Transformative agency and desistance from crime

Introduction

Many recent accounts have suggested that desistance from crime involves an interplay between agency and the socio-structural context within which the individual lives (Bottoms et al., 2004; Farrall and Bowling, 1999; Farrall and Calverley, 2006; Giordano et al., 2002). However, despite this recent focus within the literature, accounts of agency remain vague and under-explored (Healy, 2010), and there has been relatively little attention paid towards the effect of changing socio-structural contexts (Farrall et al., 2010). Furthermore, the exact configuration and interaction of individual and structural factors is uncertain. LeBel et al. (2008) outline three possible models to explain this interplay. Within ‘strong-subjective’ models of desistance the emphasis is upon the motivation and willpower of the individual concerned: ‘One need only decide to change and envision a new identity for oneself in order to go straight’ (LeBel et al., 2008: 138). However, this proposition suggests a concession to voluntarism, implying that individuals are capable of transforming or reproducing their social environment with free will, which is not the case (Elder-Vass, 2007: 26). On the other hand, ‘strong social’ models of desistance emphasize the importance of life-course turning points, such as marriage or employment, and ‘it is the arrival of these events, which are largely outside of an individual’s control, that will best predict success … rather than the mindset of the individual’ (LeBel et al., 2008: 139). Similarly, Farrall and Bowling (1999), drawing upon structuration theory, argued that ‘critical situations’ can have an effect on providing the structural changes which could precede desistance from crime. Several authors have argued that interactional models of desistance may offer the most effective explanations of desistance (for example Farrall, 2002; Maguire and Raynor, 2006; McNeill, 2003; McNeill and Whyte, 2007). LeBel et al. (2008) conclude by suggesting that a ‘subjective-social model’, offers the best explanation of desistance from crime. They argue that ‘subjective changes may precede life-changing structural events and, to that extent, individuals can act as agents of their own change’ (2008: 155). In this regard it can be argued that human agency includes qualities of ‘transformational power’ (Simmonds, 1989: 187), such that individuals can exert some degree of control over the future direction of their lives.

However, any account of human action needs to give proper consideration to agency, but also needs to account for the role of structure in enabling or constraining that action (Marshall, 2005: 69). Explanations of desistance should also incorporate an understanding of both social and subjective factors. Indeed, McNeill and Whyte (2007) argue that desistance will be more likely if there is correspondence between social and subjective factors, each moving away from crime, but that desistance will be less likely if there is dissonance between these areas. They suggest that the more common experience would be some correspondence and some dissonance between each area (2007: 148). An individual may be motivated and determined to move away from a criminal lifestyle, but may lack employment opportunities or pro-social relationships at that time in their life. Conversely, an individual may have access to potentially life-changing social bonds (what Giordano et al. (2002) refer to as ‘hooks for change’), but may lack the necessary attitudes or motivation to take advantage of these.

The central challenge for would-be desisters, therefore, is to develop the necessary strategies that will assist them in developing their personal and social contexts in ways that will enable them to move away from crime and (re)integrate into mainstream society. This article proceeds from this starting point, and seeks to offer an explanatory account of how individuals develop these strategies and the effect that this has upon the actual behaviour that individuals enact. The article begins by exploring how it is possible to conceptualize desistance from crime as beginning with particular intentions for the future. These intentions may frequently entail the envisaging of a new identity, one which is different from the past self. The article then proceeds to a discussion of the difficulties associated with agency in relation to processes of individualization within a context of structural uncertainty. It is argued here that desistance from crime necessitates a particular temporal orientation of agency which allows the individual to conceive of the future in positive and efficacious terms. In order to achieve this, the article draws upon the theoretical work of Emirbayer and Mische (1998) to outline how agency can be conceptualized in processes of desistance, and two case studies are provided to illustrate this.

Desistance Intentions

Sampson and Laub’s (1993, 2005) structuralist argument essentially contends that desistance occurs as a series of side-bets, without the individual committing to any particular course of action. It has also been suggested that social changes that aid desistance may simply be chance events that occur without any impetus from the individual concerned (Laub et al., 1998: 225). If this is the case then desistance can be regarded as the outcome of various structurally induced turning points removing criminal opportunities without the would-be desister realizing it (Vaughan, 2007: 391). In other words, the would-be desister is one who does not create their own life, but one who merely responds to the circumstances that surround them. Indeed, Laub and Sampson (2003) argue that desistance most frequently occurs by default, with little reflection and without any meaningful intention (2003: 278–279). This suggests that agency is reactive, rather than purposive, which echoes Turner’s (1976) theory of the impulse self. Within this theory, individuals react to particular circumstances that they encounter, but do not develop any institutional commitments and there is, therefore, no reason to maintain a social bond if an alternative is more alluring. However, as behavioural intention is central to actual behaviour (Ajzen, 1991; Forste et al., 2010; Serin and Lloyd, 2009), it is unlikely that would-be desisters act in such passive and structurally submissive ways. Indeed, it is argued that desistance involves at least some form of conscious deliberation about future goals, and some evaluation of present contexts to determine how these could be achieved. As Maruna (2001: 147) states, successful desisters ‘had a plan and were optimistic that they could make it work’.

The existing literature, therefore, would suggest that the nature of a future orientation entails the envisioning of an alternative future self and new social roles that the would-be desister seeks to occupy (Giordano et al., 2002; Graham and Bowling, 1995; Maruna, 2001; Shover, 1983). Farrall (2005) has suggested that alterations to important social roles can play a crucial part in the development of a new sense of self. Similarly, Giordano et al. (2002) argue that desistance is achieved as the result of a series of cognitive shifts, in particular relating to attitudinal change and the envisioning of an alternative ‘self’. These authors suggest that agentic moves are a fundamental aspect of the desistance process. The individual needs to be ready to change and needs to be exposed to ‘hooks for change’, such as employment or marriage. Giordano et al. (2003), observing the interplay between marriage, peer association and persistent criminality, found that in the absence of motivation and commitment to a new ‘self’ individuals were more likely to ignore their spouse or leave the relationship altogether. Rumgay (2004) also argues that changes to personal identity may be strongly related to successful desistance, and that attempts to desist may result from the tension between the subjectively perceived past self and the desire to adopt a more conventional, pro-social identity.

This tension between past/present and future identities has been explored more recently by Paternoster and Bushway (2009), who argue that desistance requires the individual to discard their own identity in favour of an alternative one. This begins when an evaluation of the present self leads the individual to regard criminality as more costly than the benefits obtained. When this occurs the individual will actively seek to make changes to their identity and to their personal and social contexts. They argue that there are two potential identities that provide a motivational and a deterrent effect. The ‘possible self’ is the person that they could become if they make changes in their lives, and this is the identity that they are motivated towards. The ‘feared self’ is the person that they could become if they continue to offend, and this identity deters the individual from persistent offending behaviour. As opposed to Maruna (2001: 87), who argues that desisters ‘rewrite a shameful past into a necessary prelude to a productive and worthy life’, Paternoster and Bushway (2009: 1107–1108) argue that would-be desisters need to ‘cast off’ old identities in favour of alternative versions of the self. The perception of a ‘feared self’ provides the initial motivation to change, which will lead to a change in social preferences more concomitant with an emerging new identity. For Paternoster and Bushway (2009: 1152), therefore, human agency provides the all important ‘upfront work’ necessary to start the process of desistance. This will then lead to changes in social networks and social roles, which the authors refer to as ‘structural supports’ (2009: 1129) which will assist the would-be desister in making the transition towards conformity, as these structural supports will be more conventional and pro-social than the individual themselves. These structural supports (whether they are social roles, institutions or new social networks) will then provide the individual with the necessary pro-social opportunities to become the ‘possible self’.

This, then, raises the questions of how would-be desisters acquire these new identities, and how they learn how to act in role appropriate ways. Earlier accounts of desistance suggest that explanatory accounts can be found within rational choice theories (Clarke and Cornish, 1985; Cornish and Clarke, 1986; Cusson and Pinsonneault, 1986), such that: ‘the probability of desistance from criminal participation increases as expectations for achieving friends, money, autonomy and happiness via crime decrease’ (Shover and Thompson, 1992: 97, emphasis in original). However, while choice is one aspect of agency, such explanations do not adequately account for the complexity of desistance processes, or of the constraining effects of social and cultural contexts upon such decision making. While Paternoster and Bushway (2009) do not discount the effect that social structures, processes and relationships have upon the transitions towards and maintenance of desistance, it is made explicitly clear within their proposition that individuals intentionally and actively decide to cease offending and acquire new identities and social roles. Indeed, for them human agency is ‘intentional self change’ and individuals create their own futures, they are not coerced by external structures (2009: 1149–1150). In this respect, these authors argue, individuals acquire new identities by wilfully dedicating themselves to them. In this particular conceptualization of desistance, individuals decide that they no longer want to be ‘an offender’, and instead choose an alternative, ‘non-offender’ identity for the future, which may entail changes to socio-structural roles. Paternoster and Bushway (2009) argue further that desistance cannot take place in the absence of the prior agency work undertaken to identify a new replacement identity. This replacement identity should be more specific in nature, as opposed to a more generic ‘I want to be a good person’ vision, and as such the identity itself will provide the individual with the knowledge of how to act in role appropriate ways:

The possible self of a non-offender provides a current offender with a specific and realistic route, roadmap, or strategy to take in order to realize that self, and steps that can be deliberately taken to change one’s life in a way consistent with that self. (Paternoster and Bushway, 2009: 1149)

Rumgay (2004) placed greater emphasis on structural forces leading to positive identity change. She states that there need to be opportunities to fulfil new identities available within the immediate social environment, and that such opportunities need to be accessible and more highly valued than present circumstances (see also Giordano et al., 2002). The subjectivity of life-course events, such as marriage or employment, is also recognized by Farrall (2002: 11), and it is likely that recognition of opportunities and the value placed upon them is conditioned by motivation and self-confidence. These characteristics are strongly related to successful desistance (Burnett, 1992; Burnett and Maruna, 2004; Farrall, 2002), but it is likely that motivation and confidence are themselves conditioned by the social context. As such, persistent offenders who are more deeply embedded in criminal contexts are more likely to face circumstances which do not present many recognizable opportunities to fulfil new roles and identities and, therefore, do not generate the motivation and confidence required to desist. For Rumgay (2004: 408), recognizing new social roles as an opportunity to change one’s identity ‘marks the beginning of active attempts at personal change’. New social roles offer a ‘skeleton script’ (Rumgay, 2004: 409), providing the individual with behavioural cues as to how to act appropriately within the social role and how to act in accordance with the new identity. Giordano et al. (2002: 1055, emphasis in original) refer to these scripts as ‘a detailed plan of action or fairly elaborate cognitive blueprint’. While it is most likely the case that certain identities or social roles offer behavioural ‘cues’ or ‘roadmaps’ for how to act, the approaches referred to above do not adequately account for how individuals’ strategies for desistance develop during times of institutional or structural uncertainty, or in light of the challenging social contexts that have been well documented in previous research (Harper et al., 2004; Mair and May, 1997; Rex, 1999).

Such complex and challenging conditions have implications for some of the more common social relations associated with desistance – employment, relationships and contact with criminal justice agencies – and it is suggested that this may have deleterious effects upon the likelihood of desistance for many individuals (Farrall et al., 2010). However, the counterargument to this would be that under the same conditions possibilities of choice are more prevalent than ever before, and it is choice which is ‘implied in all definitions of agency: alternative courses of action are available, and the agent could have therefore acted otherwise’ (Hays, 1994: 64). Large scale institutional changes in areas such as welfare regimes, employment, education and family (Dean, 2003; Hamnett, 2003; Lash and Urry, 1994; Wacquant, 1999) have provided individuals with less predictability over the configuration of their life course. In order to navigate the life course, individuals need to construct actively and reflexively their own narratives in order to maintain desistance from crime. However, the availability of choices over the life course may be regarded in overly optimistic and universal configurations (Jamieson, 1999; Skeggs, 2004), and it is argued here that would-be desisters, who are likely to face significant personal and social challenges, may not be able to embrace the process of reflexivity in the manner which may be implied in some accounts of agency, identity change and desistance.

Reflexivity, Agency and Desistance Intentions

This article draws upon findings from research which explored the initial transitions towards desistance of a group of 20 individuals under probation supervision between 2009 and 2010. One of the primary concerns of the research was the exploration of the role of agency in primary desistance. As ‘qualitative data expose human agency in the process leading to individual change’ (Laub and Sampson, 1998: 229), a qualitative approach was deemed relevant for this study, and in-depth, face-to-face, semi-structured interviews were selected as the appropriate method. In order to understand the role of agency in transitions towards desistance the research was designed to elucidate how individuals perceived their personal and social contexts, their experiences of probation and the subsequent impact upon their prospective accounts of desistance. Thus, the research follows other recent studies of desistance which seek to capture subjective accounts of desistance, often through the use of narrative or in-depth interviews (for example Farrall, 2002; Farrall and Calverley, 2006; Giordano et al., 2002; Maruna, 2001). While some have criticized this approach for its lack of generalizability beyond the sample of interviewees (Massoglia and Uggen, 2007), its suitability can be found most readily in exploratory studies such as this.

Before interviewing began, meetings were held with the three probation teams who participated in the study, to recruit participants for the study and to discuss the various issues that had arisen from the literature review and the design of the research questions. These meetings acted as pseudo-focus groups and were used as an approximation of ‘interview trialling’ (Gillham, 2000), in order to highlight key issues and to make alterations to the interview schedule as necessary. Interviews were conducted at the probation office that the probationer would normally attend, and were digitally recorded and transcribed with the informed written consent of all participants. Interviews ranged in duration from approximately 25 minutes to 105 minutes.

One of the central aims of the research was to explore how individuals develop strategies to sustain desistance. In part, this reflects the aforementioned link between intentions and behaviour (Ajzen, 1991), but also because I contend that such strategizing is a fundamental aspect of primary desistance, which is itself an underexplored phenomenon. Typically, primary desistance refers to a crime-free gap or a lull in offending, while secondary desistance involves the assumption of a new non-offender identity (Maruna et al., 2004). It is argued here that research conducted with individuals who may considered to be primary desisters can provide a useful insight into the strategies and intentions held by individuals in this stage of the desistance process (a discussion of the merits of researching primary desistance is beyond the remit of this article, but see Healy, 2010; Healy and O’Donnell, 2008 for a rationale for this research).

In order to explore the extent to which would-be desisters embrace the reflexive self, a conceptualization of agency is required which can account for the effect of structural forces without reducing agency to a structurally determined phenomenon. Furthermore, a methodological approach is required which can distinguish between reflexivity at the time that decisions are made, and reflexivity which is rationalized after the event – a shortcoming of Giddens’ (1991) account of individualization over the life-course. It is argued here that these two processes need to be clearly distinguished, as reflexivity in the present tense will reflect, to a greater or lesser extent, the influence of structural inequalities and barriers in the social context. As such, prospective strategies for desistance were examined utilizing Emirbayer and Mische’s (1998) conceptualization of agency to analyse probationers’ narratives.

For Emirbayer and Mische (1998) agency is not something which is possessed, but rather a quality of engagement between the actor and their structural context. It involves three dimensions – ‘habit’, ‘imagination’ and ‘judgement’ – and, depending on the configuration of these dimensions, can reproduce or transform the structural context in response to particular problems presented by the social environment (Emirbayer and Mische, 1998: 970). Agency, therefore, is something that people achieve through their action, and it is the interplay between structure and action which determines the orientation of agency that will dominate at a given point in time. Indeed, ‘agentic processes can only be understood if they are linked intrinsically to the changing temporal orientations of situated actors’ (Emirbayer and Mische, 1998: 967). Crucially, these orientations are only analytical distinctions and each is present within all action, although not in equal measure, and to understand agentic processes necessitates an understanding of the interplay between these dimensions, and how the interplay varies according to the structural context of action. It is argued here that the temporal orientation of agency will alter depending upon the intentions of the individual and the circumstances of their social context. Hence, an intention to change coupled with a social context which includes enabling structural properties is more likely to encourage imaginative and future oriented agency.

The ‘iterational’ orientation of agency is where individuals perform actions based on routine. For Giddens (1984), routine practices are the cornerstone of all habitual action and provide individuals with a sense of ontological security. According to Giddens (1991) individuals have to locate constantly their own sense of ontological security, a point which is shared by Beck and Beck-Gernsheim (2002: 31) when they write that ‘modern guidelines actually compel the self-organization and self-thematization of people’s biographies’. Individuals will routinely try to maintain a sense of ontological security, or else they would be paralysed by anxiety, and this is most effectively achieved by establishing circumstances of familiarity and routine (Liddle, 2001: 56). The ‘projective’ dimension of agency is concerned with the intention of self-change. Drawing upon Mead’s notion of ‘distance experience’, the projective dimension of agency explains how individuals are able to distance themselves from past repertoires of action, including the habits and traditions of social structures that constrain their identities and activities. It is this capacity for distanciation that allows for individuals to undertake new schemas of action and alter life-course trajectories. The projective dimension of agency is, therefore, future-oriented as individuals ‘construct changing images of where they think they are going, where they want to go, and how they can get there from where they are at present’ (Emirbayer and Mische, 1998: 984).

The ‘practical-evaluative’ dimension of agency relates to the actions of individuals in response to the contextual demands of the present. In this respect, the practical-evaluative dimension of agency can be understood as the manner by which individuals incorporate knowledge of past actions, and their orientations towards the future, into the circumstances of the present. As Emirbayer and Mische (1998: 994) argue, ‘relatively unreflective routine dispositions must be adjusted to the exigencies of changing situations; and newly imagined projects must be brought down to earth within real-world circumstances’. Thus, crucially, the practical-evaluative dimension is not merely an adaptation to certain problematic situations, but encompasses the judgement and deliberation that individuals exercise in their decision-making strategies. Reflexive deliberations which are concerned with generating alternative future possibilities are more likely to lead to the individual exercising transformative agency. This is because, in undertaking these reflexive deliberations, the individual wishes to distance themselves from their present context. However, individuals may wish to maintain their contexts, to retain a sense of ontological security, and will therefore act to achieve this by undertaking habitual action.

The similarities between this conceptualization of agency and that offered by Paternoster and Bushway (2009) are that in both instances for desistance to occur the individual needs to distance themselves from their past identity and commit to an alternative future self. This is achieved by envisioning a new identity that is related to particular social roles. Both theories also contend that agency is developed over time, which helps to explain the notion that desistance is a journey which is likely to include setbacks (Burnett, 2004). There are, however, two crucial distinctions between the two theories. First, Paternoster and Bushway (2009) do not adequately account for how individuals make decisions about how to act in the present, whereas the chordal triad of agency offers an explanation of this by referring to the evaluative orientation that individuals may adopt. Second, Paternoster and Bushway (2009: 1133) account for setbacks by suggesting that these occur if the individual fails to predict future preferences accurately. I do not disagree completely with this argument – if individuals envisage and commit to an identity or future self that is unrealistic then this is likely to lead to dissatisfaction and relapse. However, the chordal triad of agency accounts for this by showing that individuals envisage future possibilities, but then evaluate these against what is possible given the social context of the present. In other words, variations in the ‘agency effect’ can be explained by the social structure of opportunities and life chances (Marshall, 2005: 68–69), and the differential distribution of power and resources as socially inherited from past structures (Archer, 1995). For this reason, it is more likely that setbacks will occur as a result of inequalities or structural barriers within the individual’s social context.

While it is important not to overestimate the role of economic and material resources in the construction of individuals’ biographies (May and Cooper, 1995: 79), recent policies, underpinned by the individualization thesis, neglect the structural constraints that inhibit identity construction and, in particular, the importance of access to economic resources and cultural spaces that make self-realization a possibility (1995: 80). Empirical support for the social and cultural marginalization of would-be desisters (see Farrall et al., 2010) brings approaches to desistance as underpinned by claims about individualism, liberty and choice under close scrutiny. Farrall’s (2002) sample suggested that they wanted to stop offending and, while many showed signs of desisting by the end of the study, there was a significant difference between the experiences of those who faced structural obstacles and those who did not. Indeed, individuals who faced no obstacles were significantly more likely to be able to desist. This suggests that offenders are likely to be excluded from the economic, social and cultural resources required to achieve the self-realization that can lead to successful desistance. While much of the recent desistance literature recognizes that human agency does play an important role in the desistance process, it is important not only to retain some focus upon the influence of structure, but also to conceptualize agency as intrinsically linked to structure.

The argument presented thus far states that desistance can be achieved as individuals reflexively consider their objective social contexts, and devise courses of action accordingly. Would-be desisters will identify a future ‘self’ that is different from their past ‘self’, and to which they will strive. This future self may either be that of ‘non-offender’, or some other identity that is incongruent with offending. The pathway chosen to become the future self will depend on how the individual mediates their objective social contexts, and the structures within it that impinge on them. Further, the particular courses of action that the individual takes along this pathway will depend upon their temporal orientation at that time – whether anticipating the future, drawing upon the past, or responding to the immediate demands of the present – and this orientation is conditioned by the interplay between action and structure. In many ways this may be considered to be an ‘ideal’ pathway to desistance, but it should be noted that the process of reflexive consideration is constant and ongoing, and that not devising possible future courses of action is not necessarily unreflexive, but rather an alternative configuration of reflexivity. In fact, given the social and cultural contexts that many desisters will be challenged with the possibility of conceiving of future courses of action is, at times, likely to be somewhat limited. In other words, the social and cultural contexts within which individuals live will influence the nature of reflexivity, but not preclude it. In a similar way, the impact and influence of socio-structural turning points is context-specific, and can only be examined with recognition of this (Carlsson, 2011).

Two case studies have been selected for discussion here, as they are regarded as exemplars of how agency in the desistance process is conceptualized in this article – in other words, they provide an empirical investigation of the chordal triad of agency in the desistance process. The two men satisfied the sampling criteria of having at least two previous convictions and experience of probation supervision for at least one month, but less than one year within the current sentence. Neither of the men had offended (self-reported or officially recorded) for at least six months, and were therefore considered in this study to be primary desisters. Both men reported that they might offend again in the future, but they were still considered to be primary desisters as this phase of desistance has been conceptualized here, and elsewhere, as the transitional phase between crime and conformity – the ‘liminal’ stage of desistance (Healy, 2010; Healy and O’Donnell, 2008). In this regard, the transitional phase of primary desistance provides a potentially fruitful area of research as it may help to shed light on the mechanisms by which individuals leave the ‘liminal’ stage and either revert to crime or achieve long-term desistance.

Case Study 1: ‘Ryan’

Ryan had served a short custodial sentence following breach of conditions for conviction of a violent offence. Following his release from prison, he had been under probation supervision for approximately six months at the time of interview. He had been convicted on several occasions prior to this, for drugs offences, common assault, carrying offensive weapons, criminal damage and theft. His offending had begun during his teens after he had been excluded from two schools and developed friendships with a group of youths who were also not in education. Ryan had begun smoking cannabis at 14, and by the time he left school he was smoking on a daily basis. During the interview, Ryan suggested that his cannabis addiction was, directly or indirectly, related to his offending. Some of his violent offences were related to the group that he spent time with, who encouraged each other to commit acts of violence. The same was true of his criminal damage offences. Other violent offences were the result of owing someone money for drugs he had bought from them, or where people owed him money for drugs. His convictions for theft were also the result of trying to obtain money to buy drugs.

Ryan had left school without any qualifications, and had subsequently never had what he called a ‘proper job’. Instead, he had sporadically worked with his step-father doing removals and house clearing on a cash-in-hand basis. Ryan had grown up in an area with rates of unemployment and crime higher than the national average (according to Office of National Statistics Neighbourhood Statistics data), and continued to live there at the time of interview. He had been out of work altogether for approximately one year, and suggested that he spent most of his time with friends smoking cannabis:

Like I said, a lot of the time I’d just be round my mate’s house smoking weed, but we can’t always do that so we just hang around the estate or go up town. There ain’t much to do up town, it’s shit, but it makes a change from hanging round the estate, so. I prefer to go round my mate’s house, but we’d be hanging round the estate quite a bit as well, there ain’t nothing else to do. (Ryan)

Ryan’s story, therefore, was one that is, unfortunately, all too common among young men who have been excluded from school, become involved in drugs and who have not been able to engage with formal labour markets. The outcome for Ryan (as well as many other young men in his circumstances) was to become involved in a variety of offending behaviours. He said that the reason he had begun offending was because he and the group he associated with did not have jobs and had lots of free time to spend smoking cannabis and getting into trouble. He referred to some of these other people as ‘triggers’, stating that when he was with them they could push him towards doing things he would later regret. He had articulated that this series of events had led to him committing various criminal acts, and that if he was unable to address issues in relation to his peers, drugs and employment then it would be likely that he would commit another offence if an opportunity presented itself.

During the interview, Ryan had stated that he wanted to stop offending, and he was asked about how he intended to achieve this. He explained that giving up cannabis would be a positive step towards moving away from crime, but that this was unlikely because of the proclivity among his friendship group for smoking cannabis. He said that if he gave up he would become ‘a loner’, and that he could not stay within his friendship group and give up because he said it would ‘drive me mad’. He explained how he did not have any money to participate in any other activities: ‘I’m only on the dole. I get £90 a fortnight. I spend £40 of that on weed, that leaves me with £50. I give my mum £30, and that leaves me with £20 for the fortnight.’ Ryan also said that he would like to get a job, and that this would help him to stay away from crime. However, he was quite pessimistic about his options: ‘Obviously, like I say, I left school with nothing and you can’t really get a job without it, so, erm, I’m a bit stuck with that really.’ Instead, he stated that he would try to obtain work through his step-father again, performing manual labour for cash-in-hand. He said that ‘most of the blokes round my way do it’, and that he would follow suit by asking people in the local area and that, provided he could prove he was a hard worker, someone would offer him work, even if only on a casual/informal basis. He did say that he would be willing to try an alternative route but, again, he did not seem to think that he would be successful in pursuing other options:

I’ve been waiting for the job centre to set this [interview] up. I was meant to have one before, but when I got there no-one had heard anything of me, so that was a waste of time. If they mess me around like that again I won’t bother with them no more. It’s daft really. A waste of my time, especially if I’m walking all the way from [home]. (Ryan)

Finally, Ryan stated that he would like to move away from his local area and that he would be more likely to resist temptations to offend if he were able to. Again, however, his perception of the likelihood of this occurring was somewhat pessimistic. He said that he did not have any opportunities to move elsewhere, partly due to a lack of money, but also because the only areas that were realistically available to him were similarly rife with drugs, unemployment and crime. He said that he knew most of these areas well, and that in every area he knew a drug dealer. When asked about whether he thought he would be able to stay away from crime, Ryan responded positively. He said that he thought it was quite likely, provided he was able to stay away from some of the people who might ‘trigger’ him into offending again.

Case Study 2: ‘Alan’

Alan was serving his second community sentence, for an assault, drink-driving and possession of drugs. He had previous convictions for assault and drink-related offences, and during the interview it became quite clear that drinking had been a problem for Alan for some time and that his offending history was related to alcohol. As part of his sentence, Alan had completed the Drinking Impaired Drivers programme to try to address this aspect of his behaviour, and he and his supervising officer had shared a number of discussions about his drinking. During his previous sentence, Alan had completed the Low Intensity Alcohol Module (LIAM) to try to challenge his drinking and alcohol-related offending. When asked why he thought he had re-offended following this, Alan said that it was ‘pure apathy’ on his part.

Since the offence, Alan had lost the job as a printer that he had held for approximately 25 years, and had also moved back in with his parents following the breakdown of his marriage. He said that getting work and controlling his drinking would be crucial to him stopping offending. He had been in contact with an employment agency and had spoken with a debt counsellor, both of which had offered him ‘good advice’. Despite being willing to apply for a job doing ‘absolutely anything’, Alan remained uncertain about the prospects of obtaining work:

I’m not [confident] at the moment. I’d like to be, but I’m not at all. You’ve only got to read the paper every day. So I’m not. It’s everywhere you look, there’s no jobs for people who want one, and people who’ve got one are being laid off every day. (Alan)

At the time of interview, Alan had not applied for or inquired about any jobs, because he said there was simply nothing available. He said that he was making efforts to curb his drinking and that his supervising officer was providing a great deal of support with this.

When asked about his strategies for moving away from crime, Alan was quite clear that dealing with his drinking and obtaining work were the two most important factors, yet when asked further how he intended to address them he was unable to determine any course of action to overcome these barriers:

There’s just too much going on with me at the moment. I’m worried about where my money’s going to come from, worried about my ex and keeping my home, when I’m going to see my son. It’s just hard to focus. My life’s just chaotic, it’s hard to juggle it all. (Alan)

Alan spoke about how he was spending his time, and it became clear that much of it was being spent in the same pubs that he frequented prior to the offence. He said that there was a lot of alcohol involved and that it often became an ‘emotionally strained situation’ that would frequently lead to violence. He said that he went there because he knew everybody there, and, despite them being what he called ‘naughty people’, they were his friends and he described them as being supportive. As a result, Alan stated that it was possible that he would offend again given his current circumstances and the manner in which he was spending much of his time: ‘If alcohol’s involved and, like I say, emotions run high, or what-have-you, you don’t know what might happen.’

Discussion: Uncertainty and Reflexivity

In both cases outlined above there are clear indications of the temporal orientations described by Emirbayer and Mische (1998). Both men stated that they wanted to stop offending, and both were able to provide suggestions as to what would need to happen for them to be able to achieve this. Perhaps unsurprisingly, given the wealth of evidence within the desistance literature, both Ryan and Alan stated that employment was of paramount importance. Relationships, accommodation and drugs/alcohol were also key concerns, which also reflects the prominence of these issues in the existing literature. This is suggestive of a projective orientation of agency when considering the future self in relation to desistance. Both men were able to envisage new social roles and future identities that would distance them from their past selves in an attempt to alter their life-course trajectories. This could also be explained by Paternoster and Bushway’s (2009) notion of the future self, as both Alan and Ryan were able to consider an alternative future that would not involve offending. However, the theoretical conceptualization of agency offered here provides a better explanation of how the two men considered the possibility of this future self in relation to their present social contexts.

Ryan and Alan were clear about the changes that would help them to desist from crime, yet when the discussions moved towards exploring how they would achieve this both men gave somewhat frank appraisals of the likelihood of these changes occurring. Thus, in both cases there appeared to be a tension between, on the one hand, wanting to desist and being able to identify the necessary changes to support this and, on the other hand, being aware of the realistic opportunities available within their immediate social contexts. In Emirbayer and Mische’s (1998) terms, this tension resides between the ‘projective’ and the ‘practical-evaluative’ orientations of agency. Within the former, the would-be desisters demonstrate self-change and identity distanciation, yet within the latter these future aspirations are ‘brought down to earth’ as the two men consider their social contexts.

The reality for many would-be desisters (illustrated in the accounts of the two men above), therefore, is that the institutional commitments to which they aspire are fairly conventional in relation to employment, family life and wider living arrangements. Beneath the surface of these conventional goals, however, is a social context characterized by complexity, uncertainty and exclusion. This institutional uncertainty is caused, in part at least, by the increased fragmentation of traditional institutional arrangements (see Bauman, 2002; Beck, 1992; Beck and Beck-Gernsheim, 2002; Giddens, 1991), such that patterns of work and family life are fundamentally different to previous (albeit relatively recent) historical periods (see also Farrall et al., 2010). There exists, therefore, a tension between wanting to desist and seeking traditional institutional arrangements to achieve this, and the process of identifying various structural barriers which lead to exclusion for the individual. In part, this may be the consequence of individuals having relatively narrowly defined aspirations based on traditional notions of masculine identities (see also Fletcher, 2007). However, it is also likely that, at least some, would-be desisters envisage future identities and new social roles which are more concomitant with contemporary late-modern society, but are excluded from achieving them as a result of the unequal distribution of power and resources within their social contexts. The outcome of this, it is argued here, is that many would-be desisters in this situation will revert to an iterational orientation of agency. The two case studies described above are illustrative of iterational agency in this sense.

In Ryan’s case, he envisages alternative living arrangements and employment in particular as being necessary for him to desist, but he suggests that such opportunities are unavailable to him. Consequently, he reverts to the iterational orientation of agency. With regards to employment, he suggests that he would not be able to obtain formal, mainstream employment, and instead acts on routine and habit by seeking informal work through his step-father. In this respect, routine is enacted as a means of trying to achieve goals and aspirations where new or alternative forms of social action appear to be unobtainable. Ryan also appears to revert to the iterational orientation of agency with regards to his cannabis use. Despite identifying a direct or indirect link between his drug use and his offending, Ryan suggests that he will continue to smoke cannabis, yet it is important to note that this could quite easily be explained as Ryan simply choosing to do something which he enjoys. It is possible that Ryan’s discussions of future employment and his desire to desist from crime reflected the circumstances of the interview, insofar as it had been arranged by his supervising officer and took place at the probation offices. However, this does not necessarily negate the findings reported here, and this second illustration of iterational agency could have resulted from Ryan encountering the uncertain and complex reality of contemporary life. Thus, it is argued here that would-be desisters may seek refuge in iterational orientations of agency, which are associated with strictly defined and highly familiar roles, in response to institutional uncertainty and complexity, and to the demands of constructing a ‘do-it-yourself biography’, to use Beck’s (1994) terms.

It could also be argued that aspects of Ryan’s case are illustrative of the practical-evaluative orientation of agency. He recognized that his drug use could lead to further offending, and identified alternative social roles that would help him to desist, yet he maintained that he would most likely continue to smoke cannabis. However, he also suggested that his offending was strongly associated with a particular peer group, with whom he had curtailed his association, and stated that he spent time with a new peer group who did not engage in offending behaviour. In this respect, it could be argued that he has considered his future ‘non-offending’ self in relation to his present circumstances and determined a realistic and appropriate course of action which accommodates the two.

Alan’s case is similar to Ryan’s in many respects. He recognizes that employment, housing and family are important in his journey towards desistance, and also that alcohol is a key trigger to him offending. He also identified a number of similar barriers that would prevent him from achieving the goals that he believed would help him to desist, most notably a lack of employment opportunities. For Alan, however, the process of reflecting on his present circumstances had clearly had a deleterious effect upon his sense of self. Within the spheres of work and family life, Alan had recently faced two stark realities of modern life. The employment sector that he had worked in for 25 years was in decline and it was unlikely that he would work in this area again, and his marriage had ended and he would have to come to terms with the fact that he would possibly see his son on an infrequent basis. In this regard, Alan’s iterational agency had led him to seek habit and routine by continuing to drink, despite his own recognition that this could lead to further offending. In this regard, the uncomfortable nature of living with institutional uncertainty could coerce some would-be desisters towards enacting old habits and routines, and the iterational orientation of agency provides a means of explaining this.

Of course, alternative explanations exist. Although my research cannot say whether or not Alan and Ryan experienced setbacks, or whether or not they successfully desisted, it would appear from their own evaluations of the likelihood of them achieving these goals that setbacks would be likely. It is possible that Paternoster and Bushway’s (2009) argument that such setbacks occur when individuals inaccurately predict future identities and roles could explain this. However, given the relatively mundane and conventional nature of the future roles and aspirations aspired to by the two men in the above case studies, it would seem to be more likely that such setbacks would occur as a consequence of the unequal distribution of power and resources in society resulting in various structural barriers within their immediate social contexts.

Conclusion

The central argument presented in this article is that existing conceptualizations of agency in the desistance process are, generally, too vague and do not adequately account for the role of structure in conditioning agency. The conceptualization of agency articulated by Emirbayer and Mische (1998) has been offered as means of contributing to this gap in the desistance literature. It has also been argued that many accounts of agency, both within the criminological and sociological literature, rely too heavily upon the individualization thesis, which suggests a degree of freedom and choice which is unavailable to many would-be desisters, given the socio-economic conditions of contemporary late-modern society. The purported range of opportunities available as a key characteristic of the individualization thesis are too heavily rooted in middle-class experiences, and do not incorporate a recognition of the structural reality that many would-be desisters face in their everyday lives.

One of the key conclusions drawn from this article is that when would-be desisters encounter institutional uncertainty or structural barriers they may tend to revert to the iterational orientation of agency, which underpins routine or habitual action. This may be routine action in an attempt to achieve certain aspirations or goals, or it may be an attempt to restore a sense of ontological security. The latter is a key characteristic of life in late-modernity, whereas the former may have a number of significant implications for the desistance process. First, as argued towards the beginning of this article, the existing literature suggests that desistance is more common when individuals envisage a new future identity or social role that is fundamentally different from their ‘offender’ identity. This, it is argued here, reflects the projective orientation of agency. However, when faced with circumstances which may appear to be unmanageable, or situations characterized by institutional uncertainty, would-be desisters may revert to routines or habits because they provide a rigid and highly familiar set of social roles and relationships. However, even though these routines or habits may not entail offending behaviour directly, they may be less likely to lead to desistance because such routines are unlikely to be stable in the long term given the contingent and changeable nature of institutions in contemporary society. Indeed, late-modern institutions demand reflexivity, a willingness to change and adapt, and the capability of acting flexibly. Adhering rigidly to routines and habits does not sit easily with this version of late-modern individualization.

Second, iterational agency, routines and habits for many would-be desisters may be likely to entail less conventional or pro-social networks and relationships. Social networks may prove to be an essential aspect of providing the opportunities necessary for desistance to take place (see McNeill, 2009), and where these networks are less pro-social such opportunities are likely to be diminished. Routine social networks may be likely to involve problematic drinking or drug taking, informal and casual employment, and volatile and unstable relationships. Related to this is the third possible implication for desistance – that is, that iterational agency, routines and habits may entail offending-related behaviour. In many cases this may likely involve alcohol or drugs, but there are also numerous other factors that may act as ‘triggers’ to further offending, such as peer group association.

Central to the concept of agency in desistance is the notion of intentional self-change. However, this should not be interpreted as the individual simply deciding what it is that they want to do or who they want to become, but rather it should entail an understanding of how the individual devises workable means of achieving these aspirations and, moreover, of how they determine realistic, yet desistance-focused, alternatives under conditions of institutional uncertainty and structural change. It was argued earlier in the article that any account of agency should incorporate an understanding of the influence of social structure, and it is the role of structure in enabling or constraining agency that is at the heart of this interplay. Would-be desisters are likely to face a number of various constraints within their immediate social contexts, and it is how the individual responds to these that is central to the argument presented here. Responding to the constraints of institutional uncertainty and social change is a key component of the individualization thesis, but not all individuals are equally well equipped to deal with this. The outcome for many would-be desisters is that such responses may entail a response characterized by routine and habit, as opposed to the more optimistically grounded contentions of many individualization theorists.

Any apparent lack of continuity towards achieving the goals and aspirations that many would-be desisters hold is not necessarily a reflection of a lack of commitment towards them, or a failure to predict realistic aspirations in the first instance. Rather, it is likely to reflect the fact that decisive moments in making the transition towards desistance are highly structured by the individual’s social, cultural and material resources. Where individuals are disadvantaged in this respect, as many would-be desisters are, this is likely to lead to a more disjointed transition towards future aspirations and, ultimately, a non-linear transition towards desistance. This may suggest that the sequential delineation of primary and secondary desistance often recounted in the desistance literature may not reflect the true reality of the desistance process. That is to argue that individuals may adopt new identities or social roles through a projective orientation of agency prior to abstaining from offending in the longer term. Offending behaviour may continue in some way or another as the iterational orientation of agency takes prominence in response to encountering uncertainty, but the identity and social roles envisaged through the projective orientation of agency will remain. A full consideration of this is beyond the remit of this article, but further discussion should be welcomed.

This article has shown that while agency is a crucial aspect in the desistance process, it is highly conditioned by structure, and the conceptualization of agency provided by Emirbayer and Mische (1998) provides a useful analytical framework for exploring this dynamic. This has a number of implications for policy and practice, both within a custodial and a community setting. First, the individuals in this article had both appeared to have adopted a reflexive stance in relation to their strategies for desistance, yet there was not, perhaps, an easily identifiable epiphany in either case. This raises questions around what prompted these individuals’ reflexivity, and if this could be identified more readily then alternative ways of working with would-be desisters could be implemented. Further research into this area should be welcomed. Second, it has been argued within this article that agency is multi-contextual – that is, that different contexts influence how individuals exercise agency by providing conditions which enable and constrain agency which, in turn, influences the possibilities of action for particular individuals at a given time. The various contexts that would-be desisters encounter solicit alternative temporal orientations of agency which can enable, constrain or suppress these possibilities of action. Approaches to working with offenders should, therefore, take greater consideration of individual needs, contexts and aspirations. This may not be straightforward in a climate where workloads and pressures are increasing, and opportunities for practitioners to develop individualized relationships may become more limited. However, to ensure longer-term desistance these kinds of relationships are highly important, and demand greater attention from policymakers, practitioners and academics alike.