Criminological Highlights Vol. 15, No. 6 – June 2016

View this issue as a PDF: PDF iconCrimHighlightsV15N6.pdf

This issue of Criminological Highlights addresses the following questions:

  1. Why should suspensions and expulsions from school be used sparingly?
  2. Would building a wall on the Mexico-US border reduce the number of illegal Mexican workers in the US?
  3. Can street crime be substantially reduced by implementing a policy that encourages ‘stop, question, and frisks’ of ordinary citizens?
  4. Why is it in the public interest to help former young offenders get jobs?
  5. Is good housing policy for the mentally ill also good crime policy for everyone?
  6. If your local police service wants funds for a crisis intervention team to deal with its interactions with the mentally ill, what would be an evidence-based response?
  7. Do young people understand police warnings?
  8. Do body-worn cameras and other forms of surveillance of police reduce the use of force by police?

Suspending or expelling children from school increases their subsequent criminal justice involvement.

Previous research has suggested that children who miss school because of a school suspension or expulsion do less well academically than comparable non-punished peers, are more likely to commit offences, and are less likely to complete high school (Criminological Highlights 8(5)#4, 15(1)#1, 12(6)#1). 

In addition, in the US, data suggest that “non-Whites are much more likely than Whites to be removed from school for less serious rule violations, including tardiness, disrespectfulness or classroom disruption….  In contrast, White children are more likely to seek and receive therapy or medication than Black and Hispanic children” (p. 117).  This study looks at the relationship between early school punishment and involvement in the criminal justice system, as well as the relationship between early therapy/medication and criminal justice involvement, during late adolescence and young adulthood.

As part of the US National Longitudinal Study of Youth, mothers of young boys were asked if their children had ever been suspended or expelled from school between ages 6 and 14. Similar questions were asked about involvement in therapy or use of medication for troubles in school.  These children had been followed from birth onwards; hence it was possible to get measures of behaviour in childhood and performance in school before any suspension or expulsion might have taken place. 

Various differences among the children prior to the age at which they might have been suspended/expelled were controlled for.  These included school performance measures (e.g., test scores, repeating a grade) as well as race/ethnicity and poverty. In addition, measures often shown to relate to involvement in crime for young adults were controlled, including attitudes toward taking risks, whether the youth had graduated from high school, income, and depression.  

Above and beyond the effects of these childhood and young adulthood variables, being suspended or expelled from school increased the likelihood of involvement in the criminal justice system (conviction for charges other than minor traffic violations).  However, being suspended/expelled was unrelated to subsequent involvement in the mental health system. On the other hand, those youths who received therapy or medication for school-related problems were more likely subsequently to have some involvement with the mental health system, but were not more likely to be involved with the criminal justice system. 

The effects of school punishment are equally strong for Blacks, Hispanics and Whites. However, the relationship between early therapy or medication and subsequent involvement with the mental health system is stronger for Whites than it is for other groups. 

Conclusion: “By creating the impression that a boy is a ‘troublemaker’ and cutting him off from important educational resources, including classroom time and extracurricular opportunities, school punishment can contribute to a process of criminalization that significantly increases the odds of involvement in the criminal justice system” (p. 135).  Receiving treatment for school problems, however, is unrelated to subsequent criminal justice involvement.  “This process of criminalization, but not medicalization, contributes to what many legal scholars call the ‘school to prison pipeline’ in which children with behaviour problems are essentially tracked out of school and into the criminal justice system instead of alternative social control institutions like the mental health system” (p. 135). 

Reference: Ramey, David M. (2016). The Influence of Early School Punishment and Therapy/Medication on Social Control Experiences During Young Adulthood. Criminology, 54(1), 113-141.

Back to top

In order for a country’s border enforcement efforts to be effective in limiting the number of illegal migrant workers, it is first necessary to understand something about the flow of these workers into and out of the country.

It is estimated that between 1986 and 2008 the number of undocumented residents of the US increased from about 3 million to about 12 million, “despite a fivefold increase in the number of US Border Patrol officers, a fourfold increase in hours spent patrolling the border, and a twenty fold increase in nominal funding” (p. 1557-8).   

This paper suggests that “the unprecedented militarization of the Mexico-US border not only failed in its attempt to reduce undocumented population growth but backfired by increasing the rate of undocumented population growth into the country and turning what had been a circular flow of male workers going into 3 states into a settled population of families living in 50 states” (p. 1558).  This study is based in part on detailed histories of border crossings collected, since 1982, from household surveys in 143 selected communities in Mexico as well as interviews of those who settled in the US but originated in these same communities. 

In 1965, the US began placing limits on Mexican immigration and, over the next 15 years, increasingly limited the number of temporary workers in the US.  In the late 1950s, it is estimated that approximately 450,000 Mexicans entered the US as temporary (largely agricultural) workers. By the late 1970s, this program had been completely eliminated.  Migration, however, did not stop; “it simply continued under undocumented auspices” (p. 1559).  However, during this period the migration was “overwhelmingly circular” (p. 1560).  What had changed is that the workers were no longer legal.  Public sentiment was aroused; in 1985 President Ronald Reagan declared the undocumented workers “a threat to national security” (p. 1561).  

The militarization of the border, created with an approximately 10-fold increase in border control funding between 1980 and 2010, changed the locations and, most importantly, the risks of crossing from Mexico to the US.  Crossings became both more expensive and more hazardous. The likelihood of apprehension (and deportation back to Mexico) increased, though this could be reduced if potential migrants spent additional money on (illegal) help in crossing the border.  Many, however, simply tried again.  It is estimated that of those who tried to enter before 2008, 95% eventually succeeded.   Circular migration – entering the US temporarily to work and then return to Mexico – became less attractive, given the costs and risks of entry to the US.  

Not surprisingly, Mexicans were most likely to enter the US illegally when there was demand for their labour, Mexican wages were low, and when they had connections with those already in the US.  However, the age of those entering the US illegally increased between 1970 and 2010.  Few returned to Mexico, given the costs and risks involved.  In 1980, it is estimated that approximately 45% returned to Mexico within a year. By 2010, it is estimated that this had dropped to close to zero.  

Conclusion: The massive increase in US Border Control expenditures after 1986 “had no effect on the likelihood of taking a first undocumented trip northward or the odds of gaining entry to the US on such a trip, but it did have a strong effect in decreasing the likelihood of returning to Mexico once an entry had been achieved” (p. 1588).   “The combination of increasingly costly and risky trips and the near certainty of getting into the US created a decision-making context in which it still made economic sense to migrate but not to return home to face the high costs and risks of subsequent entry attempts” (p. 1590).  Dramatically increased policing of the border, then, had a net effect of increasing the number of undocumented workers in the US.  

Reference: Massey, Douglas S., Jorge Durand, and Karen A. Pren (2016).  Why Border Enforcement Backfired. American Journal of Sociology, 121(5), 1557-1600. 

Back to top

Police interactions with ordinary citizens involving ‘stop, question, and frisks’ appear to have very little effect on crime. 

High rates of policing of locations known to be high in crime (crime ‘hot spots’) appear to have a modest impact on crime in that location, but the impact is apparently temporary and may only be effective if high density policing is coupled with certain types of police activity (Criminological Highlights 12(3)#3, 13(3)#2, 14(5)#3).  

This study looks at the impact of “stop, question, and frisks” (SQFs) of ordinary citizens in New York City (NYC).  SQFs often take place in locations identified by the police as ‘hot spots.’ This and other forms of active policing are sometimes seen as the cause of the drop in homicides in NYC.   It is estimated that in 2003, there were approximately 160,000 SQF stops and 597 homicides. The number of SQF stops increased such that by 2011, there were 685,000 and ‘only’ 515 homicides.  After the courts ruled them unconstitutional in 2013, the number of such stops dropped to about 47,000 in 2014 and homicides also declined to 333. This study attempts to see whether SQFs – carried out at high rates before they were found to be unconstitutional - actually have an impact on crime.

The method was to look at crime on ‘street segments’ – essentially a city block, including the two intersections – during the years (2006-11) when an unconstitutionally high number of SQFs were carried out.  The goal was to see if an SQF carried out on a given street segment in a given week had any effect on crime on that same street segment the following week.  Looking across all 5 NYC boroughs, the data suggest that an SQF one week reduced non-traffic related crime on that street segment in 3 of the 5 boroughs. 

However, “in the peak years of SQFs in NYC, almost 700,000 SQFs would lead to only a 2% decline in crime” (p. 47).  Attributing the decline in crime solely to the SQF is problematic in that it is impossible to separate out the effect of the SQF on crime from the mere presence of police. In addition, attributing this modest drop in crime to the SQF ignores the “degree that SQFs are coupled with other policing strategies” (p. 49).  Specifically, “[i]n light of research findings on the effectiveness of directed patrol, the prolonged presence of police in a crime hot spot might very well be the active ingredient of SQFs, as opposed to anything that the police were doing” (p. 61).  As one commentator noted “the efficacy of the SQF tactic, at least from the standpoint of marginal deterrence, is considerably more ambiguous than its advocates might like to admit” (p. 62). Finally, even if there is a small effect, it is impossible to know whether this effect relates only to only certain types of SQFs (e.g., those involving actual offenders).  

Conclusion: Although the data suggest that stop, question and frisks (SQFs) may be associated with small reductions in crime in the location in which the police stop took place, one has to consider the other effects of SQFs: “The aggressive use of SQFs could erode citizens’ willingness to report crime to, or to cooperate in investigation or intelligence gathering with, the police. In a recent survey…  young respondents who were stopped more frequently reported less willingness to report crimes even when they were the crime victims” (p. 63).  Even if it could be shown that the apparent effects of SQFs on crime are due to SQFs and not some other correlated factor, “[t]he question is whether this approach [SQFs] is the best one for crime prevention at hot spots and whether its benefits are greater than its potential negative impacts on citizen evaluations of police legitimacy” (p. 50).

References:  Weisburd, David, Alese Wooditch, Sarit Weisburd, and Sue-Ming Yang (2015). Do Stop, Question, and Frisk Practices Deter Crime?  Criminology & Public Policy,  15(1), 31-56.  Apel, Robert (2015). On the Deterrent Effect of Stop, Question and Frisk. Criminology & Public Policy, 15(1), 57-66.

Back to top

For young people who were incarcerated as youths, gaining employment as adults reduces involvement in crime. 

The transition from adolescence into the adult workforce is particularly difficult for those with criminal records. This study attempts to understand the apparent reduction in crime that often takes place for those who do manage to find employment.  Simply put, is it the financial support that is critical or is the reduction in crime due to other aspects of the employment relationship?

The study examines the lives of 270 boys and 270 girls who experienced an average of about 21 months in a juvenile justice institution in the Netherlands.  Their employment and criminal justice involvement were then followed until age 32.  The goal was to understand the impact of employment and other forms of income support on offending. Under the Dutch system, income support could be in the form of payments while the person was unemployed, public assistance, or payments because of a disability.  Hence, for 14 years (from age 18 to age 32), the effect of employment and income support on offending could be estimated for each individual.  In addition, controlling for various other personal and demographic factors, the effect of employment differences on involvement in crime across individuals could be estimated. 

During this 14 year period, 75% of the men and 40% of the women were convicted of at least one offence. Most (about 85%) were employed at least once, though typically for about a quarter of each year.  For two types of analysis (the effect of different levels of employment across individuals and the effect of changes in employment on offending within individual), being employed was associated with decreased offending.  This was equally true for both men and women.  For men – but not for women – receiving public assistance appeared to reduce offending.  When property offending was examined alone, it appears, once again, that the impact of employment was consistent across statistical models and gender.  Income support, once again, only seems to reduce property offending for men. In addition, when looking across individuals, it appeared that the length of employment for men was associated with reduced violent offending.

Employment, in general, was more strongly associated with reduction in offending than was income support suggesting that in addition to “merely receiving an income, the nonmonetary aspect of work, such as social control or reduced opportunity to engage in crime because of changes in daily routine activities, are important in reducing criminal behaviour” (p. 563).

Conclusion: The findings suggest that “informal social control that [is a normal part of] employment is important in explaining the reduced offending rates for both men and women, even though [Dutch] women are more likely to work part-time and might attach less value to work than men” (p. 563).  More generally, however, findings such as these suggest that programs for those returning to the community from prison may have an importance and impact that is larger than one would expect simply from the effects of providing additional support.

Reference: Verbruggen, Janna, Robert Apel, Victor R. Van Der Geest, and Arjan A. J. Blokland (2015). Work, Income Support, and Crime in the Dutch Welfare State: A Longitudinal Study Following Vulnerable Youth into Adulthood. Criminology, 52(4), 545-570.

Back to top

Being randomly assigned to receive housing and support services from Housing First – a Vancouver (Canada) program that provides housing and mental health services to people who are homeless and who suffer a mental disorder – reduced subsequent criminal convictions. 

The combination of being homeless and mentally ill is associated with relatively high rates of offending.  Not surprisingly, therefore, prisons tend to have a disproportionate number of people suffering from a mental illness.  This study addresses the problem of the mentally ill in the criminal justice system by attempting to deal with, simultaneously, their housing and mental health needs.

Adults with a diagnosed mental disorder who were homeless or precariously housed (i.e., they lived in transitional housing after at least two periods of homelessness in the previous year) and who had at least one prior contact with the British Columbia (Canada) justice system were randomly assigned to one of three types of treatment: 
(a) Scattered site (SS) treatment in which individuals were offered choice of housing (typically among 2-3 available units) in buildings scattered across the city of Vancouver. Approximately 20% of the units were occupied by the previously homeless.  The study participants assigned to this treatment type were provided with access to psychiatric and primary healthcare services by a treatment team, 24-hours a day, 7 days a week. 
(b) Congregate housing (CONG) in which individuals were placed in a former hotel in which all residents were study participants. All residents had access to psychiatric and primary health services at any time. In addition, there were attempts to promote a community among residents through various programs and activities.  
(c) A treatment-as-usual (TAU) group of individuals who received the normal services available to those experiencing homelessness and mental illness. 

Conviction data – drawn from the Provincial Courts in British Columbia – were obtained for all participants for a period that varied between 9 and 24 months.  Those assigned to the “Scattered Site” treatment were significantly less likely to be convicted of any subsequent offences than the treatment-as-usual group.  The Congregate group was marginally less likely to be convicted than the treatment-as-usual group. 

Obviously, it is not possible from this study to disentangle the separate and combined effects of the housing and mental health services.  It is also not clear why there was apparently a difference between the Congregate and the Scattered Site conditions. 

Conclusion: The results of this random assignment experiment “demonstrate that Housing First [housing plus mental health support] produces significant reductions in reconvictions compared to usual care.  People who are both homeless and mentally ill are frequently in contact with the justice system, a process that is both destabilizing to the individual and costly to society… Housing First has been shown to improve housing stability and health service involvement.  [These results show] improvements in public safety and reductions in crime” (p. 6).

Reference: Somers, Julian M., Stefanie N. Rezansoff, Akm Moninzzaman, Anita Palepu, and Michelle Patterson (2013).  Housing First Reduces Re-offending among Formerly Homeless Adults with Mental Disorders: Results of a Randomized Controlled Trial. PLOS|One

Back to top

We do not have evidence that crisis intervention teams – designed to help police deal more effectively with the mentally ill –  are effective in reducing the use of criminal charges and in minimizing the use of force by police.

With the reduced use of secure mental health facilities to house the mentally ill, the police in many jurisdictions are called upon to respond to various forms of disruption and crime involving people apparently suffering from mental illness. Crisis Intervention Teams (CITs) are designed to help the police link the mentally ill with community programs, thus providing an alternative to criminal justice processing. This paper examines the currently existing data on the effectiveness of CITs in achieving this goal.

One model that has been used (the “Memphis Model” of CITs) has been to provide training to police officers about mental illness and in handling mental illness crises. The focus is on de-escalation techniques, providing resources within the community for the mentally ill, and on ensuring the police know about these resources so that mentally ill people can be treated in the community rather than being arrested.

Various outcome measures are relevant in determining whether CITs ‘work.’ These include the possible reduction in charges, decreased use of force, increased officer safety, and reduced time spent by police officers processing those with mental illnesses who come in contact with the police.  This review examines the existing research on arrests, use of force, and police officer safety. In order for a study to be considered as providing evidence on one or more of these factors, it was necessary that there be a comparable control group where the CIT was not available. 

Perhaps the most discouraging finding is that out of 820 papers reporting ‘evidence’ about CITs, only 8 studies were carried out that both had an appropriate control group and had measures of one or more of these outcomes. Seven were US studies, one was Australian.  In understanding the ‘outcome’ findings, one needs to remember that there is no ‘standard’ definition of a CIT.  Hence differences across studies may well reflect differences in procedures used by the CIT or differences in other aspects of the program and its clients.  

There was no overall impact of the availability of CITs on arrests, though the results of the five studies were “significantly heterogeneous” (p. 85) suggesting that the different patterns may reflect more than just random variation.  Various different measures (self-reports, official records) were used in different studies on officer use-of-force. None of the five studies showed favourable impacts on use-of-force and the combined effect of the use of a CIT across the eight studies, once again, was not significant.  Injury to police officers, as it turns out, was rarely reported and could not be assessed adequately. Data are not available on other possible benefits such as the reduction of stigma and time spent in processing cases. 

Conclusion: In this review of the “highest quality studies available of the effectiveness of CIT (Crisis Intervention Team) models” (p. 90), there was no convincing evidence that they reduce arrests or officer use of force.  One possible explanation for the ‘null’ effects is that, in recent years, even officers working in environments without formal CIT programs may have become more willing to look for alternatives to arrest.  But the simple conclusion that the CIT model is “effective” and should be implemented more broadly cannot be sustained on the basis of available research.  On the other hand, police services should be encouraged to experiment with – and evaluate – what the most effective approach is in responding to calls involving the mentally ill in their communities.

Reference: Taheri, Sema A. (2016). Do Crisis Intervention Teams Reduce Arrests and Improve Officer Safety? A Systematic Review and Meta-Analysis.  Criminal Justice Policy Review, 27(1), 76-96.

Back to top

It is unlikely that warnings from Canadian police given to youths are adequately understood by them.  

As with adults, Canadian police must inform accused youths of their legal rights. For example, youths have a right to legal counsel; they can consult parents or other adults, and youths can choose to have parents and/or counsel present during questioning.  They are not required to make statements to the police. 

The Youth Criminal Justice Act specifies that the warnings to the youth of the youth’s rights must be “in language appropriate to his or her age and understanding” (s. 146(2)(b)). However, the form of the various warnings is not specified in the legislation. Hence police services have developed special formal warnings for youths that can be read to youths when they are arrested.  This study examines a simple question: Is it likely that youths understand these warnings?

There is a good deal of research (e.g., Criminological Highlights 11(3)#7) that suggests many police warnings designed for ordinary citizens are often written in ways that make them difficult to understand.   In this study, 50 warning statements from the Royal Canadian Mounted Police and from at least one police service in each province and territory were studied. There were 31 “unique” warnings. (The remainder were used by more than one police service). The police services that provided the warnings for the study are responsible for policing at least 90% of Canada’s population. 

Two very different forms of assessment of these 31 different warning forms were employed. First, the ‘complexity’ of the forms was assessed.  For example, long statements are generally more difficult to understand. The warnings varied in length from 239 to 1192 words. The statements were also assessed according to standardized measures of the school grade level needed to comprehend the statements as well as the grade level at which specific important words would be likely to be understood.  The estimated grade level needed to understand the warnings ranged from approximately Grade 6 (age 12) to Grade 12 (age 18). Many of these warning statements contained words that would not be expected to be understood by those well over the age of criminal responsibility in Canada (age 12).

In a separate study, 32 students, aged 15-17, in St. Johns, Newfoundland & Labrador were read the forms used by their local police service. They were asked to indicate whether they understood what was read to them and their confidence in their comprehension of the warning. They were then asked to write down the information that they would give to a close friend who was arrested.  Broadly speaking, the students were, not surprisingly,  able to recall more of the warnings when tested after each section was read to them than when they were asked to recall what they had been told after all information had been delivered. However, only 22% of the respondents recalled more than half of the information contained in the warning.  

Conclusion:  Perhaps because warnings are often long and written in language that is difficult for youths to understand, it is it not surprising that young people do not fully understand the warnings that are normally used by police.  “Also of importance was the fact that participants [who were read the warnings used by their local police] reported high levels of confidence in how much they understood and almost always confirmed that they understood the rights that were presented – despite the overall low level of comprehension.  This finding suggests that simply asking youths whether they understood the rights is not a useful procedure for ensuring that youths actually understand their rights” (p. 821). But, in addition, other research would suggest that even if they understand the ‘words’, youths may not be able to resist the pressures to make statements to the police (Criminological Highlights 13(4)#2).

Reference: Eastwood, Joseph, Brent Snook, and Kirk Luther (2015).  Measuring the Reading Complexity and Oral Comprehension of Canadian Youth Waiver Forms. Crime & Delinquency, 61(6), 798-828.

Back to top

 Police officers are less likely to use force against civilians when they believe that their actions are being recorded.

has made it possible, at least in theory, to monitor more closely how officers respond to difficult interactions with civilians.  Results from three recent papers suggest that police officers, when they believe their actions are being recorded, are less likely to use force with civilians.

Body-worn cameras (BWCs) are the most recent piece of technology whose effects on police behaviour have been examined.  The challenge, of course, is not just to monitor the behaviour of police officers when they are wearing BWCs, but to find an appropriate comparison to know whether any change is really caused by wearing the BWC.  Said differently, control groups are necessary. 

In a study carried out in Orlando, Florida, police officers volunteered to be part of a study of the effects of BWCs.  Eighty-nine officers volunteered for the study and were randomly assigned either to use the BWC during all shifts for the next 12 months or to carry out their usual duties without the BWC.  The main measures used were reports of incidents where force was used with a civilian. Citizen complaints were also monitored.  The results show that the two groups were similar on these two measures before the 12-month test period, but that during the 12 months of the study, the officers with the BWCs were less likely to have used force with a civilian at least one time and were less likely to have been the subject of a civilian complaint.

In another study in a small California city, BWCs were employed on a random basis by all front line police officers on about half of 988 police shifts during a one year period.  In other words, on a random basis, all police officers during a shift would be wearing BWCs or none would.   The ‘shift’, then, was the unit of analysis and the research question was whether there was a change in the number of use-of-force incidents during BWC-shifts as compared to control shifts.  During this period, there were more incidents of police use-of-force during control shifts than during the BWC-shifts. There were very few complaints about police behaviour in either type of shift.  More interesting, perhaps, is the fact that the overall number of recorded uses of force in the police service dropped dramatically during the study (from about 65 per year to 25).  The number of complaints also dropped – from an average of 40 to 3.  Since these decreases are too large to have been the result of having cameras during only some of the shifts,  it may be that simply carrying out the experiment and the discussion and training that it entailed made officers more cautious about the use of force, whether or not they were wearing a camera at the time. Alternatively, this could be an effect of other events and concerns happening at the time.

In a third study, a former Canadian police officer interviewed 231 police officers in two cities about the impact of potentially having their interactions with civilians captured by video cameras on citizens’ smart phones. Most (94%) reported that they had, at some point, been aware of being recorded by a citizen.  Most (84%) reported high level of awareness, during their normal activities, of the possibility that their actions would be recorded.  Most importantly, most officers (74%) reported that they thought that their behaviour had changed because of the possibility that it would appear on social media. 

Conclusion: It would appear that police officers are, not surprisingly, affected by the possibility that their everyday behaviour and their use of force will become an ‘objective record’ for the public to assess.   The data suggest that body-worn cameras may reduce the use of force by police officers in their interactions with the public.

References: Jennings, Wesley G., Mathew D. Lynch, and Lorie A. Fridell (2015). Evaluating the Impact of Police Officer Body-Worn Cameras on Response-to-Resistance and serious complaints.  Journal of Criminal Justice, 43, 480-486. Ariel, Barak, William A. Farra, and Alex Sutherland (2015).  The Effect of Police Body-Worn Cameras on Use of Force and Citizens’ Complaints Against the Police. Journal of Quantitative Criminology, 31, 509-535. Brown, Gregory R. (2016). The Blue Line on Thin Ice: Use of Force Modifications in the Era of Cameraphones and Youtube. British Journal of Criminology, 56, 293-3.

Back to top

Publication Type

Volume Number


Issue Number