et al., 2002). To date, these replication studies are mostly descriptive in nature.10

What Has Been Learned?

While broad support for the pulling-levers approach may be justified for many reasons, the committee found modest scientific evidence that demonstrates whether these types of targeted policing programs can effectively lower crime and violence. Clearly, there was pronounced and important change in the youth homicide rate in Boston over the period of the intervention, some of which was arguably due to Operation Ceasefire, some due to secular changes in youth homicide, and some due to other (and perhaps unknown) factors. The particular effects of this intervention, however, are unknown. Furthermore, in the committee’s view, the existing data and methods make it difficult to assess how Operation Ceasefire and other similar policing programs affect crime. Researchers cannot hope to credibly control for the many confounders that influence violence and crime using simple time-series comparisons. With similar policing programs being adopted in a number of other areas, there may be opportunities to combine data from these sites to provide more persuasive estimates. Invariably, however, researchers will be confronted with the fact that the programs were not randomly adopted, the trends in violence are influenced by a multitude of factors, and the dynamics of crime and violence are highly complex.

The lack of research on this potentially important intervention is an important shortcoming in the body of knowledge on firearms injury interventions. These programs are widely viewed as effective, but in fact knowledge of how, if at all, they reduce youth crime is limited. Without a much stronger research base, the benefits and harms of these policing interventions remain largely unknown. The committee recommends that a sustained and systematic research program should be conducted to assess the effect of targeted policing aimed at high-risk offenders. Additional insights might be gained by using observational data from different applications, especially if combined with thoughtful behavioral models of policing and crime. An alternative means of assessing the impact of these types of targeted policing interventions would be to run randomized experiments, similar in spirit to those described above. Using this framework, one might hope to disentangle the effects of the various levers and more generally assess the effectiveness of these targeted policing programs.


McGarrell and Chermak (2003) recently completed an unpublished study of the Indianapolis pulling-levers intervention. Using time-series analyses, they found a 42 percent reduction in homicides associated with the implementation of the intervention and found that homicides were less likely to involve firearms, groups, and drugs.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement