Deterrence's Difficulty

Abstract

A standard assumption of criminal punishment is that it will create deterrence. This Article superimposes a few complications on this premise through the vehicle of substitution. Beginning with law and economic approaches to criminal law, it contends that the notion of marginal deterrence should be extended to cover other instances of substitution. For example, substitution might suggest that harsh penalties on crack cocaine might have increased consumption of heroin (heroin is punished with a sentence that is 20 to 400 times lower than a comparable crack sentence). Considerations about optimal penalties, therefore, cannot revolve solely around the harm of the activity being punished, as they must examine the futility of likely substitutes, and discount the benefits accordingly. To minimize harmful substitution effects, overarching approaches to criminal sentencing (such as the Sentencing Guidelines) might take substitution effects into account in formulating penalties. Several other complications the paper develops takes up questions about income effects, and how strategies based on such effects can maximize deterrence.The paper then relaxes the assumption of fixed preferences, and suggests that criminalization may shape tastes in several ways. Building on work in sociology, it contends that a spectacle of punishment and other lore-creating mechanisms of criminal law may be designed not to price conduct, but to change people's attitudes towards crime. Yet there is potential complementarity from criminal sentences because they may produce stigma and mark offenders. The stigmatization of various subgroups may lead to internal norms about crime that may further marginalize its members, thus producing more crime. In addition, the paper suggests that recent work in cognitive psychology undermines the notion of rational-choice between alternatives, and that decision making between criminal options might be influenced by technically irrelevant alternatives (via extremeness aversion and compromise effects). As a result of these substitution effects and cognitive biases, it might be worthwhile for criminal law to create "less bad" alternatives so that substitution can be used to detract lawbreakers away from serious crimes, and channel their criminal activity into less harmful areas.