Criminological Highlights Vol. 1, No. 6 – September 1998

View this issue as a PDF: PDF iconCrimHighlightsV1N6.pdf

Why should we have a separate youth justice system?  It’s not just a question of being less harsh: it’s a better way of reducing the social cost of youth crime.

Separate youth justice systems make sense for a number of quite distinct reasons including the following:

  • The involvement of young people in crime, and their decisions as defendants reflect cognitive and psycho-social immaturity.  “Scientific authority indicates that, in general, the cognitive capacity for reasoning and understanding of preadolescents and many younger teens differs substantially in some regards from that of older teens and adults” (p. 160).   Furthermore, youth are more influenced by peers than are adults.  They have different attitudes and perceptions of risk (p. 163). And they seem “to discount the future more than adults do, and to weight more heavily short-term consequences -- both risks and benefit -- a response that in some circumstances can lead to risky behaviour” (p. 164).   In short -- perhaps in part because they have had fewer life experiences -- their decision making is different from that of adults.  In deciding whether to join in a robbery, a focus on the short-term may make the fear of peer rejection more salient than possible long term consequences of being involved in an offence with a firearm.  And, for the youth as compared to the adult, the excitement and possible immediate rewards of the crime are more relevant than are the possible consequences to himself or others of participating in the offence.
  • “Delinquent adolescents are at risk of being less competent participants in their defence than are adults, and... this risk is especially great for youths under the age of fourteen” (p. 168-9).
  • Policies concerning the punishment of adult offenders are inappropriate for youth since most youth “grow out of” crime.  Thus policies based solely on the harm done by offenders are not likely to be sensible ways of reducing crime.  “[T]he future educational, employment and social productivity of those youths whose crimes are adolescent-limited behaviour is likely to be negatively affected [by imposing adult type approaches on youth]” (p. 179). 

The trend at the moment is to a shift toward an adult system for dealing with young offenders.  This may occur in terms of the actual legislation dealing with young offenders (e.g., by making it more punitive in its orientation) , or, as in the United States, by shifting more and more young people into the adult justice system. This would appear to be counterproductive.

Conclusion. “The most effective means to implement the lessons from developmental psychology is to maintain a system of adjudication and disposition that is separate from the adult criminal justice system” (p. 188).  “Commitment to the development of intensive rehabilitative interventions for young offenders and to the protection of the future prospects of youths whose crimes are adolescent-limited offers the long term promise of lowering the social cost of youth crime more effectively than the blanket punitive policies that are now in vogue” (p. 188-9).

Reference: Scott, Elizabeth S. and Thomas Grisso. The evolution of adolescence: A developmental perspective on juvenile justice reform.  The Journal of Criminal Law and Criminology,  1998, 88 (1), 137-189.

Back to top

Crime prevention programs should be shown not only to be effective, but “cost effective” in terms of the reduction of costs -- criminal justice and social.  Such programs do exist for youth.  Investment in them will not only reduce crime, but will save money.

Background.  Finding programs to treat delinquent youths is not difficult.  Finding programs that are effective is somewhat more difficult.  And finding the “best” way to invest money in reducing crime is even more difficult.  This study, carried out by the Washington State Institute for Public Policy, examined programs for youth where there was sound research to examine their costs and outcomes. 

Results.  Programs do exist, but their impact on youth is often “modest.”  They certainly will not guarantee success.  “The best interventions for juvenile offenders lower the chance of re-offending by about 40%” (p. 7).  Typically, the programs reduce rates of recidivism by about 20-30%. This is important to keep in mind, because it means that the graduates from the best known programs will often re-offend.  It is also relevant when one hears claims in the media about “quick fix” interventions.  But these modest impacts -- e.g., a reduction of reconviction rates from 45% to 27% (a 40% reduction) for probationers in some locations -- may still be worthwhile. 

The question, from a public policy perspective is simple: If a program is likely to reduce recidivism by only modest amounts (20-30%), is it still worth it?  The answer is “yes” -- sometimes.   First of all, one has to ask whether one is interested only in public costs -- typically “criminal justice system” costs.  Some programs do not show a savings on criminal justice costs alone, but do show savings if the costs to victims of crimes are included. Also, for some programs (e.g., early intervention programs directed at health or education issues), other benefits of the program to society can be measured.

But for many of the sixteen programs that are examined in this paper, there are criminal justice savings that can be shown within a year or two.  For example, in a “program for first time minor offenders on diversion where youth appear before a community accountability board shortly after committing an offence”  (the Thurston County FastTrack Diversion program), there is a 29% reduction in offending, with a savings to the criminal justice system of about $2700 per participant after one year.  In large part, this savings may come from the fact that its taxpayers costs are low ($136 per participant).  Other intensive programs funded solely with public money take longer to show criminal justice savings.

But there are some expensive, and thoroughly evaluated,  programs that will never show any kind of benefit when one looks at a measure like “felony reconvictions by age 25.” Juvenile boot camps are one notable example.

Conclusion: Cost effective programs exist for reducing recidivism of juvenile offenders and to prevent delinquency.   They are not necessarily cheap to implement but when considered as investments, they are sensible. Some of the intensive supervision programs, for example,  cost $4500-$6000 per participant and take a few years to show criminal justice savings.  A program for chronic juvenile offenders including a home placement with trained foster parents and other treatment and probation services was quite expensive, but showed benefits to victims and for criminal justice budgets.  Evaluated solely in terms of changes in recidivism rates, however, these programs might be seen as having only modest benefits.  However, as investments to achieve victim and criminal justice savings, they were very effective.

Reference: Aos, Steve, Robert Barnoski, and Roxanne Lieb.  Preventive programs for young offenders: Effective and Cost effective.  Overcrowded Times, 1998 (April), 9 (2), 1,7-11.

Back to top

Crime can be reduced by the collective action of those who  live or work in local city blocks that have drug and  disorder problems.  The police can help by supporting groups on the block and by coordinating services that address non-crime problems on the block.

What can be done to reduce crime in an urban area?  Those living in a neighbourhood have little direct control over who lives in their neighbourhood.  Similarly,  it is difficult for people to create “cohesive” or “caring” neighbours.  But people can do some things to reduce crime in their neighbourhoods.  Police statistics are sometimes used to identify “hot spots” -- where crime and disorder are likely to occur.  Typically, these “hot spots” are single city blocks which acquire characteristics that are conducive to crime.  And city blocks, even more than “neighbourhoods,”  turn out to be sensible sociological, as well as geographic, units to examine when attempting to prevent crime.  For a crime to take place, one needs an offender (without controls) a victim (without protection) and an appropriate location. 

This study identified city blocks in Oakland, California, that clearly had crime and disorder problems. On-site observations were made, and resident “place managers” were interviewed. “Place managers” are people who “live or work near problem places and who, by virtue of their proximity and interests, may have primary or personal responsibility to the street block” (p. 383). Typically four place managers per block were interviewed. 

Implementing crime prevention strategies.  City blocks were randomly assigned to receive special attention from a unit of the Oakland police.  This unit worked with individual citizens, coordinated visits by other local government agencies (fire, public works, rodent control officers, utility companies, etc.) to ensure that all building, safety, etc., codes were enforced, as well as to ensure that owners of problematic properties were made aware of the problems (e.g., the selling of drugs).   The “control” blocks got standard police patrols.

The results demonstrated two independent types of effects. First, “collective action” on the part of place managers -- meeting with community groups about problems, working with the police or community groups about problem areas, participating in a neighbourhood cleanup, participating in neighbourhood or block watch programs, etc., -- had positive effects.   Signs of disorder were reduced. The number of people observed selling drugs was reduced on the blocks where there was more collective action taken by place managers.  Second, above and beyond these effects, those areas targeted (on a random basis) for the police department’s “special attention” in coordinating other city services, showed positive change on these same measures.   Individual action by place managers (e.g., simply calling 911 or the drug hotline, talking to building managers or tenants) did not have a positive impact.

Conclusion: Collective action by place managers seemed to be effective in reducing crime and disorder as was support from the police in dealing with aspects of the block that made it an inviting site for problems.  As the authors point out “place managers play an important role in controlling drug and disorder problems... and may be most effective when they are more socially integrated with their neighbours on the street block and when they are involved in collective, rather than individual, problem solving efforts” (p. 397). Individual actions, such as simply calling the police, did not seem to be effective. “Police efforts that build working relationships with a core group of place managers may have a greater likelihood of long term success than police building one-on-one working relationships with individual place managers.”  Collective neighbourhood actions appear to be important. 

Reference: Mazerolle, Lorraine Green, Colleen Kadleck, and Jan Roehl.  Controlling drug and disorder problems:  The role of place managers.  Criminology, 1998, 36 (2), 371-403.

Back to top

Alleviating the poverty of those in society who are most in need helps everyone: property crime levels go down.  

Background.  The view that “poverty causes crime” is challenged, sometimes, by those who argue that welfare undermines families and economic values and is, itself, a cause in crime.  There has been, recently, a good deal of work on this subject. A paper summarized in Highlights reported that locations that had low welfare rates had high homicide rates.  The underlying assumption behind these findings was that more generous support payments reduced stress on people which resulted in lower levels of interpersonal violence. 

This paper looked at 437 metropolitan counties in the U.S. with populations over 100,000 where reasonable estimates of property crime rates (burglary, larceny, motor vehicle theft)  could be obtained.  The main measure of poverty was the amount received per person from the national program “Aid for families with dependent children” which, though national in scope, has rates set by the state.  These measures were then corrected for cost of living by using a measure of the average rent paid by the poor.  A list of other factors were controlled for to see whether welfare rates had an impact above and beyond these well known correlates of crime.  These “control” factors included the following: The percentage of families that were poor and headed by a female, the percentage of women who were divorced, unemployment rates, average income, and the percentage of the population age 16-24 (the highest property-crime prone age). 

Findings.  The standard predictors of crime generally did predict each of the crimes that were examined (burglary, larceny, and auto theft).  But above and beyond each of these, the two welfare measures -- the percent of poor families who actually receive welfare, and the welfare payment per person adjusted for cost of living -- had a significant impact on each of the crimes: higher welfare rates, or a higher proportion of people actually able to receive welfare led to lower crime rates. 

Since it is sometimes suggested that high welfare rates “serve as a disincentive for creating and maintaining married couple families among the poor,” the authors of this study also examined the predictors of divorce rates and proportion of families in the county that were poor and headed by a female.  As it turns out, high welfare rates were associated with lower divorce rates and lower rates of poor female headed families.

Conclusion. Those who argue that high welfare rates undermine traditional values and thus encourage crime are simply wrong.  High welfare rates (amount given in welfare benefits or rules which allow access to welfare) lead to low crime rates.  These findings are not rhetoric: the study used data from hundreds of metropolitan counties in the U.S.

Reference: Hannon, Lance and James DeFronzo.  Welfare and property crime.  Justice Quarterly, 1998,  15 (2), 273-287.

Back to top

Communities are often convinced that there is an unprecedented “gang” problem.  The result is that resources are mobilized even though the problem may be more in our heads than on our streets.  But the effects of these kinds of panics are real and long lasting.  

Context.  A recent survey of prosecutors in the U.S. found that most saw “gangs” as a growing problem in their communities.   The problem is that the evidence supporting the view that gang crime (and gang violence in particular) is growing is suspect. Some social scientists have pointed out that seeing “gangs” (or crime generally) as an increasing problem serves the needs of criminal justice groups  -- particularly in times when there is competition for legitimacy and resources.  This paper looks at a sudden increase in concern about gangs in Las Vegas, Nevada in the late 1980s and early 1990s.

History of gangs in Las Vegas.  Before the mid-1980s there appeared to be no gang problem. Nevertheless, in 1985 two police officers were assigned to gather evidence on gangs.  These officers announced in 1986 that there were 4000 gang members in the city involved in crime.   Media coverage of gangs skyrocketed from fewer than 25 stories about gangs per year from 1983-7 to approximately 140-170 per year in 1988-91.  A poll in 1989 showed that most residents (89%) thought that gang problems were worsening.  Police sweeps were authorized and, patrols (often by undercover police) of schools began.  New statutes were introduced; consideration was given to banning gang membership, and penalties for “gang-benefiting” crimes were increased.  By 1992, the police began to declare a victory over the gangs and, as laws were passed that gave police additional powers and large increases in police budgets were approved, the gang “problem” disappeared from public view. 

What happened.  Police data suggested that during this period police recorded charges against those identified by the police as being gang members increased from about 3% to 7% of those charged, but most of the increase occurred late in the period -- around 1992 or so. However, even prosecutors were not comfortable with the labeling of gang members, suggesting that the statistics of gang membership might be vastly exaggerated.  When the concern erupted, “known gang members accounted for only about 3% of all reported violent crime” in the county.   The anti-gang legislation was, with a couple of exceptions, almost never used against gang members.

Conclusion:  A moral panic suggesting that “gangs” were a serious and growing problem in Las Vegas in the late 1980s was created where there was no real evidence of any change in the prevalence of gang problems.  But such events do not occur spontaneously.  When do panics occur?  The group that benefited most from the view that “gangs were out of control” -- the police -- appeared to have created the moral panic. Stories of gangs came, not surprisingly, at a time when there was a budget crunch and when the legitimacy and fairness of the police were being questioned (because of allegations of brutality). Police spoke of the growing threat from gangs, the “fact” that the police were “out-gunned” by the gang members, and needed new resources and new legislation. The police presented a “four year plan” for increased resources to combat gangs. In the end, the panic disappeared: newspaper articles about gangs dropped off dramatically by 1994.  But the police got their resources and their laws, and attention was diverted from ongoing police scandals.   But throughout the whole panic period, even using the police department’s own statistics, gang activity, if it increased at all, never accounted for more than 5-7% of crimes.  Gangs, even though they were largely imaginary, had an impact, but on the criminal justice system, not on crime. 

Reference: McCorkle, Richard C. and Terance D. Miethe.  The political and organizational response to gangs: An examination of a “moral panic” in Nevada.  Justice Quarterly,  1998, 15(1),  41-64.

Back to top

The ultimate result of punishing the disadvantaged -- “deviance” becomes normal. A study of African American men ages 18-35 shows that half were under some form of criminal justice control on any given day in 1997.  Sometime during their lives, 80% of African American men will spend at least one night in prison.

Background:  Recent studies in U.S. cities have found that a high portion of black males are likely to be subject to some form of criminal justice control.  Some attribute this to high levels of crime; some to enforcement strategies. There is no doubt that different types of activities are likely to be differentially enforced.  For example, the street selling or possession of certain drugs in some cities is the subject of large scale enforcement activities, whereas the selling and use in private dwellings is not seen as worthy of notice.  Similarly, under U.S. federal sentencing guidelines, possession of crack cocaine results in a dramatically harsher criminal justice response than does possession of one of its main ingredients (cocaine in powder form).  88% of those subject to the harsher penalties for crack are black (and most of those subject to penalties for cocaine in its natural state are white). Such policies obviously explain some of the problem.

This study had a very simple purpose: to do a “census” of black men in the criminal justice system as a first step toward understanding the impact of the system on this group of people.   The focus was Washington, D.C.  When looking at the criminal justice figures, it is important to remember that Washington D.C’s population is about two-thirds (65%) black.  

Almost everyone who is being held in the D.C. Department of Corrections is black.  Corrections had 8153 blacks and 135 whites.  Within the D. C. Department of Corrections, “The fraction of African American men in treatment facilities ranges between 64% and 77%... depending on the facility, but in custodial facilities, the range is between 94% and 99%...” (p. 359).  Figures for other parts of the criminal justice system are similar.  In total, “On an average day in early 1997, 50% of the African American men between the ages of 18 and 35 were under justice control... This number is troubling both for its size and its irrelevance to public safety” (p. 365).

The result of criminalizing a substantial portion of a large population is, of course, that funds are not available for much else.  In 1996, Washington, D.C. spent $842 million on criminal justice (17% of its budget). Less than that ($795 million) was spent on public education. “During the 1980s, corrections spending increased at a rate almost 7 times that of higher education spending” (p. 362).

What can be done?  There are many ways in which the situation in Washington could be addressed.  The lessons are important for any jurisdiction worried either about the over-use of the criminal justice system or about over-incarcerating specific populations.

  • Use prison largely for serious violent offences.  In particular, look for alternative approaches to dealing with drug offenders.
  • Develop a larger range of intermediate sanctions.
  • Develop better ways of dealing with social problems outside of the criminal justice system.
  • When policies are being developed, examine impacts on specific (racial) groups.

Conclusion:  A criminal justice system that has half of a vulnerable population under its control in a city with high levels of crime and very high levels of fear can be described as a system that is not accomplishing its goals.  Washington serves as the example of what happens when a single simplistic approach (criminal justice sanctioning) to a complex problem (crime) is allowed to get out of control.  And it serves as a lesson to those who argue that what we need is “more” criminal justice.  Clearly “more” does not work.  

Reference:  Lotke, Eric.  Hobbling a generation: Young African American men in Washington, D.C.’s criminal justice system -- Five years Later.  Crime and Delinquency, 1998, 44 (3), 355-366.

Back to top

Mandatory minimum sentences for drug crimes are shown to be less effective than treatment in reducing the use of cocaine.  

Background. Legislatures love mandatory minimum sentences.  The suggestion is that with a flick of a pen, the change in the law will reduce the level of the problem.  Mandatory minimum sentences are popular for certain kinds of crimes -- in particular those that frighten the public.  Gun crimes constitute one example, drugs another. Of course, there is almost no limit on how high the mandatory minimum might be.  If “a few years” is said to reduce the level of the problem somewhat, then “more years” should reduce it more.   

But there is another problem.  Many criminal justice programs, if implemented with huge amounts of resources, would be shown to be effective.  Putting a few more police on the street may not affect street crime, but putting 100 police officers into a small area probably would affect crime in that area.  Incapacitating people provides another example: locking up a few more people will not have a measurable impact on crime, but if a substantial portion of the population is in prison, there are, quite simply, fewer people on the street available to commit crime.  

The argument that “if one life is saved” the crime control strategy is “worth it” is a dangerous and short-sighted approach.  In an era of limited public funds for crime prevention or control, the question is not whether “one life might be saved” but whether “more lives might be saved” using a different approach.  A more socially useful approach is to ask “how many lives are saved (or whatever benefit one wants to examine) per million dollars spent” on each of a number of different programs. 

This study, carried out and published by the Rand Corporation, examined the impact of mandatory minimum sentences for drug crimes in comparison to treatment.  The “outcome measure” that was used was “kilograms of consumption prevented per million dollars spent” on each of a number of different approaches to reducing cocaine use.  Note, however, that this is not a “value free” outcome measure: if one’s interest was largely the punishing of cocaine dealers, reduction of cocaine use would be less important.  The second point that one must consider is that some programs -- like effective treatment programs -- are more likely to show their effects in the long term.  

Results.   Comparing various criminal justice approaches with the “treatment of heavy cocaine users” leads to findings easy to summarize:  Treatment of heavy users results in higher levels of “kilograms of cocaine consumption prevented per million dollars spent” (103.6 kg/$1 million over a 15 year period) than does longer sentencing (12.6 kg/$1 million) , or conventional enforcement of the drug laws (27.5 kg/$1 million).   Varying the assumptions behind these models does not appear to make much difference: treatment appears to be the most effective approach, using this definition of “effective.”    It should be pointed out that if long sentences could be directed solely or largely on very high level suppliers of cocaine, and there was little or no replacement for this activity in the market, there would, obviously, be a point at which longer sentences would be cost effective. However, “it is not plausible that these [approaches and results] could pertain to the average federal mandatory minimum defendant, but there may be individuals who do meet the criteria” (p.62).

Conclusion.  “Spending additional money on arrest, prosecution, and conventional sentencing is more cost-effective than spending additional money to extend terms served [for those who are already arrested]... Treatment is more cost-effective than either enforcement approach [conventional enforcement or mandatory minimum sentences] at reducing cocaine consumption and cocaine spending” (p. 51). 

Reference:  Caulkins, Jonathan P., C. Peter Rydell, William L. Schwabe, and James Chiesa.  Mandatory minimum drug sentences: Throwing away the key or the taxpayers. money.  Rand Corporation, Drug Policy Research Centre, 1997.

Back to top

Arresting domestic violence offenders does not seem to be a good strategy to get them to stop offending.  Neither does prosecuting those who are arrested.  

Background.  For about 20 years, the criminal justice systems of many jurisdictions have been examining their policies on what to do about domestic violence.  In the mid-1980s, there was some weak evidence that “arrest” was a good policy.  Careful replication of these early findings took place and the cautiously stated conclusion was that “the clear superiority of arrest in the [early] work was not replicated... [and in those studies that did have an impact] any deterrent effects of arrest were short-lived and in some cases reversed with the passage of time” (p. 435). 

Arrest is not, however, everything.  Does it matter if there is a follow through by the criminal justice system and the person accused of domestic assault is prosecuted?   The evidence from earlier studies suggested that prosecution, like arrest, is no silver bullet. Various prosecutorial policies do not appear to affect the likelihood that an offender will repeat his offence.

This study examined the outcome of the court hearing on recidivism in 669 domestic violence cases. The outcome measure was fairly simple:  was the defendant arrested again within six months of the handing down of the final court decision.  There were three sets of predictors: was the case dismissed in court, was probation imposed, did the accused receive a jail sentence.  Various factors were “controlled for” including the nature of the original charge, criminal record of the defendant, the relationship of defendant to victim, and whether the victim was injured. 

The results are simple to describe.  The offender’s criminal record predicted subsequent offending, but whether the case was dismissed, the offender put on probation or placed in jail had no impact on recidivism.   There is “no basis to conclude that recidivism was deterred by any of the [court outcomes]” (p. 440-441).    These results are “consistent with others in the field, none of which has demonstrated a deterrent effect of prosecution” (p. 441).  

Conclusion.  The authors suggest that there are good reasons to suspect that the criminal justice system may be ineffective in deterring domestic violence.  Previous work has shown that arrest does not appear to deter and this study, along with others, shows that different court outcomes do not have any differential impact on the likelihood that an offender will re-offend.  Such behaviour may, simply, not be under criminal justice control. “We should not be surprised if criminal justice intervention is not the controlling factor in interpersonal relationships governed by complex forces” (p. 442).   Arrest and prosecution, then, cannot be justified in terms of their effects on recidivism.  Policies that endorse such approaches need to be examined if they are based on the idea that such criminal justice approaches will protect victims.

Reference:  Davis, Robert C., Barbara E. Smith, and Laura B. Nickles.  The deterrent effect of prosecuting domestic violence misdemeanors.  Crime and Delinquency,  1998, 44 (3), 434-442.

Back to top

Publication Type

Volume Number


Issue Number