Setting a Higher Standard for the Evaluation of Problem-Oriented Policing Initiatives

Approximately 30 years have passed since Herman Goldstein (1979) first articulated the basic principles of problem-oriented policing. During those years, problem-oriented policing emerged as a widely practiced approach to crime prevention in police departments throughout the United States as well as throughout the world. Problem-oriented policing enjoys broad support from federal agencies, professional policing groups, and a small cadre of scholars interested in effective crime prevention practices. Given this support, it was surprising to learn that Weisburd, Telep, Hinkle and Eck (2010, this issue) identified only ten evaluations that met the minimum methodological standards of their systematic review—a comparison group that did not receive the problem-oriented policing treatment. Their systematic review also identified 45 before–after intervention studies without a strong control or comparison group. To some observers, a ratio of one strong evaluation to every four or five weak evaluations might not sound overly concerning. Weak evaluations, unfortunately, provide less valid answers to policy questions when compared with well-designed quasi-experiments and randomized controlled trials (Campbell and Stanley, 1966; Cook and Campbell, 1979). Several crime and justice scholars suggest that a “moral imperative” exists in pursuing the most rigorous evaluation designs to discover whether a program is effective (e.g., Boruch, 1975; Weisburd, 2003). Isolating the effects of treatments or programs from other confounding aspects of selection or design is viewed as one of the evaluator’s most important obligations to society. When the evaluation evidence base is informed largely by weak designs, practitioners risk implementing certain treatments or programs as effective crime prevention practices when they are not, which can lead to significant economic and social costs.