Criminological Highlights Vol. 16, No. 4 – April 2017

View this issue as a PDF: PDF iconCrimHighlightsV16N4.pdf

This issue of Criminological Highlights addresses the following questions:

  1. Why are girls in bail court perceived to be more likely to be in need of ‘treatment’ than boys?
  2. Does wearing a hijab or niqab make it more difficult to tell whether a woman is telling the truth?
  3. Can Australia reduce its imprisonment rate?
  4. How are the use of foster care for children and imprisonment for adults linked?
  5. How does an arrest during high school affect a youth’s educational prospects?
  6. Does employment reduce offending for ‘high risk’ youths?
  7. Is the disruptive impact of imprisonment on employment related to time spent in prison?
  8. Does the imprisonment of a child’s father have an impact on the likelihood that the child will commit an offence?

Girls in Canada’s youth courts are much more likely than boys to have broad treatment conditions imposed on them as a condition of pretrial release.

Although there has been a substantial decline in the rate that youths are brought to youth courts in Canada, the rate for one offence – failure to comply with a court order (typically the charge of failing to comply with a condition of pretrial release imposed in bail court) — has not shown a similar decline.  Currently these charges are the most serious charge in 17% of youth court cases.

To some extent, the cohrts themselves are responsible for this large number of ‘status offences’ in Canadian youth courts: previous research (Criminological Highlights 12(5)#3, 13(1)#1, 13(5)#5, 15(3)#1) has demonstrated that imposing large numbers of conditions – many of which have little relationship to the offence – on youths who are released, especially when combined with long waits until their cases are disposed of,  increases the likelihood of youths failing to comply with their bail conditions.

This paper looks carefully at the bail conditions imposed on boys and girls in one of Canada’s largest youth courts.  Although the Youth Criminal Justice Act states that youths cannot be detained in custody as a substitute for social welfare purposes (s. 29(1)), there is nothing in the Act that explicitly prohibits courts from imposing treatment or welfare-based bail release conditions.  Although higher courts have decided that it is not sufficient simply to impose conditions on a youth just because someone in the courtroom at the time the decision is made thinks it might be a good idea, it would appear that there are few restrictions on conditions imposed on youths.

This paper looks at whether or not treatment orders were included in the bail conditions imposed on a random sample of youths – 425 boys and 75 girls – who were released by a Toronto court between 2009 and 2013. Girls were significantly more likely, overall,  to have treatment orders imposed on them (70% of girls and 45% of boys received treatment conditions). Typically if a youth received a treatment order it was rather broadly defined.  For example, it might require a youth to be assessed by a doctor or to “follow a doctor’s orders” or to attend a program such as counselling, anger management, or substance abuse.  Usually the conditions included both attending a treatment program and “being amenable” to treatment.

For boys, treatment orders were significantly more likely to be imposed if the most serious offence involved violence, if the youth was facing more than one charge and if the youth had a previous charge.   Girls, for each category of each of these variables (e.g., cases involving violence and cases without violence), were more likely to receive treatment orders than boys.  More importantly, however, none of these three variables was significantly related to whether a girl received a treatment order.

Conclusion: Simply being a girl, it would seem, was sufficient for the court to require a treatment order in 70% of the cases, quite independent of the number and nature of current charges and whether or not there had been previous charges. Boys not only were less likely than girls to receive treatment orders, but the likelihood of receiving a treatment order for boys appeared to relate to features of the case.  Since youths when they appear in bail court have not been assessed to see if there is a need for treatment and they have obviously not been convicted of anything, it is curious that girls in bail court appear to the court almost automatically to ‘need’ treatment.  “Are we observing anachronistic vestiges of the view that females in conflict with the law must be either ‘mad’ or ‘bad’?” (p. 94)

Reference: Sprott, Jane B. and Allan Manson (2017) YCJA Bail Conditions: “Treating” Girls and Boys Differently.  Canadian Criminal Law Review, 22, 77-94.

Back to top

Ordinary people are better at determining whether female witnesses are telling the truth when the women are wearing a hijab or niqab than when they are wearing no form of face or head covering.

The wearing of a niqab (a veil that covers a wearer’s face except for her eyes) is contentious in many communities.   In court, such face coverings are seen as problematic because of the belief that the trier of fact – a judge or a jury – must see the witness’s face in order to determine whether she is telling the truth.

Determining whether such beliefs have merit starts with the simple fact that people, in general, are not very good at telling whether what a stranger says is true.  But in addition, “A meta-analysis of 50 studies revealed that overall lie detection accuracy was similar, whether observers received audio (i.e., more restricted) or audiovisual (i.e., less restricted) information” (p. 402).

In this paper, two studies, using the same basic design and materials, were carried out first with Canadian university students (Experiment 1), and then with university students from Canada, the Netherlands (using only those with relatively good English language skills) and the UK.   Female students observed a video of woman who had been asked to watch a stranger’s bag.  Half of the students saw a video in which the woman stole something from the bag; half saw a video in which the woman did not steal anything.  The students were then told to imagine that they were called as a witness for the accused – to state that they did not see her steal anything.  Hence half were asked to lie; half were told to tell the truth.   They were then randomly assigned to wear a black niqab, a black hijab or to remain unveiled.  A black shawl was used to cover their clothing.  A trained research assistant arranged the veils and shawls.   Each “witness” was then examined and cross examined.  From these videos, sets of 20 clips – 10 with the witness telling the truth and 10 with her lying (in random orders) – were created for each of the three veiling conditions (niqab, hijab, no veil).

Male and female students were then shown one set of video clips and were asked to indicate whether, in each clip, the witness was telling the truth.  Note that a response bias – toward saying that the witness was lying or telling the truth – could not increase or decrease accuracy.

The results from the two experiments were very similar across countries.  “Contrary to the assumptions underlying [some court decisions suggesting that being able to see the face is important for determining veracity], lie detection was not hampered by veiling.  In fact, observers were more accurate at detecting deception in witnesses who wore niqabs or hijabs than those who did not veil.  Discrimination between lie- and truth-tellers was no better than guessing in the latter group…. It was only when witnesses wore veils (i.e., hijabs or niqabs) that observers performed above chance levels” (p 407).

Conclusion:  In this study, then, wearing a veil actually improved lie detection.  It would appear likely that when the witness they were observing was wearing a veil – either a niqab or a hijab – those charged with the responsibility of determining if the witness was telling the truth focused on the actual statements coming from the person they were observing.  “Seeing a person’s entire face does not appear to be necessary for lie detection; banning the niqab because it interferes with one’s ability to determine whether the speaker is lying or telling the truth is not supported by scientific evidence” (p. 408).

Reference: Leach, Amy-May, Nawal Anmar, D. Nicole England, Laura M. Remigio, Bennett Kleinberg, and Bruno J. Verschuere. (2016). Less is More?  Detecting Lies in Veiled Witnesses. Law & Human Behaviour, 40(4), 401-410.

Back to top

Australia, like some other countries, has experienced a large increase in imprisonment rates in the past 30 years.  There appear to be significant costs of the increased rates of imprisonment and few benefits.  It may, however, be possible to reverse the trend.

Between 1918 and 1984, the (national) Australian imprisonment rate averaged about 60 per 100,000 population, varying between about 45 and 81.  After 1984, however, it increased almost every year to a high of about 142 per 100,000 in 2013.  The increase occurred in each state and in the Northern Territory, though there were differences across states (see Criminological Highlights, 15(4)#5).  The increase “has been particularly marked in the case of Indigenous Australians and women” (p. 138).

Prison was used more sparingly before 1984.  For example, in 1980, 29% of those convicted of break-and-enter and theft were imprisoned; in 2015, 50% received a sentence of imprisonment.  Prior to 1984, there was a presumption of bail for most offences.  There was, as well, in at least one state, a presumption in favour of release on parole.  But during the 1970s and 1980s, crime was increasing; and punitive policies quickly followed (e.g., mandatory minimums, increased maximum penalties and reduced opportunities for release).  But since 2000, crime in Australia has decreased.  Imprisonment rates – and policies supporting high rates – have not.   The bulk of the reduction in crime is exceedingly unlikely to have been caused by the increased imprisonment since the most optimistic ‘incapacitation’ estimates would suggest that imprisonment could not account for more than 10% of the reduction in crime that Australia experienced.

Data from Australia mirror what is known in Canada for youths (Criminological Highlights, 4(4)#3): “Courts are much more likely to send a person to prison if he or she has already been imprisoned, even after adjusting for other factors that influence the risk of a prison sentence” (p. 145).  Hence, to some extent, imprisonment leads to more imprisonment (at least among repeat offenders).

A number of suggestions have been made for addressing the challenge of “imprisonment growth in an era of decreasing crime”.  A focus on increased “offender rehabilitation” programs seems unlikely to affect anything since few prisoners, now, receive any program known to have favourable effects.  Reducing or abolishing the use of short sentences would reduce prison admissions dramatically, but would have less impact on prisoner counts.  Finding alternatives to imprisonment for non-criminal breaches of restrictions placed on those being punished in the community would be helpful, just as would finding alternatives to prosecution for minor offences.  Australia, like Canada and some other countries, also has a problem of a large pretrial detention population. Addressing this could have a dramatic impact on the size of the prison population.  Finally, it is suggested that “toning down the political rhetoric on law and order” might “offer Australians a more rational, more considered approach to law and order [that would allow the spending of monetary savings] on things they really want” (p. 148-9).

Conclusion:  The growth of imprisonment in Australia is not caused by crime, but rather by the beliefs and values of those making decisions about how to punish those who have offended.  Reducing the use of imprisonment to levels that served Australians well for decades, therefore, is not going to be the result of finding a single solution.  Rather it will probably involve multiple approaches and a change in the views about how best to respond to crime.

Reference: Weatherburn, Don (2016). “Rack ‘em, Pack ‘em and Stack ‘em”: Decarceration in an Age of Zero Tolerance.  Current Issues in Criminal Justice, 28(2), 137-156.

Back to top

Child protection intervention in the form of admissions to foster care is most likely to occur in US states with punitive social policy regimes such as high incarceration rates and a high proportion of cases resulting in death sentences.

Large numbers of children – 5.9% of all US children, 11.4% of African American children, and 15.4% of Indigenous children – enter the foster care system at some point in their lives. The variation across states is also dramatic – from a low of 1.42 to a high of 10.46 foster care entries per 1000 children in the population per year. This paper examines the variation, across states, in the use of foster care in relation to other policies, including criminal justice policies.

Entry into foster care can be seen as “a regrettable but necessary component of a broad and generous welfare state…” or as “a vehicle through which states monitor and punish parents and families who fail to conform to hegemonic parenting standards” (p. 578).  However, since poverty is “among the key factors [predicting] child maltreatment” it follows that entry into foster care should be expected to relate to poverty levels. The study looks at rates of entry into foster care as a function of poverty measures as well as various other social policies.  These include programs designed to address poverty as well as criminal justice policies.  Various other social policies are included in the predictive models as well. Estimates of these indicators from the 50 states for 10 years (2002 to 2011) were included as predictors of the use of foster care.

The findings suggest that the rate of entry into foster care (per 1000 child population) was lower in states with more generous welfare systems and in states with relatively lower crime rates.  However, even after crime rates were controlled for, adult incarceration rates and the number of death sentences per prison admission were positively associated with high rates of entry into foster care.   As an example of how entry into foster care is related to other social policies, it appears that, holding all other variables constant, states with the most generous welfare rates tended to have lower foster care entry rates.

In addition, however, entry into foster care was related to criminal justice policies.  “A state with a criminal justice regime that is less punitive than the average state… is expected to have 3.4 foster care entries per 1000 children… whereas states with broad and punitive justice regimes are expected to place 4.9 children per 1000 into foster care.”

Conclusion: “After controlling for relevant demographic, social, and political contexts, states with expansive and generous welfare regimes place fewer children into foster care than do states with expansive and punitive criminal justice system” (p. 576).  States with high crime and high imprisonment also tended to have high rates of entry into foster care. It would appear that “disruptive methods of child protection [e.g., placing a child in foster care] are institutionally aligned with punitive forms of social control” (p. 577). In addition, given the negative relationship between welfare generosity and the foster care institutionalization rate, it would seem that states are making an implicit choice in how they deal with children growing up in poverty: to invest in more generous welfare programs or to invest in programs that move children from their families into foster care.

Reference:  Edwards, Frank (2016). Saving Children, Controlling Families: Punishment, Redistribution, and Child Protection.  American Sociological Review, 81(3), 575-595.

Back to top

Youths who are arrested during high school are less likely ever  to enroll in a four-year postsecondary education program than equivalent youths who were not arrested.

There is substantial evidence that contact with the criminal justice system can have negative effects on youths (e.g., Criminological Highlights 14(6)#1, 11(4)#3).  This paper examines the long-term impact of arrest on an important aspect of youths’ lives: the likelihood that they will enroll in a 4-year post-secondary education program.

An arrest can have a direct impact on a youth’s chances of being accepted into a college or university program since many educational institutions require applicants to provide information about their discipline history and criminal record. Surveys of college admissions officers indicate that “two thirds would consider denying admission to a student who had been convicted of marijuana distribution and half would consider denying [admission] to a student who had ever been [simply] arrested for the same crime” (p. 641).  But in addition, there can be indirect effects: “the stigma of an arrest may trigger social exclusion at school” and “teachers or guidance counselors [may] perceive arrested students as poor investments” (p. 624) and not encourage them to do what is necessary to gain admission to a 4-year program.

The study used data from a large (US) longitudinal study. Youths, who had graduated from high school and who had been arrested during their first 3 years in high school, were matched with very similar students who had not been arrested.  A statistical technique (propensity score matching) was used to create equivalent groups of youth who had the same likelihood of being arrested (given their background characteristics on 59 variables such as involvement in delinquency, substance use, friends who engaged in anti-social behaviour). In this way 162 arrested youth were matched with one or more control youths (who had not been arrested).

31% of youths who had not been arrested in high school enrolled in a four-year postsecondary institution within 9 months of graduation, compared to only 21% of those who had been arrested.  There was, however, no (significant) difference between the groups in the enrollment rates in 2-year colleges.  The difference in enrollment rates in a 4-year program between those arrested and those not arrested persisted for 10 years after graduation.  At that point, 50% of those who had not been arrested, but only 41% of those who had been arrested ever enrolled in a 4-year post-secondary education program.

A substantial amount of the difference in the likelihood of enrollment in a 4-year postsecondary educational institution for those arrested and those not arrested appears to be due to two factors: grade point average and enrollment in advanced courses in the fourth year of high school.   Arrest did not significantly predict college entrance exam scores and suspensions from school did not predict enrollment in a 4-year postsecondary program, so neither factor could be said to account for the difference in 4-year postsecondary enrollment.

Conclusion: Youths who are arrested in high school had a substantially lower likelihood of ever attending a 4-year post-secondary educational institution.  Given that prior to the arrest, the youths were equivalent on many relevant dimensions including cognitive abilities and involvement in crime, it would appear that being arrested sets in motion a set of processes that has a permanent negative impact on youths’ educational attainment.

Reference: Widdowson, Alex O., Sonja E. Siennick, and Carter Hay (2016). The Implications of Arrest for College Enrollment: An Analysis of Long-Term Effects and Mediating Mechanisms.  Criminology54(4), 621-652.

Back to top

Regular – but not temporary – employment or receiving public assistance can reduce property offending by youths released from a treatment facility at age 18.

The effect of employment on crime is not straightforward. Although it might be logical to assume that crime – especially property crime – would be less frequent if a person was employed, this relationship is not as simple or universal as some might think (see Criminological Highlights, 4(3)#6, 6(4)#5, 8(6)#4, 10(2)#3, 13(3)#5, 14(2)#7, 14(5)#7, 15(3)#3, 15(5)#3).

The analysis in this study allows one to estimate not only the effect of employment on crime (property and violent), but also the effect of engaging in crime on subsequent employment.  In addition, it allows one to separate out the effect of third variables that might lead to both unemployment and crime, for example “individual preferences and abilities that select individuals into offending [and] … into unemployment” (p. 183). Such preferences and ability could suggest, spuriously, that there was a causal relationship between unemployment and crime.

In this study, 270 male youths who spent at least 2 months in a ‘judicial treatment institution’ in The Netherlands were followed from age 18 to 32.  92% had committed at least one offence prior to age 18.   Monthly data were collected from government files on temporary or continuing employment, as well as unemployment insurance, disability insurance and public assistance (a welfare payment providing the minimum income needed for subsistence).  Offending was measured as convictions for offences (identified by month of the offence) that took place during this 14 year period.  Controls were included for whether the person was incarcerated, married, or a parent.

As with other ‘high risk’ offenders, this group had a relatively low employment rate: on average only 31% were employed in any given month, with about 74% of those who had any employment having regular employment.   In the sample as a whole, property crime declined with age; violent crime did not.

Regular employment (where a person was on the payroll of an employer) significantly reduced a person’s subsequent property offending.  Temporary employment had no effect on a person’s subsequent property offending. Neither regular nor temporary employment had an effect on violent offending.

Whether or not a person received any type of social welfare (unemployment insurance, disability insurance, or public assistance) had no overall effect on subsequent property or violent offending.  However, receiving public assistance – subsistence payments requiring no proof other than need – did appear to reduce subsequent property offending.

Conclusion.  It is possible that the stability that regular employment demands in a person’s life explains why regular, but not temporary, employment appears to reduce property crime.  That apparent stability did not carry over to violent crime: employment of any kind had no impact on violent crime. Receiving public assistance – a form of social welfare provided to the most disadvantaged residents of The Netherlands where this study was carried out – did reduce property offending. However, the fact that no type of social welfare program or employment affected violent offending suggests that violent offending is largely a function of factors other than a person’s financial circumstances.

Reference: Mesters, G., V. van der Geest, C. Bijleveld (2016). Crime, Employment, and Social Welfare: An Individual–Level Study on Disadvantaged Males.  Journal of Quantitative Criminology, 32, 159-190.

Back to top

Being imprisoned for six to twelve months has more dramatic negative impacts on employment prospects than imprisonment for less than six months.

Those who have a criminal record have a difficult time getting a job (Criminological Highlights 4(3)#6).  But there are other reasons that those who spend time in prison may be very likely not to have work.  Those spending substantial periods of time in prison also run the risk of losing relevant work skills, becoming detached from the institution of work, and developing closer ties to those more likely to be involved in crime than with employment.

This study examines the relationship between the time that a sample of 702 men spent in prison and their likelihood of being employed after release.  Sentences of a year or less are common in many countries.  In Canada, for example, approximately 90% of those released after serving a sentence of imprisonment were in prison for a year or less.  This paper, therefore, looks at those released from Dutch prisons after serving a year or less to determine what impact it has on the their employment record.

The paper examines the employment records of those who were in prison for at least a week and up to a year.  “Time in prison” included both pretrial detention and sentences served immediately after conviction.  Five groups of prison lengths were created: 1-6 weeks, 3 weeks to 3 months, 3-4 months, 4-6 months, and 6-12 months.  Groups were equated by using a ‘propensity score’ technique such that pre-existing measures on each person were used to predict their likelihood of serving each of these sentence lengths.  In this way, the researchers were able to create groups of people who spent varying lengths of time in prison but who, generally, were very similar on relevant dimensions.  Employment was measured through an interview that took place approximately 6 months after release. Two different self-report techniques (which produced very similar results) were employed.

There were no consistent differences among the four groups that had spent less than 6 months in prison.  In the first month after release, about 33% of those who had been in prison for less than 6 months were employed.  This increased to about 40% employed 6 months after release.  However, the time in prison did not seem to matter for those who were in prison for 6 months or less.

The group that was different from the 4 groups who spent less than 6 months in prison were those who had been imprisoned for a relatively long period: 6-12 months.  In the first month after they were released, only 20% were employed.  Six months after release only about 27% were employed.  Their rate was obviously considerably lower than the employment rates for the groups imprisoned for less than 6 months.  It appears that those who spent 6-12 months in prison were both less likely than those who spent less time in prison to regain their pre-prison employment or to find new employment.

The differences in employment rates did not appear to be related to treatment programs in prison or recidivism rates. Those who spent very short periods of time in prison were less likely than others to participate in any rehabilitative prison programs.  In addition, recidivism rates did not differ significantly across the groups: about one quarter of the study participants reported committing at least one offence during the 6 month follow-up period.

Conclusion: It would appear that those who are imprisoned for 6 months or more are less likely to be employed after release than those who spend less time in prison.  Six months for these prisoners appeared to be an important threshold: for those who spent less time in prison, the ‘time in prison’ did not seem to matter.  But being in prison for 6 months or more did appear to reduce employment prospects.  Hence, “although employment is largely insensitive to imprisonment length among short term prisoners [i.e., those with less than 6 month prison stays], an apparent threshold exists at approximately 6 months.  Beyond 6 months, longer imprisonment corresponds with incremental deterioration in employment prospects” (p. 422).

Reference: Ramakers, Anke, Robert Apel, Paul Nieuwbeerta, Anja Dirkzwager, and Johan Van Wilsem (2014).  Imprisonment Length and Post-Prison Employment Prospects.  Criminology, 52(3), 399-427.

Back to top

Legal changes can reduce crime: A change in the law in Denmark that reduced dramatically the risk of incarceration for those convicted of certain crimes also reduced dramatically the risk that the sons of fathers convicted of these crimes would commit offences.

There is substantial evidence that there are negative impacts of incarceration on the families of those sent to prison (e.g., Criminological Highlights 14(2)#1, 12(6)#7, 12(5)#1, 13(1)#7, 12(6)#8, 9(5)#6, 13(3)#3).  This study examines a legislative change in Denmark in 2000 that dramatically expanded the use of probation with community service as a penalty for certain offences rather than imprisonment. Looking at those not sentenced to fines (which were not targeted by the legislation) the use of imprisonment for the remaining offenders (the “treatment group”) dropped dramatically as a result of the reform.

The study looks at the offending of children who were born in Denmark and were age 12-18 at the time of their fathers’ conviction. The study sample included fathers whose conviction took place in the year prior to the legislative change or the year after the reform that increased the use of probation/community service.  The dependent measure was whether children of these fathers who were convicted during this period (the year before and the year after the change in the law) were charged with a crime. A 10 year follow-up period was used (after the father’s conviction).  Various statistical controls were introduced (in three different ways) in order to eliminate effects that might be due to socioeconomic factors, family characteristics, and previous criminal justice involvement of the father.  Data were collected from centralized government registries.

About 87% of the children in the pre-reform ‘treatment’ group experienced the incarceration of their fathers, compared to only about 24% of the children in the post-reform treatment group.  Hence the children in the study had quite different experiences pre- and post- reform. Within 10 years of the conviction of the father, 64% of the male children whose fathers were convicted before the reform (where the father was likely to go to prison) had been charged with an offence.  For those boys whose fathers were convicted post-reform (where the father was unlikely to be imprisoned), only 57% were charged with an offence within 10 years.

In order to test whether other changes in Denmark might be responsible for any effects of the legislative change, the offending rates of children of fathers convicted of crimes (before and after the reform date) but whose offences were not covered by the legal reform were also examined.  The incarceration rate for fathers convicted of these other offences was relatively stable across the period being examined. Male offspring of these fathers, however, were charged at a slightly higher rate if the father was convicted after the reform date (a 60% offending rate) rather than before (57%).  Various statistical techniques were used to determine whether the effects of the reform for male offspring of fathers who were convicted were stable.  Looking across these analyses, it would appear that incarceration of a male child’s father increases somewhat the likelihood that the boy will commit an offence.

There were, however, no consistent effects for female offspring of incarcerated fathers. Girls were considerably less likely than boys to be charged with an offence, but the incarceration of the father appeared to have no significant effect.

Conclusion:  As with previous studies, it would appear that the incarceration of a parent can have harmful – in this case criminogenic – effects on children in the family.  In this case, boys – but not girls – were more likely to commit offences in the 10 years after a father’s incarceration.  When one considers, then, the intergenerational transmission of crime, it would appear that at least part of it has to do with criminal justice policies.

Reference: Wildeman, Christopher and Signe Hald Andersen (2017). Paternal incarceration and children’s risk of being charged by early adulthood: Evidence from a Danish policy shock. Criminology, 55(1), 32-58.

Back to top

Publication Type

Volume Number


Issue Number