Foundations of the Workfare State – Reflections on the Political Transformation of the Welfare State in Britain

The British ‘welfare state’ has been transformed. ‘Welfare’ has been replaced by a new ‘workfare’ regime (the ‘Work Programme’) defined by tougher state regulatory practices for those receiving out-of-work benefits. US-style mandatory community work programmes are being revived and expanded. This article, therefore, considers shifting public attitudes to work and welfare in Britain and changing attitudes to working-age welfare and out-of-work benefits in particular. It also considers the extent to which recent transformations of the state may be explained by declines in traditional labourist politics and class-based solidarity. Thus, we attempt to develop a richer understanding of changing public attitudes towards welfare and the punitive regulatory ‘workfare’ practices engaged by the modern state in the liberal market economy; reflecting on the nature of the relations between ideology, party policies, popular attitudes and their political impact.


Introduction
The contradiction is that while capitalism cannot coexist with the welfare state, neither can it exist without the welfare state. (Offe : , original emphasis) This article considers how far patterns of change in British public attitudes to work and welfare, and the apparent shift from 'welfarist' to 'workfarist' values, may be explained by declines in traditional labourist politics and class-based solidarity and the repositioning of the British Labour Party regarding welfare in the s. Drawing on trend data from the British Social Attitudes (BSA) survey (e.g. Park et al. ), the first section examines the transformation of the British 'welfare state ' (q.v. Wincott ) and the relevance of social attitudes in the policy process, before describing the study was redefined in order to secure labour market participation under the heading of 'activation'. New training programmes and strategies were devised to promote employability (as long-term unemployment can deskill) and to increase the job-readiness of unemployed workers (Paz-Fuchs ). This approach aligned with the goal of workfare: not simply to reduce unemployment but also to tackle the wider cultural problem of 'worklessness'. The policy paradigm has therefore firmly shifted away from protecting incomes in favour of welfare conditionality and activation, as the workfare state in Britain attempts to enforce work for all.
Workfare policies introduced by the UK coalition government, including the new punitive 'Work Programme' (Cm  ), are designed to solve a problem about the effects of 'passive', social citizenship rights-and-eligibilitiesbased welfare. 2 This problem arose, it was argued, because too many citizens had been reduced to 'welfare dependency' (Wiggan ). Supporters of activation argue that the new conditions impressed on welfare benefit 'claimants' -roles and responsibilities, duties and obligations -are perfectly and morally justifiable as the social contribution now owed by citizens to society (Mead ). Labour exchanges have long helped to solve the problem of economy and subjectivity, as Rose () observes, but in the workfare state, the 'jobcentre' takes on new significance in 'activating' the labour power of workers.
Activation policy in the UK is now actively moving more people off welfare and into paid work; the effects are being felt across the different groups of benefit claimants, including people with a disability or health condition, single parents and two-parent families with children (table ). Under the new sanctions regime (Welfare Reform Act ), claimants are no longer exempt from looking for paid employment (unless they are deemed to be lead carers or are deemed unable to work after a medical assessment). 3 Harsher sanctions have been introduced to help to ensure that jobseekers comply with the new workfare regime. To maintain eligibility to benefit support, jobseekers must actively look for paid employment or face benefit sanctions; they must take up any offer of paid employment or face sanctions; and they must undertake all specified work preparation activities demanded or face sanctions (table ). If Table  Workfare conditionality in Britain • Group : Full conditionality: jobseekers; all work-related requirements apply. • Group : Work preparation: people with a disability or health condition who have a limited capacity for work. • Group : Worked-focused interview: lone parents or lead carer in a couple with a child aged one to four; keeping in touch with the labour market. • Group : No conditionality: people with a disability or health condition which prevents them from working; lone parents, or lead carers with a child aged under one.
Source: SSAC : .  conditions are not met, claimants can now find their benefits suspended for up to three years. New options are also available for mandating 'work-shy' claimants into unpaid work. Welfare recipients deemed to be lacking an adequate 'work ethic' (or those who display work-aversion tendencies) now face 'Mandatory Work Activity'. This is an important development in light of the assertions made by Jobcentre Plus (JCP) staff that the vast majority of people claiming out-of-work benefits are not interested in working for a living (Dunn ). Adapting to the demands of changing circumstances, the state has become much more active in regulating the behaviour of citizens, government is about ways of conducting conduct after all but the issues involved are not solely about 'legitimacy'. Conditionality has always been a part of the Beveridgean welfare state but the intensification of work-conditionality raises a number of interrelated normative and empirical questions for social policy that are discussed elsewhere (Deeming a; Whitworth and Griggs ), particularly concerning the ethics of work-conditionality for different groups of benefit claimants and the use of 'evidence' used to drive through ideologically motivated workfare reforms. Much of the empirical evidence emerging from the 'Work Programme' evaluations seems patchy at best, and auditors report poor returns on investment for the British taxpayer. We might well argue that the new techniques of statecraft being introduced to manage the economy and society are made possible only if they achieve a real level of public acceptability and 'legitimacy' in democratic politics and policy processes (Kumlin ). Public opinion is politically important in terms of justifying changes to the social security system (Griggs and Bennett ) and, while the British public may be concerned about rising inequality and the needs of the very rich, they seem to manifest less concern for people on low incomes (McKay ; Rowlingson et al. ). Nevertheless, 'social problems' such as 'poverty' and 'unemployment' and measures for addressing them are constructed and represented by policymakers for policy-making (Stevens ; Wiggan ). Thus, 'welfare dependents' and 'jobseekers' are a particular category of the population whose conduct is increasingly governed by the state's coercive powers through such moral 'problematizations' of individual character (Dean , ).
The power of myth both encourages and hinders our understanding of the social world, famously documented by Richard Hoggart () and more recently by Dean (). Increasingly, the media report stories about the state failing to 'look after us' or claiming that welfare has become a 'soft touch' for 'scroungers' and the 'lazy', 'benefit cheats' or 'welfare tourists' -all of which undermines trust and weakens collective solidarity for the principles of social protection. Similarly, politicians now capitalize on the promise to 'make work pay' for the 'hardworking majority' of striving 'middle-class' voters (Parker ). The government creates and reinforces distinctions along the work/ non-work axis in policy and rhetoric. State benefits are now said to support the 'shirkers' (living on benefits has become a 'lifestyle choice', it is claimed), a narrative that reflects, at least in part, contemporary prejudice against working-class citizens. Low-skilled workers, for example, face a different and altogether higher risk of unemployment than the nation's 'strivers'. Stories in the popular press capture the imagination of members of the public, unem-ployed citizens are branded 'skivers' who 'free ride' on the actions of others in society, it is claimed, especially relying on the endeavours of the 'hardworking majority'.
In the empirical analysis, we therefore examine welfare attitudes in British society at the aggregate level, but we also require a better understanding of attitudes defined by traditional indictors of welfare allegiance, such as social class and identification with political parties.

Methods
The most reliable and consistent source of survey data on public attitudes to social welfare in the UK comes from the BSA (Park et al. ). This nationally representative, repeated cross-sectional survey (described in more detail in the Appendix) has explicitly set about capturing public views on issues relating to welfare services in Britain annually since . Thus, it offers a sound basis for investigating the relationships among social attitudes and welfare provision in Britain. In the context of recent policy developments, we are particularly interested in public attitudes towards unemployment protection and social security benefits that aim to alleviate poverty. The BSA, for example, solicits views about benefit levels for unemployed people, asking respondents to judge whether out-of-work benefits are too generous or too low and could cause hardship. Further questions ask whether benefit levels discourage unemployed workers from finding paid work and whether the conditions placed on claimants and jobseekers are too weak to secure successful 'back-to-work' outcomes. The survey asks respondents to consider whether unemployed people in their area could find paid work if they wished, thus tapping into the extent to which the public subscribes to either individual or structural explanations for the causes of unemployment. The survey also solicits views about welfare spending on poor people more generally and asks whether the social security system itself is a priority for extra public expenditure. The focus throughout is on attitudes towards work and welfare, and income security for the working-age population. Attitudes towards retirement incomes and benefits, for example, or public services such as healthcare are tasks for another day.
In the analysis, we employ a range of descriptive statistics and multivariate regression techniques, including logistic regression for survey responses in binary form (Hosmer and Lemeshow ). Using the core questions and variables relating to welfare and work, we consider long-term attitudinal trends and views defined by class and allegiance to political party (restricted to the three main political parties in British politics, Liberal, Labour and Conservative). As Crompton () observes, 'social class' remains a contested concept in the social sciences. In this analysis, we are forced to rely on occupation (which provides a good indication of class and employment relations), adopting the standard occupational classification scheme applied to social surveys in Britain (Rose et al. ). According to this perspective, classes are aggregations of positions within the labour market. In Britain, we observe the following class structure: traditional higher professional occupations ( per cent of the British working population belong to the professional class); the managerial and technical 'white-collar' roles (one-third of the British popula- VOL. 49, NO. 7, DECEMBER 2015  tion in employment belong to this class); one-fifth of workers are in skilled non-manual occupations (the lower-middle-classes). Turning now to the working-class, or working-classes, about one-fifth of the working population are in skilled manual occupations. Fifteen per cent of workers are in semiskilled occupations and  per cent of the working population are unskilled labourers, the smallest category. The reduction of class to occupation threatens to leave out of the model those citizens who are 'unemployed' (the benefit-dependent group), and those citizens who are 'economically inactive': the retired population, students or unpaid family carers. In the analysis, however, other socio-demographic variables are included to produce a multifactoral approach to class and social relations (including labour force status, see table ).
Multivariate logistic regression models are used to assess the socioeconomic and demographic characteristics that help to explain or predict attitudes to welfare provision in the national population. BSA explanatory and outcome variables are shown in table . The odds ratios in the results tables (tables  and ) show the strength and the direction of the independent predictors, and asterisks indicate the level of significance ('independent' here means after taking account of all of the other demographic and socioeconomic variables in the model). Much of the discussion focuses on the multivariate results in table , with cross-referencing to the interaction effects for class, education and political party allegiance shown in table . All study calculations are weighted to correct for differential and non-response bias in the survey data.

Descriptive analysis
The social survey evidence points to a fundamental shift in public views on welfare provision over the past three decades. In figure , we find -throughout much of the s and s -a widespread belief that out-of-work benefits were set at such derisory levels that they caused significant hardship for people living on them. In , for example,  per cent of the British population said that out-of-work benefit rates were too low. By , that figure had peaked at  per cent (figure ).
Then at the start of the st century, a distinct attitudinal shift begins to emerge. People no longer accept that benefit levels are inadequate: only one person in five said they were set too low in  -just  per cent of the British population. Congruent with this trend is the growing popular belief that the standard of living for claimants on welfare is too high. Generous benefits are now seen to discourage work and encourage 'welfare dependency'. According to the latest data,  per cent of the British population now believe that out-of-work benefits are too generous and promote the 'dependency culture' that is now seen to exist in the UK. Attitudes towards unemployed people are clearly changing and hardening fast. The attitudinal trends reported in figure  are all the more striking when we consider their economic context. The real value of unemployment benefit in Britain has changed very little over   this period (if anything, there has been a marked decline against average earnings, figure ). Many working-class families struggle to manage on benefits that are kept below any social standard of decency in order to strengthen the incentive to find work in the growing number of low-paid, low-skilled jobs that involve poor-quality work (Shildrick et al. a). 4 With the growth in    casual and temporary (non-standard) employment, many families are now trapped in a cycle of low pay/no pay poverty as citizens move in and out of the labour market.
The visibility of the reaction against social protection for women and men of working-age increases with the addition of the linear trend lines, as shown in figure . The trend lines help to expose the clear and growing divergence in the attitudinal survey data and the popular drift away from support for better welfare (because benefits are too meagre and cause hardship) towards the now popular belief that they are, in fact, too generous and thereby encourage 'welfare dependency'. There is significance in the point of intersection in the diverging trend analysis, seen around the year . In the approach to the  general election, the Labour Party was making the case for active labour market policies, a feature of its general election manifesto (Labour Party ). Tony Blair now spoke about actively reforming the welfare system, under Labour 'welfare will be a hand-up, not a hand-out', he declared (Blair ). Following communitarian thinking, Labour now argued that a good society combines respect for individual 'rights' with the principle of 'reciprocity', that members must accept responsibilities for their families and the wider community at large (Gilbert ). The 'New Deal' policies, introduced by Labour in , set out the workfare strategy, which emphasized that welfare would now be more clearly defined by labour market participation and mandatory work-related activities. In British social policy terms, the 'New Deal' schemes represented a radical ideological departure from the traditional role of social security, which had hitherto focused on redistributing provisions to support those without work.
Our interest in the analysis concerns the public standing of welfare intuitions, questions over 'legitimacy' and 'trust' arise. In figure , we see the   number of people who believe unemployed workers and benefit claimants could find work if they really wanted to has steadily increased over the study period. Over half the population now say claimants could find work if they wanted, compared to a quarter in the early-s. Activation workers (JCP staff) also mistrust claimants, who they generally consider and label 'work-shy' (mentioned earlier). Thus, it could be argued that the new punitive policies currently being implemented by the coalition government are a necessary response to the 'legitimacy crises' of the welfare state. New policies must look to address the individualized pathologies of the 'dependency culture', if indeed it exists in Britain on the scale that some politicians would have us believe (research suggests the 'culture of worklessness' thesis is overstated in the UK, see Shildrick et al. a), but policymakers certainly must do more to tackle social and labour market disadvantage.
Unemployment continues to be a real and persistent risk facing workers (particularly low-skilled workers) in the advanced economies (OECD ). In the UK, we also find generational and regional divides in opportunities to find paid work (Whitworth ). Increasingly, however, the survey data suggest declining solidarity with unemployed citizens; perhaps those in employment care much more about their own job security than the unemployment of others, or perhaps most British citizens no longer recognize or accept unemployment risks themselves: risk is increasingly concentrated at the bottom of the class structure, after all. In the growing climate of distrust, it is also possible that more affluent citizens may be content to abandon collective forms of social protection. The pursuit and understanding of security has increasingly become an 'individualized' if not altogether private matter. Self-insurance and individual adaptability to labour market risks are the only options that remain in the face of declining solidarity. Remarkably, negative attitudes towards unemployed workers continued to hold firm with the economy in recession during - and - (figure ). Such well-formed opinions are likely a reflection of the continued 'moralising' over the 'social problem' of persistent unemployment. The media tends to blame, shame and stigmatize unemployed people and welfare claimants (Taylor-Gooby ; McEnhill and Byrne ). Citizens who fail to secure paid work are the social differentiated 'other' (not-me) it seems; once again defined as the 'undeserving poor' members of British society as Coats () observes.
Attitudes towards welfare benefits for poor families appear broadly consistent across the classes. The distribution of views according to class found in  is broadly similar to that seen in , albeit at a much lower level (figure ). In ,  per cent of working-class respondents said government should spend more on welfare benefits for poor families, compared with about half of middle-class respondents. In , only one-third ( per cent) of working-class respondents agreed that the government should spend more on welfare benefits for poor families (a decline of  percentage points), compared to one-quarter of middle-class respondents. The fall in the mid-s is a likely response to New Labour repositioning itself during this period. 'Third Way' reforms replaced welfare with workfare. It appears that traditional Labour voters followed the direction of their party on this issue, adopting some of New Labour's values (Curtice ). The shift in political discourse -and the 'Third Way' politics associated with Tony Blair -proved popular with the electorate and working-class voters. Under Blair, Labour went on to win its second and third consecutive victories in the  and  general elections; however, the longer-term costs to collective welfare services and unemployment protection are only now becoming apparent. Social security is now less of a public priority than it was in the early-s, and marked class differences in support of additional social security spending are diminishing (figure ). By , we find that just  per cent of working-class representatives and just  per cent of middle-class voters regard social security as a priority for government spending.
In figure , we see that supporters of the Conservative Party are the least likely to perceive that government should provide better welfare benefits to help poor families; Labour Party supporters have traditionally been the most likely to hold pro-welfare-state values, compared to supporters of the Conservative or Liberal Parties. However, we also observe that support for welfare among Labour voters has been in steep decline over the last two decades. In , for example,  per cent of Labour Party supporters agreed that the government should spend more on welfare benefits for poor families, compared with just over one-third ( per cent) in  (a decline of  percentage points). Among Conservative voters, support for collective action to alleviate poverty has been relatively low in comparison and the decline over the two decades more gentle (figure ). The extent to which the long-running downward trend will continue as we approach the  general election remains to be seen but a reversal of the trend seems most unlikely, given the direction of workfare policy in the UK. In the next section, we seek to model explicitly    some of this heterogeneity in welfare attitudes to understand the changing nature of welfare provision in Britain.

Multivariate analysis
Regression models are used to explain or predict attitudes to social security provision in the national population. Negative views appear widespread, but some members of society may be more critical of welfare than others. Here we want to identify more clearly those who claim that out-of-work benefits are too generous, those who think that unemployment benefits disincentivize paid work, and those who believe that work conditionality in the welfare system is too weak.
In the results (table ), we observe the importance of age, gender, ethnicity, education, labour force status, household income, occupational class, allegiance to political parties and region of residence in shaping public attitudes towards social protection for the working population. Age influences views about welfare. Young adults (aged -), for instance, are significantly more likely than older adults to believe that out-of-work benefits in Britain are set too high (table : Model a). Other things being equal, the odds of a person aged - years claiming unemployment benefits in Britain are too generous are more than three times greater than for an older person aged  or above. It is fascinating to learn that young adults are taking a tough stance on welfare compared to older adults, given that they are arguably experiencing the brunt of the social reforms relative to social protection policy in later life for pensioners.
Jobseeker's Allowance (JSA) and housing benefits could be withdrawn for the under-s who are not in work, education or training under new proposals announced in  by Prime Minister David Cameron. 6 Around one-fifth of young adults (aged -) are unemployed, well over twice the national average. Young adults are also more likely to believe that work conditionality in the British welfare system is weak, as are adults aged - years (table : Model a). The relative odds of a young adult reporting the social security system suffers from weak conditionality is more than three-and-a-half times greater than an older person aged -plus. Men are  per cent more likely to believe that benefits are inadequate, compared to women. Ethnicity also appears to shape attitudes towards the welfare state. Belonging to a black and minority ethnic (BME) group also significantly increases the relative odds of reporting that out-of-work allowances in Britain are too generous (table : Model a) and deter claimants from working (table : Model a). Citizens from BME backgrounds, for example, are twice as likely to say that state benefit levels are too high and are nearly three times as likely to say that they disincentivize paid work, compared to the majority 'White' British population.
Education is an important predictor of welfare attitudes. Citizens with low education are significantly more likely to be dissatisfied with the British social security system than those with higher levels of educational achievement. With educational attainment below the level of university degree, for example, the relative odds of someone asserting benefit levels are too generous are    per cent greater than for those with a university degree. People with educational attainment below university degree level are also  per cent more likely than graduates to believe the current benefit system is a disincentive to paid work. Those without formal qualifications are twice as likely as graduates to report state benefit levels are too generous (table : Model a), even though this group is most at risk of unemployment. They are also four times more likely than those with a degree to say benefits act to deter paid employment (table : Model a). There is also a clear income effect in the data. The higher the level of household income, the more likely a person is to believe that out-of-work benefits are too generous. Class appears to be an important contributory factor here, but the 'trade union effect' is surprisingly weak, after controlling for all of the other variables  in the model. People in professional and managerial occupations, for instance, are nearly three times as likely to believe benefit levels are too generous, compared to unskilled labourers. Professionals are also significantly more likely than unskilled workers to say the social security system suffers from weak conditionality. Skilled manual workers are three times more likely than unskilled manual workers to believe out-of-work benefits are too generous, but not skilled 'white-collar' workers. The interpretation is clear. Compared to skilled manual workers, who find it easier to maintain continuity of employment in better paid roles, low-skilled workers (if they are employed) are more likely to experience precarious employment and non-regular work, including casual and temporary employment. Job insecurity and the experience of being pushed out and pulled into the labour market are defining features of precariousness in modern societies, as Shildrick et al.( b) observe. Labour force status is also a particularly strong determinant of attitudes towards welfare after controlling for other factors. Compared to unemployed adults, for example, people in paid employment are five times as likely to say that welfare benefits are too generous (table : Model a) and they are three-anda-half times more likely to believe that the conditions placed on claimants are too weak to secure successful back-to-work outcomes (table : Model a).
Citizens who are economically inactive (covering those who have retired from the labour force or people involved in home duties) are also significantly more likely to say that unemployment benefit levels are too generous and the conditions placed on jobseekers are too weak, compared to adults claiming out-of-work benefits. Social security against the adverse effects of unemployment is clearly of value to the British citizens who are protected (at least in part) by social policy. Partisanship helps to account for welfare attitudes. People who vote for the Conservative Party are significantly more likely to believe that out-of-work benefits are too generous, compared to citizens who vote for the Liberal Democrats (table : Model a). They are more likely to say that the social security system fails to properly incentivize work (table : Model a) and that conditionality is too weak (table : Model a). The relative odds of a Conservative voter saying that benefit levels are too high are nearly five times greater than for a Liberal or Labour voter. Conservative voters are also four times as likely to say that welfare deters paid employment and are twice as likely to believe that workfare conditionality in Britain is too weak, compared to citizens who vote for the Liberal Democrats (suggesting Cameron's new 'Work for the Dole' policy, discussed later, will appeal to the Conservative Party faithful).
Religious beliefs have some impact; people with religious views assert that the new workfare conditions attached to benefits are too harsh (table : Model a). Lastly, we observe some spatial patterning in the survey data. People living in central London, who experience a high cost of living and will bear the brunt of the new Housing Benefit cap, are significantly more likely than people living in Outer London to say that benefits are inadequate and, as such, do little to deter paid work. People living in Inner London, for example, are four times more likely to believe benefit levels are inadequate, compared to people living in Outer London. People living in Yorkshire are weakest members of society), however, policymakers appear to have lost sight of important structural and economic drivers, and employment constraints. In the process, politicians are actively failing to engage the public on important issues of social policy: the problem of persistent 'structural unemployment', labour market fluctuations and labour market inequalities, the lack of jobs (particularly in some areas), the lack of 'decent work' (which is not just about 'paid employment') more generally (Dean ). Some groups have higher unemployment rates than others: policymakers have clearly misunderstood the complex structural causes of youth unemployment (with unemployment levels amongst young adults at record highs of  per cent in the UK). Policymakers can no longer ignore particular issues facing single parents (work-life balance and childcare are among them), in-work poverty (especially among working families with young children), and the social insecurity of people who cannot find work in the open labour market (adequate social protection can help to reduce poverty and this is part of the unresolved contradiction between the human right to welfare and the realities of human suffering created by government policies that tolerate poverty and exclusion). 7 In the political circumstances, fundamental changes in the dominant social values in Britain may not be a surprise, as they are evidenced by the declining solidarity with unemployed citizens and the growing lack of concern for the material hardships faced by unemployed workers and people living on low incomes than there was  years ago when the BSA first began. The findings also point to a growing lack of 'trust' in the population. Most people now firmly believe that JSA claimants could get a job if they really wanted one. The  recession dented this long-term trend somewhat but not substantially. The survey findings suggest a fundamental shift in views on the underlying causes of unemployment. The survey data clearly reveal a departure from societal or structural explanations for unemployment towards individualist interpretations with an emphasis on human agency. The British public now sees work aversion and the declining 'work ethic' as one of the main issues facing society. Coupled with this trend is a growing belief that out-of-work benefits are now too generous and act to promote the 'dependency culture'. This view is widely held, despite evidence to suggest the real value of unemployment benefit in Britain has changed hardly changed over the past  years.
Some of the largest attitudinal shifts have occurred among sections of the British population that have traditionally supported the principles of the welfare state: Labour voters and representatives of the working-class. Declines in support for welfare amongst Labour voters are particularly marked. The findings for 'social class' (defined by occupation) are perhaps more curious. The figures show significant declines in working-class support for collective welfare insurance to protect workers from unemployment risks, a possible reaction to 'scroungerphobia', as working-class citizens attempt to dissociate themselves from popularized notions of the 'undeserving poor' (Shildrick and MacDonald ). On closer examination, however, we find a more gradated field of class relations, with different (but recognizable) stratifications of unemployment risk reflected in the survey data. In the regression analysis, we find that only manual workers in skilled occupations -about one-fifth of the working population -now believe that benefits levels are too generous. Other things being equal, unskilled and semi-skilled workers currently make no such assertions. Low-skilled workers who experience greater job insecurity appear to have a stronger desire to see the welfare safety net more securely woven to protect their living standards in the event of unemployment.
A number of more general conclusions can be drawn from the British experience. First, we can observe the changing point of political gravity towards the right. 'Third Way' ideas have moved the 'traditional' left to the centre -welfare policies of the centre-left have become similar to the centreright. There now appears to be a growing political consensus around the merits of the workfare project, a recognizable attempt by the state no less to settle capitalism's contradictions and the (ongoing) legitimacy crisis of the (welfare) state (Offe ). Here we may also draw attention to the role of political discourse and the actions of political parties in shaping public attitudes in the policy process, leading to the growing acceptance and likely embedding of workfare values and principles in the collective psyche. British public attitudes appear to have followed the logic of government policy. Labour supporters, in particular, appear to have accepted the workfare line promoted by Labour under Blair. The present Conservative-led coalition government has pursued further workfare reforms and reductions in welfare spending with little political or public opposition. Second, there appears to be a new and more divisive type of 'welfare politics' emerging in Britain and other advanced liberal democracies (Deeming c). In the political contest for votes, workfare ideology not only divides workers but also further depoliticizes the relationships between capital and labour in the process. Policy representations by political parties and divisive politics played out in the media are implicated, as is the objective situation whereby workers compete for shares of a diminished pie of scarce resources. Recent scholarship (Nolan et al. ; Piketty ) reminds us that, as more of society's resources are distributed upwards to the wealthy members of society, the middle stratum of skilled workers (manual and non-manual) compete with low-skilled labourers and unemployed workers at the bottom of the class structure.
The Conservatives under Cameron are very much on the side of 'hardworking people' (their political slogan at the  Party Conference) and now seek to extend the principles of workfare on social justice grounds. New mandatory work conditions will apply to long-term unemployed people. 'There is no option of doing nothing for benefits', claims the Chancellor, George Osborne, 'no something-for-nothing any more'. 6 Under the 'Work for the Dole' programme, those in receipt of JSA will be required to give something back to their community in order to maintain eligibility to benefit support: making meals for older people, clearing up litter, or working for a local charity are among the options currently being discussed (such workfarist regulatory principles can be traced back to the postwar period, particularly in the USA, see Mencher ; Piven and Cloward ). Whether 'Work for Benefit' policies represent a responsive 'thermostat' (Herbst ) gauging the temperature of public opinion or a 'weathervane' (sensing the shift) is not altogether clear. In other words, will the political project 'workfare' gain momentum? Will the British public accept or demand more workfare measures in response to what policymakers are doing -or fewer? Particularly telling, however, is that the Labour Party has mounted little opposition to the latest welfare reforms and certainly offers no prospect of reversing the workfare trend as we approach the next British general election in .I n response to new voter demands, Labour no longer appears to promote the interests of 'labour' (including unemployed workers and those people precariously employed), but instead seeks to attract the votes of 'hard-working families' and remains committed to raising their living standards. (The term 'hard-working families' appears to have entered the political lexicon in the UK during the  election, appearing in Labour's manifesto, Labour Party .) The notion of 'hard-working' citizens is not simply a rhetorical device aimed at a political constituency -middle-class voters. It reflects a range of normative assumptions about the way in which life should be lived in stcentury capitalist society. Work is seen as the best form of welfare, not only because work pays better than welfare but also because it promotes participation, inclusion and well-being. Thus, 'workfare' policy follows a certain economic and political (but socially divisive) logic, with the main political parties in the UK appealing to the same political constituency. The challenge for social policy, as Rueda () argues, is to overcome such 'dualizing' processes and 'dualistic' political tendencies that inhibit the development of a more inclusive society. A few preliminary thoughts on how we might actively pursue this agenda for change follow.

Limitations and future research
The majority of the British public now believes that life on out-of-work benefits has become too easy and that the British social security system may need (further) reform. Whether the opinion polls amount to anything other than market-testing for the effects of political rhetoric is unclear. Certainly, this is a criticism that could be levelled at the BSA data. Respondents may not be informed about the issues at stake: they may have little experience of budgeting on benefits, for example, or may not be aware of the actual sums JSA claimants receive in benefit.
It appears that the British public favours social security reform, but again, with current survey data, it is not altogether clear how we should interpret the public mood. The public may prefer to see social insurance principles reasserted in British welfare policy, rather than US-style mandatory work programmes. Labour recently announced plans to strengthen the link between contributions and benefits, which they hope will restore public faith in collective insurance against unemployment to provide a decent standard of living above the means-tested minimum safety net (the British social security system has a much weaker contributory element than many other countries such as Germany for example, see Clasen, ). 4 The reciprocal principle (which is the basis for public support for the welfare system) has always been strong in British social policy, from Beveridge's early report, 'benefit in return for contributions, rather than free allowances from the State, is what the people of Britain desire' (Cmd  : ). Unfortunately, however, this is unlikely to help the many carers and the growing numbers of people now leading  precarious and uncertain lives, trapped in the cycle of low pay/no pay poverty under the ever-watchful eye of the punitive (workfare) state.
Further research might therefore explore some of the issues emerging from the present analysis. We need to know whether (or the extent to which) the BSA is capturing or reflecting real and deep-seated value changes. Qualitative inquiry is likely to prove the more reliable option for this task. Qualitative research might also consider the prospect of restoring or strengthening the contributory principle in Britain's system of welfare, and the ethicality of the Conservative's more radical American-style community work programme. Lastly, there are limits to the value of cross-sectional surveys, such as the BSA survey, in terms of understanding how individuals themselves change their views in response to government policy. Panel surveys (such as the UK's 'Understanding Society') capture this dynamic; they can shed more light on those individuals who have consequently accepted that there is no alternative to workfare policy. Comparative analysis may shed further light on the social structure of attitudes to welfare, and provide a deeper understanding of the relations between ideology, political parties, public opinion and public policies. This is a topic that merits further investigation. changes over time. The BSA includes a series of standard questions on political party affiliation and political values. Regarding 'social class', the BSA employed the Goldthorpe-Heath class scheme up to , and the National Statistics Socio-Economic Classification class scheme from . and Iain Duncan Smith, Secretary of State for Work and Pensions,  Party Conference speech, http://www.conservativepartyconference.org.uk/Speeches/ _Iain_Duncan_Smith.aspx (accessed  June ). . For example, Article  of the  UN Universal Declaration on Human Rights states that, 'everyone, as a member of society, has the right to social security', and Article  states that, 'everyone has the right to a standard of living adequate for the [ir] health and well-being', http://www.un.org/en/documents/udhr/ (accessed  June ).