For most long-term crackdowns having apparent initial deterrence, however, the effects began to recede after a short period, sometimes despite the continued injection of a police presence or even an increase in police sanctions. Five studies involving post-crackdown data showed continued deterrence well after the crackdowns ended. In two cases, such residual deterrent effects lasted for a longer period than the crackdown itself. These findings of initial waning of deterrent effect and residual deterrence suggest that crackdowns might be more effective if they are limited in duration and rotated across targets. Although such a policy raises ethical and legal questions, these should not preclude experimental research to establish the residual deterrent effects of short crackdowns across a range of police tactics, offense types, and places. This study proposes a research design for testing the intermittent-crackdown hypothesis and provides an example of its implementation in Minneapolis. 4 figures, 1 table, 72 references. (Author abstract modified)