This issue of Criminological Highlights addresses the following questions:

1. How do police networks affect police misconduct?
2. Do Black Americans want to reduce police funding?
3. How can the design of a prison affect prisoner self-harm, violent tendencies, and overall well-being?
4. Do “liberal” bail laws contribute to crime?
5. Are very short prison sentences a good substitute for longer periods of probation?
6. How do long prison sentences have especially punitive impacts on Black prisoners?
7. Why do Black women achieve higher levels of education attainment than Black men?
8. How does a nearby homicide affect young women’s lives?

Item 1

Police officers in Chicago with strong ties with another officer who was injured in the course of their duties are more likely than other similar officers not associated with an injured colleague to be cited for misconduct after these events. This finding was more pronounced when the person suspected of doing harm to the police officer was Black.

Police tend to see their work as inherently dangerous, often leading to an us-versus-them approach to their interactions with people while doing their jobs (Criminological Highlights 21(2)#2). This paper examines the impact, on an officer’s behaviour, of having a close police associate injured as a result of contact with a civilian. Specifically, the study looked at whether having an associate injured increased the likelihood of a police officer being the subject of a civilian complaint or being cited for excessive force.

The paper begins with the hypothesis that threats to the police are often perceived to be coming from Black and Hispanic men in high-crime neighbourhoods. Hence a police officer may be more likely to believe that there is a “War on Cops” (p. 165) if a person close to them experiences harm from a Black civilian. Although there is “some evidence that a small number of [police] officers generate a large number of [citizen] complaints” there seems to be little agreement about what predicts police misconduct other than the fact that “female [police] officers are overall less likely to be involved in contentious situations” (p. 161).

The study uses data (obtained through freedom of information requests) on complaints against active Chicago police officers, and personnel records, etc. for the period 2004 to 2015. The relationships among police officers were defined by whether police officers were ‘partners’ (working together on at least 2 arrests in the previous year) or ‘collaborators’ (1 arrest), or where they were in the same ‘cohort’ (they had trained together). The focus was on whether a given police officer had a partner, collaborator, or cohort member who had been injured by a civilian. Those who had such a contact were compared with officers who did not have a partner, etc., who had been injured. Various control factors (e.g., misconduct history and recent injury to the officer, race, age, and sex) were included.   

If a police officer had a partner (but not a ‘collaborator’ or ‘cohort’ member) who had been injured, they were more likely, subsequently, to be cited for misconduct and/or the use of excessive force. This result held even when previous misconduct by the officer and being injured by civilians were controlled for. In other words, the vicarious experience of violence by an officer was associated with misconduct and excessive force even when controls related to the police officer were included. Furthermore, the “[r]esults show the effects of peer injuries on increased misconduct are most striking for injuries attributed to Black suspects” (p. 174). Most of the effect of injuries to partners is accounted for by those incidents with Black suspects.

Conclusion: These results suggest that to fully understand police misbehaviour and use of force with civilians, one has to look beyond the characteristics of the officer. The close contacts that the officer has with others – specifically those who have been injured in their work – appear to be important factors in understanding complaints against an officer for their use of force and other misconduct.

Reference: Zhao, Linda and Andrew V. Papachristos (2024). Threats to Blue Networks: The Effect of Partner Injuries on Police Misconduct. American Sociological Review, 89(1), 159-195.

Item 2

Black Americans’ views of the police are more nuanced than some might think. Although Black Americans are more likely than White Americans to believe that police racism is increasing, the vast majority of Black Americans do not want to reduce the amount of police patrolling in their own neighbourhoods, nor do the majority of Black Americans want spending on the police to be reduced.

Black Americans are considerably more likely than White Americans to report that they are afraid of the police. They are also more likely not to trust or have confidence in the police and to believe that significant changes in policing are needed. This paper examines a finding seemingly inconsistent with these views: that the vast majority of Black Americans do not want to reduce funding of local police nor do they want to reduce police presence in their communities.

One possible explanation for this set of findings is that although Black Americans are afraid of the police, they also are afraid of, and are more likely than Whites to suffer from, crime. This study examined views of policing of Black and Non-Black Americans who were asked to assume either that crime was increasing or that it was decreasing. They were also asked to assume either that the treatment of people in their neighbourhoods would improve or that it would not. As it turned out this “context” information had little, if any, impact on people’s views of police. 

Most White (93%) and Black (85%) respondents to a national survey wanted police to spend at least as much time patrolling their neighbourhoods as they currently do. Likewise, most respondents (White: 92% and Black: 84%) wanted police spending in their neighbourhoods to be the same or higher.  Those who have high levels of fear of crime (and relatively low levels of fear of the police) are most supportive of police patrols. “When it comes to police patrol, all Black respondents do not speak with one voice” (p. 5).

One possible explanation for the high levels of Black support of current levels of police budgets and police patrols of their neighbourhoods is that Black Americans are more likely than Whites to experience violent victimization: Black Americans were the racial/ethnic group that was most likely to witness gun violence and to experience a shooting near their residence. [These findings are similar to data for Black Canadians who constitute 4.3% of Canada’s total population but, in 2021, constituted 15.2% of homicide victims. Black Canadians are also twice as likely as non-Indigenous, non-racialized Canadians to report that they have little or no confidence in the police.] “Black Americans may prefer to maintain police patrol and spending even when crime declines and even absent significant policing reform because they may assume that there will come a day when criminal offending rates go back up, especially if policing is rolled back. Indeed, evidence exists that Black Americans view policing as a cost-effective way to control crime” (p. 9). At the same time, the data suggest that “policies that reduce Americans’ fear of the police and improve police-civilian relations would bolster civilians’ support for local policing” (p. 9).

Conclusion: The results of this study suggest that although the Black American community has serious concerns about the actions of their local police, they also have serious concerns about the nature of crime in their communities. “Most Americans, irrespective of their race, prefer to maintain, or even increase, local police presence and spending. Among Black Americans, this preference for maintaining or increasing police presence and spending is robust; it is not significantly affected by information about crime trends or policing reforms… The implication is that Black Americans, just like non-Black Americans, are supportive of policing – they want police departments to be well-funded and officers to be out on patrol” (p. 9). Support for the police also would increase if Black Americans’ fear of being mistreated were reduced.

Reference: Balcarová, Linda, J.T. Pickett, A. Graham, S.P. Roche and F.T. Cullen (2024). On the Robustness of Black Americans’ Support for the Police: Evidence from a National Experiment. Journal of Criminal Justice, 92. [Canadian data added by the Criminological Highlights group.]

Item 3

Prisons that are designed such that prisoners can experience higher concentrations of greenspace (vegetated landcover) within the walls of the prison are associated with lower rates of prisoner self-harm, less violence targeting other prisoners or prison staff, and higher rates of prisoner well-being.

Most studies of the conduct and well-being of prisoners focus on the characteristics of the prisoners, programs offered in prisons, the deprivations that prisoners experience, as well as the management styles and daily routines of the prisons. This paper goes beyond these “standard” explanations for behaviour in prisons and looks at the association between the presence of greenspace – ordinary green landcover – within the prison and prisoner well-being, controlling for the more ‘standard’ explanations.

In this study, the percentage of greenspace within prison walls was estimated using data from maps and aerial photographs of prisons in England & Wales. Data were available that allowed the researchers to estimate the proportion of greenspace within the prison walls for almost all prisons (89 to 96 prisons depending on the analysis) in England & Wales. A range of dependent variables was used, including prison records of self-harm and violence involving prisoners as well as some prisoner self-report measures of their well-being. Characteristics of the prisons and prisoners were used as control factors. These included such factors as the age, ethnic diversity of prisoners, the percent of the prisoners who were British, the average sentence length, the security level of the prisons, whether the prison was built in the 19^th century, the rate of overcrowding, and the size of the prison. 

The prison was the unit of analysis. In other words, the study looked at the effects of the characteristics of each prison on the behaviour and reports of prisoners – averaged across all prisoners in each institution. The results were consistent across measures. Each of the dependent variables – the rate of occurrences of prisoner self-harm, prisoner-on-prisoner violence and violence toward prison staff – was negatively correlated with the amount of vegetated landcover within the prison walls even when other factors were controlled statistically. Greenspace “exercises a significant and dampening effect on both self-harm and violence… supporting the notion that greenspace is important for well-being in the prison system of England & Wales” (p. 310). 

In addition to these official measures of well-being, self-reported measures were examined: the percentage of prisoners in each prison who reported mental health and/or emotional problems, a more general measure including self-reported drug habits or misuse of medication in the prison as well as a measure of the proportion of prisoners reporting that they do not feel respected or safe. All measures showed the same general effect: the amount of greenspace in the prison predicted favourable outcomes. There was some evidence that “lower rates of self-harm and violence are more strongly associated with larger amounts of greenspace in prisons with a high percentage of unsentenced prisoners” (p. 314). In addition, the effects appear stronger with young prisoners and in prisons with a high level of overcrowding.

Conclusion: The presence of a higher proportion of greenspace within the walls of prisons “is significantly associated with [lower rates of] self-harm and prison violence even when [controls are included] for the effects of both the characteristics of prison and their populations” (p. 317). These findings suggest that broad characteristics of prisons that are not obviously linked to prisoner characteristics may be important in promoting the mental health and well-being of prisoners who are likely, within a relatively short period of time, to be released back into the community.

Reference: Moran, Dominique, Jacob A. Jordaan, and Phil I. Jones (2024). Greenspace in Prison Improves Well-being Irrespective of Prison/Prisoner Characteristics, with Particular Beneficial Effects for Younger and Unsentenced Prisoners, and in Overcrowded Prisons. European Journal of Criminology, 21(2), 301-325.

Item 4

A bail reform initiative in New York State that resulted in an increase in the number of accused people released into the community without special “bail” conditions had no measurable impact on crime rates in the state.

It often happens that “liberal” pretrial release laws or practices are blamed for crime. When a person awaiting trial in the community is charged with a new offence, or when a high-profile accused is released, there are often cries that pretrial release is a major cause of crime that could easily be avoided if the person had been detained. This paper looks at the impact of a liberalization of New York State pretrial release laws on crime.

The problem in assessing the impact of pretrial release policies on crime is that even if those released have a very low likelihood of committing an offence while on pretrial release, increasing the number of people being released will, by definition, increase – in the short run – the likelihood of one or more of them committing an offence. This focus on the pretrial detention period (or the period during which a person might have been detained) ignores a critically important fact. Compared to release in the community, being detained in custody has been shown to increase, in the long term, the likelihood of a person committing an offence (see Criminological Highlights 17(3)#1, 17(2)#7, 17(5)#3) after they are finally released. Similarly, placing unnecessary conditions on those on pretrial release can have negative long-term impacts (21(3)#7).

In 2019, New York State bail reform legislation modified the law by requiring courts to release accused people on their own recognizance or under non-monetary conditions unless they were charged with certain offences. This list of offences was modified, to some extent, in the years following.

To assess the impact of the change in the law on crime, crime rates for seven serious offences (including 4 violent and 3 property offences) were assessed monthly. Two separate approaches were used to determine the impact of the legal change on crime: (a) an interrupted time series which examined whether there was a measurable discontinuity in crime rates associated with the change in the law for each of these seven offences, and (b) a “synthetic control” approach where a weighted measure of crime rates from other states (in which the law did not change) was compared to New York’s crime rates. The synthetic control was created by giving weights to the crime rates in other states so as to approximate the crime rates prior to the legal change.  That way, changes in crime rates in NY State occurring after the legal change could be assessed against this control.

The interrupted time series – which obviously had no “control” comparisons – showed a mixture of findings. When looking at the analyses with the synthetic control, the results showed some minor variability in crime rates across offences when NY State changed its laws. However, these differences were not significant. It is reasonable then to conclude that “the effect of bail reform on crime rate increase is negligible” (p. 393). Hence the findings are consistent with the conclusion that bail reform of this kind does not necessarily lead to more crime. The paper also underlines the difference between looking at reoffending rates by those affected by a legal change and looking at overall crime rates as they would affect society more generally. 

Conclusion: It is likely that crime by those on pretrial release in many jurisdictions accounts for only a small portion of total crime.  But studies that focus only on crime associated with those on release do not necessarily answer the question that people are interested in: Does a broad policy affect the overall rate of crime? The results of this study suggest that overall crime is not likely to increase measurably if the presumption of release is extended to those who have committed many different offences.

Reference: Wu, Sishi and David McDowall (2024). Does Bail Reform Increase Crime in New York State: Evidence from Interrupted Time-Series Analyses and Synthetic Control Methods. Justice Quarterly, 41(3), 371-399. 

Item 5

A program designed to eliminate the consequences of the revocation of probation for minor offenders by automatically substituting very short (e.g., 2 day) prison sentences as soon as possible after the first court hearing has been touted by its founders and others as an effective liberal reform. The only problem is that it doesn’t seem to be reliably better than the practice it replaced.

One problem with a probation sentence is that if a person does not abide by its conditions, they may be jailed for failure to comply with the order even if they would never have been jailed for the original offence. This paper examines an attempt to get around this problem by imposing very short jail sentences early in the process and then releasing the offender without conditions.

Early evidence from a highly cited but unpublished study suggested that making punishment for minor offences “swift, certain and fair” (with a short prison sentence after every violation) led to fewer arrests and fewer days of incarceration than probation as usual. The program was first publicized in 2004 and by 2016 had been implemented in 160 jurisdictions in 31 US states. It was often funded by the federal government and was seen as a way of reducing the overall use of imprisonment. This paper looks at the evidence available from 24 evaluations of these programs.

About a third of these evaluations used a randomized control design. In most of the studies, the sanction was imposed within 7 days but in most cases, the period in jail averaged more than 2 days. The target populations varied, but most involved adults who would otherwise have received probation. Some offenders in some studies (e.g., violent or sex offenders) were excluded from consideration for the program. The effects on recidivism were individually converted to correlation coefficients. The overall results suggested that the “swift, certain and fair” program reduced imprisonment by a tiny amount (r = -.058). Equally worrisome is the fact that there was no significant effect for the 5 studies that used the best designs – those with randomized control groups that received treatment as usual.

Equally important is the finding that most of the variation in the size of the effect on recidivism was due to factors other than whether the person was in the swift-certain-fair group or the treatment-as-usual group. Furthermore, “the extent to which programs adhered to the swift, certain, and/or fair principles had no impact on effect sizes” (p. 60).

This model of holding offenders accountable by way of swift, certain, and fair sanctions is obviously an attractive one. “When initially formulated, …and now, two decades later, this intervention continues to receive government funding and to be implemented in the United States and beyond. Despite the common-sense appeal of the program… the empirical evaluations of this model have been unfavourable” (p. 63).

Conclusion: The results of the 24 evaluations of this program show, at best, an average of a tiny impact on recidivism. But this average includes 5 studies in which an increase in offending and/or imprisonment was associated with assignment to the program. Indeed, even the studies – disproportionately those with less than ideal research designs – that suggest a reduction in recidivism demonstrate that the size of this effect is very small. Given the size of the overall effect and the variation across studies, it is clear that anyone contemplating implementing a program such as this one needs to ensure both that it is implemented adequately, and that the implementation is subjected to an adequate evaluation.

Reference: Pattavina, April, Joshua S. Long, Damon M. Petrich, James M. Byrne, Francis T. Cullen, and Faye S. Taxman (2024). Revisiting the Effectiveness of HOPE/swift-certain-fair supervision programs: A Meta-Analytic Review. Criminology & Public Policy, 23, 45-76.

Item 6

In general, Black Americans are more likely to experience incarceration and to have lower incomes than White Americans. In addition, the longer a Black person spends incarcerated in a given year, the lower their income the next year. This was not the case for Whites. These differential impacts of incarceration on Blacks and Whites contribute to the income gap between Blacks and Whites.

In the US, Whites working full time earn, on average, about $200 a week more than Blacks. This income gap could “influence incarceration by shaping a suspect’s ability to post bond and to hire a private defense attorney” (p. 840) which, in turn, could influence the disposition of subsequent contacts with the criminal justice system. Incarceration also influences the ability of a person to get employment, housing, etc. This paper examines the reciprocal effects of incarceration and income on Black and White Americans.

The study uses data from a longitudinal study of a representative sample of Americans born between 1980 and 1984 who were interviewed regularly until they were about 33 years old. As part of each interview, respondents were asked to report their total income and the number of months in which they experienced one or more days of incarceration. Blacks reported receiving considerably less income and were more likely to be incarcerated than Whites in each of the 14 interviews.

Looking across individuals, it appears that “time spent incarcerated had a stronger influence on future earnings for Blacks when compared to Whites” (p. 854). For Blacks, time spent incarcerated reduced income throughout the period of the study. “Formerly incarcerated Whites do not experience reductions in their income the following year when compared with Whites who were not incarcerated the prior year, suggesting that incarceration directly exacerbates the income gap between Blacks and Whites” (p. 854).

Conclusion: Time spent incarcerated seems to have a larger impact on earnings for Blacks than for Whites. “For every month a Black individual [experienced incarceration during a given year] they appear to earn between $300 and $600 less than non-incarcerated Blacks, an effect that was not observed for Whites” (p. 855). “These findings suggest that racial disparities in incarceration contribute to racial disparities in income in the US, the effects of which likely contribute to racial disparities in other life-course outcomes that have yet to be examined” (p. 855).

Reference: Silver, Ian A., Christopher D’Amato, and John Wooldredge (2024). Exacerbating Inequality over the Life-Course: Examining Race Differences in the Reciprocal Effects between Incarceration and Income. Social Forces, 102, 839-860.

Item 7

Black women achieve higher levels of education than do Black men. About a third of this difference is caused by different levels of involvement in school disciplinary practices and the criminal justice system.

The gender gap in educational attainment is larger in magnitude for Black men and women than it is for people of other races in the US. Black women have earned college degrees at rates higher than Black men since at least the 1940s. Given that they grow up, on average, in similar households and neighbourhoods, it is important to know why this is. 

This study looks at data from a large longitudinal survey of 1600 Black children born between 1982 and 1993 in the US. They were at least 4 years old when the first data were collected and at least 25 when educational achievement was assessed. Girls were more likely to receive a high school degree than boys (85% vs.71%). Boys were more likely than girls to have been convicted of a non-traffic criminal offence (24% vs. 7%) and boys were almost twice as likely as girls to have been suspended or expelled from school at some point in their lives (46% vs. 24%). Girls were more likely than boys to have received a 2- or 4-year college degree.

Not surprisingly, the gender gap in graduating from high school is explained in part by differences in achievement in school: girls do better than boys. However, the size of the effect of exclusionary school discipline on educational attainment is roughly the same, and criminal justice exposure accounts for much more of the difference between boys’ and girls’ likelihood of high school graduation. Controlling for other factors, punishment (by the school or criminal justice system) accounts for about a third of the difference between girls’ and boys’ high school achievement. For both boys and girls, being suspended/expelled or having criminal justice exposure reduced the likelihood of graduating from high school.  Girls, however, are much less likely than boys to receive these punishments. When looking at the likelihood of earning a 4-year college degree, “gender differences in criminal justice contact explain 49% of the… gender gap” (p. 943). What is especially important about these findings is that they apply even when behaviour problems (reported by the child’s mother on three separate occasions between when the child was 5-6 to 12-13) are controlled for. The difference in educational achievement appears to be substantially a result of society’s response to youths rather than their backgrounds or early behaviour. 

Conclusion: “Early observed differences between Black boys and girls have long-lasting implications for their educational pursuits…. Across several key moments in students’ educational trajectories, Black boys and men experience a disadvantage relative to Black girls and women due to comparatively higher levels of experience of exclusionary school discipline and contact with the criminal justice system” (p. 946).

Reference: Thompson, Marissa (2024). Examining the Black Gener Gap in Educational Attainment: The Role of Exclusionary School Discipline & Criminal Justice Contact. Social Forces, 102, 926-951.

Item 8

The effects of homicide go far beyond those immediately associated with the crime. This study demonstrates that young women who lived within about 400 metres of the location of a homicide had an increased likelihood of getting pregnant in the week following that homicide.

It is well established that homicides and other serious violent crimes affect children’s and adolescents’ sleep, attention span, impulse control, school performance and, perhaps, other factors. This paper goes beyond those findings and examines the impact of nearby homicides on young women’s risk of pregnancy.

This study relies on data from 375 women, 18-19 years old, who lived in Flint, Michigan between 2008 and 2012. They were interviewed weekly “about their pregnancy status and pregnancy-related behaviours and desires” (p. 865). For each woman, an assessment was made as to whether a homicide had taken place within about 400 metres of her residence in the previous 7 days. Flint’s homicide rate (per hundred thousand residents) in the middle of the data collection period (2010) was 59.5. This is very high compared to the US homicide rate that year (4.8) or Canada’s (1.6). 23% of the women were exposed to a homicide at least once during the course of the study.   



Each week, it was determined whether the respondent had any type of (male) partner and had sexual intercourse during that week. Women were also asked questions about the specific form of contraception they used. For simplicity’s sake these were divided into categories: “long acting reversible” contraceptive devices such as IUDs that did not require frequent or daily attention, and “short-acting contraceptives” such as birth control pills that did.

23% of the women became pregnant during the study. Because interviews were carried out often (weekly) and respondents provided the expected birth date, the researchers were able to estimate when the respondent became pregnant. And because they knew the date of the homicides in Flint, they could determine if the pregnancy occurred within a week after a homicide took place within about 400 metres of where the respondent was living. 

Nearby homicides had no impact on the likelihood that a woman had sex that week though homicides were associated with the women reporting an increased desire to have sex. A nearby homicide was not associated with an increased desire on the part of the woman (or her partner) to get pregnant. Nevertheless, the women were less likely to use any form of contraception during the week after a homicide took place near to where they lived. The effect was largely due to a reduction in the regular use of birth control pills (the most commonly used method) at the time of a nearby homicide. Birth control methods that require user effort infrequently were not significantly affected by nearby homicides.

When asked about the fact that they didn’t use contraception at the time of the nearby homicide, the women reported that it was not because they forgot to use it, or couldn’t access it, or their partner did not want them to use it. The women’s explanation was most likely something like they “just didn’t” [use it] or “just because” or “neither of us bothered” or “we didn’t care” or there was “no reason” (p. 891).

Conclusion: Clearly, homicides that occur close to where a person – in this case, a young woman – lives can have life-long consequences for people who have no direct relationship to those involved in the homicides. “The only explanation [the authors] found consistent support for [that might explain the reduced use of contraception] is a cognitive pathway…. that nearby homicides… reduce [a woman’s] contraceptive vigilance” (p. 886). The results “exemplify how [the effect of homicides] can extend from one generation to the next” (p. 888) and can affect, in profound ways, how homicides (which typically occur in socially disadvantaged neighbourhoods) can have unpredictable effects on certain groups in society.

Reference: Weitzman, Abigail, Jennifer Barber, Jusin Heinze, Yasamin Kusunoki, and Marc Zimmerman (2023). Exposure to Nearby Homicides and Young Women’s Reproductive Lives during Transition to Adulthood. American Journal of Sociology, 129(3), 856-906.

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