View this issue as a PDF: CrimHighlightsV17N1.pdf
This issue of Criminological Highlights addresses the following questions:
- Do non-profit community organizations help reduce crime in our neighbourhoods?
- How can ‘work release’ programs improve the safety of our communities?
- Does immigration make our neighbourhoods safer?
- Does reporting intimate partner violence to the police reduce reoffending?
- How does the use of ‘big data’ change the nature of policing?
- How is society harmed by the imprisonment of those under age 25?
- Should prison life be made harsher so that prisoners will be deterred?
- What kinds of people think that the criminal justice system should be made more punitive?
One way to reduce crime is to increase the number of local non-profit organizations that focus on improving life in urban communities.
Communities often look to criminal justice organizations – most notably the police and courts – to reduce crime. The effect that the justice system organizations have on crime is, however, quite limited. This paper investigates the effects other organizations – local non-profits – have on crime.
Beginning in the early 1990s, a rather dramatic decrease in crime occurred in the US. Also during this time, there was growth in the number of community non-profit organizations in many US cities. One possibility is that non-profit organizations that “influence the level of social cohesion within a neighbourhood and the degree to which communities are able to solve problems” (p. 1215) may be responsible for at least some of the reduction in crime that occurred. The challenge, then, is to see whether the growth in the number of these kinds of non-profit organizations contributed to the reduction in crime or whether the relationship between these two factors (growth in the number of non-profit organizations and reduction in crime) might be due to some other factor (e.g., broad economic prosperity).
Community oriented non-profit organizations often devote themselves to community issues broader than just crime reduction. Thus they might focus on general neighbourhood development, substance abuse, workforce development, or youth.
The specific methodological challenge is that the growth of nonprofits in a community may well relate directly to crime problems: a city may invest in nonprofits because of a crime problem. Hence it is important to look at changes that occur within cities rather than across cities. It turns out that the growth in community nonprofits is correlated with growth in nonprofits that are not community oriented (i.e., those focusing on the arts, medical, etc.). However, growth in the density of these arts, etc., nonprofits is not related to crime reduction.
This paper looks at the effects of changes in the density of community non-profit organizations in the 300 largest cities in the US on murder, violence and property crime rates, controlling for other factors (e.g., poverty, employment, race, etc.). The findings suggest that the addition of community oriented nonprofits to a city resulted in reduction in all three measures of crime. It would appear that nonprofits that focused largely on substance abuse programs and workforce development had the largest impact on crime. However, the other types of community focused nonprofits also appeared to have a crime reduction impact. Given that there were more organizations aimed largely at neighbourhood development and youth, the impact of these other organizations was important. Various other analyses were carried out in an attempt to strengthen the argument that the effects are causal, not just correlational.
Conclusion: The findings supported the inference that investment in community oriented non-profits was effective. This is not to say, of course, that any non-profit that purports to focus on communities is crime reducing. Rather, it suggests that organizations whose primary goal is not necessarily crime reduction (e.g., those that focus on substance abuse and workforce development) can also have crime reducing impacts. Obviously other factors contribute to the reduction of crime rates. The findings here do not challenge these other successful approaches. However, they do suggest that since many theories of crime relate to community factors, it is important, if one is interested in reducing crime, to consider investing in organizations whose purpose is to improve life in communities.
Reference: Sharkey, Patrick, Gerard Torrats-Espinosa, and Delaram Takyar (2017). Community and the Crime Decline: The Causal Effect of Local Nonprofits on Violent Crime, American Sociological Review, 82(6), 1214-1240.
A broad integrative work-release program for prisoners which combines work release and other integrative therapeutic programs can reduce reoffending.
It is often assumed that ‘getting a job’ after prison will help prisoners stop offending. But the relationship between work and reduced offending is complex (see Criminological Highlights 16(6)#3). The employment programs related to prison that are effective in reducing reoffending appear to be those that involve more than just getting prisoners jobs (Criminological Highlights 15(3)#3).
This study examines work-release programs in Israeli prisons. In addition to allowing certain low risk prisoners with 6-42 months left on their sentences to leave prison during the day to work at a paying job, the program is “integrated into a broader rehabilitative framework” such that it “serves as a bridge between life in prison and life in the community.” It does this by providing programs designed to help prisoners “develop personal and social skills that are believed… to enhance reintegration into the community following release” (p. 245). Prisoners are housed in an ‘open door’ part of the prison and are allowed to travel unsupervised to and from work.
The program, lasting 6-12 months, involves weekly group counselling sessions. Prisoners also typically receive cognitive-behavioural group therapy. In addition, they are allowed three days a month of unsupervised absence from the prison. The jobs they work at are typically in production or manufacturing plants.
About 30% of prisoners are dismissed from the program before completion. However, since the study was designed to evaluate the program per se those prisoners who failed the program are considered to be part of the treatment group even though they did not complete the program.
To create a comparison group, the study used “propensity score matching” – a technique designed to identify those who look just like those who received the treatment but did not receive it. The matching procedure took into account various socio-demographic factors, the prisoners’ offences and incarceration histories, and various details of their experiences in prison.
Re-arrest and re-incarceration rates for the treatment and control groups were examined during a 5-year follow-up period. At all points during the follow-up period, those in the control (non-treated) group were more likely to be re-incarcerated. After 5 years, 22.5% of the work-release participants were re-incarcerated compared to 33.1% of the untreated control group. Re-arrest rates showed a similar pattern: 32% of the work-release sample was rearrested compared to 46% of the control group.
Conclusion: Clearly the work-release program reduced the rate of re-offending leading to re-incarceration. The size of the favourable effect of the program is unusually high which may reflect “the broad integrative approach of the work release program in Israel” (p. 256). “Work release programs… must deal with the broader environment of prisoners, and the more general problems and difficulties that prisoners face with reintegrating into society” (p. 257).
Reference: Weisburd, David, Badi Hasini, Efrat Shoham, Gali Aviv, and Noam Haviv (2017). Reinforcing the Impacts of Work Release on Prisoner Recidivism: The Importance of Integrative Interventions. Journal of Experimental Criminology, 13, 241-264.
When the proportion of immigrants in neighbourhoods of two Australian cities increased between 2001 and 2011, violent crime in those neighbourhoods tended to decline.
Recent research on the relationship between immigration and crime, carried out largely in the US (see Criminological Highlights 5(4)#6, 8(6)#3, 10(6)#7, 11(1)#4, 11(2)#1, 13(6)#7, 16(1)#2) suggests that increases in the concentrations of immigrants in cities does not result in increases in crime. Indeed, increases in the number of immigrants may be responsible for some of the ‘crime drop’ that has occurred in recent years.
This paper looks, across time, at the impact of increases in concentrations of immigrants on violent crime in 882 neighbourhoods in two Australian cities (Brisbane and Sydney). Sydney has served as a “gateway city” for immigrants for many decades. Brisbane, on the other hand has, since 2001, seen substantial growth in its immigrant population – both in terms of numbers of immigrants and the diversity of country of origin. Neighbourhoods in these two cities were the unit of analysis. The study looks at changes within neighbourhoods in the concentration of immigrants and the relationship between these changes and violent crime rates. The main independent variable of interest in the study is the percent of foreign born neighbourhood residents. Data for 2001, 2006, and 2011 were obtained for 580 Sydney and 302 Brisbane neighbourhoods.
In both cities, looking at changes across time, neighbourhoods in which the number of immigrants grew tended to have decreases in violence. However, these effects did not seem to be as large the effects of some of the relatively stable characteristics known to relate to crime rates in neighbourhoods (e.g., economic disadvantage).
White English-speaking immigrants (largely those born in the UK or New Zealand) comprise the largest group of foreign-born immigrants in Australia. To determine if other immigrant groups who may face more economic and other challenges in Australia show different patterns, the effects of changes in the concentrations of the three other large groups of immigrants (Indian, Chinese, and Vietnamese) were examined. The effects were a bit inconsistent across groups and cities. In a number of instances increases in the percent of one of these three groups was associated with significant decreases in violent crime (e.g., as the percent Chinese in Sydney neighbourhoods increased, violent crime went down). However, in no case were increases in the concentrations of any of these three groups in either city associated with significant increases in violent crime.
Reference: Increases in the concentration of immigrant groups in both Brisbane and Sydney were generally associated with decreases in violent crime. There was no support for the hypothesis that increasing concentrations of immigrants increase crime. Unfortunately, these findings, like much of the research literature on ‘immigration and crime’ do not “directly test the effect of immigrant concentration on the neighbourhood processes important for the regulation of crime” (p. 706). “While the general public may fear that increased immigration will increase crime rates…, this study suggests that such concerns are largely unjustified, at least in the Australian context” (p. 708). These Australian findings are, however, quite consistent with US findings.
Reference: Sydes, Michelle (2017). Revitalized or Disorganized? Unpacking the Immigration-Crime Link in a Multiethnic Setting. Journal of Research in Crime and Delinquency, 54(5), 680-714.
Victims of intimate partner violence are less likely to suffer from repeat victimization if the incident was reported to the police and the victim received help or advice from a victims’ services agency. The arrest of the perpetrator, however, had no protective impact for the victim.
Intimate partner violence (IPV) accounts for a substantial portion of violent victimizations of women and some, but a much smaller portion, of violent victimizations of men. There is little agreement, however, about how society can best protect those who have experienced IPV from repeat victimization.
The goal of this paper, using longitudinal survey data, was to examine the impact of three separate actions on repeat victimization: (1) reporting an IPV incident to the police; (2) the police decision to arrest the offender; and (3) victim contact with a social agency. The study used the large US national victimization survey in which survey respondents were interviewed every 6 months for three years. Of the 2,221 IPV victims, 83% of whom were women, about 20% experienced repeat victimization within the period of the follow-up. The analyses that were carried out used ‘propensity score matching’ – a technique designed to equate groups that either received or did not receive each of the three ‘treatments’ (reporting to police, arrest of the offender by the police, and contact with a victims’ services organization). Separate matching was carried out to ensure that those being compared on the impact of each of these three ‘treatments’ were equivalent to those who did not receive the ‘treatment.’
In 26% of the cases of IPV, the police were contacted (by the victim or someone else). In the matched sample (in which the two groups of victims – those in contact with the police and those not in contact – were matched), those whose crimes were reported to the police had a lower risk of victimization throughout the follow-up period (of up to 40 months). For example, 6 months after the victimization, 16% of those whose victimization had been reported to the police reported a repeat victimization. 27% of those whose victimizations had not been reported to the police experienced a repeat victimization.
However, the police decision – to arrest the accused person – was unrelated to repeat victimization.
In addition, however, “the cumulative probability of repeat victimization was significantly lower for victims who received help from victim agencies than for those who did not” (p. 35). Looking at equivalent groups of victims who did or did not access victim services, the study shows that the probability of repeat victimization within 6 months was 15% for those who accessed victims’ services and 21% for those who did not.
Conclusion: “Arrest does not produce the desired effect of keeping victims safe from abuse in this… diverse, nationally representative sample” (p. 358) of victims of intimate partner violence. Additional analyses failed to discover any sub-sample of victims for whom arrest reduced re-victimization. On the other hand, reporting the incident to the police and receiving victims’ services (perhaps at the suggestion of the police) did reduce re-victimization. Unfortunately, it cannot be determined why reporting the incident to the police is effective, nor can it be determined what kinds of victims’ services are most effective.
Reference: Xie, Min and James P. Lynch (2017). The Effects of Arrest, Reporting to the Police, and Victim Services on Intimate Partner Violence. Journal of Research in Crime and Delinquency, 54(3), 338-378.
Big data surveillance is changing the nature of big city policing.
The use of ‘big data’ in many parts of modern society has changed the day to day operations of many professions. This paper suggests that in some, but not all, ways, the use of big data has changed the nature of policing.
Big data analytics has shifted policing practices to varying degrees: (1) Individual criminal risk assessment is quantified, (2) Data are being used for more predictive rather than reactive or explanatory purposes, (3) Dramatically more people are being made the target of surveillance, (4) People who have no direct contact with the police are now routinely included in the data used by police; (5) The police are using data collected for non-police purposes in order to try to identify possible offenders. It is suggested that these changes “have implications for inequality, law, and organizational practice in a range of institutional domains” (p. 978). In some cases big data analytics “is associated with mere amplifications in prior surveillance activities, but in others it is associated with fundamental transformations in surveillance activities and daily operations” (p. 985). This study relies on interviews and observations of officers in the Los Angeles (California) Police Department as well as other related organizations.
One of the most straightforward uses of ‘big data’ is the “quantification of civilians according to risk” (p. 936) on the basis of contact with the police, thus creating a list of people who are classified as ‘high risk’ on the basis of information that, at best, will be an imperfect predictor of what they might do in the future. Lists of people calculated to be ‘high risk’ were given to Los Angeles police officers. Clearly this is simply a more sophisticated version of the traditional focus of patrol officers on those they thought to be likely to offend. A more important change that has occurred is the use of data to predict where offences might take place – a move from purely reactive policing to a predictive approach.
One of the fundamental transformations of policing is the use of non-police data to identify links with other people, cars, phone numbers, and addresses. In combination, for example, with automatic license plate readers, the police link offences, people, and locations. However, the data used in these systems include such non-police indicators as foreclosures, collection agency data, and utility bills. Moves are being made to include, in a single data base that would be available to the police, an individual’s interaction with law enforcement as well as with agencies involved in social, health, mental health, and child and family services.
Although it could be argued that the use of big data might “replace unparticularized suspicion of racial minorities”, these approaches obviously “may be implicated in the reproduction of inequality… by deepening the surveillance of [certain] individuals…, widening the criminal justice dragnet unequally, and leading people to avoid surveilling institutions that are fundamental to social interaction” (p. 997).
Conclusions: Predictions and data analytics might appear at first blush to be impartial. But the data that are used determine, to some extent, the outcome. Once a person is ‘in’ the system and has ordinary links to those whom the police see as suspicious, that person is more likely – correctly or not – to be seen as suspicious. Unequal rates of database inclusion that will necessarily occur with the inclusion of external non-police (e.g., welfare) data, can have unanticipated negative impacts on those who are then subject to increased surveillance. “The burden of new surveillance practices is not borne equally, nor is the error they produce” (p. 999). Clearly, legal frameworks need to be developed to take account of the changes that are occurring.
Reference: Brayne, Sarah (2017). Big Data Surveillance: The Case of Policing. American Sociological Review, 82(5), 977-1008.
Offenders under age 25 sentenced to prison have a lower likelihood of completing secondary school than those sentenced to house arrest enforced with electronic monitoring.
There is a substantial amount of information suggesting not only that imprisonment does not reduce subsequent offending but that it also can have harmful effects in other domains of life (e.g., Criminological Highlights 16(4)#5, 14(6)#1, 11(4)#3). This paper examines the impact of imprisonment in comparison with electronically monitored house arrest on the completion rate of secondary school by young men in Denmark.
In 2006, Denmark amended its legislation to allow prison authorities to substitute periods of electronically monitored house arrest for short periods of imprisonment. The goals were simple: to maintain labour market participation and educational enrollment. From 2006 onwards, those sentenced to prison for up to 3 months are allowed to apply to serve their sentences at home, enforced with electronic monitoring. Assuming certain conditions (e.g., having a permanent address and consent from others living there), they are allowed to serve their sentences in the community. However, they must also attend a crime prevention program, allow unannounced visits from correctional workers, and agree to drug and alcohol testing. Denmark, like Canada, makes frequent use of relatively short sentences (In Denmark 61% of prison sentences were under 4 months; in Canada 77% of prison sentences were 3 months or less).
This study looked only at those offenders who were enrolled in an education program at the time of conviction and were sentenced to prison for 3 months or less between 2006 and 2009. The house arrest sample consisted of 443 offenders who met the criteria for release (even though only 63% actually received this treatment). The data were analyzed conservatively: the house arrest-electronic monitor group consisted of all those eligible for this program whether or not they applied for it. The ‘prison only’ group consisted of those convicted prior to the implementation of the reform who, therefore, served their sentences in prison. The two groups were matched on over 20 variables (e.g., criminal history, offence type, educational attainment, etc.). All offenders were followed for 3 years after their release from prison or the electronic monitoring program.
There were no short term (3 months to 1 year) differences between the two groups. However, two years after the end of their sentences (prison or electronically monitored house arrest), those given the opportunity to serve their sentences in the community were less likely to have dropped out of school and more likely to have completed their post-secondary education.
Conclusion: Participation in an electronically monitored house arrest program, as compared to normal imprisonment, has long term beneficial effects on secondary school completion rates. The fact that these effects showed up in the long term (2 to 3 years after the completion of the sentence) suggests that the overall program, part of which required attending school, was more effective than standard imprisonment even though the duration of the formal program was, at most, 3 months. It may have been effective, in large part, because those who participated in the electronically monitored house arrest program did not have their lives (and education) disrupted by imprisonment.
Reference: Larsen, Britt Østergaard (2017). Educational Outcomes After Serving with Electronic Monitoring: Results from a Natural Experiment. Journal of Quantitative Criminology, 33, 157-178.
Are those who don’t find prison to be very aversive less likely to be deterred by a prison sentence? Not if one controls for their offence and criminal record.
It is sometimes suggested that if prison life were made more unpleasant, those subjected to it would be less likely to commit offences that would put them in jeopardy of returning to prison. This paper investigates the relationship between subjectively experienced severity of imprisonment and recidivism.
There is evidence that compared to those given non-custodial sentences, those sentenced to prison are not less likely to reoffend (see Criminological Highlights “The Effects of Imprisonment”). In addition, those receiving longer sentences do not appear to be more likely to be deterred than those receiving short sentences. What has not been investigated is whether those who experience imprisonment as particularly harsh are more likely to be deterred than those who find imprisonment to be not so aversive.
This study examines the initial experiences of male prisoners in a pretrial facility in the Netherlands who subsequently were sentenced to prison. Data were collected three weeks after the prisoners entered the facility. They were asked to rate the severity of their experience, how much it felt like a punishment, and the degree to which they saw detention as harder than they had thought it would be. Because prisoners varied considerably in their backgrounds, a number of control variables were included. These included the sentence they received, the offence type, criminal history including whether they had previously been imprisoned, as well as a number of personal characteristics (e.g., age, ethnicity, education, employment, family situation).
Without taking into account the control variables, the analysis supported the hypothesis that subjectively experienced severity acted as a deterrent: those who indicated that they found imprisonment to be more harsh were less likely to reoffend within 6 months. The actual length of imprisonment also had an independent impact on recidivism. However, when offence type as well as other characteristics of the offender (previous convictions, etc .) were controlled for, the effect of the subjectively experienced severity of imprisonment disappeared.
It turns out that those who report that imprisonment is particularly difficult tend to be those with no convictions in the past 5 years, and those who, prior to being incarcerated, were employed and who had partners. In addition, those who were convicted of sex, weapons, or public order offences were particularly likely to see imprisonment as harsh.
Conclusion: The study demonstrates how easy it would be to find support for the view that “harsh prison conditions deter.” If standard predictors of reoffending (e.g., criminal record) are not taken into account, one could easily – but erroneously – conclude that subjectively experienced harsh prison regimes keep people from reoffending immediately after they are released. But when standard predictors of reoffending are taken into account, the subject experience of imprisonment is not related to reoffending rates.
Reference: Raaijmakers, Ellen A. C., Thomas A. Loughran, Jan. W. de Keijser, Paul Nieuwbeerta, and Anja J. E. Dirkzwager (2017). Exploring the Relationship between Subjectively Experienced Severity of Imprisonment and Recidivism: A Neglected Element in Testing Deterrence Theory. Journal of Research in Crime and Delinquency, 54(1), 3-28.
Americans who believe that offenders should be given harsh sentences are also likely to resent special favourable treatment of Black Americans and government attempts to help the poor.
Previous research has suggested that punitive criminal justice views relate to at least three separate concerns: that social solutions to crime do not work, the sense that society is experiencing a moral decline and those who engage in crime threaten the moral order, and that Blacks, who are perceived to be disproportionately responsible for crime, are inappropriately getting special favourable treatment in other areas of life.
This paper examines these explanations as well as one additional explanation: “that American punitiveness is rooted in the neoliberal ‘war against the poor’” (p. 937). The paper uses data from the U.S. General Social Surveys in 2000 and 2014 and focuses on two measures of punitiveness: support for the death penalty and whether respondents think that the courts are too lenient in the sentences they hand down.
The study measured “concern about crime” by asking people whether sufficient resources were being allocated to law enforcement and the problems of crime. “Social anxiety” was assessed with three questions related to whether people, in general, can be trusted, whether people generally try to be helpful, and whether people try to be fair. “Anti-Black racial attitudes” were assessed with questions on matters such as whether Blacks should be expected to succeed in society without special programs. “Animus toward the poor” was assessed with questions related to income disparity and whether the government has a responsibility to help improve the standard of living of the poor. Various other factors (e.g., age, race, political orientation, religiosity) were controlled for.
In both 2000 and 2014, support for the death penalty was related to social anxiety (the view that most people cannot be trusted), racial resentment (that Black Americans are inappropriately receiving special treatment) and negative views of the poor (that the poor are responsible for their own position in life).
When looking at support for punitive sentencing, both racial resentment and negative views of the poor predicted support for harsher sentences in 2000 and in 2014. The other predictors were not consistent over time.
Conclusion: “Both racial resentment and animus toward the poor are powerful predictors of punitive views, despite recent decreases in racial intolerance and the overall prevalence of punitive views. These results… point to the continued salience of race and poverty in conceptions of the threatening ‘other’.” (p. 956-7). These results are also consistent with other research “suggesting that anti-Latino sentiment [in the US] is linked with support for aggressive policing” (p. 957). More generally, the results suggest that “the social sources of punitive views have not shifted fundamentally” (p. 957) in the US between 2000 and 2014.
Reference: Brown, Elizabeth K., and Kelly M. Socia (2017). Twenty-first Century Punitiveness: Social Sources of Punitive American Views Reconsidered. Journal of Quantitative Criminology, 33, 935-959.
- Regular Issue