Social media: A challenge to identity and relational desistance

Society has witnessed a rapid growth in the prevalence and use of social media. The influence and impact of this expansion has sparsely, if at all, been considered within the context of desistance from crime. This article draws upon the narratives of male and female service users subject to community supervision by a Community Rehabilitation Company, collected as part of a doctoral thesis. Evidence demonstrates how social media plays a largely negative role, with some gendered difference on a service user’s identity and relational desistance, alongside individual agency and wider social structures. This enables the article to develop the evidence base of interactionist perspectives of desistance, while also recognising gender.


Introduction
Society has witnessed a rapid growth in the prevalence and use of social media, creating interactive online spaces where news is instantaneous, articles can be shared, and users have the ability to interact and comment on content. The research underpinning this article shares the narratives of individual service users during a defined period of community supervision. The narratives voiced were part of wider research within a doctoral thesis. This considered the fundamental role of positive, valued, and quality relationships in the process of desistance and the enabling and constraining factors they provided for co-produced rehabilitation, while recognising gendered difference. Focus here is given to an original finding of this wider research drawing attention to the role of social media in the desistance process of individual service users. It draws particular attention to identity and relational desistance and the links between agency and structure to theoretically develop interactionist perspectives which also recognises gender. To date, literature and research on desistance has sparsely, if at all, considered how and why social media may influence the desistance process. Taking a gendered comparative approach, the narratives of men and women highlight the negative effect of labelling and stigmatisation; the limited agency that individuals possess when their offending, or personal information, is shared within online platforms; and the challenges faced when social media undermines a sense of belonging and access to social capital in wider structural environments. However, the article also evidences that for some, social media can provide a supportive place where identities can be re-established and positive, valued, and quality relationships can be created or maintained. Therefore, recognition of individuality and the unique nature of an individual's process of desistance is important, in line with previous research (Maruna, 2001;McNeill, 2019;B. Weaver, 2019). Through raising these research findings and shedding light on this new and important research area, it is hoped that further exploration can be undertaken to develop both academic and practice-based understanding, which considers the role of social media in desistance and how this might differ through individual intersectionality.

Expanding our understanding of desistance
Theoretical work on desistance offers explanations of how and why an individual service users criminal behaviour changes overtime. However: . . . the process by which an individual desists from offending is invariably difficult to unravel and understand, and encompasses a whole range of personal and contextual changes. (Priestley and Vanstone, 2019: 335) Commonalities highlight how it is a continuous process (B. Weaver, 2019) which is not focused around a single change in behaviour and recognises that 'one person's reason for changing their life . . . might be another person's reason to escalate offending' (Maruna, 2001: 25). An essential common element of an individual service user's desistance is their ability to learn from mistakes and failure, while at the same time building on their success (Priestley and Vanstone, 2019). What is of fundamental importance in developing desistance-based literature is a consideration of how positive change occurs for individual service users (Maruna and LeBel, 2010;McNeill, 2006). This is especially when considering how changes can be enabled, or in fact constrained, and it is important to integrate findings across gender to create a stronger understanding of desistance and the role of agency and structure (Rodermond et al., 2016).
As the process of desistance is far from linear, Nugent and Schinkel (2016) have identified 'act desistance', relating to non-offending behaviour; 'identity desistance', which considers the internalisation of a non-offending identity; and 'relational desistance', demonstrating how change is recognised by others (Nugent and Schinkel, 2016: 570). Much of the previous work on desistance focuses on a male perspective (Barr, 2017;Harding, 2017;Rodermond et al., 2016); therefore, this article considers both identity and relational desistance, across gender, in attempts to expand current knowledge and understanding. Identity desistance stresses the role of an individual's identity and their capacity to exercise agency (Paternoster et al., 2016) and how this influences one's ability to assume their role as a nonoffender (Nugent and Schinkel, 2016) through their redemption scripts (Maruna, 2001). For women, their capacity to exercise agency is suppressed through patriarchal oppression, evident within both wider society and the criminal justice system (Barr, 2017;Hine, 2019). This also draws links to women's perceived vulnerability (Corston, 2007;Ministry of Justice, 2018), resulting in their limited choices (Schwartz and Steffensmeir, 2017). This is a vital consideration as individual's actions and their choices are regarded to be expressions of their identity (Paternoster and Bushway, 2009); therefore, it is important to recognise how this is then reflected through an individuals offending (Maruna, 2001).
However, desistance is not solely reliant on the individual, with their identity also being shaped by wider external and social factors. A service user's relationships can provide a supportive 'hook', and the scaffolding for significant life changes (Giordano et al., 2002), through offering effective opportunities for relabelling (McNeill and Weaver, 2010;B. Weaver, 2012). These hooks can also be defined through, for example, the role of legitimate employment (Farrall et al., 2010;Giordano et al., 2002) which enables access to social capital, and therefore positive opportunities which can influence desistance. For women, research demonstrates that to access available hooks, identities must fit within normative social behaviours in relations to gender (Broidy and Cauffman, 2017) with structural and patriarchal barriers in place (Barr, 2017(Barr, , 2018Brown and Bloom, 2009;Corston, 2007;Singh et al., 2019). Relational desistance recognises how the process is one which can only be understood within the context of relationships (Farmer, 2017;McNeill and Weaver, 2010;McNeill et al., 2012) and is therefore most likely achieved within them (Maruna, 2001) through feelings of inclusion (McNeill, 2009). However, it is important to recognise the challenges presented in the relationships of women within the criminal justice system and their differing links to offending behaviour. Every individual has a unique set of relationships (Borgatti et al., 2009), which are accumulated, maintained, challenged, and developed in a unique way dependent upon the individual service user (Rutter, 2019). Relationships form part of everyday life (Smart, 2007), involving interactions that are both enabling and constraining (Kadushin, 2012), impacting opportunities to connect with others (Mental Health Foundation, 2016).
When considering the role of relationships, in the process of desistance, it is important that the interplay between agency and structure is also recognised, drawing together identity and relational desistance and building upon an interactionist perspective. This emphasises; the significance of subjective changes in an individual's sense of self and identity, and as part of that, their aspirations, in response to their (changing) social contexts. (B. Weaver, 2019: 643) An important recognition is that agency is considered as a dynamic interaction between an individual and their social world (Healy, 2014;Giordano et al., 2002), dependent upon situational circumstances (King, 2014), and an individual's sense of belonging in a moral community (Burke et al., 2019). However, research evidences differing levels of agency and access to structure dependent on gender (Barr, 2017;Schwartz and Steffensmeir, 2017). Therefore, any attempts at understanding the process of how and why desistance occurs needs to be considered within individual, relational, cultural, and structural contexts (B. Weaver, 2019). Here the role of social media has yet to be explored.
Crime reporting is a long-standing news staple (Greer and McLaughlin, 2017), with movement from more traditional print-based material, to online reporting and sharing of crime news stories on corresponding social media platforms which are then shared by people not usually involved in the production of news. Across these changing media platforms, there is a continuation of repetitive messages surrounding common criminality and those who commit crime (Hayes and Luther, 2018). What is disseminated among the media is seen to shape experiences (Presser and Sandberg, 2015), as coverage generally pays more attention to punishment than prevention (Hough and Roberts, 2017), or rehabilitation and therefore is limited in its potential to support desistance. As a primary source of information, the changing nature of reporting and sharing news highlights the importance of considering questions of why and how the media socially construct crime and how this effects public understanding (Baranauskas and Drakulich, 2018). Media reports are increasingly available to view online, with this news source consistently and substantially on the rise creating long-lasting effects where judgement and discrimination is felt by individual service users long after conviction (Stacey, 2017). The labelling of an individual (Becker, 1963) as an 'offender' creates a false image of a permanent 'outsider' who is unredeemable (Priestley and Vanstone, 2019: 338). It also challenges the opportunities for individual service users to internalise positive changes to their identity but also challenges their social integration (B. Weaver, 2019). Through the formulation of 'criminal women' (Carlen, 1985), there are cultural assumptions about acceptably feminine behaviour (Walklate, 2011) which results in higher levels of societal stigmatisation (Estrada et al., 2012) a more arduous journey for women (Barr, 2017). The public exposure of previous criminal behaviour carries a specific type of shaming and stigmatisation, that is ambiguous and flexible, and evidences a long-lasting violation of privacy which removes an individual's control over their identity and therefore a degree of agency (Lageson and Maruna, 2018). This then impacts on both identity and relational desistance and how positive change can be recognised by others (Nugent and Schinkel, 2016), highlighting the medias influence of public images of desisters, alongside crime and the criminal justice system more widely (Hayes and Luther, 2018). However, the media is seen to distort the true picture of crime, contradicting official statistics and leaving the detrimental impression that crime is on the increase (Greer and McLaughlin, 2017).
Social media has provided a platform which enables and encourages public integration (Jewkes, 2015) but can also leave individuals feeling more isolated, especially when they have been subject to negative experiences (Primack et al., 2019). Through social media platforms, the wider public have the ability and opportunity to integrate themselves in real-time comment threads and discussion groups, which results in their direct involvement in the news process (Greer and McLaughlin, 2017), from the privacy of being behind a screen (Lageson and Maruna, 2018). The online spaces and interactions created by social media can challenge the prosocial identities established by individuals engaged in a process of desistance, both internally and externally. Involvement, and interaction, on social media platforms range from comment and sharing features to single click reactions that enable uses to express an emotional response without text (Spottswood and Wohn, 2019). The online naming and shaming of individuals convicted of crimes subjects them to embarrassment and condemnation, representing an example of digital vigilantism, albeit an area of limited research (Dunsby and Howes, 2019). This demonstrates features of online trolling, defined as a 'deliberate, deceptive and mischievous' attempt to provoke reaction (Golf-Papez and Veer, 2017: 1339), causing substantial and negative consequences (Howard et al., 2019). Online abuse, made possible through the interactive nature of social media, has been fundamental in shaping its contemporary landscape and is recognised, through a dystopian perspective, as inherently harmful, by lending itself to this (Salter, 2017). If these negative experiences of social media are felt by service users undertaking community supervision, and within the criminal justice system more widely, individuals may face difficulties and challenges in internalising a non-offending identity and near impossible to have this reflected back to them.
However, social media can also provide a space which can offer an alternative way for service users to shape their own identities (van Dijck, 2012) and therefore demonstrate agency, connect with people (Jewkes, 2015), and have their identities recognised and accepted by others. Through a utopian account, social media is regarded as an inherently beneficial force (Salter, 2017). Within these online platforms, individuals are able to reimagine their identities through a participatory strategic performance (Siapera, 2018), and when considering the interlinking role of agency and structure, can support in the process of desistance. The participatory relationship enabled through social media evidences how it may provide a platform for interaction that can either challenge identity and relational desistance or more positively create a space where desistance can be validated and reflected back to individual service users.
Despite the growing societal dependence of digital technologies, community supervision alongside the criminal justice system largely ignores the impact of social media when attempting to encourage desistance from crime (Reisdorf and Rikard, 2018). There is limited understanding of how social media can be an additional area of navigation for service users, highlighting the minimal acknowledgement or support available through community supervision (Hayes, 2015). To ensure professionals provide 'thick' supervision, which is embedded in the community to help enable service user's build their social capital and relationships (Dominey, 2019), a recognition of social media is important. In addition, society has a degree of responsibility in removing the public stigmatisation attached to service users of the criminal justice system (Seaman and Lynch, 2016), and therefore those subject to community supervision. If achieved, this would support a reflective process of relabelling (B. Weaver, 2012) to occur to enable identity and relational desistance. Societies response provides the opportunity to 'bear witness' to desistance, which does not have to happen face-to-face, a vital component for individual service users to become active agents in their own desistance narratives (Anderson, 2016). Therefore, there is the need to develop stronger evidence to support a practicebased understanding of the role social media may play in challenging, or encouraging, the process of desistance across gender. It is also important to consider how it may alter the relationships surrounding an individual service user to develop both theoretical and practical understanding of identity and relational desistance and their links to agency and structure.

Providing a voice through narrative expression
The underpinning research design was qualitative in its approach, focusing on the experiences of participants (Sword, 1999), with deep immersion into the social scene of both community supervision and the lives of participants. This produced 'rich', 'holistic' data (Maruna, 2010: 12) through the narratives of service users subject to community supervision. The underpinning philosophy of the research placed the experience of the service user at centre stage, recognising that practitioners and researchers need to hear and attend carefully to their views in any attempts to influence (Burke et al., 2019). Overall, the participants shared difficult and emotional narratives about their involvement in the criminal justice system, drawing upon the complex relationships that surrounded them as individuals.
Larsson (2019) has identified how narrative research has increasingly gained the attention of criminologists interested in understanding criminal behaviour from a service user perspective. Narrative interviewing is common within desistance-based research (Maruna, 2001), with the voice of those holding experiences of the criminal justice system being integral to studies of desistance from crime (A. Weaver and Weaver, 2013). As individuals, our narratives are central to human experience and constantly change and evolve over time (Presser and Sandberg, 2015), mirroring the dynamic process of desistance. Storytelling through narratives is an important methodological approach, but it can also be conceived as an intervention which supports the process of desistance (Bove and Tryon, 2018). This is because it provides an opportunity for individuals to shape their own identity, supporting them to escape hardship (Presser and Sandberg, 2019).
Fieldwork was carried out in one community supervision office, and corresponding women's centre, within a Community Rehabilitation Company in the north of England. It focused on the narratives of 25 service users, over a longitudinal time frame between late 2017 and early 2019. Opportunistic sampling was adopted, using board criteria which only required that participants were subject to a community order, with at least 6 months remaining at the point of engagement. Participants were gradually recruited as they expressed interest (Teddlie and Tashakkori, 2009) in the research themselves or through professionals working with them. The participants comprised of 12 men and 13 women to enable integration across gender (Rodermond et al., 2016). At the point of engagement, demographics and offending histories were similar, although this was not an explicit aim. Male service users involved in the research were aged between 19 and 49, with an average age of 38. Their offending background ranged from 0 to 36 previous convictions, averaging at 11. The female participants were aged between 18 and 58, with an average age of 36. Their offending histories ranged from 0 to 38 previous convictions, averaging at 10. The most recent offence categories were also similar across gender and included, for example, shoplifting, common assault and battery, and possession of an offensive weapon. It is important to note that participants were assessed as being low risk because their supervision was conducted by a Community Rehabilitation Company. Also because of the longitudinal design underpinning the research participant demographics changed over time. There are limitations of the small sample size often involved in qualitative research and it should not be assumed that the sample was representative of individuals on community supervision. In addition, those individuals difficult to reach, and maintain contact with, are likely to be unintentionally excluded.
When considering the role as a researcher, it was important to recognise and take responsibility for this position, and the effect this may have had on the creation of knowledge and possible research biases (Berger, 2015). Awareness of this position in the lives of participants, and the research field-sites, evidenced a continual negotiation of the 'insider'/'outsider' boundaries (Jewkes, 2014). The substantial amount of unstructured data collected required preparation prior to analysis (Boulton and Hammersley, 2006). Therefore, the narrative of each research participant was audio-recorded and fully transcribed. To ensure careful analysis of data, NVivo was used (Foster, 2006), which supported organisation and management, resulting in surfacing ideas being formulated into a series of codes (Gibbs, 2007). These were grounded in the data through open coding and an inductive approach to analysis (Lewins and Silver, 2007) to which social media emerged as an influencing factor within identity and relational desistance.
The research was conducted after the implementation of Transforming Rehabilitation. This context should be considered with regards to service delivery, and the challenges evidenced by previous research within this working environment (Albertson and Fox, 2019;Canton and Dominey, 2018;HMIP, 2019;Robinson et al., 2016;Sukhram, 2015). Ethical approval was granted through the host university, and all relevant risk assessment paperwork was completed to satisfy the Community Rehabilitation Company and ensure potential harm avoidance to research participants. Prior to agreement to take part in the research, participants were provided with all relevant information about its purpose; how the information they provided would be recorded, stored, and used; and their right to withdraw from the research at any time. Each service user who agreed to take part in the research had the opportunity to decide on their own pseudonym, as this avoided the act of power recognised when renaming research participants (Guenther, 2009). These chosen pseudonyms have been used in this article to ensure confidentiality and anonymity.

In a world where you can be anything #bekind
The narratives of individual service users provide evidence to highlight that social media influences the process of identity and relational desistance, which, in turn, impacts on both individual agency and wider social structure, for both men and women. Some narratives also draw reference to more traditional forms of printbased media. However, there are some overlaps in discussion due to print-based articles also being published and shared within online spaces and corresponding social media platforms. This enables public interaction, evidenced to influence identity, agency, wider social structures, and sense of belonging. When in conversation, participants of both genders shared narratives and experiences of their involvement, and engagement, with social media platforms where news articles about them were visible. This influenced their sense of self, their confidence in their wider community, employment opportunities, and loss of relationships in which they deemed positive and that previously provided them with support (Rutter, 2019).
Social media created opportunities for labelling to occur demonstrated in the narratives of individual service users. Specifically, when their personal details were shared on social media platforms experiencing feelings, and an identity, of being an unredeemable outsider (Priestley and Vanstone, 2019 Narratives also demonstrated the limited public sympathy for individuals involved in criminal behaviour (Lageson and Maruna, 2018) when displayed across media platforms.
obviously when I got put in the [local paper], I got a lot of hate then . . . like obviously people that didn't know me was giving me a lot of s**t . . . but then I just changed me name on Facebook . . . it's the way they put in the paper want it, like it said 'out-of control mum of one', omg b**tard. (Karen) The narratives of Steven and Karren demonstrate a reiteration of gendered identity, which can suppress opportunities for desistance; each narrative demonstrates evidence of agency; however, Karen shows a stronger ability to make surface-level changes to her identity through changing her name of social media platforms. When doing this, she also unfriended and blocked numerous contacts to enable herself to be surrounded by online support and relationships that reinforced Karen's identity and relational desistance.
The prominent influence of social media platforms created an additional area of navigation as the labels and stigmatisation that were created extended beyond those attached by the state through conviction and sentence and were very much in the public domain. For the individuals involved in the research, this often presented an additional barrier in being able to internalise a positive identity, assuming their role as a non-offender and also having this identity reflected back and legitimised by others (Nugent and Schinkel, 2016;B. Weaver, 2012). The reason behind Karen's changing approach to her use of social media.
For Elvira, her sense of self and ability to establish a prosocial identity which was reflected back to her was challenged through the manipulation and creation of rumours on social media platforms. These rumours reaffirmed her offending identity and demonstrated stigmatisation. Similar to Steven, Elvira's experience evidences how social media removes an individual's agency, as they have no choice or control over what is being said or shared about them within these online spaces. Nor do they have any influence over how others may interact or comment on these public posts, or articles, demonstrating limited control over identity, alongside the ambiguity and flexibility of online criminal labels and stigmatisation (Lageson and Maruna, 2018 For Elvira, her narrative not only evidences the challenges to her identity desistance, alongside her agentic capacity, but also her access to wider social structures. Online activity created difficulties in Elvira's relational desistance. Her offending identity was reinforced and reflected back to her, having a direct impact on her employment, a known hook for change and desistance (Giordano et al., 2002).
I did get a job after it . . . but when it all came out in social media and everything what I had done, they had to let me go . . . there was too much in the social media . . . I was all online . . . and a lot of people put where I worked and stuff . . . I just wish that it hadn't gone in the paper and I might still be working where I was working. (Elvira) Elvira's narratives emphasise the simultaneous role of agency and structure, developing gendered recognition within interactionist perspectives. The concerns expressed through individual narratives were not only about how service users felt within themselves but also about the impact of online posts and the interactive nature of social media on their positive relationships. This added an additional strain to their process of identity and relational desistance, by limiting their ability to achieve a prosocial identity and negatively impacting their sense of belonging within both their community and in specific relationships. The narrative of Sarah summarises this.
So, nothing had been in the paper and over a week had gone by and suddenly it was in the paper. So, nobody knew about it and they [extended family] had given them all this information . . . they had put this sort a half a page thing in the paper with some really old photographs . . . so they did this spread in the paper and obviously it was extremely upsetting and I think I was more worried about, not me . . . coz at the end of the day I've done it and if anyone wants to confront me about it . . . they want to ask me I will tell them the truth and if they want to ignore me then that's up to them . . . I was more concerned about the influence on me husband at work. (Sarah) The ease of sharing information on social media also resulted in participants becoming victims to online trolling, and experiences of humiliation, which is complicated by uncertainty of what information may be shared of revealed about them (Lageson and Maruna, 2018). The narratives shared evidences of the negative impacts on a prosocial identity through reinforcing criminal identities and the creation of additional barriers to both identity and relational desistance. Within each of these contexts, it is important to consider an individual's sense of self and their confidence, relating to their capacity to exercise agency, expanded through the narrative of Sarah.
Yeah but it affects just normal things in your life . . . it's like I didn't go out, I wouldn't go out . . . I wouldn't go out for a meal I wouldn't go out for a drink because I just felt really self-conscious of me self because I thought what if everyone is looking and talking . . . because they had put that s**t in the paper . . . but you know what I mean it was that . . . what if I go somewhere and there is somebody else in there that knows me and they are looking at me . . . so I was extremely self-conscious even about going to the shop, going to the supermarket . . . . (Sarah) The narratives of participants demonstrate media stigmatisation, evidenced as a pain of desistance which is recognised to be unsupported by the practices of community supervision (Hayes, 2015). In addition, society has a role to play through accepting and supporting individuals in their process of change, within prosocial communities if desistance is to be at all possible (McNeill et al., 2012;McNeill and Farrall, 2015;McNeill and Weaver, 2010;Shapland et al., 2016). This is challenged by social media. The analysis of participants' narratives provide evidence to show how the role of both agency and structure are interlinked by the influence of social media. This impacts on the process of both identity and relational desistance, highlighting some differences in gender but fundamentally reiterating the importance of individual intersectionality.

Some publicity is good publicity
Participants also described, albeit not as frequently, how social media provided a positive space for them as individuals. Liz explained why and how she engaged with particular online platforms, which evidenced the wider role these spaces played for individuals.
[I] enjoy socialising on Twitter or Facebook, reconnecting with old friends, it's nice to see how people have done in life or sharing positive things about my life and people been happy for us. Love following the celebs on Twitter and also the news from all round the world. I learn a lot from social media and it broadens what I know. (Liz) When talking about the role of social media in their lives, a small minority of participants felt it supported them to develop a prosocial identity. Karen's narrative has already demonstrated how her interaction with social media enabled her to shape her own identity. To expand, Karen's identity was reaffirmed by others enabling her to publicly bear witness to desistance (Anderson, 2016) and gain support and encouragement in challenging stigmatisation and recognising her possible self through the participatory relationships created. This evidences changes in both identity and relational desistance. Karen: I think as well because I am doing good and people are seeing good I seem to be getting rewarded with things more . . . like respect, like I am getting a lot more respect. Natalie: Are you getting that from a lot more people? Karen: Yeah loads of people, like people in the street have come up to me that I didn't even know who have been on me Facebook and said how well I am doing. And I'm like WOW . . . got support of people that I don't even know.
Across gender, participants also expressed a range of narratives evidencing how social media provided a mechanism which enabled them to establish a sense of belonging and maintain positive personal relationships, known to facilitate desistance. They expressed feelings of being part of a community, albeit an online one. Here there are some clear gendered differences in how participants discussed their positive use of social media. Women focused on how it provided them with a mechanism to maintain contact with people already within their lives. For men, social media was used differently, in an attempt to extend their relational networks, or metaphorically as a possible means of escape.
When service users had positive experiences of social media, it provided them with a space which offered a way for them to shape their own identities (van Dijck, 2012), connect with people (Jewkes, 2015), and have their identities recognised and accepted by others. Within this online space, some participants were able to reimagine their identities through a participatory strategic performance (Siapera, 2018), evidencing agency, and access to wider social structures. This participatory relationship evidences how social media may provide a platform for an individual service user to make positive changes in their identity and relational desistance.

Supportive service user interactions with social media
This research suggests that critical thought needs to be given to the role of social media in lives of those undertaking community supervision and involved within the criminal justice system more widely. Practices of community supervision should work to develop how they can consider the impact of social media on service user relationships and the process of identity and relational desistance across difference intersectionalities.
Research findings, suitably anonymised, were used to structure feedback sessions with the hosting Community Rehabilitation Company. Local management recognised that they had not thought about social media within current service delivery but agreed that it was an important issue. This reiterated the work of Hayes (2015) which identified a pain of desistance relating to media stigmatisation which was unaffected by community supervision. It was suggested by Community Rehabilitation Company management that the role and impact of local and/or social media could be considered within a defined area of their service delivery. Discussions also highlighted how work to develop this could be included in a Rehabilitation Activity Requirement, a contested replacement of the Supervision Activity Requirement introduced under Transforming Rehabilitation (Robinson and Dominey, 2019). If put into prescriptions for practice, the aim would be to support service users who felt they were negatively affected by the increasing changes from traditional print-based material, to online reporting and sharing of crime news stories by corresponding social media platforms and the wider public.
The narratives of participants in this research have demonstrated the far-reaching impact social media has on their lives through their capacity to exercise agency, their relationship dynamics, and their wider structural situations through access to social capital. Social media most frequently reduced participants' capacity to exercise agency as they had limited control over public media outlets and what was published or shared about them. Participant narratives have also highlighted how media reporting and perceptions have shaped and altered their relationships with the wider community. Participants described invasions of privacy where their pictures were shared alongside media articles. They also expressed how this negatively influenced their access to social capital, through for example employment. All of which influenced their identity and relational desistance.

Concluding reflections
This article has highlighted the limited research into the role of social media on the process of desistance for individual service users during a period of community supervision. The narrative data collected as part of a doctoral study have evidenced the negative effect of labelling and stigmatisation; the limited agency that individuals possess when their offending, or personal information, is shared within online platforms; and the challenges faced in an individual's sense of belonging and access to social capital in their wider structural environment. This demonstrates the barriers faced in both identity and relational desistance. In addition, it is important to note that participant narratives did not demonstrate any direct links between their use or engagement with social media and their offending behaviour. This provides an additional area of research in developing wider understandings of the role social media plays in both offending behaviour and its influence on desistance. Narratives have also shed light on how social media can provide a supportive place in which identities can be re-established and positive relationships can be built or maintained. Therefore, recognition of individuality and intersectionality is important, in line with previous and developing research into desistance.
Social media is part of the fabric of modern life, so it is unrealistic to expect those undertaking community supervision not to engage with online spaces and platforms. It is important to harness the positivity that can be created and received on social media and within online spaces if desistance is to be supported, but the damaging impact of negative social media interactions also need to be addressed. As researchers, and practitioners, working in criminal justice, it is our responsibility to begin to recognise, understand, and support the harm, pain, and hurt caused by public labelling and stigmatisation that takes place within online spaces and the challenges this creates for an individual's agency and their wider structural circumstances. Further exploration of the role of social media in the process of identity and relational desistance is needed, and taken through an interactionist perspective would enable the recognition of the interlinking elements of agency and structure and have the potential to offer prescriptions for practice.

Acknowledgements
Firstly, I would like to thank the individuals whom shared their stories with me. In addition, I thank Professor Chris Fox for the continued encouragement and support for this article alongside the insightful comments provided by the reviewers.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received financial support for the research, authorship, and/or publication of this article: The author received joint funding from Manchester Metropolitan University and Interserve Justice to undertake the doctoral research underpinning this article. No financial support was received for the authorship or publication of this article.