The Policy Evaluation and Research Unit at Manchester Metropolitan University carried out a Rapid Evidence Assessment, to systematically review and summarise evidence about the effectiveness of probation supervision on reducing reoffending (Smith et al., in preparation). The review highlighted the positive effects of various forms of probation supervision on reducing reoffending, but found limited evidence relative to the prevalence of probation supervision globally and in Europe.
What is it?
Probation supervision is an important part of the Criminal Justice System, in the United Kingdom and in many countries worldwide. Community sentences often comprise a number of components such as unpaid work, a curfew or prohibition requirement, and drug or alcohol treatment (Sentencing Council, 2017). Probation supervision is central to co-ordinating these, with co-ordination activity taking place within one-to-one probation supervision meetings.
Probation supervision therefore comprises of a series of meetings between the offender and offender manager. In addition to co-ordinating the sentence, supervision meetings serve a number of purposes, which in the UK may include:
• Communicating the rules of probation, including the ‘sentence plan’.
• Clarifying what will happen if the offender does not comply with the terms of his/her order.
• Agreeing the dates, times and places of future meetings.
(UK Government, 2017)
Supervision may also be oriented towards the delivery of some interventions within sessions. These might include short interventions relating to substance misuse or debt; or so-called ‘talking interventions’ such as brief solution focused therapy (BSFT) or cognitive behavioural therapy (CBT). As with any professional one-to-one relationship, specific probation supervision relationships may vary according to a number of factors, including the perceived depth and quality of the relationship and the skills demonstrated by the offender manager. The offender manager’s skills may vary due to the training and performance of the individual offender manager, and different practises and skills may be employed depending on the client group or organisational and political context of the work. For example, specific approaches to probation supervision (e.g. Interserve’s Interchange model, ‘Citizenship’, developed in the north east of England) impose a theoretical orientation and practical structure on the supervision, and consequently a degree of consistency in the application of offender management.
In general however, probation supervision has a long history, both in the United Kingdom and worldwide, and has therefore seen a diverse range of practices and skills employed. Despite the generally consistent purpose of probation supervision, practice varies. This poses difficulties when trying to understand if and how probation supervision consistently works to reduce reoffending due to the lack of a body of high quality evidence which addresses similar approaches. For example, Trotter (2013) notes the diverse range of skills identified in studies on probation supervision, and the consequent lack of consensus on the efficacy of these skills in reducing reoffending.
In a recent Rapid Evidence Assessment (REA; Smith et al., in preparation), it was found that supervision styles varied greatly throughout the Global North, and that this impacted on the level of success in reducing recidivism.
Should it work?
To answer the question of whether a policy or practice should work, it is necessary to consider its’ theoretical basis. By understanding the mechanisms of how a policy or practice is expected to effect change (the ‘intervention logic’ or ‘theory of change’), we begin to understand whether and how it should work. Therefore, we should seek to understand the “conceptual link from an intervention’s inputs to the production of its outputs and, subsequently, to its impacts on society in terms of results and outcomes (European Commission, 2004). More specifically, the question is whether, or not, the widespread practice of probation supervision reduces reoffending (recidivism). In order to answer this, it is helpful to therefore consider how exactly probation supervision is theoretically expected to reduce recidivism.
Such theoretical mechanisms have been specified. For example, the principles of offender risk, criminogenic needs, and responsivity to different types of ‘treatment’ have informed approaches to probation supervision for some time (Andrews and Bonta 2006, Lipsey and Cullen 2007). Another theoretical perspective is desistence theory, which would acknowledge the role which supervision may play in reducing reoffending. This however is somewhat of an underdeveloped theory as it relates to probation supervision, perhaps due to the fact that desistence journeys are unique to the individual offender (McNeill and Weaver, 2010).
In addition to criminological perspectives which may help to understand the theory of how probation supervision should work, one might look to other professions, in which a one-to-one relationship is central to the provision of the service. Guidance suggests that one-to-one relationships are more likely to yield positive results, when they are perceived by offenders to be good relationships (McNeill and Weaver, 2010). Similarly, a review of the quality of probation supervision (Shapland et al., 2012), highlighted generic factors that represent quality probation supervision. The review found that supportive relationships are important in the journey towards desistance, and highlight a number of ways in which supervision might work to reduce reoffending, such as developing and maintaining motivation and hope, promoting offenders’ strengths and resources (human capital), working collaboratively with offenders, and developing their social capital.
The evidence reviewed in Smith et al.‘s (in press) REA could be described as atheoretical. It relied on randomised controlled trials and statistical techniques to ascertain causality (i.e. did probation supervision result in a consequent reduction in reoffending), but did not articulate a theory as to how this might be expected to happen, or to explain the observed effects. However, the discussion highlights how qualitative interviews may help to deepen understanding of the difficulties inherent within these one-to-one relationships, in a way that statistics is unable to measure.
Does it work?
To understand whether probation supervision works to reduce reoffending, we undertook a systematic search for studies in which offenders were randomly allocated to conditions (e.g. probation supervision vs a custodial sentence), or for which there was a well-matched or unmatched comparison group. We followed a Rapid Evidence Assessment methodology, identifying 837 studies from an initial search, and after screening them on exclusion criteria, we found 13 studies. The included studies addressed probation supervision both as a community sentence and as part of a license (parole). Examples of specific probation supervision interventions studied include:
• Low intensity supervision
• Discussion of procriminal attitudes in supervision
• Use of cognitive techniques in supervision
• Reduced officer caseloads
• Staff Training Aimed at Reducing Rearrest (STARR)
• Local initiatives (e.g. the Citizenship programme; the Serious and Violent Reentry Initiative; the Minnesota Comprehensive Offender Reentry Plan; Maryland Proactive Community Supervision; the Statutory Early Release programme)
After identifying the evidence, we conducted meta-analyses on the results. The studies that were identified in the REA included dichotomous outcome measures (e.g. did the individual reoffend within a specified period?), and continuous outcome measures (e.g. how long was it before an individual reoffended?). As such, we conducted two separate meta-analyses, one for each type of measure.
Both meta-analyses indicate that probation supervision, overall, has a positive effect on reducing reoffending. The likelihood (expressed as odds ratios or hazard ratios) of reoffending is shown to be likely to be lower for those offenders who have been exposed to some type of supervision (with reference to the specific supervision interventions summarised above). This finding is consistent with theories which posit an effect of probation supervision on desistence from crime (such as the examples outlined above), and a number of previous studies of probation supervision and its effect on reoffending (e.g. Lipsey and Cullen, 2007).
As our systematic search identified and included a number of studies which examined different aspects of probation supervision, it is necessary to be cautious when interpreting and understanding any single, overall observed effect. The heterogeneity of the evidence is consistent with the diverse global and historical contexts of probation supervision, and the many variables which determine supervision.
Whilst all of the summarised probation supervision interventions show a positive effect between probation supervision and reducing reoffending, it is important to note that a number of these are programmes which were designed for implementation within a specific and local criminal justice system (often in the United States), and that the observed causal relationship may therefore not generalise to the United Kingdom.<
How strong is the evidence?
A systematic search for evidence of the effect of probation supervision on reoffending identified 13 studies which were based on random allocation of offenders to a treatment group, and/or a matched control group. However, given the long history and prevalence of probation supervision worldwide, there were relatively few studies which met the criteria for inclusion. Furthermore, the identified studies largely focused on unique and localised probation supervision policies and practices. This makes it difficult conclude that there is a consensus of evidence which identifies the effect of specific probation supervision practices on reoffending.
Is it worth it?
None of the studies included in the Rapid Evidence Assessment featured an economic analysis, and we are therefore unable to provide an answer to this question based on the recent REA.
Some previous reviews (e.g. Aos et al., 2006) have estimated the economic costs and benefits of various specific instances of probation supervision, but have not found enough evidence to estimate the economic costs and benefits of supervision in its broadest sense.
Can it be implemented?
The supervision of offenders within the probation system is an integrated and well-established part of the United Kingdom’s criminal justice system.
Specific supervision practices (e.g. the use of cognitive techniques) and policies (e.g. reduced officer caseloads and low intensity supervision) could potentially be implemented, notwithstanding financial and operational considerations, including staff development. However, without a clear consensus of evidence that these specific elements would reduce reoffending in a context different to that of the original study, caution should be taken when making decisions about whether to incorporate them into probation supervision practice. A further consideration is whether the observed effects from the studies would also be observed in a less well controlled operational environment. For example, the effect of the use of cognitive techniques in supervision would depend on staff consistently demonstrating the use of the techniques, and fidelity to the theoretical approach.
What's missing from the evidence?
There are a small number of studies documenting the effect of probation supervision on rates of recidivism which meet the criteria for inclusion in this review. Additional studies employing robust randomised and matched comparison group designs are required to investigate some of the specific elements of probation supervision. These elements would include those identified by the studies included in this REA (thus testing their characteristics in different contexts), and those which are theorised to be ‘active ingredients’ in probation supervision, such as the quality of the relationship, or the degree to which the relationship develops the human and social capital of the offender.
In particular, more high quality evidence of probation supervision in the United Kingdom and Europe would make a distinct contribution to the evidence base, as well as further evidence on whether probation supervision reduces reoffending for according to variation in offender characteristics (e.g. by risk, gender, offence).
In short more evidence is required which helps to develop theory, and understanding of which aspects of probation supervision work to reduce reoffending, in what way, and for whom.