Criminological Highlights Vol. 6, No. 5 – August 2004

View this issue as a PDF: PDF iconCrimHighlightsV6N5.pdf

This issue of Criminological Highlights addresses the following questions:

  1. Do restrictions on gun markets make a difference?
  2. Can social policies that focus on pregnant women and young children reduce crime?
  3. Does taking young offenders to court reduce subsequent offending?
  4. Why does Japan have low rates of violent crime?
  5. Can we predict rearrest of women on probation or parole?
  6. Are police officers who endorse traditional police culture more likely to use force with citizens?
  7. Does the type of job an adolescent has affect the youth’s involvement in crime?
  8. Why do some victims of wife assault want their assailant arrested while others do not?

The more firearms there are in a community, the more likely it is that high school students in that community will carry them.

Background. “When it comes to gun policy, one of the few uncontroversial assertions is that unsupervised adolescents should not carry them in public” (p. 27).  Two broad approaches have been used to limit the use of firearms by adolescents: reducing demand (the motivation to carry or use a firearm) and reducing supply (the availability of guns).   Some prominent scholars have suggested, however, that criminal gun use by teens is not affected by supply because those who want a gun can always find one. This “futility hypothesis” would suggest, then, that “restrictions on gun markets do not reduce the misuse of guns” (p. 28). Consistent with the “futility hypothesis” is the finding that many youths report that it is easy for them to find (illegal) guns. 

This study, using a national survey of adolescent males carried out in 1995, examines gun-carrying among 15-17 year olds, 10% of whom report having carried a gun at least once in the month prior to the survey.  Typically, teens who report having carried a gun indicated that they did so to protect themselves.  

Gun ownership rates vary enormously in the United States.  The estimated rates of ownership of guns in households in Louisiana, Alabama, and Mississippi (over 55%) are more than 4 times those of Hawaii and Massachusetts (about 12%).  In this study, local gun prevalence was estimated by using a measure that has been shown to have a very strong relationship to the rate of gun ownership in a community: the proportion of all suicides committed with guns.  The basic analysis, then, looks at the relationship between this measure of gun ownership and reported gun carrying by adolescents.   Various other factors were statistically held constant –   crime rate in the community, socio-economic status, age and race of the youth, whether the youth lived in an urban area, as well as measures of neighbourhood disadvantage.  The most notable finding is that “controlling for individual characteristics, the likelihood of gun carrying is positively related to gun prevalence in the county, and strongly so” (p. 40).   The findings suggest that “a 50% increase in county gun ownership is associated with approximately a doubling of the prevalence of teen gun carrying” (p. 41).  

It could be argued, however, that places where guns are prevalent are simply more dangerous, and, therefore, youths are more likely to arm themselves for self-defence purposes.  Two facts argue against this as an alternative explanation for these findings. First, the violent crime rate was controlled for (using robbery rate as a proxy for violent crime).  More persuasive was the finding that “Gun prevalence has little effect on the likelihood that the teen carries any type of weapon, either a gun, knife, or something else…. While [gun prevalence]  does not affect the likelihood that a teen carries a weapon, the availability of guns clearly increases the likelihood that those teens who do carry weapons choose guns” (p. 41). 

Conclusion:  It appears that the “futility hypothesis” is not confirmed: There is value in trying to reduce the availability of guns, since it appears to relate to the likelihood that teenagers will carry a gun.  “Where guns are prevalent, adolescents will find it easier to borrow or steal or buy them from family members or other people” (p. 49).  “Supply-side interventions [do not appear to be] inherently “futile” and should not be ruled out a priori by policymakers” (p. 50).

Reference:  Cook, Philip J. and Jens Ludwig (2004). Does Gun Prevalence Affect Teen Gun Carrying After All?  Criminology, 42 (1), 27-53.

Back to top

Policies relating to support for pregnant women and young children can reduce the likelihood of anti-social behaviour.

Background.  There are data suggesting that physical aggression in children peaks at age 2 and declines thereafter. “This is an indication that most humans learn to control the use of physically aggressive behaviour before they enter the school system… Those who have not learned to control their aggressive reactions by the time they enter the school system enter a vicious circle of negative interactions, where rejection from their peers, because of their aggressive behaviour, leads to more reasons for aggression” (p. 207).  Hence intervention (prevention) programs during early childhood would appear to stand a good chance of having a substantial impact on future aggression.

This paper reviews a number of different interventions in which delinquency or socially disruptive behaviour was typically the outcome variable.  Most of the studies involved randomized control design and had follow-up periods (after the intervention) of at least a year.  Some of the studies used delinquency as the outcome measure; others used measures that have been shown to be related to delinquency (e.g., the prevention of inadequate parenting). Some of the positive outcomes that have been found include the following:

  • The well known evaluation of the “High/Scope Perry Preschool Project” was based on the theory that preschool programs addressing educational needs would aid in the youth’s cognitive development   Though the intervention typically occurred only when children were 3-5 years old, by age 27 these children had experienced fewer adult arrests than the untreated controls.  
  • Similarly, in the “Syracuse University Family Development Research Program” help was given to parents during the first five years of the child’s life that focused in large part on the nature of the interaction between the parent and the child.  The intervention also included a component whereby the child attended an enriched day care during the first 5 years of his or her life. These youths, when 13-16 years old, were less likely than untreated children (6% vs. 22%, respectively) to have been involved with the justice system. 
  • A nurse home visit program during pregnancy and the child’s first two years of life reduced subsequent delinquency (see Criminological Highlights, 4(2)#1). 

In all, 28 studies are reviewed in this paper.  As a whole, the review demonstrates that “interventions with high-risk families can change the parenting behaviour which many theories identify as the first stop in a chain of events that can lead to anti-social behaviour” (p. 233). 

Conclusion. “The review of the 28 prevention experiments [suggests] that early childhood interventions can have a positive impact on the three most important risk factors for juvenile delinquency: disruptive behaviour, cognitive skills, and parenting.  Furthermore, experiments with long-term follow-ups which have targeted at least two of these risk factors in childhood have shown a significant impact on criminal behaviour. From these results, it can be concluded that early and intensive preventive interventions can have the desirable impact which appears to be so difficult to achieve with disruptive elementary school children and juvenile delinquents” (p. 237). 

Reference: Tremblay, Richard E. and Christa Japel. (2003) Prevention During Pregnancy, Infancy, and the Preschool Years.  In Farrington, David P. and Jeremy W. Coid. Early Prevention of Adult Antisocial Behaviour.  Cambridge University Press. 

Back to top

Contact with the formal juvenile justice system increases the level of criminal activity in early adulthood.

Background. The labelling perspective suggests that contact with the justice system increases the likelihood of further delinquency.  One explanation for this effect is that formal contact with the justice system has a “negative impact on conventional opportunities… and leads to cumulative disadvantage in future life chances…” (p. 1288).  

This study followed a sample of Rochester, New York, youths through adolescence until they were 21-22 years old. Youths were asked if they had been arrested or had other contact with the police and whether they had experienced other, more formal, juvenile justice intervention.  When the youths were young adults,  self-report offending records were obtained on seven relatively serious offences (e.g., robbery, attacks with a weapon, break and enter, car thefts) as well as the youth’s involvement in drug sales.  

The results demonstrated that a youth’s likelihood of graduating from high school was lowered as a result of police or juvenile justice involvement even after controlling statistically for previous offending, parental poverty, and school ability (at age 12). A separate analysis found that “experiencing official [criminal justice] intervention in adolescence is significantly associated with reduced odds in favour of staying in school in a subsequent period” (p. 1301).    An analysis of self-reported criminal activity at age 19-20 demonstrated that police or juvenile justice intervention earlier in adolescence was associated with increased criminal behaviour in early adulthood.  The effect of police or juvenile justice intervention  “has stronger crime amplification effects among the disadvantaged [African American youths living in poverty]” (p. 1306).  The effects of juvenile justice interventions on drug selling were quite similar: police or juvenile justice intervention increased the likelihood of drug selling at age 19-20.  And again, “the effect of juvenile justice intervention on drug selling is stronger among those from impoverished family backgrounds” (p. 1306). 

Looking at criminal involvement at age 21-22, it appears that an earlier intervention by the police increases crime rates generally through its effect of decreasing the likelihood of graduating from high school and increasing the likelihood of unemployment at age 19-21.  Juvenile justice intervention appears to have a direct effect in increasing drug selling and general crime at age 21-22, but also has an indirect effect by way of decreasing the likelihood of graduating from high school which, in turn, increases the likelihood of unemployment at age 19-21.

Conclusion.  It appears that police or juvenile justice intervention with young people has a reasonable likelihood of increasing the probability that the youth will, as a young adult, be involved in crime and/or drug selling.  Generally speaking, these negative impacts on youths are more likely for those who come from impoverished backgrounds or are black.  The argument, therefore, that it is important to apprehend and prosecute young people in order to hold them accountable for their actions should  be questioned.  These data suggest that for many youths – especially those from impoverished backgrounds – the best strategy may be to do as little as possible and wait for them to outgrow their criminal behaviour.

Reference:  Bernburg, Jön Gunnar and Marvin D. Krohn (2003).  Labelling, Life Chances, and Adult Crime: The Direct and Indirect Effects of Official Intervention in Adolescence on Crime in Early Adulthood.  Criminology, 41 (4), 1287-1318.

Back to top

Why does Japan have low rates of violent crime?  Variation in Japanese rates of violence in the second half of the 20th century is associated with factors such as economic conditions that are known, in western cultures, to relate to crime rates. Cultural characteristics unique to Japan that are related to informal social controls do not seem to be very important. 

Background.  Impressed with the obvious cultural differences between Japan and western countries, many commentators have suggested that Japan’s low and generally decreasing crime rate is due to such factors as the manner in which Japanese social organizations (e.g., family, workplace, schools) highly value interdependency and thus achieve informal social control.   The only problem with this explanation is that while crime rates dropped between 1950 and 2000, “family and community ties, the supposed root of informal social control in Japan, have weakened…” (p. 181).  Divorce rates, the number of women in the work force, and urbanization have all increased dramatically in Japan during this period.

This study examines Japanese homicide and robbery rates as a function of divorce, urbanization, females in the work force,  as well as traditional determinants of crime rates in the west: economic stress (income inequality, unemployment, and poverty),  the likelihood of apprehension for crime (as reflected in police clearance rate), and the proportion of young males in the population.  

The results suggest that violent crime in Japan is controlled in large part by economic factors: higher homicide and robbery rates during this period are associated with periods of higher income inequality, higher unemployment and poverty as well as with low apprehension rates and higher proportions of young males in the population.   One of the measures of informal social control (female labour force participation) was associated with the robbery rate, but not with the homicide rate.  A separate analysis, looking at individual areas of Japan across this same period had similar findings: changes in divorce rates, urbanization, or that proportion of women in the work force had no consistent impact on crime. 

Conclusion. “The results indicate that a comprehensive explanation for Japanese crime rates should include measures of economic stress, age structure and certainty of punishment.  The findings also suggest that social disorganization explanations, closely linked to the idea that Japan’s crime rates are due to its unique cultural characteristics, should not be overemphasized in accounts of Japanese crime trends” (p. 200-1).  These findings might be seen as suggesting that it is possible for other countries to follow the Japanese example and achieve relatively low rates of violence: “Other countries probably cannot reproduce Japan’s cultural milieu, but they may be able to create more favourable economic trends” (p. 201).  Nevertheless, for Japan, the economic changes that are taking place (e.g., the end of guaranteed lifetime employment) “threaten to substantially increase these [crime] rates” (p. 201). 

Reference: Roberts, Aki and Gary LaFree (2004).  Explaining Japan’s Postwar Violent Crime Trends.  Criminology, 42(1), 179-209.

Back to top

A study of women felony offenders in the U.S. shows that a standard indicator of risk – the Revised Level of Service Inventory (LSI-R) – does not add to one’s ability to predict rearrest or supervision violations of women on probation or parole once the financial status of the woman is taken into consideration. 

Background.  What factors predict a woman’s risk of recidivism?  A commonly used instrument, the LSI-R (Level of Service Inventory-Revised) has been used with some success in order to predict the recidivism of men.  It is said to be an effective measure of “comprehensive risk and criminogenic needs” (p. 189) and was developed “to assist correctional administrators in distributing scarce resources across offenders based on a combination of recidivism risk and need for services” (p. 189).  However, there has been some dispute about the measure’s usefulness with respect to women, based in part on ideas concerning the “gendered nature of offending [suggesting that ] female pathways to crime are different from those of their male counterparts” (p. 187).  In particular it is argued that “the influence of poverty on recidivism among women offenders” (p. 187) has not been given sufficient attention. 

This paper looks at the recidivism risk of 134 female felony offenders on probation or parole in Minnesota.  Two dependent measures were examined: rearrest for any offence and reports by the women as to whether they had violated conditions of probation or parole.  The LSI-R had been administered to the women.  In addition, the woman’s financial status was assessed. 

The results show quite clearly that the LSI-R is related to both rearrest and parole or probation violations.  In a similar fashion, whether or not the woman was living in the community in poverty also predicted rearrest and parole or probation violations.  What is most interesting, however, is that above and beyond the woman’s poverty status, the LSI-R added nothing to the overall predictability of these two measures of recidivism.

Looking only at those women whose income put them below the “poverty line,” it was found that two forms of state-supported help  – publicly subsidized housing and life skills programming  which included “assistance in areas such as identifying potential employers, filling out job applications and developing interview skills” (p. 200) – significantly reduced the likelihood of the woman being rearrested and/or violating probation or parole conditions.  Those women living in poverty without such assistance were three times more likely to recidivate than were those who did receive assistance (45% vs. 14%).  Looking only at those women living in poverty, receiving state support reduced recidivism, but the LSI-R score did not  help predict recidivism once the receipt of state aid was taken into account.

Conclusion. It would seem that the most commonly used measure of  “risk” of recidivism (LSI-R) does not help predict recidivism above and beyond a woman’s economic status.  More importantly, however, the study focuses attention on the importance of the economic well-being of women serving sentences in the community and the importance of  “state-sponsored assistance to meet critical short-term economic needs” (p.202) of these women.  Investments by the state in reducing poverty (directly or indirectly through employment assistance programs or housing) appear to be good investments in reducing recidivism as well as improving the lives of an impoverished group.

Reference. Holtfreter, Kristy, Michael D. Reisig, and Merry Morash (2004). Poverty, State Capital, and Recidivism Among Women Offenders. Criminology and Public Policy, 3 (2), 185-208. 

Back to top

Police officers who reject the traditional norms of the “police culture” are less likely than police who hold traditional police attitudes to use coercive behaviour in their interactions with citizens.

Background.  “Police culture” is usually defined in terms of a number of quite separate attitudinal clusters:  the view that citizens are not to be trusted; that citizens would not support them if help were needed; that police supervisors are more interested in procedural rules and regulations than in the officers’ well-being; and that procedural rules (e.g., relating to search and seizure) sometimes need to be broken.  

This study examines the relationship between the strength of a police officer’s adherence to traditional police culture and the officer’s use of coercive techniques in their interaction with citizens.   A total of 638 police officers in Indianapolis, Indiana and St. Petersburg, Florida were interviewed.  Each officer was also observed during 12 separate shifts in areas where police citizen interactions were thought to be most frequent.  The analysis of the officers’ style of interaction with citizens is based on encounters with 3223 citizens whom the police identified as suspects (wrongdoers, peace disturbers or people who had been the subject of a citizen complaint).   Police officers were classified as adhering to traditional police culture on the basis of an analysis of clusters of attitudes.  Generally speaking, for example, if they tended to distrust citizens, hold negative views of supervisors and police management, endorse the view that sometimes rules had to be broken and that patrols should be carried out aggressively, they would be classified as endorsing a traditional police culture. 

In the observed encounters between police and citizens, use of force was scaled from none, through verbal commands or threats, to physical constraints (e. g. pat downs and handcuffing) through “impact methods (pain compliance techniques…)” (p. 1019).   Generally speaking, pro-police culture officers used each type of force more often than did those officers who did not endorse police culture.  In order to control for other (legal) determinants of the use of force, independent observers recorded whether the suspect showed resistance, the strength of the evidence against the accused, whether the suspect was subsequently arrested, whether the citizen showed disrespect to the police, as well as other factors that might explain the use of force by the police officer (e.g., whether the suspect showed evidence of the effects of drugs or alcohol). 

Results.  While many of these control variables (e.g., whether the suspect resisted the police officer, the race and gender of the suspect) had effects on the use of force by the police officer, above and beyond these “control” factors, the use of force was positively associated with the officers’ adherence to traditional police culture.  “Pro [traditional police] culture officers in both cities were significantly more likely to use force compared with the [anti-police] culture group…” (p. 1026). 

Conclusion. It is clear that those officers who endorse what might be called the “police culture” are more likely to use force (verbal and physical) in their interactions with citizens.  On the other hand, it was clear in these two U.S. police forces that a substantial number of officers did not hold these views.  The notion of a “universally shared police culture [may, therefore,] have been overstated by police scholars” (p. 1030). 

Reference: Terrill, William, Eugene A. Paoline, III, and Peter K. Manning (2003). Police Culture and Coercion. Criminology, 41 (4), 1003-1034.

Back to top

There is not a simple relationship among adolescents between working and deviance (e.g., alcohol use, arrest for a crime).  The effect of working on deviance depends on the nature of the job: some jobs increase involvement in crime; others decrease criminal involvement.

Background. The relationship between work and crime appears, to some extent, to be age related. “Although even marginal employment reduces the likelihood of adult crime, working more than 20 hours per week appears to increase delinquency and problem behaviours for adolescents” (p. 264).  Previous research has suggested that the peers one encounters at work may affect delinquency levels (see, for example, Criminological Highlights, 6(4)#5).  However, little research has been done on the impact of different kinds of work on delinquency. 

This paper analyzed data from a longitudinal study in Minnesota in order to try to understand the impact of work on three measures of adolescent deviance during the final year of high school: breaking school rules, alcohol use, and arrests.  The work characteristics that were examined were wages, whether youths viewed their jobs as giving them higher status among their friends, whether youths felt that their jobs provided learning opportunities, whether the youth was highly supervised at work, work stress, and whether youths thought that their jobs increased, decreased, or did not affect their school marks. 

The results showed that various work characteristics did have impacts on the various measures of deviance.  For example, looking at arrests in Grade 12, those youths who thought that their work reduced their grades in school were more likely to report being arrested than were those youths who reported that they had a great deal of autonomy in the work place.  Alcohol use in Grade 12 was associated with longer work hours, a belief that the job did not provide useful skills, more autonomy in the workplace, a belief that the work was interfering with grades, and the belief that the work enhanced one’s reputation with peers.  Misbehaviour in school was most common among youths who were involved in long hours of work, who saw work as interfering with their grades, and whose work provided a great deal of autonomy in the workplace.   

Clearly it is not simply the number of hours that is important in understanding the impact of adolescent work on deviance.  Certain characteristics of the work environment are important in understanding whether it will increase, decrease, or not affect adolescent deviance.  But in addition, there are “certain dimensions of adolescent employment [that appear] to exacerbate the “harmful” effects of long hours on adolescent delinquency” (p. 280).  Although long hours of work (e.g., more than 20 hours per week) generally appear to increase deviance in high school students, long hours of work that helped the students’ grades appears to decrease use of alcohol in Grade 12. 

Conclusion. “Job conditions are likely to exert very different effects in adolescence and adulthood.  To reduce delinquency, ‘good jobs’ in adolescence must support rather than displace academic roles and offer genuine opportunities to learn something useful.  Such jobs should also provide extensive controls, with circumscribed levels of autonomy, wages, and status among peers” (p. 283).  “Work experiences that are expected to decrease arrest, recidivism, and substance use in adults…. actually increase delinquency in adolescence” (p. 284).  

Reference: Staff, Jeremy and Christopher Uggen.  The Fruits of Good Work: Early Work Experiences and Adolescent Deviance (2003).  Journal of Research in Crime and Delinquency, 40 (3), 263-290.

Back to top

Not all women victims of spousal assault want the police to arrest the man who attacked them.  Generally speaking, disadvantaged women who have been victims of more severe violence are more likely to want the police to arrest the man who attacked them. 

Background.  It is now reasonably well established that a large proportion of women victims of spousal assault do not call the police when they are assaulted. Furthermore, of those women who do call the police, only a portion want the police to charge and prosecute “their” offender (See, for example, Criminological Highlights, 3(2)#8 and 3(5)#6).  Instead, they may have called the police for other reasons – e.g., to get their attacker to stop. “Whether it is desirable to remove all victim input from decision making [on filing and dropping charges] is a subject of dispute” (p. 316).  However, it is reasonably clear that arresting the offender cannot be counted on to have a deterrent impact on his future behaviour.

This study examines the views of 419 women in Charlotte, North Carolina, who were cohabiting with a male and had been assaulted by him.  In all cases, the police had been called in to deal with what was deemed to be a misdemeanour spousal assault.  Some days after the incident in which the police had been called, the woman was interviewed.  The women tended to be poor and black. About half were married and half were cohabiting. 

The results are consistent with previous studies: only about 30% indicated that they wanted their offender to be arrested.  Most wanted him taken away or warned.  A victim’s desire to have the offender arrested was associated with the victim being poor and black, having been assaulted many times by him in the previous 6 months, living with an offender who had previously been arrested, and being the victim of a more severe attack.  In other words, women who had been attacked more severely or who had been attacked by someone with a longer history of violence were more likely to want him to be charged. 

Previous research has shown that victims are just as good as actuarial methods in predicting future offending in circumstances such as these (Criminological Highlights, 3(2)#7).  Consistent with this finding is the fact that women in this study who wanted their offender to be arrested were more likely to be re-victimized than were victims who did not want arrests to be made.  Seven different measures of “re-victimization” were examined (e.g., a subsequent threat to the victim or her family, a subsequent attack on the victim or her family, etc.).  A fairly consistent predictor of re-offending was the number of times the victim had been previously hit.  Hence it is not surprising that it was these cases in which women were most likely to want “their” offender to be arrested. 

Conclusion. As with previous studies on this topic, it appears that a woman’s views on what should happen to the man who victimized her in domestic assault cases are not random.  Instead, such decisions appear to be predictably related to the facts of the case.  It can be argued that these data “support the argument that officials should pay attention to the preferences of victims of domestic violence when deciding whether to effect an arrest… Victims may possess knowledge that is of importance in predicting future danger” (p. 332). 

Reference: Hershel, David and Ira W. Hutchison. (2003) The Voices of Domestic Violence Victims: Predictors of Victim Preference for Arrest and the Relationship Between Preference for Arrest and Re-victimization. Crime and Delinquency, 49 (2), 313-336.

Back to top

Publication Type

Volume Number


Issue Number