View this issue as a PDF: 
CrimHighlightsV2N6.pdf
Three factors have apparently contributed to a decline in intimate partner homicides: fewer long-term intimate relationships, improved economic status of women, and a growth in domestic violence resources. But the benefits flow largely to males: fewer women kill their male partners.
The Context. Killings that involve intimate partners have been declining in the United States during the past several decades. It has been well established that the death of an intimate at the hands of a partner is commonly the culmination of ongoing violence in a relationship. Thus, research suggests that factors that facilitate exit from a violent relationship or inhibit the development of such relationships should reduce the rate of intimate partner homicide by a simple mechanism – the reduction of exposure to a violent partner. Research on domestic violence research has also demonstrated that victim precipitation is common in cases where women kill their intimate partners, whereas it is unusual when men kill their female partners. Moreover, intimate violence is far more likely to result in serious bodily injury of the woman than the man. These findings are crucial to understanding the effects of exposure reduction factors on male and female intimate partner homicide victimization.
This study examined the contribution of three exposure reduction factors (i.e. decline in long-term intimate relationships, the improved economic status of women, growth in domestic violence resources) on the decline in intimate partner homicide rates. The authors hypothesized that exposure reduction would have a larger impact on the rate at which women kill their partners than on the rate at which men kill their partners. This is a result of women’s greater risk of serious injury from domestic violence and, in turn, their more active search for ways to reduce or escape this violence. The analysis was based on a panel data set of 29 large U.S. cities for four periods, spanning 1976-1992. Various exposure reduction factors were examined for their effects on rates of male and female intimate partner homicides. Measures of domestic violence services included the presence of hotlines, counseling services, legal advocacy programs and shelter availability. The rate of marriage and rate of divorce for each city and time period were included to capture declines in long-term intimate relationships. Finally, two measures of the economic status of women were included: relative educational attainment and sex differences in earnings.
The Results. Each of the three factors contributed to the reduction in the intimate partner homicide rate, most prominently to the rate at which women killed their male partners. Domesticity, as measured by marriage and divorce rates, was found to be related to the rate of spousal homicide: the greater the marriage rate, the higher the rates of both husband- and wife-perpetrated homicide whereas a high divorce rate reduced the rate at which wives killed their husbands. As expected, the declining marital rate was also associated with an increase in the rate of unmarried males killing their partners. The analysis also showed that the improved educational status of women was associated with a reduction in the rate at which females killed their unmarried partners. Finally, there was a statistically stable and negative impact of hotlines and legal services on the rate at which wives killed their husbands.
Conclusions. From a policy perspective, these results suggest that an important criterion for evaluating policies meant to prevent violence among intimates is the extent to which they reduce exposure in violent relationships. However, it is important to recognize that these exposure reduction factors vary for males and females. The benefits of exposure reduction in this study arose primarily from averting male victimization with little evidence found for exposure reduction effects on the female victimization rate. Other approaches must be found for reducing male-perpetrated intimate homicide.
Reference: Dugan, L., Nagin, D.S., & Rosenfeld, R. (1999). Explaining the decline in intimate partner homicide: The effects of changing domesticity, women’s status, and domestic violence resources. Homicide Studies, 3 (3), 187-214.
Why does the U.S. incarcerate so many people? It is the unique combination of “American moralism” and the structure of American government that appears to be the most plausible explanation.
Background. U.S. imprisonment rates (668 per 100 thousand in mid-1998) are unprecedented in American history or in other western democracies. The rates, of course, are consistent with harsh U.S. policies such as the death penalty, life-without-the-possibility-of-parole laws, extensive mandatory sentencing laws, and strict guidelines. The U.S. has not always been like this: in the 1930s U.S. incarceration rates were comparable to those of many European countries.
Examining the explanations. Two explanations that have been offered which do not, on their own, account adequately for this phenomenon:
High U.S. crime rates. For the most common crimes (property crimes) U.S. rates do not appear to be higher than those of most major western countries. Where the U.S. stands out is in firearms violence, but these, numerically, cannot account for prison populations. Crime trends -- increasing recorded violence rates beginning in the early 1960s -- parallel increasing prison populations in the U.S. but not in other countries (e.g., Finland or Germany (p. 423-4)). Unlike other countries, American politicians chose deterrence and incapacitation as strategies to address the crime rate, but these strategies “were adopted long after crime rates began to fall” (p. 422).
The U.S. public wants harsher penalties. The research in the U.S. mirrors that in Canada: “ordinary people have the same complex and ambivalent attitudes to criminals that judges and lawyers do...”(p.424). More importantly, “public support for harsh laws typically follow[s], not precede[s], media and political preoccupation with crime” (p.425).
Partisan politics has moved “crime” from the status of a difficult, but unglamorous, public problem to the centre of partisan politics: crime policy (along with welfare and affirmative action policies) became useful issues for defining differences between parties. Crime policy became crude sound-bite slogans presented in “simplistic and emotional terms” (p.427). As a consequence, crime policy became a metaphor “for broad concepts like personal responsibility...” (p.428). The “right” won (and the left has tried to recapture the territory by “out-righting the right”).
The federal government’s role has expanded directly and indirectly into social policy arenas and many of these new roles have not constituted conspicuous successes. Confidence in government decreased, allegiances were formed on the basis of single political issues, and, as a result, “policy initiatives that command the broadest support -- such as harsh crime, welfare, and immigration policies -- are understandably important to elected politicians.... Harsh policies on crime and welfare can be debated in moral terms, respond to broad-based anxieties and empathies, and affront no powerful constituency” (p. 430).
Historical trends in crime rates in many western countries rise and fall independent of crime control policies. Presently, they are falling. However, harsh policies (e.g., drug policies) appear after the problem (e.g., drug use) has started declining. This is understandable, in part, because decline in drug use may follow acceptance of the destructive or “moral” concerns about drugs (or crime). Few political leaders are willing to argue against policies aimed at reducing immorality. Similarly, now that crime is seen as a moral (individual responsibility) issue rather than an issue of social disadvantage, few will speak out against punishing the immoral. In addition, whether the policies “work” or not is unimportant: harsh approaches (to crime, welfare, immigration, etc.) are favoured by politicians and (subsequently) the public because these policies define and locate the problem in a particular personal and moral space.
Reference: Tonry, Michael. Why are U.S. incarceration rates so high? Crime and Delinquency, 1999, 45¸ 419-437.
Three strikes laws have had no impact on crime levels. More surprisingly, they have generally had little impact on the criminal justice system largely because they represented, to a great extent, more symbolic than real changes. Even in California, projected impacts were much less than expected. The major impact was predictable: prosecutors have increased their control over the criminal justice process.
Background. Three-strikes legislation has taken America by storm. Even Canada has its own “little-3-strikes” provision in the proposed Youth Criminal Justice Act -- three strikes and you’re presumptively an adult (for sentencing purposes). The U.S. laws vary considerably (p. 134-6) on what constitutes the first two strikes, and what offences can be considered a third strike. Finally, the consequences of being “out” vary across states. Critics suggested that the legislation would have an enormous impact on the criminal justice system; supporters suggested that the legislation would reduce crime. Neither occurred.
The general finding in most states was that the law had little impact for a simple predictable reason: “the vast majority of the targeted offender population was already serving long prison terms for these types of crimes... The three strikes law movement is much ado about nothing and is having virtually no impact on current sentencing practices” (p. 142).
California is the exception. California is unique in the baseball justice world because its “third strike” can be any felony. It also created an unusually harsh second strike provision and courts have decided, among other things, that certain juvenile adjudications can count as strikes (p.144). Although the law explicitly prohibits plea bargaining, it allows the prosecutor to “discount a prior conviction...” if the prosecutor believes that a baseball sentence would not be “in the furtherance of justice” (p. 143). Nevertheless, there have been some interesting 3-strikes sentences including:
27 years to life (to be served in prison) for attempting to sell stolen batteries (value $90).
Minimum of 5 years (to be served) for selling $5 worth of marijuana.
25 years to life (to be served in prison) for, after failing to stop at a stop sign, failing to stop when the police tried to stop the offender. A chase occurred but there was no accident and no injuries except to the car’s tires which were shot out by the police.
The law had an impact initially on the number of preliminary hearings, though this increase did not last long. Trial rates for second and third strike cases are dramatically higher (4% of non-strike felony cases go to trial compared to 9% for second strike and 41% for 3rd. strike cases) but there are few 2nd. and 3rd. strike cases. Counties varied dramatically on how strictly they implemented baseball sentencing. Some (e.g., San Francisco) discounted baseball penalties for those charged with non-violent offences. Generally speaking, crime went down in California as elsewhere in the U.S. and the decrease was uncorrelated with the zeal in implementing baseball justice. The impact on prison populations was much less than initially projected, though there were about 10,000 baseball sentenced offenders admitted to prison each year, and the prison population increased between spring ‘94 (the first inning) and spring ‘98 by almost 30,000 (about 27%). Because of the uneven implementation, however, this was dramatically less than had been projected. Judges, it seemed, also tended to undercut the impact of baseball sentencing (p.152).
Crime, of course, was unaffected by the changes in the law. Those states with and without baseball sentencing and those counties in California which embraced or did not embrace baseball sentencing are indistinguishable when looking at crime rates.
Conclusion: Baseball sentencing rules have symbolic impact but have less impact than originally thought on criminal justice processing. They have no impact on crime. They can, however, “enlarge the discretionary powers -- and hence sentencing powers -- of the prosecutor at the expense of the judge” (p. 158).
Reference: Austin, James, John Clark, Patricia Hardyman, and D. Alan Henry. The impact of ‘three strikes and you’re out’. Punishment and Society, 1999, 1, 131-162.
The method of supervising drug abusers on parole can make a difference: Social support programs which addressed the needs of the parolee combined with drug testing created better results than routine parole (with or without frequent drug testing).
Background. Drug abusers are often seen as being poor candidates for community corrections programs because of fears that their addictions will lead them (back) into criminal behaviour. Given that they do eventually return to the community, the question that should be asked is what can be done to maximize the likelihood that they will live a peaceful and drug free life?
This paper reports a study in which offenders with a history of narcotic addiction and/or heavy cocaine use were randomly assigned to one of three parole supervision conditions: routine parole (where urine testing for drugs occurred infrequently), routine parole with weekly drug-use monitoring, and intensive social support combined with weekly drug-use monitoring. This involved, among other things, identifying the needs of the parolees and helping to link parolees with available services in the community.
The results -- on such measures as parole violations, re-arrests, re-convictions -- were clear: social support along with drug testing was superior to drug testing alone (p. 173). Furthermore, within the “routine parole” condition, some parolees obviously received treatment. The outcomes for this group were more favourable than for those who received no treatment during their parole.
When one looks at “drug use” as an outcome measure, however, the results are somewhat different since the groups were differentially tested. Nevertheless, the results are clear: the more one looks for drugs the more one finds. Hence “combining monitoring with counseling can be problematic.... If... positive results lead to termination of counseling and/or incarceration... urine monitoring can undermine the objective of maintaining clients in treatment as long as possible” (p. 176).
Conclusion. Measures aimed at addressing the needs of parolees with a history of substance abuse clearly can be effective in reducing the likelihood of offending. More generally, however, this study demonstrates that how one supervises those on release from prison can be just as important as who is being released and supervised. A releasing authority, such as a parole board, truly interested in “managing” risk would, therefore, examine carefully what it can do to help reintegrate a prisoner into society. It would not limit its attention, when considering release, solely to the characteristics of the prisoner before it.
Reference: Hanlon, Thomas E., David N. Nurco, Richard W. Bateman, and Kevin E. O’Grady. The relative effects of three approaches to the parole supervision of narcotic addicts and cocaine abusers. The Prison Journal, 1999, 79, 163-181.
Doing crime or having contact with the criminal justice system as a young person has long-term harmful effects. Above and beyond any background factors (SES, family structure, race, cognitive skills, etc.) “delinquency in adolescence… reduces educational and occupational attainment in young adulthood especially among males” (p. 269).
Background. What happens to delinquents when they grow up? Most young people commit offences and a fair number of them become caught up in the youth justice system. When they stop committing offences, does their “delinquent background” hurt them in other ways? Most research that looks at school or occupational attainment has examined individual, family, cultural, or attitudinal variation in young people as predictors of who succeeds and who does not.
This paper looks at delinquent experiences as a possible impediment to later success. Specifically, using an American longitudinal sample of youth who were 14-17 in 1979, it examines educational and employment outcomes for those interviewed in 1990-92 when they were 25-30 years old. A number of variables – family socio-economic status, family size and structure, race, a cognitive skills measure (at age 14-17), and educational expectations when an adolescent – were controlled for statistically. Delinquency at age 14-17 was measured by self-reported drug use, involvement in property crime or violence, or contact with the criminal justice system.
For males, controlling for background and educational measures (cognitive skills and educational expectations), being delinquent in adolescence was associated with a lower likelihood of graduating from high school or getting a university degree as well as lower occupational status and a higher likelihood of being unemployed at age 25-30.
For females, it appeared that being delinquent as a youth had little direct effect on occupational achievement at age 25-30, but instead hindered educational attainment. Being delinquent as a youth had no impact on female unemployment at age 25-30.
Conclusion: “Delinquency in adolescence… reduces educational and occupational attainment in young adulthood, especially among males… for ‘ordinary kids’ as well as their more disadvantaged counterparts” (p. 269). These findings “testify to the importance of avoiding trouble when young: early deviance, both directly and indirectly has lingering effects that negatively influence life chances…. [The findings] cast doubt on the notion of benign or innocuous wrong-doings” (p. 269). Of course, “responses to delinquency may have an effect on life course trajectories…. Research indicates that adults who are officially identified and processed as criminals find it harder than other job seekers to secure employment” (p. 270). Hence, although we do not know exactly why deviance has the impact it does, these findings suggest that the impact of “doing crime” or being caught for it can have long term consequences.
Reference: Tanner, Julian, Scott Davies, and Bill O’Grady. Whatever happened to yesterday’s rebels? Longitudinal effects of youth delinquency on educational employment. Social Problems, 1999, 46, 250-274.
Young people are blamed for more crime than they really do. One reason is that young people are more likely to be caught.
Background. Each year in the U.S. (and to a lesser extent in Canada) when crime statistics are released, people look at the proportion of those arrested who are juveniles. When the proportion of arrested people who are juveniles is high, the assumption is that juveniles are responsible for a high proportion of crime. For example, in the U.S. when the 1997 FBI statistics were released, the fact that 30% of those arrested for robbery were juveniles was interpreted as meaning that 30% of the robberies were committed by juveniles. This ignores the fact that fewer than 20% of robberies are “cleared”, and, therefore, in most robberies there are no arrests.
This paper demonstrates that inferences about who commits crime based on who is arrested for it are likely to be wrong. There are reasons to believe that juveniles are more likely to be caught than adults: they are less experienced and they are more likely than adults to commit offences in groups. Law enforcement personnel may also be more motivated to locate and arrest juveniles.
This study examined robbery data from 7 American states, and used “incident based” data where victims’ perceptions of the age of the offenders were recorded. Compared to incidents apparently involving adult offenders, those involving juvenile offenders were:
- more likely committed by more than one offender,
- more likely to take place outside rather than inside a building,
- less likely to use a weapon such as a gun, club, or knife,
- less likely to have an adult victim, and
- more likely to result in the offender being arrested.
A more sophisticated analysis showed that “controlling for other incident characteristics, these data find that juvenile robbery offenders are 32% more likely to be arrested than are adult robbery offenders” (p. 157). In addition, the presence of a weapon increased a juvenile robber’s probability of arrest, but not that of an adult. “This is consistent with the national concern surrounding kids and guns…. [However] the relative seriousness of the offence is less of an issue when handling juvenile offenders. If these biases reflect the attitudes of the public at large, not only are juveniles more likely than adults to be arrested for similar crimes, but juvenile crimes may be reported to law enforcement [agencies] at a higher rate. This would add to the distortion of the juvenile crime component of crime that flows from law enforcement data” (p. 160).
Conclusion: “Based on these analyses, both juvenile arrest and clearance statistics over-represent the juvenile responsibility for robbery in the United States…. This bias may have increased in recent years as a result of the intense national focus on juvenile violence and the related increase in law enforcement attention to juvenile violent crime” (p. 160). More generally, the study demonstrates how cautious one has to be in interpreting reports of crime or arrests as they are contained in official records of crime.
Reference: Snyder, Howard. The overrepresentation of juvenile crime proportions in robbery clearance statistics. Journal of Quantitative Criminology, 1999, 15, 151-161.
Why do demographers predict that there will soon be a demographically driven crime wave? Two reasons: they don’t look at the data, and they don’t understand crime. Data do not support the notion that crime rates will increase because of changes in the age distribution.
Background. Demographic determinists in Canada and the U.S. have suggested that because of an increase in the number of children just entering the adolescent years, “the next crime wave will get so bad that it will make 1995 look like the good old days” (p. 582). The problem with such predictions is that they are based on an assessment of a single variable: the number of people of crime prone years. They do not take into account two things: projected increases in the size of the overall population (thus not taking into account the size of the denominator in calculating crime rates) and they do not take into account age specific crime rates and changes in these rates.
This paper attempts to divide “changes in crime rates into two parts: that which is attributable to shifts in the age structure and that due to all other sources of variation” (p. 585). When this is done, one starts with, for example, the fact that U.S. murder rates rose, from 1960-1980 from 5.08 to 10.22 per 100,000 in the population. This was the period, of course, when the baby-boomers aged into the crime prone years. However, only about 20% of this increase is attributable to changes in the age structure. Looking at age-specific rates, one finds that if age-specific crime rates had stayed the same, the U.S. murder rate would have increased during these 20 years only from 5.08 to 6.08 per 100,000 in the population.
The results of this analysis demonstrate that “changes in the age structure through 2010 will work very weakly in favour of lower crime in stark contrast to the frequently made assertion that the United States is in the beginning stages of a demographically driven crime wave” (p. 592). The experience in recent years of dramatic changes in the age-specific crime rates of certain groups suggests that it is these changes that are much more important in understanding changes in overall crime rates. Social programs which have an impact on the level of involvement in crime of those entering the crime-prone years are, obviously, much more important than the shifts in the age structure of our society.
Reference: Levitt, Steven D. The limited role of changing age structure in explaining aggregate crime rates. Criminology, 1999, 37 (3), 581-597.
People are generally not good at determining whether suspects are lying when they are being interrogated about crimes. More serious is the fact that formal training in detecting liars makes people less accurate.
Context. Statements made by accused people during police interrogations are obviously very important determinants of what happens to them and how a police investigation proceeds. Police are often trained not only in interrogation techniques, but also in how to detect whether a suspect is telling the truth. Surprisingly, there are few studies evaluating the usefulness of such training.
This study took 8 undergraduates and got them to agree to “commit” one of four “crimes.” For all practical purposes, these were real crimes: the only difference was that all of those involved knew that it was part of an experiment. Eight other undergraduates did similar, but innocent, acts. Interrogations were then carried out and these interrogations were videotaped. These tapes -- of “offenders” and innocent people -- were then shown to other people. Half of these other “observers” had been given training in “the Reid technique” of detecting deception (from verbal and non-verbal cues) and half had not.
The results were quite straightforward. Thirteen of the 16 suspects in the study assumed that the interrogator would be able to judge them accurately as being “guilty” or “innocent” as the case may be. The overall level of accuracy was at chance level: flipping a coin would have been just as effective in determining whether a “suspect” was “guilty” or “innocent.” However, the untrained observers were slightly, but statistically significantly, more accurate (56%) than the observers who had received one of the standard police training packages. Nonetheless, although they were more often wrong than not, the “trained” observers were able to cite more reasons for their decisions. These findings are not unusual: other studies as well have shown that people are not good at detecting liars and that training does not significantly improve their ability. Part of the problem in the real setting is that “research shows that police officers hold the same erroneous beliefs as laypersons do about the verbal and nonverbal indicators of truth and deception…. [and that] even professionals who regularly make such judgements for a living, such as detectives…, are highly prone to error” (p. 511). “In this vein, it is important to keep in mind that police and other experienced law enforcement professionals have not outperformed novices in related past studies” (p.513).
Conclusion. The problems inherent in any interrogation are increased dramatically because interrogators are not good at determining whether suspects are telling the truth. This paper demonstrates that training does not help interrogators determine whether the truth was told. Indeed such training may provide the interrogator with convincing “reasons” for believing or disbelieving a suspect unrelated to whether the suspect is telling the truth. In short, this paper reminds us of how easy it is for erroneous judgements to have devastating impacts on the way in which a police investigation proceeds.
Reference: Kassin, Saul M. and Christina T. Fong. “I’m innocent!”: Effects of training on judgments of truth a deception in the interrogation room. Law and Human Behavior, 1999, 25, 499-516.
Publication Type
- Regular Issue