‘Walking on ice’: The future of parole in a risk-obsessed society

Introduction

Risk assessment, for the purposes of identifying and classifying individuals or groups, has been a feature of governance for hundreds of years. Within criminal justice, the first risk prediction tool was piloted in the early 1930s (Harcourt, 2007), and today risk assessment tools and techniques abound as a means of targeting the most effective and efficient criminal justice interventions at the most dangerous offenders/offence groups. Those high risk prisoners eligible for parole (the early release from prison on licence to incentivize good behaviour in prison and to prepare for re-entry into the community), and in particular those who have served long-term custodial sentences for sexual and violent offences, are now less likely, however, to be granted—or to sustain the requirements of—parole. However, where parole is granted, such licensees are also increasingly likely to be recalled to prison because of failing to comply with supervision requirements or because of tensions among professionals tasked with the risk classification and management of such groups in the community.

This article argues that the term ‘high risk’ offers a justification to parole decision makers (for whom public protection is the primary aim) to take pre-emptive action to more closely monitor and more readily curtail supervision in the community for high risk individuals. While the parole board may commit to releasing high risk categories of prisoner, other professionals within the community (namely police and probation) may have a lower tolerance of managing that risk once those prisoners are released. In seeking to contain—or more ambitiously to eliminate—risk at the expense of enabling rehabilitation and reintegration, parole practices may prove counterproductive to public protection since released prisoners without constructive support are more likely to fall back on illegal means of survival (Harcourt, 2007; Robinson and McNeill, 2004; Weaver and Barry, 2014a; Werth, 2011).

First, the literature on risk, notably its history and rise in criminal justice, is explored—particularly in respect of parole. This article then highlights the lived experience of high risk individuals subject to parole supervision and the views of professionals overseeing the release of such prisoners into, and supervision within, the community. The implications for parole’s future of stringent licence conditions and professional propensities to breach and recall are that the population of eligible prisoners for parole will become increasingly high risk and high profile (Harcourt, 2007) as to make parole potentially unsustainable in the longer term.

The rise of risk in criminal justice

The concept of risk is not new in criminal justice, and risk categorization has long been seen as necessary, albeit controversial. Indeed, more than 250 years ago, Beccaria, cited in Hudson (2003), suggested that treating certain offenders more harshly, based on risk, was likely to be counterproductive to reducing crime. While originally used in criminal justice to gauge the probability of recidivism, and thereby to maximize on cost-efficiency (Werth, 2019), risk assessment has increasingly been used to gauge the probability of harm, despite Padfield (2002) suggesting that accurately predicting harm is highly unreliable.

Since the Second World War, rapid social and economic change has created uncertainty and insecurity about increasingly tenuous social discipline, interpreted by successive governments as reflecting moral decline and the need for greater public protection (Garland, 2001). This resulted in the view in the 1970s that ‘nothing works’ (Martinson, 1974), resulting in a growing shift from a more individualistic, rehabilitative stance to one of collective managerialism and punitivism, not least for high risk offenders: ‘from a focus on the singular case to group classification [. . .] [and from] knowing and normalizing individuals to a focus on identifying and managing populations’ (Werth, 2019: 5). The pre-1970s ‘Old Penology’ (Feeley and Simon, 1992: 451) focused on individual rehabilitation of offenders whereas the ‘New Penology’ (1992: 449) focused on the efficient assessment and social governance of the risks posed by both individual, rational actors (O’Malley, 2009) as well as high risk groups such as sex offenders (McAlinden, 2011). McAlinden (2011) notes that the highest risk offenders are indeed sex offenders, despite those offenders having a comparatively low probability rate of recidivism (Brayford et al., 2010; Brown et al., 2008; Erooga, 2008).

Kemshall and Wood (2007) suggest that the main precursors to the New Penology involved focusing on dangerous or persistent offenders, easing economic pressures on prisons and responding to a declining confidence in rehabilitation. Some commentators have questioned the extent of the shift away from rehabilitation for probation and social work supervisors (e.g. Lynch, 2000; Robinson, 2008; Werth, 2019), suggesting that these workers retain traditional concerns with re-entry and reintegration of ex-prisoners (Kemshall and Wood, 2007; McNeill, 2004; Ugwudike, 2011). Indeed, Garland (2001: 176) suggests that a focus on risk is not incompatible with rehabilitative concerns: risk control has shaped and reframed how rehabilitation is understood, but rehabilitation still exists ‘as a means of managing risk’. However, one underlying premise of the New Penology is that imprisonment and probation are no longer intended to rehabilitate offenders. Indeed, crime becomes ‘tolerable’ (Feeley and Simon, 1992: 455) and prisons become ‘a warehouse for the highest risk classes of offenders’ (1992: 460). Thus, in the latter half of the 20th century, a rise in crime made the primacy of rehabilitation untenable (Young, 1999); the need to expand the range of criminal justice interventions made greater assessment and categorization of risk advisable (Harcourt, 2007); and a rise in the prison population made the development of a parole system essential (Nellis, 2007).

The governmentality literature (for an overview, see Rose, 2000) likewise focuses on the retreat from welfarist, government-led risk reduction towards more punitive policies based on the management and individualization of risk. Hudson (2003) suggests that risk categorization, when coupled with blame and stigma, has exacerbated the individualization of risk. Risk categorization derives from what Young (2007) cites as a ‘weak’ rather than a ‘strong’ version of social exclusion, where the former blames the excluded and the latter blames the excluders. The more the excluded (or in this case, the criminalized) are blamed, the more likely such groups are to be denied inclusion (see also Harcourt, 2007). Rose (2000) argues that advanced liberal societies comprise two methods of controlling problematic populations, based on the responsibilization of citizens and the securitization of consumption. These are ‘circuits of inclusion’ (2000: 324) which regulate and control conduct by incorporating marginalized individuals, and ‘circuits of exclusion’ (2000: 330) which manage more ‘dangerous classes’ such as poor, homeless and workless individuals as well as known offenders. Some professionals in criminal justice seek to re-affiliate the excluded and to reattach them into circuits of inclusion (notably many probation and social work practitioners in criminal justice who encourage reintegration approaches within supervision). However, there are also professionals who deem reintegration or re-affiliation untenable for certain individuals (for example, those defined in insurance terms as ‘bad risks’ (2000: 326)), preferring to neutralize and contain these ‘anti-citizens’ (2000: 330) through imprisonment.

Contemporary thinking on rehabilitation versus punitivism, however, fails to fully differentiate between what I would argue are two distinctive and competing demands within criminal justice—community supervision as an alternative to a custodial sentence and community supervision as part of a custodial sentence, the latter in the form of parole and non-parole licences and the subject of this article. There is also a tension between the respective roles and responsibilities of the various professionals supervising high risk released prisoners in the community, roles which have various capacities for discretion in decision making (Lynch, 2000; Robinson and McNeill, 2004). Lim (2011: 17) highlights institutional attempts to both create and tame risk and that risk governance ‘serves particular interests’. The current, increasingly punitive and incapacitative strategies relating to the management of risk from parolees in the UK are a case in point. Indeed, nowhere within the UK criminal justice system is risk more acutely felt than in the system of parole, where the target population is already proven in law to be high risk through conviction for a serious offence and confirmed as such because of a sentence of long-term imprisonment.

The concept of parole is a relatively recent phenomenon in the UK, being introduced only 50 years ago through the Criminal Justice Act 1967. Two of the defining features of the early parole systems across the UK were assisting the rehabilitation of released prisoners on licence and protecting the public from serious harm (Justice, 2009; Kincraig Committee, 1989). However, the Kincraig report identified a stark difference between Scotland and England and Wales in terms of parole being granted: in 1987, 57% of prisoners eligible for parole in England and Wales were granted it, compared with only 28% in Scotland (Kincraig Committee, 1989). While the Kincraig Committee offered no explanation for this divergence in two jurisdictions with otherwise similar crime and sentencing levels, the inference is that Scotland’s parole board members are more risk averse than their counterparts south of the border. Certainly, that figure of 28% receiving parole in Scotland has not changed considerably in the ensuing 30 years (Parole Board for Scotland, 2017), and is even lower for high risk prisoners, with, for example, only 13% of those on life licence in Scotland being granted parole in 2015–2016 (Van Zyl Smit, 2018).

Scotland has recently witnessed a significant increase in the number of recalls to prison following breach of licence conditions. In 2012–2013, of those licensees who had been breached post-release, the Parole Board for Scotland recalled 69% of those on parole licence; 63% on non-parole licence; 71% on extended sentences; and 62% on life licence (Parole Board for Scotland, 2014). Three years later, those figures had increased to 78% on parole licence; 89% on non-parole licence; 85% on extended sentences; and 67% on life licence (Parole Board for Scotland, 2017). In the seven years from 2004/2005 to 2011/2012, the recall rate in Scotland doubled from 351 to 702 prisoners, and in 2015 it was not projected to change dramatically over the following seven years (Scottish Government, 2015). For those high risk prisoners released on life licence in Scotland, recall figures were only 10 in 2002/2003 but increased to 89 in 2016/2017 (Van Zyl Smit, 2018).

In England and Wales, the Prison Reform Trust (2018a) notes that while less than 1% of released prisoners on parole have committed a further serious offence, the number of determinate prisoner recalls has more than doubled between 2012 and 2016 and despite the abolition of Indeterminate Sentences for Public Protection (IPP) in 2012, the number of recalls from such sentences continues to rise (Beard, 2019) and to cause concern (Prison Reform Trust, 2018b).

The two key drivers to these rises in the recalled prison populations, not only in the UK but also in the USA, include: (1) an increase in the number of ex-prisoners being breached because of an inability to comply with increasingly stringent licence conditions (Padfield, 2013; Scottish Government, 2008; Solomon et al., 2014); and (2) an increased propensity on the part of parole boards, police officers and probation officers to recall following breach (see, for example, Hucklesby, 2017; Parole Board for Scotland, 2017). These two drivers to increased recall rates are explored further below through the views and experiences of licensees and parole professionals.

Methodology

With few measures of effectiveness available from the extant research within the UK in respect of parole decision making and management, and with a dearth of empirical evidence of the impact of the New Penology on practice (Ugwudike, 2011), this article is informed by two recent studies, one in England (Weaver and Barry, 2014a) and the other in Scotland (Barry et al., 2015) to explore the above two drivers to increased recall rates, drawing on the views and experiences of licensees, parole board members and supervising police/probation officers. The English study (quotations below referenced ES) explored the process and outcomes of community supervision within Multi-Agency Public Protection Arrangements (MAPPA) ¹ in an English Probation Trust and involved interviews with 26 high risk (i.e. levels 2 and 3) MAPPA offenders on post-release supervision in early 2013. Face to face interviews with these service users took place in probation offices, or in approved premises, alongside interviews with 26 parole professionals (nine probation officers, five police officers and 12 policy makers—the latter comprising four senior probation staff, two senior police officers, four MAPPA Strategic Management Board members and two National Offender Management Service staff). However, all licensee identities were protected and therefore the matching of licensees to supervisors was not feasible. The licensees ranged in age from 22 to 71, with an average age of 42, and two were female. These respondents were on orders such as Life Licence, Imprisonment for Public Protection (IPP), Parole and Non-Parole Licence, Community Supervision and Sexual Offences Prevention Orders (SOPOs). ²

The Scottish study (quotations below referenced SS) explored perceptions of compliance with, and breach of, community-based supervision in Scotland and included interviews during 2013 and 2014 with 69 ex-prisoners or currently recalled prisoners convicted of mainly serious violent or sexual offences. These respondents were part of a wider study of 250 service users (the remaining 181 in the sample were on community payback orders and other alternatives to custody) and 125 professionals, the latter including sheriffs, lawyers, senior management and basic grade staff in criminal justice social work ³ and senior management in the Scottish Prison Service, the police and the Parole Board for Scotland. The 69 ex- or current prisoners, all of whom were male, ranged in age from 20 to 73 with an average age of 41. Again, all licensee identities were protected and therefore the matching of licensees to supervisors was not feasible. Through purposive sampling, there were similar numbers of compliers (32) and breachers (37). A total of 35 of the 37 breachers had been recalled to prison at some point since their order began. The majority of respondents were interviewed face to face in their own homes or in social work offices, but 14 were interviewed in prison. These Scottish respondents were, or had been prior to recall, on orders such as Life Licence, Parole and Non-Parole Licence, Extended Sentences ⁴ and SOPOs.

Perceptions of the management of licensees

The following two sub-sections discuss the two drivers to increased recall rates based on professional and licensee attitudes and practices. The article then explores the implications of recall processes/outcomes for the future management of high risk released prisoners, concluding that parole’s future is in danger of being eroded by the increasing use of containment through selective incapacitation as the default position for managing high risk offenders.

Conclusion: The effectiveness of parole in managing risk

Examples have been given in this article of how the system of parole in the UK has become a potential site of conflict between released prisoners and the various MAPPA professionals tasked with supervising them in the community, thus undermining its effectiveness. For licensees, a seemingly arbitrary, inconsistent and ill-defined approach to risk management in the community prompted over four-fifths of licensees at interview in Scotland to comment on the perceived unfairness of the criminal justice system (see also Digard, 2010; Morgenstern et al., 2018). Such unfairness was particularly apparent in respect of the often indeterminate nature of recall—‘once you get recalled, it’s hard to get out’ (Probation Officer 3, ES), irrespective of whether they complied with or breached their conditions. Licence conditions and the constant fear of recall also tended to frustrate access to employment and social networks, leaving licensees being told what they cannot do in the name of public protection rather than what they can do in the name of reintegration (Hudson, 2003). Harcourt (2007: 28) describes a ‘ratchet effect’ which, when predicting high risk offences and profiling those for selective incapacitation, creates a concentrated high risk prison population and is counterproductive to re-entry and reintegration, since such high profile supervision ‘reduces work opportunities, breaks down families and communities, and disrupts education’ (2007: 29). Harcourt (2007: 188) anticipates that the prison population will increasingly comprise the highest-risk, highest-profile offenders and that already ‘the prediction of future dangerousness has begun to colonize our theories of punishment’. Taken to its logical conclusion, it could be argued that the prison population in the future will comprise such high risk and persistent offenders as to make parole unsustainable as a risk management tool.

For parole board members, the aim is to ensure public protection by releasing early on supervision only those prisoners for whom ‘the level and nature of risk are deemed to be manageable’ (Parole Board for Scotland, 2018: 9). This potentially creates a tension between deciding which prisoners parole board members consider to be a management risk if released into the community, and which prisoners MAPPA considers to be a management risk once released into the community. For the police at least, the end result of parole could be argued to be recall in the name of risk elimination. Indeed, parole itself could be argued to be a risk indicator for certain police colleagues. And for probation officers/social workers who continue to regard reintegration as an aim of supervision, their ability to make a constructive difference to the lived reality for released prisoners is minimal within a criminal justice system that focuses increasingly on risk elimination at the expense of risk management (Hudson, 2003).

These differing views and practices of the respective criminal justice agencies, coupled with the doubts that many practitioners often have about being able to alleviate poverty and marginalization through criminal justice policy and practice, have led to wider social policy concerns from such professionals about parole’s role in assisting reintegration. Goffman (2014), for example, highlights the fact that police in deprived areas within the USA are unable to alleviate social problems facing many high risk offenders and that their primary role is mere containment. Several probation and social work practitioners in the two studies illustrated in this article also had reservations about being able to offer constructive support to released prisoners in the current socio-economic climate, leaving the onus on released prisoners themselves to effect their own reintegration (Barry, in press) despite barriers imposed by licence conditions.

Feeley and Simon (1992) note that high recall rates used to be seen as a sign of failure but are now seen as a sign of success for the parole system. However, both Padfield (2013) and the Prison Reform Trust (2018b) argue that high recall rates should, on the contrary, be seen as a failure on the part of the criminal justice system rather than a success on the part of the parole system. Within the USA, Simon (1993: 201) argues that parole supervision had become ‘a temporary space between initial release from prison and inevitable return to prison’. In the UK, the burgeoning cases of re-arrest and re-incarceration of released prisoners for breach of licence conditions are often not subject to the due process of legal scrutiny or a fair trial because, as illustrated above, the situation for licensees being recalled to prison is justified merely on the balance of probabilities and processed, without legal scrutiny, as an administrative rather than a legislative exercise. As seen in the empirical research drawn on above, because of increasingly stringent conditions, licensees are more likely to breach or to be breached, and because of their increasingly risk averse practices, police and probation/social work supervisors and parole board members are more likely to recall.

Based on the statistics cited earlier, the increasingly cursory use of parole and the arbitrary use of recall with high risk offenders in the UK currently presents not only a legislative and practical barrier to ensuring justice for high risk prisoners on release, but also an ethical and human rights dilemma because reintegration is undermined by rigorous conditions, and rights to due process and liberty are undermined by the over-use of recall. Gelsthorpe (2007: 486) suggested that by the end of the 20th century, probation officers in England and Wales had become ‘screws on wheels’, enforcing rather than facilitating compliance. The Human Rights Act 1998 could go one of two ways, in Gelsthorpe’s view, either to encourage a rights culture or to fire-watch cases being brought that could undermine the status quo. Due process rights are ‘fundamental to the rule of law’ (Gelsthorpe, 2007: 497) and, as Hudson (2003: 28) suggests: ‘should not be compromised or suspended for reasons of utility’. And yet utility, in a neo-liberalist context, trumps due process in the name of risk elimination. Safety is more important than equality, security more important than liberty. Hence Clear and Cadora’s (2001: 59) contention that risk control is indeed about ‘containment, not change’.

This culture of containment is experienced by supervisees as unfair: it undermines the values and principles surrounding incentivized early release and disrupts ex-prisoners’ reintegration back into society. As a result, it has been argued here that parole has moved precipitously away from a combination of rehabilitation, reintegration and public protection, towards a focus solely on public protection with the use of recall becoming the default position for high risk offenders on licence. Given Harcourt’s (2007) ratchet effect of targeting and imprisoning the highest risk offenders, the capacity and propensity of parole board members to release prisoners on licence could increasingly be undermined. Indeed, it could well lead to the eventual unsustainability and hence demise of parole itself because of the negative impact of supervisory requirements on released prisoners’ ability to desist and reintegrate. Released prisoners by and large want to reintegrate into mainstream society, whereas many of the professionals tasked with their management now want to contain them. As Padfield (2013: 43 citing Maguire and Raynor, 2006) suggests: ‘perhaps the most fundamental question is not whether society can resettle prisoners (probably it can), but whether it really wants to’.