View this issue as a PDF: 
CrimHighlightsV4N2.pdf
Do good social programs reduce crime? Clearly, the answer is yes. A randomized trial study of the impact of a nurse home visitation program shows that crime can be substantially reduced by providing nursing support to mothers during pregnancy and for the first two years of a child’s life.
Background. It is well known that early experiences in a child’s life affect the likelihood that he/she will be involved in crime. Thus, factors such as low birth weight (especially when combined with a disadvantaged childhood), harsh parenting styles (or practices of rejection) and association with deviant peers will increase the likelihood of offending in adolescence and adulthood. The policy question which needs to be addressed is a simple one: how can communities intervene, in positive ways, to decrease the likelihood of later offending?
This paper uses the “gold standard” for attributing causality – the randomized trial experiment – to determine the effects of broad-based intervention in a child’s life – home visits by a nurse before and after the birth of a child – on offending behaviour during adolescence. Mothers in their first completed pregnancy who were “at risk” (i.e.: young, single and/or of low socio-economic status) participated in the experiment. These women were assigned to different groups on a randomized basis. As such, the groups can be considered to be equivalent for all practical purposes. For some of the mothers (the control groups), the program simply provided assessment and referrals for treatment. For one “experimental” group, they received this same assessment and referrals, but a nurse also visited them an average of 9 times during pregnancy. The nurse promoted positive health-related behaviours during pregnancy and the early years of the child as well as general help to the mother (e.g., family planning, getting a job, etc.) during these visits. For the second experimental group, this monthly support visitation program continued until the child was two years old.
The results are simple to summarize. The nurse visitation program, especially when the monthly visits continued until the child’s second birthday, reduced the incidence of involvement with the police, arrests, and contact with the child welfare system as a “person in need of supervision” during the child’s early adolescent years (up to age 15) (p. 1242). “Adolescents born to nurse-visited women who were unmarried and from low-SES families had fewer episodes of running away from home, arrests, convictions and violations of probation than did their counterparts in the comparison group. They also had fewer sexual partners and engaged in cigarette smoking and alcohol consumption less frequently” (pp. 1241-2).
Conclusion: Social programs designed to promote healthy children can reduce crime. If a community wants to be tough on crime and simultaneously promote good health in children, it can do so by providing public health services to mothers during their pregnancy and to mothers and children in their first few years of life.
Reference: Olds, David, Charles R. Henderson Jr., Robert Cole, John Eckenrode, Harriet Kitzman, Dennis Luckey, Lisa Pettitt, Kimberly Sidora, Pamela Morris, and Jane Powers. Long-term Effects of Nurse Home Visitation on Children’s Criminal and Antisocial Behaviour. Journal of the American Medical Association, 1998, 280, 1238-1244.
When the state delegates the administration of punishment to private prison companies, it may, in fact, be unwittingly delegating the allocation of punishment as well.
Background. Generally speaking, our society assumes that the state has a central role in the allocation of punishment for wrongdoing. The traditional notion is that the prison system – public or private – administers the punishment but does not decide how much of it to apply. This presupposes that incarceration can be categorized “as a simple monolithic process” (p. 81). When prisons are run by the state, there are still concerns about how prison officials determine the amount of punishment that a prisoner will experience. However, when the prison is run by a private company, these concerns cannot be ignored. If the administration of punishment cannot be completely separated from its allocation, “how far should the judicial ‘power to punish’ be privatized?” (p. 80).
This study examined private prisons in Australia and began by noting that prison systems exercise discretion over prisoners (p. 84). As the author notes, there are “routine, quasi-judicial decisions that affect the legal status and well-being of the inmates” (p.84). For example, prison systems, by their very nature, are often called upon “to use force – perhaps serious or deadly force – against a prisoner” (p. 84). Further, prisons make recommendations to parole boards, they decide on the use of solitary confinement, and they bring charges against prisoners for institutional infractions. These clearly involve the allocation of punishment, not merely the administration of it. The concern is understandable: what if the private company is, in effect, in a conflict of interest? There are obvious possibilities of pecuniary interests being relevant to decisions about prisoners (for instance, where a prison is paid according to its in-custody count or the situation in which a particular prisoner is expensive to house (e.g., for medical reasons)).
In Australia, where disciplinary hearings are held within the (public or private) prison, it is sometimes exceedingly difficult to know whether a conflict of interest exists. Does a prison official feel that a particular outcome would benefit the prison (private company)? “There is no such implied suspicion of bias in the operation of a public correctional centre (unless evidence establishes the contrary) because public correctional officers are assumed not to hold a private commercial interest that conflicts with their disciplinary role…. Impartiality is crucial in the process of deliberations… in the corrections system” (p. 86).
Clearly, these concerns must be addressed. Whether they can be overcome is yet to be seen. Institutional transfer hearings (e.g., where a troublesome inmate is seen as being inappropriate for a particular institution) are an obvious example of conflict. A private prison manifestly has an interest in ensuring that certain prisoners are given security ratings which would require them to be housed in their institution – or elsewhere. If higher levels of security imply higher levels of punishment (e.g., fewer privileges, more difficulty getting released) then security decisions (on which institutions have a great impact) become questions of the “allocation” or punishment and not simply its administration.
Conclusion: The notion that private prisons simply administer punishment but do not allocate it does not, in practice, fit the facts. Too many “administrative” decisions, in fact, involve the nature or amount of punishment that a prisoner will experience. As jurisdictions move into the private prison world, these questions will need to be addressed.
Reference: Moyle, Paul. Separating the Allocation of Punishment from its Administration. British Journal of Criminology, 2001, 41, 77-100.
Provinces in Canada and states in the United States which have the highest degree of economic inequality also have the highest homicide rates. Much of the divergence in the homicide rates of these two countries appears to be attributable to differences in their levels of economic inequality.
Background. A number of different theoretical approaches suggest that “inequitable access to goods provokes antisocial behaviour and violence” (p. 220). For example, the finding that economic inequality appears to be more important than other variables, including poverty rates, as a predictor of homicide supports this hypothesis. However, it is difficult to separate poverty and economic inequality in the United States since low average income is associated with high income inequality. In contrast, those provinces in Canada with the lowest average income (the Atlantic provinces) also tend to have the lowest levels of economic inequality. The reason for this difference relates to Canada’s social programs. Employment insurance substantially reduces financial inequality in the poorest (i.e., Atlantic) provinces while having limited impact in Quebec and little, if any, effect west of Quebec (p. 232). Furthermore, those provinces with the highest level of economic inequality in Canada have the highest homicide rates. This study typically describes economic inequality in terms of income after tax and after social transfers.
Results. Perhaps the single most important finding is that when household income is statistically controlled for, high levels of economic inequality are associated with high homicide rates. Further support for the hypothesis that homicide rates are related to economic inequality comes from various analyses (pp. 228-231) which demonstrate the strong, positive association between economic inequality and variation in homicide rates within provinces across time. Median household income did not produce this same relationship. Finally, when one looks at the U.S. and Canada simultaneously, it becomes clear that not only do Canadian provinces generally have lower homicide rates, but they also have lower levels of inequality than do American states. In fact, “[t]he results suggest that [state and provincial level] income inequality may explain a substantial proportion of the large and much-discussed difference between these neighbours’ national homicide rates” (pp. 230-231).
Conclusion. In the U.S., violence appears, at first blush, to be associated with poverty. These data suggest that poverty is not as important as the level of economic inequality in a community in understanding homicide rates. In Canada, high levels of economic inequality are also associated with relatively higher homicide rates. Further, as economic inequality rises over time, homicide also increases. Conversely, when economic inequality drops, homicide rates decline as well. Given that economic inequality is partially under the control of social policies (e.g., employment insurance, social assistance), these results suggest that government social (and economic transfer) policies may be important tools in addressing levels of violence in our society.
Reference: Daly, Martin, Margo Wilson, and Shawn Vasdev. Income Inequality and Homicide Rates in Canada and the United States. Canadian Journal of Criminology, 2001, 219-236.
Schools can affect the likelihood that students in them commit criminal offences. This impact is evident above and beyond the effects of the type of children attending the schools. Well-run schools can reduce crime.
Background. It is difficult to examine the effects of schools on crime because youths tend to go to local schools. Schools in dysfunctional communities may appear to be associated with crime, but this association may have more to do with the neighbourhood or the characteristics of the youths attending the schools than with the school itself. This paper looks at the effects of different types of schools while attempting to control for neighbourhood and student differences. In other words, it asks whether the manner in which a school is organized and run affects crime. Other research has shown that neighbourhoods have impacts on crime above and beyond the characteristics of the students themselves.
This synthesis of the literature suggests the following conclusions (controlling for community and student characteristics):
- Teacher victimization appears to be associated with larger schools (for middle schools) and larger numbers of students taught by the average teacher (for high schools), punitive teacher attitudes, more ambiguous or punitive rule enforcement, and “democratic attitudes of teachers” (the view that students and parents should have a say in the way in which schools are run) (p. 71).
- Student victimization is associated with perceived unfairness and lack of clarity of school rules.
These factors clearly constitute effects of the ways in which a school is run. Good working conditions, clarity of academic and behavioural requirements, good classroom management, and an administration in which teachers feel their views are heard contributed to lower delinquency rates of the students.
Conclusion: School characteristics have a moderately large impact on crime by students. Administrative and management practices, clear communication and goal setting, fair procedures for students and teachers, as well as consistent, although not punitive, enforcement of rules contribute to a reduction in student offending. “The research implies that principals and teachers control behaviour by setting rules, communicating clear expectations for behaviour, consistently enforcing rules, and providing rewards for rule compliance and punishments for rule infractions…. By maximizing student learning and engagement, schools increase commitment to education and attachment to school. By modeling appropriate behaviour and establishing a fair and just discipline system, school staff enhance student beliefs in the validity of rules and laws” (pp. 90-91). Put simply, well run schools have impacts far beyond their immediate educational goals.
Reference: Gottfredson, Denise C. “Chapter 3: School Effects” in Schools and Delinquency. Cambridge University Press, 2001.
It is not the experience of being brought up by a sole parent that affects the likelihood of delinquency. Rather, that which happens after the disruption in the family is crucial: high conflict and being separated from the mother appear to be the most important variables.
Background. “The topic of broken homes has been a central part of delinquency theory since… the nineteenth century” (p. 22). Research using both official measures of delinquency and self reports have typically found correlations between broken homes (or disrupted families) and delinquency. The problem is that both of these terms (broken homes and disrupted families) “include many types of family experience” (p. 22). In particular, these terms ignore many pre-disruption (e.g., family conflict) and post-disruption (e.g., subsequent family reconstitution) effects.
This paper looks at the relationship between “family disruption” and delinquency, but characterizes the family in terms of whether the breakdown followed high levels of family conflict and whether the child, after the family disruption, was still subject to the stabilizing influence of the mother. The study uses the longitudinal data from the Cambridge study in delinquent development which followed 411 boys born in the mid-1950s for several decades. Both self-report and official data were examined. Interestingly, the results using these two types of measures were, for the most part, identical.
The results replicated previous studies in that delinquency was more prevalent among the boys from disrupted families. Independent of the family structure variable, however, family conflict – defined as “chronic tension or disagreement between parents in many fields” (p. 30) – also predicted delinquency. Yet, when one looked at conflictual families only, boys from those families which were subsequently “disrupted” were no more likely to be delinquent than boys from conflictual, but intact families. Similarly, families that were disrupted because of disharmony were more likely to be criminogenic than families disrupted by death. Further, children who were subsequently with their mothers after separation were also less likely to be delinquent. Instability (measured by the number of changes in the operative parents) also predicted delinquency.
Conclusion. It is clear that “family structure” is important as a predictor of delinquency. However, being from a disrupted family, per se, does not seem to be as important as the nature of the disruption. Highly conflictual, “in-tact” families appear to be just as criminogenic as highly conflictual, disrupted families. Thus, the cause of the disruption matters. Similarly, that which occurs after the disruption also matters. Being separated from the mother, or having instability in the subsequent family structure appears to increase delinquency. “Single parents” per se do not seem to be responsible for delinquency.
Reference: Juby, Heather and David P. Farrington. Disentangling the Link between Disrupted Families and Delinquency. British Journal of Criminology, 2001, 41, 22-40.
Why do judges not use draconian measures to deal with apparently dangerous offenders? Because these measures are in “conflict with fundamental principles and approaches of the common law… in particular proportionality, discretion, and natural justice” (p. 66).
Background. Many western countries, including Canada, have special, typically draconian, laws to deal with dangerous offenders. These include indefinite sentences, mandatory sentences and special principles of sentencing (e.g., incapacitation) for certain types of offenders. The question asked by this paper is a simple one: why are relatively few offenders sentenced under these laws? In Canada, for example, in the first 20 years of such legislation, only about 150 “dangerous offenders” on indefinite preventive detention out of a potential pool of approximately 14,000 prisoners had accumulated in Canada’s penitentiaries. Hence, the number admitted each year constitute only a tiny proportion of the roughly 4400 prisoners admitted to federal penitentiaries. “Though highly symbolic of the state’s power to act, and of considerable sociological significance, [laws focusing on dangerous and/or recidivist offenders are] in practice irrelevant, offering little effective protection to the public…. [In addition], throughout their histories, [these laws] were often strongly and continuously opposed by those charged with their implementation – the judiciary” (p. 52).
Judicial responses to these laws have typically been less than enthusiastic. Over the years, judges have found numerous ways of circumventing mandatory penalties. Few of those who qualify for special legislative provisions receive these penalties. Australian habitual offender legislation and sex offender provisions, for instance, have been imposed on only a tiny proportion of all those who might qualify (pp. 56-7).
It is probably inappropriate to imply that judges have consciously “resisted” the imposition of these provisions. The judicial actions which result in their non-use “have not been planned or plotted” (p. 59). “The bedrock of resistance to special laws is the principle of proportionality whose origins have been traced back to the Magna Carta, 1215…” (p. 59). In addition, “where the judiciary suspects or believes that legislation represents an attack on fundamental human rights, it will strictly construe the language of the statute in order to give it the narrowest interpretation consistent with the intention of the statute and the preservation of those human rights” (p. 61). Due process requirements and narrow interpretations of the meaning of “prior convictions” have also restricted the use of these laws. “Judicial discretion… is central to the self-concept of the judiciary… Courts… have generally been distrustful of unaccountable administrative authority with the result that attempts to restrict the amount of judicial discretion have been strongly resisted” (pp. 62-3). Finally, at least in Australia, the courts have been “reluctant to make too much of probabilistic data” (p.65). “The liberty of the subject under the common law is not to be set at hazard upon a statistical probability, nor curtailed in the expectation, no matter how well grounded, that an agent of the Executive Government or a Parole Board will choose to set him free before the law’s sentence has run its course” (p. 65).
Conclusion. When special provisions of criminal laws are in conflict with standard criminal law principles, the latter appear to generally win out in jurisdictions with common law traditions. Notwithstanding this fact, “the danger [of special draconian laws is] an emerging philosophy of despair” whereby these laws “are no longer justified on the basis of the gravity of the offence or even on the basis of the offender’s future conduct, but simply on the basis that the offender has forfeited his or her right to participate in society” (p. 67).
Reference: Freiberg, Arie. “Guerrillas in our Midst? Judicial Responses to Governing the Dangerous” in Mark Brown and John Pratt (eds.) Dangerous Offenders: Punishment and Social Order. New York and London: Routledge, 2000.
Criminal justice systems, as organizations, have a way of adapting to changes in the law in a manner that seems to maximize organizational needs. Consequently, the explicit goals intended by modifications in legislation can be undermined. Changes in sentencing laws will be ignored if their adoption represents serious organizational costs.
Background. There have always been concerns that attempts to stifle the use of discretion in one part of the criminal justice system (e.g., sentencing) would lead to its displacement to another decision point (e.g., charging practices). For example, it has been demonstrated that in several sentencing guideline systems which have forbidden the use of personal characteristics such as race and gender, the criminal justice organization has adapted to these restrictions. More specifically, these same factors are seen to re-appear elsewhere in the system as important determinants of the ultimate outcome. There are, in fact, three models of how the justice system might react to legal change: (a) a formal legal model which would predict simple compliance; (b) a “substantive rationality model” which suggests that the system would adhere to established norms, and (c) an “organizational efficiency model” which submits that the prime goal is to process cases quickly by way of guilty pleas.
The state of Washington introduced sentencing guidelines in 1984 which were based largely on the offence and the offender’s past history. In 1988, sentencing law changed such that first time drug dealers would, for the first time, presumptively go to prison for 12-14 months. In 1990, presumptive sentences for certain drug offences were dramatically increased. In 1992, a court determined that offences such as “conspiracy to deliver narcotics” would no longer be treated as a less serious version of the substantive offence, but could be subject to a non-prison sanction.
This study looked at sentencing data from 1985 to 1995. The pattern of results suggested that an “organizational efficiency” model fit the data best. Pleading guilty clearly led to less severe sentences, partly because “offenders who plead guilty [were] consistently convicted of less serious charges” (p. 1380). The first set of changes in the law (eliminating the automatic advantage of being a first time offender) had little overall impact, but the subsequent changes dramatically increased the difference in sentence between those who pleaded guilty and those who went to trial. Indeed, going to trial increasingly “cost” the convicted offender in terms of the final sentence. “The severity of charges at conviction changed significantly following each change in the law, which suggests the manipulation of charges (and subsequent sentences) rather than a strict application of charges to the crimes committed… The overall impact of [the changes in the law] is substantially greater for offenders convicted at trial than for those who pled guilty” (p. 1384).
Conclusion. The hypotheses that either changes in the law will be automatically followed or that punishment will be determined by established norms were contradicted by these findings. Indeed, the “adaptation” to the changes in the law occurred only for those who pleaded guilty. These results supported the “efficiency” model: “charging practices will change as options for inducing guilty pleas change. Most importantly those changes are entirely contingent upon offenders having pled guilty” (p. 1385).
Reference: Engen, Rodney L. and Sara Steen. The Power to Punish: Discretion and Sentencing Reform in the War on Drugs. American Journal of Sociology, 2000, 105(5), 1357-95.
The saying “garbage in, garbage out” needs to be applied to meta-analyses conducted in criminology. This meta-analysis of 35 separate studies shows that an apparent positive effect of family treatment on delinquency completely disappears when one examines only methodologically sound studies. Generally speaking, the more rigorous the study, the less likely it was to show a positive impact of family treatment on the subsequent recidivism rates of youths.
Background. Meta-analyses of research findings have been used in several areas of correctional research to demonstrate the apparently positive impact of certain kinds of correctional interventions. Essentially, the technique allows the researcher to combine, into one analysis, the findings from a large number of discrete studies in order to determine if a relationship exists between the intervention and some outcome (e.g., recidivism). If there is a relationship, a meta-analysis allows one to estimate its size.
This study is based on the underlying fact that the combination of the results of a large number of methodologically flawed studies (using meta-analyses) simply produces one large meta-analytically flawed result. It studies the impact on recidivism of interventions involving families in the treatment of young offenders. Thirty-five studies examining this question were found. They were graded in terms of their methodological rigor, whereby a point was given for each of four positive attributes of the study (e.g., random assignment, follow-up of at least one year). The simple results of the meta-analysis could be described in very impressive terms: the studies involved 17,141 young offenders and found a very significant positive treatment effect.
The size of the effect of treatment can be scaled such that zero signifies that there is no effect, a negative number means that the treatment increased recidivism and a positive number indicates that the treatment reduced recidivism. The overall treatment effect was 0.15. More important was the relationship between the apparent effect of the treatment and the methodological rigor of the findings. Studies that only had 1 of the four criteria of methodological rigor had a “treatment effect” of approximately 0.33 while those that met two of the four criteria had an effect size of about 0.18. For studies with three of the four criteria, the treatment effect was 0.12, and for those attributed all four points, it was approximately zero.
Conclusion. The findings were clear: “studies using poorer experimental designs displayed significantly better results than studies using stronger experimental designs” (p. 246). Thus, although a “standard” meta-analysis would have led one to conclude that family interventions had a meaningful impact on subsequent recidivism for young offenders, this study shows that this apparent effect is likely to be the result of methodological flaws. The lesson is a simple one: fancy or sophisticated statistical techniques cannot overcome basic methodological weaknesses. In other words, “garbage in, garbage out”.
Reference: Latimer, Jeff. A Meta-Analytic Examination of Youth Delinquency, Family Treatment, and Recidivism. Canadian Journal of Criminology, 2001, 237-253.
Publication Type
- Regular Issue