J Donald Moon. Handbook of Political Theory. Editor: Gerald F Gaus & Chandran Kukathas. 2004. Sage Publication.
The term ‘the welfare state’ came into common usage in the middle of the twentieth century. Its use reflected the growth in Western democracies of governmental responsibility for, and programmes addressing, an extensive range of human needs, such as education, health care, housing, child care, and economic security for the elderly, the unemployed, and the disabled. Some of the programmes of the welfare state, such as public schools and old age pensions, were first developed in the nineteenth century, but what might be called the ‘institutional’ welfare state did not fully emerge until after World War II, when most democratic countries adopted a more or less integrated range of programmes of welfare provision and policies of economic management. The institutional welfare state is characterized by a range of programmes designed to meet different needs and to provide security against various contingencies. Depending upon what ‘category’ one falls into, one would be eligible for different types of benefits. Thus, elderly people would be eligible for pensions, sick people for sickness benefits and health care, unemployed people for unemployment compensation, young people and those without marketable skills for education or job training, etc. At least as an ideal, as Brian Barry (1990) points out, the institutional welfare state would not even require a general safety net, since specialized programmes would cover all of the different conditions that prevent people from meeting their needs. In reality, of course, there will always be some who fall between the cracks, and so the welfare state must have a programme of ‘social assistance’ to cover residual cases. The emergence of the institutional welfare state is reflected in the enormous growth of government expenditures to finance its programmes, both in absolute terms and in relation to national income. In the UK, for example, social expenditure increased from less than 6 percent of GNP in 1920 to 25 percent in 1996-7 (Barr, 1998: 171).
There is no standard definition of the welfare state, and there are major national variations in the forms it has taken, reflecting the different sequences through which welfare states emerged, the social forces that advanced or resisted their creation, and the various political cultures and institutional frameworks. Students of the welfare state have offered a variety of classifications of welfare regimes, and disagree among themselves even about whether particular countries (notably, the US) even qualify as welfare states. Some students of welfare politics emphasize the difference between selective and universal welfare states (e.g. Rothstein, 1998); others discern liberal, corporatist, and social democratic regimes (e.g. Esping-Andersen, 1990); while yet others distinguish among social democratic, Christian democratic, liberal, and wage-earner welfare states (Huber and Stephens, 2001). More philosophically oriented theorists place the welfare state in the context of different traditions of political thought, and different ideals and/or patterns of justification. Thus, some discuss the minimal state and the arguments for and against it (e.g. Nozick, 1974; Schmidtz and Goodin, 1998); others consider the ‘residual’ versus the ‘institutional’ welfare state (e.g. Barry, 1999); yet others find four distinct strands, laissez-faire, feminism, socialism, and Fabianism (Clarke, Cochrane and Smart, 1987). While most recognize that class is a major concern of the welfare state, an increasing number of theorists see that gender is at least as important (Gordon, 1990; Fraser, 1997).
As a political formation the welfare state tends to divide theorists who in other respects share a view of politics. Thus, defenders and critics of the welfare state include people who identify themselves as (inter alia) conservatives, liberals, communitarians, socialists, and postmodernists, and so both its critics and its defenders find themselves with strange allies and opponents. There is certainly no single, unified theory of the welfare state setting out its fundamental principles and institutions. In the untidy field of political theory, theorizing of the welfare state is particularly untidy, and any general survey—including this one—must be highly selective not only in the issues it covers, but also in the way in which it constructs the topic itself.
One natural way of conceptualizing the welfare state is to view it in instrumental terms, as the organization of society to promote ‘welfare’ or well-being. The roots of this view can be found in Jeremy Bentham’s political thinking, particularly his insistence on rationalizing law and political institutions to make them maximally effective in realizing the interests of the community, which is to say, in his words, ‘the sum of the interests of the members who compose it’ (1948: 126). Different accounts of the welfare state on this view would project different accounts of what constitutes ‘welfare’ or ‘well-being,’ how the welfare of different individuals should be aggregated, and what policies and institutions are most effective in promoting welfare so understood. Although this instrumental approach captures many important issues, particularly justifications of the welfare state in terms of ‘efficiency,’ it sits uncomfortably with other important theories of the welfare state, particularly those that see it as ‘expressing’ or embodying the requirements of social solidarity and democratic citizenship.
In this chapter I will proceed nominalistically so to speak. In spite of the great variability mentioned above, welfare states share important features; four of the most important are a democratic political system, a largely private market economy, a wide range of public programmes that provide monetary support or services as a matter of right, and an active role for the state in managing the economy to dampen the business cycle and to regulate economic activities. It is important to emphasize the third feature, since it distinguishes the welfare state from an earlier tradition of relief for the poor. In the welfare state, receiving benefits does not undermine one’s citizenship; social provision is not an act of charity or a mere exercise of the state’s police powers.
Theorizing about the welfare state has tended to develop in response to its emergence, and the political conflicts and unanticipated consequences it had, rather than being the object of a particular political programme or philosophy. In an earlier time the welfare state may have provided a clearer target for study and criticism. Some may have bemoaned it for its role in shoring up capitalism. Others may have envisioned it as a possible road to a socialism, gradually contributing to the decommodification of labour and the collectivization of consumption and, eventually, investment. In this age of muted expectations, few see the welfare state as having such power, and attention has shifted to the different forms it has taken, and to the ongoing dilemmas with which it struggles. The hopeful visions of a T. H. Marshall or a R. H. Tawney, who imagined that the institutionalization of the social rights of citizenship would pave the way for a genuinely inclusive, democratic society, have given way to agonizing struggles over what appear to be permanent dilemmas of social policy. The very institutions that make the relief of destitution possible, at the same time create new forms of marginalization; the promise of freedom is accompanied by the reality of new and not so new forms of discipline, measures to increase efficiency and responsiveness at the same time raise troubling questions of equity and access. In the next section, I will set out a line of argument, appealing to the value of ‘efficiency,’ justifying one of the welfare state’s principal features—the collective, mandatory organization and provision of certain services. I will then address the ‘redistributive’ function of the welfare state, which is justified by appeal to values such as rights, solidarity, and social justice. In the last section I will discuss some of the ways in which both left-wing and right-wing critiques of the welfare state have converged in recent years.
Efficiency-Based Accounts
The task of meeting our needs in areas such as education, health, and old age security is obviously not uniquely assigned to the state. Even in countries with extensive social programmes, most of the ‘labour’ involved in providing welfare services, such as care for the young, the elderly, and the ill, is provided informally in households, through kinship networks, and by volunteers in religious and other organizations in civil society. In addition, many welfare services are provided through market transactions, such as the purchase of life or medical insurance. Why, then, should the state be involved in providing welfare, either directly in the form of specific services (such as health care or education) or in the form of resources or income to enable people to meet their own needs? Government programmes, after all, both involve an element of coercion and impose uniformity. Social insurance for unemployment or pensions, for example, is based on mandatory ‘contributions’—i.e. taxes and sets uniform benefit schedules, retirement ages, and other requirements. Allowing people to meet their own needs permits people to shape their lives according to their own priorities.
The alternative to state provision is often taken to be the market, where profit-seeking firms provide consumers with goods and services. But this is an oversimplification, as families and voluntary associations also play key roles. Prior to the rise of the welfare state, at least in Britain and America, people formed voluntary organizations to cover contingencies such as illness, disability, death, and old age. Individuals and families, wishing to insure themselves against various contingencies, could often choose among a variety of groups, or in some cases commercial firms, offering protection on different terms, and could make arrangements that were more or less tailored to their own circumstances and aspirations. The rise of the welfare state with its compulsory programmes has led to the demise of many of these voluntary associations and private firms, reducing citizens’ autonomy and imposing uniformity on them. The more extensive the welfare state, the more it has displaced other welfare institutions.
One reason for substituting state for private provision is that state provision (either of services or of resources) can sometimes be more effective than private provision, either because it can provide services or resources more cheaply, or because private provision is incapable of providing an optimal (or even adequate) level of services. In such cases, public provision may be justified on the grounds that it corrects some form of what is called ‘market failure.’ A standard example of market failure is public goods, such as national defence. In such cases, providing the good for one member of the group is impossible without providing it for all. It is difficult for a group to provide public goods for itself voluntarily, because each member has an incentive to free ride on the efforts of others, with the result that the good in question is not provided at all (or is provided only at a less than optimal level).
But are typical welfare goods public goods in that sense? To some degree, perhaps. If we all wanted to live in a society where no one suffered from destitution, and were willing to pay something to see that achieved, then we would all be better off if the government provided a safety net. Reasoning along these lines, Milton Friedman (1962: 191) has argued for a minimal welfare state, in which a ‘negative income tax’ would be employed to provide a subsistence income to people without other means of support. The minimal welfare state would not, however, be an ‘institutional welfare state,’ since its main concern would be to ensure that everyone had enough income to avoid destitution. Presumably, it would also provide other public goods such as public health and sanitation, for each of us is better off if others are inoculated against infectious diseases, or if the town disposes of every household’s sewage and garbage in a sanitary manner. But many welfare programmes do not seem to provide public goods: the principal beneficiary of an old age pension is the pensioner, the principal beneficiary of a high school or college education is the student whose skills are improved and whose life is enriched, the principal beneficiary of open heart surgery is the patient whose life is saved, and so forth.
Even when the institutional welfare state does not provide public goods, strictly speaking, there are other limitations of the voluntary model it can overcome. For example, private firms and voluntary organizations are poorly equipped to protect individuals from income loss due to unemployment. Non-governmental risk-pooling schemes work best when the chances that one person will suffer a given condition—say disability or death—are more or less independent of anyone else’s chances, and when the overall risks facing the group are known. Under these conditions, each individual can pay into the fund, which can accumulate enough to provide benefits to the unfortunate. But if the risks in question are not independent, if one person’s suffering increases the likelihood that others will suffer as well, then a private scheme may collapse, as more and more people shift from being contributors to being claimants, and the group’s reserves are depleted. Unemployment is (in part) cyclical, which means that in a downturn some people lose their jobs, and as a result reduce their consumption, thereby leading other firms to lay off workers, in an expanding cycle. Thus, a private firm or voluntary association offering unemployment insurance would run the risk of going out of business as fewer and fewer people held jobs (and so paid into the fund) and more and more people lost their jobs, and so became claimants. Because state-sponsored schemes, unlike private associations, are able to run deficits, and to the extent that these deficits actually contribute to expanding demand and so reducing unemployment and stabilizing the economy, they can deal with problems that non-state schemes cannot.
Voluntary welfare provision may also be unable to cover everyone in a society. Many people in the heyday of mutual aid societies were not members, and non-members were often among the least advantaged, those without steady jobs and a secure place within the community. And it is easy to understand why. Organizations offering protection recognize that those most likely to need protection have the greatest incentive to seek it, and so to join a mutual aid society or to purchase insurance, while those facing the lowest risks have an incentive to stay out. As a result of this process of ‘adverse selection,’ risks tend to be spread over a smaller and smaller part of the population, and premiums must rise accordingly. This process of adverse selection can continue to the point where most of those in need of protection are unable to afford it, because premiums have to rise so high that all but the most vulnerable drop out. The welfare state can combat the problem of adverse selection by making membership compulsory: ‘because low risks cannot opt out, it makes possible a pooling solution’ (Barr, 1992: 755).
Adverse selection is reinforced by a second process or condition, called ‘moral hazard.’ People who are insured against a certain risk may be more willing to take chances than they would be in the absence of insurance. Knowing that if I get sick or injured, my medical bills will be covered, may make me more willing to engage in risky behaviour, such as downhill skiing. To the extent that this occurs, organizations may face higher claims, thereby forcing them to raise their charges, and discouraging others from purchasing protection. More obviously, unemployment insurance schemes are subject to moral hazard, for knowing that I will be covered in the event that I am unemployed, I have an incentive to quit (or arrange to be fired) and/or not to seek or accept employment. Of course, state schemes are subject to moral hazard as well, but the key point is that if the genuine risk of losing one’s job is to be covered at all, it must be covered through a public programme (see Barr, 1998: 190-2).
A related problem is a tendency for people to overuse services when they are free at the point of delivery. If my purchasing an insurance policy or joining a mutual aid society gives me the right to a free service, I may be tempted to take advantage of that opportunity to a greater extent than I would be if I had to pay for it each time I used it. In that case, costs would tend to escalate and the group as a whole may end up paying more for the protection than would be optimal from their own point of view.
For all of these reasons organizations offering protection will try to limit use, to prevent too many high risk people from joining, and to charge them more in order to hang on to their other members. In the case of voluntary groups, such as neighbourhood-, work or craft-based mutual aid societies, informal patterns of social surveillance and affinity may function to exclude outsiders and others who are thought to be especially likely to need benefits. Similarly, private firms may use various underwriting mechanisms to screen out high risk individuals or groups. The overall result may well be that certain groups may receive no or inadequate coverage, and the cost of services may be much greater than they would be if they were provided through a compulsory plan that spread risks more widely and rationed services to avoid overuse.
It is important to stress that state provision is not necessarily superior to private provision. Even if there are clear examples of ‘market failures,’ areas in which voluntary provision is incapable of providing an optimal level of services of one sort or another, it does not follow that government action will be superior. Just as real-world markets are subject to market failure, so real-world governments are subject to non-market failure. For example, while mandatory programmes can avoid the problem of adverse selection, by requiring low risk individuals to participate in the risk-sharing scheme, they may exacerbate the problem of moral hazard, by giving individuals incentives not to provide for themselves (e.g. by reducing their savings rate, or not taking a job) and relying upon the public programme of pensions or unemployment compensation to meet their needs. And government provision is subject to its own limitations. For example, government programmes can be run for the benefit of the bureaucrats who administer them, at the expense of the clients they are supposed to serve, or may be captured by special interest groups, who succeed in diverting resources to their own ends. Moreover, markets are often able to provide public goods or deal with externalities in effective ways. What is required, then, is a balancing of the relative costs and benefits of different forms of provision for different kinds of contingencies, and in different settings.
The recognition that public provision can involve greater costs than voluntary programmes has led to calls for ‘privatization’ of some welfare state activities during the past 20 or 25 years. Different groups have advocated devolving to private parties those activities once performed by the state, ranging from the sale of nationalized industries to contracting with private firms to provide public services, such as running schools or supplying cleaning services to a government bureaucracy. In a similar vein, recent years have seen efforts to increase choice and simulate market processes within public programmes, such as the use of vouchers in public education, or the ‘internal market’ in Britain’s National Health Service. In all of these initiatives, the hope is to increase efficiency, to make service providers more responsive to clients, and to enable people to receive more individualized services, reflecting their specific needs and interests. On the other hand, these developments raise the concern that even ‘quasi-market’ choice in areas such as pensions or education will adversely affect disadvantaged groups. For example, when the successful school in a system relying upon vouchers or other ‘parental choice’ mechanisms is able to attract more students than it has space for, the fear is that it may respond by excluding ‘problem’ children, possibly leaving them even worse off than before. Whether the issue is pensions, education, health care, or other areas of the welfare state, efficiency arguments for public versus private provision involve a balancing of their relative costs.7
The Welfare State and Redistribution
A second line of argument supporting the welfare state appeals to the idea of justice rather than efficiency. The policies of the welfare state do not simply make it possible for individuals to realize their own interests more effectively, but generally redistribute income. Efficiency-based arguments normally take the outcome produced by market exchange, prior to governmental taxation and transfers, as their baseline, and show that a particular policy can at least in principle make everyone better off than they would be given that baseline. But to the extent that welfare policies deliberately redistribute income, those whose income goes down would normally (though not necessarily) be worse off; such policies could be justified, then, only by invoking values other than efficiency.
More important, the appeal to efficiency is itself problematic, in as much as the pretax/pretransfer baseline it takes for granted must be justified. There are some risks which we face, when we think of our lives taken as a whole, that cannot be covered by any form of private provision, because they reflect conditions into which we are born, such as congenital handicaps, genetic predispositions to certain diseases, and the cultural and economic disadvantages one’s parents may suffer. Because of these conditions, those who are fortunate have no incentive to join a risk-sharing scheme to compensate those who are not. Any private system of provision is limited to pooling the shared risks that people face in the future, and so presupposes a ‘baseline’ of a given distribution of advantages and disadvantages. But from a larger point of view, this restriction to a given status quo is arbitrary.
Even using the term ‘redistribution’ may be misleading to the extent that it seems to presuppose that the initial ‘distribution’ is somehow morally privileged, so that deviations from it—re distributions must be ‘justified.’ But a moment’s reflection should be sufficient to see that this presupposition is false, and that any distribution of ‘the advantages of social co-operation’ must be justified, whether it results from market transactions or from welfare state policies specifically designed to redistribute income.
The presumption that distributions that result from ‘government’ action must be justified, and that pretax and pretransfer distributions are presumptively just, appears to be widespread at least in America, leading to hostility on the part of some towards the welfare state. Strong libertarians like Nozick hold that taxation to redistribute resources from some taxpayers to others is not only presumptively but actually unjust because it violates citizens’s property rights. This critique obviously presupposes that the right we have to our property, including income from employment or business activity, is not created by the state, but exists in some sense ‘prior’ to political life, and so limits what governments may legitimately do. If such a theory of natural or prepolitical rights could be vindicated, it would block redistributive welfare state programmes.
Welfare Rights
It goes beyond the scope of this chapter to examine the case for natural or prepolitical property rights, but it is worth pointing out that many of the considerations that can be invoked to support strong property rights also support welfare or ‘positive’ rights, and so can be used to justify the redistributive activities of a welfare state. When we think about why we are attracted to the idea that humans have rights at all, including a (defeasible) right not to be coerced by others, the reasons we are likely to come up with will support the idea that people ought to be accorded certain basic welfare rights, rights to goods and services necessary for human functioning. For example, Nozick refers to the idea that people are capable of leading meaningful lives, and so they have (or should have) a right against being coerced by others because such a right is necessary to protect that fundamental human capacity. I can only create projects for myself, and organize my life to realize those projects, and thus find meaning in my life, if I am free from coercion by others: they can’st force me to do their bidding rather than fulfil my own aspirations.
This is a powerful argument, but it is equally true that to live my own life requires not only protection against interference from others, but also access to the resources necessary to life itself. If those resources can be appropriated as private property, then a person could be deprived of anything resembling a decent life, or even life itself, because she lacked the necessary resources. Jeremy Waldron (1993: 309-38) gives the example of a homeless person, in a setting in which all land and other amenities, such as toilets or sleeping places, are privately owned. Under those circumstances, she would not be able to live, or at least to live without violating someone’s ‘rights.’ But what reason would she have to acknowledge a duty not to take what she needed, when her life depended on it? It is hard to see why people, recognizing the possibility that they might become impoverished, would have reason to accept a system of property rights that could leave them in such desperate straits. As Waldron (1993: ch. 1 and passim)argues, the only system of property rights that all have a reason to endorse would be one that ensured that no one need be deprived of essential resources, and the obvious way of achieving that would be to make property holdings subject to taxation, so that the state could provide essential goods and services, or at least a minimum income, when necessary.
This line of argument supports what might be called a social minimum state, not necessarily an institutional welfare state. The core argument is that some fundamental human values—the idea of a meaningful life, personal autonomy, or life itself can be realized (or at least guaranteed) only if there are government programmes providing enough income at least for subsistence. F. A. Hayek, for example, is renowned as a critic of the welfare state, but he accepts the idea of a social minimum, arguing that citizens may feel that there is ‘a clear moral duty of all to assist, within the organized community, those who cannot help themselves,’ and so the society could provide ‘a uniform minimum income… outside the market’ to those who are indigent (1976: 87).
Others have argued that people have prepolitical welfare rights, on all fours with the ‘negative’ rights to non-interference such as the right to bodily integrity, and that it is the government’s responsibility to secure those rights. A just society, then, could only be a society in which those rights are secured, and so only a welfare state could be a just state. To the extent that welfare rights are rights to specific resources, such as health care, education, and housing, fulfilling them may require or at least justify an institutional welfare state, not just a redistributive tax and transfer system. More plausibly, arguments from positive or welfare rights might be combined with the efficiency-based arguments surveyed above to justify an institutional welfare state.
The view that we have welfare rights that are, in some sense, prepolitical, which require that the state provide various goods and services, is subject to well known difficulties. The standards defining the scope of such rights claims are notoriously vague. Raymond Plant et al. (1980), for example, base positive rights claims in ‘needs,’ but what are the boundaries of need? I may ‘need’ an enormously expensive kind of medical treatment in order to prolong my life, if only for a few days, but is it plausible to say that I have a right to such treatment? Ronald Dworkin argues the traditional practice of medicine may be based on the ‘rescue principle,’ which answers that question affirmatively: ‘it says we should spend all we can [on health care] until the next dollar would buy no gain in health or life expectancy at all,’ but he insists that ‘No sane society would try to meet that standard’ (2000: 309): it would require sacrificing too many competing goods, including other rights claims, like the right to an education or a minimal standard of living.
Alan Gewirth views positive rights claims as implicit in the commitment to human agency, a commitment one necessarily undertakes in performing any intentional action, because doing so presupposes that one views oneself as an agent, and so is implicitly committed to those conditions necessary for the exercise of agency, which include access to certain resources. But who is responsible for ensuring that I have access to the resources necessary to exercise agency in my own case? Gewirth holds that when I cannot meet my needs through my own efforts, others have an obligation ‘positively to assist’ me (1978: 134). But what standards are they to use to determine what constitutes a reasonable effort on my part?
These concerns may not be decisive to reject the idea of basic welfare rights, but they do mean that specifying them is impossible in the absence of some political process through which the standards governing responsibility and trade-offs among conflicting uses can be determined (see Holmes and Sunstein, 1999). And because these rights cannot be specified except through a political process, it is implausible to view them as establishing a prepolitical standard of justice to which that political process must conform.
Equality of Opportunity
A second justice-based argument for the welfare state appeals to the idea of ‘fair equality of opportunity,’ to use Rawls’s phrase. Fair equality of opportunity requires not only that there be no ‘arbitrary’ barriers to the life choices one may make, such as restrictions on occupational or educational opportunity based on race or gender, but that everyone has access to the resources and experiences necessary to qualify for the different positions and careers that exist in society. To the extent that one’s chances in life are determined by the class position into which one is born, then fair equality of opportunity is denied. Arguably, fair equality of opportunity supports not only a social minimum state, but an institutional welfare state, in which education, including perhaps early childhood education, and medical care are provided on a common basis for all. But, like welfare rights generally, the requirements of fair equality of opportunity cannot be specified except in specific social contexts; the kind of educational opportunities necessary in a largely agrarian society, to take an obvious example, are very different from those required in a postindustrial setting. And once again it is necessary to make tradeoffs between equal opportunity and other values, such as the privacy and autonomy of families.
Membership and Solidarity
Because concepts of positive rights and equal opportunity are not well defined outside of specific social contexts, they are often combined with arguments appealing to ideals of citizenship and social solidarity. The basic argument is that the welfare state should guarantee the inclusion of all citizens as full members of a democratic society, which requires that an extensive range of social rights be provided. The reasoning is fairly straightforward: just as citizens must have civil and political rights, they must be guaranteed certain social rights if they are to be full members of a society, and specifically if they are to participate in democratic politics. The key premise in this argument is that citizenship must be universal. All who are capable of intentional or responsible action must be full citizens. The only legitimate basis for exclusion is incapacity for responsible action.
T. H. Marshall (1977) offers a classical account of the welfare state as the necessary result of the universal extension of citizenship. He traces the emergence of universal citizenship by observing three successive phases, the first involving the general extension of civil rights, the second the universalization of the suffrage, and the third the growth of the welfare state and the creation of the ‘social rights of citizenship.’ Social citizenship is essential for democratic equality; people who are destitute, or who lack access to essential resources such as medical care, or who do not have at least a basic education, cannot stand with others on an equal plane. If citizenship is to be universal, the state must guarantee that everyone has access to these essential goods. The democratic state must thus be a welfare state.
There are a number of variants of this argument, but a common theme is a deep suspicion of the market and at least certain forms of individualism. Whereas arguments from efficiency take the market as a baseline, and justify social policies on the ground that they can correct market failures, arguments from solidarity begin with something close to the opposite assumption—projecting an ideal in which all activities are organized through collective associations, in which individuals are oriented principally towards common needs and aspirations. Richard Titmuss (1972) extols the ‘gift relationship,’ and David Harris (1987) speaks of the family as a model for social life. More concretely, Claus Offe (1984) and Gosta Esping-Andersen (1985) once expressed the hope that the growth of collective consumption and other forms of decommodification will eventually displace capitalism, leading to a socialist order of society.
Harris offers a communitarian version of the argument from solidarity. He argues that ‘full membership’ in a society requires that each person be able to enjoy ‘a certain style of life’ and ‘certain life chances’ (1987: 147). Although he recognizes that modern societies include a plurality of different groups, he insists that there are more or less common standards of what an individual must be able to do and how one must be able to live if one is not to be excluded or socially marginalized. These standards determine the needs of members of that society, and should be equally available to all citizens as a matter of right, for only in that way can the equal status of members be recognized and respected (1987: 154-7). This line of argument supports the institutional welfare state in which services are provided in kind in part because ‘citizens have a right to that specific resource,’ such as ‘education,’ rather than a right ‘to income which may or may not be spent on education’ (1987: 150). Further, the universal provision of certain services is expressive of, and may contribute to, a sense of community and equal citizenship. Finally, providing services in kind may be a form of ‘justified paternalism’ to the extent that ‘some persons may be imprudent or wasteful or be unable to make adequate use of cash’ (1987: 150-1).
Harris’s account relies upon an analogy between political society and the family: just as we have obligations towards, and rights against, members of our family, irrespective of what they may have done for us individually, so we have obligations towards, and rights against, our fellow citizens. The stress on obligations is crucial, for the possibility of enjoying one’s rights depends upon the willing support of social policies on the part of the citizenry, and to claim one’s rights one must be prepared to fulfil the ‘system of duties’ that ‘underlies the structure of citizen rights’ (1987: 160). Thus, a person ‘who is genuinely and personally responsible for his condition’ has no rights-based claim to assistance, for he has not fulfilled his ‘duty to maintain himself as an independent member of society’ (1987: 160). However, Harris goes on to argue that the pragmatic difficulties involved in determining whether someone’s unfulfilled needs are a result of his own choices are so great that we should presume that there are no such cases, and should rely upon a ‘sense of duty or community… to prevent or minimize abuse of the system’ (1987: 161). Of course, this same sense of community is necessary if those relatively advantaged citizens, who are net contributors to welfare schemes, are to be willing participants in the process.
These arguments are subject to a number of obvious reservations. In the first place, the founding of rights and obligations on ‘membership’ is deeply problematic, in as much as it begs the question of whether the social order of which we are to be members is just. To say that we are all equally members, and so should be equally entitled to the resources necessary to be full members, is not to say that as members we are equal, as Harris seems to suppose. In a caste or feudal society, everyone might equally be a member, and membership may well carry with it certain welfare (and other) rights, but these rights are not equal but differentiated by status, and the distribution may be deeply unjust. The general argument from membership does little to support a modern welfare state in the absence of a larger theory of justice. I will return to this point below.
Second, and ironically, welfare states have a systematic tendency to undermine the very communitarian sentiments and relationships that would support the values of solidarity and equality. Although participating in a common programme, such as a national health service or medicare, may give rise to feelings of solidarity with others, what people actually experience may often be quite different. In many cases it is more like being reduced to the status of a client, attempting to meet one’s needs through an impersonal and unresponsive bureaucracy. Far from contributing to a sense of community, public provision (which is almost inevitably bureaucratic provision) may disrupt the communitarian forms through which needs may have been met in the past, replacing personal relationships which engender obligations and mutual identification with legally prescribed associations of strangers.
Further, the commitment to equality can sometimes sit uneasily with the commitment to democracy. Consider, for example, Albert Weale’s argument for earnings related welfare state schemes, such as social security in the US. Weale argues that such schemes increase the total volume of government transfers, thus leading to greater ‘egalitarian effectiveness.’ Weale explains this egalitarian effectiveness in part as follows:
Of course, there is no necessary incentive to redistribute savings in the public earnings related system, but equally there is little practical opportunity to resist any modest redistribution that managers of the public scheme determine. Denied the ‘exit’ option of shopping around, the typical citizen is confronted merely with the costly ‘voice’ option of changing the terms of the public scheme. Since people are often highly ignorant of the details of pension schemes, participation to change their terms is extremely costly. (1990: 481)
In short, because democratic control is difficult, popular opposition to redistribution will be ineffective, allowing elites to achieve greater ‘egalitarian effectiveness’ than citizens would be willing to support directly.
Solidaristic Conceptions of Justice
An adequate account of the welfare state, one that can justify its redistributive aims, must ultimately be based upon a theory of justice, and the most promising theories are those which Phillipe Van Parijs calls ‘solidaristic conceptions of justice’ (1995: 28), such as those offered by Rawls, Dworkin, Amartya Sen, and Van Parijs himself. Solidaristic conceptions of justice are based upon a commitment to ‘equal concern’ for the interests of all, and to ‘equal respect, that is, the view that what counts as a just society should not be determined on the basis of some particular substantive conception of the good life’ (1995: 28). The ‘liberal’ commitment to equal respect in solidaristic theories of justice underlies their support for the standard ‘negative’ and democratic rights characteristic of the welfare state, and the commitment to equal concern underlies their accounts of social justice and so the redistributive elements of the welfare state.
Different solidaristic theories provide different accounts of social justice, and support different institutions. No theory, by itself, directly supports the institutional welfare state. Van Parijs, for example, rejects it in favour of a system providing the highest possible basic income for all, and Rawls explicitly rejects the welfare state on the grounds that it tolerates the highly unequal distribution of wealth produced by a capitalist society, and so undermines democracy by concentrating too much economic and political power in a wealthy elite. Still, solidiaristic theories can supply the deficiencies, noted above, in justifications of the welfare state that appeal to membership and solidarity, and to the baseline problem in efficiency-based arguments. With regard to the appeal to membership, solidaristic theories of justice provide grounds for the value of social inclusion on a principle of equality. And they address the serious lacunae in efficiency-based arguments, specifically the fact that they take a market generated outcome as their starting point, and ask whether that outcome could be improved through some government policy. But because there is nothing privileged about market generated outcomes, market institutions and the ‘initial’ distribution of resources must themselves be morally justified, and solidaristic theories of justice address that problem.
While solidaristic arguments do not necessarily justify the welfare state as the ideal regime, they do provide grounds for central welfare state policies. Rawls’s ideal regimes, a property-owning democracy or market socialism, would have to be welfare states in the sense I have used the term here: that is, they would have to have social policies that would collectively provide for certain needs, justified in terms of efficiency and their redistributive consequences. And Van Parijs allows significant scope for collective provision including the area of medical care.
Paradoxes of the Welfare State
I have asserted that the justification of the welfare state rests upon arguments invoking both efficiency and justice. Arguments for redistribution or equality do not necessarily support the collective provision of services or categorical forms of income support characteristic of the welfare state, but must be supplemented by other considerations. These can include the ‘expressive’ value of common provision stressed in some social democratic and communitarian accounts, with their direct appeals to solidarity and membership, or to analyses that show that collective provision of services (e.g. medical care) or social insurance is superior to voluntary and market provision in terms of enabling individuals to meet their needs.
A common theme in justifications for the welfare state is the need for social provision if freedom is to be effectively realized for all citizens. Ironically, this very commitment to freedom and solidarity is subject to a certain inner tension. One way to think about this is to see that the effort to empower individuals and to promote social inclusion also leads to new forms of marginalization and control. One major reason for this is that the goal of social inclusion cannot be achieved merely by providing people with resources, for what is ultimately required is that people be able to participate effectively in the social and political ‘life of the community’ (Sen, 1992: 5). Thus, we must be concerned not simply with citizens’s command of external resources, but with what they can do with them, as Amartya Sen argues. He distinguishes between ‘functionings,’ which ‘represent parts of the state of a person—in particular, the various things that he or she manages to do or be in leading a life,’ and a person’s ‘capability,’ which ‘reflects the alternative combinations of functionings the person can achieve’ (Sen, 1993: 31). Functionings can be very complex performances or states of being, or ensembles of performances and states of being, such as ‘being in good health’ or ‘achieving self respect or being socially integrated’ (1993: 31).
In spite of the importance of functioning as opposed to possessing, most evaluations of welfare state performance focus on what people have, rather than on what they can do. One common measure, for example, is ‘percentage of poor households lifted out of poverty as a result of taxes and transfers,’ where poverty is defined as having an income below 50 percent of adjusted median household income of the country in which one lives (Rothstein, 1998: 183 4). But if the objective of the welfare state is to enable citizens to participate effectively, this measure is problematic because income, or income alone, does not provide the capability to achieve many of the most important functionings. In a recent study aptly titled What Money Can’st Buy (1997), Susan Mayer has examined the ‘functionings’ of children, adolescents, and young adults, and correlated them with family income. Her findings, consistent with Sen’s general argument about the relationship between resources and functionings, are that, above a basic level, in most cases ‘additional parental income does not improve children’s chances for success’ (1997: 2). Mayer hypothesizes that the reason that income has such a limited effect is that other parental characteristics, such as ‘skills, diligence, honesty, good health, and reliability, also improve children’s life chances, independent of their effect on parents’ income. Children of parents with these attributes do well even when their parents do not have much income’ (1997: 3). The more general point here is that functionings that are important for full membership or citizenship depend upon internalized dispositions and skills and not merely on access to external resources. Thus, it might be concluded, ensuring equal citizenship requires programmes that go beyond the provision of external resources.
Welfare and Work
The argument about the necessity for effective functioning, as opposed simply to having access to resources, has been most heated in the area of work. If democratic citizenship requires that all be enabled to participate fully in society, then people must have not only certain resources, but also certain capacities, skills, and dispositions. Recently, however, the issue of duties of citizens has become more urgent, as all welfare states have seen the emergence of a class of citizens who are dependent on the state’s welfare programmes for their survival, and who do not provide for themselves and their families through their own labour, or through benefit programmes such as social security in which benefits reflect their previous efforts and earnings. In much popular discourse, the welfare-dependent population is stigmatized as exploitative, irresponsibly taking advantage of the social safety net to avoid work and provide themselves with more leisure than they could otherwise afford, giving rise to demands for punitive measures to limit access to such benefits. But even people who are sympathetic to the plight of those who more or less permanently rely upon public assistance have cause for concern. One can acknowledge that people rely upon ‘welfare’ because their options are so limited, and so their condition represents an indictment of the society rather than the individuals concerned, but the fact remains that receipt of social assistance does not enable one to attain full citizenship or membership in society. It simply sustains one in a marginalized condition. Social inclusion requires more than receiving benefits.
This line of argument has been advanced by a number of ‘conservative’ critics of the welfare state. Lawrence Mead (1992), for example, argues that the character of poverty at least in America has changed in the past several decades, and that the social exclusion represented by poverty reflects the inability of poor people to act as rational agents in pursuit even of their own interests. The key to overcoming this exclusion is to inculcate in the passive poor the capacities for agency, for acting to promote their own interests and to control their own lives, by imposing adequate disciplinary controls on them. If poverty creates social exclusion, and so is a barrier to citizenship, then the state must ensure that its citizens develop the capacities that enable them to escape poverty. The key policy, in Mead’s view, is workfare; the poor must be required to work as a condition of support, for unless they develop the discipline and sense of accomplishment that work involves, they will be unable to escape the conditions of dependency. Social policy must take on an explicitly ‘paternalistic’ character, and the state self-consciously assume a tutelary role. Mead holds out the possibility that ‘public paternalism might help regenerate… informal [social] controls, by involving community organizations in directive programs’ (1997: 27-8). In that case, ‘paternalism in its public sense might not have to be permanent,’ but only because the necessary disciplines are imposed through other social agencies.
Nikolas Rose has pointed out that the emphasis on paid employment is not a monopoly of the right: ‘From the “social democratic left”, too, work [is] now seen as the mode of inclusion, and absence from the labour market the most potent source of exclusion’ (1999, 163). In some solidaristic accounts, the emphasis on work invokes an older language of duties. In Harris’s account, for example, the duties correlative to our welfare rights are ‘strict obligations’ and may be enforced by ‘coercion’ (1987: 161). In this, he echoes Marshall, who looked beyond the social rights of citizenship to consider the duties of the enriched and inclusive model of citizenship he advocated, including ‘the duty to work,’ which he thought was of ‘paramount importance.’ Similarly, Amy Gutmann and Dennis Thompson offer a justification for enforcing work obligations that draws on the idea of citizenship, arguing that ‘work should be seen as a necessary part of citizenship’ (1996: 293), because it is ‘essential to social dignity.’ Since ‘earning is not only a means of making a living but also a mark of equal citizenship,’ paid employment has a ‘political dimension’ that ‘provides a further justification for the obligation to work’ (1996: 302).
But this obligation to work is not, or is not merely, a demand to be made on the individual, one which he might reasonably wish to resist, for ultimately it is rooted in an ideal of social inclusion and active citizenship through which the individual’s own interests and needs can be realized. Anthony Giddens sounds this theme in his call for ‘the positive welfare society,’ in which ‘the contract between individual and government shifts, since autonomy and the development of self—the medium of expanding individual responsibility become the prime focus’ (1998: 128). Replacing the traditional ‘welfare state’ with the ‘social investment state,’ the task of government would be to invest in ‘human capital’ rather than ‘the direct provision of economic maintenance’ (1998: 117). Although he allows that full employment might not be realized, he calls for the redistribution of work to include as many as possible, and various forms of payment for participation in the ‘social economy,’ the sphere of civil society traditionally maintained by voluntary work. As Rose puts it, the contemporary ‘organization of freedom’ views individuals as best able to fulfil their political obligations in relation to the wealth, health and happiness of the nation not when they are bound into relations of dependency and obligation, but when they seek to fulfil themselves as free individuals’s, which depends ‘upon the activation of the powers of the citizen’ (1999: 166).
There is a certain irony in this development. It is not just that the welfare state’s commitment to providing the resources necessary for everyone, including the most disadvantaged, to make their formal freedom effective comes with new forms of social control. That could hardly be avoided. Rather, the irony is to be found in the way in which this current in thinking about the welfare state appropriates and reverses the emphasis of those who first resisted it. Long before the development of the welfare state, von Humboldt argued that ‘The true end of Man… is the highest and most harmonious development of his powers to a complete and consistent whole’ (1969: 16), a condition which can be realized only if the authority and action of the state are limited to protecting individuals against violations of their rights, for any effort on the part of the state to advance the ‘positive welfare of the citizen is harmful’ (von Humboldt, 1969: 3). State action ‘invariably produces national uniformity’ (von Humboldt, 1969: 23), and leads to the ‘deterioration of the moral character’ of citizens because they are deprived of the opportunity and need to manage their own concerns, and the stimulus to intellectual and moral development that such management provides. Von Humboldt’s argument inspired important strands in John Stuart Mill’s thinking, and his ideas are reflected in the work of many nineteenth-century liberals, who often defended the elimination of outdoor relief and similar policies on the grounds that welfare undermines capacity for autonomy and other virtues of its intended beneficiaries. Today, in the focus on social inclusion and the recognition that this involves not just having rights and resources, but fulfilling duties and exercising capacities, the welfare state is being reshaped (or at least readvertised) as a device for shaping citizens and inculcating virtue, including the virtues necessary to ‘fulfil themselves as free individuals’ (Rose, 1999: 166). That is not, I hasten to add, a reason to reject the welfare state, but only a reason to recognize both the causes of its persistence in the face of continual ‘crises,’ and its limitations.