Punitive benefit sanctions, welfare conditionality, and the social abuse of unemployed people in Britain: Transforming claimants into offenders?

A defining feature of U.K. welfare reform since 2010 has been the concerted move towards greater compulsion and sanctioning, which has been interpreted by some social policy scholars as punitive and cruel. In this article, we borrow concepts from criminology and sociology to develop new interpretations of welfare conditionality. Based on data from a major Economic and Social Research Council-funded qualitative longitudinal study (2014 – 2019), we document the suffering that unemployed claimants experienced because of harsh conditionality. We find that punitive welfare conditionality often caused symbolic and material suffering and sometimes had life-threatening effects. We argue that a wide range of suffering induced by welfare conditionality can be understood as ‘ social abuse ’ , including the demoralisation of the futile job-search treadwheel and the self-administered surveillance of the Universal Jobmatch panopticon. We identify a range of active claimant responses to state perpetrated harm, including acquiescence, adaptation, resistance, and disengagement. We conclude that punitive post-2010 unemployment correction can be seen as a reinvention of failed historic forms of punishment for offenders.


| INTRODUCTION
Punishment pervades the 21st century British social security system, which is designed to deter citizens from relying on the state (Department for Work and Pensions [DWP], 2011). Claimants are punished and controlled through poverty, conditionality, and surveillance (Fletcher, 2015). The "cruel, inhuman, and degrading" sanctions system (Adler, 2018) is the second harshest in the world (Immervoll & Knotz, 2018), with penalties of 100% removal of benefit income for up to 3 years. 1 Claimants who miss one Jobcentre appointment have their payments reduced or removed for 28 days. If they make the same mistake three times, the penalty is 91 days. Nearly a quarter of all jobseeker's allowance (JSA) claimants were sanctioned between 2010 and 2015, sometimes for trivial reasons (National Audit Office [NAO], 2016). The working-age social security budget has been slashed by £37 billion (Butler, 2018). The value of JSA, already low by international standards, was cutfirst by a 1% below-inflation limit to annual uprating (2013)(2014)(2015)(2016) and then by freezing benefit levels from 2016 (Dorey & Garnett, 2016). In 2018, JSA was worth only 12.8% of median wages (IPPR, 2019). Many unemployed people are also affected by wider cuts to housing benefit, the "bedroom tax" (spare room subsidy), and the household benefit cap. The benefit claim process has also become "digital by default," which erects a digital barrier that obstructs individuals' access to their entitlements (Alston, 2018). All of which has succeeded in instilling "fear and loathing" of the system in many claimants (Alston, 2018, p. 6).
Given the ineffectiveness of the British sanction system in securing job outcomes (Taulbut et al., 2018), recent studies have focused on the violent, rather than regulatory, nature of punitive welfare reforms (Burnett & Whyte, 2017;Grover, 2018;Pring, 2017). International theorisations of the punitive aspects of contemporary welfare reform have presented "prisonfare" and "workfare" as a "double regulation of the poor" (Wacquant, 2009) that is a unique product of neoliberal statecrafting. Similarly, recent interpretations of British welfare reform present "institutional violence" (Burnett & Whyte, 2017) and "social murder" (Grover, 2018) as characteristic of the "age of austerity." However, the British state has a long history of punishing unemployed people, particularly in times of economic crisis (apparent, e.g., in the 1930s labour camps, Fletcher, 2015;Fletcher & Wright, 2018).
Here, we explore how claimants experience punitive social security and focus on three key questions: what suffering is caused by welfare conditionality and sanctioning; what strategies have been deployed by claimants to minimise state perpetrated social harms; and how can this be conceptualised? First, we outline the qualitative longitudinal research design, methods, and sample. Second, we discuss conceptualisations of the regulation and punishment of unemployed people. Third, we outline the British policy context. Fourth, we identify claimant experiences of punitive British social security policies. Fifth, we present a range of claimant responses to punishment. Finally, we conclude with reflections on what this in-depth analysis of the punitive dimensions of welfare conditionality for unemployed people reveals as new or uniquely neo-liberal in character.

| METHODS
This article presents new evidence from the first major independent study of "Welfare conditionality: Sanctions, support and behaviour change" (Economic andSocial Research Council-funded, 2013-2019). The research aimed to explore the efficacy and ethicality of welfare conditionality in principle and practice and involved research teams at six U.K. universities (Glasgow, Sheffield Hallam, York, Heriot Watt, Salford, and Sheffield), investigating multiple policy fields. This article draws on data from the jobseeker stream of the research, consisting of three waves of qualitative longitudinal research (2015)(2016)(2017) with JSA claimants interviewed after periods of 6-9 months. Purposive nonrandom sampling techniques were used to recruit 65 claimants in Edinburgh, Glasgow, Bristol, London, Manchester, Peterborough, and Sheffield (see Wright & Stewart, 2016, for full first wave findings). Participant profiles at Wave A are detailed in Table 1, below. At Wave B, 43 remained in the sample reducing to 33 by Wave C. In total, 141 interviews were completed, with a retention rate of 49%. At the first interview, jobseekers were of working age: 9% were aged under 18, 15% were aged 18-24, 43% were aged 25-49, and 32% were aged 50-64.
At the time of the first interview, 95% of the sample were unemployed (one had recently moved into full-time work and two engaged in permitted part-time work of short or variable hours). At Wave B, more than two thirds of the sample were unemployed, and 12 interviewees (28%) were in paid work, of which eight were full-time, three were part-time, and one was self-employed. At Wave C, just over two thirds of the sample were unemployed, and 10 interviewees (30%) were in employment, of those six were full-time, two were part-time, and two worked variable hours. Four of those who had found work at Wave B were unemployed again at Wave C. These figures relate to snapshots of employment status at the time of interview and do not include experiences of movements in and out of work between waves. The employment status of those who left the study after the first or second interview is unknown; 43% of the sample were disabled at one or more waves, including some who had been transferred to JSA after failing a work capability assessment for the main disability benefit Employment and Support Allowance.
The volume and complexity of multisite and multiteam research design necessitated a highly systematic and structured approach to data management and analysis, using QSR NVivo software (Saldana, 2003). Researchers assigned attributes for transcripts of interviews they conducted. A two-tier approach to coding was used, with coding officers applying the first tier of framework matrix coding (Corden & Nice, 2007;Lewis, 2007) across the sample.
The matrix coding was assembled inductively by a working group of the project PI, a Co-I, and researchers drawn from a range of the institutional teams. The second tier of coding was conducted by the authors on the JSA subset. Key themes were identified from a close reading of a selection of transcripts, which were then coded across the sample. This was supplemented by text searches to verify the representativeness of findings and to identify data that did not fit the main trends.
3 | RE-CONCEPTUALISING UNEMPLOYMENT " CORRECTION" Social security systems, past and present across the globe, operate to regulate recipients (Byrne, 2005;Jones & Novak, 1999;Peck, 2001): Relief arrangements deal with disorder, not simply by giving aid to the displaced poor, but by granting it on condition that they behave in certain ways and, most important, on condition that they work.
Any institution that distributes the resources men and women depend upon for survival can readily exert control over them. (Piven & Cloward, 1971, p. 22) For Piven and Cloward (1971), this regulation expands and contracts cyclically in response to political attempts to quell civil uprising and serves capitalist economies by reinforcing work norms. Wacquant (2009) sees the disciplinary reforms of several welfare systems from the late 1990s onwards as a criminalising strategy that is uniquely neoliberal, involving a shift: from protective welfare, granted categorically as a matter of right, to corrective workfare, under which social assistance is made conditional upon submission to flexible employment and entails specific behavioural mandates. (Wacquant, 2012, p. 72) Gustafson (2009, p. 715) concludes that this "criminalization of poverty highlights economically and legally institutionalized ideologies of neo-liberalism, racism, sexism, and the dehumanization of the poor." Soss, Fording, and Schram (2011, p. 6) see neoliberalism and paternalism as joint forces that "have redefined poverty governance around a disciplinary agenda that emphasises self-mastery, wage work and uses of state authority to cultivate market relations," extending into business and civil society. Wacquant (2009) presents traditional forms of labour regulation as outmoded, whereas Soss et al. argue that the "old tools"-"barriers to welfare participation, low levels of aid, stigmatising rituals" -are still used and "have been augmented by changes in the basic goals and operations of welfare programmes" (p. 7). Anglophone countries have led the downgrading of social citizenship, replacing status with contract (Handler, 2003) and substituting entitlements for conditional rights (Dwyer, 2018;Dwyer & Wright, 2014).
The new justification rests on "a conception of fairness that is instilled in the notion of reciprocity", where "responsibilities and obligations counter-balance rights" (Paz-Fuchs, 2008, p. 1).
The increased use of tougher sanctions for widening claimant groups has been central to this international sweep of reforms (Immervoll & Knotz, 2018). U.S. evidence shows that sanctions trigger benefit exits, but many sanctioned benefit "leavers" do not receive other essential support they are entitled to and are less likely to subsequently thrive (Curtis, 2002). Sanctions can stimulate job entry to unsustainable low-quality poorly paid jobs, but people often return to benefits (cf. Fording, Schram, & Soss, 2013). Sanctioned claimants are more likely to be disadvantaged, for example, homeless (Reeve, 2017), than those who are not sanctioned and genuine barriers to employment, rather than resistance, often prevent compliance with requirements (Hasenfeld, Ghose, & Larson, 2004). Wheelock, Wald, & Shchukin, (2012, p. 1) show that in the United States, "racial attitudes seemingly link support for punitive approaches to opposition to welfare expenditure." In welfare settings ranging from work-first Englishspeaking systems to human-capital Nordic regimes, sanction rates are influenced by external pressures like targets and benchmarking and local organisational practices, with limited caseworker discretion at street level (ibid.; Caswell & Høybye-Mortensen, 2015;Lens, 2008). Lens (2006, p. 573) shows that "sanctions were based primarily on attendance records and became a paper-processing function"; interaction between claimants and advisers was "routinized and mechanical, resulting in improper and arbitrary sanctions." Since 2010, British welfare reform has followed a "criminalisation strategy, which is sensitive to central government control" (Fletcher & Wright, 2018, p. 12). This fits within, but brutalises, a longer term preference for punishing claimants instead of meeting their needs (Knepper, 2007). A major impoverishment offensive was launched by the U.K. coalition government from 2010 onwards. The "great British sanctions drive" was a central line of attack (NAO, 2016;Webster, 2017). From 2012, ordinary service sector occurrences, like missing an appointment, were criminalised using language of "offences," "transgressions," and "serial and deliberate breaches" to constitute a "huge secret penal system, rivalling in its severity the mainstream judicial system but without the latter's safeguards" (Webster, 2014). The sanctions system has impacted detrimentally on millions of benefit claimants (Adler, 2018, p. 51). This sharp punitive turn disrupted the United Kingdom's post-war welfare reform trajectory. The scale and depth of the resultant "avoidable diswelfares" (Grover, 2018) are difficult to interpret as anything other than the state acting against the interests of disadvantaged citizens. This reveals the fact, often overlooked in mainstream policy process literature, that "governments can be so incredibly inhumane" (Rummel, 1994, p. 26). Social security can involve state cruelty. However, there is no agreed conceptual framework or typology for identifying, classifying, or explaining non-lethal harm perpetrated by the state against citizens, even when it is large scale. Grover's (2018) Marxist conceptualisation of "violent proletarianisation" comes closest by identifying deaths caused by destructive policy decisions as "social murder." Similarly, Cooper and Whyte (2018, p. 2) develop the concept of "institutional violence," from Galtung's (1969) idea of "structural violence," to refer to the routine and detached administration of policies, implemented by public and private authorities, that produce acute physical and psychological violence.
This focus on violence helps to highlight the insidiousness of British welfare cuts and enables a reconceptualisation of government action as state crime (cf. Box, 1983;Green & Ward, 2004;Monaghan & Prideaux, 2016). Burnett and Whyte (2017) use it to show that workfare is a violent form of forced labour that also permits employers to breach health and safety laws with impunity, injuring and killing unpaid workers. However, the exclusive focus on violence works better as a metaphor than a concept when injuries are non-lethal, non-physical, or difficult to evidence. For example, is violence necessarily the most fitting description for "the lived experience of feeling humiliated, anxious and vilified" (Cooper & Whyte, 2017, p. 23)? The concept of state-inflicted social harm (Hillyard & Tombs, 2004) can usefully cast light on the role of unemployment "correction" as part of a re-emergence of mass "socially mediated harms" (Pemberton, 2015, p. 7).
Here, we are concerned with the consequences of the punitive turn for unemployed claimants. Suffering has been understood as a product of the way in "which people are painfully disposed to grasp the disciplinary force of powerful social structures upon their lives" (Wilkinson, 2012, p. 189). Bourdieu's (1999) concept of social suffering is useful for understanding the relative social dimensions of "positional suffering" (p. 4), occurring from occupying lower positions within hierarchical structures and shows how "inequalities are materialised in the body and lived" (McNay, 2012, p. 234). Being "inscribed on the body" (Frost & Hoggett 2008, p. 452) means there is a strong emotional dimension to the suffering. It encompasses not only material poverty but also the "lived experience of domination and repression, including feelings-humiliation, anger, despair, resentment" (Frost & Hoggett 2008, p. 439). We identify social security as a field where a variety of suffering has traditionally been experienced, within a highly differentiated society that has "multiplied the social spaces (specialized fields and subfields) and set up the conditions for unprecedented development of all kinds of ordinary suffering" (Bourdieu, 1999, p. 4). However, we suggest that the British sanctions regime has changed the mode of domination. In the post-war welfare system, much of the violence associated with unemployment benefits was symbolic-"the gentle, disguised form of violence" (Bourdieu, 1990, p. 133). However, post-2010, more direct forms of material violence have dominated.

| POLICY CONTEXT
The British system of support for unemployed people has a long and complex history of reform (Brenner, Peck, & Theodore, 2010;King, 1999;Whiteside, 1991). The roots of contemporary reforms lay in the perceived need to regulate the use of social security in the context of mass unemployment in the 1980s. The Conservative government's introduction of JSA in 1996 strengthened behavioural conditionality, surveillance, and some forms of disciplinary discretion. Furthermore, the "cumulative effect of serial restrictions on benefit eligibility, the tightening of 'availability for work' tests, and straightforward downward pressure on dole payments was to force hundreds of thousands of people off welfare" (Peck, 2001, p. 283). Many went into lowwaged work, others into various forms of economic inactivity, and some into the informal economy. The deterrent function of JSA was demonstrated by antecedents such as Project Work and Restart, which entailed mandatory job-related interviews.
The preoccupation with the size and cost of the welfare state was transformed by New Labour in the late 1990s into a concern with its incentive structure. Welfare "dependency" rather than poverty per se became the explicit focus of policy. Peck (2001, p, 333) argues that U.K. workfare has "emerged as one of the organizing frameworks for a putatively transformative 'reconstruction' of the welfare state." Labour's approach introduced work-related conditionality, especially from 2002, to new groups like lone parents and ill or disabled people via work focused interviews. However, Labour's welfare changes (1997-2010) also included new measures to "make work possible" (e.g., financial assistance with the high cost of childcare) and "make work pay," for example, the national minimum wage and tax credits to top up wages (Millar, 2002).
The Conservative-led U.K. coalition government drove through a much harsher approach from 2010, with "austerity" cuts followed by an abrupt extension and intensification of JSA sanctions in 2012 (Dwyer and Wright, 2014;DWP, 2010). Penalties shot up from a maximum of 28 weeks 2 to 100% reduction for 3 years (Webster, 2014). Open-ended sanctions were introduced, lasting for a discretionary period "until compliance." The tiered system has different penalties for "low" level "offences," such as non-attendance, for which the first-and secondtime penalty is 28 days, rising to 91 days on the third occasion. "Medium" level sanctions are for failing to be available for work. "High" level sanctions are for failing to apply for a job, ranging from 91 days for first and second occurrences to 1,095 days for a third. A single JSA claimant aged 25 years or over loses £300 for a 4-week sanction (NAO, 2016). Benefit sanctions differ from other financial penalties such as court fines in that they take immediate effect and individuals lose their only or main source of income causing disproportionate hardship (Adler, 2016).
From 2013, unemployed people began to claim the new working-age benefit-Universal Credit-alongside claimants who would previously have been eligible for dedicated disability or in-work payments (working tax credit). Universal Credit reinforced the decline of insurance-based benefits and began to operate as a catch-all social assistance scheme, which also incorporates housing and childcare payments. Part-time workers are expected to dedicate their non-working time to job search, up to 35 hours per week. This allows conditionality and harsh sanctions to be applied to claimants in a wide range of circumstances.
Policy reforms have been supported by reforms of the welfare delivery system-principally, the creation of Jobcentre Plus in 2001, serving unemployed claimants throughout England, Scotland, and Wales (Wiggan, 2007).
Regular mandatory contact with front-line staff was a core element of the new approach. The type of support provided by Jobcentre Plus has been severely constrained by very low levels of funding (Bonoli, 2010), the hegemony of work-first approaches, and the lasting influence of benefit off-flow targets, which influenced some staff to discourage people from claiming benefits (House of Commons Work and Pensions Committee, 2016). Jobcentre Plus reinforced a move away from training and public employment programmes towards low-cost measures for immediate job entry (Fletcher, 2000).

| FINDINGS
This section presents claimant experiences of punitive British social security policies, identifying Jobcentres as a site of social suffering in relation to sanctions, the treadwheel 3 of futile job search, and self-administered surveillance via the digital panopticon. Despite their weak social position, claimants are active agents who respond to punitive practices, often mirroring offender adaptations to prison. Jobcentre visits were experienced by most of our interviewees as "social suffering" (Bourdieu, 1999) and "ritualised humiliation" (Charlesworth, 2000, p. 81), where the "symbolic politics of myth and ceremony" (Handler, 2003, p. 235) were enacted punitively. Claimants usually regarded Jobcentres as benefits processing centres intent on compliance, lacking any real resources to help find work. They often felt humiliated, angry, despairing, and resentful.

|
They treat you like scum, like you're nothing, because obviously you're a benefit scrounger. (WSU-

PE-JM-023)
These symbolic dimensions of suffering relate in part to the well-documented social and psychological harms of unemployment. Coming to the Jobcentre means confronting a spoiled identity (Goffman, 1963) with feelings of shame and failure (Wright, 2003). However, what is new is that the symbolic violence (Bourdieu, 1990) is now inseparable from material domination. There is nothing "gentle" or "disguised" (ibid.) about the prospect of destitution.
Most participants were broadly supportive of the principles of conditionality and reciprocity. There was a general agreement that some form of sanction system was needed in principle to deter non-compliant or fraudulent "others." However, in practice, benefit recipients usually felt that the threat of sanctions was unnecessary or unjust in their own case. Half of the sample (33) had experienced a sanction, whereas half (32)  Anne (WSU-SH-JM-015), a 60-year-old former carer, who left work because she could no longer manage the strain, found being sanctioned "very harrowing." Jobcentre visits were so stressful that she often took a sleeping tablet the night before. Anne had been sanctioned for insufficient job search activity but had been pressurised to apply for jobs that she was not capable of doing: "They just seem to want you to apply for any job." She could not bring herself to tell her sister about the sanction, despite it being overturned on appeal. The appeal took several months and eroded her trust in work coaches. Anne felt angry and "a little bit ashamed that it's happened […] perhaps people with think that 'no, you haven't been doing what you should be doing'." The sanction had compounded her financial difficulties since: "there is no way you can manage without no money coming in." Instead of prompting claimants to find work, punitive conditionality created destructive conditions that could demotivate to the point where some lost all hope: I just got really upset, depressed and I gave up a bit and I was drinking a lot and I didn't care anymore. Here, in the context of the full set of transcripts from all three waves of the study, the language of "killing" is used more metaphorically. Although David did experience a range of harms, here, he is expressing deteriorative demoralisation. This type of long-term mental anguish can be understood as "social abuse," a form of "social harm" (Pemberton, 2015).

| On the treadwheel of perpetual futile job search: Relentless social suffering
Both the threat and application of sanctions caused social and material suffering that can be interpreted as social abuse. In addition to ensuring attendance at appointments, claimants reported that sanctions were threatened for insufficient job search, judged against the claimant commitment and discretionary work coach expectations.
Unemployed people are therefore disciplined by fear, towards perpetual job search motion, almost all of which is fruitless: You're not getting the support and help you need to get any work. You feel as if you're just going in a circle, wasting your time doing job searches.

AS-012)
For example, Ken (WSU-BR-AS-011), was typical of many older jobseekers we spoke to. He was in his early 50s and "desperate for a job," but despite his extensive experience in warehouse management and forklift driving, applying for three jobs every day did not result in any employment offers for a year. He relied on public transport, which made a long commute prohibitively expensive, but was sanctioned for 4 weeks for refusing to apply for a job approximately 100 miles away. He was "frightened" and felt that the sanction was unjust and profoundly undermining: I had no money, no food and it put me in debt with the risk of losing my abode [a council house].
Ken's experience of Jobcentre Plus was as an enforcement agency constantly "checking up on you." Because he was adamant, "I do not want to get sanctioned again," so I was applying for jobs that I was never going to get because you had to apply for jobs. This arduous and unrelenting futile activity can be interpreted as a reinvention of a historical form of physical punishment that operated in British prisons in the 19th century. The Victorian "treadwheel" was a punishment for convicts, introduced in 1818 and commonplace by the 1890s. Prisoners were forced to turn the steps of a "never-ending staircase" with their feet while gripping a bar to keep them upright (Ignatieff, 1977, p. 177). The gruelling labour was a daily routine for inmates, 7 a.m. until 5 p.m., who had to step 48-50 times a minute to turn the wheel and maintain perpetual motion. The sheer uselessness of the treadwheel-some ground corn but most did nothing but "grind the air"-was a salutary warning, and it was a humiliating punishment that was finally outlawed in 1902 for its excessive cruelty. In the 21st century, unemployment intervention reflects many of the same themes and rationales of the Poor Law era including deterrence, morality/social control, economy/efficiency, and reciprocity (Paz-Fuchs, 2008;c.f. Foucault, 1995). Furthermore, the British conditionality regime sentences unemployed benefit claimants to a labour in vain on a treadwheel of relentless futile job search, reinventing brutal Victorian penal technology towards the same ends-"grinding the air" is degrading social abuse intended as a deterrent.

| Universal Jobmatch as digital panopticon: Self-administered social suffering
During the period of our fieldwork (2014-2017), the job search treadwheel was self-administered online using the DWP job vacancy portal Universal Jobmatch, which was the main form of support available to benefit recipients in the United Kingdom. "Do it yourself" job search was usually combined only with cursory work coach contact, who where the observed (originally prison inmates 5 ) must never know whether he is being looked at any one moment; but he must be sure that he may always be so […] totally seen without ever seeing.
For Foucault, panopticism is a "modality of power" and a technique of coercion. Similarly, Universal Jobmatch panopticon is used as a corrective tool "a way of defining power relations in terms of the everyday lives of men [sic.]" (ibid., n.p.). Its self-administered design is also recognisably panopticon: He who is subjected to a field of visibility; and who knows it, assumes responsibility for the constraints of power; he makes them play spontaneously upon himself; he inscribes in himself the power relation in which he simultaneously plays both roles; he becomes the principle of his own subjection. (Foucault, 1975, n.p.) Many interviewees questioned the efficacy of Universal Jobmatch, reporting duplication of other better online vacancy sites, out of date vacancies, limited capability, and inoperable clunky interfaces between the parts of the system designed for job applications and the parts designed for DWP monitoring and work coach interaction (e.g., online job searches on other sites had to be logged manually). These design flaws and administrative inadequacies meant that using Universal Jobmatch was often associated with a sense of futility that required claimants to behave in certain ways despite those behaviours offering little prospect of desirable outcomes: You're applying for the same jobs.
[…] But the system isn't up-to-date, so people are applying for jobs that don't exist.
[…] Universal Jobmatch isn't keeping up with what real jobs are there. (WSU-BR-AS-014) The digital system also redefines looking for work as an online activity, delegitimising face-to-face and paper-based forms of job search (although a minority of interviewees were required to complete paper and online activity logs, which created duplication and confusion). Moreover, this reliance on digital systems made it harder for some individuals to provide the necessary evidence of job search: Everything is on computers now. If you're not looking for work on the computer it doesn't go through onto the internet, you're not looking for work are you? (WSU-BR-AS-010)

| Claimant responses to punitive conditionality: Mirroring adaptations to imprisonment
Punitive welfare conditionality, like work-first activation more generally, is designed on the false assumption that unemployed claimants are naturally inactive either because of their perceived incompetency or immorality (Wright, 2016). However, participants in this study were neither passive nor helpless. Positioning the state as the perpetrator of harm against disadvantaged citizens does not, however, mean that claimants can simply be understood as victims.
Those we spoke to were active in seeking solutions to the challenges that confronted them. Claimants, despite their profoundly weak social position, responded to punitive social security policy by acquiescing, adapting, resisting, and disengaging. These responses "create small and necessary spaces of personal control and autonomy" (Gilliom, 2001, p. 7) and reflect the ways in which offenders adapt to incarceration.

| Acquiescence
Most interviewees complied with the demands made by Jobcentre staff: "You mostly just did what they told you to do" (WSU-PE-JM-023). When Jenny found work as a care assistant, she struggled to follow the logic of being told to keep applying for jobs until her pre-employment checks were complete. Nevertheless, she complied: "I've paid money to go to these job interviews, even like train journeys. And that made me skint as well" (WSU-PE-JM-023).
Another young man, with a background in retail, exceeded his job search target: "you need to apply for like 3 jobs a week […] I look for 6 or 7 every week" (WSU-PE-JM-025). Similarly, many indicated that they acquiesced in the imposition of their claimant commitment.
More generally, interviewees commonly regarded Jobcentres as being primarily focused on ensuring compliance with the mandatory benefit claim conditions. Pressure to achieve more demanding job search requirements coupled with recipients' strong desire to avoid the punitive effects of a sanction resulted in people applying for jobs they had no realistic chance of getting. Intensifying welfare conditionality therefore encouraged a culture of counterproductive compliance, which had little impact on moving people closer to the labour market.

| Adaptation
The criminalisation of claimants meant that compliance was often grudging or tokenistic. Behavioural changes mirrored those made by individuals entering the prison system (see Haney, 2012). This was evident in three key adapta- Second, prisoners learn to control and suppress their own internal emotions and reactions to events around them. Emotional over-control and a generalised lack of spontaneity may result in the development of a "prison mask" (Crew et al., 2014;Haney, 2012;Jewkes, 2005) that is unrevealing and impenetrable. Similarly, emotional overcontrol sometimes characterised claimant behaviour. Many interviewees were acutely aware of the danger of revealing too much about their personal circumstances or emotional state to work coaches fearing that this might lead to benefit sanctions, which undermines trust and further limits the possibilities of support. The need to suppress emotions was most apparent in situations where claimants felt that they were being deliberately provoked by work coaches in order to justify sanctioning activity.

Resistance
A minority of interviewees resisted surveillance and punishment. Some lied about the extent of their job search activity. For example, one lone parent in her late 40s admitted: "I'm not actually applying for many jobs. I mean I do say I am but realistically I'm not" (WSU-ED-BW-023). Active resistance was more common among the minority who were shielded from the financial consequences. For example, one former construction worker admitted: "I did deserve it. I was doing nothing to find a job… It [sanction] did not matter because I was making money [from drug dealing]" (WSU-BR-PD-005a). He explained: "the only reason I'm signing-on is so there is a paper trail to be perfectly honest with you." Young men, especially those living at home, were another such group. One 25-year-old man (WSU-LO-BW-009) was, for example, living with his parents in East London at the time of interview. He had had several jobs since leaving school including warehousing, retail, and labouring in the construction industry: "I've done all sorts of work but it's never been permanent work." Consequently, he had made several separate claims for JSA. He had been sanctioned on "three or four" separate occasions during his time claiming JSA. The most recent sanction was reported to be due to insufficient job search activity, which he felt was unfair. "I thought I'd done everything that they wanted me to do. It was just my advisor's eyes at the appointment who thought I wasn't doing enough." Despite this, he had not appealed the sanction because "I cannot be bothered." When asked about the consequences of the last sanction he replied: "Parents helped me out but then that caused friction between me and my parents because I was supposed to be giving them housekeeping. They're actually keeping me for the next four week [sanctioning period] and it was hard." Nevertheless, he was unconcerned about receiving further sanctions reporting that the experience had "just given me more hate for the Jobcentre."

Disengagement
Many interviewees thought that a primary purpose of welfare reform was to facilitate disengagement: "I think it is just to make it as unbearable as possible so you do not want to be on it anymore" (WSU-PE-JM-023). A combination of low benefit levels, increased surveillance, and the sanctioning of claimants led some to disengage from the benefit system: "I stopped signing on JSA for a while because I got fed up with it" (WSU-BR-PD-005). This was largely due to the poor relationship that this former construction worker had with his work coach, which had led to a sanction for a lack of job search. "I found it was down to your personal adviser as how hard it was for you…. I would turn up and I'd applied for two or three jobs and they'd be really on my case." Interviewees highlighted several strategies to cope with the loss of benefits. Most borrowed money from family and friends, visited food banks, or engaged in informal work. A few resorted to survival crime.
I got arrested for shoplifting because I was so hungry I needed to feed myself. (WSU-PE-JM-002)

| CONCLUSION
In this article, we have demonstrated that unemployed interviewees experienced claiming benefits and using employment services in Britain as predominantly punitive and criminalising in character. There has been a fundamental shift from providing help with accessing jobs and careers to monitoring compliance with behavioural rules enforced by sanctions. Claimants felt criminalised by harsh sanctions and related conditionality tools, which dispensed a great deal of social and material suffering. Symbolic violence (Bourdieu, 1990) was already widespread, but the post-2010 punitive regime established the prominence of overt material domination, in the form of realistic fear of deep poverty, unmanageable debt, loss of home, and destitution. Welfare reforms have made the process of claiming benefits socially abusive. Benefit sanctions sometimes caused harm so extreme and intense that it could be described as "social murder" (Grover, 2018) because it threatened life. There were also examples of physical harm that warranted description as "institutional violence" (Cooper & Whyte, 2018). However, for most interviewees, over the course of three waves of research, the suffering incurred because of the application and threat of sanctions and the conditional system of welfare reform can be described as social abuse, a form of social harm. Importantly, it is the U.K. government that is the perpetrator of this harm.
We likened the punitive nature of relentlessness mandatory futile job search to the Victorian treadwheel, which was used in prisons as a cruel punishment. Like convicts on the treadwheel, benefit claimants must spend full weeks "grinding the air," in useless perpetual motion. This humiliating ritual of labour in vain also serves to deter others from subjecting themselves to it. Relatedly, we see the online job vacancy portal, Universal Jobmatch, as a digital panopticon that uses technology to self-facilitate social abuse, particularly for those disadvantaged in the labour market, because it is identifiably punitive and enmeshed with the sanctions system, which operates through fear. These insights were based on claimants' accounts of their lived experiences of sanctions and support to a developing field of critical analysis by applying concepts from criminology, social policy, and sociology to show how the British state punishes unemployed people and perpetuates social harm (Pemberton, 2015).
Nevertheless, a key problem with conceptualisations, which stress the violence of welfare reform, is that they can be one dimensional and ignore the agency of those that are being harmed. Claimants are not simply passive victims of "institutional violence" and do not meekly accept their "social murder." Despite their weak social position, interviewees responded to state perpetrated harms via acquiescence, behavioural adaptations, resistance, and disengagement. We argue that as Jobcentres have become more dangerous places, where individuals are at increased risk of losing their only source of income and being pushed into deep poverty, behavioural responses have often mirrored the adaptations made by the incarcerated. The irony is that this puts some at an increased risk of being punished by the criminal justice system. Consequently, punitive welfare conditionality has the potential to result in state perpetrated harms to some of its citizens in both the welfare and criminal justice systems. Although analysts have tended to portray punitive unemployment correction as a feature of modernity, neo-liberalism (Wacquant, 2009), or the "age of austerity" (Cooper & Whyte, 2017;Grover, 2018), we conclude that social and material suffering is a long-standing feature of the British approach. Post-2010 sanctions-based welfare conditionality reinvents failed forms of 18th and 19th century punishment and surveillance that the imperial British state used to govern through brutality and fear.