Probation and Community Sanctions: A Collection of Research Findings from Criminological Highlights

In 1996, Canada introduced legislative reforms to amend s. 718 of the Criminal Code which sets out the purposes and principles of sentencing. In addition to the six objectives of sentencing that were already in place, namely denunciation, deterrence, incapacitation, rehabilitation, reparation, and promoting responsibility/acknowledging harm on the part of the offender, s. 718.2 enumerated additional principles that a court should take into consideration when imposing a sentence. Specifically, s. 718.2 (d) and s. 718.2(e) provided directive to exercise restraint in the imposition of any sentence, which state that:

(d) an offender should not be deprived of liberty, if less restrictive sanctions may be appropriate in the circumstances; and

(e) all available sanctions, other than imprisonment, that are reasonable in the circumstances and consistent with the harm done to victims or to the community should be considered for all offenders, with particular attention to the circumstances of Aboriginal offenders.

Subsection (d) outlines a general principle of restraint to adopt the least restrictive sanction possible while subsection (e) specifies restraint in terms of the use of custodial sentences.

Although codified as a sentencing principle in 1996 in Canada, the idea that we need to reduce reliance on imprisonment as a sanction is not new and has persisted over the better part of a century as seen in multiple reports and commissions (see Archambault Report 1938, Ouimet Committee Report 1969, Canadian Sentencing Commission 1987). Pitched as a means of reducing imprisonment, probation and community sanctions are often conceived of as “alternatives” to imprisonment. An inherent problem in thinking of probation and community sanctions as “alternatives” to prison is that it reflects a presumption in favour of imprisonment as the first, default or standard punishment (Cole forthcoming[1]), thereby furthering the conceptualization of probation and community sanctions as being lenient, not punitive enough, or somehow deviations from a standard punishment.

Rather than as an alternative to punishment, probation and community sanctions should be understood as a form of “constructive punishment” that is “justified and administered as a punishment” in their own right, that also goes beyond the mere infliction of pain in rehabilitating the offender by encouraging responsibility for the offence (p. 13, 6.1.4)[2]. In order to mitigate the presumption of imprisonment as the default punishment, Cole, forthcoming, suggests that “Parliament should consider amending s. 718.2(e) of the Code by inserting the word ‘first’ between ‘should’ and ‘be’, so that the subsection would read ‘all available sanctions, other than imprisonment…should first be considered.’” Imprisonment should be the last remaining choice after other options have been exhausted. In other words, rather than considering whether probation and community sanctions might be appropriate instead of sending someone to prison, probation and community sanctions should be considered in terms of whether—in and of themselves—they are appropriate in a particular case.

From a numerical perspective, community sanctions that are typically administered or supervised by probation officers clearly outnumber custodial sentences for adults and especially for youths as can be seen in the following tables for correctional admissions in Ontario and Canada.

Source: Statistics Canada, CANSIM (downloaded 12 April 2021)

Note: The use of “other community programs” as a descriptor apparently varies across the country

This overview will discuss four broad themes in the research literature that are pertinent to explaining why probation and community sanctions may be preferable to imprisonment in many circumstances and how these community-based options may be further improved for greater effectiveness.

1. In comparison with imprisonment, probation and community sanctions are just as effective—if not more effective, in some cases—at reducing recidivism, while also being more cost effective. Furthermore, probation and community sanctions have less deleterious consequences on education and employment than the consequences associated with imprisonment.
2. However, not all aspects of probation and community sanctions are effective at reducing recidivism or promoting better outcomes in other areas of civic life. Rather than assuming that a particular approach will work, it is important to conduct careful and thorough evaluations of a specific program as it is implemented in practice.
3. Approaches that focus more on surveillance and control tend to be less effective. Higher levels of supervision, higher numbers of and more restrictive conditions are associated with increased violations, higher recidivism, and more rapid reoffending. Overall, approaches that:
   • (a) strike a balance between enforcement/control and help/rehabilitation;
   • (b) have a clear goal;
   • (c) identify and target offenders who are more at risk of reoffending; and
   • (d) provide services that are specifically suited to the offender’s criminogenic needs tend to be more effective in reducing both breaches of conditions and re-offending.
4. Research on perceptions of sentencing and community sanctions suggest that offenders often view community sanctions as being more punitive than short stays in imprisonment. Moreover, the public is quite supportive of the use of community sanctions. This generalization holds even among those who say that they think that sentences are not harsh enough, if adequate information is provided about real cases.

a) Comparable or Better Reductions in Recidivism

Although it would be tempting to view incarceration as a way of reducing offending, existing data would suggest that those who are imprisoned rather than being given community sanctions are, if anything, more likely to re-offend (see another “Criminological Highlights” collection: The Effects of Imprisonment: Specific Deterrence and Collateral Effects (https://www.crimsl.utoronto.ca/research-publications/faculty-publications/effects-imprisonment-specific-deterrence-and-collateral).

Often, a fundamental challenge in trying to compare the effects of imprisonment versus community sanctions on recidivism is to have two groups that are equivalent in all other characteristics but for the sentence. The only way to guarantee that this, so that we can isolate the cause of a difference in recidivism (if we see one) to the difference in sentences, would be to have offenders randomly assigned to imprisonment or community sanctions. This type of randomized controlled experiment is considered the “gold standard” in evaluation research. In many cases, though, this is not feasible and may raise ethical complications given what we know about the harms related to imprisonment. Therefore, not surprisingly, these kinds of randomized experimental studies are fairly rare.

More commonly, the next best thing is to create two groups of people who are as similar as possible on all characteristics except for the sentence they received. Often, this is done by finding pairs of people who are identical on variables known to be associated with recidivism, except for the fact that one went to prison and the other was given community sanctions. An alternative approach is to create an overall measure of the likelihood of receiving a community sanction using all of the background information available, and then matching on this ‘propensity score’ those who received community sanctions with those who were sent to prison.

One study was able to do both using data for offenders sentenced in the Netherlands in 1997. They took pairs of people whose backgrounds would make them equally likely to receive community service, but only one actually did. They also matched on other characteristics, such as age, sex, and the relative length of the sentence. Offenders could receive up to 240 hours of community service or 6 months in prison. Only those who had never been sentenced to either option were examined to rule out any carry-over effects from previous experiences with these sanctions. The results showed that those who were sentenced to prison had higher recidivism rates in terms of average annual rate of convictions for all crime and separately for property and violent crimes (p. 14, 11.5.2). To put this in perspective, in 2018, 73,951 offenders[1] were sentenced to custody in Canada, with 74% them (or 54,891 offenders) being sentenced to 6 months or less in prison in Canada. Although not all these offenders would have met the criteria for this study since some of them had already experienced imprisonment or a community sanction, this study would suggest that an option of up to 240 hours of community service would have been an effective way of reducing recidivism.

Similar results have also been demonstrated in the UK using the same ‘propensity score matching’ method. This study identified equivalent pairs of male offenders who could have received a prison sentence or a community sanction. Of more than 5,000 men for whom data were available, reasonably equivalent matches were found for 2,324 of them. All were supervised in the community: either following a prison sentence of at least one year or just sentenced to a community sanction. After following them for a year, the study found that those sentenced to prison were significantly more likely than those given a community sanction to be convicted of an offence; and to re-offend sooner. Overall, the differences between the imprisoned and community sanction group were largest for those who were least likely to go to prison at the time of sentencing, e.g., those with the least serious offences and/or least serious criminal records. In other words, prison seemed to have the greatest negative impacts on those who were the most likely candidates for a community sanction (p. 15, 16.2.7).

Because these two studies used the ‘propensity score’ method to statistically create two similar groups, one could raise issue with the fact that perhaps the propensity score missed or lacked data on other important variables on which the two groups differed. In other words, it could be suggested that it was this missing other factor that caused the differences in recidivism between these two groups, not the difference in sentence. This is why a randomized controlled experiment, where offenders are randomly assigned to community sanctions or imprisonment, is really important. These designs help to ensure that the two groups are equal but for the sentence. As noted above, these studies are rare due to feasibility and ethical considerations. However, it was possible in a Swiss study that compared the recidivism rates of offenders randomly assigned to community service orders or to a short period of incarceration (up to 14 days). If an offender sentenced to a short stay in prison were eligible for community service, the offender was given the option of being assigned, on a random basis, to community work rather than prison. The results showed no significant difference between those assigned to community service and those sent to prison on the likelihood of being re-convicted or the average number of convictions within 24 months of the sentence. However, for re-arrests, those assigned to community service were less likely to be re-arrested than those who served their sentences in prison. This study shows that a community service order can be just as effective, if not more effective, in reducing recidivism than a “short sharp shock” in prison (p. 16, 3.4.4). By echoing similar findings from studies using less rigorous designs, this study helps to lend support and credence to those findings since this study used the “gold standard” of evaluation research—a randomized control experiment.

These findings are particularly impressive for community sanctions given the types of offenders who are more likely to receive probation versus those more likely to be precluded from probation. Often, the seriousness of the offence is used as a barometer for whether a community sanction is appropriate, with those convicted of a serious violent offence being ruled out, even if they are not at a high risk to reoffend and the offence itself may have been uncharacteristic. In fact, a “seriously recidivist” shoplifter is more likely to get probation. Given that many “high risk to reoffend” people are given probation, it is surprising the community sanctions are at least as effective or more effective in reducing recidivism than imprisonment (p. 17, 2.3.3).

While being just as effective or even better at reducing recidivism than imprisonment, community sanctions also tend to be more cost effective than prison (p. 17, 2.3.3). Even in cases where community sanctions do not reduce recidivism more so than imprisonment, their use could be justified solely on the basis of reduced incarceration costs in the future alone. For example, a study examining the impact of a therapeutic community work release program in Delaware found that factoring in costs such as prosecution costs, criminal justice costs, and additional social benefits (e.g., more stable employment, less drug use, less dependence on social welfare), for every $19 USD invested in this particular community program, 1 day of incarceration was avoided. It is important to remember that not all treatment programs based in the community are cost effective, but this study suggests that a carefully implemented community sentence can pay for itself solely in terms of reduced subsequent incarceration costs (p. 18, 6.4.1)

Overall, if community sentences are just as likely, or sometimes more likely, to reduce recidivism than imprisonment, and it is more cost effective, then community sentences should be more preferable than imprisonment in appropriate cases.

b) Less Deleterious Impact on Education and Employment

Moreover, probation and community sanctions tend to have less deleterious effects on education and employment compared to imprisonment. In 2006, Denmark allowed prison authorities to substitute periods of electronically monitored house arrest for short periods of imprisonment in order to maintain labour market participation and education. From 2006, those sentenced to prison for up to 3 months were allowed to apply to serve their sentences at home with electronic monitoring, as long as they attended a crime prevention program, allowed unannounced visits from correctional workers, and agreed to substance testing. These offenders were matched with those who served their sentences in prison on over 20 variables including criminal history, offence type, educational attainment, etc. They were followed for 3 years after the end of their sentence. After two years, those given the opportunity to serve their sentence in the community were less likely to have dropped out of school and more likely to have completed post-secondary education, in large part, because they did not have their lives disrupted by imprisonment (p. 19, 17.1.6).

In addition to education, employment and work force participation are less severely disrupted by community sanctions compared to imprisonment. A study carried out in Sweden examined all males born in 1975 and 1980 who were residents of Sweden by age 16. Because it is possible to link administrative data on Swedish citizens, not only were employment records available but other relevant data (e.g., education, immigration, mental health, criminal record, etc.) were also available. Using propensity score matching (described in the previous section) based on these background data, the researchers created two equivalent groups: those sentenced to either prison or probation for the first time at age 20-25. Prior to being sentenced, there were no differences in employment records between these two groups. However, for the first three years after being sentenced, those who received a prison sentence were less likely to be employed than those who received probation (p. 20, 17.6.2).

Similar results were also found for young adults in the US. Using a subset of respondents from the US National Longitudinal Survey of Youth, this study matched those who were convicted and incarcerated and those who were convicted but not incarcerated on 30 different variables (e.g., family structure, educational background, arrest history, etc.). The people in this study were, on average, only in prison for a little more than 4 months for their first time. Even this relatively short period of incarceration had a significant impact on their employment. Incarcerated young adults were less likely to be in stable employment, more likely to be consistently out of the work force, and more likely to not employed and not looking for work. Since all of the people in this study were convicted, it is clear that there is an additional deficit created by incarceration, in addition to conviction itself (p. 21, 11.4.4).

With fewer negative consequences for education and employment associated with community sanctions, those who are given community sentences are also less likely to be dependent on welfare benefits after their sentence. A study examining male non-traffic offenders sentenced after April 2006 (when electronic monitoring at home was introduced as an option for non-traffic offenders) in Denmark found that those punished in the community instead of in prison were less likely to dependent on welfare benefits after the completion of their sentence (p. 22, 14.6.5).

However, not all practices related to probation and community sanctions are effective at reducing recidivism or promoting better outcomes in other areas of civic life. Although programs that address the factors that former prisoners face (e.g., employment, housing) can often be justified on humanitarian grounds, it does not mean that they will be effective in addressing those factors and/or that they will reduce re-offending. There is no single program that is going to work for everyone. Evidence on effective programs is fairly thin; therefore, careful and thorough evaluations need to be built into the implementation of all programs. Even if there is an effective program, it is risky to generalize about what might be effective to a new setting or to different kinds of offenders. It is important to recognize that “[g]ood intentions may have disappointing results” (p. 23, 16.2.6). Rather than assuming that a particular approach will work, it is important to conduct careful and thorough evaluations of a specific program as it is implemented in practice.

Vocational, educational, and life skills programs have been staples of the correctional system for years. However, there have been relatively few adequate assessments of their impact. Looking at the research literature that does exist, the overall finding seems to be that the results are going to depend on the exact program in place, the characteristics of the offenders in the program, the outcome measures of interest, and the rigorousness of the evaluation methodology. Although prevalent, one cannot assume that these ‘staple’ programs will work because a similar program or approach had positive impact elsewhere (p. 24, 3.4.2). 

For example, it is often assumed that “more” is better when it comes to providing services, especially for youth who are involved in the formal youth justice system. A study examined youths (12-18 years of age) in California’s Ventura County who were apprehended either for an offence or a probation violation and who were deemed to be “moderate” risk to re-offend. Because they were randomly assigned to receive either ordinary probation or a form of “intensive” supervision, the two groups are presumptively equivalent but for the type of supervision they received. The ordinary probation involved minimal intervention: monthly contact with the youth, referrals to outside agencies, and no particular special services. The intensive probation involved at least one hour of face-to-face contact a week, at least two family contacts a month, programming to address various “risk” in the family (e.g., substance issues, parenting, etc.) as well as restorative justice programs. After 18 months of follow-up, there was no difference between the two groups of youths in terms of re-offending. Providing services does not, in and of itself, guarantee that these services will be effective (p. 25, 7.3.2).

Another approach that is often assumed to be effective at reducing re-offending is electronic monitoring. As a technique for knowing someone’s whereabouts, electronic monitoring has been around since the mid-1980s. In an evaluation of electronic monitoring programs in British Columbia, Saskatchewan, and Newfoundland, this study found no reduction in recidivism that was not attributable to differences in risk levels. Furthermore, the availability of electronic monitoring has not invariably reduced the reliance on prisons (p. 26, 3.2.4). Similar findings were reported in Scotland, with no effect on offending rates, either in the short or long term than any other penalty. Moreover, it did not appear to deliver promised cost savings and ended up increasing technical breaches of conditions (p. 27, 4.3.7).

Based on the evaluative evidence on probation and community sanctions, it seems that approaches that focus more on surveillance and control tend to be less effective. In contrast, approaches that: strike a balance between enforcement/control and help/rehabilitation; have a clear goal; identify and target offenders who are more at risk of reoffending; provide services that are specifically suited to the offender’s criminogenic needs tend to be more effective in reducing both breaches of conditions and re-offending.

a) What Does Not Work

It is often assumed “more is better” when it comes to supervision and conditions. This assumption has been the basis for the proliferation of intensive supervision practices. However, intensive supervision with more conditions and more restrictive conditions is a complex way of setting people up to fail. Research has documented that intensive supervision leads to more technical violations than ordinary supervision. This then results in increases in technical charges (e.g., administration of justice offences, court order violation, etc.) that actually lead to increased imprisonment (p. 28, 1.2.7).

In a study of more than 250,000 people on parole supervision in California during 2003 and 2004, those subjected to high levels of supervision were more likely to be found to have violated parole, controlling for characteristics of the offender. Whether or not an offender is found to have violated their conditions has as much to do with the nature of the supervision as it does with who is being supervised. Often, intensive supervision tends to focus on violent and sexual offenders. This in turn makes these offenders appear more likely to violate conditions, which is then used to justify even more supervision, perpetuating a problematic cycle (p. 29, 12.1.2).

On the other side of the spectrum, for low risk offenders, anything more than very occasional meetings with probation officers appears to run the risk of being a waste of resources. A study carried out in Philadelphia identified low risk offenders who were randomly assigned to low intensity supervision (supervised by probation officers with large caseloads with an average of 2.4 contacts per year with the offender) and normal supervision (receiving an average of 4.5 contacts). The two groups did not differ on whether or not the offender recidivated within one year. In addition to not providing any criminological benefit, intensifying supervision of low risk offenders may be a misallocation of resources (p. 30, 11.4.8). Even when offenders violate conditions, requiring them attend a “day reporting centre” appears to be no better, if not worse, in terms of re-offending or violations of conditions, than simply allowing them to continue in the community with ordinary supervision (p. 31, 13.6.4).

There exists a delicate balance between support and surveillance with community supervision. A shifting of that balance to favour over-reliance on surveillance may create more problems than it resolves. For example, a study examining the impact of frequency of drug testing of youths on parole under the control of the California Youth Authority found no evidence of increased frequency of drug testing lowering recidivism (measured by re-arrests). Youths were randomly assigned to one of five conditions which varied the frequency of drug testing so the groups can be considered equivalent on all dimensions except for the frequency of drug testing. High levels of drug testing increased rather than decreased arrests for violent and total offences in the 42-month follow-up period (p. 32, 5.3.3).

Finally, community supervision programs are less effective than they could be in cases where probation officers do not adhere to basic principles of effective supervision. In order to understand why probation services may not be as effective as it could be, a case study was carried out in Manitoba, Canada, examining probation files and audio recordings of probation officer—client meetings. This study starts with the principle that there is no point in focusing resources on those offenders not in need of intervention; therefore, offenders who are at a high risk of reoffending need to be identified and targeted. The next step is to clearly identify the goals of the intervention; and provide effective interventions to meet those goals. The study found that “probation officers can learn to do more and do it better” (p. 33, 9.6.3). Most probation officers administered risk assessments but there was a corresponding intervention plan for only 40% of the identified needs. In addition, analysis of the audio files of conversations between the probation officer and offender indicated that criminogenic needs were not discussed in the majority of cases. More broadly, community supervision may be limited by the availability of programs. For example, 40% of adult offenders in the study had employment problems noted but only 10% of these cases had an action plan designed to address this (p. 33, 9.6.3).

b) What Can Work

What is encouraging is that “doing supervision better” can have an impact on re-offending. In a study of community correction officers in Victoria, Australia, officers were given a training course in which they were encouraged to: 1) promote and reward “prosocial” actions (e.g., keeping appointments, doing community work); 2) focus specifically on problems that appeared to be related to offending; and 3) try to see things from the offender’s perspective. The clients of these trained officers were compared with the clients of officers who did not follow these principles. The results were quite positive, particularly with the focus on “prosocial” orientation of community correctional workers being associated with fewer breaches of conditions and fewer offences, one and four years later. This study shows that, even without “special”, “expensive”, or “new” treatment, particular orientations or supervisory styles can affect the outcome of community corrections (p. 34, 1.1.8).

In addition to promoting a “prosocial” orientation in supervision, moving away from a strict enforcement/control model seems to be more effective in reducing recidivism. In a study of the ‘Proactive Community Supervision’ (PCS) model in Maryland, community supervision programs that provided motivational enhancement, more conducive social learning environments and a targeted emphasis on core criminogenic needs were found to be more effective at reducing re-arrests than providing ordinary probation (p. 35, 9.6.4). The targeted nature of this intervention in focusing on core criminogenic needs played an important role but this type of targeted intervention delivered in a “correctional milieu where offender change is supported” (p. 35, 9.6.4)—not simply demanded and enforced—is what had positive impacts on offenders.

Similar findings were also reported in an evaluation of the ‘Intensive Surveillance and Supervision’ Program (ISSP) in New Jersey. In addition to providing more appropriate services to the offenders, the program tested whether parole officers were categorized as being law enforcement, social work, or balanced in their orientation. Overall, the rate of recidivism for a new conviction or revocation of parole for any reason was lower for the ISSP group receiving more appropriate services for their needs than the ordinary parole group. Moreover, technical violations/revocations varied by the orientation of the parole officer. Revocations for a new conviction was least likely when the parolee was being supervised by a parole officer with a balanced orientation. Technical violations were, not surprisingly, most likely noted for parolees being supervised by officers with a law enforcement orientation (p. 36, 8.2.2)

However, balancing control/enforcement with help/rehabilitation is not sufficient. In addition, community sanctions and programs need to be clearly conceptualized and properly implemented. Programs with clear goals of what the program should achieve while targeting a specific type of offender that is supposed to benefit from the program are more effective at lowering recidivism rates than a “one size fits all” type of approach (p. 37, 2.2.2). 

Having said that, there should be a caution against using commercially marketed prediction instruments in identifying and targeting a specific type of offender who is more likely to re-offend. Although prediction instruments are attractive to criminal justice professionals because they give a veneer of objectivity to criminal justice decisions, popular prediction instruments are no better at predicting whether someone will re-offend than ordinary people’s risk predictions using readily available indicators (e.g., age, sex, criminal history). This has been documented for COMPAS, formerly Northpointe (p. 38, 17.2.1), LSI-R (p. 39, 8.4.6), and Youth Level of Service Case Management Inventory (YLS/CMI) (p. 40, 9.6.5). In addition, these risk prediction tools tend to disadvantage Black defendants with false positives (p. 38, 17.2.1), have limited predictive ability for women whose pathways to crime are gendered (p. 39, 8.4.6) and African-American and Latino youth (p. 40, 9.6.5).

Overall, approaches that have been shown to be successful and effective in reducing breaches and recidivism tend to: emphasize a prosocial orientation in supervision style; strike a balance between enforcement/control and help/rehabilitation; provide appropriate services that are specifically targeted to the offender’s criminogenic needs; have a clear goal; and effectively allocate resources by specifically targeting offenders who are at higher risk of reoffending.