‘21st Century Welfare’ in Historical Perspective: Disciplinary Welfare in the Depression of the 1930s and Its Implications for Today

Abstract

Since 2010, UK governments have intensified conditionality as part of a programme of ‘welfare reform’. Social scientists have undertaken much critical analysis but less attention has been paid to possible historical parallels. This article sheds new light on welfare reform through comparison with the depression of the 1930s. It undertakes a documentary analysis of policy in the 1930s informed by a governmentality perspective. In both periods, governments committed to liberal orthodoxies and feared the unemployed would become vulnerable to ‘demoralization’ and ‘dependency’; their behaviour and character were determinant of their rights to support. However, there are notable differences in what interventions have been considered appropriate. The article assesses the significance of continuities and contrasts, and argues in particular that the severity and ubiquity of behavioural regulation employed today is even greater than that seen in the ‘dark decade’ of the great depression.

Introduction

[B]y reforming the welfare system, including benefits reform, we will take welfare into the 21st century.

Iain Duncan Smith (Cabinet Office, 2010)

The post 2010 programme of ‘welfare reform’ has been described as the biggest change in the UK’s social security system since Beveridge: ‘a vision for a new welfare settlement; a welfare state fit for the 21st century’ (Department of Work and Pensions, 2014). It has been as marketed as a great innovation, as a novel project informed by the behavioural sciences and implemented through modern digital technology. It has continued the trend ‘from a more enabling programme of employment assistance to a more punitive and controlling activation strategy’ (Edmiston et al., 2017: 254) via increases in the conditions attached to the receipt of benefits. Dwyer and Wright (2014) have demonstrated that an increasingly ‘ubiquitous conditionality’ defines the treatment of those receiving state support and that this has chipped away at the basis of social citizenship. As more and more entitlements become conditional, vulnerable groups are increasingly excluded (Reeves and Loopstra, 2017). This article furthers these observations by placing these escalations in a historical perspective. It contends that neoliberal welfare policy can be compared productively to the pre-welfare state liberalism in existence during the inter-war years, and that such a comparison demonstrates the extremity of the present-day conditionality regime in a way that should alarm contemporary observers.

A number of recent articles have examined escalations in behavioural conditionality. Fletcher and Wright (2017) have argued that ‘welfare reform’ has been marked by a growing authoritarianism in the benefit system, where behavioural conditionality was continually escalated. This has been seen as a means of enforcing labour discipline beneficial to capital for example, by Wiggan (2015). These measures have been implemented in a context of a growing hostility to the poor. For example, welfare reform has been justified with narratives about ‘workless’ communities and with a new right critique of the welfare system as generating ‘dependency’ (Deeming, 2015; Lehtonen, 2018). This critique has involved a heavy stigmatization of benefit claimants who were framed as problematic individuals in need of discipline and control via an escalated regime of conditionality and sanctions (Redman, 2019). Elsewhere, however, attention has been paid to possible historical parralells. Welshman (2006, 2017) has highlighted similarities between the ‘underclass’ discourse underpinning ‘welfare reform’ and the ‘Social Problem Group’ concept in the inter-war years highlighting that the ‘current interest in concentrations of worklessness, and in intergenerational poverty, also carries striking echoes of earlier debates’ (Welshman, 2006: 604). This has also been discussed by Taylor (2018) who discusses the restoration of interest in the Victorian concept of ‘character’ under austerity. This ‘character’ has become an important aspect of employability discourses with elites attempting to design policies to inculcate desirable qualities in the unemployed. As in the Victorian era, ‘character’ is intended to provide the lower classes with qualities which would enable them to endure hard times. While Fletcher (2015) has compared the operation of 21st century ‘workfare’ schemes with the Work Camps run for the long-term unemployed in the 1930s. ¹

This article sets out to provide a more in-depth understanding of the historical antecedents of ‘welfare reform’, and their significance. It does this by looking at the treatment of the unemployed during the great depression of the 1930s. By doing so, it shows that many new techniques of behavioural regulation represent revivals of practices from the past. Similar preoccupations with the behaviour and attitudes of claimants existed in both time periods with policy makers creating mechanisms to distinguish those deserving of aid from those likely to exploit public funds. However, the article demonstrates that there are significant observable differences between the way conditionality was structured in the two periods, with different groups of claimants receiving very different treatment. The article uses a governmentality analysis to examine the way conditionality was justified and enacted.

Policy makers have long sought to distinguish the ‘worthy’ from the ‘problematic’ claimant ‘to create legal borders to map the moral boundaries distinguishing “deserving” from “undeserving” claimants’ (Whiteside, 2015: 151). However, for the most part contemporary social scientists charting the emergence of neoliberal welfare and the post 2010 project of ‘welfare reform’ have paid less attention to the possibility of drawing longer historical parallels, apart from references to the 19th century Poor Laws. The use of history by politicians and think tanks in these policy debates has often been restricted to discussions of the Beveridge Report of 1942 (e.g. Hughes, 2014; Mulheirn and Masters, 2013). Less attention has been paid to the inter-war years despite some significant similarities. The two periods were both characterized by economic crisis (Eichengreen, 2015) and by a dominant liberalism which presented austerity in public finances as crucial to national recovery (Blyth, 2013).

This article is structured as follows: the next section sets out the theoretical and analytical framework used. The article then moves on to discuss first the way the unemployment problem was constructed in the policy discourse in the 1930s. Second, the article will explain the way in which behavioural conditions were applied to different groups of unemployed. The focus then moves back into a discussion of what instructive similarities and differences can be found between the 1930s and the period since 2010.

Governmentality and ‘welfare reform’

Governmentality literature in social policy follows Foucault by focusing ‘on the intervention of the state through organizations, procedures, rules and categories’ (Boland, 2016: 3). Of central interest is how administrators or governors perceive subjects and populations and how ideal outcomes and subjects are constituted in discourse and translated into techniques of governance.

Disciplinary power ‘‘makes’ individuals; it is the specific technique of a power that regards individuals both as objects and as instruments of its exercise’ (Foucault, 1977: 170). It is a matter ‘of “making up” citizens capable of bearing a kind of regulated freedom’ (Rose and Miller, 2010: 272). With regard to unemployment benefits, disciplinary measures seek to ‘make up’ a subject that can be (and seeks to be Peeters, 2013) re-integrated into the labour market and no longer dependent upon state funds (Dean, 1995). Foucault’s recently published (in English) lectures On the Punitive Society (2015), delivered in the run-up to the publication of Discipline and Punish (1977), provide an interesting supplement to the latter. They contain a greater focus on the development of policing, driven by developing capitalist production. They show that disciplinary social policy is not neutral regarding the labour market; economic liberalism ‘and disciplinary techniques are completely bound up with each other’ (Foucault, 2010: 66). Conditionality is a capitalist technology seeking to enforce ‘the worker’s duty to offer his labour-power on a free market’ (Foucault, 2015: 173).

The conditionality increases seen in the present day depend on a view of passive subjects, ‘workless’, and ‘dependent’, who require a firm guiding hand after a period of over-indulgence. This view also provides a justification for escalations which are represented as in the ultimate interests of the claimant who has neither the capacity to refuse (as an irrational subject) nor the right to do so (as the morally indebted recipient of aid).

But these interventions are not applied evenly: their application is dependent upon a system of classifications. This differentiation and ranking of individuals can entail very different forms of treatment. ‘The distribution according to ranks or grades has a double role: it marks the gaps, hierarchizes qualities, skills and aptitudes; but it also punishes and rewards’ (Foucault, 1977: 181). In the UK, the different conditions placed upon those receiving payments, such as demands that they train, search for work, or carry out unpaid work, are stratified by the individual’s status within a liberal ‘moral order’ (Boltanski and Threvenot, 2006).

Documentary sources

This article uses original documents from the 1930s. In the research project, almost 50 files held at The National Archives (TNA) of UK were examined, 15 files held by Birmingham City Council and five TUC files pertinent to benefit administration. TNA holds the files of central government departments. Sources of particular use included the AST7 file series of the Unemployment Assistance Board (UAB), a body established to administer benefits for those whose national insurance contributions had expired. In addition, the MH57 file series of the Ministry of Health’s Poor Law Department. The Ministry of Health was historically responsible for the Poor Law and retained responsibility for overseeing localized benefit schemes.

Another useful source was the Royal Commission on Unemployment Insurance, a cross-party investigative body appointed by the government to examine wide-ranging questions about the nature of the unemployment problem. However, while the remit was wide ranging, it started from the premise that insurance should be financially self-supporting. The document is useful because of its central role in the formulation of policy responses to unemployment, the commission’s recommendations were largely adopted by the National Government in an act of 1934.

However, significant parts of the UK’s welfare system during the period were run by local government, especially the most punitive parts of the system for the uninsured. Birmingham was chosen as a local case study for this research. As it provides an exemplar of how policy was supposed to work. The city was governed along classical liberal principles and was praised by government for its implementation of policy without unrest or defiance. The purpose of this case study is to show the translation of policy objectives into practice.

TUC files provide a counterweight to official sources. In the aftermath of the 1926 general strike, the TUC was concerned to inform government policy using its own social affairs department, which collated evidence from its branches, to inform its national arguments.

Documents were selected through searches of the archive catalogues of these institutions. This search was conducted purposively; it was exploratory and a ‘saturation’ criterion was used to determine when to stop searching files. The selection of documents is structured by what survives and what does not, records being always partial and incomplete (Scott, 1990). Present-day documents were selected purposively, and government papers which set out the case for and provided justifications for welfare reform have been examined.

This article is concerned with how the unemployed were and are characterized in policy discourses and how this was translated into administrative devices. Documents were read – as well as sampled – purposively with the aim of shedding light on present-day policy; hence, aspects of policy resembling the present day were sought out. History has been studied with the present firmly in mind.

Representations of the unemployed

Fears of the moral decline of the unemployed on benefits are not new. If the problem is represented to be (Bacchi, 2009) one of ‘worklessness’ and a culture of ‘welfare dependency’ in our own time, how was it conceptualized in the past? We can see a strong precedent in the fears of ‘demoralization’ in the 1930s (Jahoda et al., 1972; Welshman, 2006).

Demoralization as a concept was steeped in the individualist assumptions of liberalism (Walters, 2000). It was a disciplinary problematic which understood the loss of values and dispositions favourable to liberal labour markets as the consequence of prolonged unemployment. Consequently, it provided a justification for interventions. The lack of purposeful activities structuring the working day, as well as other socially integrative functions of employment, would lead to demoralization. The concept should be read as de-moralization: the loss of morale and morals was essentially the same thing. It was seen as a serious problem: in 1929, two surveyors, Drs Pearse and Lowry, claimed that ‘every thoughtful person with whom we have talked has expressed greater concern at the destructive effect of idleness upon the character and morale of the unemployed’ (Harris, 2004: 200) than at their decline in physical health. The Pilgrim Trust (1938), an organization founded in the 1930s to provide charitable relief but also to undertake research into solutions for social problems, was concerned that the ‘unemployed community’ helped the unemployed adjust to regarding life without work as a normal one. This would be likely to cause a long-term problem, especially for ‘young men for whom the lessons of self-discipline and the influence of working experience have not come with any great force. They are a problem now. They will be more of a problem as they grow older’ (Bakke, 1933: 188). The implications of such a problematic were that it was

important in the interests of the community to avoid a system of unconditional relief. The state of continued unemployment has almost inevitably a demoralising influence and, if prolonged, may result in a definite decline of initiative and independence. (Royal Commission on Unemployment Insurance, 1931: 119)

This highly moralized view of work as a fundamental good beyond the receipt of wages strongly parallels that seen in the present day by writers like Friedli and Stearn (2015) and Cole (2008). Psychological views of the therapeutic functions of work served and serve to support moralism in social policy.

However, in contrast to the present, while ‘no work’ was seen as a problem in the 1930s, so was ‘bad work’. Since the late 19th century, casual labour had been problematized by social investigators and governments (Whiteside, 2014). Casual work was seen ‘as expensive, inefficient and a source of social and moral degeneration (poverty breeding criminality, sickness and incapacity)’ (Whiteside, 2015: 155) which undermined future working capacity. This view of casual work was echoed in the social surveys of the time; the Pilgrim Trust judged that ‘spells of casual work which a “general labourer” finds for himself are not enough to give men the habit of working’ (Pilgrim Trust, 1938: 151). Casual workers were feared to be happy to subsist on allowances and subsidies, picking up small bits of work at their convenience to enable a rootless and antisocial lifestyle. However, this work was also seen as a feature of poorly organized labour markets, and as a waste of human capital, for example, by a younger William Beveridge (Phillips and Whiteside, 1985). Early proposals for Employment Exchanges aimed to eliminate casual work and so render these industries more efficient. The Exchanges were to be used to identify and exclude the least efficient workers. In 1909, Churchill had argued that it was not possible to divide the ‘vagrant and the loafer’ from the legitimate workman without a set of work tests, that is, the offer of permanent work (King, 1995); by offering these tests, the Exchanges sought to impose the discipline of regular work.

We can see many resemblances between this view and that of the present. The post 2010 programme of ‘welfare reform’ is justified by a critique of the welfare state as ‘a vast, sprawling bureaucracy that can act to entrench, rather than solve, the problems of poverty and social exclusion’ (Department of Work and Pensions, 2010a: 15). Benefit rights and the absence of intensive behavioural regulation are framed as ‘writing off’ claimants. Claimants’ culture and choices are seen as explaining their failure to benefit from opportunities assumed to have been abundant. This is an individualizing responsibilizing discourse of personal deficiency, failure, and recovery which downplays structural or labour market causes for unemployment (Pantazis, 2016). There is less official acknowledgement of these causes even than was present in the 1930s. These increasingly stigmatizing ‘scrounger’ narratives have become not only prevalent in public discourse, but institutionalized in policy (Patrick, 2016). The view of claimants held by policy makers often vacillates between one of cynical manipulators of entitlements taking advantage of public funds, and a different view of claimants as rendered incapable by the dependency breeding effects of the welfare state (Fletcher et al., 2016). There has been a centring of attitudinal problems as a cause of unemployment.

These arguments make individuals the object of intervention, reframing structural crises of capitalist economies as times when ‘families [fail] to enjoy the financial and non-financial benefits of paid employment’ (Gregg, 2008). These social psychological arguments are employed in a paternalistic fashion to justify the coercions of active labour market policy. Official policy is to promote work as a means of social improvement (Friedli and Stearn, 2015). De-commodification (Offe, 1984) of wage relations has been framed as dangerous to the human spirit. Consequently, conditionality is argued to be a device of social inclusion. It is argued to be in the interests of the poor to behave in certain ways whether or not they see it this way. This resembles the poor law administrator, ‘a guardian acting in the interests, but not necessarily in accordance with the wishes, of his wards’ (The National Archives (TNA), 1933, MH57/7 Public Assistance Conference. Draft of the minister’s speech).

Therefore, it can be seen that the view of the problem was characterized by a concern about the deterioration of workers as moral subjects. The next section looks at the measures brought in to correct this.

Conditionality in the 1930s

In the present day, there is a strong consensus among policy makers around activation measures and the enforcement of active job search. Given the similarities in the preoccupations of policy makers with ‘dependency’ and ‘demoralization’ in both periods, it is appropriate to ask what sort of behavioural devices were applied to the potentially ‘demoralized’ to maintain the norms of waged work? Today, there exists an extensive tiered system of punishments. For low-level offences like lateness to appointments or not doing enough to find work (for the ‘able-bodied’ without caring responsibilities, doing enough means demonstrating 35 hours per week of job search) there are 4-week sanctions (Reeves and Loopstra, 2017). ‘Higher level’ ‘failures’ like the refusal of a job or failure to apply for a specific vacancy to which they are directed by the jobcentre can mean a 3-month sanction increasing from 6 months to 3 years for subsequent refusals (Dwyer and Wright, 2014). What measures were applied to meet these objectives in the 1930s? The answer varies a great deal between the claimants of insurance, extended insurance, and Poor Law successor benefits.

The history of support for the unemployed in Britain is marked by a tension between those policies which saw the unemployed as victims of economic circumstance and those which saw them as willing to exploit the generosity of the wider community (Brown, 1990). In the 1930s, the help available reflected a sort of ‘moral ordering’, institutionalized in the divisions between insurance-based benefits and the successors to the Poor Law. Relief for the able-bodied unemployed was divided into three separate systems. First, there was Unemployment Benefit (UB) administered by the Ministry of Labour and paid as a right for fixed periods to workers in ‘insured trades’ upon the basis of their contributions when in work. Second, there was an intermediate system of extended benefits for those who had claimed unemployment benefit for more than its 6-month limit, but who were judged to be unemployed only due to the economic downturn and who were ‘normally’ in insured work. Third, for those unable to satisfy contribution requirements (and for some excluded from assistance for behavioural failings), there was Public Assistance (PA).

The insurance system

Insurance-based benefits had very different conditionality regimes from the present day. In the 1930s, benefit was paid to insured contributors as of right and was not subject to means tests:

There is no interference with, or enquiry into, his domestic economy; he retains full control over his private purse; and the wages and conditions of employment which he may be called upon to accept as suitable are safeguarded. (Royal Commission on Unemployment Insurance, 1931: 157)

The actuarial and contractual nature of these benefits enabled recipients to be accepted as conforming to liberal virtue. After all, ‘[o]nly the client’s subscription ensures the provident cover: it therefore implies regular work, ordered time, disciplined consumption, individual responsibility’ (Defert, 1991: 231). The mechanisms of fixed benefit in return for fixed numbers of contributions had originally been envisaged as making the system self-policing. In Winston Churchill’s phrase, it was the ‘morality of mathematics’ (Deacon, 1976), which eliminated the dangers of demoralization and abuse.

There was relatively little conditionality in the insurance system. Insurance rewarded work discipline if it could be evidenced by contributions, but the behavioural conditions seen today were largely absent at least for 6 months until they were required to claim extended benefits.

Extended benefits

The extended system was founded on the view that if

these workers are capable of work and show that they will normally seek to obtain their livelihood in industrial employment, the community has a duty to provide for their needs during unemployment, and to take such steps as are practicable to check the deterioration, in employability and character, which may follow. (Royal Commission on Unemployment Insurance, 1931: 118–119)

Those ‘normally’ in respectable and regular work (TNA, 1933–1937, AST7/20, Bullard’s minute of 14 September 1934) were to be distinguished from the ‘ordinary poor’ (Lowe, 1986). This principle started with Out of Work Donation, introduced after World War I to maintain ex-soldiers and their families, and extended to former munitions workers (Lynes, 2011). However, the treatment they could expect fluctuated as government sought to balance on one side the cost of the scheme and desire to defend the contributory principle and on the other the political impossibility of dealing with mass unemployment using the Poor Law. This issue remained unresolved throughout the 1920s as governments of varied stripes continued to perpetuate extensions of insurance benefits under a range of different labels and subject to a variety of conditions – not least the Household Means Test that provoked mass protest (TNA, 1931, MH 57/14, Minutes of the Transitional Payments Conference: General Administration), over its impoverishment and humiliation of previously independent skilled workers (Garside, 1990; Hannington, 1977). In 1934, the system became Unemployment Assistance (UA) run by the central Unemployment Assistance Board, a centralization of power from the previous incarnation; Transitional Payments which had been administered by Local Authorities.

For claimants of extended insurance benefits, something like modern job search conditionality had been applied in the Genuinely Seeking Work Test of the 1920s. These tests were designed to trap or catch out claimants whose status and attitudes aroused suspicion. Their purpose was to determine ‘the state of the applicant’s mind’ (Deacon, 1976) that is, his disposition towards the labour market. Political pressure from workers, unions, and the unemployed led to the test being scrapped. The tests were replaced with the Capable and Available criteria which put the burden of proof upon the authorities to prove the claimant’s unwillingness to take an offer of work (TNA, 1934–1938, AST7/64) by providing evidence that a claimant had been notified of a vacancy by the Employment Exchange but had refused it. ‘Availability’ was understood to be a ‘passive state’. The Royal Commission on Unemployment Insurance (1931) ensured that the Genuinely Seeking Work Test was not reinstated and regarded the new criteria as sufficient to deal with very small numbers of the genuinely workshy through a ‘maximum disqualification’ of 6 weeks benefit.

To some extent, the object of concern was still to exclude the scrounger. After all, ‘the Board would not be carrying out its duty if it shut its eyes to repeated unreasonable refusals’ (TNA, 1934–1939a, AST7/76 Note of 20 November 1934). Officials continued to debate what could be done about claimants who refused work and who were suspected of being a ‘troublesome type’ (TNA, 1934–1939b, AST7/79). However, this could not mean falling back on the Poor Law methods of ‘the work centres or the workhouse, [as] one can imagine the outcry’ (TNA, 1934–1939b, AST7/79 Reid to Eady 14 January 1936). The option remained to consider claimants not available for work and to rule them ineligible (a 6-week disallowance), but there was reluctance to resort to this measure. Instead, recommendations were issued to officials to warn claimants not to continue to fail to ‘avail themselves’ of opportunities. While this resembled the Genuinely Seeking Work Test, there was an aversion to reviving it and to appear to be too close to its methods. There appears to have been internal debate within the UAB on the question of even how to interpret the refusal of a job. Myrddin Evans, the Deputy Chief Insurance Officer (Saville, 2004) at the Ministry of Labour, went as far as to say,

there may be isolated cases, of course, but our experience shews that the reasons put forward for refusing to accept offers of employment are, in the main, genuine rather than frivolous and not such as to raise a resumption of non-availability. (TNA, 1934–1938, AST7/64 Myrddin Evans to Bullard 20 November 1935)

However, it was still possible to ‘reasonably draw the inference from that person’s conduct that he had so hedged about the field in which he was bona fide seeking employment that there was no reasonable prospect of his obtaining work at all’ (TNA, 1934–1938, AST7/64 Note of 4 October 1935).

Overall, a very different political climate prevailed in the 1930s than today, despite the political pressures to restrict spending coming from the Treasury (Peden, 2000). Attempts to treat the ‘respectable’ skilled working-class man as a pauper provoked mass protest. Today’s regime demands that the claimant prove themselves through a demonstration of constant effort to seek employment but attempts to do something similar in the Genuinely Seeking Work Test had failed by the 1930s.

Public Assistance

In addition, there was PA: the successor to the old Poor Law, shunned by ‘respectable’ workers as a system for the relief of ‘paupers’ and administered by Local Authorities under the supervision of the Ministry of Health. The PA as a localized system was highly variable in its conduct. While some Labour controlled authorities defied governmental authority by refusing to implement the means test and granting the maximum payment to the unemployed, others retained the Poor Law tradition of less eligibility and deterrent conditions, including detention in workhouses and unpaid work.

The Public Assistance conference of 1933 affirmed paternalistic principles of administration:

relief is not a matter of right of the individual but a matter of duty on the part of the community; the poor law administrator is not a cashier paying on demand but a guardian acting in the interests, but not necessarily in accordance with the wishes, of his wards. (TNA, 1933, MH57/7 Public Assistance Conference. Draft of the minister’s speech).

The traditional administration of the poor law was based on a few fundamental principles: discretionary relief (i.e. no legally guaranteed right to relief), deterrence, and ‘less eligibility’. Deterrence meant that the conditions attached to the receipt of relief would be sufficiently unpleasant as to ensure only the genuinely desperate would apply. ‘Less eligibility’ was the principle that the relief on offer should provide a significantly worse standard of living than enjoyed by even the worse off ‘independent labourer’ (Burns, 1941; Fox-Piven and Cloward, 1972; Polanyi, 2001); in the present day, this is called a ‘work incentive’. Less eligibility can be seen in the means test and lack of right to any guaranteed scale of payment. Deterrence can be seen in the conditionality attached to PA. The 1930s was a time when authorities were attempting to rationalize the poor law to reorient it to the constructive purposes of employability: its ‘discipline should be that of the school rather than that of the prison’ (TNA, 1933, MH57/7 Public Assistance Conference. Draft of the minister’s speech). Policy makers sought to modernize practice and aimed for PA to become based on individual assessments of claimant’s employability needs (TNA, 1930, MH57/1, descriptive report of Health Minister’s Policy).

While workhouses continued to exist, a more significant form of deterrence in PA was the use of enforced work placements, Local Authorities were obliged after 1930 to set to work, training or instruction able bodied men in receipt of PA. The Ministry of Health reported that in December 1931 alone over 14,000 men had been sent to compulsory work or training (Ministry of Health, 1932). Files from the TUC show that this work was often hard manual labour (Warwick MRC Trades Union Congress (TUC), 1924–1933, responses to call for evidence on Test and Task Work of 3 August 1933). This work would attempt to arrest demoralization by subjecting workers to full working weeks of activity to maintain working discipline. This aim was summarized by a draft of the Minister of Health’s speech to the 1932 Public Assistance Conference which said that

[T]he Minister attaches great importance to this method of securing that the able-bodied do not merely remain able-bodied, in the full sense of physical and mental fitness, but should even improve his capacity as a potential worker. While any form of discipline is resented by the few, there is every reason to believe that the many prefer occupation to idleness. (TNA, 1933, MH57/7, Public Assistance Conference. Draft note of the minister’s speech)

Discussion: what can this tell us about ‘welfare reform’ today?

The study of liberal welfare measures in the 1930s provides lessons for understanding today’s ‘welfare reform’. The continuities identified between technologies of power employed in the pre-welfare state era undermine the innovative and modern picture of ‘21st century welfare’ painted by its advocates. For all that government promises ‘21st century welfare’, many current measures have a long history. These continuities show that the political rationalities (Brown, 2015) and pathologies of the UK’s governing liberalism have remained quite continuous between eras.

The article has found that in many ways, present-day regimes of intervention are more intrusive and punitive than those from the 1930s. This comparison should be disturbing, in that it shows the extremity of the current social policy regime. A particularly disturbing aspect of this is the trajectory of ‘reform’. As the 1930s progressed, benefits were becoming more inclusive and conditionality less severe, and there was a shift away from the highly stigmatizing regimes of behavioural assessment associated with the poor law. Eventually, these reforms were to contribute to the development of a new welfare regime in the UK. The opposite has been the case in the present day; since the 1980s, the system has become progressively more punitive and disciplinary.

There are many differences, not least a very different set of assumption about desirable labour markets which can be seen in several areas. We might contrast the critique of casual labour in the inter-war years with the rhetoric accompanying the launch of the government’s new benefit Universal Credit which pledges to be able to incentivize and to mandate that claimants take even very small amounts of work (Department of Work and Pensions, 2010b). In the 1930s, casual work was seen as a source of social dependency, insufficient to give men the steady habits which would enable them to fulfil their roles as the heads of self-sufficient families (Walters, 2000). Permanence and regularity were seen as guarantors of good character. There was a highly gendered ideal of male breadwinner families as normal and ideal, a view institutionalized in policy and which structured policy problematics. The view of insecure work in today’s policy is very different. It perhaps reflects the shift from employment security to employability security (Chertkovskya et al., 2013) amid an ethic of ‘projective’ work (Boltanski and Chiapello, 2007) and adaptable self-management (Worth, 2003). The valorization of work in modern policy discourse (Cole, 2008) contrasts to the increasingly exclusionary nature of contemporary employment (Bailey, 2016). Rather than representing an inefficient failure of industrial organization, today precarious work is seen as ‘flexible’ and a sign of a market-mediated efficiency within current regimes. Irregularity has been reconceptualized as a way by which the individual is expected to secure their independence through developing their capacity to adapt quickly to new roles, growing their human capital in each post in order to advance to the next (Boltanski and Chiapello, 2007). This new view of insecure work as a bulwark against dependency rather than its cause is reflected in extensive systems of subsidies such as the Tax Credit system and the variable rates of other benefits which all serve to facilitate low hours of work and which are now integrated into Universal Credit. Even the most extreme forms of labour market insecurity such as the Zero Hours contract, analogous to early 20th-century ‘casualism’, have been seen as acceptable by policy makers. This can be seen, for instance, in the Taylor report of 2017, which proposed only very minor changes and regarded ‘flexibility’ as a virtue of the modern UK economy.

In the main, the severity of current conditionality stands out as a contrast. Conditionality has become more and more severe and far reaching in recent years (Dwyer and Wright, 2014), whereas in the 1930s, behavioural conditions of the kind seen today were becoming less and less acceptable. The Genuinely Seeking Work test had been abolished amid mass protest and the evidence presented shows that there was little appetite for its reintroduction. In the 1930s, the burden lay upon the Employment Exchange to offer claimants work if they wished to stop paying benefits to them. Today under Universal Credit, claimants are

required to treat looking for work as their full-time job. . . . As a minimum therefore, claimants should be engaged in work search for at least the number of hours we expect them to be available for work. For many claimants this will be the equivalent of a full time job. (Department of Work and Pensions, 2011)

This ‘full time job’ has been codified as an expectation of 35 hours per week of job search. Claimants are obliged to report on this activity every 2 weeks and can as part of this requirement be directed to an ephemera of activities such as applying for specified vacancies, producing CVs, acquiring email addresses, setting up accounts on the government’s Universal Jobmatch website or attending talks and workshops, all under the threat of a ‘sanction’ for non-compliance. This activity is disciplinary in Foucault’s sense; the subject is rendered a visible and observable object by such a series of mechanisms of assessment which seek to identify deficiencies and bring the subject up to standard (Foucault, 1977). In the 1930s, this sort of activity was seen as burdensome and inefficient; today it has been reconceptualized as labour, virtuous in and of itself, but also as a moral obligation to ‘taxpayers’.

The sanctions introduced in October 2012 are the most severe to date. A ‘low level’ violation such as non or late attendance for a Jobcentre appointment (where job search is inspected, and benefits are paid out), or presenting what is judged to be an insufficient work search over the preceding period (recall the 35-hour standard) results in 4 weeks without benefit. A more serious ‘offence’ – such as failure to apply for a specified vacancy notified by the authorities – results in a 3-month sanction. The maximum available penalty ‘for repeated “high level” non-compliance is complete withdrawal of benefits for three years’ (Watts et al., 2014: 3). In the 1930s, disallowances were not as severe as they are today: 6-week disallowances were the maximum, and (again) the burden of proof was placed upon the authorities.

In the benefit system in the 1930s, many punitive measures existed which, like voluntary imprisonment in the Workhouse, do not have equivalents today. Unpaid work also has some different manifestations but is motivated and justified by similar disciplinary problematics. While today’s Mandatory Work Activity claimants were not subjected to hard manual labour as were the men claiming PA, the placements have been still been in less desirable sectors of the contemporary economy characterized by low pay and widespread insecurity. In the 1930s, these practices were reserved for the pauper and regarded as inappropriate for respectable insured claimants. Under the Work Programme, this unpaid work had a broader application, being applied to the ‘long term unemployed’, but also to lone parents and many claimants of disability benefits (Rafferty and Wiggan, 2017).

But this is not the whole picture: conditionality can be structured and differentiated. In the 1930s, there was the higher status of the insured contributor to consider, distinct from the Poor Law claimant, and subject to fewer of the behavioural tests demanded of either them or the Jobseekers Allowance or Universal Credit claimant today. Then, if one could demonstrate one’s worth through one’s contribution record, one could receive 6 months’ benefit as of a right. We might say that the ‘moral ordering’ (Boltanski and Threvenot, 2006) in use in the 1930s permitted a greater variety of statuses, from contributor to pauper. Today, the field is being flattened: the long-term decline of contributory benefit receipt is now joined with the ‘ubiquitous conditionality’ referred to by Dwyer and Wright (2014) erasing differences in treatment in favour of a universal escalation of disciplinary measures to push claimants towards the labour market. Policy has functioned to ‘remove the long-standing distinction between the respectable receipt of benefits for unemployed workers with contributions records and the more needs-based principle for other groups’ (Fletcher and Wright, 2017: 6). Social insurance is a disciplinary and exclusionary technology as well, which ‘pegs entitlements and benefits to employment, work performance, and contributions’ (Defert, 1991: 231), but it is one that implies a greater status for the contributor. In different periods of time, different criteria predominated for assessing the ‘worthiness’ of individuals (Boltanski and Threvenot, 2006). While both are liberal in their objectives, the means by which ideal subjectivities were to be achieved has changed. Today judgements of the virtue of the unemployed are made not on the basis of whether they can claim ‘membership of the insured class’ (Beveridge, 2015), but on the basis of behavioural assessments. In this flattened model stigmatization is increased, the destigmatized insured claimant having been abolished.

While an understanding of welfare benefits as disciplinary technologies of power is applicable in both periods, the he application of such technologies appears dependent upon the social status or ‘worthiness’ (Boltanski and Threvenot, 2006) of subjects and consequently the justifiability of intervention.

Conclusion

This article has shown that 1930s liberalism and current-day neo-liberalism formulate similar problematics and attempt to solve them in similar ways. Both moralize the economic and place an injunction on the unemployed to behave as ideal liberal subjects (Brown, 2015). A governmentality analysis of conditionality regimes as disciplinary technologies of power has allowed this article to identify similar methods used to discipline unemployed people in two eras. Changes in the accounts and justifications presented by policy makers of the problem of unemployment as well as the treatment of groups of unemployed can show larger changes. The ways in which these regimes differ show the advance of behaviourism over any system of status-based rights. What can be observed in recent years is, in effect, the restoration of the dominance of the Poor Law view of provision as a social evil to be minimized, and the corresponding exclusion of the Social Insurance view of provision as aid to an efficient labour market. The view of the unemployed as subjects has become less differentiated and cruder, and the ‘respectable’ unemployed have largely disappeared. This may suggest that after a social democratic interlude, UK social policy may be reverting to type.

Comparison with the 1930s provides a great deal of insight into the welfare system and the regulation of unemployment today. This article has made a contribution to current debates on the ‘reform’ of social security systems, by bringing in a historical perspective which problematizes current understandings of society and opens up new avenues for critique. In policy terms, this should contribute to criticism of the utility and justice of conditionality. It has shown that the current system not only has precedents in regressive institutions, but in many respects exceeds them in its punitivity.