What Works to Reduce Reoffending: update of the evidence on imprisonment and community disposals

Chapter 2: Effectiveness of Interventions

2.1 Effectiveness of interventions

The link between criminal justice interventions, reoffending and desistance is complex. Even the term desistance is contested across the literature, with some studies defining desistance as a permanent cessation of offending over several years, whilst others accept that instances of reoffending may occur within a longer period of what they would term ‘desistance’. For the purposes of this review, desistance and other key terms are defined in the Glossary on Pages 3 and 4.

Before considering the various interventions available and their impact on reoffending, we must consider the definition of ‘effectiveness’ within this context. Across criminal justice jurisdictions, different interventions have different aims and, therefore, different parameters for being deemed ‘effective’. Different aims, or positive aspirations, for criminal justice interventions include: ‘desistance’, ‘rehabilitation’, ‘reintegration’ and reduced ‘reoffending’ (sometimes known as ‘recidivism’).

Given the various legitimate aims of each particular intervention, the criteria for what makes an intervention effective can vary markedly. For example, as a review by the Sentencing Council states, a sentence focused on rehabilitation may seek to achieve a number of objectives: “helping to reduce reoffending, promoting desistance from offending, or facilitating reintegration into the conventional social world, beyond the simple act of reintroducing into the community by virtue of their release” (Sentencing Council 2019). As set out by the United Nations Office on Drugs and Crime (UNODC) in the Kyoto Declaration, full reintegration would be the gold standard for a positive criminal justice outcome (UNODC 2021), although this would be hard to achieve in reality. Pragmatically, sentencing policy may narrowly focus on reducing reoffending because it is a more manageable and measurable aim, while at the same time reducing the number of offences being committed or more serious offences being committed (Sentencing Council 2019). In line with the Sentencing Council, this review will use the aims of deterrence, desistance, and rehabilitation and reintegration, when determining the effectiveness of different interventions.

Additionally, it is worth noting that there is strong evidence that those interventions that are most effective are based on risk, needs and responsivity (RNR), where the level of intervention should be matched to the risk of reoffending, the person’s criminogenic needs^[2] and, delivering support in a way tailored to maximise their personal responsiveness. This typically involves cognitive elements, and therefore often uses CBT or other evidence-based psycho-social treatments.

It is clear from the evidence discussed in this paper that desistance, for people who have been involved in persistent offending, is a difficult and complex process, with likely lapses and relapses. It is an inherently individualised process, meaning that one-size-fits-all interventions are unlikely to be effective. It is this challenge that the criminal justice system must meet.

2.2 Specific interventions

2.2.1 Imprisonment

Imprisonment has a limited impact on reoffending while an individual is serving their sentence, due to their incapacitation and because there are fewer opportunities to commit crimes in prisons. However, similar to findings in the 2015 What Works Review (Scottish Government, pp 23-26), the evidence relating to reoffending after release is mixed overall. Research has found that imprisonment has a weak or null effect (Petrich et al. 2021), or even a criminogenic effect, as well as a damaging effect on inmates’ mental health and wellbeing (Edgemon & Clay-Warner 2019; Fazel et al. 2016). Where there have been rehabilitative effects observed from prisons, that have led to lower reoffending, these typically relate to specific settings that emphasise and adequately resource rehabilitative programming.

The following section examines the available evidence on the impact of imprisonment on rates of reoffending.

Incapacitation

It is sometimes argued that imprisonment is effective at reducing reoffending through incapacitation effects, for the simple reason that while an individual is imprisoned they are less likely to reoffend (Al Weswasi 2024). However, it is important to acknowledge that the incapacitation effect differs for offenders with different risk levels. A recent study in Sweden found that the incapacitation effect of first time imprisonment (specifically how many offences were prevented by imprisoning an offender over an alternative sentence in the community) was limited, averting 1.17 offences for males and 0.78 offences for females annually. However, the authors highlight a range of challenges in both comparing to and generalising from the results, such as variation in sample composition, incapacitation measurement and, differences in offender types. Taking the data and the limitations into account, the study concludes that the incapacitation effect of incarceration is modest, and when offenders are categorised by risk category, there is substantial heterogeneity in outcomes and their desistance patterns will vary according to different levels of need and risk. For example, the results show that for a first-time incarcerated offender in the highest-risk group, 1.22 convictions are averted annually and 2.55 offenses are averted annually that could have resulted in a conviction. For offenders in the medium-risk groups, the corresponding figures are 0.31 convictions and 0.68 offenses. Therefore, and recognising the pressures on Swedish prisons, the limited crime-preventative effect for low-risk, first-time offenders may mean that non-custodial sentences should be considered as an alternative (Al Weswasi 2024)

However, other studies found that prison may in fact increase the risk of reoffending (Wermink et al. 2023; Petrich et al. 2021; Damm & Gorinas, 2020). As one study asserts, prisons are effective sites of offender convergence, providing an environment conducive to enabling crime, as well as contacts, skills and opportunities to allow offenders to engage in future criminal behaviour once they are released (Damm & Gorinas, 2020). Studies suggest that this is due to negative peer association within prison and disruption of employment, housing and social bonds. Moreover, the economic impacts of incarceration may increase the financial incentive to engage in criminal behaviour upon release, while the social impacts provide access to opportunities to engage in further crime (Wermink et al. 2023; Ministry of Justice 2023; Prison Reform Trust 2022). For these reasons, the incapacitation effects of short term imprisonment may need to be considered alongside negative impacts that may increase the risk of reoffending upon release.

Deterrence

Deterrence refers to the extent to which a sentence/punishment might deter people from offending. The evidence for the deterrent effect of prison is mixed. One review comparing perceived risk of punishment against actual risk of punishment, found that there may be some baseline deterrent effect generated by the criminal justice system. Whilst an additional, albeit limited, deterrent effect may also exist for short periods of time when punishment events, such as lengthy prison sentences for homicide, are highly publicised. However, this is only likely to be present for as long as the news cycle. The review highlighted that to produce an adequate deterrent effect it may be more effective to target very specific deterrence messages at audiences who are at especially high risk of committing a crime in the near future (Kleck 2016). Two other reviews found limited evidence that prison can persuade some offenders not to reoffend in the short-term, although this was largely attributed to the effects of incapacitation (Loeffler & Nagin 2022; Petrich et al. 2021).

Another study found no discernible impact of the deterrent effect of imprisonment when compared to non-custodial sentences (Villettaz et al. 2015). This finding was reinforced by a more recent systematic review of the literature, which found that most studies surveyed showed that deterrence had little impact on the probability of recidivism (Wermink et al. 2023).

A review that evaluated 16 systematic reviews and meta-analytical studies of sentencing and deterrence in the criminal justice system, found that there is limited evidence on the specific impact of deterrence on reoffending. The review notes that since many crimes are committed when an individual’s reflective capacities are severely compromised, some crimes cannot be deterred because it would require an individual to reflect on the consequences before a crime is committed. The review also highlights the need for further research into awareness of sentencing policy in the general population, as without such information it is difficult to draw conclusions around whether changes in sentencing policy (such as increased sentence length) have a resultant impact on deterrence (Perry 2016).

Rehabilitation

Some studies have highlighted the potential rehabilitative effects of imprisonment, to the extent that it is used for educational, vocational or treatment purposes. Criminogenic effects, i.e. an increased likelihood of offending, were principally found in settings where rehabilitative programming was not emphasised (Loeffler & Nagin 2022). Similarly, another study in the US found that the impact of incarceration on the likelihood of reoffending was linked to the availability of rehabilitative programming (Duwe & Clark 2016). The evidence suggests that the effectiveness of imprisonment, as a measure to reduce reoffending, is predicated on the availability and quality of rehabilitative programmes. Therefore, an important consideration within the context of an increasing prison population is the capacity to provide the robust and high quality rehabilitative programming required to reduce reoffending. Should capacity to provide this programming be reduced, the evidence suggests this may contribute to a rise in reoffending. However, there is some evidence that these positive rehabilitative effects that can reduce reoffending may be overstated and indeed, the criminogenic factors noted above, such as negative economic impacts and the influence of anti-social peers, may mean the collateral effects of custody off-set some of its potential beneficial effects (Wermink et al. 2023).

Loeffler & Nagin (2022) also found that pretrial incarceration appears more likely to generate adverse recidivism impacts than post-conviction incarceration. Reasons for this seem rooted in the way that pretrial incarceration differs from post-conviction incarceration, where offenders generally have less access to rehabilitative programming compared to convicted prisoners and less access to substance misuse and mental health treatment resources. They also serve shorter terms of incarceration and they experience an environment typified by higher levels of population turnover and higher levels of stress and mental illness. Further, evidence suggests that incarceration with access to rehabilitative programming is more likely to have a positive impact on recidivism than incarceration that is merely punitive in nature.

Sentence length

As with other interventions, the effectiveness of short-term prison sentences (up to 12 months) will be different for each offender and will vary across demographics, circumstances and the type of offence (Ministry of Justice 2023). However, a review by the UK Sentencing Council found little evidence for effective or beneficial outcomes in respect of short term prison sentences (Sentencing Council 2022), whilst studies like Wermink et al (2023) (discussed above) also highlight concerns about short sentences in particular.

Imposing longer prison sentences has also been investigated to determine whether it leads to reductions in reoffending. Research that explicitly addresses the effects of sentence length is more limited than literature comparing the impacts of imprisonment against non-custodial sentences. Those studies that do exist show inconsistent findings and complex patterns, and primarily focus on the US, making it difficult to draw overall conclusions with specific relevance to Scotland. For example:

• One natural experiment in France appeared to indicate that increasing sentence lengths did not initially deter offending, but people who experienced the longer sentences did then become less likely to offend again (Philippe 2023).
• However, a systematic review of studies found that, overall, the effect of incarceration length on recidivism appears too heterogeneous to draw universal conclusions and the strength of findings are inconsistent across studies due to methodological limitations (Berger & Sheidigger 2022). Of the studies reviewed, some suggested that longer sentences provide additional deterrent benefits in the aggregate, though some studies also had null effects. None suggested a strong aggregate-level criminogenic effect.
• Wermink et al. (2018) studied the short-term effects (6-month follow-up period) of sentences up to 15 months, and that were on average 4.1 months in length, and found that length of sentence does not have a significant effect on recidivism in the first 6 months after release. The study found that this conclusion holds true across various measures and types of recorded recidivism, including reoffending, reconviction, or reincarceration.
• Using data on individuals released from Florida prisons, and using matching techniques, Mears et al. (2016) found that longer periods in prison were initially associated with a greater risk for reconviction, but that these effects disappeared approximately 2 years after release, underscoring the importance of longer follow-up periods, since effects may dissipate over time.
• One study in Sweden concluded that an increase in incarceration time could allow for further rehabilitative interventions for some at-risk individuals serving shorter sentence lengths. However, this presupposes that the environment in which they are incarcerated has the resources to correctly identify at-risk individuals, sufficient social and health programs to address their needs, and support to maintain improvements when they re-enter the community (Al Weswasi et al. 2022).

A striking conclusion with regard to the more recent quasi-experimental studies on sentence length is that although research designs have improved, the overall evidence remains unclear, with some studies yielding null effects of sentence length on reoffending and others pointing to the existence of a minor deterrent effect with a longer sentence. Further, there are also issues regarding the generalisability of recent studies. Not only have few studies been conducted outside the US but the findings are also often based on specific populations of offenders, such as juveniles or inmates sentenced within a specific court system.

2.2.2 Remand

This review found limited evidence published since 2015 that specifically focuses on remand and its impact on reoffending. Like custodial sentences, remand can, to some extent, prevent reoffending in the very short-term through incapacitation effects. However, it can also impact negatively on aspects of an individual’s life that have been shown to improve long-term desistance, such as physical and mental health, employment and family relationships.

In 2021, the Howard League published a report highlighting the negative impacts of remand in Scotland. Evidence suggests that time spent on remand in Scotland is often unproductive, with limited opportunity to engage with support services, such as education and rehabilitation. In addition, although remand prisoners are allowed more personal visits than convicted prisoners, these visits can place increased pressure on families to provide support and to visit regularly, often at considerable expense. Furthermore, the negative consequences of remand are well documented, with impacts on physical and mental health, employment, housing and family relationships (Howard League 2021). As noted in the previous section on prison sentences, there is evidence from the literature on short-term sentences that these consequences of custody can have a criminogenic impact and may make reoffending more likely in the long term.

2.2.3 Bail

Evidence suggests that the effectiveness of bail supervision is largely dependent on the offender’s willingness to comply with bail conditions and to desist from further offending.

Recent research commissioned in Scotland found that there were mixed views on the effectiveness of bail supervision among professionals working with offenders. The research found that its effectiveness was largely dependent on the accused person’s ability/motivation to engage and to desist from further offending. Further, it was noted that although some people complied well with supervised bail, they still ended up on remand after committing further offences, even though they had otherwise complied well with their bail conditions (Scottish Government 2023). Overall, those interviewed for the research felt that they were unable to comment on the effectiveness of bail supervision due to a lack of readily available evidence or statistical data. However, the main potential benefit of offenders being released on bail, as opposed to being kept on remand, was the stability it offered to the accused: it allows the accused to keep their employment, their tenancy, and maintain family connections. The maintenance of these benefits was seen to prevent reoffending (Scottish Government 2023).

A review of the research conducted by the Australian Government found that bail support programs can reduce remand populations, reduce reoffending among participants, and improve sentencing and long-term outcomes for both accused persons and the criminal justice system. Findings indicate that bail support programs should prioritise support over supervision, as well as treating and responding to an individual’s criminogenic needs over monitoring. Further, evidence suggests that undertaking treatment and accessing programmatic services while on bail can demonstrably reduce reoffending. For example, supporting the accused person to meet basic responsibilities, such as appearing in court when required, can ensure they receive the benefits of bail, are able to remain in the community and maximise their chances of receiving a non-custodial sentence (Willis 2017).

2.2.4 Early release measures

This section presents findings from the evidence base on early release measures, focusing on both electronic monitoring and parole.

Electronic monitoring

Offenders released under electronic monitoring (EM) may be less likely to reoffend than those released without monitoring. However, some studies found no difference in offending behaviour between those released with EM and those without. The mixed evidence may be a consequence of the tension between EM deterring further offending behaviour, but also increasing the probability that the offending that does take place will be noticed and potentially result in the individual returning to custody. Similar to findings from the 2015 paper, this review found that EM is likely to be most beneficial when delivered alongside support to address the person’s needs.

Evidence suggests that there are mixed results regarding reoffending and reconviction with use of EM. There is some evidence that indicates that reconviction rates for monitored people who serve their full sentence on EM are lower, or similar, compared to matched groups who serve their full sentence in custody. However, as with other non-custodial sentences, caution must be taken when comparing the reconviction rates of those released with EM on Home Detention Curfew (HDC)^[3], to those who have been released from custody. It is likely that those on HDC have a lower risk of reoffending in comparison to other offenders, so the rates may not be directly comparable, although this risk is controlled for in some of the studies reviewed.

A 2019 Scottish Government report reviewed the strengths and weaknesses of EM for offenders in the UK as well as reviewing the international evidence base on this topic. The review found that there may be differences in reoffending between people released on EM and those released without EM following a prison sentence. Although both groups have an equal opportunity to reoffend, there may be a higher level of deterrence for those monitored via EM, when compared to those released from prison without EM. Firstly, because detection of an offence is more likely while being monitored and second because a breach of EM conditions is likely to result in revocation and imprisonment. Additionally, those subject to EM have been subject to a risk assessment process by the judiciary, prison and probation authorities meaning that it is typically only lower risk offenders who are released into the community. Williams and Weatherburn (2019) also found that those subject to EM are likely to differ from other offenders in observable ways, such as committing less serious offences and in unobservable ways such as the individual’s level of remorse. Such factors are also likely to influence reoffending outcomes. The Scottish Government review concluded that in those instances when offenders were released on EM, this was shown to be most effective when delivered alongside support to address the individual’s needs.

A meta-analysis of 17 studies, which assessed quantitative data on reoffending, concluded that EM did not have a statistically significant effect on reoffending overall. Although most individual studies did show an effect in the positive direction and as such, it is important to note no ‘significant effect’, is not the same as ‘no effect’. Of those that showed a positive effect on reoffending, 3 specifically compared EM to imprisonment and found reductions in reoffending when compared to the alternative of imprisonment (Belur et al. 2017). Another study, included within the meta-analysis, conducted a randomised control trial on offenders in Germany and found that EM did not have a recidivism-reducing impact. However, the study was supplemented by qualitative analysis which indicated that some interviewees did experience a deterrent effect but only for the duration of the EM. In addition, some interviewees reported benefits of being subject to EM, in that it allowed them to maintain employment and social contacts (Meuer & Woessner 2020)

Parole

Evidence on the efficacy of parole on reoffending behaviour is mixed, with effects differing across jurisdictions and likely dependent on the level of support provided to the offender.

Typically parole^[4] is considered to have two primary functions: to provide treatment to meet the needs of the parolee by facilitating continued access to appropriate services, and to monitor the offender’s behaviour with the aim of reducing the risk of recidivism (Vito, Higgins and Tewksbury, 2017). Although parole is widely used, both in Scotland and in international jurisdictions, evidence on its effectiveness to prevent reoffending is mixed. However, it is possible that the inconsistent results obtained in studies of parole effectiveness simply reflect differences between jurisdictions in the level of supervision and support provided to parolees. Given the non-random nature of the conditional/unconditional release decision, it is also possible the conflicting results are a result of selection bias (Wan et al. 2015)

One review that compared recidivism outcomes for offenders released early from prison to supervised parole with those released at the end of their sentence, with no supervision, found that although findings in this area are mixed, there is some evidence to support the efficacy of supervised release as a means of reducing rates of reoffending (Sydes, Eggins and Mazerolle 2018). Another study evaluated the effectiveness of parole supervision in Australia using propensity score matching, a statistical approach used when controlled experiments aren’t possible. The study found that offenders who received parole supervision upon release from custody were less likely to commit a new offence, took longer to commit a new offence, and committed fewer new offences than those who did not receive supervision (Wan et al. 2015).

One study in the US, conducted in 2015, also found mixed results when comparing supervised release with no form of supervision. The researchers found that offenders released to any form of post-prison supervision were approximately 11 to 20% less likely to be arrested for any crime (felony or misdemeanour, excluding technical violations of supervision) and 30-44% less likely to be convicted for a felony offence after release from incarceration. However, the researchers urge caution, as these positive findings are somewhat mitigated by negative effects of supervision on arrest for felony and return to prison. These findings indicate that offenders who were subject to post-prison supervision were also 7 to 20% more likely to be arrested for a felony, and 67 to 360% more likely to be returned to prison than those with no supervision requirements. The researchers state that the significant increase in the likelihood of reimprisonment may be due, at least in part, to the nature of supervision – whereby supervised offenders may be returned to prison for technical violations of their supervision order, in addition to as a result of a new sentence. In sum, those offenders on supervision were not only more likely to be caught if they committed a new offence, they were also more likely to commit an offence due to technical violations of their supervision. One key limitation of this study is that it includes all movements into the Florida prison system and does not isolate those returned to prison due to a violation compared to those with a new offence (Clark 2015)

Another study in the US, which looked at whether there was a causal effect of prison and parole time on recidivism, found that time on parole has no significant effect on recidivism. Possible reasons for this outcome were cited as the declining use of parole discretion, where Parole Boards previously assessed how dangerous offenders are and the likelihood of possible reoffending. The researchers also suggest that the null effect of parole on recidivism could be driven by the (in)effectiveness of supervision policies following an offenders release from prison (Zapryanova 2020).

2.2 Community disposals

Overall the evidence indicates that community sentences are more effective in reducing reoffending than short-term prison sentences and may provide greater opportunity for rehabilitation. However, this opportunity for rehabilitation is largely dependent on the offender’s ability to access rehabilitative programmes and the quality of such programmes.

Typically studies compare the impact of short-term prison sentences and community sentences on reoffending. In Scotland, reconviction rates are lower for those receiving community sentences compared to those released following a short-term prison sentence (12 months or less). Although it should be noted that reconviction rates between those receiving custodial sentences and community sentences are not directly comparable, due to the different types of offender associated with different offences (Scottish Government 2024b). Further, the PASS in Scotland means that a court must not pass a sentence of imprisonment for a term of 12 months or less unless it considers that no other method of dealing with the person is appropriate (Scottish Government 2022). Typically, short custodial sentences are received for the types of less serious crime, or chronic low-level offending, that tend to also have high reconviction rates, while long sentences are received for more severe offences which, on average, tend to have low reconviction rates. For example, the reconviction rate for custodial sentences of 12 months or less was 50.2%, compared to 7.1% for sentences over four years. For those given a sentence length of between 3 and 6 months, the reconviction rate was 52.9% and it was 53.7% for 3 months or less (Scottish Government 2024b).

In England and Wales, the impact of short sentences, community orders and suspended sentences on reoffending was assessed by the Ministry of Justice Analytical series (2015 and 2019). Both of these publications found that custodial sentences of 12 months or less were associated with statistically significant higher rates of reoffending than community orders and suspended sentences. The analysis from 2019 found that the one year average number of reoffences per sentencing occasion was also higher following short-term custodial sentences than if a community order had instead been given (by around 65 reoffences more per 100 sentencing occasions). Further analysis from the 2019 report indicated that when the one year reoffending impact of short-term custodial sentences was compared to community orders, the reoffending impact was 4% higher, regardless of whether the short-term custodial sentence length was less than 3, 6, or 12 months (Ministry of Justice 2019). However, the analysis also showed that offenders sentenced to immediate custody or to a community order were more likely to reoffend than an offender receiving a suspended sentence order. The Sentencing Academy of England and Wales identified similar findings, but stated that it remains unclear why suspended sentence orders result in better reoffending outcomes than short-term prison sentences (Sentencing Academy 2021). However, it should be noted that suspended sentence orders are currently not a sentencing option available in Scotland.

Yukhenko et al. (2019) carried out a systematic review of 28 studies, focusing on recidivism rates internationally. The review found that community sentences were typically associated with lower recidivism rates than custodial sentences, particularly in those individuals with low to moderate risk levels, and specialised community sentences are more likely to benefit individuals in certain offender groups in relation to reoffending outcomes.

It is important to acknowledge, however, that this review makes clear that there are some significant difficulties and barriers to researching and comparing the impact of community disposals on reoffending across different jurisdictions. The researchers found wide variation in recidivism rates between countries. This effect can be explained by differences in judicial practices, definitions used and the operationalisation of recidivism measures. Additionally, the actual sanctions that comprise community sentences are different depending on the legal system. For example in certain jurisdictions, the use of fines are included in analysis of rates of recidivism. Furthermore, there are many different approaches to defining the start point of a follow-up period in an individual’s given community sentence. In many reports, the distinction between a follow-up period and a period of supervision was not always clearly described and accounted for. For example, in some reports the follow-up period started after the completion of community service, i.e. after the end of supervision. This approach ignores a significant amount of time when a sentenced individual is at risk of committing a new offence. Finally, there may be differences in the quality of supervision, availability and effectiveness of rehabilitation programmes, availability of vocational training, and access to healthcare services. These are factors that researchers and practitioners are keen to examine; however, because of differences in reporting practices, it is difficult to compare these factors internationally. For these reasons, the researchers advocate for the introduction of reporting guidelines for recidivism rates for community sentences (Yukhenko et al. 2019). Such guidelines would enable us to better understand the impact of community disposals on reoffending.

In addition to the positive impact community disposals have on reoffending when compared to short custodial sentences, recent research has found that short custodial sentences may also be less effective for reintegration and rehabilitation outcomes than other community sentences, though this can vary across different offender characteristics and offence types. Opportunities to address underlying criminogenic needs for offenders such as mental ill health or substance use, may also be greater in the context of community sentences (Ministry of Justice 2019). For example, a 2018 study on the UK’s Durham Constabulary ‘Checkpoint’ Programme encompassed a randomised control trial and adult deferred prosecution scheme involving 521 offenders entering the criminal justice system. Those in the treatment group were offered entry into an offender programme focusing on aspects of accountability, where they are expected to confront the consequences of their behaviour. The study concluded that there was a reduction of over 10% in both the number of individuals who went on to reoffend, and the number of reoffences committed, over a 24 month period follow up period, compared to control groups (Weir et al. 2021).

2.3 Age

Evidence on the impact of age on reoffending suggests that motivations for offending change with age and that young people typically have the highest rates of offending, but also the most likelihood of desisting over time as they mature. On the other hand, older people who commit crimes have less neurological development and social maturation ahead of them, potentially making it more difficult for them to change their patterns of behaviour. While the general pattern established earlier in this report - that most effective responses to offending typically keep people in their communities and address their underlying criminogenic needs – holds for people of all ages, the differences in risk and motivation mean that tailoring support and intervention options according to the age of the offender is likely to improve the impact on reoffending.

Age has consistently been recognised as a relevant risk factor in terms of reoffending. This link between age and crime is known as the age-crime curve, where children commit crimes at higher rates into their teens, peaking at around the ages of 18-20, then desisting as they reach adulthood (Sentencing Council 2019). Whilst ageing itself cannot be affected by sentencing options, it has been argued that “we should understand (changes in the level of reoffending) not as a spontaneous and inevitable physiological and psychological process associated with ageing, but rather as a social process which can be enabled or impeded by a person’s associates and environments.” On this basis, by enabling social maturation, and avoiding interventions that impede it, sentences may be able to reduce the volume of offending in a criminal career or reduce the length of a criminal career (Sentencing Council 2019).

A more recent review by the Scottish Sentencing Council was carried out on neurological and neuropsychological evidence pertaining to the age at which adolescents achieve cognitive maturity. The review found that immaturity of cognitive regions along with overactivation of emotion and reqard-related regions contributes to adolescents finding it difficult to think rationally and critically before making complex decisions. Evidence suggests that adolescents find it difficult to override their drive towards short-term gratification, particularly in males, where, in comparison to females, higher levels of sensation-seeking and lower levels of impulse control are observed. The presence of peers has also been observed to exert an influence on decision making either towards anti-social or prosocial behaviour although the mechanism for this remains unclear (Sentencing Council 2024).

Significantly, evidence supports theoretical models that position poor decision-making and increased risk-taking in adolescence as the result of typical maturational processes rather than solely related to preference or personality. It would appear therefore that the consideration of culpability, and by extension sentencing, in both adolescents and young adults should include appropriate regard to their cognitive maturity (Scottish Sentencing Council 2024). For these reasons, rehabilitation should be a primary consideration when sentencing young people, given their relatively lower level of maturity and, a greater capacity for change and rehabilitation than an older person.

2.4 Female desistance journeys

There is a degree of overlap between the criminogenic needs of men and women, although some factors appear to have greater impact on women than men, and female offenders may also have higher rates of non-criminogenic social needs (HM Prison and Probation Service 2018; Rodermond et al. 2016). Women may also experience these needs differently, due to structural differences in, for example, economic independence or trauma related to gender-based or domestic violence (Rodermond et al. 2016). This has implications for the effectiveness of interventions. In terms of physical and mental health, the negative impacts of imprisonment in particular, may be greater for women (SCCJR 2023).

Much of the evidence base on reoffending focuses on the how and why men desist from offending. Far less is known about the desistance process for women. A recent Scottish Government review of quantitative evidence on women in the justice system indicated that there are differences in the patterns of reconvictions when broken down by gender. On average, females are reconvicted less often than males, although the gap has closed in recent years (Scottish Government 2022). In 2020-21, the reconviction rate was 27.5% for males and 23.5% for females. Over the longer-term, reconvictions for males have gradually decreased, but have remained mostly stable for females.

The prison population in Scotland is largely comprised of men aged 21 and over. Women have consistently constituted a very small proportion of the prison population, at around 5-6% since 2001-02. In 2019-20, this figure stood at 4.9%, equating to around 400 women in prison on an average day.

A review that incorporates 44 studies on female desistance (Rodermond et al. 2016), indicates male-based theories of desistance generally apply to female offenders, albeit with some differences. For example, the review indicates that the prevalence of particular criminogenic needs tend to be different in women and men. For women, having children and supportive relationships are more strongly correlated with desistance than they are for men. For men, being employed is a stronger factor in desistance, while having criminal friends is more likely to inhibit desistance. In studies focused on women, positive factors in predicting desistance are economic independence, absence of problematic drug use and having individual agency.

The presence of risk factors for criminal behaviour varies to some extent by gender. A study by HM Prison and Probation Service, that examined the nature of the most prevalent and predictive risk factors for reoffending among men and women, found that the most prevalent criminogenic needs for women were poor problem solving, impulsivity, and unemployment. These were also the most prevalent needs for men. The prevalence of other needs not only differed somewhat by gender but also by risk and offence type. Of those sampled, 74% of women with a current conviction for acquisitive crime had a Class A drug problem compared to 54% of men in the same risk band and offence type category. This difference was not present with those at lower risk of reconviction. Binge drinking was also identified as a major predictor of women’s reoffending but was less strongly linked to reoffending for men (HM Prison and Probation Service 2018).

Although there is overlap between the criminogenic needs of men and women, where differences do exist, this can have implications for the effectiveness of interventions and what works to reduce reoffending. In terms of physical and mental health, the negative impacts of imprisonment in particular, may be greater for women. For example, in England & Wales, statistics indicate that women are committing self-harm at concerning rates. In the 12 months to March 2024, the rate of self-harm in the female estate (5,577 incidents per 1,000 prisoners) was more than eight times higher than in the male estate (652 incidents per 1,000 prisoners). Although this may be a reflection of the relative rates of self-harm for men and women in the wider community (Samaritans 2022), the number of incidents and rate of self-harm in both the male estate and the female estate are now at the highest level in the time series (Ministry of Justice 2024).

International evidence also shows that people in prison are significantly more likely to take their own lives than the general population. It is difficult to estimate the risk of suicide in the women’s estate reliably as females generally make up relatively small numbers in the prison estate, but there is evidence to suggest that females in prison are up to 20 times more likely to take their own lives than females in the community (Samaritans 2022). In Scotland, there are no current, robust figures for the prevalence of mental illness within the prison population although it is thought to be higher among women than men in the prison estate (Scottish Government 2022).

As a share of their respective prison populations, the proportion of women on remand has consistently exceeded that of men. From 2021-22 to 2022-23 that gap widened: in 2021-22, 29% of female prisoners and 25% of male prisoners on an average day were remanded whereas in 2022- 23 the proportion of women on remand increased to 33%, while for men it fell slightly to 24% (Scottish Government 2023). In a Scottish study, wellbeing has also been shown to be significantly lower among people on remand compared to those sentenced. Factors which are related to concerns around poor conditions, included restricted access to support services and limited opportunities for purposeful activity (Tweed et al. 2019). For women in particular, evidence shows that prison is damaging in a number of ways as they may experience the impact of being separated from their children, or losing a job or tenancy, more acutely, and as a result may experience a severe impact on their mental health (SCCJR 2023). Given these effects, the disproportionate use of remand for women is concerning.

For these reasons, several studies advocate that a holistic approach to reducing female reoffending must be taken. For women, the aim should be to achieve positive outcomes in multiple areas, for example health, education, relationships, resilience, and social integration. In this way women can stabilise their lives and promote readiness to change (Prison Reform 2017).