Terry Stops and Frisks: The Troubling Use of Common Sense in a World of Empirical Data

The investigative detention doctrine first announced in Terry v. Ohio and amplified over the past fifty years has been much analyzed, praised, and criticized from a number of perspectives. Significantly, however, over this time period commentators have only occasionally questioned the Supreme Court’s “common sense” judgments regarding the factors sufficient to establish reasonable suspicion for stops and frisks. For years, the Court has provided no empirical basis for its judgments, due in large part to the lack of reliable data. Now, with the emergence of comprehensive data on these police practices, much can be learned about the predictive power of suspect conduct and other predicates for law enforcement interventions. And what has been learned calls into question a number of factors that have been credited over many years. No observer of the legal system can fail to notice the growing role of data and empirical analysis in the courts. A disparate set of cases have turned in large part on rigorously analyzed data. Yet this trend has not taken root in an important set of cases involving the widely used practice of stop-and-frisk. When stop-and-frisk practices become the subject of litigation, courts generally either have no data to review or have failed to engage in empirical analysis of the data that are available and which could be used to test the claims of reasonable suspicion. Rather, the courts invoke the conventional wisdom that as a matter of common sense certain conduct, for example, furtive movement, flight, bulges in clothing, and suspect location, indicates criminal conduct. We have no argument with common sense propositions; we have no aversion to clear, straightforward thinking. But what this phrase often reflects is a set of unexamined (even if widely held) assumptions. The proliferation of data on these basic questions provides the means for empirical analysis, and it is our argument that courts should do so in assessing reasonable suspicion factors in the same manner that they have engaged in empirical judgments, using both big and targeted data, in other areas.