Criminological Highlights Vol. 22, No. 6 - February 2026

This issue of Criminological Highlights addresses the following questions:

1. How do Black and White Americans’ views of the criminal justice system differ?
2. Does a more lenient system for pretrial release lead to increases in crime?
3. Why is it in a prison system’s interest to try to attempt to reduce the security level of prisoners?
4. How can conditions of pretrial release set people up to fail?
5. What was the effect of “defund the police” movements in the US on police budgets?
6. How did Islamic terrorism increase anti-immigrant attitudes towards other non-Islamic minority groups?
7. How do programs designed to help people from acquiring criminal records disadvantage certain groups?
8. Do short sentences interfere with the ability of offenders to get employment?

Item 2

An administrative order that increased the likelihood that an accused would be granted pretrial release did not, during the period prior to the disposition of the case, increase the likelihood of new criminal activity or failures of the accused to appear as required in court.

In the early days of the COVID-19 pandemic, many jurisdictions looked for ways of reducing the size of their prison and jail populations in order to reduce the spread of the disease. Because these changes were made suddenly and for reasons unrelated to criminal justice goals, it was possible to estimate their effects on the operation of the system.

Previous research (see Criminological Highlights special issue on bail) has reported that accused people who are detained (e.g., they are processed by “tough” rather than “lenient” judges) are likely to have higher conviction, incarceration, and recidivism rates. “These deleterious outcomes are disproportionately experienced by members of racial and ethnic minority groups…” (p 605).

This study looked at people who were charged between 1 October 2017 and 30 June 2021. An administrative order made on 10 April 2020 (soon after the recognition of the COVID-19 pandemic) reduced required cash bond amounts (from $3000 to $1000) for certain felonies in one Florida county and made it possible for people to be released without having to appear before a judicial official. Because the change in policy was sudden (and without much notice), it was possible to divide accused people into two groups: those for whom pretrial release was handled as it normally had been and those who benefitted from the new more lenient policy. The effect of the sudden change, if any, associated with the change in policy could be differentiated from any long-term trend.

No “pretrial risk assessment” was carried out for those processed before or after the change in policy. However, to ensure that any changes in outcome were not attributable to changes in the nature of accused being considered for release, various measures were controlled in the main analyses including the offence type, prior record and failures to appear, race, age, gender, whether the accused was born in the US, education and whether they were married.  Clearly those dealt with under the new rules had lower cash bond amounts imposed on them; they were detained in custody for fewer hours; and they were more likely to obtain pretrial release. 

The multivariate analyses that included various control factors suggest that those who benefitted from the administrative order allowing easier release back into the community “exhibited [significantly] lower odds of new criminal activity… and failure to appear… than their counterparts charged with third-degree felonies during the period preceding the administrative order” (p. 612).

The administrative order mandating earlier release into the community had “a greater effect on defendants’ odds of pretrial release among Black individuals compared to their White counterparts” (p. 614).  But there was no evidence for any group – Black, White, or Latino – of any increase in failures to appear in court or in the occurrence of new charges.

Conclusion: A policy whose explicit purpose was to speed up the pretrial release process and to ensure that more people were released while awaiting trial was effective in accomplishing these goals. It is important to note that this “bail not jail” policy did not contribute to any increase in crime. In fact, it was the opposite: new criminal activity and pretrial failure were lower among those whose offences took place at a time where the lenient policy was in place.

Reference: Copp, Jennifer E. (2025). Cash Bail and Pretrial Compliance: Evidence from a Court-Imposed Policy Shock. Criminology & Public Policy, 24, 601-622.

Item 3

Placing prisoners in a high security prison (compared to a medium security prison) or in a medium security prison (compared to a low security prison) leads to higher rates of violent misconduct than would be expected if they had been placed at a lower level, even when individual characteristics are taken into account.

One theory of placing prisoners in high security prisons is that they can be better monitored and controlled and will have fewer opportunities for rule breaking. This study examines the rate of violent misconduct of first-time male US federal prisoners from the time that they were first placed in a cell related to their security classification until they were reclassified.

The study estimates the impact of prison security level on prisoner behaviour by looking at the relationship of a prisoner’s assessed “risk” score to their violent misconduct in prison. Almost all prisoners were placed in a facility matching their risk scores. With a continuous quantitative risk measure, this means that those at one side of a cutoff point (e.g., between medium and high security) are very similar to prisoners who have very similar scores but are on the other side of the cutoff point and are, therefore, placed in a different security level facility. The study uses this discontinuity to examine the impact of the prison placement independently of the risk score. Other factors (e.g., age, race, sentence length, offence history, etc.) that were not an explicit part of the “risk score” calculation were also controlled.  Since prison staff can modify placement for perceived safety or management reasons, the data were examined both including and excluding those whose placement wasn’t solely determined by the risk score. 

The main analyses looked at the relationship between risk score and the proportion of people with each risk score who had violent misconducts during their first prison placement. If prison placement makes a difference above and beyond the numerical risk score, one might expect a substantial change in the relationship between risk and misconducts right at the point where the prison placement changes (e.g., from low- to medium- security).

If higher security placement increases misbehaviour, it could be termed an “inflator” effect “where higher security increases misconduct” (p. 307). The change was clear: there is “a discontinuous jump in the misconduct prevalence as one crosses the low/medium-security cutoff into medium security… [just as there appears] to be change in behaviour starting at the medium/high cutoff” (p. 312-313). In other words, medium and high security placements had the effect of increasing violent misconduct compared to the behaviour of prisoners in the neighbouring lower security levels. This increase – determined by a technique known as a “regression discontinuity design” – was significantly larger than would be expected from changes in risk values of those in the lower and higher security groups. Said differently, risk scores did predict misconduct, but the effect of security placement was evident above and beyond the impact of the risk scores themselves.

These results look only at the impact of prison placement during, roughly, the first 7 months of a prisoner’s federal prison sentence. The effect of the security level placement when “minor misconduct” (e.g., refusing staff orders, refusing a work assignment) was examined was smaller and less consistent. The effect on women was not examined because almost all women (of a much smaller sample) were in low or minimum security institutions.

Conclusion: The results of the study strongly suggest that one of the impacts of placing men who have just arrived in a prison in higher security levels in prisons is that they will engage in higher rates of violent behaviour than they would if they were placed in lower security prisons. In other words, violent misbehaviour on the part of some prisoners relates in part to the decisions of prison authorities on where to place newly arrived prisoners.

Reference: Gwinn, Jason D., Miles D. Harer, & Neal P. Langan (2025). Effects of Prison Security Level on the Violent Misconduct for New Federal Prison Entries. The Prison Journal, 105(3), 305-329.

Item 4

When an accused person in Canada is released prior to trial and subject to a bail supervision program, the conditions of release can easily set accused people up to fail.

Canada is facing many challenges in the manner in which it treats people charged with offences and awaiting trial. Most obvious, perhaps, is the fact that (in 2023) almost half (49.3%) of Canada’s adult prisoners were awaiting trial rather than serving sentences. In an increasing number of cases, the accused must demonstrate that pretrial release is justified and, if they are released, must obey conditions that not only are punitive, but also may be difficult to comply with.

The number of people identified by the police as having committed the offence of “failure to comply with a court order” (largely those charged with violations of terms of pretrial release) is 51% higher in the most recent 5 years for which we have data (2020-2024) than it was 20 years earlier. During this same period, the number of criminal offences recorded by the police decreased by 14%.  

Previous research (e.g., Criminological Highlights 21(1)#7, 13(5)#5), 16(6)#4) has suggested that the conditions faced by accused people who are released while awaiting trial often create unnecessary punishments for them.  This paper examines these conditions from the perspective of those who have the direct responsibility to supervise many of those on pretrial release in the Province of Ontario – i.e., people who are employed by “Bail Verification and Supervision Programs.” These programs typically are used when sureties are unavailable and support and control of accused people in the community is seen as being necessary. 33 employees of 19 such organizations (of 22 across the province) were interviewed.  Although the sample may not be completely representative of all employees, the basic views were similar across a substantial portion of the respondents.  

Although conditions of release are “ordered” by the court, in most cases accused people do not contest conditions suggested by the Crown because doing so may put them at risk of being detained rather than being released. “Bail at all costs” is the overwhelming priority for the accused and, therefore, the lawyer representing the accused. Not surprisingly, therefore, almost half of the bail supervisors agreed with the proposition that bail conditions set the accused up to fail, since the conditions are often difficult or impossible for the accused to follow. A number of the supervisors noted that in recent years conditions are worded in a more sensitive (and sensible) manner. For example, the condition of abstinence from drugs and alcohol may sound sensible, even though it is likely to be ineffective to “order” accused people to end their addictions. But some of these bail supervisors noted that abstinence conditions are now often restricted to the accused’s behaviour outside of their residence.  However, sometimes conditions simply don’t make sense (e.g., curfews for people without stable residences or requirements that the person live in a shelter when shelter space cannot be guaranteed). In some instances, where multiple conditions are imposed by different courts, nobody has bothered to combine them into one record to ensure that they are not inconsistent. Similarly, non-communication orders are rather difficult when they are imposed on only one of two co-accused. Finally, it appears that in some instances simple errors (e.g., in addresses) are made in the required conditions. Such errors can result in an accused being arrested because they are living in a residence not specified on the release order. 

Conclusion: Canada’s law specifies that, in most cases, conditions of release should only be imposed if necessary. These conditions are presumably imposed to ensure that an accused will not commit offences while awaiting trial and will appear in court as required. Although some courts appear to be increasingly sensitive to these restrictions, recent legislative proposals seem to be designed to increase the detention of accused people and, in some instances, to impose additional restrictions on those released. What clearly is needed is some systematic investigation of the effects and effectiveness of different release conditions so that conditions of release have empirical justification.

Reference: Yule, Carolyn & Laura MacDiarmid (2026). ‘It’s a Set Up’: Examining the Relationship Between Bail Conditions and the Revolving Door of Justice. Current Issues in Criminal Justice, 38, 40-57.

Item 5

In the US, “Black Lives Matter protests did not lead to the defunding of municipal police budgets” (p. 1211). Indeed, in cities dominated by Republicans, these protests may have had the effect of increasing police budgets.

The suggestion that there should be targeted defunding of police emerged in the early 2010s. However, in 2020, when “defund the police” Google searches were prevalent in the US, this policy proposal was opposed by more Americans (58%) than supported it (31%). This paper examines the effect on police budgets in different cities and finds that the impact of protests against the police is not a simple one. 

To understand the impact of protests on the funding of local police, the city budgets and the approved number of police personnel of 264 of the largest cities in the US were examined for 3 years (2018/9 through 2020/1).  The focus of the study was the share of the city budget assigned for policing. The study examined the impact on police budgets of the number of protesters in each city from 25 May 2020 (the date of the police murder of George Floyd) to 31 December 2020 before the city budgets were approved.

Overall, there was no relationship between the size of the Black Lives Matter protests and the proportion of city budgets transferred to the police or to changes in the number of authorized police positions. However, when the researchers examined the relationship of the amount of Black Lives Matter protesting separately for cities with varying proportions of voters who supported the Republican presidential candidate in 2020, the effect of having large numbers of Black Lives Matter protesters predicted increases in police budgets. For cities with low Republican vote share, there was essentially no consistent impact (positive or negative) of the strength of the protest on police budgets. Other analyses – looking at the strength of police funding at an earlier time (2018-9) – suggest, however, that this “policy backlash to protest” (more protests against policing funding, more funding of police) “may stem from an entrenched conservatism in Republican-governed cities” (p. 1210).  Obviously, these data looked at the immediate impacts of the protests.

It seems plausible that “policymakers’ own conservatism and right-wing influences in Republican cities might explain the policy backlash in Republican cities” (p. 1211). “Cities with large Democratic vote shares neither experience a policy backlash nor reduced their police budgets” (p. 1211).

Conclusion: The immediate (or short term) impact of Black Lives Matter protests arguing for reductions of police budgets do not appear to have been successful in changing political actions on the hiring of police or on police budgets. In Republican dominated cities, protests seem to have been used by city politicians to show their support of the police. Cities that supported Democratic politicians seem not to have been affected in any consistent manner. 

Reference: Ebbinghaus, Mathis, Nathan Bailey and Jacob Rubel (2025). The Effect of the 2020 Black Lives Matter Protests on Police Budgets: How “Defund the Police” Sparked Political Backlash. Social Problems, 72, 1198-1215.

Item 6

Terrorism by an Islamic group increased not only anti-Islamic sentiment in Europe, but it also increased anti-immigrant attitudes toward Jewish and Roma minorities.

In January 2015, a group of Islamic terrorists attacked the offices of the Charlie Hebdo magazine in France and killed 19 people in two related attacks. Its effects were felt all over the world. This study suggests that events like this alter symbolic boundaries that define who is included in one’s group and who is excluded. Religion – whether actively practiced or simply the basis of the origins of national institutions and identities – and race are likely to be two important factors defining a person’s identity.

At the time of the attack, data for the European Social Survey were being collected. Hence it was possible to compare the responses of those immediately (30 days) before and after the attacks. The focus of this study is on the attitudes of the “native population” of various European countries toward particular immigrant and minority groups: Muslims, Jews, and Roma. Because of sample size limitations, the focus was on three European countries that are traditional immigrant destinations (Germany, France and Belgium) and three that are not (Finland, Czech Republic, and Ireland). 1,814 people from these countries were interviewed before the attack and 1,638 after. The main measure of immigrant sentiment were questions that asked whether members of these groups should be allowed to come and live in their country. Respondents were also asked their views on specific characteristics that the respondent believed a hypothetical immigrant must have (White, Christian background, work skills, commitment to the host country’s way of life, etc.). The focus of the study is on change in these measures associated with the Charlie Hebdo attacks.

In Europe, it would appear that the primary characteristics that define symbolic boundaries between the native population and ‘others’ are race and religion. After the Charlie Hebdo attack, there were significantly higher anti-Muslim, anti-Jewish, and anti-Roma sentiments expressed in European countries. Clearly, the anti-outgroup sentiment spilled over to “out-groups” that were not involved in the attack. Compared to before the attack, higher numbers of respondents said, after the attack, that their country should “allow none” of each of these three groups to come to their country. When looking at the details of whom people felt should be allowed to move to their country, it appeared that it was race and religion that were most important in explaining the growth in the symbolic boundaries against immigrants after the Charlie Hebdo attacks.

Increases in anti-immigrant sentiment were especially strong in the non-traditional immigrant destination countries. “Countries with a long-standing immigration flow moderate immediate effects of terrorism on anti-immigrant sentiment” (p 56).

Conclusion: “The discriminatory processes of specific native European populations are not necessarily manifested through precise social categorization but rather [they are] based on blurry symbolic boundaries that define ‘us’ or ‘them’…. Narrowly focusing on groups [whose members may have been involved in particular events] may overlook the challenges such extreme vents pose for broader minority populations” (p. 60).

Reference: Ramirez, Daniel and Joeun Kim (2025). “Not One of Us”: Anti-Immigrant Sentiment Spread to Multiple Immigrant Groups in Wake of Islamic Terrorism. Social Forces, 104, 44-66.

Item 7

A procedure that allows judges effectively to nullify a conviction after guilt has been established was designed for various purposes including avoiding the consequences of a felony conviction. Unfortunately, racialized and lower socioeconomic status defendants do not get the same level of benefits as do White and wealthier people.

“Deferred adjudication… offers legally guilty defendants protection from the mark of a conviction conditional on the completion of community supervision” (p. 154). This paper examines whether race and socioeconomic status influence who gets a chance, through this procedure, at avoiding a criminal conviction. 

Many jurisdictions have procedures whereby people who admit to committing an offence can avoid a formal conviction by completing a period of supervision in the community. (In Canada, for example, there are “alternative measures” provisions in S. 717 of the Criminal Code that allow someone who admits to an offence to avoid formal conviction in certain circumstances.)  This paper examines the deferral process at two points. First, it examines judges’ decisions (usually on recommendation from the prosecutor) to defer adjudication after guilt has been determined. Second, it examines the final decision of whether the charges should, in the end, be dismissed (without ever entering a finding of guilt) rather than convicting the person. Previous research has found that “deferred defendants have better employment outcomes and lower rearrest risks than otherwise similar defendants [who were convicted]” (p. 158-9).  Hence full implementation of the provisions can improve the life chances of an accused person.

This study was carried out in Texas (the US state with the largest prison population), a state in which those convicted (but not those deferred) can have imposed on them lifetime bans on certain welfare benefits. In addition, in Texas, costs for programs imposed on deferred accused people (e.g., required classes, drug testing, supervision fees) are paid by the defendant, thus creating structural barriers to success for those with few economic resources as well as difficulty for those whose cases are deferred contingent upon their involvement in certain programs.

Data from 69,264 cases in one Texas county filed between 1995 and 2014 were examined. Race and ethnicity were available from case files. Socioeconomic status was estimated from the neighbourhood median income and whether the accused had a court-appointed or privately-retained lawyer. Control factors about the case included sex, age, offence type, severity, number of charges, the accused’s criminal record as well as the identity of the judge and prosecutor. A statistical technique was used to control for the non-random selection of cases for deferral.

Controlling for characteristics of the case, Black men were less likely than White men to have their cases deferred. Independent of race, “defendants who reside in wealthier neighbourhoods are significantly more likely to receive a deferral than those who reside in poorer neighbourhoods” (p 164). White men benefitted most from having a hired attorney. For women, the impact of a hired lawyer did not differ by race.

Turning to the findings on whether a person who was deferred had their case dismissed after their period of supervision in the community, the data show that Black and Latinx men and women were less likely to have their charges dismissed. Racial disparity was more pronounced at this final stage – case dismissal – than it was at the stage of offering to defer the charges.

Conclusion: The findings of this study demonstrate that procedures designed to “sever the mark of conviction from the legal establishment of guilt [can] exacerbate racial and socioeconomic disparity, even when they also benefit people from marginalized groups” (p. 171). If one were looking at formal records of criminal convictions, therefore, it would appear that a procedure designed to avoid the stigma of a conviction accentuates racial and socioeconomic differences across groups.

Reference: Bing, Lindsay and Carmen Gutierrez (2025). Who Gets a Second Chance? Compliance, Classification, and Criminal Conviction. Social Forces, 104, 154-176.

Item 8

For people who were employed before they were arrested, receiving a prison sentence of more than two months reduces significantly their likelihood of getting a job after serving their sentence. Equivalent people who were imprisoned for less than 2 months were not as likely to be disadvantaged in getting employment.

After a period of incarceration, people have special difficulties getting jobs. This paper examines this relationship in detail, looking not only at the length of the incarceration, but also at various comparison groups to try to understand the nature of this relationship.

Most incarceration periods are quite short. (For example, recent Canadian court data suggest that 58% of prison sentences are for one month or less and 19% are between 1 and 3 months in length.) When looking at the effects of these periods of incarceration on issues like employment, it may be important to differentiate “long” from “short” periods of incarceration. But in addition, it may be important to think carefully about which groups to compare. This study examines a number of different comparison groups. In this summary, however, we will focus on people who had been employed before they had been arrested and for whom this was their first adult arrest (that may or may not have resulted in a prison sentence). More specifically, the focus is on the impact on employment for two distinct groups: those who experienced imprisonment for 2 months or more and those who were imprisoned for less than two months. The comparison group, for each of these groups, was those who were arrested, but never incarcerated. 

The data from the study came from the US National Longitudinal Survey of Youth. Survey participants had first been interviewed when they were youths. They were then interviewed regularly well into their adulthood. Because they had been interviewed so often, “control variables” (demographic, behavioural, educational, etc.) relevant to their experiences in the labor market were available.

The focus was on their labour market experience during the 6-month period immediately following each person’s first adult incarceration or, in the case of the comparison group that was not incarcerated, the labour market experience following their first adult arrest that did not result in incarceration. All of the groups described here had been employed prior to their arrest.

On all the measures of labour market success – whether they were employed 6 months after arrest, the proportion of the time in that first 6 months that they were employed, or whether they were, for any reason, out of the labour market – those who were incarcerated for more than 2 months were less likely to be employed in the 6-month period following the end of their incarceration.  However, those who received short sentences (2 months or less) did not show this same deficit. (On one of 5 employment measures – being out of the labour market – those who received short sentences showed a deficit, but it was significantly less than the deficit of those who had been incarcerated for more than 2 months).

Conclusion: Experiencing a short period of incarceration certainly disadvantages people in various ways compared to those who are not involved in the criminal justice system at all, but the impact is not generally apparent when comparing this group to those arrested but not incarcerated. It would appear that the impact of short periods of incarceration should be examined separately from long periods of incarceration for at least two reasons: Short periods of imprisonment may involve very different challenges for the prisoner compared to those imprisoned for longer periods, and by combining short and long periods of incarceration into one group, the impact of long periods of incarcerated may be less evident.

Reference: Siennick, Sonja E. and Alex O. Widdowson (2025) Do Short Spells Matter? Conceptualizing and Testing the Effects of Brief Incarcerations on Labour Market Outcomes. Journal of Research in Crime and Delinquency, 62, 856-899.

This issue of Criminological Highlights was prepared by Anthony Doob, Rosemary Gartner, Maria Jung, Tyler King, Jane Sprott, Danielle Van Wagner, and Bond Zhang