swedish governments protected and expanded the welfare state during the 1970s recession, reflecting a progressive interpretation of Keynes, unlike most other capitalist welfare states. Ironically, for all but one year of the 1975-81 period the SAP was in opposition, though the centre-right coalition did not deviate significantly on welfare spending policy. Public expenditure as a whole was down from a peak of 68 per cent of GDP in 1982 to 61 per cent in 1989. According to Woodall (1990: 15), the burden of public-spending cuts since 1982 has fallen particularly heavily on investment projects, and public infrastructure is falling apart in some cities. There are also serious staff shortages in education, health care and the police. Common complaints are that there arc too few state nursery-school places and that waiting lists for operations such as hip replacements or eye cataracts can be more than two years long. True, compared with London, say, most public services seem to be running fairly well . . , But Swedish voters perceive a deterioration in their services. The Swedish welfare state is of course financed by very high levels of personal and corporate taxation compared with most other Western states. Although income and wealth taxes are steeply progressive, a skilled industrial worker on an average wage in the late 1980s incurred local and national income tax at about 40 per cent of gross income (Swedish Institute, 1988), a figure which has gradually come down from around 50 per cent in the late 1970s. Employers pay a 43 per cent payroll tax for social welfare contributions, which it may be argued either is effectively a deduction from wages or inflates prices. Social insurance contributions by employees are only significant in financing unemployment benefit. Otherwise the welfare state in Sweden is financed largely from direct central government, local government and payroll taxes (Olsson, 1989: Graphs 2 and 3). A massive tax cutting reform is due to be implemented in 1990 and 1991, which the government claims will preserve its progressive effects. Indirect and wealth taxes are due to be increased, but the tax reform will undoubtedly have to be paid for by more cuts in public spending. The future of welfare expenditure in Sweden will no doubt reflect the outcome of the ideological struggle within and beyond the SAP as discussed above. Sweden: Social Democratic Welfare State 35 i wmiae Maintenance Policies and Outcomes . . me Inequality and Poverty • .' e A.2 gives an indication of the comparatively enormous significance ■ ■. tsh benefits from the welfare state as an element of gross incomes in n.-, ien. In 1980 cash benefits contributed 29.2 per cent of average gross ...-me, compared to 17.2 per cent in the UK, 16.5 per cent in the FRG it a mere 8.0 per cent in the US. To pay for this, income tax took 28.5 per cent of average gross income (that includes public and private transfers, property income and so on, as well as income from employment), compared to between 13 and 17 per cent in the other three states. Evidence for the comparatively progressive effects of this income taxation and income maintenance regime in Sweden comes from statistics on the distribution of disposable household income, adjusted for household size (Tables A.3 and A.4). In 1972/3 and 1980 the bottom 20 and 40 per cent of the distribution had very much higher proportions of total disposable/net incomes than in most of the other capitalist welfare states, while the top 20 per cent had significantly smaller proportions. Gini coefficients (Table A.5) also indicate that Sweden has consistently had one of the most equal disposable income distributions in the West, very significantly more equal than the other countries in this study. Recent data from Olsson (1986: Table 23) and Äberg et al. (1987: Table 8.11) suggests that the Swedish welfare state, at least up to the early 1980s, continued to redistribute incomes on a relatively progressive basis. Äberg et al. (1987: 140) conclude that 'the tax and transfer systems seem to have a more redistributive effect in 1980 than was the case in 1967', but most of the change occurred in the late 1960s and early 1970s in the wake of the tax reform of 1971 and the development of the state earnings-related pension. From the mid 1970s to the early 1980s (and probably since then) there was little significant change in income distribution (Olsson, 1986: 58; Áberg et al., 1987: 151). The professional and managerial classes have been able to maintain their position, in part because of the large and increasing proportion of women in this class who are in full-time paid employment, compared to the other occupational social classes (Äberg et al., 1987: 136). Using an indicator called 'the household consumption surplus' which 'indicates how much money households have available after deducting from disposable income those expenditures required to maintain a certain basic standard' of welfare including housing, Äberg et al. show that the distribution of consumption has become less equal since the 1960s. The professional and managerial class has benefited increasingly and differentially from various tax reliefs, particularly on mortgages, to enhance its living standards relative to the lower classes, over and above the basic necessities. When wealth and consumption are taken into account the professional and managerial class 'despite a long-term income equalization, has succeeded 36 Divisions of Welfare in holding on to its solid lead in economic resources . . . probably nv.o due to the use of tax relief and the high level of inflation during the 1 -'i-% (Äberg et al., 1987: 150). The same is likely to be true for the 1 ■*.. probably more so. Table A.6 indicates that in 1980 5.0 per cent of the Swedish popul ■ ■ i. were poor, using the economic distance poverty measure. Not surprisr . this was the lowest amongst the four states examined here, though no !i ■ much lower than the FRG at 6.0 per cent. There is apparently little < h . data on poverty in Sweden. There are no national scale rates for irk -tested social assistance, and anyway, as Olsson (1987: 73-4) points c i ,- relation to Sweden, 'the humiliating nature of means-tested social a----- ance makes it. . . extremely difficult to use as an indicator of both rel.. . and absolute poverty'. Unlike the situation in many other welfare si i.* most lone mothers and low-income old people in Sweden do not ha »■ ■ resort to social assistance because of the development of the pension benefits systems. Nevertheless these groups together with the long- . 11 unemployed are in relative poverty in modern Sweden. Here we i examine the position of social assistance claimants, the unemployed „::„ women in the income maintenance system to analyse the position oi the relatively poor within the welfare state. Old Age Pensions The statutory old age pension schemes accounted for 85.5 per cent of old age pension payments in 1980 in Sweden, which is unsurprisingly a much higher proportion than the average amongst capitalist states (Table A.8). On top of the statutory pension, about half of Sweden's old age pensioners receive payments from an occupational or private insurance pension. There are two statutory pension schemes, a basic flat-rate payment supplemented by an earnings-related scheme for higher earners called ATP. All Swedish citizens and foreigners settled on a long-term basis are entitled to the basic, flat-rate pension which was introduced in 194S; a much lower, universal statutory pension had been in existence since PM3. The flat-rate pension was annually indexed to prices as early as 1951, but in fact its real value against retail prices increased by 350 per cent between 1949 and 1984 (Olsson, 1986: Table 8). In this important respect, Sweden contrasts vividly with the other countries in this study in having a universal payment to old people which lifts them out of primary poverty. This has been- of particular significance for women, who have greater longevity and are unlikely to have had an occupational pension until recently. I he earnings-related statutory pension (ATP) was passed by a single parliamentary vote in 1959, after a long and furious conflict, still perhaps ihe most significant debate in the history of the modern welfare state in Sweden (Heclo, 1974). It was a victory for the notion of a universal, state earnings-related pension as advocated by the Social Democrats and the Sweden: Social Democratic Welfare State 37 union movement, and against the notion of state regulated oecupa- 1 and/or private pensions, as advocated by the centre-right parties. . social insurance schemes dominate pensions provision much more .."■ Ictciy in Sweden than in the other three states examined in this book, private pensions accounting for less than 5 per cent of total pensions " -..iditure in 1980 (Table A.8). The movement for the ATP, whose ",' litarian basis many socialists opposed, helped to forge the integration ■,.; w hite collar unions into Swedish Social Democracy during the 1950s. 11, .. the SAP managed to extend its support amongst the growing ranks of ■ !„ -ie\v middle strata' and give them a material stake in the welfare state ■ s ,- aliens, 1979: 179). The distinction between tabourism and socialism is .■-■■■ated very clearly by the 1959 pension reform, which showed that . „ i welfare reforms that have no redistributive function may still have .; .ii ilising effects that strengthen labour's position' (Pontusson, 1984: 80). : ,• \ IP only started to make a significant contribution to old people's , , i iL"i in the 1970s. It gives an index-linked payment of 60 per cent of % sic earnings in the worker's fifteen most lucrative years up to a certain ^....:!i. provided there have been thirty years' pensionable earnings. The payment is reduced pro rata for less than thirty years' service. Nonetheless this level of payment and the eligibility conditions are much more generous than most statutory pension systems. By the mid 1980s 59 per cent of pensioners were receiving an ATP pension. A supplement to the basic pension was added in 1969 to help those with low or no ATP. By the early 1990s over 90 per cent of men and 65 per cent of women should be receiving the full ATP (Olsson, 1986: Table 10). Notwithstanding the universality and generosity of the statutory pensions schemes, there are of course significant, structural inequalities in incomes amongst old age pensioners in Sweden. The most affluent are those private sector white collar employees covered by occupational or private pensions, followed by public sector employees with occupational pensions, and then by skilled workers with occupational pensions. On the lower rungs are the pensioners solely dependent on the statutory schemes, with those at the bottom on just the basic pension. These latter include a number of self-employed people, farmers and others who have not fulfilled the eligibility conditions for ATP. Nevertheless compared to the UK and the US, old people in Sweden as in the FRG derive relatively little (11.1 per cent) of their income from paid employment (Table A.9). Perhaps the single most striking achievement of the Swedish benefits system is that only 0.1 per cent of elderly people are poor, after transfers are taken into account (Table A.6), compared to 9.3 per cent in the FRG, 18.1 percent in the UK and 20.5 per cent in the US. Not surprisingly with the development of ATP and occupational pensions, income inequality amongst the elderly has increased in recent years, particularly between men and women. As Äberg et al. (1987: 130) explain, many 'women in particular receive lower pensions as a rule because they have often been housewives . . . [and] do not reach the maximum 30-year 38 Divisions of Welfare gainful employment rule.' Having said that, the relatively high level women's earnings compared to men's in Sweden and the relative gen osity of the pension eligibility rules put women in a comparatively nu better position in Sweden than in the other countries in this study. Aber; al. (1987: Table 8.7) compare the incomes of Swedish pensioners by gen> and occupational social class in the years 1967 and 1980. The clas:-differentials in pension income amongst women remained largeh unchanged in the period. Amongst men the pension gap between hlu< collar and white collar workers narrowed, but the gap between these l\u classes and the professional and managerial class remained more or less tru same. Clearly as long as wage and salary differentials are structured alone class and gender lines, structural differences in income in old age wil! remain while pensions are paid on an earnings-related basis. Hence the Swedish pensions system reflects and maintains class and gender inequal ities to a considerable extent, but blue collar male pensioners havt improved their relative position in the last twenty years. Social Assistance Means-tested social assistance is administered by local authority social services departments as part of the social work system. Rates of payment are determined locally and benefit is frequently part of a casework package. The scheme was known as poor relief until 1956 when it was reformed and renamed social assistance; in 1982 it was reformed once again and renamed 'Socialbidrag' (SB, Social Benefit). In 1963 ihe proportion of the population claiming social assistance fell to an all-time low of 3.5 per cent, since when it has fluctuated between 4 and 7 per cent. Average levels of benefit are modest at between 3 and 4 per cent of the average gross industrial wage, a quarter of the basic pension level, though supplemented by child and housing allowances (Olsson, 1987: Table 4). Until the last twenty years, most claimants were elderly or poor families, but by the 1980s claimants were predominantly single, young and often childless. Hence the poverty rate amongst non-elderly households without children at 7.0 per cent is highest in Sweden amongst our four welfare states (Table A.6). Amongst the factors prompting claims for SB, social workers cite unemployment, mental and/or physical illness and alcohol/ drug problems. According to Gould (1988: 104), one other factor associated with the increase in SB claimants [between 1981 and 1985] was that of refugees and immigrants from outside the Nordic area. While some of the local authorities reported around 10 per cent of all households in this group, in Stockholm it was claimed that it accounted for 40 per cent of the actual growth in numbers. Ethnic minorities are likely to be more dependent on SB because of iheir differential vulnerability to unemployment and their failure sometimes to fulfil eligibility criteria for mainstream benefits covering sickness, old age. unemployment and disability. SB payments can be made to supplement Sweden: Social Democratic Welfare State 39 , wages, and for families with only one wage earner 'it is becoming more -ult ... to avoid relativistic poverty' (Nasenius and Veit-Wilson, ;), even if lone parent benefits lift many of these families above the .. al assistance level. _^ould (1988) gives a fascinating account of a major dispute over the a(jministration of SB in a poor neighbourhood in Stockholm in 1984. it concerned the age-old dilemma in the administration of poor relief between, on the one hand, disciplining and deterring the claimant in order to sever their dependence on the state and encourage independence and self-help, and, on the other hand, the more benevolent, libera! view that claimants should have an inalienable right to the means of subsistence. The 1982 reform of SB explicitly sought a middle way between the two views, which seems to have been interpreted with different emphases in different social work departments. In one Stockholm office under a lot of pressure from claimants, the social workers opted for a more deterrent approach, encouraging many claimants to empower and organize themselves and their lives without resort to SB. Many of these claimants allegedly had drug and/or alcohol abuse problems. In the context of Sweden's very strong puritan, temperance tradition, the view that deterrent state 'pressure and even force might be the best way of getting them to help themselves' (Gould, 1988: 118) had considerable popular support on the left as well as the right. Nevertheless resistance from claimants prompted an official inquiry, which produced a compromise report. By 1986 it seems that the social workers had moderated their strategy, the number of successful claims had increased and the number of social work posts had been increased. Thus this appears to be a classical example of successful resistance by poor people to disciplinary, deterrent relief policy, and an equally significant example of resistance by welfare workers to under-resourcing of the welfare state. The social assistance system in contemporary Sweden functions both to discipline and to maintain an underclass, apparently of quite a disparate social composition - young post-materialists or drop-outs, older long-term unemployed single people, ethnic minorities and others who have slipped through the statutory income maintenance system. Unemployment and Labour Market Policies At first glance unemployment in post-war Sweden has been exceptionally low, and by any measure has been consistently lower than in the other countries studied here (Table A. 10). The official or 'open' unemployment figures are based on the number of insured unemployed as a proportion of the workforce insured against unemployment, but do not include either the insured or the uninsured on training, job creation and other labour market schemes. This is a remarkably narrow definition of registered unemployment, because there are young people, women and others who have not worked in insured employment long enough, if at all, to come into the 40 Divisions of Welfare reckoning. 'Open' unemployment in the post-war period has fluctut between 1 and 2.5 per cent except in the years 1982-4 when it reachec all-time peak of 3.5 per cent. The world economic downturns in the 1970s and early 1980s produced comparatively modest and short-li increases in 'open' unemployment in Sweden. However if workers cove by the various state labour market schemes are included, the unemp ment rate is more than tripled. In 1988 the 'open' unemployment rate 1.6 per cent, but if those on labour market schemes are included, unemployment rate rises to 5.0 per cent (OECD, 1989b: Diagram 22)." still does not include those who are seeking or available for { employment but are ineligible for the benefits and not catered for by labour market schemes, if one includes early retirees, 'discoura workers' who do not register and reluctantly part-time employees, OE (1989b: 62-3) suggest that 'the pool of under-utilised labour was abou per cent of the labour force in 1984' though it has declined significa since then. Labour shortages in some industries and some parts of country were met by the arrival and recruitment of migrant workers in 1950s and 1960s, and by the recruitment of many more women into \ employment since the 1960s (see below). Until recently at least, a key component of Swedish employmenl pi has been the 'solidaristic wages policy' adopted in the 1930s by the ir unions in their national, centralized bargaining with the employers. principle is that 'all wage differentials motivated by differences in pr.;iii "wage-paying ability" between regions, industries or firms should removed' (Bjorklund and Holmlund, 1990: 23) or 'equal pay for e< work' regardless of the profitability of a firm or an industry. I inefficient firms, propped up by low wages, are driven out of biMii while efficient and profitable firms are not burdened with wages pushei by aggressive trade unionism. With more than 80 per cent of the lar force organized in trade unions, the annual wages bargain has playi central role in employment policy. During the 1980s the solidaristic w; principle has to some extent fallen by the wayside in a context of incre; unemployment, faltering real wages and more aggressive strategies t employers and trade unions. An inevitable consequence of wages solidarity is that the redum workers from inefficient firms should be retrained or redeployed with assistance of the state. Hence in the 1940s the state Labour Market Bi (AMS) was established, an agency which helped smooth the way for dynamic restructuring of Swedish industry by private capital, partial 1 in the 1950s and 1960s. The AMS has presided over a range of measii First, the employment service matches the unemployed to vacancies ( national basis. All vacancies and layoffs have to be notified by law to service which pays substantial mobility and relocation grants. The I; increasingly took the form of serious financial inducements until being back in 1987. A substantial proportion of the unemployed have ľ encouraged to migrate to the boom areas in the south of the country, o Sweden: Social Democratic Welfare State 41 ... ■ reluctantly, with the consequent break-up of established commun-'. .. Second, there are AMS training schemes (beyond the in-house ■ ■ig by employers) organized in training centres and schools. Third, \\:^ oversees job creation measures including relief-work projects, .....n-.-rary employment subsidies and recruitment subsidies. Finally, there '"re extensive special measures for the occupationaliy handicapped, often 'nvolving the subsidization of wages in the open labour market. The expenditure on these 'active' labour market programmes is more than double the expenditure on unemployment benefit and amounts to nearly 2 pCr cent of GDP, the highest level of such spending in the OECD. A very substantial proportion of the expenditure on active labour market measures, fluctuating between 30 and 50 per cent during the 1980s, has been devoted to the occupationaliy impaired. Another unique result of these measures in comparative terms is that only 8 per cent of the unemployed in ]988 were jobless for more than a year, compared to about 50 per cent on average in the European Community. The unemployment insurance benefit (UIB) system is a clear exception to the universalism and direct public sector provision generally characteristic of the Swedish welfare state. UIB is administered by voluntary societies under the control of the trade unions, and financed by government, trade union and employer contributions. In 1950 only a third of the labour force was thus covered, but this has risen to over 80 per cent in the 1980s. Benefits are paid for up to twelve months under fairly tight eligibility conditions; contribution and benefit levels vary considerably from one UIB society to another. For example 'members with higher unemployment risks (e.g. musicians) pay higher contributions than those with lower unemployment risks (e.g. metal workers)' (Kerans et al., 1988: 136-7). With increased government subsidy of the societies inaugurated in 1974, eligibility rules have become less severe, UIB society provisions more uniform, and membership of a society became compulsory for all trade union members. Although UIB is both taxable and earnings-related, Sweden is only surpassed in its generosity to the insured unemployed by Luxembourg and Denmark according to OECD (1989b). Hence disposable income for UIB recipients is reduced by between 12 and 29 per cent by unemployment (OECD, 1989b: 82). In 1980 average UIB payments were 78 per cent of average gross industrial wages (Olsson, 1987: Table 4). UIB remains under voluntary administration by the unions for a number of reasons. First, the generally low levels of unemployment combined with the high levels of union membership mean that pressure for a government takeover of the societies is not significant. Reflecting this, second, the UIB societies have been seen by the unions since the nineteenth century as an important means of recruiting and keeping members. Hedo (1974) describes the central role played by the Swedish trade unions in the development of support for the unemployed over the last century in comparison with the relative negligence of the British unions. From the trade union viewpoint the advantage of the voluntary system is that 42 Divisions of Welfare members are served by work colleagues whom they trust, the unions go out their way to render a service to their members, and beneficiaries are likely perceive themselves as ongoing members of their respective union. The unior in turn, are able to keep a close tab on the working conditions of their memhei as well as having regular contact with the unemployed. (Kerans et al., 19H 132-3) The trade unions continue to defend the system because 'in a country wi no tradition of a closed shop, services to members are very important ft the maintenance of union support and identification' (Kerans et al., 198 138). In 1974 an unemployment assistance benefit was introduced for tl uninsured; with certain eligibility conditions, it is only payable for up to s months and the fiat-rate payment is very modest. Nevertheless by the 19H the number of beneficiaries was more than double the number of those ( UIB, the great majority of these being young people. As Vogel (1987: 26 suggests, there is a growing number of young people in their twenties ;u of women of all ages who are in and out of paid unemployment, ar ineligible for UIB, though they may claim unemployment assistant benefit. Over the last twenty years Swedish governments have built up Ian public sector deficits to finance the 'full employment' commitment \ direct subsidies to industry, expansion of employment in the welfare slu and the active labour market policies, all of which absorbed a great deal or unemployment. By the late 1980s with rising inflation and industrial militancy, the commitment to 'full employment' through measures such as we have described has been questioned more confidently by the neu realists. OECD (1989b: 55) attribute Sweden's higher than average inflation to the low level of unemployment and the tight labour market, which gives the trade unions enhanced wage bargaining power. They suggest that an increase of 1 per cent in the official level of unemployment (a massive 67 per cent increase in terms of people) would reduce inflation in Sweden to the OECD average (OECD, 1989b: 71). Woodall (1990: 9) concludes that 'to anyone else but the Swedes, that would seem a small price to pay', but it is not yet clear whether this trade-off of unemployment and inflation is so precise, and even if it is, what interpretation will be put on it by Swedish governments in the 1990s. Women, the Labour Market and Income Maintenance The position of women in the Swedish labour market is distinct in several respects, as documented for example in OECD (1985a). The female proportion of the paid labour force (Table A.12), from being below the average amongst the capitalist states after the war (26.3 per cent in 1950). rose to being the second highest after Finland in the 1980s (48.0 per cent in 1987). By 1987 79.4 per cent of adult women below retirement age wore 'economically active' in paid employment or training, a ratio approaching that of men. This is by far the highest level of female economic activity Sweden: Social Democratic Welfare State 43 mongst our four states (Table A. 13). The lower economic activity of omen in the immediate post-war years in part reflects the fact that veden was a non-combatant in the war, so that the demand for women's viid employment was much lower. Since the 1960s women have been -ccnly recruited into paid employment, not least through the active labour nai'ket measures. In 1970 married women constituted 46 per cent of those ,n AMS training schemes, compared to 14 per cent ten years earlier Wilson, 1979: 79). Since the early 1970s women have had parity with men though no more than that) in terms of participation in and support from he various AMS schemes described above. Women's hourly earnings as a proportion of men's have risen steadily in he post-war period to reach 90 per cent in 1981, by far the highest in the ")KC'D states. This data however comes from the LO, the blue collar trade inion federation, and does not cover by any means all sectors of women's jaid employment. According to Ruggie (1988: 183), 'using samples of full-ime. full-year workers in all unions, women's pay as a proportion of men's ippears to be lower than previously thought - in 1981 it was 80.5 per cent iverall.* In 1981 46.4 per cent of women employees worked part-time, the ,ccond highest proportion of female part-time employment among the )F(D states (OECD, 1985a: Table 1.3). Social benefits and employment irotection are reduced significantly for those who work less than twenty-two hours a week, but it is not clear what proportion of women part-time workers are affected by this. With the growth of part-time employment women's weekly earnings in comparable occupations may be as much as 30 per cent less than men's, before taking occupational segregation into account. Registered unemployment has been considerably higher amongst women than men, a gap that increased in the 1970s and narrowed in the 1980s, as women benefited from labour market measures and unemployment hit male-dominated industries. Comparison of unemployment levels between men and women is problematic, as hidden unemployment is much higher amongst women. Nevertheless on the OECD basis, female unem-plounent has been consistently higher than male unemployment throughout the post-war period (Table A.14). An official survey found that 90 per cent of women aged between twenty and fifty-nine wanted paid employment, but only 80 per cent had achieved it (Vogel, 1987: 270). The final important feature of women's position in the Swedish labour market is vertical and horizontal occupational segregation, which follows a pattern in some respect similar to that of other Western countries: 89 per cent of secretaries, 94 per cent of nursing auxiliaries, 78 per cent of shop assistants and 90 per cent of cleaners are women (Scriven, 1984). Compared to other states, however, women are much less prominent in 'unskilled' factory work in the private sector and extremely prominent in public sector employment in the welfare state. According to Esping-Andersen (1990: 202). in the period 1965-85 'women accounted for 87 per cent of total heahh-education-welfare employment growth in Sweden'. He concludes that the past two decades have seen the upgrading of women's employment 44 Divisions of Welfare from 'junk-jobs' in private industry to higher-status and better paid --K. the welfare state. This has however resulted in increased hoi .■<.. occupational segregation by gender, with women increasingly conce si. ;. in clerical work and welfare work. In this respect, when compared i-> i|.. USandFRG, Sweden emerges as the most gender-segregated among the three cí m,: t More than half of the women are locked into typical female jobs, while ■■ * ■ ; women have penetrated the sanctuaries of male dominance !' .»■ industrialization in Sweden only augments the problem. (Esping-Ai ■! :* ■ 1990: 212) It would therefore seem that women in Sweden have not been i. .,.ij * much as substitutes for male workers in industry. They have bun i-^ beneficiaries in paid employment terms of the expansion of '■-..■ ■ --services in the 1960s and 1970s as part of counter-cyclical Ke*-n-uneconomic policies. Perhaps above all as an employer the Swedish i* fc h.i-. state has affected women's welfare over the past two or three d.<..iv* Positive discrimination in the active labour market measures, sex t , ■• 11. legislation and belated trade union pressure have had some sui»,.— ;■ improving women's wages and salaries relative to men, but such pi ^ v have had little positive effect on horizontal and vertical occup ■ i'-iii! segregation (Ruggie, 1984: Chapter 4). The increase in part-tír^ p i-employment of women, much of it in low-paid service sector jobs, s ■ ■»»■ ■ reflect a growing marginalization of women workers who are 'fun< ! • n.ij once again as an industrial reserve army . . . and this fact is being 11 -kv by the rhetoric of family values' (Ruggie, 1988: 185) which is discussc«1 further below. The extent of paid leave (for sickness, holidays, parenting etc.) is : r: outstanding achievement by Swedish workers, with women taking [wo v three times as much paid leave as men. Esping-Andersen (1990) an.l Woodall (1990) refer to this rather pejoratively as 'paid absenteeism'. (»;: any given day 20 per cent of women workers are on paid leave of one kŕ i" or another, reaching a level of 30 per cent in the public sector. For motht i-with children under three, the proportion was 47.5 per cent, four times tlk national average for all workers. Esping-Andersen (1990: 156) on this hy-i-suggests that in Sweden 'a very large share of what normally is regarded i-labor time is in fact "welfare time" ' illustrating that 'the employe:-' control of the purchased labor-commodity is heavily circumscribed' However the increasing amount of sickness leave taken by women is probably also a direct result of stress-related illnesses including alcoholism and depression, as described by Ruggie (1988: 186). The levels of sickness leave amongst both men and women are much higher than in most other capitalist states, possibly reflecting a relatively generous and open-ended sick pay scheme. In 1990 the Social Democrat government announced a reform to make employers responsible for the first two weeks of sickness to cut down alleged abuse, despite strong opposition from the trade unions. Sweden: Social Democratic Welfare State 45 . ,.f .veiling the links between women's position in the labour market . -vir eligibility for income maintenance benefits is never easy. ".\.iiiioless, given Swedish women's overwhelming representation in ■ed paid employment, their eligibility in their own right for pension, í less, unemployment and other welfare benefits linked to that employ- V v must be commensurate, and much better than in states with a .„ icantly lower level of female paid employment such as the FRG. The vely advantageous position of Swedish women compared to their ,;j s in other countries is still limited of course by their lower earnings ., higher levels of unemployment compared to Swedish men and the high ,,, of part-time paid employment amongst women. I ihnlc Minorities and the Welfare State i.' „■• ition j „■ Swedish people are widely described as having one language, a single I .Hieran religious tradition and a common culture and history, without .__ntly being colonialists or being colonized. Native Finns, Lapps, Jews ■i.: Gypsies form long-established, relatively small ethnic minority com-■ n .uties. Recently the Lapps have fought a losing battle with the state over _!.i' .ig rights and access to land in the far north of the country. Prior to the ■'i . Sweden experienced large-scale emigration, especially from poor, n.i . areas, particularly to the United States. In the 1930s the Social Demo-% .-i. ideology of the People's Home and expansionary socioeconomic policies were legitimated in part by their declared purpose to reverse the falling birth rate. The building of the welfare state was clearly linked to the recovery of national identity and purpose. The fear of national population decline mingled with 'a mild sort of nationalism', according to Gunnar "y Myrdal (1938: 204), the influential Social Democratic sociologist. Writing <-0 for an American audience in 1938 he was highly ambiguous on the question of whether immigration would be part of a solution to 'the population question' (Myrdal, 1938: 203). In the same context, he also wrote that víc are just as much, and even more, interested in the physical, intellectual and moral quality of the population as in its quantity. Now at least in Sweden with its homogeneous population, quality does not depend on racial differences. (Myrdal, 1940: 203) Myrdal later became a very significant advocate of liberal race relations reform in the US. Here in the Swedish context he was hinting at a social eugenic attitude to race and immigration which was then, and is still today, a significant element of Social Democratic ideology both in Sweden and elsewhere, even if it is rarely made explicit. From the mid 1950s to the late 1960s migrant workers came to Sweden, mostly from Finland but also from Yugoslavia, Greece, Turkey and Italy. According to Widgren (1982: 153) most of them came on their own initiative, presumably referring to the Finns. Finnish workers gained the 46 Divisions of Welfare right of unhindered entry to Sweden in 1954, when Iceland, Denmas* Norway, Finland and Sweden established the Nordic free labour marK Swedish offices for the collective recruitment of workers were estahlish,-in Ankara, Athens, Rome and Belgrade. These were all closed down '•, the end of the 1960s, except the Yugoslav one which closed in 1977. Fp j-1967 onwards the government used existing immigration laws to lin-i. increasingly severely, the number of non-Nordic immigrants, so that 'si 1974 there has been practically no immigration into Sweden of non-Ncr ■. workers' (Widgren, 1982: 153). Non-Nordic immigration since 1974 :... consisted of the families of already-settled immigrant workers and poliu . refugees (defined much more liberally than most states) from many pari- .\ the world, including Latin America and South Africa. The regulations « admitting relatives of foreign residents are much more liberal than t host ■■■ the UK and the FRG. Also in the early 1970s immigration from I;inl;:i.; was put on a more controlled basis, which in effect means that a Finns-: migrant worker has to have a job in Sweden fixed through the sf. i. employment service before being allowed entry. The recession of the ein!-. 1970s led to a considerable exodus of Finnish workers; according i Widgren (1982: 149), '36 per cent of all Finnish immigrants who hi arrived in 1970 left Sweden in 1972.' Thus by the mid 1970s, as elsewhcr, .i: Northern Europe, a formidable immigration control system had K ^i, established, allowing entry to a regulated number of Finns, relative; :■! established immigrants and political refugees, but nobody else. This poi,. i was subsequently consolidated into Immigration Policy Acts of 1975 . :".i 1976. According to Hammar (1984: 29), in the day-to-day practice ■•■ immigration control at the borders, immigration officers have Ivi.i instructed since 1976 to identify non-Nordic people by their physit.. appearance, clearly an explicit form of institutional racism. The legitim. -tion of immigration control is presented in terms of enabling 'undcrei. -ployed national groups to have access to the labour market' (Widgre:. 1982: 151) and 'the protection of the Swedish labour market' (Hamm; ■ 1984: 28). Yet these considerations emerged well before the modo increases in open unemployment in the 1970s, and could equally well ha1-* applied in the 1950s for example to the recruitment of women. Anoih\ -official argument for immigration control made in the late 1960s was thai 'ii the number of resident immigrants became too large it would becoi::.-impossible to guarantee them a reasonable standard of living' (Hamm;-: 1984: 41). implied here are fears about the possible creation of hi underclass which might allegedly overburden the welfare state. The Status of Ethnic Minorities in 1981 about 12 per cent of the Swedish population were of mign i origin, of whom about half were Finns, and projections suggest thai "by ď-" year 2000, 20 to 25 per cent of the Swedish population will consist of pco| I* closely linked by origin to other countries' (Castles, 1984: 64). Will: .i Sweden: Social Democratic Welfare State 47 , ■ --ze permit, the status of the migrant worker and his/her family is !-. -.cure. The number of migrants without legal status is probably small *.'. all residents have to have an identity number, and the extent of the *.,, .-l.l economy is limited by the efficiency of the tax gathering system. ' . .. -lice, however, have been criticized for stopping people of non- v-i.::«. appearance for identity checks. After five years' residence in VA..V "' a foreigner can apply for and get Swedish citizenship (two years . \.i-dic citizens). Naturalization policy is much more liberal than most i.ipitalist states, though few of the non-Nordic residents apply for it, ., 11,. -i-, because they feel reasonably secure without it and because dual ■!.,-. p. hip is not allowed. I ľ io the mid 1960s there were no positive social policy measures .■■.. I * illy directed towards migrant workers and their families, who it was ,,i. k d would assimilate with the support of the welfare state. Between I ;.... i d 1974 a long process of public discussion and consultation took ■•! i.- ■■! whether the assimilationist ideology should continue or whether ,t iIiMilturalism should be adopted. Eventually the Immigration Policy V ■' 1975, which closed the doors to non-Nordic workers, embraced a ! im. . multiculturalist ideology towards Sweden's immigrant and ethnic nr.-ii'Y communities. The policy has three goals: equality of living *- ■■,.!. ids, cultural freedom of choice, and political solidarity between the ■h-i.\--ous Swedes and ethnic minorities. Thus the state has promoted -,- ■■ k ninority political, social and cultural organizations and activities, ■\j .'■ ig on a limited scale bilingualism in schools, and in 1976 foreigners Aiili ■ 'east three years' residence were given the vote in local and regional .L.i ■ .-s. In terms of equality in the labour market, in living standards and m we u are, it is extremely hard to assess the relative position of the ethnic minorities in Sweden. Survey evidence cited by Widgren (1982: Table 9) shows that foreign workers experienced much more dangerous and unpleasant working conditions; in the 1980s the open unemployment rate amongst foreign workers was double the rate for the labour force as a whole. There is little evidence of effective measures to achieve equality of access to and use of the welfare state in relation to the needs of the ethnic minorily communities. There is no equivalent to the Commission for Racial Equality in Britain, although racial discrimination was explicitly outlawed in the new constitution of 1976. However, ethnic minority trade union representatives and local politicians are becoming increasingly numerous and vocal in pressing forward such issues. At the same time in recent years, there has been a marked increase in overt racism and racist politics in Swedish society which has been documented and described by Larsson (1991). This includes racial attacks and harassment as well as discriminatory behaviour by the police, public officials and landlords. There is therefore some evidence that the new ethnic minorities in Sweden, mostly originating from the migration of workers in the 1950s and 1960s, do constitute a racialized underclass in Sweden, particularly the Southern Europeans. The extent to which the welfare state has furthered 48 Divisions of Welfare or undermined this process is unclear. Certainly immigration con. . policies developed since the late 1960s, while being more libera! viy-ii ■■ those already settled, are institutionally racist in discriminating ag, ,iv non-Nordic and, to a lesser extent, Finnish migrants and also in i-i,, requiring visas for entrants from Asia, Africa and Eastern Europe. Women and Family Policies Ideology and Family Policy Reform Swedish social policy is famous for its relative 'liberalism' (as opposu! >. conservatism) vis-a-vis some of the key aspects of family policy, refleu::;, widely held, long-established values amongst the population. For exan.:» .■ the legal status of illegitimacy was abolished in 1917, liberal divorce re'-'iT was introduced in 1920, homosexuality was decriminalized in 1944 . .ii compulsory sex and birth control education was introduced in schoeK ,i, 1956. Historically public support for lone mothers and their childrei -. been comparatively generous, and cohabitation has been treated rejatr-- ■■ tolerantly. Compared to most other Western states, defenders •> traditional patriarchal view of marriage and the family as an al'.i •• sacrosanct private institution have had much less influence. The origin i these liberal values lie deep in Sweden's cultural and socioeconomic hi--:""-(Myrdal, 1945: Chapter ill). By the 1980s, over 35 per cent of chi .' were born to unmarried parents, 30 per cent of families with children u:il!j eighteen were lone parent families (90 per cent being lone mother families and over one in two marriages ended in divorce, twice the rate in Britain Swedish family policies have also been shaped by a mixture of othe ideological traditions, similar to those which have shaped policy in othe states. These include the advocacy of birth control particularly for th working class as a means of combating poverty associated with larg families. Such neo-M alt husi an ideas had enormous influence around th turn of the century in establishing widespread use of modern birth con t re techniques by the early 1920s. The mixture also includes different variant of the feminist tradition, bourgeois, libertarian and socialist, which hav had two great waves of political influence in Sweden as elsewhere. The first wave peaked with the achievement of women's suffrage in 1919, and the second wave emerged in the 1960s. In the inter-war years, pronatalist thinking came to prominence as Sweden's birth rate became the lowest in Europe. As already discussed in relation to emigration and immigration, the declining population size in these years was a major political issue linked directly to fears of national decline. Swedish pronataiism does not embrace a rejection of birth control, but alongside the notion that 'every child should be a wanted child" the view was widely accepted that the state should try to encourage people to have a modest number of children. After coming to power in 1932 the Social Democrats began to integrate aspects of these various ideologies Sweden: Social Democratic Welfare State 49 heir own response to the 'population question' as it was called. This jeical synthesis was performed by Alva and Gunnar Myrdal in 1934 in s ■ book Crisis in the Population Question, which was later published in -panded English version as Myrdal (1945). The authors argued that , .. ation policy and socialist socioeconomic policies should be inextric- ' entwined so that the nation and the family would protect each other ,' "hrive. The fall in the birth rate was attributed by the Myrdals to ..r:-ived birth control, the decline of child labour and above all the „',- ised and unjust financial burden of child rearing on parents. The % i were portrayed as contributors to the upkeep of the nation who were . v .verished by that contribution. The goals of social policy should be to it.ise fertility by 25 per cent, to encourage 'medium-sized families', to .;■■■ (ive the 'quality' of the next generation by radical improvements in j |- welfare, to reduce illegitimacy and to encourage marriage, while at ..■ same time emphasizing the voluntary nature of parenthood in a ,.,■:■■'cratic society. These goals were planned to be achieved by a very ,,, .v range of collectivized services and benefits in all areas of human i. : re. The Myrdals take pains to distance their ideas from the policies ». i -.■ implemented at the very same moment in Stalinist Russia and Nazi u-nany. Indeed initially 'the Social Democratic press attacked the \!\ 'als' position as nationalistic and even militaristic, and some i ■ .>.;rvatives [saw] ... a backdoor effort to promote socialism' (Adams m Winston, 1980: 183) but pronatalist concern prevailed in the end. liiwii programme unquestionably put a strong emphasis on universal social ■velfare measures, not to be directed just at the working class or poor ■-■eople alone, yet they also expressed a strong commitment to using family velfare benefits and services to break down class inequalities, particularly mongst children. The Myrdals' approach to family policy is predomi-.anlly child-centred and resembles the view of children as valuable human .iipital in which the state should invest heavily to secure its future. The ■■rogramme was certainly not feminist in orientation and the effects of :imily policy on women's position in society are not directly confronted. Nevertheless Myrdal (1945: 121) acknowledged that ddending the right of the working woman to marry and have children becomes a protection of, and not a threat against family values . . . The forced celibacy or sterility among wage-earning women is a sign of society's incapacity to adjust itself to modern conditions. A population policy of democratic vision thus creates a new stronghold for married women's fight for their right to work. Thus the Myrdal programme of reform included parental leave and benefits, socialized day care and encouragement of men to take a greater share of domestic labour. Many elements in the programme thus went way beyond the confines of the Social Democrats' actual policy making in the 1930s and 1940s. The importance of the Myrdals' book was that 'population policy suddenly became a lever for socialists for far reaching sociopolitical reforms in collaboration with the conservatives, who were prepared to bank on the preservation of the Swedish national stock' 50 Divisions of Welfare (Liijeström quoted by Scott, 1982: 13). Population policy alom>. Keynesian counter-cyclical public expenditure policies became the le-i mation for the expansion of the welfare state, and to a large extent i» remains true today. Family policy reforms in the 1930s in the wake of the Myrdals' re;-. -included free maternity care in public clinics, special public hou'-i-.' schemes for large families, rent rebates according to family size, chile \ allowances, marriage loans, guaranteed maintenance for some ! ■■„ mothers and employment protection for mothers. In 1948 universal c1 benefit was introduced to replace the inegalitarian child tax allowan-.. Child benefit is not statutorily indexed to inflation, but it has largely %.; pace with retail prices (Olsson, 1986: Graph 20), adding about 5 per <.■.. per child to the gross earnings of an industrial worker. In the 1 ■*■! statutory provision of birth control and family planning advice by -I;., health service and statutory maternity leave were introduced. Not n absent from this list is liberal abortion reform, and day care and nui\ ■ education for the under sevens (primary school starts at the age of sewi both of which represented too much of a challenge to traditional ideol.-.. The effects of the reform on the birth rate are debatable; it stabilize.1 . elsewhere during the mid 1930s, probably linked more to econ< n recovery and the fall in unemployment. A second wave of discussion of family policy reform began in the ■ u-, 1960s. Pronatalism remained on the agenda, particularly amo'i» conservatives concerned that most parents were increasingly limiii:i. themselves to having one child. Family poverty was also of gro"\ concern, with the rediscovery of poverty amongst families with depem .-i children, particularly amongst lone mother families. Of much gre ik significance than these two factors and in contrast to the 1930s, the .hj feminism of the women's movement together with concern about i': implications of and needs created by the growth of women's |\i.. employment were both firmly on the political agenda (Liijeström, 15 "* These two pressures, which of course are closely linked, contributei i-what is called the 'sex-role equality' movement, an element of the u iL movement for social equality of outcome in the welfare state wi.i radicalized the Social Democrats in the 1960s. Beginning in the early 1960s a largely female group of sociolofK* economists and psychologists documented the effect of women's p i-employment on the family and on children, and also the impaci ■ patriarchy in the home and in paid employment on women's job oppoi" i ities. They advocated the break-up of the sexual division of labour ir i" home, and policies of equal opportunities and affirmative action ou'-1: the home in the welfare state and in employment. The Social Democ:.---women's organization took up these ideas in 1964, and they v.-eventually incorporated into the party's Programme for Equality ado"K-in 1969. The document was by no means, however, a feminist manif< *: for as Scott (1982: 7) explains, Sweden: Social Democratic Welfare State 51 ,■ as Sweden's early social measures were accepted because they were '..'".mted as the solution to a 'population crisis' ... so the sex-role equality ',. -am was assimilated because it was part of an 'equality program' that had '".. thing for everybody . . . Thus some of the more fiercely debated measures * ng to the status of women (or the status of men) were packaged so that the ..\' iion of women was only implicit. ,.,..-t the many policy reforms ensuing from the sex-role equality ■wi.ient have been liberal abortion law reform, extensive statutory n"i.nisi leave, and a great expansion of day care, all of which are \ n. ied in detail below. Also unemployment benefit was extended to '..n ...intributor housewives. Possibly most significant of all, in 1971 .. ^.-vndent taxation rules were introduced for men and women, which ', j, t 'more profitable for a wife to work even at a part-time job than for : s ■ sband to take on a few extra hours' (Scott, 1982: 72). : .-■. Mothers ; k position of lone mothers provides a good illustration of the achieve-■■i,-::i- and the limits of family policy in contemporary Sweden. In 1983 19.0 i-.i cent of families with children under eighteen were lone parent families .. hL- A.16), around 90 per cent being lone mother families. The .i .i.-iii-tion of lone parent families is much higher than in Britain and the I \u i. but somewhat less than in the US. Only 9.2 per cent of lone parent ľ :lk:s in Sweden are poor, which is a very much lower proportion than in :hc ■ ■ her welfare states examined here. In 1979 86 per cent of Swedish lone :i nil. rs had paid employment, a very much higher proportion than in the ,: .-. countries studied in this book, and much higher than the proportion . . ncourages the establishment of lone parent families. On the contrary ^-ii'i-ys and statistical analyses show unequivocally that lone parents are «VM'-mically disadvantaged in many respects' (Kindlund, 1988: 76). In I'M 20 per cent of lone mother families were dependent on social ■w*i. nee, compared to 9 per cent of lone father families and 3 per cent of ■"■a » went families. There is no guaranteed minimum income for lone ;■ -. .is, but Advanced Maintenance payments were introduced in 1937 i:d subsequently considerably improved. The index-linked, flat-rate pay-■iv ■ «s about one and a half times the rate of child benefit but 'payment is »'n n ional on the custodial parent assisting in efforts to establish ľ-M lity' (Kindlund, 1988: 89), so that by no means all lone mother l in1: es benefit. About 14 per cent of Swedish children in both lone parent ■»•I ■ iconstituted families benefit from maintenance advances, which are ľ h- vith child benefit. The government recovers about a third of the 52 Divisions of Welfare expenditure from fathers. Lone mother families in Sweden are tin probably more generously supported by the welfare state than m an-country in the world, yet they are far from having achieved equali two parent families. The welfare state merely succeeds in lifting r them out of dependence on means-tested assistance. Abortion Until the liberal reform implemented in 1975, the struggle over a I rights was waged as fiercely in Sweden as in most other countries. I amidst great concern about apparently increasing numbers of abortions, a government committee suggested that in certain circum abortion should be decriminalized. These circumstances included n cases of rape and hereditary problems, but also social indications where mothers in extreme circumstances were 'worn-out', living poverty', likely to suffer career damage or likely to be 'disgraced' shame of unmarried motherhood, or where the birth of a child ihrt a permanent relationship. However the Population Commission si translate the Myrdals' ideas into policy, rejected the Abonioi: mittee's report, arguing that the social problems confronted by r should be mitigated and prevented by other benefits and sen ice?; welfare state, particularly birth control. The only social indication ľ abortion which the Population Commission approved was for the out mother', for whom the birth of another child would be life-thre; The Population Commission view prevailed and legislation in ll.) decriminalized abortion where there was serious danger to the it life or health, which was interpreted by doctors very conservatively same year, 1938, legislation forbidding the sale and distribu contraceptives was repealed, which prompted one of the pioneer; education, Elise Ottesen-Jensen, to comment that 'the beautiful p of voluntary parenthood which the Population Commission uphold contraceptive question kicks the legs out from under itself in the a question' (quoted in Liijestrom, 1974: 43). The very limited socio indication for legal abortion was relaxed slightly in 1946 in the con t sharp struggle over the issue between the Conservatives and some I on one side, and women's organizations and the blue collar trad federation on the other side. The number of illegal abortions was es at more than double the number of legal abortions in the 1940s am but in order to get a legal termination a woman had to convince the doctor ... of her own weakness and inability to handle affairs. She had to bow down to the diagnostic rules of the game and asf role of an insufficient, depressive or neurotic . . . She could not be coi responsible and mature. (Liijestrom, 1974: 59) During the 1960s with the rise of the women's movement and c social and medical attitudes, the number of legal abortions wit indications began to rise, until by 1974 as the reform legislation w; Sweden: Social Democratic Welfare State 53 ..i. parliament, illegal abortion had all but disappeared. Yet it took '. ,■„ for the Swedish government to enact a liberal reform of the ^r law. In 1965 a government commission was appointed, which in ,. j-, rted that women should have the right to early, safe and legal i-i'ions, but that abortion should be regarded not as a form of birth .;. : ut rather as 'a last resort'. Eventually legislation went through [Ticiit in 1974 and came into force at the beginning of 1975. The law women the right to a termination up to the eighteenth week of yncv unless a doctor believes that it would put the woman's life or i at risk. After the twelfth week a social worker has to investigate the n's circumstances in order to inform the doctor's decision. A doctor's t] to perform an abortion before eighteen weeks is always referred to oard of Health for review. After eighteen weeks, all decisions are ed to the Board of Health, informed by a social work investigation, .ion is illegal if the foetus is viable, except where the woman's life is ened, but the definition of viability is left to medical discretion. : 1974 abortion law reform appears to have won wide acceptance in ;n, and the issue seems to have disappeared from the political agenda ;suh. The number of abortions has remained roughly the same since ; the Act, at a level of twenty per thousand women aged fifteen to ("our, comparable with that in the UK. A very significant achievement ; legislation and of the health service is that 95 per cent of rations are performed by the thirteenth week of pregnancy, which is safer for women and much cheaper for the health service. This ucs very favourably with the situation in other states. ''are and Parental Leave rowth of women's paid employment from the 1960s onwards and the mitant rise of the sex equality and women's movements have made a i» impact on social policy concerning the parenting and care of ju. particularly pre-school children, that is those under seven years iy 1983 the great majority (82 per cent) of mothers of pre-school jn had paid employment, in sharp contrast to the situation in the three states examined here (Table A.15). Policy has developed in rections, first by greatly increasing public investment in day care and .1 by giving parents very substantial statutory rights to paid parental to care for children. These two obviously complement each other, ne supporting out-of-home care particularly for parents in paid ynient, and the other supporting parental care in the home. It is able that both policies are designed primarily to benefit parents in employment and their children, rather than being universally of t lo all, whether the parents are in paid employment or not. This no reflects the very high proportion of women in paid employment in :n, though it is perhaps surprising that the child care debate has not nore child-centred. 54 Divisions of Welfare The Parental Insurance scheme was introduced in 1974 and haj- ! improved regularly since. The basic entitlement is twelve months1 k , 90 per cent of gross earnings funded by social insurance, a period whk 1.1 . be shared between the parents as they wish. The parent has to haveu. -,.". six months for the same employer, or twelve of the preceding twem-.-i.." months to qualify. Parents cannot take leave simultaneously but it t . , . divided up into intermittent blocks. According to Rapaport and ,\:..., (1989: 10), a further .six months of unpaid leave is available and the entire-éighteen .. ,■ will become paid at 90 percent earnings from 1991; in-effe'et this will ex .■„■ around twenty months because people on parental leave are also ent: . ■ annual leave ... in practice most parents took leave in a full-tirm " . _. immediately after the birth of their child. Since July 1989 the time limit fos ■ "■ leave has been extended until a child is eight. The government would like parents to spread leave over a longer iv. and/or use it on a part-time basis. Other leave entitlements at 90 per i.\ m,. gross earnings are 'fifty days pregnancy leave, ten days paternity k .-.^ ninety days per year per child under twelve to care for a sick chiK i-i ■ care for a child if his/her normal caregiver is ill [average leave in 19>>- ".... seven days per sick child], and two days leave per year to visit a ch .-. nursery or school' (Rapaport and Moss, 1989: 11). All these other v • _. are funded by social insurance, and are available to all employees w rii ■ ■ qualifying restrictions. Finally parents have the right to work a six hoi. ., in paid employment until the child is eight. Unfortunately there u i:- official statistics on the take-up rates of parental leaves, but they . -v assumed to be high, as reflected in the 'absenteeism' figures di- -.«^ above. In 1986 83 per cent of fathers took paternity leave, but fathe' .-* n\ of the other leaves is modest, except for leave to care for sick chilo . -i ■: which fathers took 44 per cent of the days in 1985. Hence in 1984 in >' -. six months of the child's life, fathers took only 1.9 per cent of th. . ■■* days, rising to 9 per cent in the period after the first six months. B-i^-i! 1974 and 1980 fathers' usage of parental leaves increased relative to mothers', but since then there has been a modest decrease. Sundqvist (1987: 151) attributes this to 'the economic situation', that is falling real incomes and less employment security. Role reversal in Swedish families is clearly very unusual. Rapaport and Moss (1989: 38) gained the impression that 'fathers in Sweden take a greater share of family responsibility than ... in the UK . . . [but] there is insufficient evidence to confirm (his.1 Government equal opportunities policy in the 1990s will attempt to encourage more take-up of parental leave by men. Overall it is very clear that Swedish social policy is comparatively interventionist in encouraging parental care of children, in contrast to mere ideological exhortation by policy makers and politicians more typical of other states. Public provision and funding of day care for children, both after-school for older children and pre-school for children under seven, has probably been the most significant new service development in the Swedish welfare Sweden: Social Democratic Welfare State 55 0ver the past two decades. As usual with day care, changing j0ns and parameters make precise quantitative measures of the sion impossible. In 1967, according to Berfenstam and William-., (1973: Table 13), children under seven with parents in paid vtiwit were overwhelmingly cared for privately, mostly at home by •h, relatives or partners, but also by private childminders. Only 16 nt of them attended full-time or part-time nurseries or nursery pre-Si By 1987 of all children under seven, 53 per cent were still cared for »ly. again largely at home, but the other 47 per cent were cared for tried childminders or municipal day care (Broberg and Hwang, 1991: 5.d). In 1987 34 per cent of the under threes received care in ipal nurseries or from salaried childminders. The government aims )1 to provide a public day care or salaried childminder place for all ;n over eighteen months old whose parents are in paid employment, ng or training. ind these figures lie the two major developments of the last two ;s. the development and expansion of both an administratively :ited municipal day care service and a municipally salaried child-r service. The day care service includes nurseries, mothers' clubs and me kindergartens. Provisions vary widely amongst local authorities, of whom provide few or no places for the under threes. Most local -dies also organize salaried childminding schemes: tlminders get a fixed monthly salary, but on condition they look after at least children full-time or provide the equivalent hours of care for children idisig part-time; many have to look after eight-ten part-time children to get ■ salary. One reason so many children attend organised childminding mes on a part-time basis is that many municipalities only admit children ling full-time care to nurseries, arguing that nursery places are too costly U) be used part-time. (Broberg and Hwang, 1991: 78) ccount certainly gives the impression that municipal childminding is icly hard work, done (almost exclusively by women) for a very modest salary. Fees paid by parents cover between 10 and 15 per cent of the cost of municipal nurseries and salaried childminders, and are usually income-related. Lone parent families are prioritized for nursery places, but children of immigrant parents are underrepresented in municipal day care (Broberg and Hwang, 1991: 90). The dual provision of salaried childminding and municipal day care reflects a long-established political division of opinion, which has remained prominent on the political agenda over the last two decades (Adams and Winston, 1980: 99; Broberg and Hwang, 1991: 96-8). On the political right both the Conservative and Centre Parties have opposed the expansion of municipal day care, regarding the public investment in it as discriminating against parents who stay at home with their children. These parties have argued that the state should encourage more parents to stay at home with a universal child care allowance. Parents could then make a choice between staying at home or buying day care, only modestly if at all supported by 56 Divisions of Welfare public subsidy. Non-socialist local authorities have tended to salaried childminding as a more flexible and familial form of da\ . .. There has been considerable tension between more traditionally n,i.,. * local authorities and the central government over the expans ■ .. municipal day care (Ruggie, 1984: Chapter 6). The non-socialist i* ■-. have focused on the class inequalities which have emerged in the ^ municipal day care. According to Broberg and Hwang (1991: 130 ■,, higher the parents' educational and occupational status, the earlii ■■- . place their children in day care outside the home.' The blue collai • union federation (LO) considers this is due to the limited opening h .■ ;i (6.30 a.m. to 7 p.m.), and have long argued for twenty-four hour mi i U| day care to be available. In fact, however, these class differences in * ._ probably reflect more a difference in cultural attitudes towards par.vi-..-to which the non-socialist parties have appealed with some success. • 'n other hand middle class parents in rural and/or traditional local au . ■■■■■ areas complain about shortage of nursery places, where priority for p. ^ goes to lone parents and poorer children. The expansion of publicly funded day care in Sweden, lik. ■., development of the welfare state, has thus brought into focus a nun '■. ■■ class and gender issues. Publicly funded day care is by no means unn „-i'.. . available to all who want or need it; much of it is on a part-time basi- uui ■ is not clear whether this is what mothers want or whether it is deteiu'iii,.. by employers' demand for part-time workers and the limited ' ■., resources for day care. The New Patriarchy? A number of contrasting analyses of Swedish family policy have bu-i p forward. Most analysts agree that the reforms since the 1960s are '■ 'L ■. directly with the demand and opportunities for women's paid emplo-. ■■ ■ r Ruggie (1984: 299) for example explains the expansion of day * * i Sweden compared with its stagnation in Britain in terms of 'the diffi - .-,k^ in the role of the state toward women workers' linked to 'the resp ■ •■ . status of labor and its relation with the state'. Ruggie stresi - i'--significance of Swedish corporatism, the partnership of the labour movi ment, the employers and the state, in the promotion of employnicn-opportunities for women. As a result 'perhaps the most significai. development concerning women workers in Sweden is the tendency towai ' "universalization" of the category of worker . . . the distinctions amor., workers based on class, occupation and sex, are breaking down am becoming less determinant of workers' opportunities and rewards' (Rtiy gie, 1984: 340). This scenario is perhaps somewhat overoptimistic in tk light of the evidence on gendered segmentation of the labour force, but -reflects a widely held view that the forward march of women in Sweden h" been closely intertwined with the forward march of socialism and th.-labour movement. Ruggie (1984: 24) dismisses feminist interpretation* Sweden: Social Democratic Welfare State 57 ie says cannot 'adequately account for the differences that are n policies for women in Britain and Sweden'. Adams and Winston iowever explain such differences between the US and Sweden in ľ different feminist strategies in the two countries. They describe . women as having pursued 'social feminism' through social policy i while women in the US have pursued 'equal rights feminism' i legal and constitutional reform. Adams and Winston suggest that women have chosen to advance their cause from within estab- d political and social organizations rather than through more auto- ous forms of organization. Social feminism in Sweden has thus played -niiicant role throughout Swedish civil society, not just in the Social Hieratic and labour movements. As Adams and Winston (1980: 157) t out, 'it can hardly be argued that the welfare state is a response to the Is of working women because most of the Social Democrats' welfare rammes were introduced before women entered the labor force in ficant numbers.' Women's activism in shaping the population policies \c 1930s so that they were implemented without a strong ideology of esticity has been a feature of Swedish family policy, which dis- lishes it significantly from the Stalinist and Nazi variants being eniented at the very same time. ie successful incorporation of social feminism into mainstream politics have held back the development of a more critical and more radical nisi movement. Scott (1982: 158) suggests that self-organized, auto-ous feminist initiatives such as rape crisis centres, shelters for battered en. self-help women's health clinics are rare in Sweden. The word iniMrť is more widely associated with a revolutionary, 'separatist' view in most other Western countries. Yet the need for such organizations ars to be just as strong, and they have begun to emerge in the 1980s ' :gan, 1984: 661-3). As Scott (1982: 157-8) says, > a paradox of the Swedish welfare state that it has absorbed so many feminist iiuinds, and yet women in Sweden appear to find it even more difficult than :\ do elsewhere to keep 'from being devoured by equality under the terms of a ik value system' . . . Men feel that they have 'done a great deal for women' . hut they expect it all to be done in the framework of existing male m-itilutions. Such analysis of the limits of the Swedish welfare state in undermining patriarchy has been extended recently by a number of Scandinavian women writers. Hemes (1987) sees Scandinavian welfare corporatism as a reconsiitution of patriarchy in male-dominated corporate institutions, upon which women are particularly dependent. Women are much more reliant nn the welfare state for paid employment than men, and their daily lives depend much more directly on the services and benefits of the welfare state i han do men's lives, because women still carry the greater burden of unpaid domestic and caring work. Yet real power in the upper reaches of welfare administration and in the wider corporate institutions is predominantly in the hands of men. Hemes (1987: 76) suggests therefore that 'one 58 Divisions of Welfare can describe the Scandinavian state form as a tutelary state for . since they have had a minimal role in the actual decisionmaking concerning distribution.' Borchorst and Sum (1987: 154) also suĽ»t as a result of the substantial gains for women from the developmer welfare state, patriarchal power is now maintained by 'the instituti tion of women's dual roles as mothers and workers', which has 1 free to consolidate their power in public and private corporate tnsti The Scandinavian welfare state has thus had a contradictory im women, sweeping away some of the traditional patriarchal assu about the primacy of women's unpaid domestic work, but esiabi new form of patriarchal citizenship for women in partnership \ welfare state as both employer and provider (Sum, 1987 and llJ8f The Health Care System Finance and Administration Technically speaking Sweden does not have a national health ser\ to that of the UK or Italy. This is because the bulk of health care and finance is not under direct central control, but is under the a the upper-tier, elected local authorities - the county councils. In out-patient consultations and GP services are financed through i semi-autonomous sickness insurance funds. AH residents of the < whether contributors or not, are entitled to hospital, out-pati primary care. Patients pay a fee for GP and out-patient consulta 1984 the charge was £5 for a private doctor (registered with the ir fund) and £4 for a public doctor. There are prescription charj subsidized fees for dental care. Hospital care is free, though 1c chronic patients have to pay £3 a day. Navarro (1975: 74) is sun correct to point out that 'it is misleading to refer to the fundin Swedish health system as an example of national health insurance' access to the system is not based on contributory eligibility rules ar cent of the funding comes from local and national taxation' Using a different parameter one could argue that public sector heall: care is more comprehensive than in Britain. Thus in 1984 public expenditure on health care in Sweden constituted 91.4 per cent of total health c;u\ spending, the second highest amongst OECD countries, compared to 88. ■ per cent in the UK (Table A. 19). There are only two small private clink* offering acute medical treatment, and only 5 per cent of doctors operate«: totally in the private sector in 1985. As M Gordon (1988: 210) explains 'although patients ... go to private practitioners, the great majority of tlu-practitioners were associated with hospitals or public dispensaries, nnu their private fees supplemented their basic salaries from public source««' Britain and Sweden were amongst only six (out of twenty-three) advance-1 capitalist states offering 100 per cent public coverage for hospital care ;■-early as 1960. Most striking in comparison with Britain is the proportion of Sweden: Social Democratic Welfare State 59 resources devoted to health care (Table A.20). In the UK in 1984 re spending accounted for 5.9 per cent of GDP, compared to 9.4 in Sweden, a proportion only exceeded by the US. In terms of jth care spending per capita (adjusted for purchasing power weden spent $1445 in 1984, compared to an OECD average of iCD, 1987: Table 20); see also Table A.21. In the years 1960-75 ih of health care spending as a proportion of GDP growth was Sweden amongst eighteen OECD countries, but by the 1980s it :■. igst the lowest (Table A.22). According to the Swedish National Health and Welfare, in the 1980s 'the State has imposed heavy i restraints on the county councils [who provide the bulk of health n the effect that medical spending has been cut from 9.7 per cent 1 [in 1982] to 9.1 per cent in 1986' (quoted by Diderichsen and , 1989: 222). This suggests that like the FRG and Britain, Sweden : quite successful in containing health care costs in the context of crisis of the state, niinant administrative and financial power in the Swedish health ;m lies with the twenty-six county councils who run the hospitals i of the primary care through out-patient clinics. The post-war ars saw a massive investment in new hospitals; hospital-based medicine was hegemonic, with much less investment in primary preventive health measures. In order to rationalize this growth, in ral government introduced legislation which prescribed in detail rules and standards for the provision of hospital care. Govern-its for new hospital building and training of physicians were used 'age compliance. By the late 1960s, 'primarily because of con-unrest among the populace over ever increasing local taxes' , 1975: 5) which largely finance the hospitals, the health care machinery was strengthened and began to have significant effects. ,e slowdown in growth since the early 1970s. In 1982 central itic control of the county councils was significantly diminished, legislation was abandoned in favour of decentralized administration, in which detailed central regulation and close supervision is . . . replaced by broad planning and coordination . . . The state now relies on two means of directing the activities of the county councils: by giving special grants for the kinds of health programmes it wishes to see . . , and by framing special conditions in the slate insurance system which make the county councils more interested in certain programmes . . . State grants have been channelled into psychiatric care as well as preventive care. (Lane and Arvidson, 1989: 86-7) This reform diminishes the direct power of central government over the management of health care, but, of course, in this system local politicians arc under acute pressure to limit costs, because of their impact on local taxation. Hence during the 1980s there have been few new programmes and an increasing emphasis on economic efficiency. There is a move towards making clinicians directly accountable for their budgets, called the 60 Divisions of Welfare 'base unit system'. This may generate better financial discipline accountability, while 'the autonomy of the health professionals seems t retained if not increased at the clinical level' (Lane and Arvidson, ] 94). The complexities and contradictions of decentralization and consciousness in health care are enormous. Power Struggles in the Health System On the face of it health care issues in Sweden seem to be very depolitic in the sense that overt political or social struggles around health care ij seem to be rare. This may reflect the fact that, comparatively unusi. the health service is formally under the control of locally elected p cians. Inevitably this means that the consumer, trade union and fessional politics of health care is carried on within established local j political structures. The politics of health care is therefore dir incorporated into mainstream politics. To the outside observer, it apj in fact that decision making about priorities, resourcing, industrial tions and so on are highly bureaucratized and largely dealt with by ce and local government administrators in partnership of course wit! medical profession. Ever since the ascendance of Social Democracy in the 1930s then been an ongoing struggle between government and the doctors ove status and remuneration of the profession, behind which lies the univ conflict over clinical autonomy and attempts by the state to co professional behaviour. The Social Democrats came to power in the with a commitment to establish a national health service in a come which only 30 per cent of doctors were employees of the county co hospitals. Medical care was dominated by private practitioners oper from their own offices or from the public hospitals on a fee-for-sc basis. By the 1970s, as mentioned above, the great majority of dc were salaried employees of the public health care system. This transft tion did not come about without conflict, which came to a head ir instances. First the government commissioned a report, published in on out-of-hospital and out-patient medical care. In fact the five memb' the commission could not agree, which is extremely unusual in Sw< but the minority report by the Director General of the National VU Board, Axel Höjer, reflected Social Democratic thinking. Höjer's \ic his own words, was that all medical care services needed by the individual should be offered Ii charge at the time of treatment. It should be the duty of the community m i this by means of an extended and (comprehensively) regulated organh including and coordinating public health, hospital care, preventive medic an individual basis and ambulatory medical care. Especially ambulator) m care, which has largely been left to medical doctors' private initiative ... i be extended and regulated by public provision, (quoted by Serncr, 1980: In other words Höjer wanted a fully salaried, publicly employed m< profession, a system which integrated preventive, curative and Sweden: Social Democratic Welfare State 61 ;s, free at the point of consumption. His vision was not unlike that of rttemporary in Britain, Aneurin Bevan. Like Bevan, his proposals o severe opposition from the medical profession, represented by the 5h Medical Association (SMA). As Anderson (1972: 78) explains, ■ was a declared Socialist, and even in Sweden, such a label does not ifortably with the medical profession.' The Social Democrats did not . ze sufficient public and trade union support for the Höjer proposals; iehind-the-scenes, bureaucratic strategy for reform failed. The SMA nore successful in rejecting Höjer than the BMA were in attempting ; ;ct Bevan. Hence when finally in 1955 the government implemented ilsory health insurance to cover ambulatory care, the out-of-hospital , s continued to work on a fee-for-service, private basis with patients irsed for a certain proportion of the fee from the insurance funds. ling the 1950s and 1960s the nature of the medical profession ;d; the old guard who stoutly defended solo practitioner status illy retired. The concentration of new investment on in-patient care citadels of the system, the county hospitals, led to the atrophy of it practice. More and more doctors were hospital-based and salaried. ,al conflict in the SMA centred around income inequalities among ral doctors, particularly between those who benefited from the fee-ľvice work and those who did not. This coincided in the 1960s with "lergence of the 'equality movement' amongst the Social Democrats. in 1970 the corporate rationalizers in the health care bureaucracy ded in making the hospital doctors totally salaried and in abolishing 3-for-service basis for ambulatory care. This was called the 'Seven is Reform', named after the original flat-rate fee of around 70 pence. t a two tier flat-fee system emerged, with a higher charge to see a ; on a 'private' appointment basis, and a lower fee to queue to see a on a 'public' basis. As already discussed, since 1970 the great ity of the medical profession have in effect become salaried civil ts. Indeed in 1971 the doctors, alongside other members of the sector professionals union, SACO, went on strike very briefly in a struggle with the government, which was broken by emergency tion. The doctors had finally entered traditional collective bargain-id so some of Höjer's principles had come to fruition. Inequality in to ambulatory care was sharply reduced, as the cost barriers which ed those on low incomes came down. Salary differentials amongst lists were diminished, to give more priority to lower-status specialties s psychiatry and geriatrics. The Seven Crowns Reform was one of ajor achievements of the equality movement within the Social cratic Party, and its successful implementation 'demonstrated how tower the Swedish medical profession exercised' (Starr and Immer-)87: 237). ing the 1970s and 1980s the relationship between Social Democracy e SMA appears to have been more harmonious, though not without :ant conflicts. In 1979 the counties sought to force doctors to move to 62 Divisions of Welfare underresourced areas, to which the SMA responded with a campaig 'the protection of free enterprise1, but in general doctors seem to collaborated with the planners. Most recently in 1982 the SMA rei aspects of the decentralization of health service management, desc above. An important element of this managerial reform was the separation of management responsibility from clinical responsibility the notion that final medical responsibility lay with the county cc rather than with particular physicians. The SMA lost this battle paradoxically, with the virtual freeze on real growth in health spending in the 1980s, the county councils have had limited root manoeuvre in shaping clinical priorities and detailed spending. The cc between the corporate rationalizers (the politicians and administr: and the medical professionals continues in the county hospitals' con of power. Health Status and Health Inequalities The health status of the Swedish people is unquestionably one c highest in the world. The infant mortality rate has consistently amongst the lowest in the Western world, 6.8 per thousand live bir 1985. Life expectancy is amongst the highest in the West, signifii better than that in the other capitalist states focused on in this bool these parameters Sweden performs about the same as Switze (OECD, 1987: Tables 8 and 9). Using a composite ranking of seve age- and sex-specific mortality rates for ten advanced capitalist states mid 1970s, Maxwell (1981) ranked Sweden in first place, some way ; of Switzerland which is a comparatively low spender on health care ( A.23). Such statistics are perhaps largely a reflection of the very high standard in Sweden rather than the welfare state in general or the 1 care system in particular, though Maxwell's findings indicate some pc differential impact of the social democratic welfare state. As in Britain, during the 1970s three critical and interrelated issues to prominence in health policy in Sweden: diminishing resources, technological demands and costs, and faltering performance in ter health status and health inequalities. The 'technocratic ideology' emphasizes a hospital-based, curative, high-tech approach to health t particularly hegemonic in the Swedish health service. Hence, in-p hospital care is much more predominant than in the US or Britair ambulatory care correspondingly much less common. This has bej; change with out-patient primary care doubling its budget share i 1980s. The emphasis on hospital technology may be explained in p the keen commitment of the Social Democrats to technological rat ity as the handmaiden of industrial and social progress applied to ! care (Diderichsen, 1982: 195). By the mid 1970s political opposit such assumptions, applied to nuclear power for example, was g momentum in the shape of the green movement. The Social Deiru Sweden: Social Democratic Welfare State 63 ■rconfiticnt commitment to nuclear power was a major factor in their eat in 1976. ■l0 the familiar dilemma of increasingly articulated health needs and wííiľ pressure to limit resources led not only to administrative ■cntrali/ation in 1982, but also to a renewed attention to preventive ilth care and equality of health outcomes. Unlike in Britain, it does not lear th^1 these issues were particularly pressed forward by the public tor unions and the left, because they have never been particularly .niim-'iil in public debate (Carr-Hill, 1989: 40). Nevertheless social MilNt* and epidemiologists in the 1970s discovered significant inequal-s in mortality and morbidity. Le Grand (1989) examined the variation in rialilv ;tinongst the population as a whole in a large number of industrial ieiics. He found that although Sweden had relatively low overall qiialiu of mortality amongst children and infants, for adults the Swedes no better than the average variation in mortality in thirty-two mines. Data on social class differences in mortality is inadequate since trove r n ment does not recognize such parameters, but on the basis of the ited evidence Erikson (1987: 56) concludes that 'there are some, albeit !ier ^mall, differences in mortality between social classes in Sweden.' Ikonen (1989) suggests that adult mortality in six countries including sden is correlated with levels of education for both men and women, posting another aspect of the effects of class. Using data from the Living ňdarďSurveys of 1968, 1974 and 1981, Kjellström and Lundberg (1987) nd significantly higher incidence of health problems amongst women, eiiy people and the blue collar working class, with their degree of advantage remaining largely unchanged since 1968. As in other indus-.1 societies, women in Sweden suffer more ill health than men, despite ir greater life expectancy. The use of health care services by the less ihhy groups in the population increased significantly over the period, ;gesling perhaps that inequalities would otherwise have widened, verlheless access to health care remains more difficult for pensioners, isewives and the unemployed compared to those in paid employment iderichsen, 1982: 194). This may reflect the fee-paying basis for bulaiory care and the fact that poorer people have to queue and wait ger lb r appointments, than those who can afford the higher fees to see a :tor on a 'private' appointment basis. The limited evidence suggests that welfare state as a whole may have had only a limited effect in curbing ■>s and gender inequalities in health status. "Hiring the early 1980s expert reports commissioned by the government these issues prompted the passage of the 1985 Health Policy Bill which ogni/od that pcciiil attention should be paid to those groups of society who are at greatest i^k . . . manual workers and the low-level salaried employees; immigrants, ^»jí-tĽiin unemployed, people living alone, divorced men and among children >t purem s with low incomes and social and mental problems, (quoted by iiililprcn and Diderichsen, 1986: 536) 64 Divisions of Welfare Surprisingly the health problems of women are not specifically men The Bill also advocates extra resources for prevention, public epidemiology and primary care. The Bill is a rather uneasy comj between three radically different forms of policy prescription - c ative, social democratic and a radical public health approach. C atives, including the Swedish Conservative Party, emphasize comi and privatization to increase real resources and individual responsi!: health problems related to lifestyle. Mainstream Social Democr emphasize the benefits of scientific medicine allied with ration effective planning and management in the public sector, and the ance of medical technology and drugs for Sweden's exports. Impro of economic growth would generate more resources for health c lower unemployment, seen as a major contributor to health problei radical public health movement advocates a fundamental switch c ities towards primary care and preventive measures, targeting healthy groups. They would shift the policy emphasis from im lifestyle changes towards highlighting corporate and public respo for ensuring informed choices for all groups. The first Public Report subsequent to the Bill (reprinted in Diderichsen and Li 1989) hesitantly moves a little way towards the radical position seems unlikely that the Bill signals a really dramatic shift in healt and priorities in Sweden. Conclusion Since the 1930s, the Social Democratic hegemony has shaped the I welfare state so that it is very significantly different from the other states examined in this book. The hegemony is based on a de collaboration between organized labour, capital and the state wl not existed in other welfare states. In this partnership, private continues to exercise control over industry, finance and commerce much threat of political interference. The Swedish economy is ow controlled by private interests to an astonishing extent, when one o how frequently Sweden is described as a socialist country. One fan Wallenbergs, has effective control of more than a third of the total the Stockholm stock exchange. Yet this extremely dynamic c economy has thrived alongside the development of an extensive state. Over the years since the 1930s, the welfare state, as we ha' has succeeded in ameliorating class and gender inequalities in though also reconstructing and even sustaining forms of class, ger racial inequalities and oppression. The Social Democratic he would suggest that the successful development of the welfare stať the capitalist economy is a symbiotic process, but this is impoi prove. Since the 1970s, as in other countries, private capital and ev Social Democrats have come to believe that the welfare state is £ drag on further economic growth. On the left, it is argued Sweden: Social Democratic Welfare State 65 »ed activity of private capital has created extra burdens for the state in terms of unemployment, retraining, social dislocation, stirement and so on. Unquestionably in the 1990s the Social ratic hegemony is going to bear its greatest pressures since the This became clear in February 1990 when the Social Democrat nent resigned after failing to get an austerity package through ent, which included a two year wage freeze and a ban on strikes in ie for continued support for the welfare state. The government ■ w the wage freeze and strike ban, but still now proposed higher taxation, cuts in social benefits and postponement of the planned i )n of parental leave in 1991. Behind these struggles He a number of imilar to those facing other welfare states. there is the question of the impact of the welfare state on the r class. There seems to be a growing perception amongst the !i working class that the welfare state may differentially benefit the onal and managerial groups and that the working class may pay in disproportionate contribution towards the costs of the welfare I his argument is commonly made about the welfare states of Eastern and elsewhere. An understandable reaction is to call tor wage ;s and tax cuts to put money into workers' pockets with which they ' the welfare services and goods that they need without having the ;tate to them how their needs should be fulfilled. Such an argument ie to the fore in the day care debate for example. Another related the question of the remoteness and authoritarianism of the welfare I his has come to prominence around questions such as the rights of ssistance claimants, drug addicts, parents of children taken into 'Ould, 1988) where the Swedish welfare state has tended to adopt uthoritarian practices. There are increasing pressures for more s welfare rights, better advocacy for clients and more consumer ation. This again is related to a third aspect - the emergence of the :ial movements over the past twenty years. To some extent the li of the Social Democratic hegemony may have held back some of novements, for example the autonomous women's movement as •á above. Nevertheless the anti-nuclear-power movement and the novement have had an increasing impact on the political scene in 'ears. In 1988 the Green Party exceeded the threshold of 4 per cent ote for the first time and thereby achieved parliamentary represen-Their philosophy of decentralization and consumer power may have ict on the Social Democrats. Finally it is apparent that Swedish tions are investing more and more overseas, as a reaction to the di levels of corporate taxation at home and also to be established inside the single European market of the EC. iocial Democrat government of 1982-91 attempted to respond to assures, while at the same time trying to counter inflation and to the welfare state, which is unquestionably still very popular in its form. It is therefore extremely premature to talk of the crumbling 66 Divisions of Welfare of the Social Democratic hegemony, though clearly a significant adapta has already taken place in that the growth of the welfare state has t ended in the 1980s. The new centre-right government formed in 1991, it predecessor in the 1970s, will probably not lead to radical change. In long term, however, the likelihood is that the hegemony will adapt lo capitalist and international realities, moving towards the new realism o1 social market economy model in the absence of a feasible socialist m c 3 I i k1 Federal Republic of Germany: the Welfare State in the Social Market Economy .eral Republic of Germany (FRG) came into being in 1949 with the i of the Basic Law, a strong written constitution. The occupying allies, particularly the United States government, exerted great o over the shape of this constitution and of post-war West German The constitution puts firm emphasis on the rule of administrative ■ regulation for the governance of social policy. Tt also devolves sable power within the welfare state to the regional states (Länder) . 1 government. There is a continuous debate about the extent to ■e Basic Law could meet socialist or other radical aspirations, but in ■iation so far it has certainly furthered a 'private capitalist philo- ith a certain social leaning1 (Sontheimer, 1972: 34), Politically the nt force in the FRG has been the Christian Democratic Party with their more conservative Bavarian partners the Christian Social f ,; ,, JCSU). They have led post-1949 federal governments, except for ■'/if iod of Social Democratic Party (SPD) and Free Democrat Party ^ ^ coalition between 1969 and 1982. All governing parties have ed the idea of the Social Market Economy (see below) as an ica! shell. political economy of the FRG may be divided up into three ogical periods. Up to the mid 1960s rapid economic growth (the lie miracle) took place in a context of considerable austerity and tion for the mass of the population until the late 1950s. There was ly a marked absence of political conflict over social welfare issues, high levels of unemployment in the 1950s and the pressing needs of i of incoming refugees from the East. Capital enjoyed 'the presence atively technologically advanced production apparatus . . . [and] extraordinary weakness of the working class and its organisations eir devastation by fascism' (Hirsch, 1980: 116). Conservative CDU olicy reform in the 1950s won sufficient support from the demora-3rking class to preserve their hegemony. The period from the mid ) the mid 1970s saw a resurgence of industrial conflict and struggles ; welfare state, leading to the partial eclipse of a more conservative nation of the Social Market Economy. The SPD/FDP coalition 'd tripartite corporatism in response to growing pressures from the