Criminological Highlights Vol. 10, No. 1 – December 2008

View this issue as a PDF: PDF iconCrimHighlightsV10N1.pdf

This issue of Criminological Highlights addresses the following questions:

  1. Did Canada’s most recent (2003) youth justice law succeed in its goal of reducing the use of court?
  2. Does alcohol use cause youths to offend, or is it simply that those youths who drink are also more likely to offend?
  3. Why did U.S. imprisonment rates increase so dramatically since the 1970s?
  4. Are arrest and incarceration stigmatizing for poor urban African-Americans?
  5. Are youths identified as psychopaths especially likely to re-offend?
  6. Do restorative justice programs for adult offenders reduce re-offending?
  7. What use do juries make of an accused person’s criminal record?
  8. Are ‘teen courts’ effective at reducing offending?

Canada’s Youth Criminal Justice Act dramatically decreased the likelihood that police who apprehend a youth would refer that youth to court.

Canadian police have always had the authority to decide whether to charge someone who apparently had committed an offence.  Under the youth justice law in place until 2003, they were explicitly told that ‘taking no measures or taking measures other than judicial proceedings… should be considered for dealing with young persons who have committed offences.’  But under the law that came into force in April 2003,  police were told, in the legislation, that non-court measures were often preferable, and that they were obligated to consider non-court approaches in every case (though there were no consequences of ignoring this requirement).

This study answers a simple question: Did this aspect of Canada’s Youth Criminal Justice Act (YCJA)  have the impact that it was designed to have – to reduce the use of court for offences brought to the attention of the police?  Using police data on the number of youths apprehended and not charged, and apprehended but charged, it would appear that the desired impact was achieved.  The data supporting this conclusion are as follows:

  • Looking at the proportion of those youth who were apprehended by the police and subsequently charged, the drop in the charge rate that occurred when the YCJA came in force (from 56.4% to 44.6%) was dramatically larger than the change in previous years (an average decline of about 1% a year).  
  • This change occurred in the three months immediately following the date on which the YCJA came into force.  No comparable drop occurred in any other period. 
  • Between 2002 and 2005, there was an overall drop of 23% in the proportion of young persons charged. Consistent with what one would expect from the act (e.g., stronger pressure to use non-judicial approaches for minor property and for other minor offences), this decreased charge rate was much larger for minor property crimes (e.g., minor thefts: a decrease of 42%) and drug crimes (a decrease of 33%)  than it was for the more serious violent crimes (9% decrease) and major property crimes (14% decrease).  Minor assaults were between these two groups (23% decrease).  There was only a small decrease for probation violations and other administration of justice offences (5% to 8% decrease). 
  • The large decrease in the charge rate for youths occurred in all regions of Canada. 

Conclusion.  “One of the main objectives of the YCJA was to reduce referrals to youth court in Canada…. The statute makes clear that measures other than laying a charge… such as taking no action, giving an informal warning or formal caution, or diverting to a program are entirely appropriate forms of law enforcement with young offenders” (p. 362).  The changes that took place were consistent with the intent and timing of the legislation. It would appear, therefore, that the law was remarkably successful in achieving its goals. “Part of the reason for the success of the YCJA lies in the explicit and creative drafting of the sections of the statute dealing with [measures for responding to offending outside of the court system]” (p. 363). But in addition, it may have been that this was an idea whose time had finally arrived: charge rates had been drifting slowly downward in the years preceding the implementation of the YCJA.

Reference: Carrington, Peter J. and Jennifer L. Schulenberg (2008).  Structuring Police Discretion: The Effects on Referrals to Youth Court.  Criminal Justice Policy Review, 19(3), 349-367. 

Back to top

Some, but not all, of the relationship between alcohol use and crime among adolescents is due to the fact that, whether or not they have been drinking, youths who drink alcohol are more likely to commit crimes than youths who abstain from alcohol use.  

There are two separate explanations for the relationship between alcohol use and crime.  On the one hand, it is possible that alcohol intoxication has a causal relationship with certain types of crime, especially violence. For example, alcohol impairs judgement. For example, it has been suggested that alcohol causes people to focus on the immediate situation that they are in and to ignore future costs.  Such explanations would suggest that alcohol use could play a causal role in non-violent crime since the consequences of crime would be less salient. On the other hand, it has been suggested that alcohol use among adolescents may be the result of the same risk factors (e.g., socialization factors, having delinquent friends) that make certain adolescents more likely to commit crime. If this is the case, alcohol use per se may not be a cause of crime.

In a survey of 5,142  15-16 year-olds in 55 representative schools in Finland, respondents were asked whether they had engaged in a number of different offences (beating someone up, fighting, destruction of someone else’s property, graffiti, shoplifting, stealing from home, and car theft). If they had committed a particular kind of offence, they were asked if they had been drinking, when they committed the most recent incident,. They were also asked how often they drank alcohol.  

If there were to be a relationship between how frequently a youth drinks, generally, and how frequently that youth commits an offence when sober, such findings would suggest that at least some of the relationship between alcohol use and offending is due to a more general factor such as the youth being the ‘type’ of person who engages in both behaviours (drinking and offending).  On the other hand, if the relationship between frequency of drinking and total offending (offending when drinking and when sober) is stronger than the relationship between frequency of drinking and sober offending, it would suggest that alcohol use alone has a causal impact on offending. 

For all offences that were examined, the prevalence and frequency of drinking was related to committing offences when sober.  Drinkers committed more offences when sober than non-drinkers, and those who drank more frequently committed more offences when sober than those who drank less frequently.  For two offences – shoplifting and stealing from home – drinking measures were just as strongly related to offending while sober as they were to total offending, suggesting that drinking per se did not increase the likelihood of these two offences. In contrast, the results suggest that “intoxication has causal effects on adolescent violence and vandalism [in that the] drinking measures are related much more strongly to [total] delinquency than to sober delinquency for these offences” (p. 799).  

Conclusion. The evidence suggests that, for certain property offences, the relationship between drinking and offending is spurious – that the relationship is not causal.  For these offences, youths who drink may be more likely to offend, but it is because of other related factors, not the use of alcohol.  For other delinquent acts – most notably violent offending – it would appear that drinking increases the likelihood of offending.   These latter results are consistent with experimental studies and with some studies of men’s violence toward their intimate partners.

Reference:  Felson, Richard, Jukka Savolainen, Mikko Aaltonen, and Heta Moustgaard (2008).  Is the Association Between Alcohol Use and Delinquency Causal or Spurious? Criminology, 46 (2), 785-808.

Back to top

Why has the US imprisonment rate increased so dramatically over the past three decades?  It may relate more to the structure of government than to the attitudes of ordinary citizens toward crime or punishment.

It is often assumed that the reason US imprisonment rates increased by about 500% between the mid-1970s and mid-2000s is that ordinary citizens became upset with crime and demanded harsh penalties.  Conservative theorists suggested that “there is growing, justified outrage at what is happening to modern American society” (p. 68) that led people to demand harsh penalties. They suggest that “public opinion was well ahead of political opinion in calling attention to the rising problem of crime” (p. 268).  The only problem with this suggestion is that it does not fit the facts: “Repugnance toward serious offenders has been conspicuously common in human history, and that ‘state of nature’ cannot explain the recent emergence of the large and widening divide in harshness between the US and other democratic societies” (p. 270).  Aside from anything else, members of the public in almost all western countries hold the belief that penalties are not harsh enough. 

Why then, in most democratic societies, aren’t governments much tougher on ordinary crime? One possibility is that the problem of crime (which tends to affect the poor and disenfranchised more than it does those in positions of power) “is not an especially important issue in most advanced democratic systems” (p. 272).  In addition, in most democratic societies, “those places in government that set general crime policy are usually removed from review in elections and even from review by legislatures. They are located instead in parts of the executive and judicial branches of government that are distanced from direct or repressive democratic accountability” (p. 273).  Hence punishment is professionalized. The “criteria for making such decisions are regarded as involving principles that require professional judgment” (p. 273).   Furthermore, under systems that allow decision makers discretion in deciding on punishments, the decision is made “after persons to be punished become known to decision makers, [and, therefore, offenders] and their interests are transformed from abstractions into persons” (p. 273.)

It is argued that the “distribution of governmental power in American criminal justice… helps to explain the salience of criminal justice as an issue….” (p. 276).  Most punishment policy in the US is state policy.  For state governments, “crime policy looms large… because it has little with which to compete” (p. 276).  Lack of confidence in government may have contributed to high rates of imprisonment since one of the important changes in US punishment policy was to move toward fixed (offence based) punishment rules and away from individualized discretionary decisions. 

Conclusion. “Much of the punishment hardware that facilitates leniency depends on trust in government’s expertise and benevolence.  Citizens are restrained from acting on emotions and ‘throw away the key’ sentiments when they believe that there are principles of punishment – legal proportionality, predictions of dangerousness, responsiveness to treatment – that require governmental expertise.  As soon as the claim of expertise is discredited, people on the street (or their state representatives) are every bit as expert as judges, parole boards, or correctional administrators” (p. 276). “What has always distinguished the governance of punishment in the United States from other advanced democracies is a structural vulnerability to democratic pressures that arises out of federalism, the election of prosecutors and judges, and high levels of life-threatening violence.  These enduring features have coexisted with hostility toward criminals and enthusiasm for punishment that seem typical of other advanced democracies. The combination of higher salience and distrust of government increased punishment directly and produced structural changes in sentencing that made punishments even harsher” (p. 278). 

Reference: Zimring, Franklin E. and David T. Johnson (2006). Public Opinion and the Governance of Punishment in Democratic Political Systems. Annals, AAPSS, 605, 266-280.

Back to top

Arresting and attempting to stigmatize minority youths from high-poverty urban neighbourhoods in the U.S. has little impact on them. Mass incarceration of poor urban Blacks has already made being incarcerated the norm rather than the exception and as a result it is not stigmatizing.  

Labeling theory suggests that “formal sanctions often reinforce the very delinquent behaviours they seek to extinguish” (p. 576).  This is a particular concern because thousands of youths in poor urban neighbourhoods are arrested and incarcerated. Nevertheless, it is possible that wholesale criminalization of urban black youth has, in effect, diluted the impact of contact with the criminal justice system. 

As part of a larger study on the impact of juvenile sanctions on attitudes, behaviour, and school performance, twenty 18-19 year-old youths with an average of almost six officially recorded juvenile arrests were interviewed.  Most (16) had, on at least one occasion, been incarcerated. Ten of these 20 youths reported that the police had falsified evidence in their cases. Seventeen were African-American, two were of Latino descent, and one was of mixed race.  Close to half (47%) of the students in the schools attended by these youths had been involved in the juvenile justice system. Hence, being labelled as a delinquent was hardly exceptional for these youths. 
 
For these youths, being arrested was “no big thing.” “Arrests generally caused little stir outside of informants’ immediate family” (p. 585).  Part of the reason for this may be the high rate of arrest in the community.  In one school, for example, 66% of the black male youths were arrested in one year.  Although families of arrested youths were often disappointed when a youth was arrested, peer-group reactions were minimal.  Indeed, only three of the youths reported any negative impacts among their peers of being arrested.  “Thus, the first two preconditions for a ‘labeling’ effect to occur following arrest – a negative meaning endowed to the arrest and confirmatory treatment from significant others – did not always occur” (p. 590).  Only three of the 20 youths showed signs of diminished self-esteem or shame as a result of their arrests. 

Conclusion. “By the time of their first arrest, youths… viewed arrest as a normal part of adolescence in their communities.  The high prevalence of juvenile arrests in the community and beliefs in the pervasiveness of unwarranted arrests likely predispose many community members to take news of routine arrests lightly” (p. 594).  In effect, these youths’ crimes “were not… considered a grave violation of local norms” (p. 594).  In the poorest urban neighbourhoods, “justice system labels are not merely sewn onto individuals, but they are etched into the social fabric of the ghetto. Heavy police presence and high rates of incarceration are now fixtures of community life…. In communities devoid of meaningful educational opportunities, organized sports, jobs for youth, and valued extra-curricular activities, the justice system offers accessible, institutionally mediated means of individuation and status attainment. Accordingly, anecdotal evidence suggests that some youth, especially gang members, seem to anticipate or even value arrest or juvenile detention” (p. 596).  “Perhaps youth in some isolated neighbourhoods have developed a collective deviant or ‘oppositional’ identity as an adaptation to collective stigmatization within the larger society” (p. 597).  It would appear that if society wants youths to fear being arrested, brought to court, and incarcerated, it should not promote policies that create these as normative treatment that most youths can expect to experience. 

Reference: Hirschfield, Paul J. (2008). The Declining Significance of Delinquent Labels in Disadvantaged Urban Communities.  Sociological Forum, 23 (3), 575-601.

Back to top

Youths who are labelled as psychopaths by various psychological measures are no more likely to reoffend than are  youths with low scores on measures of psychopathy once the standard social and personal predictors of recidivism are taken into account.

The search for personality measures that predict recidivism in youths is popular in large part because many youth justice systems use the likelihood of future offending as a factor in deciding what should happen to a youth who is implicated in crime.  Although there have been serious concerns raised about the adequacy of the concept of psychopathy in describing adolescents (see Criminological Highlights, V5N1#3), the fact that measures of youth psychopathic tendencies predict, to some extent, future offending lends credence to the argument that these measures are useful.

The important question, when employing a construct such as psychopathy, is whether it adds to our ability to predict who will re-offend once less exotic predictors (e.g., age, past criminal behaviour, substance related problems, mental health problems such as attention deficit hyperactivity disorder) are taken into account.  In this paper, a group of 85 young offenders (42 sex offenders and 43 other offenders) in a secure remand facility in Florida were assessed on a number of measures of psychopathy among youths.  This is obviously the kind of population – people for whom there is special concern about re-offending – for whom careful assessments are most likely to be important.  In addition, measures were obtained of ordinary predictors of criminal activity (e.g., past offending, etc.).  Youths were tracked for re-offending for an average of about 2 years. For some analyses, a uniform time-at-risk of re-offending of about 20 months was used. Measures of violent, non-violent, weapons-related, and overall recidivism were obtained. 

Two of the overall measures of psychopathy were strongly correlated (r=.79) but neither of these two measures correlated well with the third adolescent psychopathy measure (r’s of .27 and .35).  The various sub-scales of each measure produced similar patterns.  More importantly, even before the standard predictors of recidivism were taken into account, none of the overall measures or their subscales showed substantial correlations with any of the measures of recidivism (no correlation exceeded 0.32, and many were not significant).

More important was the finding that once the standard predictors of recidivism -  such as past history of property offending, alcohol abuse, and age were controlled for statistically, none of the psychopathy indices added to the ability of the standard predictors to predict any form of recidivism.  The overall correlations between this set of standard predictors and recidivism were in the range of 0.60 to 0.65  -- dramatically higher than the correlations involving psychopathy alone. 

Conclusion. For a personality measure to be useful in predicting recidivism, it must add to our ability to identify those youths who will re-offend above and beyond the predictors of recidivism that are well known both to criminologists and to those working with youths (e.g., past criminal behaviour, alcohol use, age).  These three standard measures of psychopathy consistently failed this “incremental” test.  Said differently, the study suggests that there is no justification for using any of these psychopathy measures to predict future offending, given that ordinary measures are far better for this purpose. 

Reference: Douglas, Kevin S., Monica E. Epstein, and Norman G. Poythress (2008). Criminal Recidivism Among Juvenile Offenders: Testing the Incremental and Predictive Validity of Three Measures of Psychopathic Features. Law and Human Behaviour, 32, 423-438.

Back to top

Although victims and offenders report being very satisfied with the outcome of restorative justice programs for adults and the programs, overall, reduce the frequency of offending, there does not appear to be convincing evidence that these programs consistently reduce the likelihood or the severity of offending.

Restorative justice programs typically have a number of different goals, only one of which is to reduce recidivism (see Criminological Highlights, V6N5#1). More generally, restorative justice has been defined as “a process whereby parties with a stake in a specific offence collectively resolve how to deal with the aftermath of the offence and its implications for the future” (p. 1).  This paper looks at three restorative justice programs, dealing largely with adult offenders.  The restorative justice process used in these programs was in addition to normal criminal justice processing.  Hence, the restorative justice procedure (e.g., mediation or a conference) might take place after a finding of guilt or after a final warning had been given.  Thirteen separate sub-programs involving restorative justice interventions were assessed In some of the sub-studies, participants were randomly assigned – after agreeing to participate in a restorative justice conference – either to receive the conference or not. In other programs, participants were matched with comparable offenders who had not been given the opportunity to participate.  

Offenders were followed for two years after the restorative justice intervention took place. One difficulty in assessing these 13 interventions is that most of them had few participants.  Hence, for an effect to be statistically significant, the impact of the intervention on recidivism would have to have been quite dramatic or consistent.  In fact, in terms of the reduction in offending – defined as the presence of at least one reconviction within two years or the average severity of the reconvictions – there was only one program (out of 13) in one location in which the restorative intervention reduced offending. That program involved a conference that took place after a guilty plea and before sentencing for property offenders. Summing across all 13 of the interventions, there was no consistent impact of the restorative justice intervention on whether or not the person reoffended or on the severity of the subsequent offences. However, there was an overall significant decrease in the frequency of reoffending.  Moreover, for the most successful of the programs, the savings to society from reduced offending were larger than the cost of mounting the restorative justice programs. Furthermore, given that restorative justice programs are well received both by victims and by offenders, it is also important to note that there were no significant findings suggesting any criminogenic effects of restorative justice programs. 

The study attempted to discover the types of cases for which restorative approaches were most effective.  Unfortunately, “it is not possible to predict, from this evaluation, that one offender will be more likely to benefit from restorative justice than another on the basis of their prior characteristics” (p. 67). 

Conclusion:   The restorative justice programs examined in this paper were designed to accomplish a number of goals – reducing recidivism was only one of them.  The adult offenders in these programs, many of whom had been found guilty of serious offences tended to commit fewer offences in the two years following the intervention than did those who did not experience the restorative justice program. However, the programs did not reduce the likelihood that offenders would offend at least once, nor did the restorative justice intervention reduce the severity of reconvictions.  In other published reports on these same evaluations, very high levels of satisfaction were found for both offenders and victims.  It may be that the findings on satisfaction of the participants, alone, justify continued interest and investigation of restorative justice programs even though their impact on recidivism is less consistent across measures.

Reference: Shapland, Joanna, Ann Atkinson, Helen Atkinson, James Dignan, Lucy Edwards, Jeremy Hibbert, Marie Howes, Jennifer Johnstone, Gwen Robinson, and Angela Sorsby (2008). Does Restorative Justice Affect Reconviction? The Fourth Report from the Evaluation of Three Schemes.  Ministry of Justice, United Kingdom.  Available from www.justice.gov.uk.

Back to top

Juries use an accused person’s criminal record to convict defendants when the evidence against the accused is weak.

A principal reason for criminal defendants’ decisions not to testify in jury trials is that they have criminal records. Jurors typically only hear about the prior criminal record of accused persons if they testify. The fear is that the jury would convict, not because of the evidence, but because the defendant was already a proven criminal.  In some jurisdictions, judges must decide on the admissibility of a criminal record of an accused person by balancing two things:  the value of the evidence of the criminal record of the accused person in determining the credibility of the accused and the possible prejudicial impact of the criminal record on the verdict in the case. 

This study examined 358 jury trials in 2000-2001, largely involving minority (Black or Hispanic) defendants, in four large urban locations in the U.S.. Case information was gathered from the court clerk (on such factors as whether the accused testified and whether the jury found out about the criminal record, the jury decision, and information about the accused). Before the jury returned with a verdict, trial judges rated such matters as the strength of the evidence and what their own verdict would have been. After their deliberations, jurors gave ratings of various parts of the case.

In cases in which defendants testified, judges reported that the testimony of the accused was – on average – more important than all other evidence except that of the victim and eyewitnesses. Not surprisingly, defendants were more likely to testify in cases in which they did not have a criminal record (62% testifying) than when they had a criminal record (45%).   This is largely due to an effect for minority defendants. 62% of minority defendants without a criminal record testified compared to only 43% of those with a criminal record.  The comparable figures for whites were 67% and 61%, respectively.  When the accused had a criminal record and testified, the jury found out about the record in 52% of the cases.  If the accused did not testify, the jury seldom (9%) found out about the record.

One would expect that the criminal record of the accused would have the strongest impact for cases in which the evidence against the accused was weak.  In cases in which the evidence is very strong, one would expect that the criminal record would not matter.  Consistent with this, in cases with relatively weak evidence in which the jury heard of the accused criminal record, the likelihood of conviction was higher than if the jury did not hear about the record.  In relatively strong cases, only the strength of the evidence predicted whether the defendant was convicted.

Defendants with criminal records were rated as being just as believable as defendants without criminal records. This result is similar to findings from experimental research. The impact of criminal record does not, therefore, appear to operate through the legally permissible mechanism of defendant credibility.  Rather it appears to have a direct impact on the likelihood of a finding of guilt. 

Conclusion. It would appear that “One could view the prior record as ‘making up’ for evidentiary deficiencies” (p. 30).   It may be that jurors infer guilt directly from the knowledge that the accused has been found guilty at least once in the past. Or it could be that the threshold necessary for a finding of guilt drops in cases in which the accused has a record, on the assumption that it matters less if an innocent accused with a criminal record is found guilty than if the accused does not have a criminal record.  Finally, it is possible that the criminal record changes the meaning of the evidence against an accused. The same evidence may be seen as being more incriminating if the accused has a record.  In any case, it appears that a defendant’s criminal record promotes findings of guilt in exactly the cases – those with weak evidence – in which wrongful convictions are most likely to occur.

Reference: Eisenberg, Theodore and Valerie P. Hans (2007). Taking a Stand on Taking the Stand: The Effect of a Prior Criminal Record on the Decision to Testify and on Trial Outcomes. . Cornell Law School, Legal Studies Research Paper No. 07-012.

Back to top

Another quick-fix crime prevention program is shown to be ineffective. Teen courts – in which decisions concerning juvenile offenders are made by peers – may even increase the likelihood of future offending.

Teen courts are designed to divert youths who have apparently committed minor offences from traditional youth courts into settings in which other adolescents take an active role in deciding what should happen to the offender. It is estimated that there are now over 1000 such programs operating in the U.S., processing nearly 100,000 cases.  One reason that they may be so popular is that they are designed – consistent with many restorative justice programs – to try to make the offender aware of the impact of the offence on the victim. 

Though there are many stories of successes, the teen court literature suffers from the absence of comparison groups.  At best, it could be said that there have been some encouraging results in some locations, but not others. In this study, youths in Maryland who met the eligibility criteria for teen courts were randomly assigned either to be processed in the normal way or in a teen court. They were not given a choice as to which program they participated in, but youths in both groups had the choice of whether they participated in the research.  Four months after they had participated either in ‘normal’ processing or in the teen court, the youths answered confidential questions about drug use and general delinquent behaviour. In addition, official records related to police apprehensions were searched for a period 18 months after referral to the programs. 

In terms of official recidivism (measured as re-arrest), there was not a significant difference between the groups, but if anything, the re-arrest rate was slightly higher for the youths who were assigned to the teen court.  In addition, on the self-report measures, the teen court participants were, if anything, more likely to report drug use and being involved in delinquent behaviour.  On various attitude measures, the results, if anything, suggested that traditional treatment of these youths was more effective than the teen court. 

Conclusion. As the authors point out, in the US “enormous amounts of time and money are spent on teen court programs each year, without strong evidence of their effectiveness” (p. 150).  Though the effects were not strong, if anything the results suggest that “teen court youth were consistently found to have less favourable outcomes than those in the [traditional youth court] sample” (p. 151).  Such findings are not surprising and might well have been predicted from various perspectives.   It is possible, for example, that the youths feel stigmatized by the process of being ‘shamed’ in front of their peers as opposed to the more traditional anonymous court process.  In addition, it is possible that some of the sanctions imposed by the teen court (e.g., tours of detention centres – see Criminological Highlights, 6(2)#4)  may have had negative impacts.  But the point remains: giving the responsibility for sanctioning of minor offenders to other youths appears, if anything, to have negative impacts on those being sanctioned.

Reference: Stickle, Wendy Povitsky, Nadine M. Connell, Denise M. Wilson, and Denise Gottfredson (2008).  An Experimental Evaluation of Teen Courts. Journal of Experimental Criminology, 4, 137-163.

Back to top

Publication Type

Volume Number

10

Issue Number

1