RAND > Joint Research > DPRC > Center Activities > 2006 Speaker Series


AddThis Social Bookmark Button

DPRC 2006 Speaker Series

When Brute Force Fails: Strategy for Crime Control

Mark Kleiman, Professor of Public Policy and Director of the Drug Policy Analysis Program, UCLA

Thursday, February 23, 2006

Presentation Abstract:

Insofar as potential offenders are selfishly rational -- deciding to offend or not depending on the relationship between the cost of complying with the law and the risk of breaking it -- and the resources needed to detect and punish any given crime are finite, the scarcity of punishment resources will tend to create a positive-feedback effect in the rate of offending. When the level of offending goes up, the result is that the punishment per offense goes down as the limited punishment capacity is stretched over a larger number of offenses. The lowered risk of punishment, in turn, will further stimulate offending. Thus crime rates under these assumptions will tend to exhibit “tipping” behavior, with both high-violation and low-violation equilibria possible given the same underlying causal situation.

If, on the other hand, the punishment per crime is fixed and the total amount of punishment allowed to vary, then, if the “demand curve” for offending is normal, the total amount of punishment assigned will form an inverted “U” as a function of the punishment per crime; great lenity and great severity will both result in small amounts of actual punishment compared to moderation.

These considerations suggest the great importance of concentration of enforcement: by offense, by offender, and by time and place, and of the direct communication of deterrent threats in order to minimize the cost of “tipping” a high-violation equilibrium into a low-violation equilibrium. Sanctions credibility is vital. “Broken windows” policing, “Cease-Fire”-style gang interventions, and the “coerced abstinence” (testing-and-sanctions) approach to controlling illicit drug use among probationers can all be seen as applications of this simple logic. Since universal zero tolerance is never possible in a world of finite enforcement resources, targeted zero tolerance, with clear communication of precisely what will not be tolerated will outperform “equal-opportunity” approaches to enforcement.

Home About RAND Opportunities Research Areas Books and Publications View Shopping Cart