Preventing Crime

NT OF ME J T US U.S. Department of Justice R T A I P C E E D B O J C S Office of Justice Programs F A V M F O I N A C I J S R E BJ G O OJJ DP O F PR National Institute of Justice JUSTICE National Institute of Justice R e s e a r c h i n B r i e f Jeremy Travis, Director July 1998 Issues and Findings Preventing Crime: What Works, Discussed in this Brief: A con- gressionally mandated evaluation What Doesn’t, What’s Promising of State and local crime prevention by Lawrence W. Sherman, Denise C. Gottfredson, Doris L. MacKenzie, John Eck, programs funded by the U.S. Peter Reuter, and Shawn D. Bushway Department of Justice. Many crime prevention programs work. These are the major conclusions of a Key issues: What works to pre- Others don’t. Most programs have not yet 1997 report to Congress, which was based vent crime, especially youth vio- been evaluated with enough scientific on a systematic review of more than 500 lence? Out of all the hundreds of evidence to draw conclusions. Enough scientific evaluations of crime prevention different strategies used in com- evidence is available, however, to create practices. This Research in Brief summa- munities, families, schools, labor provisional lists of what works, what rizes the research methods and conclu- markets, places, police, and crimi- doesn’t, and what’s promising. Those sions found in that report. lists will grow more quickly if the Nation nal justice, which ones succeed, invests more resources in scientific In 1996, a Federal law required the and to what extent? What does evaluations to hold all crime prevention U.S. Attorney General to provide Con- the scientific evidence suggest programs accountable for their results. gress with an independent review of the about the effectiveness of federally funded crime prevention? Key findings: Very few opera- hat Works? tional crime prevention programs have been evaluated using scien- W tifically recognized standards and • For infants: Frequent home visits • For older male ex-offenders: methodologies, including repeated by nurses and other professionals. Vocational training. tests under similar and different • For preschoolers: Classes with weekly • For rental housing with drug dealing: social settings. Based on a review home visits by preschool teachers. Nuisance abatement action on landlords. of more than 500 prevention pro- • For delinquent and at-risk • For high-crime hot spots: Extra police gram evaluations meeting mini- preadolescents: Family therapy and patrols. mum scientific standards, the parent training. • For high-risk repeat offenders: report concludes that there is mini- • For schools: —Monitoring by specialized police units. mally adequate evidence to estab- —Organizational development for —Incarceration. lish a provisional list of what innovation. • For domestic abusers who are works, what doesn’t, and what’s —Communication and reinforcement of employed: On-scene arrests. clear, consistent norms. promising. The evidence is current —Teaching of social competency skills. • For convicted offenders: Rehabilitation as of late 1996 when the literature —Coaching of high-risk youth in programs with risk-focused treatments. “thinking skills.” • For drug-using offenders in prison: Therapeutic community treatment continued… programs. R e s e a r c h i n B r i e f Issues and Findings effectiveness of State and local crime • In places (such as businesses, 2 continued… prevention assistance programs funded hotels, and other locations). by the U.S. Department of Justice, “with review was completed and is ex- • By police. special emphasis on factors that relate pected to change continually as to juvenile crime and the effect of these • By criminal justice agencies after more program evaluation findings programs on youth violence.” The law arrest. are completed and reported. required that the review “employ rigorous Crime prevention programs in each of Target audience: Federal, State, and scientifically recognized standards these settings are legally eligible for Jus- and methodologies.” Framers of the and local policymakers; criminal tice Department crime prevention fund- law expected that the evaluation would and juvenile justice professionals, ing. However, because Congress requires measure: practitioners, and researchers; edu- that most funding decisions about spe- cators; and leaders of community “(a) reductions in delinquency, juvenile cific programs be decentralized to State organizations promoting preven- crime, youth gang activity, youth sub- and local governments, no detailed tion of crime, juvenile delinquency, stance abuse, and other high-risk factors; breakdown of funding is available by set- ting or by program. The review focused and drug abuse. (b) reductions in the risk factors in the community, schools, and family environ- on whether there is scientific evidence Updates: The most recent lists of ments that contribute to juvenile vio- favoring the types of programs that are what works, what doesn’t, and lence; and (c) increases in the protective eligible for funding, showing they can what’s promising are regularly factors that reduce the likelihood of de- accomplish their goals. 1 updated at the University of linquency and criminal behavior.” This Research in Brief describes the sci- Maryland Web site, http:// After an external, peer-reviewed competi- entific methodologies used to perform the www.preventingcrime.org. The full tion, the National Institute of Justice se- review as well as the limitations of the text of the 1997 report, this Re- lected the proposal of a group from the available data. It then summarizes the search in Brief, and annual updates University of Maryland’s Department of conclusions reached by the authors to de- can all be downloaded from that Criminology and Criminal Justice to per- velop three separate lists of programs for Web site. form the review. which a minimum level of scientific evidence was available: what works, what The review defined “crime prevention” doesn’t, and what’s promising. The text broadly as any practice shown to result in provides more details on the evaluations less crime than would occur without the of each type of program as well as cita- practice. It also examined any program tions to the sources of data the authors that claims to prevent crime or drug reviewed to reach their conclusions. abuse, especially youth violence, and, in Note: The page references in brackets and accordance with the congressional man- italics that follow the bibliographic cita- date, examined the effects of programs on tions refer the reader to the pages in the risk and protective factors for youth vio- printed version of the full 1997 report to lence and drug abuse. Congress where the authors discuss the topics in greater detail. Programs meeting any of these criteria were classified into seven local institu- tional settings in which these practices The science of crime operated: prevention • In communities. To most practitioners, crime prevention is an art. But as the U.S. Congress indicated • In families. in the law requiring this report, the art • In schools. of crime prevention (like the art of medicine) can be evaluated and guided by the • In labor markets. 2 R e s e a r c h i n B r i e f science of measuring program effects. about the way to test cause and effect detail in standard texts, notably Cook Scientific evaluations of crime preven- in any given study—a concept known and Campbell (1979). In the course of tion have both limitations and strengths. as “internal validity.” The rules are far preparing this review, the authors de- The major limitation is that scientific less clear, especially in social sci- veloped a shorthand means of summa- knowledge is provisional, because the ences, about how to judge how widely rizing these rules called the Maryland accuracy of generalizations to all pro- the results of any study may be gener- Scale of Scientific Methods [see pp. 2– grams drawn from one or even several alized—a concept known as “external 15 to 2–19 and the Appendix]. This tests of specific programs is always validity.” The results of a very strong, scale was modified from a similar sys- uncertain. The major strength of scien- internally valid test of how to reduce tem for coding evaluations in a major tific evaluations is that rules of science child abuse among rural, white teen- review of drug prevention work per- provide a consistent and reasonably age mothers, for example, may or may formed by the Center for Substance objective way to draw conclusions about not generalize to a population of inner- Abuse Prevention (1995) and was later cause and effect. city African-American mothers. The found to be similar to scales used to two populations are clearly different, assess the internal validity of clinical Limitations but the question of whether those dif- trials in medicine (Millenson, 1997, ferences change the effects of the pro- p. 131). These standards for assessing Scientific knowledge is provi- gram can best be answered by testing internal validity have been developed sional. The most important limitation the program in both populations. over the past century in a wide range of science is that the knowledge it pro- of fields and are directly responsive to duces is always becoming more re- There is a child abuse prevention pro- the congressional mandate to employ fined, and therefore no conclusion is gram discussed below that has been “rigorous and scientifically recognized permanent. All of the conclusions pre- found effective in both kinds of popu- standards and methodologies” in pre- sented in this Research in Brief, as in lations (Olds et al., 1988). Many preparing the report. the report to Congress, are provi- vention programs, however, have been sional—just as all scientific knowl- tested in only one kind of population.

Load more