Abstract
This article contributes to research on activation practices and inequality reproduction, analysing how class is done in Swedish municipal subsidised employment (SE) programmes. Drawing on focus groups and interviews with 45 participants, 41 social workers and four managers in six programmes, the study uses Acker’s inequality regimes theory and Offe’s distinction of class categories in the labour market. Findings reveal that the SE programmes do class by sorting and adapting participants into low-wage, precarious jobs and enforcing labour market conformity. Constructions of ethnicity and gender also shape how class is done, reflecting broader patterns of ethicised and gendered labour market segregation. Subsidised employment functions as a cost-saving measure for municipalities, reducing social assistance costs by creating a flexible, low-cost workforce with limited protections. The “municipal reserve army” supports labour market competition and wage suppression, underscoring the role of such programmes in sustaining class inequalities.
Introduction
Activation in social policy refers to policies mandating labour market participation as a condition for receiving benefits and reinforces the notion that societal worth centres on one's position (or lack thereof) in the labour market (Eichhorst et al., 2008). Studies on activation through labour market programmes consistently reveal how individuals in poverty due to unemployment face marginalisation and stigmatisation whilst being kept active, disciplined and prepared for labour force entry (Govender, 2023; van Berkel et al., 2017).
In the last decade, research on labour market programmes has begun to pay attention to how organising and practicing activation is entangled with doing inequalities based on gender and/or ethnicity (Hansen, 2018; Ullman, 2023). Previous research has shown how labour market programmes are gender-biased, as unemployed participants are offered courses and work placements in line with traditional gendered norms (Betzelt, 2008; Parsland and Ulmestig, 2022). At the intersection of gender and ethnicity, research has shown how labour market programmes reproduce refugee women as barely employable mothers and spouses, whereas the aim of the labour market programmes is to make refugee women employable (Ullman, 2023). Previous research indicates the need to study the role of labour market programmes as an extension of the welfare state in reproducing these inequalities. Yet, the role of labour market programmes in reproducing class is seldom explicitly discussed or studied. There are a few exceptions. Byrne (2005), Burrows and O’Brien (2016) and Prendergast (2019) discuss how activation aims to keep the underclass prepared for a flexible labour market while discouraging unemployment. However, questions remain regarding how activation through labour market programmes are part of doing class.
In this article, similar to other categorisations, class is recognised as being done within institutionalised settings, such as in the school system, labour market, and social services (Offe, 1996). Within organisations, such as the authorities running labour market programmes, interaction with staff can either make individuals unsuitable for employment or make them suitable for specific tasks in line with labour market needs (cf. Harrits, 2019). Hence, organisations have a vital role in reproducing classed categorisations that have an impact on individuals’ access and power over resources necessary for survival, for instance income. Acker (2006) emphasises that class inequalities are rarely reproduced alone within settings such as labour market programmes, but intersect with other inequalities, such as ethnicised and/or gendered inequalities, creating ‘inequality regimes’ within organisations. Thus, addressing class requires considering intersections of ethnicised and gendered categorisations, differentiating how unemployed individuals are activated in labour market programmes. This article seeks to contribute to the growing literature on the relationship between activation practices and the reproduction of inequalities.
The specific activation practice that has been studied for this article is subsidised employment (SE), organised through Swedish municipal labour market programmes (SE programmes). SE is widely practiced globally (Card et al., 2018; Mörk et al., 2021) and involves fixed-term employment with a set taxable pay (Forslund et al., 2019). In Sweden, SE that is provided through municipal labour market programmes typically targets unemployed individuals receiving social assistance. Examining SE through an analytical lens of doing class is particularly intriguing. SE as an activation intervention allows for a temporary shift in the access and control over the resources that are necessary for survival, to the extent that unemployed individuals receiving social assistance become (temporarily) employed, get access to the social insurance system. In that sense, SE can be understood as a tool for decommodification (Hultqvist and Hollertz, 2021). However, studies indicate that predominantly unemployed individuals with employable qualities, i.e. skills that are valued by the labour market, are selected for SE (Forslund et al., 2019; Gubrium et al., 2017). Job creation schemes, such as SE, have also been criticised for pushing individuals living in poverty towards low-paid, insecure jobs, which benefits the reproduction of labour power and the welfare state, but not the individual (Wiggan, 2015). SE programmes are subsequently interesting to study in relation to how class is done.
The aim of this article is to contribute with knowledge on the reproduction of classed inequalities through welfare organisations, by analysing how class is done in municipal labour market programmes using subsidised employment as the main intervention.
The empirical setting is two Swedish municipalities, and the empirical material encompasses interviews from six municipal labour market programmes in which SE is the primary intervention, comprising a total of 45 SE participants. 41 social workers and 4 managers employed within the SE programmes. The empirical material is analysed through a theoretical framework that combines a macro-level analysis of class stratification with a meso-level focus on how class is done in intersection with ethnicity and gender within organisations.
Background
The Swedish welfare system has long been viewed as a pioneer of decommodification, reducing workers’ dependence on selling their labour through a strong welfare infrastructure (Esping-Andersen, 1990). Social insurance covering sickness, old age, parental leave, and unemployment helped counteract proletarianisation (cf. Offe, 1984). Since the 1980s, however, welfare retrenchment and activation policies have strengthened work discipline globally, including in Sweden (Grover, 2019). Stricter eligibility criteria and reduced benefits, especially for those in poverty, have been conceptualised as re-commodification (Grover, 2019; Offe, 1984). Shaped by market liberalism and labour market flexibility (Sennett, 1998), these reforms have increased precarity, particularly among foreign-born workers and women, who are overrepresented in insecure, low-paid jobs (Ahrne et al., 2021).
Sweden's welfare system operates through a dual structure: the Public Employment Services (PES) manage national benefits, while municipalities administer social assistance for individuals outside PES coverage. Youths, foreign-born individuals, and people with health issues are overrepresented in the municipal system (Forslund et al., 2019). Access to assistance typically requires participation in municipal labour market programmes, which differ from PES-led schemes through more sanctions, reduced rights, and means-tested benefits (Govender, 2023). Around 20% of Sweden's unemployed take part in such programmes (Forslund et al., 2019). In 2020, municipalities spent about 500 million Euros on them, employing 5200 social workers and involving around 100,000 participants, focusing mainly on training, internships, matching, and subsidised employment (SALAR, 2021).
Municipalities are both welfare providers and public-sector employers in areas such as education, elderly care, childcare, and landscaping (Jacobsson et al., 2017). These jobs are often precarious, low-paid, and offer limited prospects for long-term employment or unemployment benefit eligibility (cf. Mörk et al., 2021). This dual role remains underexamined.
Subsidised employment (SE) has been central to Swedish active labour market policy since the 1980s, administered by PES through schemes for different unemployed groups. Subsidy levels vary, covering partial or full employment and supervision costs (Forslund, 2018). SE expands in downturns (cf. Sjögren and Vikström, 2015) but is also used in stable periods, consistent with activation logics stressing individual responsibility. Since 2010, PES has facilitated SE for about 130,000 people annually (Engdahl and Forslund, 2019). While official data on PES–municipality cooperation are lacking, around 40,000 SE positions have been municipal, making SE the third most common municipal activation measure (Forslund et al., 2019).
SE has predominantly been studied through effect evaluations. Quantitative studies suggest SE facilitates labour market entry more effectively than many activation measures (Card et al., 2018; Mörk et al., 2021; SBU, 2022). Critics warn of job displacement, wage suppression, and the exclusion of those furthest from employment in favour of those closer to it. Qualitative studies indicate SE can provide a sense of recognition in work, even though participants remain misrecognised in wider society due to the low-paid, low-qualified nature of their jobs (Govender, 2023; Hultqvist and Hollertz, 2021).
This article examines six SE programmes, each targeting different unemployed groups, frequently including social assistance claimants. The programmes varied in tasks, duration, and additional activation measures, see Table 1.
Overview of SE programmes.
All programmes operated under national and local regulations, as well as collective agreements with trade unions. Participants generally lacked full labour rights, such as negotiating wages or securing long-term employment. Some SE roles excluded access to unemployment benefits (PES, 2018), though participants received vacation and occupational pensions. Each programme was staffed with front-line managers and social workers, who led teams and supported participants through job coaching and job-seeking activities.
The PES wage subsidies, ranging from partial to full, were central to the financing structure of these programmes. Municipal budgets determined how many subsidised positions could be funded annually. By using subsidies from PES, municipalities could afford to include more participants overall. To illustrate, a municipality could have funding for 50 participants. By applying for subsidies from PES, additional participants could be accepted to the programmes. If 25 participants were fully subsidised by PES, then the municipality could employ 75 individuals instead of 50. In this way, PES funding enabled more people to exit the social assistance system at a lower municipal cost—illustrating municipalities’ dual function as welfare providers and employers.
Analysing the doing of class, a theoretical framework
To contribute to knowledge on reproduction of class inequalities through welfare organisations and analysing how class is done in the six SE programmes, the theoretical framework combines a macro-level analysis of class stratification with a meso-level focus on how class is done in intersection with ethnicity and gender within organisations. The theoretical framework allows for an exploration of how class is reproduced through institutional knowledge claims both implicitly in society and explicitly within organisational contexts, where the six municipal SE programmes serve as empirical examples.
On a macro level, I draw on Claus Offe's (1996) typology of ‘winners’, ‘losers’, and the ‘superfluous’ in the labour market to conceptualise class as a categorisation grounded in institutional knowledge claims of employability. Offe (1996) moves away from traditional class labels, instead defining class through types of labour market attachment. This shift foregrounds employment security and access to rights as key class indicators. Class is thus understood as systemic differences in access to resources necessary for survival (Acker, 2006).
‘Winners’ are regularly employed individuals whose skills are in demand by the labour market. This analysis, however, focuses on the distinction between ‘losers’ and ‘superfluous’. ‘Losers’ are those who are unemployed or in precarious employment but still seen as potentially employable. The ‘superfluous’ are regarded as non-employable—unable to adapt to dominant social norms or lacking the competencies required by the labour market.
Offe (1996) identifies five institutionalised criteria that shape these distinctions:
Employed or employable qualities (skills valued by the labour market) Legality (e.g. residency status and/or criminal record) Intact or recoverable health (mental, physical, and freedom from substance abuse) A socially well-organised life (e.g. stable housing and support networks) Linguistic and cultural competence and ethnic affinity (alignment with dominant norms)
These criteria are not neutral. They reflect institutional knowledge claims about productivity and worth, functioning as mechanisms that legitimise classed divisions within welfare systems. According to Offe (1996) individuals become categorised into one of the three categories when encountering welfare organisations such as schools, social services or labour market programmes since the organisations are soaked in institutional norms (see also Grover and Piggott, 2010).
Offe's (1996) typology resonates with Guy Standing's (2011) ‘precariat’, though Offe (1996) is primarily concerned with categorisation rather than defining a new class. His distinctions are perceived as a development of Marx's (1976) reserve army of labour, where ‘losers’ form a potential labour force, while the ‘superfluous’ resemble the lumpenproletariat—those outside productive work (Gubrium et al., 2017). Unemployment has become increasingly entrenched, and even those who meet employability criteria may be categorised as ‘superfluous’. This trend is reinforced by activation policies that discipline the unemployed through intensified requirements and conditional support.
While Offe (1996) offers a framework for who occupies certain classed positions in relation to institutional norms, he does not address how these categories are done within organisations. To explore this, I turn to Joan Acker's (2006) theory of inequality regimes, which enables a meso-level analysis of how class is done within SE programmes.
Class is done through everyday practices in which individuals are categorised in ways that determine their access to resources. In this study, this refers to how unemployed people are positioned as ‘losers’ or ‘superfluous’ in line with Offe's (1996) categories. As Acker (2006: 7) argues, gender and ethnicity are “integral to the creation and recreation of class inequalities and class divisions, emerging in complex, multifaceted, boundary-spanning capitalist activities.” While the doing of class is the main focus, this analysis also attends to how gendered and ethnicised inequalities intersect with it.
Acker (2006) emphasises that organisations reproduce broader social inequalities through their internal processes. Inequality regimes, as she defines them, consist of six interconnected components:
Bases of inequality Organising processes that maintain inequality Visibility of inequality Legitimacy of inequality Control and compliance Competing interests and the possibility of change
These components help explain how classed categorisations are done within SE programmes. For example, if inequality is invisible or seen as legitimate, it is more difficult to contest. Organisational practices such as eligibility assessments, behavioural monitoring, or informal norms around employability reflect and reproduce these regimes.
In summary, this framework combines Offe's (1996) macro-level categories ‘winners’, ‘losers’, and the ‘superfluous’ with Acker's (2006) meso-level theory of inequality regimes to analyse how class is done within municipal SE programmes. Together, these approaches show how institutional norms and organisational processes co-produce classed positions. In the concluding discussion, I return to the concept of the reserve army (Marx, 1976) to highlight how activation policies, including subsidised employment, contribute to class reproduction, a dynamic still underexplored in existing research.
Method
The empirical material presented here forms part of a broader study funded by Forte and approved by the Regional Ethics Review Board in Uppsala (Reg. no. 2016/173). Employing an abductive research strategy, the study began inductively and developed through iterative engagement with empirical material, theoretical concepts, and prior research, identifying underlying patterns that informed the central findings (Alvesson and Sköldberg, 2018).
The material consists of semi-structured focus group and individual interviews with 45 participants—41 social workers and 4 front-line managers—from six municipal SE programmes (see Table 2). Participants varied in terms of work experience, education, country of origin, age, and motivation to participate. Interviewees held diverse roles but are collectively referred to as social workers for confidentiality. All interviews were conducted by the author and informed consent was obtained.
Overview of respondent categories/SE programmes/interview type.
As Table 2 shows, focus group interviews constituted the primary method, chosen to facilitate discussion among respondents (Kruger et al., 2019). While similar to individual interviews in terms of content, focus groups allow for the exploration of shared and divergent experiences, revealing additional empirical layers. However, participants in Programme T and all managers (Programmes W–Z) were interviewed individually. The choice to interview participants in Programme T individually was made at their request, for convenience. Managers were interviewed separately to allow social workers greater freedom in expressing their views. Thus, the methodological design can be seen as pragmatic, shaped by respondent preferences.
This combination also enabled the collection of both individual experiences and context-dependent discussions which nuanced the analysis in how class was done (Lambert and Loiselle, 2008). Interviews were transcribed verbatim and translated into English for this article.
Analysis
Thematic analysis was applied and Bazeley's (2009) three-step model supported structure to the analysis: describing themes, comparing and reorganising empirical material, and relating themes to previous research and theoretical concepts. During the analysis, I also noted whether the empirical data was based on individual or focus group interviews in order to get a combination of individual experiences and contextualisation in the themes (cf. Lambert and Loiselle, 2008). Initial themes, drawn from Offe's (1996) differentiation between the ‘superfluous’ and ‘losers’ of the labour market, included employable qualities, legality, mental or physical health, a socially well-organised life, linguistic and cultural competence, and ethnic affinity. The concept of inequality regimes (Acker, 2006) was utilised in the final step to draw conclusions based on the analysis in view of how the classed categorisations of ‘losers’ and ‘superfluous’ was done in the SE programmes. I connected the categorisations with thematised processes within the SE programmes and noted the visibility and legitimacy of inequality and control and compliance for inequality. At this stage, a final thematisation was constructed of three overarching themes, based on the organisational processes that were analytically noted in the six labour market programmes; (I) sorting, (II) adaption and (III) profit, which together provide empirical support for how class was done.
Findings
Sorting
For unemployed individuals to be eligible for subsidised employment in the municipal SE programmes, they had to meet certain criteria, which relate to the criteria listed by Offe (1996) in the distinction between ‘superfluous’ and ‘losers’, which is understood here as a sorting process conducted by the social workers in the six labour market programmes. The sorting process was based on the target groups of unemployed people at whom the programmes were aimed, as described in the background section. This type of sorting process is common within welfare organisations (see Grover and Piggott, 2010). In the table below, I show the criteria corresponding to the sorting process. The table is based on the empirical material from the six labour market programmes when potential participants were discussed in the interviews with the social workers and managers.
The table shows the sorting process corresponding with Offe's (1996) criteria. As shown in the table, not all criteria were corresponding in all the programmes, which does not implicate that the criteria were not relevant to the programmes. Rather the lack of criteria reflects how other criteria could be more important for the programmes and were thus highlighted in the interviews Table 3.
Offe's (1996) criteria between ‘superfluous’ and ‘losers’ in the labour market corresponding to the sorting process in the six labour market programmes.
As the table shows, being employable, having a recoverable mental and/or physical health and linguistic and cultural competence were criteria being used in assessing unemployed individuals for potential participation in all the programmes. For example, unemployed individuals were considered ineligible if they did not have intact or recoverable mental or physical health and/or did not have a socially well-organised life, for example if they had experienced homelessness. As one social worker said: “Having proper accommodation, that's very important. And being drug-free” (SW4:Z). Like Grover and Piggott (2010), I interpret the sorting processes described in this article as being driven by social and economic factors, based on the cost of supporting unemployed individuals in relation to moral beliefs about the causes of their unemployment. The social and economic factors were especially clear in relation to employability. Employability was as one of the most prominent criteria mentioned in all six SE programmes and was mainly presented as motivation. Motivation is connected to moralising public assumptions about unemployed individuals being lazy and unmotivated to work (Prendergast, 2020) and can therefore be understood as an ‘employable quality’ (cf. Offe, 1996). Thus, motivation, and especially motivation to be active, is a class marker, in so forth that motivation clarifies the border between unemployed categorised as ‘superfluous’ and ‘losers’ in the labour market. In SE programme Y, the social workers conducted interviews with potential participants and afterwards labelled them, mainly depending on whether they showed motivation. SW1: We have a level division with a colour scale; green, yellow, red. Green is ready right now. We remove red for various reasons. Language weakness, health reasons, other reasons, you have some reason that makes you not fit at all. Yellow is reserve we can call them too. So they can join if we have enough places if we have the finances for it. SW4: Then if you look at the people, [participants], who are language-impaired who perhaps belong in the yellow group but still there can be a discussion. SW5: The most important thing is that the person is motivated and wants this job. (SW1, 4 & 5:Y)
The quote illustrates how motivation as an ‘employable quality’ plays a crucial role in the sorting of potential participants. Being labelled as ‘green’ indicates readiness to work and showing motivation. Motivation is deeply embedded in the work ethic underpinned by welfare states which are part of flexible capitalist systems, valuing individual effort to participate in the labour market (Sunnerfjell, 2023). By looking for motivation, potential participants are thus categorised on the basis of a class marker as either ‘superfluous’ or a ‘loser’ in the labour market, i.e. the categorisation entails that individuals who lack motivation are not needed in the labour market. In the quote above, Swedish language proficiency also emerges as a decisive factor in this sorting process, introducing an ethnicised marker in the class-based sorting process. Referencing language skills was common in all six labour market programmes when sorting participants. Often without formal requirements but rather guided by ‘[…] a sense of being able to do the job’ (SW4:W), if language skills were assessed as enough for participation in the SE programmes. Previous research (Parsland, 2023; Ullman, 2023) emphasises the centrality of language skills in reproducing ethnicised inequalities. Consequently, class and ethnicity intersect (Acker, 2006) in the sorting process, requiring foreign-born participants to demonstrate both motivation to work and to improve Swedish skills in order to be accepted.
For the participants, continuing to show motivation to work within the SE programmes also meant ‘having a roof over my head and food for the day’ (P1:T). Hence, finding and expressing motivation meant that the participants could avoid going back to the stigmatising social assistance system and being assessed as superfluous in the labour market. However, some of the participants expressed motivation to be in the SE programme and work, based on the hope that it would lead to future employment: “[w]e got a chance here in the programme. That's something, they can help us get back into a routine. How to work, how to think about, think outside the box.” (P5:U). Acker (2006) describes expressions of ‘motivation’ or ‘hope’ from individuals in marginalised positions as an internalised control that maintains inequality regimes. The internalised control can take the form of giving meaning to governing structures, such as activation being a support towards regular employment, but also serves as a check on their own behaviour (Acker, 2006). The process of sorting unemployed individuals highlights how the SE programmes do class by using criteria that align with the classed categorisations of being ‘superfluous’ or a ‘loser’ in the labour market, which can put pressure on participants to show motivation to work and pushes unemployed individuals who lack the criteria back to the stigmatising social assistance system.
Adaption
The second thematised process is adaption, which shows how the SE programmes did class by adapting the participants towards the lower segments of the labour market, even though participants had a varied background. Some participants had experience of white-collar jobs and/or university studies before migrating to Sweden. For instance, the participants in programme U shared their past labour experience in regards of one being a maths teacher, another a carpenter, a third a white-collar employee within the bank system. What they have in common is that they must adapt their previous labour market experience to the lower segments of the Swedish labour market. In the focus group interview, this caused frustration among the participants: For example, she [P2], she has been working a lot of years, she has work experience, she has worked a lot with the bank. Why not utilise [the experience] or try to teach the community here? […] After fourteen years she has been here in Sweden, she has lived here in Sweden. But when there is no one who can utilise her experience, her work experience and learn then it is no, no use at all. (P5:U)
The interpreted frustration exemplified through the quote above visualise the inequality in the Swedish labour market for people with a migrant background, when their previous work experience is not recognised but rather made invisible (cf. Mulinari, 2024). Nevertheless, the quote also provides an insight into the task of municipal labour market programmes, as the programmes don’t activate participants in relation to their previous work experience but rather in relation to the needs of the lower segments of the labour market (cf. Grover, 2019).
In order to adapt the participants towards the lower segements of the labour market, the municipal labour market programmes employed discipline, by gendering and ethnicising the class categorisation of the participants as ‘losers’ in the labour market and by making class invisible. Discipline was expressed in the programmes through the social workers’ activities that circled around pushing the participants to pursue regular employment or vocational training while in SE. “Unfortunately they settle down a bit and think that ‘now I have a job, how nice’. So you have to keep up the pressure.” (SW1:U). ‘Putting pressure upon’ participants highlight the doing of class, as part of the programme design was to re-commodify the participant through labour market discipline (cf. Grover, 2019). This discipline is evident in participants’ descriptions of adapting to labour market routines and social norms through SE. One participant made an analogy between participation in the labour market programme and doing military service when asked what purpose he thinks that the programme serves. P2: For many people it is their first job, you could say. That you learn something. The military service hardly exists now. A lot of people learned when they got there that it wasn't just, ‘mum does this, mum does that’, but that they had to make the bed and stuff like that. I: Right, some discipline? P2: That's right. (P2:T)
The quote illustrates a participant's comparison of the responsibilities gained through military service in Sweden and those encountered in the labour market programme. It highlights discipline as a way of adapting to labour market routines and social norms. Hence, adaption can be understood to strengthen the symbolic border between the categorisation of ‘superfluous’ and ‘losers’ in the labour market through discipline (cf. Tilly, 1998).
The doing of class through the categorisation of participants as ‘losers’ in the labour market reflects how work tasks in the six SE programmes corresponded to an adaption towards low-skilled work within municipal welfare services, for instance, low-skilled tasks within preschool, elderly care, serving food, janitorial services and landscaping that often include manual labour. Within these employment frames, the participants could participate in whichever work tasks they preferred, as explained by the social workers in the interviews. Then it is very important for us and for them. We start from what they are interested in. We ask them a question about whether they are interested in working in childcare or as a food steward or whatever it is we can offer them at that time, and, based on their interests, we check their options further. (SW2:Y)
This quote illustrates how SE is framed as offering participants ‘choices’, while in reality these are limited to low-status jobs available within the organisation. Participant ‘interests’ are thus aligned with existing organisational needs, masking how class is done. Inequality is rendered invisible, and SE is presented as empowering rather than reproductive of labour market hierarchies. One potential consequence of the invisibility of inequality is that SE through low-skilled work tasks is maintained as common sense, thus making it more challenging for individual participants to resist (see Acker, 2006). As participants adapt to these roles, class hierarchies are reinforced, activating the unemployed into precarious employment and securing their status as marginalised ‘losers’ in the labour market (Offe, 1996).
An intersecting gendered distribution of work tasks performed by the participants in the SE programmes was an additional aspect of how work tasks were presented as ‘options’ by the social workers. Disclosing how the adaption process towards the lower segments of the labour market in fact included gendering how class was done. The exception was labour market programme X, in which all participants had the same work tasks. Generally, women were overrepresented in SE characterised as ‘caring professions,’ while men were overrepresented in SE characterised as ‘manual professions’ (See also Parsland and Ulmestig, 2022). SW2: So there's the textiles team now and the kitchen team and the cleaning team. There are more women there. They are in those teams. And the cleaning team is spread throughout the municipality, in schools and so on. But there are many women. I: I find that very interesting. Because it follows … the gender distribution in the labour market in general. SW2: Yes, it does. And then there's the matter of which country you come from. It's difficult to get Muslim women to sit in our cars with a lot of guys. It's not possible. SW3: I don't really understand this. Because I mean, if you go to the parks department that manages urban open spaces, there are women working there. Girls. But getting female workers here […]. But it's a bit strange. It's very strange. SW2: There is something that blocks them. SW3: Yes, where is it … It's not here in the [programme]. (Social workers 2 & 3:T)
The quote illustrates how class is done through intersecting processes of gender and ethnicity. Social workers described distributing work tasks according to the gender-segregated labour market, adapting the programme to prevailing norms. This shaped participants’ SE choices, aligning certain bodies with certain jobs. Ethnified markers—such as ‘Muslim women’—also influenced this adaptation. Thus, the adaptation process involved doing class through normative assumptions about gender and ethnicity (cf. Acker, 2006). In interviews, attributes like language, religion, and cultural traditions were often linked to unemployment and framed as individual shortcomings, producing ‘culture clashes’. SW3: We have had some culture clashes here too. There's a lot about praying and ‘we always do that’. No, not here. SW1: Also, when they've come to their subsidised employment and so on. Then they've had to talk about … I'm thinking about preschool. How to talk to children or something like that. R5: That's a good example. R1: Yes, in Sweden we respect free will and all that. But sometimes you may have had, I don't know, it may be cultural. The jargon is a bit harsher. (SW3,1,5:U)
The quote shows how social workers attribute ‘cultural clashes’ during SE to foreign-born participants, framing unemployment as a matter of ‘cultural’ deficiency. Rather than excluding participants as ‘superfluous’, the adaptation process aims to reshape them, aligning with Offe's (1996) notion of labour market ‘losers’. This ethnified adaptation reinforces stereotypes, portraying foreign-born individuals as unemployed due to cultural traits, while native-born individuals are framed as victims of social conditions (Harrits, 2019), as illustrated in the following quote. I am also thinking about the rather few Swedish participants. All are Swedish, but born in Sweden, [participants] that I have had in the labour market programme. Many of them have been quite jaded and scruffy people. We talked about it the other day, [Social worker 2] said that not many [in the programme] have a Swedish-born mother and father. But those we have had have not been feeling well. (SW5:U)
The distinctions made by the social workers in the quotes emphasises how the ethnicised adaption manifests, where the ‘Swedish’ participants are individuals who are not feeling well and also are under-represented in the programme. The stereotypical categorisations legitimised a differentiation between foreign-born and native-born participants, as stereotypes often are ‘common sense’ and seldom contested (Acker, 2006; Harrits, 2019). Such framing obscures structural inequalities and shifts the focus from limited job opportunities to individual ‘deficits’, reproducing classed and ethnicised hierarchies in the labour market. The quote thus also highlights that although the classed categorisation of ‘loser’ in the labour market is similar for both native-born and foreign-born participants, foreign-born people are overrepresented in the low-status and low-paid occupations, which means a precarious position in the labour market and generally lower wages than native-born people (Ahrne et al., 2021). The type of work assigned to the ‘losers’ of the labour market is gendered, and in the adaption process, social workers attribute meaning to foreign-born ‘culture’ and individual choice, rather than seeing structural discrimination in the programmes.
Profit
The last thematised process is profit, highlighting how the interviewees perceived the SE programmes making an economic profit for the municipalities regarding reducing welfare costs in terms of reducing costs for municipal welfare services and reducing costs for social assistance. When participants in SE programmes perform tasks that are part of welfare services, it is presented as an economic benefit for municipalities. In general, participants had a monthly salary of between 1350 and1800 Euros, which is significantly lower than the average income in low-skilled employment in Sweden, which is around 2000 Euros a month (SCB, 2025). At the same time, as described by a participant in Programme T, participants risk being used as cheap labour for tasks for which they are not qualified. In the following participant's example, she is employed as a language support worker in SFI education (Swedish for immigrants). I plan the day, the schedule, lessons, how to teach them, what they need. In principle, like a regular SFI teacher. But it's called a language support worker. It may be that I also interpret sometimes when needed. […] it feels like I have, I am in chains. I'm so limited. I don't have much to choose from. So I have to make do with what I get, basically. (P4:T)
The quote highlights the misrecognition and limited agency of the participants in terms of experiencing not having any options. From the participant's perspective, their options are to continue claiming social assistance or have SE with low pay. As Govender's study (2023) shows, this limited experience contributes to unemployed individuals in SE feeling misrecognised in relation to society.
The respondents also perceived that SE could keep social assistance costs down. In the SE programmes, the primary goals of SE were in the first place to reduce the municipal costs of social assistance and in the second place to improve participants’ establishment in the labour market. “Then someone has calculated the cost somewhere too, that they [the participants] cost society quite a lot. It's not something they [the governing bodies in the municipality] say, but that's obviously also the case” (SW3:Z). The quote highlights how social workers viewed social assistance as costly, making SE a strategy to reduce municipal expenses through PES subsidies. Unlike other municipal programmes, SE was largely state-funded, rendering it a profitable option. As one social worker noted, transferring individuals from assistance to SE could even generate municipal revenue. One subsidy provided by PES, ‘Extra jobs’, was particularly lucrative. Social worker 5 in programme U explained: […] many [participants] have social assistance. So you could say it's a very profitable business for the municipality because the government grants cover 100 percent of the cost of the employment plus you have a supervisor's support. So the municipal administration makes more than 900 [SEK] per person per month. And we have perhaps 140 [participants] right now in extra jobs. Plus, social assistance goes down, families become self-sufficient, with all that that entails. (SW5:U).
The quote highlights how SE enabled municipalities to offload welfare costs by shifting individuals into low-paid, state-subsidised work. Rather than promoting long-term security, the system economically rewards the relegation of vulnerable individuals to precarious labour. This reflects a structural logic where managing and containing unemployment takes precedence over addressing inequality or supporting vulnerable individuals towards job security.
A municipal reserve army: concluding discussion
Class always indicates inequality, as the concept of class highlights the unequal distribution of resources and power between categories of individuals (Acker, 2006). As previous research has illustrated (Govender, 2023; Hultqvist and Hollertz, 2021), activation through SE can make resources available and change the stigmatised position of the unemployed in society, to some extent. Indeed, this study also highlights how resources become available for some unemployed individuals, as the participants in the studied subsidised employment programmes are provided with employment that also provides them with access to the social insurance system. However, framing activation as resource-generating rather than re-commodifying assists social policy's desire to turn unemployed people into workers, without questioning how this is done, or which jobs unemployed individuals are fit for. By analysing how class is done in six SE programmes, this article provides insight into these questions.
The main results of this article point towards how class is done by analysing the organisational inequality regimes through processes in the six studied SE programmes. Through the analysis, it becomes clear how resources were divided unequally between individuals who are accepted onto the programmes, and individuals who are assessed as ‘superfluous’ for the needs of the labour market. Nevertheless, the analysis also reveals inequalities within the programmes. The participants are adapting their behaviour, skills and attitudes to fit the needs of the lower segment of the labour market, where perceptions of participants’ ethnicity and gender appear to play a role in their adaptation to the labour market. Thus, connecting to the ethnified and gendered segregation in the Swedish labour market (Ahrne et al., 2021). The main purpose of using subsidised employment as a municipal activation intervention was formulated as reducing welfare costs, in terms of cutting social assistance and labour costs for the municipal work force.
The meaning of the categorisation of ‘losers’ is that there should be a group of people available to do the precarious jobs in the lower segments of the labour market, like a reserve army (cf. Marx, 1976). Since the SE programmes are targeted at municipal jobs in the lower segments, the doing of class can be understood as reproducing a municipal reserve army. This is evident in the organisational processes which focus on sorting and adapting participants together with activities such as discipline conducted as well as stereotypical ethnicised and gendered categorisations of participants towards that respond to the need of municipal low-wage work. The process of profit highlights how the meaning attributed to activate unemployed individuals through SE was connected to the participants carrying out work tasks to a cheaper employer cost for the municipality, but with poorer conditions than the rest of the workforce, since the participants have short-term contracts that are flexible and less protected, alongside being activated whilst working. The exploitation of unemployed individuals through the SE programmes may seem analytically clear cut but is legitimised and made invisible and thus harder to contest for participants in the programmes (Acker, 2006).
The purpose of a reserve army, according to Marx (1976), is that unemployed, but work-capable, individuals are to operate as a lever on the labour market, keeping competition high for work opportunities while keeping salary demands low. By training participants for specific municipal jobs in the lower segments, the results indicate that such a function can also be found in the six municipal SE programmes studied. The concept of a municipal reserve army highlights the convenient municipal need of having access to a labour reserve, given municipalities dual role of being both responsible for unemployed people without social security and employers in low-status and low-wage welfare occupations. Through the SE programmes, municipalities can therefore create a workforce for the temporary and precarious jobs in the municipality.
Since municipal SE programmes, vary in size, content and target group, the results of this limited qualitative study must be generalised with caution. The study underscores the necessity to further study the organisational processes that do class and the implications these have for unemployed individuals, both in the immediate and longer term. This should encompass labour market conditions, social inclusion and the ways in which class is experienced and perceived and thus, by extension, the reproduction of the reserve army through activation. The ethnicising and gendering of doing class in this study highlight the need to analyse class as ongoing relational and in constant movement, where organisational processes shift and create different experiences of class, even though the classed categorisation is common (Acker, 2006). Previous studies have highlighted the re-commodification of activation as part of reproducing the reserve army and argue that the reserve army is primarily reproduced through labour discipline (Prendergast, 2019; Wiggan, 2015). By analysing the organisational processes in the SE programmes, this result illustrates how the doing of class is interwoven in the organisation of the municipal SE programmes, which legitimises the reproduction of the reserve army in terms of cost reductions for welfare. Meanwhile, unemployed individuals continue to struggle over “the crumbs from the rich man's table” (Govender, 2023: 215).
Footnotes
Acknowledgements
The author would like to thank professor Rickard Ulmestig, associate professor Gabriella Scaramuzzino, Professor emeritus Anders Forslund and associate professor Bettina Leibetseder for helpful comments on previous drafts of the article.
Funding
This work was supported by Forskningsrådet om Hälsa, Arbetsliv och Välfärd (Forte), grant number 2016-07123.
Author biography
