‘You couldn’t have a heart and want to strike’: Mobilising workers in England’s social care sector

This article uses John Kelly’s mobilisation framework, with its foundational concept of injustice, to explore workers’ propensity towards unionism in England’s outsourced social care sector. Drawing on 60 interviews with union organisers and officers, care workers, support workers and care company managers, this research highlights the difficulties of union organising in the sector and explores theorisations of mobilising. The research contends that for mobilisation theory to provide insight into relationships between work and unionism, varieties of injustice and collectivism need to be contextualised. Paid care provision generates both employment-related injustices and care-related injustices, which lead to divergent collective identities and attitudes towards unions. An absence of a coherent entity for workers to attach blame to – within a context where private providers frequently remain reliant on state funding levels – affects whether injustice and collectivism progress to mobilisation and unionisation.


Introduction
This article engages critically with Kelly's (1998) mobilisation framework to contribute an analysis of the role of injustice and collectivism in encouraging or inhibiting unionism in England's social care sector. It focusses on mobilisation within domiciliary care, residential care and support work in England, 1 areas which UK governments have increasingly outsourced to for-profit and not-for-profit care providers. Care packages (specifying the amount of care time that an individual will receive) are now commissioned to care providers by local authorities -with the effect that private providers are often reliant on public funding. Funding levels for social care provision have been impacted by austerity policies implemented since the 2008 financial crash, leading to claims that caring labour is undervalued (Hayes & Moore 2017). The fragility of the sector appears to be worsening under the effects of Covid-19 (Bottery & Ward 2021) despite funding increases. Historically, greater funding does not necessarily translate into improvements for workers (Rubery et al. 2013) and residential homes are frequently reliant on private equity backing, with an expectation to increase returns on investments (Burns et al. 2016).
The combination of austerity practices, privatisation and financialisation have all affected the way that caring labour is carried out. Work in the sector is characterised by poor employment practices and low wages (Baines et al. 2019;Rubery et al. 2013). Tensions exist between the informal, 'intimate' requirements of care and the formal expectations of the commissioning process to deliver care within delineated, short visits (Bolton & Wibberley 2014: 683). Informal, emotional aspects of care have been relatively resilient (Baines 2016;Brown & Korczynski 2017;Johnson 2015). This is in part due to the willingness of workers to engage in unpaid labour (Aronson & Neysmith 1996;Bolton & Wibberley 2014) -a willingness which can be related to a narrative whereby caring labour is entered into 'for the love' (Folbre 1995). Restructuring in the sector has also affected the ability of unions to organise. There is a notable absence of union membership (Baines & Cunningham 2015;Hayes & Moore 2017) or other forms of collective mobilisation. Some local authorities have signed up to the trade union UNISON's 'Ethical Care Charter' (and the newer 'Residential Care Charter'). However, the charter focusses on connections between quality of care and working conditions, without explicitly mobilising workers to collectively improve conditions and become union members.
While research has focussed on the impact of restructuring in the sector on employment relations (Rubery & Urwin 2011) and the impact of austerity on caring labour (Hayes & Moore 2017), the specific obstacles to mobilisation and union organising in the sector are less explored. Mobilisation has been considered in some comparative contexts: Rogalewski (2018), for example, explores the mobilisation of women migrants working in Switzerland's care sector. Huget (2020) describes how workers face a moral conflict when deciding whether to mobilise and go on strike, or whether to ensure continuity of care and avoid strike action. This article builds upon this extant research on mobilising workers and moral conflicts using the context of England's social care sector. The 'The mobilisation framework and its critiques' section points to a need for more rigorous analysis of the foundational concepts of Kelly's framework. The 'Methods' section outlines the methods and process of analysis used in the research. The 'Findings' section shows that workers who were not union members (referred to hereafter as nonmembers) tended not to attribute blame for injustices, that collectivism among workers is shaped by the organisation of work and the structure of the sector and that a disconnect exists between generic collectivism in the workplace and unionisation. Through this analysis, the article highlights views on working conditions and unions in the context of England's largely privatised care system and evaluates Kelly's mobilisation framework -a framework central to industrial relations scholarship. Kelly's (1998) text Rethinking Industrial Relations has been a key contribution for researchers analysing and unpacking processes of union organising. The text draws upon a range of literature on social identity, social movements and collective action. In particular, Kelly draws upon the mobilisation theory proposed by Tilly (1978), which divides the mobilisation of collectives into five aspects. First, the interests of the social group, as defined by the group itself; second, the structure (or 'organisation') of a group; third, a process of mobilisation whereby a group gains collective control over resources required for action; fourth, opportunity to act -dependent on power relations between groups -and finally, the collective action. Kelly also utilises McAdam's focus on cognitive processes in his framework: McAdam (1988) emphasises that perception of injustice, breach of values, social identification among a collective group and belief that the group is entitled to (and is able to) collectively act to rectify injustices are all components of mobilisation. Drawing on the work of Tilly and McAdam, alongside using aspects of social identity theory, Kelly (1998) defines mobilisation as a 'core set of concepts, injustice, attribution, and identity' (p. 127): workers identify a shared grievance, view it as an injustice, attribute blame for this injustice to an agent (usually the employer), and form a social identity in opposition to that agent. Activists are essential in mobilising workers through these stages.

The mobilisation framework and its critiques
Empirical research has generally substantiated the framework, in particular through a focus on the role of activist leaders (Cregan et al. 2009;Darlington 2018;Jiang & Korczynski 2016;Legault & Weststar 2015). There is, however, disagreement as to what mobilisation entails. Analyses have presented mobilisation as the 'union joining' of workers (Kirton & Healy 2013), as the actions of union members (Cox et al. 2007;Johnson & Jarley 2004) or -in a more expansive interpretation -as collective resistance which might occur outside of formal union structures (Legault & Weststar 2015). There is also an ambiguous relationship between mobilisation and the 'organising model' of unionism, which proposes grassroots, bottom-up unionism focussed on building an activist base. Some studies suggest that the organising model is in practice a mobilising model (e.g. Dixon & Fiorito 2009: 173), and others view mobilisation as 'an important tool and activity within organizing [. . .] not, on its own, organizing' (Holgate et al. 2018: 4). Holgate et al. (2018) note the different conceptualisations of power behind mobilising and organising, with organising requiring a broader power base. Similarly, McAlevey (2016) distinguishes between mobilising and organising, referring to 'deep organising' as an approach requiring systemic power. McAlevey (2016) contends that mobilising is only successful in particular contexts, for example, in the case of the 'Justice for Janitors' (p. 66) campaign, where labour costs to the employer were low and demands on the union were also minimal. Deep organising, in contrast, would entail high levels of membership participation. McAlevey's approach to organising and mobilisation goes beyond Kelly's more general comments on leadership to emphasise that deep organising requires workplace leaders. McAlevey (2014) also critiques a strategic focus on 'grievances' as short-termism which primarily consolidates the power of paid union staff; a perspective which also highlights the fragility of mobilisation based on individualised injustice.
Building on this literature, this article unpacks how injustice and collectivism relate to mobilisation and then, in turn, how mobilisation relates to union membership. It explores the theoretical foundation of 'injustice' in Kelly's text; an aspect of the framework which has arguably faced the most criticism. Kelly (1998) explicitly posits injustice as 'derived from Marxist analyses of exploitation and domination within capitalist economies' (p. 64) yet it is unclear how injustice relates to exploitation. Injustice is described as a 'conviction that an event, action or situation is "wrong" or "illegitimate"' (Kelly 1998: 27) -a description which assumes a system of morality or perception of legality, which Kelly does not provide. Kelly's approach to collectivism is similarly related to a moral framework. In-groups (e.g. workers) are differentiated from out-groups (e.g. employers) because of their 'different interests and values' (Kelly 1998: 30). In response to this ambiguity, Ackers (2002) argues that without an 'ethical framework' Kelly's theory is overly reliant on Marxist concepts; injustice is 'a moral inflection [. . .] merely accruing automatically to economic militancy' (pp. 14-15). Kaufman (2018) similarly critiques the Marxism of the mobilisation framework. Kaufman (2018) argues that analyses which focus on class are 'ill-suited' (p. 59) for understanding the experiences of contemporary workers and questions whether workers feel as if they are exploited. Kelly's emphasis on injustice has also, however, been critiqued from a Marxist perspective. Cohen (2011) refers to injustice as an idealist notion which is unmoored from exploitation. Atzeni (2009) likewise argues that injustice is moralistic and individualistic. Both authors instead place an emphasis on a collectivism emergent from capitalist relations as the basis for mobilisation. According to Cohen (2011), there is more analytical value in 'rooting workplace conflict in an objective, structural trajectory of resistance triggered by the exigencies of exploitation and the capitalist labour process' (p. 374). Atzeni (2009: 18) contends that workers might 'perceive' injustices, but without activating 'moments of collectivism' related to workers' position in the labour process, mobilisation remains only a possibility. Kelly's (2018) article in the special issue on his work responds to Atzeni by arguing that mobilisation theory can simultaneously position injustice within relations of exploitation and highlight additional contingencies such as workers 'beliefs' which contribute to mobilisation. This concurrent importance of beliefs around injustice and structural exploitation goes against a more orthodox Marxist approach. For example, in Critique of the Gotha Programme, Marx (1972) decries notions of justice and ethics, notions which were more accepted among liberal economists (Bonefeld 2014). Marx's contemporary Karl Vorländer reportedly said that 'the moment anyone started to talk to Marx about morality, he would roar with laughter' (quoted in Lukes 1985: 27). Subsequent theorisations of Marxist morality have emphasised structural emancipation over moral injustice. Lukes (1990), writing about communist movements, refers to Marxism as 'a morality of emancipation [. . .]. The world from which they are to be emancipated is one of scarcity' (p. 22).
Tensions between moral injustices and structural exploitation take on a specific character within social care provision. As emphasised in the existing literature, low pay and poor employment conditions associated with economic exploitation are not always viewed by care workers as an injustice. Job satisfaction in the sector is relatively high (Hebson et al. 2015) and is often a consequence of attachments formed with care recipients (Johnson 2015). These attachments can mean that individual workers hold themselves responsible for care quality (Huget 2020) in a context where they might not have the power to instigate substantial change (Folbre 2006: 8;Ungerson 2005). However, some researchers identify potential for mobilisation within these care-based injustices (Delp & Quan 2002;Korczynski 2007;Milkman & Voss 2004). Murphy and Turner (2014) have explored coalitions between unions and community groups, looking at a successful 9-year campaign carried out in Ireland by the Services Industrial Professional and Technical Union (SIPTU). The authors used Kelly's mobilisation framework as an explanatory mechanism, finding the framework 'useful' yet lacking 'conceptual rigour': for example, 'how strong a sense of injustice, attribution and efficacy must workers experience before they become receptive to being mobilised is not specified' (Murphy & Turner 2014: 347). This article argues that propensity to mobilise in social care depends not upon how 'strong' injustice is but upon the type of injustice.
This article also explores different types of collectivism. Kelly's framework emphasises the importance of a coherent collective social identity in processes of mobilisation; wider literature suggests that collectivism in the workplace varies (Stephenson & Stewart 2001). Collectivism is impacted by social dynamics, political changes and employment relations (McBride & Martínez Lucio 2011). Furthermore, collectivism is affected by the nature of the labour. McBride and Martínez Lucio (2021: 160) note that in contexts of 'unskilled' labour, social stigma affects occupational identity: 'workers are aware and concerned of who they are as individuals and whether what they do is viewed, understood and valued -or not'. There is frequently a societal disconnect in the 'valuing' of caring labour, whereby workers find themselves morally valued while being financially overlooked. In the context of social care, whether a collective identity becomes an aspect of mobilisation therefore depends on both the broader employment relations in the sector and the societal and personal perceptions of caring labour.
Drawing on this body of literature, this research contends that the concurrent focus on idealistic notions of justice and material and structural exploitation in Kelly's framework requires further interrogation. Collectivism and injustice need to be contextualised. The article first explores workers' narratives around injustice and responsibility for injustices, distinguishing broadly between employment-related injustices and carerelated injustices. Second, it examines the forms of collectivism among workers that are generated by injustices. Third, it focusses on the relationship between workers and unions, emphasising that some workers do not view unionism as 'collectivist' and view unions as causing injustices rather than alleviating them. Kelly (1998) comments that activists must 'defend collective action in the face of counter-mobilizing arguments that it is illegitimate' (p. 35), but does not examine how forms of legitimacy and conceptualisations of injustice interact and counter each other. This article argues that these interactions are particularly distinctive in the provision of care.

Methods
The fieldwork for this research, carried out in 2018-2019 (prior to the spread of Covid-19) consisted of semi-structured interviews with 60 individuals. As a piece of qualitative research, the study does not have aims towards generalisability -the breadth of research is sacrificed to achieve a narrative depth (Hammersley 2013). Insights pertaining to mobilisation have been drawn from the data using a comparison of the perspectives of non-members and members, alongside the views of organisers and managers. The research included 17 interviews with trade union officers (including an officer from the Trade Union Congress and from the local Trade Union Council), 11 of whom were organisers holding positions at three different unions. The unions researched were GMB, UNISON and Industrial Workers of the World (IWW). All three of these unions organise within the social care sector to varying extents. UNISON and GMB are substantially larger than IWW, with the latter utilising more of an activism-based approach to organising. Organisers and officers from GMB and UNISON suggested that GMB prioritised partnership unionism (focussing on gaining voluntary recognition agreements with employers), and UNISON prioritised political campaigning and relationships with local authorities. All three unions struggled to recruit members in the social care sector.
Interviews were also conducted with 12 employees of a domiciliary care provider and the company employer/manager, 16 workers at a residential home, the manager of the same residential home and the company's HR manager and 12 support workers and former support workers. The domiciliary care provider employed over 100 workers and provided care to over 180 individuals. The residential home had 67 beds, most of which were full. The home was part of a wider company (funded in part by private equity), encompassing 80 care homes throughout the United Kingdom. At the domiciliary care company, only the employer was a union member, and at the residential home two of the workers and the manager were union members. In contrast, all of the support workers interviewed were union members. The support workers interviewed were employed at companies providing support for adults with learning disabilities and mental health needs, registered as not-for-profit organisations. Four support workers worked at company which included a day centre and residential facility, and which employed 140 people. During the fieldwork period, the workers were campaigning to meet the statutory requirements for UNISON to be recognised in the workplace. Seven support workers provided a supported living service (personal care, financial assistance, domiciliary assistance) to 26 individuals. Workers had previously been involved in workplace organising -with IWW -but these efforts had died down at the time of research. The final support worker (who had been a member of all of the three unions) worked at a company providing similar services but in a different city.
The companies all operated in the North of England and the union officers and organisers were also based in the North of England, with the exception of three national officers. Access to the workers at the residential home and domiciliary care company was gained through GMB organisers, as GMB had recognition agreements with both companies (one agreement inherited via Transfer of Undertakings (Protection of Employment), and another agreement established at the request of the employer). The support workers were accessed through union organisers and union members using an informal 'snowballing' method. The interviews lasted between 15 minutes and 90 minutes (the shorter interviews were those carried out with workers between their shifts). They were then audio-recorded and transcribed verbatim. All research participants were given information sheets describing the purposes of the study, and names and identifying features of the individuals or the companies have been taken out.
The interviews with union organisers focussed on organising strategies used by the unions, difficulties faced when organising social care in contrast to other sectors, relationships between unions and employers, and the role of member activism. The interviews with workers focussed on their experience of their job and their perspective on trade unions. The findings were coded thematically using NVivo. The analysis involved a combination of inductive and deductive analysis (Fereday & Muir-Cochrane 2006). The deductive aspect of the analysis drew upon the framework of mobilisation, as detailed by Kelly. The interviews were also analysed inductively: the NVivo 'nodes' function was used to locate emergent patterns and organise patterns thematically. These patterns were then related back to the mobilisation framework. The themes analysed in this article are: employment conditions, injustices, collectivism, individualism and attribution of blame.
The following findings sections are organised to reflect these themes. The sections explore the perspectives of workers, employers and union organisers, relating these perspectives to aspects of the mobilisation framework. The 'Employment injustices and care recipients' section examines the injustices described by workers and unpacks whether workers viewed employers -or the state -as culpable or accountable for conditions in the sector. The 'Collective identities among workers' section considers the forms of collective identity formed through and against these injustices, distinguishing between the identity of 'carer' and of 'worker'. The final findings section focusses on workers' views of unionism. The section emphasises that, without attribution of blame, a recognition of injustice and formation of collectivism remains distanced from propensity towards unionism.

Employment injustices and care recipients
The care workers and support workers interviewed described multiple aspects of their work as 'not fair' or 'wrong', including low wages, precarious contracts and inadequate sick pay and holiday pay. At the domiciliary care company workers described wages as too low -'you don't get good enough pay for it' -and criticised the organisation of their work. The commissioning practice used by local authorities relies upon domiciliary care workers carrying out a number of small visits throughout the day. This can result in large amounts of travel time and gaps between visits. Care workers argued that the pay for travel time was insufficient. The company used Google Maps to estimate travel time but one care worker argued: 'Google Maps is wrong [. . .] it's given me 10 minutes to get from [. . .] a care home down this end, to get to the other care home [. . .]. I've actually timed that walk, and it takes 22 minutes'. The worker went on: We're always told you've got to stay in call times for the full amount of time, my question is, how? Because even when you give me 15 minutes leeway either side of the call time, I'm still not getting there on time because you've not given me enough travel time to get there.
Workers at the company also complained that a significant proportion of their working day was unpaid because of the gaps between calls. One domiciliary care worker said: 'I'll be out ten, eleven hours a day and I'll only be paid for like seven or something like that, it's not right'. While the workers at the residential home did not have these unpaid gaps during their work day, like domiciliary care workers, they highlighted low pay at the company: 'I think the pay's not really that good for the work you do, like you could be stacking [supermarket] shelves and you're on like £9 an hour'. Support workers also criticised stagnant low wages. One support worker said that 'there are some who've been with [this company] for 15, 20 years . . . still grinding away for minimum wage', and another worker commented: 'I've always been on minimum wage [. . .] the only way to get by as a support worker [is] by just being extremely frugal and essentially living a semi-poverty life'.
As well as describing discontent over wages, workers referred to a pressure from management to overwork and suggested that they would enjoy their job if managers and senior staff treated them with respect. This lack of respect (interviewees often used the term 'dignity' to describe what they wanted from managers and senior staff ) was intertwined with issues connected to pay and contracts. A domiciliary care worker commented that his company's policy was to move new workers onto permanent contracts after a 3-month probationary period, yet he had worked at the company for 9 months on a 0-hours contract. He said, 'it would be nice to have some sort of security, and the appreciation as well [. . .] like "you've done a really good job, we'd like to offer you this contract", [. . .] you feel a bit hard done by'. One organiser from UNISON, commenting on the experiences of domiciliary care workers, argued: 'pay cuts hits the dignity question [. . .] non-payment of travel time is about pay but is really about dignity as well'.
Workers viewed these employment-related injustices as illegitimate -indicating a basis for mobilisation according to Kelly's framework -yet non-members also described strong emotional bonds with care recipients which offset these injustices. Domiciliary care workers commented: 'I just love sitting with them, having that companionship with them'; 'it is rewarding cos you meet so many different characters, and some of them you just . . . you do get a bit of an attachment with them, I love it'. At the residential home, a kitchen assistant said: 'I love the residents here, I think they're brilliant'. Many workers described attending the funerals of care recipients in their free time. One domiciliary care worker recalled: 'the only people you could hear in the church crying their eyes out was us lot'. These bonds with care recipients provided job satisfaction, a finding consistent with extant studies (Hebson et al. 2015;Johnson 2015) and also acted as compensation for poor working conditions. Iterations of the view that 'it's rewarding and that's why I do it, it's definitely not for the pay' were repeated throughout interviews -all of the workers using these phrases were non-members. A UNISON organiser described this view as 'just accepting their lot, really [. . .] they're in it 'cos they want to care [. . .] and they've just accepted that that comes with really crap terms and conditions and pay'.
Union members, however, were explicit in their acknowledgement that pay was the main factor motivating them to work. They viewed the widely accepted trade-off among their colleagues -that is, the off-setting of workplace injustices connected to the caring nature of the job -as a major obstacle to organising. One union member employed as a support worker commented: We are somewhat handicapped by the good-natured souls who do it for the love of it, because if we were to all speak up and say, 'actually we want the same rights as everybody else', it would be a bit more powerful.
The union members argued that because their fellow workers were focussed on improving quality of care, they did not hold managers responsible for poor working conditions and low pay. A support worker said: 'this is a private company; they can pay us what they choose to pay us. [Other workers] don't want to blame the manager, they don't want to blame the boss'. A union organiser from IWW referred to this as giving managers a 'free pass': [Some workers] really care, it's not just a pay cheque [. . .] they really give a shit about their clients, the people they're looking after [. . .] As much as you want to like [those workers] -and I understand where they're coming from -you can't just give your boss a free pass like that.
Which is ultimately what they're doing, just giving their boss a free pass for treating them like crap.
There was a divide between union members who did 'blame the boss' and non-members who did not, substantiating Kelly's (1998) assertion that it is 'vital that aggrieved individuals blame an agency for their problems, rather than attributing them to uncontrollable forces or events' (p. 29) in order for mobilisation to occur. In social care, the tendency to view problems as uncontrollable extends beyond workers; employers also attribute blame elsewhere. The manager of the domiciliary care company described this dynamic and his own role within it: 'we rip [workers] off on a daily basis and then hide and pretend that it isn't happening, or we blame the local authority and the local authority blames central government, but still nobody does anything'. Union organisers and officers connected economic injustices in the sector with austerity policies. Their strategies had increasingly focussed on local authorities or utilising political pressure, rather than confronting actions of individual employers. Among workers, there was some emphasis on practices of local authorities but this was primarily in relation to 'in house' care and support workers receiving better wages (alongside the threat that local authorities might withdraw care packages if care was inadequate). Workers did not emphasise broader dynamics of austerity, underfunding and privatisation as possible reasons for injustices. Employment relationships in the sector diffused blame, as opposed to diversifying the number of agents responsible.

Collective identities among workers
According to Kelly (1998: 126-127) mobilisation results from a social identity built around a collective interest, which has 'roots' in injustice. However, any form of collectivism among workers can be difficult to establish among domiciliary care workers and support workers given the isolating organisation of the work. The domiciliary care company was situated in a rural area, meaning that care workers would spend a large proportion of their day alone, driving to the highly dispersed homes of care recipients. A care worker commented: 'you can start to feel a bit isolated'. 'Double-ups' -care visits requiring two workers -alleviated the isolation felt by care workers. One worker said: 'it's nice to be on a double-up call even if you're in there 10, 15 minutes. It's still that little bit of social interaction with someone else'. This double-up work increased collective feeling and allowed close friendships to form between workers. A care worker said: 'as we've started working a lot together [. . .] we've become quite good friends' and another care worker commented that 'it's like a little family'.
This form of social identity was connected to the organisation of work, which involved working in close proximity, and to the form of labour, which can be emotionally draining. When workers at the domiciliary care company described a collectivism formed specifically in response to workplace injustices, it was related to bullying from office staff. Care workers avoided going to the office to pick up equipment, such as gloves, and avoided speaking to office staff on the phone. One care worker commented: 'I wouldn't ring the office 'cos they just don't talk to you right, like crap'. In particular, care workers resented that calls to the office frequently resulted in having extra shifts added to their day, and were indignant that office staff denied them sick leave and would rarely cover care shifts themselves (despite having sufficient training). Care workers' negative view of office staff created an in-group and an out-group. A care worker said that 'sometimes it's quicker just to bypass the office and sort [shifts] out between ourselves, 'cos we're all good friends, we all get along'. These office staff had more power than frontline workers in that they shaped their responsibilities and could reprimand workers, but they did not have a substantial amount of decision-making power within the company as they faced pressures to ensure that shifts were covered. Workers' views of their employer, who would have had more power than office workers, were in contrast very positive.
At the residential home, group cohesion was again related to the way that work was organised. Residential care tends to be far less isolating than domiciliary care as workers are all situated within the same building. At the home researched in this article, a collectivism was built around an interdependency among frontline workers. Workers were employed in a variety of positions including care assistants, kitchen assistants, laundry workers and domestic workers, yet tasks frequently transcended these formal boundaries. A worker employed in the kitchen provided insight into the dependence between workers: 'you depend on each other a lot, and you've got to [. . .]. Carers depend on us when there's been accidents and things, we depend on carers'. Yet the high turnover of staff meant that it was difficult to maintain relationships. A laundry worker, who had been at the company for 5 years, commented that 'you get friends with people and then they go, and I just find it a bit sad because some of the people I've become friends with and then they've left and I hardly see owt of them'.
There was no in-group and out-group at the residential home related to who had power in the workplace. Workers expressed a general ambivalence towards the manager, as management at the home changed frequently. They were unaware of (or uninterested) in the wider structure of the company which the home belonged to and which ultimately employed them. A semblance of in-group and out-group was, however, apparent in a division between workers who had been at the company for a long time and the newer workers. The latter cited bullying from the more established workers. One worker who had been at the company for a year said that her colleagues did not take a cooperative approach to labour: 'nobody's happy to work for £7.80, whatever, an hour, 12-hour shifts [. . .] that's why they should just help each other and realise that the shift would go so much better and easier if we just help each other out'. This call towards collective effort related to concern for care recipients rather than to a recognition that workers are in a similar position to each other. Another worker who had been at the company for 6 months said: 'some people are only out for themselves, in a job where everybody should be working together, and care about these guys [care recipients], like, more than themselves'. A GMB organiser emphasised that this perspective is common throughout social care: Care workers are very protective of each other and their residents [. . .] care workers will do everything they can [to] advocate for the residents, and they will make sure they're looked after. But they will look after the residents more than they look after themselves.
The social identity of workers in this instance centred on their role as individuals providing care, as opposed to as workers.
In part, as with non-members, union members' descriptions of collectivism were shaped by the organisation of their work. Workers who were required to work one-onone with care recipients referred to an isolation from colleagues, and avoided the company office. A support worker commented, 'I keep myself to myself, and [management] know that I'm good at my job, so they just kind of leave me alone'. Other support workers worked alongside colleagues within the homes of care recipients. As with domiciliary care workers' double-up shifts, this frequent small-group work strengthened connections between frontline workers: '[the workplace] was a bit like a family'. However, the connections between workers remained limited. A former support worker commented: 'you were part of a big organisation, but your world was quite small, and you didn't really have connections to other people'. Another former support worker emphasised that this insular group cohesion created an obstacle to mobilisation, but not one which had been intentionally built into the company's structure: I don't think the company are, or were, smart enough to have designed their company in such a way that it fragmented the workforce to prevent unionisation. I think that would be giving them too much credit. I think it's just the nature of the thing.
The workers from the second support provider noted that it was difficult to establish relationships with workers between the company's different services. One of the workers recalled: 'we sometimes have residential service staff come over to get some shifts in, and people are just like, "who are they?"' This divide had acted to frustrate the efforts of activist leaders to mobilise around issues. Before the fieldwork period there had been disputes over contract changes in the day centre and the residential service, mainly relating to sick pay. A worker from the day centre who had been instrumental in disputing the changes commented: 'there was other people in [residential] who were also fighting the contract because of sick leave. But they [management] don't want you to be brought together'. A drive to seek recognition through the statutory route at the company had been particularly arduous due to the fragmentation of the company's workforce. A union representative employed in the company's adult day service said, It's such a huge amount getting to 50% of staff members, when you think about how many people just come and do occasional [shifts], how many different services there are where people aren't necessarily talking to each other that much.

From mobilisation to unionisation
This section moves from injustice and collectivism to union collectivism. As noted in the 'The mobilisation framework and its critiques' section, Kelly tends to conflate collective action with unionisation. He describes the processes of mobilisation, then writes: 'out of these interactions emerges the desire for unionism, a particular form of collective representation' (Kelly 1998: 51). This research found that workers' specific understandings of unionism impeded the transition from collectivism to unionisation. Despite both the domiciliary care company and the residential home having a recognised union, workers at these sites generally had no awareness of unions. They tended to view unions as a service, which they considered to be largely superfluous, or as organisations which encouraged or compelled workers to strike. Workers drew upon the latter understanding of unionism to justify their non-membership, citing the negative effect that strikes could have on care recipients. At the domiciliary care company a member of the office staff said that anyone employed in social care could not 'have a heart and want to strike and leave somebody in their house on their own' because 'it would be them that are affected rather than us getting what we wanted, it would be them -the customers -that suffered'. A care worker at the same company reiterated this view, commenting that 'we couldn't [strike] because we care too much about the people we're looking after, and we wouldn't want them to suffer'. In the same way that non-members did not regard employers (or the state) as responsible for poor working conditions, they did not regard these actors as responsible for maintaining care quality.
Non-members therefore viewed the potential consequences of strikes as an injustice which would massively outweigh their own discontent surrounding wages and working conditions. A UNISON organiser highlighted the negative impact that this view of unionism had on mobilising: We'll go and drop some [union literature] off somewhere, [workers] will be like 'oh I'm not in a union'. We're like 'oh why not?'. 'I'll never go out on strike'. And you're like, 'well it's a bargaining tool!' Withdrawing labour, that's it. [. . .] Just the threat of strike, is sometimes enough to get managers to the table, to negotiate. So, you can never take it off the table. But yeah, people are just like 'well I'll never go on strike', and you're like, well there you go. Prepare to be exploited.
Union members emphasised that the threat of a strike was, however, difficult to uphold given that managers would know about workers' aversion towards strikes. A former support worker reflected on whether the threat of withdrawing labour could exist at all in the care sector: I think one of the issues [at the company] in people's heads, was like 'well with a union, what's the threat that you have that can kind of change things?' Potentially withdrawing labour, right? No one I worked with would have done that [. . .] Because if they did, they would have felt so responsible because the person who we were supporting [. . .] So, once you've lost that potential thing, you're kind of limited. Even if it's just the threat of that.
This issue of responsibility towards care recipients was frequently mentioned in interviews. Union members recognised that workers would feel responsible for care recipients, but they argued that they were ultimately not responsible. They situated responsibility for care within the organisational hierarchy and criticised the individualisation of responsibility towards frontline workers. A support worker commented: 'I may care what happens to [care recipients] while I'm on strike, but it's not my problem what happens to them while I'm on strike'. An IWW organiser provided a similar perspective, stressing the importance of telling workers that they are not solely responsible for the provision of care: We live in a court of law [sic], we don't live in a court of morals, and you have the right to withdraw your labour, and that could be difficult because morally you're thinking, 'is this person going to get fed, will this person get their shopping', [but] the burden of responsibility is not on you as a minimum wage person.
The organiser went on to describe the practicalities of striking. He noted that, as per the regulations on industrial action in the United Kingdom, unions would have to inform employers of strike action 2 weeks prior. The employer could therefore utilise staff from care agencies, a practice which is common in the sector even in normal circumstances. Other organisers argued that strikes play a minimal role in the tactics used by unions anyway: a UNISON organiser estimated that '90% of union members haven't been on strike, and it's incredibly hard to do it, these days [. . .] we certainly try to explain that most of our issues are resolved through negotiation'. At the residential home and the domiciliary care company however, this message did not appear to have been successfully conveyed.
The topic of strikes thus highlighted a disconnect between the recognition of injustice and the perception that unions would be well positioned to alleviate injustices. There was also a more general disconnect between unionism and collectivism, with workers viewing unionism as a self-interested pursuit. One GMB organiser noted: 'trade unions are about making things better for you, aren't they, whereas care work is about what you do for other people, so I don't know if that potentially is a hurdle'. Organisers and officers argued that unions need to emphasise to the workers that working conditions are important because the worker is important (rather than because improvements to working conditions could be a means to the end of improving quality of care). A UNISON officer said that 'you need to make sure that the carers realise that they cannot look after the residents if they can't look after themselves, it's changing that mentality, they have to look after themselves'. An organiser from GMB echoed this view: 'the carers have to value themselves [. . .], we need to say to the carers, "you are the most important person"'.
The suggestion from organisers was that union membership was low because the social identity built up between non-members was the wrong kind of collectivism. Workers were not viewing themselves first and foremost as workers. Union organisers argued that workers have to recognise that their labour is being exploited to want to join a union. A GMB organiser said, When somebody's delivering a service, they're delivering a service for that person that they're caring for, so I think there's a greater detachment from the organisation that's utilising their labour and exploiting them. It's more removed. [. . .] I think perhaps if you're working in manufacturing or retail, where your work is sort of generating a profit for somebody, even though that does happen obviously in private social care to a lesser extent, it's thought 'I'm part of this, I'm part of generating this profit, and therefore I have an entitlement to a proportion of that, and if somebody up there is determining that I've got a proportion less than I think I deserve, then it's right for me to stand up and fight for it'.
Kelly's shift from exploitation generated by the labour process towards a more subjective notion of injustice (Atzeni 2009;Cohen 2011) contrasts with the view of this organiser. The latter regarded a recognition of systemic and implicitly class-based injustices as integral to union membership, given that the structure and nature of care provision can conceal the economic exploitation in the sector.

Discussion and conclusion
This article has explored obstacles to mobilisation in the social care sector and evaluated the predictive and descriptive abilities of the mobilisation framework. It has focussed on injustice as the 'conceptual core' (Kelly 2018) of the mobilisation framework; injustices were apparent throughout interviews in the form of wage stagnation, precarious contracts and a lack of appreciation. These divergent types of injustice then appeared to relate to mobilisation in divergent ways. Non-members viewed injustices as somewhat compensated by the satisfaction gained from the labour process of caring. Emotional connections with care recipients acted as a 'moral currency of hugs and thank-yous' (Johnson 2015: 117). Furthermore, non-members seemed to view obstructions to quality of care as more compelling and immediate injustices than wage stagnation or precarious contracts, suggesting a hierarchy whereby care-related injustices were prioritised over employment-related injustices. This hierarchy was less apparent in the perspectives of members, who were often less reticent than non-members to prioritise injustices around working conditions.
Multiple types of injustice therefore shape the experiences of workers in the sector. Given this variation, injustice does not automatically accrue to 'economic militancy' (Ackers 2002). However, the findings suggested that the injustices which were employment related (and were expressed in a more economically militant way) aligned with unionism. The mobilisation framework would be improved by conceptualising injustice not as something '"wrong" or "illegitimate"' (Kelly 1998: 27) but as a precise and employment-related notion. In doing so, the framework would arguably be embracing the militancy decried by Ackers and others. Using the context of social care, this article therefore strengthens the argument that the explanatory potential of the mobilisation framework is weakened by its reliance on broad and general notions of injustice (Atzeni 2009;Cohen 2011).
This article also contributes an analysis of collectivism and social identity. Collectivism was shaped by the physical locations of the workplace, the organisation of work and the labour process of care. The labour process often meant that a 'carer' identity overrode any identity as undervalued workers. This form of collectivism was, perhaps, 'not collectivism but coping' (McBride & Martínez Lucio 2011: 803, italics in original) -connections with care recipients assisted workers in coping with low pay and insecure contracts. This collectivism with 'customers' (Korczynski 2003) adds to the trade union, workplace and social (external to work) forms of collectivism identified by Stephenson and Stewart (2001). However, this research suggests that it is a form of collectivism which does not necessarily strengthen other forms of collectivism. Furthermore, any 'worker' collective identity apparent in interviews did not neatly align with an in-group of workers and an out-group of employers. Non-members reflecting on hierarchies within the workplace tended to criticise staff who had been employed at the company for a longer period of time or to criticise office workers. These office workers had qualitatively different responsibilities to frontline workers, with aspects of managerialism such as controlling the allocation of paid working hours, but they too were under pressure. In contrast, union members were simultaneously critical of specific organisational issues relating to office workers and critical of managers and their employer. They expressed distinctions between groups as power differentials -although generally excluding state power -rather than as individualised instances of bullying or favouritism. Exploitation factored into some workers' perceptions of collectivism. While non-members did not seem to feel that they were exploited by an out-group (as predicted by Kaufman 2018), exploitation did play into unionised workers' conceptualisation of their work.
For their part, union organisers across the three unions considered workers' reluctance to recognise exploitation to be an important reason for low levels of membership in the sector. Exploitation in social care is, however, difficult to identify given that the sector combines financial extraction (privatisation, and in some instance financialisation) with an imperative to save money and depress government spending. This combination not only affects working conditions and quality of care, it also leads to an obfuscation of blame. As an employer candidly expressed it when interviewed: 'we blame the local authority and the local authority blames central government'. Kelly (1998) does acknowledge that attribution of blame can go beyond an employer/worker relationship, noting that 'normally [the] agency would be the employer, although it might also be the state' (p. 127). The findings of this article suggest though that the varieties of agents identified as responsible for injustices -this 'crucial' attribution aspect of the framework -need to be further interrogated and contextualised. This process of attribution appeared to play a more important role in mobilising workers than the extent of the injustices or the strength of collectivism; the union members interviewed were not necessarily having a worse time at work, but their willingness to blame 'the boss' was apparent.
This research also found that restructuring in the sector under austerity has individualised responsibility for care quality and intensified the notion that caring labour is primarily altruistic, further impacting mobilisation. Non-members felt that their capacity to fulfil their responsibilities would be threatened by strikes. Responsibility for care comes to be transferred onto the 'affective capacities' (Dowling & Harvie 2014) of individual workers, alongside other aspects of social reproduction. Workers then come to perceive strikes as a moral failing in a context of care (Huget 2020). Among the nonmembers interviewed for this study, this perception contributed to a disconnect between collectivism and union membership, demonstrating that collective identity is 'contested' (McBride Martínez & Lucio 2011, 2021. Individualisation of responsibility was decried by members and organisers as a major impediment to mobilisation. This impediment can be connected to divergent ethical frameworks. Legault and Weststar (2015), in their study of video game developers, emphasise that an ethical framework rewarding meritocracy can conflict with the egalitarianism of unionism. The non-members interviewed in this research similarly emphasised a different framework to economic egalitarianism, in this instance prioritising an ethic of care. They viewed union membership as an individualist self-interested pursuit which was antithetical to the particular collectivism built in the workplace. Cohen's (2006: 218) argument that it is misleading to counterpose an 'idealistic' notion of social justice against 'the existing reality of "self-interested" workplace struggles' is relevant here: '"self-interest" contains within it the core of struggle against much broader issues of injustice and inequity'. The non-members understood workplace struggles as only self-interested.
As an examination of the mobilisation model then, this research emphasises that injustice, attribution and collective identity all need to be contextualised within employment relationships, material conditions and forms of labour to evaluate how these factors relate to unionism. A capitalist mode of production generates similarities in worker experiences of exploitation and collectivism, yet this does not mean that concepts such as mobilisation can be applied generically to any context of 'work'. The research also provides insight into the organising actions of unionism. Some unions have sought to reframe workplace issues (see Murphy & Turner 2014) in an effort to align different ethical frameworks of injustice. UNISON's (2012) 'Ethical Care Charter' could be seen as an example of this, emphasising that 'working conditions are intrinsically bound up with the quality of care' (p. 6). However, there are impediments to framing mobilisation around a collectivism between workers and care recipients. Placing an emphasis on care quality can contribute to the damaging narrative that caring labour is altruistic. This view, as one non-member and care assistant expressed it, suggests that 'everybody should be working together, and care about these guys [care recipients] more than themselves'. Referring to this altruistic narrative, an organiser said that 'the carers have to value themselves' to enact change. Alongside workers valuing themselves and valuing their labour, mobilisation requires workers recognising that there are agencies responsible for undervaluing their labour -whether those agencies are employers, the state or both.
It is worth noting that it is difficult to definitively contrast perspectives of members and non-members given that the views of members are impacted by their union membership. Identifying causality becomes difficult. It is unclear whether workers join a union because of their existing views on injustices, or whether their views are formed as a result of their union membership. A key aspect of Kelly's conceptualisation of mobilising is that unions (or individual activists) can 'frame' issues to encourage mobilisation. A limitation of this research is therefore that it has not assessed workers' perspectives before or after joining a union to establish shifting perspectives on injustice and collectivism. It has also focussed primarily on injustice and forms of collectivism as opposed to other aspects of Kelly's framework. It has not focussed on, for example, the role of leaders, or on workers' opinions of the effectiveness of unions. In relation to these aspects, the findings of the broader research project suggest that often workers were unaware of unions rather than averse to them. Participation and activism among workers, viewed as essential to strong unionism by McAlevey (2014McAlevey ( , 2016, was found to be limited: the workers interviewed emphasised they were generally too 'time poor' to participate. Organisers struggled to establish networks of activists, despite the varied tactics used by GMB, UNISON and IWW. Further research could examine contexts where activism is present, and look at how leaders navigate the varying ethical frameworks and views on injustice held by workers. To conclude, this research has found that injustice and collectivism need to be contextualised to improve understanding of mobilisation. Kelly (2018) contends that mobilisation is shaped both by the 'patterns of exploitation' under capitalism and the contingencies referred to in his framework; this article illustrates how, in a context of care provision, contingencies of beliefs shape workers' experiences and their propensity towards unionism. In addition, this research suggests that some patterns of exploitation can be reinforced by -rather than challenged by -particular forms of collectivism and moral understandings of injustice.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the Economic and Social Research Council, grant ES/J500215/1.