View this issue as a PDF: CrimHighlightsV16N1.pdf
This issue of Criminological Highlights addresses the following questions:
- Can judges reduce racial disparities in the criminal justice system?
- What can be learned from Canada’s successful decarceration of youths?
- How are the experience of debt and crime related?
- Do after-school programs for youths reduce crime?
- Why do juvenile drug treatment courts not seem to work?
- What is the most recent evidence on the effectiveness of treatment programs for sex offenders?
- What are the problems in implementing restorative justice programs?
- How well do all the ‘obvious’ explanations for the ‘crime drop’ in many western countries fit the data?
Judges use different types of strategies in responding to racial disparities in the criminal justice system. One of those strategy types – a noninterventionist approach that focuses only on judges’ own possible differential treatment – helps explain the persistence, despite well-intentioned judging, of racial disparities in the system.
There is little doubt that in many countries certain groups (e.g., Blacks, Indigenous people) are over-represented in the criminal justice system. Two explanations have been offered: The disparate impact explanation suggests that it relates to offending rates (e.g., as a result of differences in economic resources or in neighbourhood and school environments) or differential susceptibility to being charged (e.g., because of differential policing). The differential treatment explanation suggests that police, prosecutors, etc., engage in “overt or implicit discriminatory treatment” (p. 333). Understanding “judges’ interpretations [of racial disparities] would improve our understanding of their role in the reproduction of racial disparities” (p. 335). In this study, most (42 of 55) of the judges interviewed offered both explanations, but 13 of the 55 attributed racial disparities to disparate impact alone.
This paper suggests that judges, guided in part by their understandings of racial disparities, have different approaches about how best to address disparities at four different stages of the justice process – arraignment (where charges are accepted and pretrial release decisions are made), plea hearing, jury selection and jury management, and sentencing.
At each of the four stages of the process, judges interviewed employed either a non-interventionist or an interventionist strategy. For example, at plea hearings involving Black defendants, judges using a non-interventionist strategy reported simply accepting the plea agreement that had been offered. Alternatively, some judges reported using an interventionist strategy – where, for example, they might reject or question the agreement if it appeared to be different from what most White defendants would receive. With respect to jury selection, judges in this jurisdiction using an interventionist strategy might choose the jury foreperson to match the defendant’s race and do what was possible to ensure that there were at least some jurors chosen of the same race as the defendant. At sentencing, judges using an interventionist strategy might attempt to craft sentences that account for “the disparate impact of poverty and restricted life chances” (p. 351) on minority defendants. A judge using a noninterventionist approach at this stage might simply strive for consistency across racial groups within his or her own sentences.
Judges’ willingness to employ interventionist strategies varied across the four stages of the court process. About a quarter of the judges employed interventionist strategies at the arraignment and jury selection stages, but only about 15% at plea hearings and at sentencing. Although about half reported using non-interventionist strategies at all stages, all but one of the remainder reported having different approaches at different stages.
Conclusion: “Judges… employ two sets of strategies – noninterventionist and interventionist – that manifest themselves differently at different stages of the court process. Noninterventionist strategies largely allow racial disparities to go unchecked with respect to possible differential treatment by other actors and with respect to the disparate impact of the criminal justice system, whereas interventionist strategies attempt to address these processes” (p. 353). Hence “even when judges are not explicitly racist and even when they acknowledge, and attempt to account for, their implicit biases, they still may unintentionally contribute to racial disparities through noninterventionist decision-making that does not account for potential differential treatment by others or for the disparate impact of poverty or facially neutral laws….” (p. 354).
Reference: Clair, Matthew and Alix S. Winter (2016). How Judges Think About Racial Disparities: Situational Decision-Making in the Criminal Justice System. Criminology, 54, 332-359.
American cities that only recently became new destinations for immigrants experienced large decreases in levels of crime.
Until the 1990s, most immigrants to the US settled in a relatively small number of cities. Since, then, however, immigrants have begun settling in cities that, before 1990, had very few immigrants. This study looks at the impact of large increases in the number of immigrants in cities that, prior to 1990, had few immigrant residents.
There is substantial evidence that large concentrations of immigrants in US cities is, if anything, associated with lower levels of crime (e.g., Criminological Highlights 10(6)#7, 11(1)#4, 13(6)#7). This study asks a related question: what happens when immigrants move into ‘new destinations’ – cities that once had low concentrations of immigrants.
Crime data for 1252 American cities and towns with populations of at least 20,000 in 2007 were obtained for two periods – 1999-2001 and 2005-2007. “New destinations” were defined as cities or towns in which the number of foreign born increased by at least 150% between 1990 and 2005 and which had at least 1000 foreign born in 2005. The analysis examines the change in crime rates between 1999-2001 and 2005-7 as a function of changes in an index that combined the percent recent foreign born (those who entered the US in the previous 5 years) with percent Latino. Various controls were included including changes in the overall population, percent of the community who were males age 15-29, measures of economic disadvantage, and household instability.
“Despite surges in immigration of at least 150%…., new destinations appear to have experienced greater declines in crime than places that did not experience such growth. This was especially the case for property crime….” (p. 38). Generally speaking, across all 1252 cities and towns, those locations in which the concentration of immigrants increased tended, if anything, to have lower levels of overall crime as well as violent and property crime. However, the decreases were not statistically significant. The most important finding comes from an examination of ‘new destinations.’ Large increases in the numbers of immigrants between 1999-2001 and 2005-7 in these ‘new destinations’ were statistically significantly associated with decreases in overall crime, as well as in violent crime and property crime.
Conclusion: Social disorganization theory “suggests that immigration is one of several large-scale social processes that can operate to foster crime by increasing social heterogeneity and residential instability….” (p. 38). The findings “contrast sharply with hypotheses drawn from [this] theory…” (p. 39). It would appear that when new immigrants settle in ‘new destinations’, their impact overall is even larger than it is in cities where their increased presence would not be as obvious.” At this point, however, it is not clear whether the crime-reducing impact of high concentrations of immigrants in ‘new destinations’ is a result of self-selection of immigrants who settle in these ‘new destinations’ or if the reduction is the result of some other factor. What is clear, however, is that current non-immigrant residents of ‘new destinations’ do not have to be concerned that the arrival of new immigrants will be associated with increased crime.
Reference: Ferraro, Vincent (2016). Immigration and Crime in the New Destinations, 2000-2007: A Test of the Disorganizing Effect of Migration. Journal of Quantitative Criminology, 32, 23-45.
Experiencing problems with debt is intimately related to crime: Having debt problems increases the likelihood of committing crimes, and committing crimes increases the likelihood of experiencing debt problems.
Crime is not evenly distributed across groups that differ in wealth. Experiencing problems related to debt, however, is quite separate from income disparity or wealth. This study examines the reciprocal relationship between experiencing debt problems and criminal offending, controlling for various social and economic factors known to be related to offending.
There are many reasons to believe that experiencing debt and offending would be related. For example, those who are characterized as having low self-control appear to be more likely to commit crime, and they also might be more likely to incur debt. Alternatively, those experiencing debt may be more motivated to commit property crimes. This paper attempts to disentangle the causal relationship between debt and crime by, among other things, looking longitudinally at changes in debt problems and offending within individuals.
The study was carried out in Finland, a country in which it was possible to link records of registered debts, socioeconomic characteristics of the population, and offending data over a seven year period. The study focuses on 150,010 individuals who were between 20 and 30 years old at the start of the key period of the study (2005). There were, in a sense, two related studies: an examination of variation in crime at one point (2005-6) in time related to the characteristics of the sample, and a longitudinal study of the relationship between changes in debt status and offending during the period 2005-2012.
“Young people with low education, low income, and limited employment history have an above-average risk of ending up with debt problems” (p. 324). However, experiencing debt problems in 2005-6 was also related to higher rates of both property and violent offending in that same year even when other individual characteristics (age, sex, education, income level, whether the person was working or unemployed, whether they owned property, and whether their parents experienced debt in 2004) were taken into account.
When variation in offending within individuals across time was examined in relation to changes in debt problems, it was discovered that the relationship between experiencing debt problems and crime was still evident, but was much smaller than when this same relationship was examined across individuals, suggesting that some, but not all, of the overall relationship is due to differences in individuals. In addition, it was found that within individuals, “the increase in levels of crime takes place almost instantly after the start of the debt default period” (p. 322) reinforcing the conclusion that the experience of debt increases crime. After that initial increase, crime did not increase. In addition, however, crime-related debt (unpaid fines) increases after conviction and remains relatively high throughout the follow-up period.
Conclusion: When looking at individuals over a 7 year period, it appears that crime was more likely to occur during periods of debt enforcement. Though this association was stronger for property crime than it was for violent crime, the association between debt and violent crime was significant. The process, however, “is reciprocal, where levels of non-crime-related debt increase slightly before the crime is committed, and crime related fines in turn contribute to levels of debt problems when they enter into enforcement after being initially left unpaid” (p. 325). If we are interested in encouraging those who have committed offences to desist from crime, we should consider the possibility that “an emerging desire to leave a life of crime behind is hampered by the burden of debt caused by the earlier criminal lifestyle and resulting periods of incarceration” (p. 327)
Reference: Aaltonen, Mikko, Atte Oksanen, and Janne Kivivuori (2016). Debt Problems and Crime. Criminology, 54(2), 307-331.
After-school programs designed to prevent youths from committing offences are often viewed favourably by youths, but a review of high-quality evaluations of these programs suggests that one cannot count on them to reduce offending.
After-school programs are seen as having the potential of reducing offending by youths. Theoretically, they could lead to a reduction in offending for a number of reasons including the possibility that they reduce the amount of time youths spend unsupervised by adults. They might also be effective if they reduced the amount of time that crime-prone youths spent with more delinquent youths. Alternatively, the activities that the youths are exposed to might themselves have crime-reducing effects.
This review examined all studies that could be found that evaluated the crime reducing impact of after-school programs. The programs that were examined were ones that targeted youths who otherwise would have been unsupervised. Outcome measures varied somewhat across studies but all involved some estimate of delinquent or antisocial behaviour.
The first challenge, however, was to find evaluations that were high quality. A key component of being high quality was that there had to be a comparison group that was similar to the treatment group. The design could involve a randomized experiment or it could include some other approach resulting in an equivalent group of youths who did not have access to an after-school program. Over 11,000 papers about after-school programs were found, though only 36 could be found that would qualify as evaluations of the programs. Of these, only 12 studies qualified as competent evaluations with sufficient data and other relevant information available to be able to understand and evaluate both the design and the results.
Of these 12 studies, only three reported significant crime reduction associated with participation in the after-school program. More concerning is the fact that two studies reported findings showing that participation in after-school programs was associated with increased offending. Pooling across all 12 studies, the only conclusion one could come to is that, as a group, after-school programs have not been shown to have favourable impacts on crime.
Intervention treatments varied across studies. Five studies involved academic programs, 2 focused on recreation, and 5 involved skills training or mentoring. Treatments took place in different contexts; some focused only on ‘high risk’ youths; some involved high school students while others involved primary school students; different measures (official reports of crime vs. self-reports) were used; some studies involved randomized experiments, others involved some type of quasi-experimental design. No grouping of studies had overall favourable effects. Hence it could not be demonstrated that some subset of after-school programs is effective in reducing offending.
Conclusion: Although one cannot conclude that after-school programs are effective in reducing crime, this does not mean that they should be discontinued. As one researcher noted, they “serve real needs for parents. Even without evidence of crime prevention effectiveness, public expenditures on [them] may be well-spent” (p. 285). After-school programs are designed with many purposes in mind. Crime prevention may be only one such purpose. What is clear, however, is that to the extent that after-school programs are justified as having favourable impacts on youths, it is necessary “to ensure that the program is implemented as planned and remains true to theoretical and operational design parameters” (p. 286). In addition, if they are justified by their effects on youth, rigorous research designs need to be built into the programs that will allow strong inferences about their effectiveness.
Reference: Taheri, Sema and Brandon C. Welsh (2016). After-School Programs for Delinquency Prevention: A Systematic Review and Meta-Analysis. Youth Violence and Juvenile Justice, 14(3), 272-290.
Juvenile drug courts vary considerably in their goals, target populations, and treatment activities. A study of nine such courts found consistency on two dimensions: there was no evidence that they reduced reoffending and they focused their efforts on youths who did not generally use drugs or alcohol in ways that were unusually harmful.
Specialty courts are popular for many reasons, among them being that those brought to criminal courts – as youths or adults – often are facing other challenges. The question, however, is not whether many who are brought to court have various mental health or related needs that might benefit from treatment, but rather whether criminal courts are well-placed to provide effective help.
This paper examines a set of 9 juvenile drug courts in four regions in the US. The courts were similar in that after identifying youths who were using illegal drugs, the courts rewarded the youths for cooperation and punished them for non-compliance. Typically youths were released when it was deemed that they had been cured of their drug problem. However, as often is the case when the focus is more on the name of the institution (‘drug courts’) rather than the substance of the treatment (what happens to the youths), there seemed to be little research-based treatment knowledge driving the operation of the drug courts.
686 youths in juvenile drug courts in 9 jurisdictions were matched with 686 ‘ordinary young offenders.’ Youths were matched on a risk assessment instrument, race, gender, and drug/alcohol abuse/dependence. Though the matching was not perfect, differences between the groups were subsequently controlled statistically. In addition, a separate analysis including only those who constituted ‘good matches’ was carried out, with substantially similar results.
One problem with juvenile drug courts was evident from the start: “The majority of [youths in drug courts] did not engage in problem [drug] use nor did they take drugs that might be considered to be more serious, such as cocaine or opiates. Additionally, [another measure obtained from the youths] suggests that these youths had fairly low treatment readiness… This might limit responsiveness to intervention” (p. 309).
Two measures of recidivism were used: arrest/referral to court and formal adjudication of charges by a court (juvenile or adult). Compared to control youths, juveniles sent to drug court were more likely to be arrest/referred to court and more likely to have their charges formally adjudicated. This was examined both for the period of time when the youths were being supervised and afterwards. Statistically, the youth’s measured risk level, age, sex, and race were all controlled in these analyses. In general, the youths in the drug courts were more likely to recidivate than ‘ordinary’ youths. Drug court youths were more likely to fail drug tests, to be found to have violated a condition of supervision or to show violations related to schools, perhaps as a result of higher levels of supervision. A separate analysis involving those who completed drug court or probation successfully also found that those going to drug court had, if anything, higher reoffending rates.
Conclusion: Although research has found that some drug courts for adults have favourable outcomes, this study underlines one of the standard effects of ‘treatments’ assigned to youths: They tend not to be effective when assigned to youths who have little need for treatment. These nine drug courts contained significantly more low risk youths than the control group. Furthermore the regime generally lasted 9-18 months. The drug of choice for the youths was generally marijuana (71% of the youths) or alcohol (24%). 67% reported using drugs or alcohol less than once a week. In addition, the treatments that were used – talk therapy and education – have been shown to have limited effects, and families and other caregivers were typically not involved. Creating a ‘drug court’ clearly is not on its own sufficient. Attention needs to be focused on who is sent to such special programs, what the treatment consists of, and how it is delivered.
Reference: Sullivan, Christopher J., Lesli Blair, Edward Latessa & Carrie Coen Sullivan (2014). Juvenile Drug Courts and Recidivism: Results from a Multi-site Outcome Study. Justice Quarterly, 33(2), 291-318.
A previous review of sex offender treatment programs (Criminological Highlights 9(5)#7) found that sex offender recidivism could be reduced with appropriate treatment. Determining the effectiveness of sex offender treatment is difficult because sex offending itself is a very heterogeneous category, treatment approaches vary dramatically, and assessing effectiveness is difficult because rates of sexual recidivism are quite low. Most studies of sex offender treatment, however, are inadequate – typically because of the difficulty of identifying appropriate comparison groups.
This paper looks only at those studies in which there were no serious doubts that the comparison group (i.e. the group that did not receive treatment) was equivalent, prior to treatment, to the treatment group. It is, therefore, slightly more selective than the earlier review in that the requirement of an equivalent non-treatment control was more rigorous. In all, after examining more than 2,000 papers, the researchers identified 27 studies (involving 29 independent comparisons of a treatment group with a control group) that were scientifically rigorous enough to be included in this review. About half were published after 2000.
The majority of the studies (21 of the 29) involved cognitive-behavioural treatment and almost all (26) involved treatment programs aimed specifically at sex offenders. They were largely carried out in Canada (11 studies), the US (8), Great Britain (3) or Germany (3). Only 6 of the 29 studies involved random assignment. Outcome measures consisted of conviction for a new offence for some studies and arrests or new charges for others. Follow-up periods ranged from one year to almost 20 years, with 24 of the 29 comparisons involving follow-up periods of more than 3 years. The 29 comparisons involved data from 4,939 treated and 5,448 untreated sex offenders.
The average sex offence recidivism rate (pooled across studies) for those who were treated was 10.1% compared to 13.7% for those not receiving treatment. Though this difference was highly significant, it is obviously not large, in part because the rates of committing new sex offences for those convicted of sex offences are so low.
Some studies also reported rates of general reoffending. The average general reoffending rate for the treated groups was 32.6%; for the untreated it was 41.2%. All of these figures should be interpreted cautiously since the follow-up periods varied considerably across studies. What is clear, however, is that the treated group generally did better.
The largest effects came from two studies involving multi-systemic therapy. The size of the effects from the 20 cognitive-behavioural therapy treatment studies was statistically significant, but much smaller. Other forms of treatment did not show statistically significant positive impacts.
Conclusion: “Taken together, the… analyses of reasonably well-controlled evaluations suggest that treatment of sexual offenders can be effective, but the results are not homogeneous. In particular, treatment in prisons and pure group formats seem to be less promising [than treatment outside of prison and treatments that include at least some individual therapy] …. Research and practice should ask more frequently what [form of therapy] works for whom, in what contexts, under what conditions, with regard to what outcomes, and also why” (p. 623). Knowing that change in re-offending rates is possible with those who have committed sex offences is a good first step, but it is only a first step.
Reference: Schmucker, Martin and Friedrich Lösel (2015). The Effects of Sexual Offender Treatment on Recidivism: An International Meta-Analysis of Sound Quality Evaluations. Journal of Experimental Criminology, 11, 597-630.
Although we know quite a bit about the effects of restorative justice conferences, there appears to be a reluctance to use these techniques on a systematic basis in many countries.
Previous research has suggested that restorative justice conferences can be useful in responding to criminal offending, though there is some evidence that suggests that they are more likely to be effective as a supplement than as a substitute for normal case processing (Criminological Highlights 9(5)#1, 10(1)#6, 15(4)#4).
This paper looks at 12 randomized experiments examining the effects of restorative justice conferences (RJC) in Australia and England. Unlike many restorative justice interventions, there was a serious attempt to test the effect of a single controlled model of delivery or conferences in which trained facilitators were “told to ask just 3 questions… and to ensure that everyone had a chance to say all that they wanted about each question. The questions were (1) What happened? (2) Who was affected by it and how? and (3) What should the offender do to try to repair the harm caused by the crime?” (p. 505).
In some of the 12 experiments, victims who participated in RJCs were less fearful of a repeat attack from the same person. Victims also tended to be more pleased with the manner in which their cases were handled if they had participated in an RJC.
For certain types of conferences, the frequency (but not the likelihood) of reoffending was reduced in the short, but not the long term. Findings from two of the studies suggested that for certain types of cases, the experience of an RJC was, at least in the short term, criminogenic, though these effects did not persist over time. Generally, however, the results suggest that there is a reduction in subsequent offending associated with participation in an RJC, though it is important to note that the effects are not uniform. Furthermore, it is clear that those offenders who did participate in the conferences were more favourable about the criminal justice process than were those who received normal criminal justice processing.
The history of these RJC random-assignment experiments was not straightforward. It was never easy to get cases referred to RJC programs, particularly in those locations in which the RJC (on a random basis) was to be used as a substitute for court processing. In two of the studies, it was estimated that only 11%-12% of ‘eligible’ cases were referred. Given this high degree of selection of cases, it is difficult to know whether the results would generalize to a less highly (police-selected) sample. Said differently, even if RJCs are effective for the type of cases that have been referred to this procedure, we don’t know if the findings would be the same for a broader selection of cases.
The researchers were successful in running actual experiments: Cases were successfully assigned at random to RJC or treatment as usual. And given that these experiments used a well-defined method of running RJCs, it means we know what treatment was actually delivered.
Conclusion: Restorative justice conferences (RJCs) clearly have favourable effects in some, but not all, types of cases. The difficulty in attracting cases to RJCs limits our ability to know whether such approaches would be effective with a more heterogeneous set of cases. One problem is that RJCs apparently are more likely to be conducted with people involved in minor matters than serious crimes, which may reflect “poor triage [since this practice means] giving RJCs to people who have little need of it, and denying it to those whose need is greatest” (p. 534).
Reference: Sherman, Lawrence W. and 10 others (2015). Twelve Experiments in Restorative Justice: The Jerry Lee Program of Randomized Trials of Restorative Justice Conferences. Journal of Experimental Criminology, 11, 501-540.
Most of the explanations that have been offered for the ‘crime drop’ that has occurred in many western countries are plausible sounding, but they are each almost certainly inadequate.
Whether one looks at the results of victimization surveys or police reported crime, it would appear that the rates of many categories of crime have dropped quite dramatically in recent decades in many countries including the US, Canada, England & Wales, Australia, and New Zealand.
Many explanations have been offered for the drop. This paper suggests that none of them is a sufficient explanation though some may offer a partial explanation for the drop in certain crime rates. Previous work has tended to suggest ‘single factor’ explanations, such as the aging of the population. However, changing demographics (see Criminological Highlights 2(6)#7, 5(4)#4) may account for a small portion of the drop for some offences but not all of it. Similarly, it can quite easily be shown that explanations based on a specific change in society, such as the availability of abortion, are almost certainly wrong (Criminological Highlights 9(6)#8).
This paper takes a different approach. It examines each of 17 hypotheses about the crime drop and uses four ‘tests’ of the ability of each explanation to account for the changes that took place, in recent decades, in crime rates. The four tests are the following:
(1) Can the explanation be applied to different countries? On the ‘abortion’ issue, for example, the crime trends for Canada and the US are very similar, but only the US had changes in the availability of abortion at the critical time in question.
(2) Before crime rates went down, they typically went up quite rapidly. Does the explanation account for this increase?
(3) Some rates for some crimes have increased recently. Can the explanation account for that variation?
(4) The timing of the crime drop. Can the explanation account for the fact that the changes in rates vary across types of crime?
Seventeen hypotheses have been proposed for the crime drop. These include the following:
changing demographics, a strong economy, consumer confidence/price inflation, laws that allow the carrying of concealed weapons, gun control laws, capital punishment, changes in rates of imprisonment, new or changes in policing strategies, more police, waning of the hard drug market, legalization of abortion, lead poisoning, immigration, civilizing processes, internet-induced changes in life styles, cell phone ownership and guardianship, and improved security systems.
As already noted, these explanations have one important thing in common: they choose a possible cause that could affect large numbers of people (e.g., the presence of lead), note that there is a change in that hypothetically causal variable, and then correlate that change in the causal variable to changes in crime (in this case, when youths who grew up in an environment when lead was presumably more likely to be in the atmosphere). The problem, however, is that the explanation may fit one set of data but not all the data. In the case of lead in the US, for example, “all proxies for lead increased dramatically from around 1910 through 1970. If the lead hypothesis is correct, then crime should have displayed a measurable increase between 1925 and 1985” (p. 451). Unfortunately for the hypothesis, crime rates were much more varied than one would expect.
Conclusion: Each of the 17 hypotheses that have been suggested as explanations for the crime drop was subjected to four separate empirical tests. No explanation for the crime drop was fully consistent with the data. One explanation – that we now have improved security (e.g., to protect from thefts of and from vehicles) – fits the data for certain crimes but is less persuasive for others. Variation in the propensity to commit crime within a society appears to be better understood than changes in crime rates within a society across time.
Reference: Farrell, Graham, Nick Tilley, and Andromachi Tseloni (2014). Why the Crime Drop? Crime and Justice: A Review of Research (Michael Tonry, ed.), 43, 421-490.