Criminological Highlights Vol. 3, No. 3 – July 2000

View this issue as a PDF: PDF iconCrimHighlightsV3N3.pdf

Systematically measured neighbourhood disorder (“broken windows”) does not cause crime in a community. “The current fascination in policy circles… on cleaning up disorder through law enforcement appears simplistic and largely misplaced, at least in terms of directly fighting crime” (p.638).   “Broken windows” may be more prevalent in high crime areas, but the data suggest that disorder is not directly responsible for crime. 

Background.  The idea that “fixing broken windows” will reduce crime has been popularized, but never demonstrated empirically, by various criminologists such as George Kelling and James Q. Wilson.  The notion they have popularized, based on the metaphor of broken windows,  is that “public incivilities – even if relatively minor as in the case of broken windows, drinking in the street, and graffiti – attract predatory crime because potential offenders assume from them that residents are indifferent to what goes on in their neighbourhood” (p. 604).  Politicians in favour of crackdowns who are looking for a political “quick fix” find “broken windows” an attractive theory. The alternate theory is that “structural constraints such as resource disadvantage and mixed land use account for both crime and disorder simultaneously” (p. 614).  

This study reports a careful examination of the “broken windows” theory of crime by first getting an independent observation by researchers of how “disordered” (socially and physically) neighbourhoods (in Chicago) actually were.  Social disorder (e.g., adults loitering or congregating, drinking alcohol in public, drug selling) and physical disorder (e.g., presence of garbage or litter, graffiti, abandoned cars) were quite highly correlated. Not surprisingly, “disordered” neighbourhoods were poorer, more likely to have high concentrations of immigrants, and lower in “collective efficacy” (willingness of neighbours to “do something” in response to problems, trusting one’s neighbours, neighbourhood social cohesion, etc.).  Collective efficacy has been found in previous studies to be an important predictor of neighbourhood crime above and beyond characteristics of the individuals in the neighbourhood.

The most important findings, however, were that measures of social and physical disorder (“broken windows”) were not related to personal violence and household burglary (assessed by victimization measures) once characteristics of the neighbourhood (e.g., collective efficacy, mixed land use) had been controlled for. “The results are consistent and point to a spurious association of disorder with predatory crime” (p. 627).  When one looks at officially recorded crime, “disorder” once again disappears as a predictor of homicide and burglary once measures of collective efficacy and prior crime rates are controlled for.  “The key result is that the influences of structural characteristics and collective efficacy on burglary, robbery, and homicide are not mediated by neighbourhood disorder” (p. 629).  The exception is the case of officially recorded measures of robbery where there is still a relationship with disorder. Whether this is due to a “complex feedback loop” (p. 637) or an artifact of official data (e.g., “citizen calls to the police or police accuracy in recording robberies is greater in areas perceived to be high in disorder” --p. 638) is not clear. 

Conclusion.  “The active ingredients in crime seem to be structural disadvantage and attenuated collective efficacy more than disorder.  Attacking public disorder through police tactics may thus be a politically popular but perhaps analytically weak strategy to reduce crime, mainly because such a strategy leaves the common origins of both [disorder and crime], but especially the last [crime] untouched.  A more subtle approach suggested by this article would look to how informal but collective efforts among residents to stem disorder may provide unanticipated benefits for increasing collective efficacy... in the long run lowering crime” (p. 638). 

Reference: Sampson, Robert J. and Stephen W. Raudenbush.  Systematic social observation of public spaces: A new look at disorder in urban neighbourhoods. American Journal of Sociology, 1999,  105,  603-651.

Back to top

Partner abuse and “general crime” are related: those likely to commit one type of crime are likely to commit the other.  However, the two kinds of criminal behaviour are different in that general crime is related to low self-control, but partner abuse is not.  These findings apply to both men and women. 

Background.  “Evidence is accumulating to suggest that many, if not most, partner abusers do not specialize, but engage in violence against nonintimates as well as a variety of non-violent crime” (p. 201).  Generally speaking,  “childhood and adolescent antecedents of adult partner violence are remarkably reminiscent of the risk factors for general crime” (p. 202).  However, there is some evidence that there are some personality factors that are different for those who commit “general crime” compared to those who commit partner abuse. 

This study focuses on two personality constructs: 
“Constraint” (where people characterize themselves as reflective, cautious, careful, preferring safe activities, etc.) or more typically what criminologists often refer to as “self-control”  and  
“Negative emotionality” (where people describe themselves as nervous, prone to worry, unable to cope with stress, etc.).    

These characteristics were measured in a representative sample of  New Zealand 18-year olds.  When these same people were 21 years old, self-report measures of general crime (theft, fraud, vice, physical force) and partner abuse (including, but not limited to physical violence) were obtained. Unlike many studies, this study examined both men and women. 
The findings suggest that, consistent with other studies, physical abuse of one’s partner was highly correlated with psychological abuse.  In addition, there was a moderate relationship (r=.39) between general crime and partner abuse.  Looking only at physical abuse of one’s partner and violent crimes against persons other than partners, for both males and females, those who commit one form of violence are likely to commit the other.

For both males and females, those high on “negative emotionality” were more likely to be involved in general crime and in partner abuse.  “Constraint,” however, was related to general crime but was not related to partner abuse.  Very similar findings were obtained when “violent crime” and “partner violence” were examined.  In other words, while “partner abuse and general crime represent different constructs that are moderately related, they are not merely two expressions of the same underlying anti-social  propensity” (p.219). 

Conclusion.  These findings suggest that certain personality characteristics – specifically, “negative emotionality” which is related to an inability to cope with stress, having a low threshold for feeling tense, fearful, and angry, etc. -- are related to violence against partners and non-intimates for both men and women.  “Constraint,”  on the other hand,  does not seem to be related to partner abuse but is related to more general types of crime.  Hence the two forms of anti-social behaviour (partner abuse and general crime) are not related to exactly the same constellation of personality characteristics.  Interventions aimed at those who commit anti-social acts against intimates and non-intimates, therefore, need to take account of both the overlap between the two and the differences.

Reference:  Moffitt, Terrie E., Robert F. Krueger, Avshalom Caspi, and Jeff Fagan. Partner abuse and general crime: How are they the same? How are they different?  Criminology,  2000, 38, 199-232.

Back to top

Most sex offenders placed on probation do not reoffend.  In particular, those with stable employment histories who receive sex-offence treatment are less likely to reoffend than those without such characteristics.

Background.  Previous research has found that “incarceration had an indirect effect on reoffending: It reduced job stability, and this instability in turn contributed to continual involvement in crime” (p. 64).  In contrast, other theoretical frameworks that focus largely on traits developed early in life view “crime as highly resistant to both informal social controls... and … the formal controls exerted by the criminal justice system” (p. 64-5).   Although sex offenders are, in many countries, increasingly subject to punitive approaches, many of those who have committed sex offences serve at least part of their sentences in the community.  Hence it is possible to see whether social bonds to employment and family can reduce crime or, alternatively, if treatment strategies are a waste of time. 

This study looked at 556 sex offenders in Minnesota from the time they were placed on probation in 1992 through June 1997.  Hence it under-represents the worst sex offenders but does include the majority of those identified as sex offenders during the period of study.  Formal social control was operationalized as drug testing, prohibitions against contact with minors, and compulsory treatment. 

Results.  Reoffending of any kind (as measured by official re-arrest measures) was more likely to occur soon after the offender was placed on probation.  After a year, 83% had not reoffended, and after two years, 75% had not reoffended.  Five years after being placed on probation 65% had not reoffended.  During the five years, only 10% were ever rearrested for any offence against persons, and only 5.6% had committed a new sex offence. Those with a long criminal history and a history of drug use were more likely to re-offend. Not surprisingly, older probationers were less likely to commit any new offence.  Even though 36% of the original offences involved family members, those living with their families during the probation period were not more likely to reoffend.

Generally speaking, job stability appeared to be related to lower rates of overall offending and crimes against persons.    Sex offender treatment, alone, did not appear to be effective.  However, for those with stable employment, sex offender treatment appeared to be useful in reducing all types of reoffending.  “In fact, for the small number of respondents whose reoffence was a new sex offence, the only factor that even marginally reduces their risk of reoffending is the combined effect of stable employment and sex offender treatment” (p. 81).  

Conclusion.  For this group of sex offenders, who were given non-custodial sentences, it appears that community approaches can be effective in reducing reoffending and that scarce treatment resources can be effectively allocated.  Most of these offenders did not reoffend and, when they did, it was most likely not a sex offence.  The findings, showing a positive impact of stable employment (especially when combined with treatment), suggest that informal social controls can be important in understanding reoffending. 

Reference: Kruttschnitt, Candace, Christopher Uggen, and Kelly Shelton.  Predictors of desistance among sex offenders: The interaction of formal and informal social controls.  Justice Quarterly, 2000, 17, 61-87.

Back to top

All conditional sentences are not created equal: The public is much more likely to accept a conditional sentence as a substitute for prison if there are conditions attached that are clearly punitive.

Background. The controversy surrounding the conditional sentence of imprisonment did not end with the recent Supreme Court of Canada decisions on the appropriateness of this sanction.  Part of the controversy involves questions of what a conditional sentence “looks like.”   It is understandable that for many members of the public a conditional sentence may look remarkably like a term of probation since, after all, both involve sanctions served in the community.  What courts seem to be saying is that a conditional sentence must involve a visible component of “punishment.”  The punishment, in turn, is meant to serve the purposes of denunciation and deterrence – sanctions typically associated with prison.

This study examined, in a national public opinion survey, the public acceptability of a conditional sentence.  The sentence was described (to different groups of people) in two different ways. One group of respondents was told that a judge was deciding whether to sentence an offender found guilty of a break, enter, and theft to a 6 month prison sentence or to 6 months to be served in the community as a conditional sentence.  The other (equivalent) group was given the same choice but was told that the conditional sentence would include, as conditions imposed by the judge, each of the following: a weekend and evening curfew, restitution and community work. 

The results demonstrate that a little bit of punitiveness went a long way.  When choosing between prison and the conditional sentence (without punitive conditions) only 28% indicated that the conditional sentence was their choice.  With the additional punitive sanctions, the conditional sentence was endorsed by 65% of the respondents. “The creative use of appropriate optional conditions can have a dramatic impact on community reaction to the imposition of a conditional sentence” (p.119).  

Part of the difficulty with conditional sentences, then, may be that punitive optional conditions are not routinely imposed.  The data from a sample of conditional sentences and probation orders in Ontario suggest that most optional conditions are as likely to be imposed in the case of probation as they are for conditional sentences.  The exceptions to this generalization are that abstaining from drugs and observing a curfew (both of which are no doubt seen as being punitive) are more likely to be used for conditional sentences.  Weapons restrictions were also more likely to be imposed in the case of conditional sentences (perhaps because of the severity of the offences involved).  There was also provincial variation in the imposition of optional conditions. 

Conclusion.  Punitive non-carceral conditions made part of a conditional sentence order can have dramatic effects on the acceptability of a conditional sentence.  While the public might not like the idea of a conditional sentence in the abstract, these sanctions can be made acceptable if conditions are attached to them that appear to be capable of fulfilling the purposes traditionally attributed to imprisonment.

Reference:  Roberts, Julian V., Dan Antonowicz, and Trevor Sanders.  Conditional sentences of imprisonment: An empirical analysis of optional conditions. Criminal Reports, 2000, 30, 113-125.

Back to top

Three strikes puts courtroom officials in the hotbox: California’s three-strikes legislation has a disruptive impact on the processing of serious cases.  

Background:  The California 3-strikes law is one of the harshest 3-strikes laws in the U.S. On the second felony conviction, offenders get twice the “normal” sentence that would apply to their offence.  The third strike gets 3-times the punishment that would apply to first offenders or 25 years to life, whichever is longer.  This third strike sentence is imposed even if the third strike is for a minor felony. Hence an offender can get 25-years- to-life for a minor felony. 

An examination of California’s 3-strikes laws shows the reality of the administration of overly harsh laws.  When the law was brought in, the assumption made by many was that the administration of the law would be undermined by judges, prosecutors, and defence counsel who would not implement the law as written. This is not what universally occurred. 

This study interviewed and surveyed judges, prosecutors, and public defenders in five large California counties.  “Both methods indicate that Three Strikes has significantly disrupted the efficiency of the courtroom and has made the prediction of case outcomes difficult” (p. 192).  For example, in four of the five counties studied, it appears that almost all prior strikes were introduced into evidence.   Plea bargaining became difficult because it became difficult to predict when prosecutors would be willing to dismiss prior “strike” allegations.   On the other hand, there was some evidence that judges were more willing to ignore prior convictions in counties where prosecutors went by the book.  

“The greatest effect of Three Strikes for workgroup (judges, prosecutors, defence) members has been an increase in trials….  Three strikes prohibits such deals [where a guilty plea is substituted for a lesser punishment].  Defendants who face extended prison terms are unlikely to agree to plead guilty…. Overall the felony trial rate is higher than before Three Strikes….”  (p. 198)

“Recognizing [the possibility of jury nullification], public defenders attempt to inform the jury that the current offence is a third strike” (p. 199).   But even though judges “routinely offered second strike defendants the lowest possible sentence, seemingly to encourage defendants to plead guilty… substantial numbers of [prosecutors, lawyers and judges] believed that they could not predict which cases were likely candidates for leniency” (p. 201).  

Conclusion: It is clear that the 3-strikes law in California is having a disruptive impact on the sensible running of the courts.  Because the law requires disproportionately severe sentences to large numbers of offenders, there are efforts (some successful) to avoid the harshness of the law.  The result is inconsistency in the application of the law and in the outcome of criminal cases in California.

Reference:  Harris, John C. and Paul Jesilow.  It’s not the Old Ball game:  Three strikes and the courtroom workgroup.  Justice Quarterly, 2000,  17, 185 –203.

Back to top

Young male and female victims of violence look remarkably like young violent offenders.  They are likely to be involved in offending; they have difficult childhoods, and they are likely to be living relatively “risky” lives.

Background.  Young people are, in general, more likely to be victims of a violent crime than are members of the general population.  This study examines who, among a representative sample of young people, are most likely to be victims of violence.

This study focuses on a group of 1024 New Zealand children born in 1977 and followed until they were 18 years old. 14% of the females and 23% of the males in this representative sample of youths reported being physically assaulted when they were between the ages of 16 and 18. Eleven percent of the assaults against the young women occurred in the context of a sexual assault.  Most of the assaults were minor: Only 2 of the 188 youths who were assaulted required admission to a hospital although 17% indicated that  they had needed medical attention.  The study examines the contextual as well as the background factors associated with victimization.

Those who had been assaulted between the ages of 16 and 18 were much more likely to report involvement in various types of offending (violent, property, drug or alcohol use). They were also more likely than those who were not assaulted to be associating with “deviant or anti-social” peers. “The psychosocial profile of physical assault victims was that of a group of young people characterized by elevated rates of risk-taking behaviour and deviant peer affiliations” (p. 246).

Young victims of violence were more likely than non-victims to come from difficult backgrounds (e.g., lower income and families characterized by parental offending, alcohol and drug abuse, family conflict, and regular physical punishment).  More detailed analyses suggest that “a history of early adolescent conduct problems and exposure to parental alcohol problems may uniquely contribute to the prediction of physical assault [victimization] even after current behaviour and lifestyle factors have been taken into account” (p.250-251).  “The risk factors and life pathways associated with physical assault were similar for both males and females” (p. 251).   

Conclusion. “For both males and females, the profile of those [youths] at greatest risk [of violent victimization] was that of a young, conduct disordered, adolescent reared by physically punitive and substance-abusing parents, who upon reaching late adolescence, engaged in antisocial and other risk-taking behaviour” (p. 254). 

Reference:  Woodward, Lianne J. and David M. Fergusson.  Childhood and adolescent predictors of physical assault: A prospective longitudinal study. Criminology, 2000, 38, 233-261.

Back to top

Canadians generally support both the Charter and the Supreme Court.  In certain controversial areas (e.g., gay rights) Canadians appear to be moving in the direction of endorsing positions consistent with those taken by the Supreme Court.

Background. The Canadian Charter of Rights and Freedoms has seen a lot of controversy in its 18 year life.  In addition,  the courts, and the Supreme Court in particular, have been criticized for, among other things, having the last word on public policy issues. The results of a recent survey suggest that Canadians support the Charter and the way in which it has changed the Canadian political landscape.

This study examined Canadians’ views of the Charter in 1999 comparing these recent views to those expressed in a 1987 survey.  Some of the highlights of the survey are as follows:

  • Most Canadians (82%) who have heard of the Charter think that the Charter is “a good thing for Canada.”  Very few (7%) think it is a bad thing; the rest are not sure or refused to answer.  These results are almost  identical to those obtained in the 1987 survey.
  • The Charter is highly regarded by a majority of those living in all regions of the country.   In no region is the percent who disapprove of the Charter higher than 10%. 
  • In 1999, as in 1987, most members of all federal parties (Liberal, PC, and NDP in 1987 plus the Reform and BQ in 1999) held favourable attitudes toward the Charter.
  • As in 1987, most Canadians (62%) in 1999 thought that courts, rather than legislatures, should have the final say on the validity of a law. There was no significant regional variation.  However, “the one group that shows marked opposition to judicial authority is Reform supporters, who split down the middle on [whether] the courts vs. legislature should have the final say]” (p. 14). 
  • There was some evidence that, in comparison with the “rest of Canada,” Quebecers were more likely to endorse limits on the Court’s power. (p. 18-19). 
  • As in 1987, most Canadians (67% to 70%, depending on the facts of the case) in 1999 disapproved of the exclusion of illegally obtained evidence in court. 
  • The overwhelming majority of Canadians (78%) in all regions of the country were in favour of providing protection for homosexuals in human rights legislation (most support in Quebec, 89%, and least in the prairies, 66%).   Politically, members of all political parties (range 79%-97%) approved this with one exception: Reform supporters were split with only 49% supporting this form of protection for homosexuals.  
  • Since 1987 support for allowing gays the right to teach in schools increased from 52% to 77%.  Support went up for people born in each decade with one exception: those in their 60s in 1999 had not become more tolerant. 
  • The linkage between views on specific cases and overall support of the Charter and the courts varies with the issue.  “The judicial decisions of which Canadians approve have greater leverage over their general feelings toward the Charter and courts than do those they oppose” (p. 53).

Conclusions.  Support for the Charter, as well as support for some controversial Charter issues, appears to be high among most geographic and political groups in Canada.  Although some opposition can be found to certain decisions, it would appear that the courts have succeeded in staying in tune with or in leading public opinion.

Reference:  Fletcher, Joseph F. and Paul Howe.  Canadian attitudes toward the Charter and the Courts in comparative perspective.   Supreme Court Cases and Court Support: The State of Canadian Public Opinion. Choices, Volume 6, Number 3.  Montreal: Institute for Research in Public Policy, 2000.

Back to top

Jurors need help to do an effective and efficient job.  Most jurors are competent to do their job, but “many of them confronted significant difficulties in doing so because they were not provided with adequate tools” (p. 89).

Background.  Juries may not be used very often, but they are seen as being a “cornerstone of the criminal justice system.”  In Canada, we know little about the operation of actual juries in part because of statutory rules prohibiting the disclosure of information about what goes on during deliberations.  

This study examines juries in 48 New Zealand trials in a variety of ways: pre-service questionnaires given to jurors, observations and examinations of transcripts,  and interviews with the judge and with jurors. 

Jurors, generally, “felt unprepared for the nature of the task and expressed concerns about the responsibilities inherent in jury duty” (p. 90-91).  One difficulty is that jurors’ jobs -- “passive observers and recorders of information who suspend judgement on the evidence and issues until they retire for deliberations” (p. 91) -- do not reflect the way in which research shows that people normally make judgements.  “It is scarcely surprising, therefore, to find that a significant number of jurors were critical of the fact that they failed to receive an adequate factual and legal framework at the commencement of the trial” (p. 91).  Various straightforward ways of providing jurors with an adequate legal framework exist but are not typically used in trials (see p. 92).  Part of the problem jurors have is in following and remembering details of the evidence.  Hence there is clear support for addressing the manner in which jurors get and retain information. Almost 80% of the jurors said that they wanted to ask at least one question but, given that they are discouraged from doing so, few did. 

The deliberations in a number of trials were described as “unstructured, disorganized and inadequately facilitated.  As a result the jury often foundered.... Success in [the role of the foreperson] rested on the extent to which the foreperson was able to bring some coherent structure to [the jury’s] discussions” (p. 96).   Compromises by jurors “to produce guilty verdicts on some charges and not guilty verdicts on other charges” (p. 97) occurred in five of the 48 cases.  At the same time, “Jurors were, with few exceptions, highly conscientious, took the role very seriously, and were extremely concerned to ensure that they did the right thing” (p. 97).  “Despite the fact that jurors generally found the judge’s instructions... clear and helpful..., there were widespread misunderstandings about aspects of the law in 35 of the 48 trials which persisted through to, and significantly influenced, jury deliberations” (p. 98).   However, “by and large these errors were addressed by the collective deliberations of the jury and did not influence the verdict of the majority of the cases” (p. 98).  It appeared, however, that in 4 of the 48 cases, the verdict was affected by misunderstandings of their legal instructions.   

Judges were asked, before the jury returned, what their verdicts would have been.  The most notable disagreement relates to 3 of the 48 trials where the disagreement was complete: in 2 cases the jury acquitted and the judge would have convicted, and in one case the jury convicted and the judge would have acquitted. 

Conclusion:  The quality of jury deliberations, and perhaps verdicts, could be improved if the difficulties facing the jurors -- difficulties which trial judges typically do not have -- were addressed.  “If [the necessary tools] were provided... most of the problems identified here would be overcome or substantially mitigated.”  It appears that legal change is, for the most part, not required.  A change in mindset is, however.

Reference:  Young, Warren, Yvette Tinsley, and Neil Cameron.  The effectiveness and efficiency of jury decision-making.  Criminal Law Journal, 2000, 24, 89-100.

Back to top

Publication Type

Volume Number


Issue Number