Police Science


Criminal Justice

Holding a Magnifying Glass Over Police Science

Malcolm Sparrow Faculty ResearcherMalcolm Sparrow,
Professor of the Practice of Public Management,
Harvard Kennedy School
Paper TitleGoverning Science

Police officers responding to an incident of domestic violence were traditionally unlikely to make any arrests, because they knew from experience that the victims of these crimes were unlikely to cooperate with prosecutors. Then the police began responding to new research that showed that an arrest, even if it had little chance of leading to prosecution, would make the aggressor less likely to be violent in the future. That’s an example of Evidence-Based Policing (EBP) — of finding out, through rigorous scientific testing, what works in crime control and prevention. But its supporters are concerned that after a bright start in the 1990s, EBP is not fulfilling its early promise, and that it deserves a more central place in policing, guiding practice and even being used to evaluate practitioners.

Malcolm Sparrow disagrees. Police science, he argues in “Governing Science,” is too important to be left to social scientists.

Sparrow, Professor of the Practice of Public Management at Harvard Kennedy School, wrote the paper at the invitation of the National Institute of Justice in order to balance the more extreme claims of some EBP champions who view social scientists as the only qualified arbiters of “what works.” The paper came out of the Executive Session on Policing and Public Safety run by the Kennedy School's Program in Criminal Justice Policy and Management with sponsorship from the National Institute of Justice.

“I believe that we are in a particularly important period in the development of police science, requiring enriched and productive relationships between police and academia,” Sparrow writes. “I also believe that much harm might result if we give EBP a dominant position in the context of that relationship.”

Equating EBP with science is “grossly misleading,” Sparrow argues. He stresses the importance of balancing the inquiry methods of the natural sciences with those of the social sciences. Social scientists focus on the question “What works?” and rely heavily on the techniques of program evaluation. Natural scientists focus more on “How does the world work?” They probe the dynamics and structure of specific crime risks, figuring out how to artfully sabotage them with incisive (and often novel) interventions. Relevant inquiry techniques, Sparrow says, include close observation, exploration, data mining, pattern recognition, analytic decomposition and disaggregation, experimentation, trial and error, and incremental adjustment.

If it were to focus exclusively on EBP, Sparrow argues, the discipline of police science would be in danger of falling into the hands of elite scholars focused on only the most demanding levels of proof, such as randomized, controlled trials. “Proponents of EBP have set the bar for ‘knowing’ so high, and made the means for generating knowledge so particular, that they end up knowing relatively little,” Sparrow writes. “Operational police need to know much more, just well enough and much sooner, in order to keep up with the pace and variety of the challenges they face.”

Sparrow points to EBP’s fundamental incompatibility with the operational problem-solving work of the police: Thorough program evaluations can take three to five years; they focus on macro-level analyses and therefore ignore distinct but lower-level components of major crime problems; they discourage creativity and experimentation by restricting operations to the use of proven methods; they perpetuate a program-centric mind-set and distract attention from a task-based approach; and, because of their focus on what is statistically noticeable, they may not recognize the swift and effective interventions that are often hallmarks of the best policing, when emerging problems are spotted and suppressed before they become statistically significant.

EBP does have a place, Sparrow concludes, but that place should be limited and clearly defined. For example, it can be used in evaluating “programs and methods that are expensive, long term, potentially permanent”— and that are deployed across departments in a sufficiently standardized way to actually be evaluated.

Sparrow, a mathematician himself, urges the police profession to embrace a much broader range of analytical and inquiry techniques, and reminds social scientists that “the majority of scientific advances benefiting humankind have arisen and become firmly established without their help.”

— by Robert O'Neill

“We are in a particularly important period in the development of police science, requiring enriched and productive relationships between police and academia. ”