View this issue as a PDF: CrimHighlightsV5N2.pdf
This issue of Criminological Highlights addresses the following questions:
- Do private prisons operate more effectively than publicly run prisons?
- Does 3-strikes sentencing legislation promote homicides?
- Are prisons criminogenic?
- Why are people afraid to live in neighbourhoods with high proportions of black residents?
- Do high crime rates account for high imprisonment rates?
- Is there a “firearms” effect, independent of criminal intent, that accounts for harm done in violent interactions?
- Does court mandated treatment of male wife batterers work?
- Which kinds of police-suspect encounters result in police use of force?
Private prisons in the U.S. appear to be having significantly more than their share of staff turnover, escapes and drug use.
Background. About one in twenty state and federal prisoners in the U.S. are housed in private prisons. The U.S. federal government and several states (as well as some provinces in Canada) seem to favour private prisons because they apparently save money in the short term. The question, of course, is not just whether this type of correctional facility is cheaper, but whether it also meets or surpasses the standards of publicly run institutions.
This study reports the results of a survey of government officials responsible for monitoring private prisons. It focuses on three issues: staff, escapes, and drug use by inmates. This research emphasizes staffing issues largely because two other areas of the survey – training and custody standards –showed no differences with publicly operated prisons. This result reflects, to a great extent, the fact that contractors adopted the same standards as those followed by the jurisdictions with which they were contracting. 81% of the privately housed inmates covered by the survey are held in prisons operated by two companies: the Corrections Corporation of America and Wackenhut Corrections Corporation. As such, the comparisons drawn in this study are largely between these private prisons and those run by the Federal Bureau of Prisons (BOP). Unlike the private sector facilities, the BOP houses many inmates in minimum security prisons without a secure perimeter. As a result, these correctional institutions were excluded from the analysis as no equivalent group exists in the private sector. Unfortunately, some bias in favour of private prisons was introduced because this sector often houses these same minimum security inmates in secure facilities. In addition, it “is further advantaged in the comparisons because the private sector held much lower numbers of high- or maximum-security inmates than did the BOP” (p.434).
The findings demonstrate that “although the overall staff-to-inmate ratios are about the same for private prisons and the BOP, private companies emphasized custody staff” (p.437). Consequently, one might expect that whatever other differences may exist between the two systems, private prisons would be good at accomplishing “custody.” However, private prison companies in the U.S. are not very successful at retaining their staff. Using a “separation rate” to measure staff turnover (i.e. an index of the number of staff who left their jobs voluntarily or non-voluntarily during a six-month period over the number of people on staff), BOP prisons lost only 4.4% of their staff. In contrast, 64 of 67 privately run prisons had rates of 10% or higher. Indeed, 22 of the 67 private prisons had turnover rates of over 50% on this measure (p.439). These differences are as true for new private prisons as older ones.
Further, staff are not the only ones who appear eager to escape from private prisons. Only one of 68 BOP prisons experienced an escape (of one person) during 1999. In contrast, 12 of 68 private prisons experienced a total of 23 escapes (p.442). Homicide rates were comparable in the two sets of prisons. Drug use - measured by random urinanalysis tests – appeared to be higher in private prisons with 66% of the private prisons having at least one prisoner who tested positive for use of drugs compared to publicly run prisons where only 38% had one or more inmates who tested positive for drug use during the same period (p.440). 11 of 68 (16%) publicly operated prisons had inmate “drug use” rates of 3% or more. In contrast, a much higher proportion (41%) of the privately run prisons had inmate drug use rates of this level or higher.
Conclusion: “Privately operated prisons appear to have systemic problems in maintaining secure facilities” (p.444). The evidence shows that the private prison industry is not successful at retaining employees. As a consequence, inexperienced staff may be responsible for (some of) the problems which seem to exist in operating these institutions. If governments turned to private prisons as a cost-savings measure, it may be that they are getting what they paid for.
Reference: Camp, Scott D. and Gerald G. Gaes (2002). Growth and Quality of U.S. Private Prisons: Evidence from a National Survey. Criminology and Public Policy, 1, 427-450.
Can 3-strikes laws promote crime? Evidence has shown that this legislation does not reduce criminal activity through incapacitation or deterrence. This study suggests that these laws can actually promote killings.
Background. Since three strikes laws became a criminal justice fad in the early 1990s, the evidence has convincingly demonstrated that they do not reduce crime. Further, they can be a significant drain on public resources as large numbers of minor offenders are incarcerated (in some states) for long periods of time.
This study, starting from a “rational decision making” perspective, examined the possibility that offenders in 3-strikes states will attempt to avoid apprehension for serious offences by acting in a rational way. More specifically, it is argued that because the penalty for an offence like robbery is, in effect, the same as the penalty for homicide for many serious offenders, the “rational” criminal may attempt to avoid apprehension by killing victims, potential witnesses, or police officers.
Using data from 188 American cities - only some of which had three strikes laws - this study examined the potential homicide promoting effects of this legislation in the period before, during, and after these laws came into effect. Other variables known to relate to homicide rates (e.g., percent African-American, percent young, percent female headed households, percent living below the poverty line, etc.) as well as changes in other measures of violence in these cities were controlled for statistically.
The results were clear: “Homicide rates have grown faster (or declined at a slower rate) in three strikes cities compared with cities without the laws” (p.408). “Passage of a three-strikes law has increased homicides, on average, by 13% to 14% over the short term, and 16% to 24% over the long term” (p.409). Finally, “there is no evidence that increases in homicide rates promote state legislatures to enact three strikes laws” (p.412).
Conclusion: “Although policy makers anticipated that [3-stikes] laws would “fix the problem” of serious crime by deterring active criminals and incapacitating repeat offenders… the climate of fear and hysteria in which the statutes were passed actually increased the likelihood of failure or negative unintended consequences” (p.418). It is clear that “policy makers should take more care to weigh, not just the potential benefits of a proposed crime control solution, but the costs as well” (p.419). “Two studies have now found that three-strikes laws increase homicide rates” (p.419). However, what is not known with any certainty is whether this effect occurs because of the hypothesized mechanism of sophisticated offenders killing innocent people in attempts to avoid detection and prosecution. Indeed, several other plausible explanations (e.g., homicide as a defiant reaction against more severe sanctioning practices) would have to be ruled out before a ‘rational model’ explanation can be accepted.
Reference: Kovandzic, Tomislav V., John J. Sloan III, and Lynne M. Vieraitis (2002). Unintended Consequences of Politically Popular Sentencing Policy: The Homicide Promoting Effects of “Three Strikes” in U.S. Cities (1980-1999). Criminology and Public Policy, 1, 399-424.
The U.S. War on Drugs and other imprisonment programs appear to ensure a continued supply of criminals. Indeed, there is “compelling evidence that offenders who are sentenced to prison have higher rates of recidivism… than do offenders placed on probation” (p. 329).
Background. “Scholarly research generally concludes that increasing the severity of penalties will have little, if any, effect on crime” (p.330). Similarly, the increase of sanctions for drug use and distribution has little (if any) effect on drug consumption. However, like many of the sentencing changes that have taken place since 1990, the War on Drugs in the U.S. is based on a deterrence model. Though much of the focus on sentencing reform has been on general deterrence, there is also a literature suggesting that imprisonment has no measurable impact on the likelihood of a punished offender committing a subsequent offence. Custodial and non-custodial sentences appear to be equally effective (or ineffective) in their effects on recidivism.
This study looked at 342 drug offenders and 735 non-drug offenders (some of whom had a history of involvement with drugs) convicted in 1993. Approximately two thirds had been sentenced to probation while the others had gone to prison. Controlling for factors known to be related to recidivism (e.g., gender, race, employment, age, prior convictions as well as factors related to the likelihood of imprisonment in 1993), the study looked at recidivism over a four-year period. Various measures of recidivism (i.e. a new charge being filed, subsequent incarceration, “time to failure”) were examined.
The results showed that “offenders who were sentenced to prison were significantly more likely than offenders placed on probation [in 1993] to be arrested and charged with a new offence…, to be… sentenced to jail or prison for a new offence” (p.342) and to “fail” more quickly. These results held for drug offenders, those involved with drugs but not convicted of a drug offence, and those without drug involvement. In all cases, those sentenced to prison in 1993 were more likely to recidivate than those sentenced to probation.
Conclusion: The authors conclude that “[t]he results… provide no support for the deterrent effect of imprisonment. Despite the fact that we used several different measures of recidivism, tested for the effect of imprisonment on different types of offenders, included a control for the offender’s predicted probability of incarceration for the 1993 offence, and examined recidivism rates during a relatively long follow-up period [48 months], we found no evidence that imprisonment reduced the likelihood of recidivism. Instead, we found compelling evidence that offenders who were sentenced to prison had higher rates of recidivism and recidivated more quickly than offenders placed on probation” (p.350). “The findings of this study cast doubt on the assumptions underlying the crime control policies implemented during the past two decades… Policies pursued during the War on Drugs have been counterproductive” (p.352). That is, unless one is in a profession that profits from high crime rates or has investments in the prison industry.
Reference: Spohn, Cassia and David Holleran (2002). The Effect of Imprisonment on Recidivism Rates of Felony Offenders: A Focus on Drug Offenders. Criminology, 40, 329-357.
It is people, not crime rates, who account for white residents’ perceptions of crime. Studies in three American cities show that one’s perception of the level of crime is associated with the proportion of young black men in a neighbourhood, even after controlling for the amount of actual criminal activity.
Background. Fear of crime is an important determinant of people’s everyday lives as well as their views about the ways in which those who offend should be handled by the criminal justice system. This recognition may be important in shedding light on the tendency of whites to avoid living in neighbourhoods with high proportions of black residents. Indeed, this behaviour raises the question of whether part of this avoidance is due to the perception that black Americans are associated with crime.
This study examines the relationship between the racial composition of a neighbourhood and the perceptions of white residents of neighbourhood crime levels. Unlike neighbourhood crime rates, “[a] neighbourhood’s racial composition is a readily observable characteristic” (p.721). Furthermore, evidence suggests that “[t]he stereotype of blacks as criminals is widely known and is deeply embedded in the collective consciousness of Americans, irrespective of the level of prejudice or personal beliefs” (p.722). Data from surveys in three cities – Chicago, Seattle, and Baltimore – were examined in an attempt to understand the way(s) in which people infer their neighbourhood’s crime rate. The actual level of crime in the neighbourhood was controlled for by examining official statistics and, in two cities, victimization measures from the survey. The effects of other factors (e.g., income, the physical deterioration of the neighbourhood) were also removed. The study hypothesized that the proportion of young black men in the neighbourhood would be used by residents as an indicator of the crime rate. More specifically, high numbers of young black men would be interpreted as indicating a high level of crime.
The results in all three cities supported this hypothesis. In Chicago, for example, both the proportion of young black men and the crime rate as well as indicators of general disorder or incivilities (e.g., noise problems and insults among persons on the street) were predictors of the perception that crime was a problem (p.740). The results for Seattle were similar. Over and above crime rates and victimization experience, the percent of young black men predicted respondents’ perception of neighbourhood crime rates. In addition, individuals who reported numerous teenagers hanging out in the street were also more likely to report that their neighbourhood had a serious crime problem (p.742). Further, the Baltimore data showed that above and beyond crime rates, the percent of black residents as well as personal victimization had an impact on perceived levels of crime. There was some evidence in Seattle and Baltimore that these effects were stronger for white residents than for black residents (p.744).
Conclusion. It would appear that “whites are averse to black neighbours in part because certain neighbourhood problems, namely crime, are perceived to be worse in black neighbourhoods” (p.748). However, the results “contradict the assumption that this perception simply reflects actual differences in neighbourhood crime levels” (p.748). Thus, it seems that whites “systematically overestimate the extent to which perceptible black and neighbourhood crime rates are associated” (p.749). Indeed, it would seem that perceptions of crime levels are still very ‘black and white’.
Reference: Quillian, Lincoln and Devah Pager (2001). Black Neighbors, Higher Crime? The Role of Racial Stereotypes in Evaluations of Neighborhood Crime. American Journal of Sociology, 107, 717-767.
What accounts for the variation in imprisonment rates across time and place? One thing is clear: it isn’t crime rates.
Background. It is well known that imprisonment rates vary dramatically across western countries. The U.S. leads the pack (with a 1997-8 rate of 645 per 100,000 people, followed by New Zealand and Portugal (with 145) and England & Wales and Canada (with 125 and 115 respectively). At the opposite end of the continuum, one finds Denmark, Sweden, Greece, Finland and Norway (with 55-65). In addition, “[s]tunning dissimilarity in imprisonment trends between countries becomes apparent when longitudinal data are examined” (p.8). Yet, the criminal justice institutions “are much the same” (p.3).
Looking, for example, at the U.S., Finland and Germany, we find that in all three countries violent crime and, to some extent, homicide rates increased fairly consistently from the 1960s to the 1990s. However, the imprisonment trends are very different. Imprisonment in the U.S. began to rise dramatically in the mid-1970s (approximately 15 years after violent crime started increasing) and continued to swell through the 1990s. In Finland (see Criminological Highlights, Volume 3, Number 5, Item 1), as in Austria (p.270), imprisonment rates decreased dramatically throughout the same period. In Germany, imprisonment varied somewhat, but decreased during this time. Self-report victimization rates across western countries also do not show any obvious relationship with imprisonment rates.
One obvious difference between European countries, on the one hand, and the U.S. (and perhaps Canada) on the other, is that in Europe “crime rates and patterns [of crime] are not regarded as something easily controlled or necessarily much affected by punishment policies” (p.4). Another distinction between U.S. and other countries’ approaches to punishment is that the former tends to have rigid provisions ensuring that legislated punitivism is carried out in full. “Populist punitivism” in the U.S. - in contrast with other nations - is allowed to run unchecked (p.18). Further differences are rooted in the existence of U.S. guidelines (sometimes, as in the U.S. federal system, more mandatory than ‘guiding’) which were created, in some instances, to ensure high levels of imprisonment. Finally, diverse responses to prison overcrowding have contributed to variation in imprisonment rates. More specifically, the U.S. simply built or contracted for more prison space as a solution. In contrast, several European countries (e.g., Denmark, Norway, and the Netherlands) established “waiting lists” for prison spaces such that offenders sentenced to prison would serve their time only when a space became available.
It would seem that European countries have been less willing than the U.S. to commit huge amounts of public funds to the punishment of offenders. Further, most countries that have become more punitive in their responses to violent offenders have tended to “reconcile populist punitive pressures with budgetary and prison-capacity limitations” (p.263) by attempting to reduce the use of expensive prison sentences for less serious offences. Despite an apparent increase in sanctions directed at those who commit “quality of life” offences (p.263), there seems to be less enthusiasm recently for straight punishment in cases in which other approaches are available (e.g., the development of drug courts).
Conclusion. It should be remembered that until the 1960s, U.S. crime policy was not dominated by short-term political interests as it is now. In contrast, “Criminal justice issues in most European countries are still relatively non-politicized and unaffected by media-driven sensationalism: Judges and prosecutors are largely career civil servants… and criminal justice policy is dominated by professional, academic, and bureaucratic elites” (p.276). The difference between these policy processes may be important.
Reference: Tonry, Michael and Richard S. Frase (eds.) (2001). Sentencing and Sanctions in Western Countries. Oxford University Press. Tonry: Punishment Policies in Western Countries (pp. 3-28). Frase: Comparative Perspectives on Sentencing Policy and Research (pp. 259-292).
The presence of a firearm in a criminal incident appears to have a complex relationship with the outcome. Firearms have different impacts on the likelihood that a conflict will lead to an attack, the likelihood of injury, and the likelihood that, if injured, the injury will be a serious one. However, their effects do not seem to be dependent on the offender’s intent.
Background. Separating the effects of firearms from those of the offender’s intent on the outcome of a conflict is difficult. Two competing theories exist: 1) firearms are more likely to be used if the offender has an intent to harm and 2) the presence of a firearm has an impact on the amount of harm done, independent of intent.
This study analyzes “more than 2000 violent and potentially violent events described by [newly incarcerated male] offenders in order to assess the role of weapons and offender intentions in the outcomes of these events” (p.271). Respondents were asked about events involving assaults, as well as incidents in which “there was a high risk of violence but actual violence by or against the respondent was avoided” (p.277). The study examines the impact of a firearm or other weapon on the likelihood of an attack, the likelihood of an injury (if an attack took place) and the likelihood that the injury (if one existed) was severe. In addition, the researchers also recorded whether or not the respondent intended to do serious harm.
The results show that the impact of the presence of a firearm or other weapon depends, to some extent, on the stage of the conflict that one is examining. Given a situation which is potentially violent, the likelihood that an attack will take place appears to be associated with the possession of a firearm or another weapon. Not surprisingly, those offenders who, at the beginning of the incident, had the intent to commit physical harm were also more likely to do so. However, the effect of the possession of a firearm or other weapon was independent of intent. Incidents in which the offender intended harm and those in which he did not were each associated with higher likelihood of attack when a weapon was available. Not unexpectedly, the likelihood of an attack was reduced if the opponent had a firearm (p.283). When one looks at the likelihood that an attack will result in injury, it appears that incidents involving attacks with firearms were less likely to involve injury than incidents without weapons. In contrast, incidents involving other weapons were more likely to result in injury. Although incidents involving the intent to injure were more likely to involve injury, the effect of intent was once again independent of the effects of firearms or other weapons. Essentially what appears to be happening with firearms incidents is that the attack is often unsuccessful: the offender misses the target. An attack with another weapon or an attack without a weapon is more likely to be successful (pp. 284-285). Finally, if an injury does effectively occur, it is more likely to be serious if it involves a firearm, though other weapons also increase the likelihood of serious injury. However, these effects are again the same for those offenders who did and did not intend to harm their opponent (p.286).
Conclusion. Perhaps one of the most important findings of this study is that the hypothesis that “weapons effects would depend on an assailant’s intent to seriously injure his opponent was not supported by the data” (p.290). Firearms have somewhat different impacts at each stage of conflicts and do not depend on offenders’ intent. Thus, attacks with firearms appear to be associated with fewer injuries than attacks without firearms, but where an injury takes place, the injury is more likely to be serious. However, there is no evidence that highly motivated offenders will exert extra effort to compensate for less effective or no weaponry.
Reference: Wells, William and Julie Horney (2002). Weapon Effects and Individual Intent to Do Harm: Influences on the Escalation of Violence. Criminology, 40, 265-296.
Court-mandated counseling for domestic violence offenders does not appear to reduce domestic violence.
Background. Victims of wife assault often return to their assaultive spouses even after they have been found guilty of the offence. Consequently, courts are often eager to reduce the likelihood of subsequent victimization. One way in which they do this is to order men to submit to a “spouse abuse abatement program” (p.344). “The most popular of these programs is a feminist cognitive psychoeducational approach, called the Domestic Abuse Intervention Program. Referred to simply as the Duluth Model, it focuses on battering as part of a range of male behaviours used to control women. The curriculum is taught in group sessions that emphasize the modification and development of cognitive techniques that batterers can use to avoid conflict” (p.344). The evaluation of these programs is important since it appears that “the most influential predictor of an abused spouse’s return to her husband is his participation in counseling” (p.345). In other words, it would seem that women return to abusive spouses in part because they think that he has been or will be “cured” by treatment.
This study took all men who were convicted of “misdemeanor domestic violence” during a five-month period in one county in Florida and placed them on probation. In addition, half of these individuals were randomly assigned to attend 26 weeks of group sessions from a local batterers’ intervention group as part of their sentence. Most men who were required by the court to participate in the assigned sessions did, in fact, attend most of them. In contrast, few of those who were simply placed on probation attended a treatment program.
The results are easy to describe. The treatment program had no impact on offenders’ attitudes toward women, wife beating, the seriousness of domestic assault, or the offender’s estimate of the likelihood of repeat violence. Since the groups were created by random assignment, it can be assumed that they were equivalent on all dimensions at the outset of the experiment. At the end of the treatment, the two groups did not differ on measures of abuse from both the victim’s and the offender’s self-reports (p.365). Official arrest statistics tell the same story: “24% of the men in both the experimental and control group were rearrested on one or more occasions during their one-year probation” (p.366).
Conclusion. There seem to be no “clear and demonstrable positive effects of this court mandated… program on offenders’ attitudes, beliefs, and behaviours” (p.371). Further, there was evidence of abuse in both the experimental and control groups. “An unquestioning acceptance of domestic violence batterers’ intervention needs to be challenged” (p.372). While intervention may appear to be a good idea, treatment of the offender – it would seem – does not necessarily lead to a reduction in domestic violence.
Reference: Feder, Lynette and Laura Dugan (2002). A Test of the Efficacy of Court-Mandated Counseling for Domestic Violence Offenders: The Broward Experiment. Justice Quarterly, 19, 343-375.
When do police use force against citizens? An observational study of police use of force in two American cities suggests that it is not the “disrespect” of the citizen which is a key variable. Instead, suspect and police officer demographic characteristics as well as factors such as the amount of resistance that the suspect exhibits appear to be important determinants of the use of coercive techniques.
Background. Although there is a substantial amount of concern about police use of force against citizens, relatively little is known. Focus has largely been on who the citizen is (e.g., race, social class, etc.) and what he/she does (e.g., displays of disrespect toward the police officer) as well as characteristics of the police officer (e.g., background and attitudes). Part of the problem for researchers has been that the interest has often centred on the use of “excessive” force which is not easily defined.
This study examines all force - looked at as a continuum - and assesses the behaviour of citizens throughout their encounters with the police. Accompanying police officers in Indianapolis and St. Petersburg, researchers observed 12,000 police-citizen interactions. A total of 3,116 of these encounters involving interactions between suspects and 270 different police officers were the focus of the present study.
Potentially the most unexpected finding was that “[s]uspects who displayed disrespectful behaviour [as coded by the researcher] toward officers were no more likely to have force used on them than were those who were respectful” (p.236). The researchers were reasonably confident that, in most instances, the police officers’ actions were not affected by their presence since many instances existed in which “the observed police behaviour could have been cause for disciplinary action” (p.227). Additional findings showed that force was more likely to be used against subjects who were male, nonwhite, young, poor, or had apparently used alcohol or drugs. Further, suspects who resisted the police officer in some way, were in conflict with another citizen when the police arrived, had a weapon, were arrested, or for whom the evidence of criminal activity was strong were most likely to be subject to force. Beyond suspect characteristics, an analysis of police officer demographic factors also showed that those who were more educated and had more experience were less likely to use force. However, certain other characteristics of police officers (e.g., whether the officer had undergone training in verbal mediation) had no effect on their use of force.
Conclusion: It would appear that who a suspect is, as well as what a suspect does, are both important determinants of whether they will be subject to the use of force. The fact that males, nonwhites, poor and young suspects are treated more harshly “irrespective of their behaviour” (p.243) is obviously a concern. In addition, while certain characteristics of the officer (education and experience) are important, others (e.g., attitudes about law enforcement) are not. Clearly, some of these factors are difficult to affect (e.g., the suspect’s behaviour). However, others – the educational level of police officers, and their experience on the job – may constitute promising areas of intervention for reshaping police departments’ approach to this problem.
Reference: Terrill, William and Stephen D. Mastrofski (2002). Situational and Officer-Based Determinants of Police Coercion. Justice Quarterly, 19, 215-248.
Publication Type
- Regular Issue