Criminal Law and Behavioral Law and Economics: Observations on the Neglected Role of Uncertainty in Deterring Crime

Abstract

Two similarly situated individuals commit identical crimes. The first is sentenced to 10 years in prison while the second is sentenced to 5 years. The disparity between the two sentences of the criminals is a reason for concern. The person who is sentenced to 10 years has a legitimate moral complaint: why was I sentenced more harshly than she was. In contrast, two individuals commit identical crimes. When the first criminal commits the crime, the police invest few resources in detecting criminals of this particular sort. Consequently, the probability that the criminal will be detected is merely ten percent. When the second criminal commits the crime, the police conduct a major campaign against crimes of this sort and consequently the probability facing the second criminal is twenty percent. The latter case although creating disparate treatment of similarly situated criminals does not raise the same type of moral resentment raised by the first case. The moral concern of the person who asks "why me?" is compelling when there is a disparity in the sentences, but has no moral force when the relevant disparity is disparity with respect to the probability of detection.

Various established doctrines and practices reflect the difference in our moral intuitions between disparity with respect to the size of the sanction and disparity with respect to the probability of detection. Sentences are public; they constitute part of our criminal law. But no criminal code provides any information with respect to the rate of detection, or conviction. Moreover, a criminal sanction cannot be imposed retroactively in our system. On the other hand, a person who committed a crime while the probability of detection was low cannot complain that she was caught simply because the probability of detection increased after she had committed the crime. Last, the probability of detection often changes due to temporary enforcement campaigns initiated by the police. No analogous sentencing campaigns during which judges impose heavier sanctions relative to the ones imposed on criminals who are tried at other times exist in our legal system. In sum, criminal sanctions are public, relatively stable and consequently relatively predictable. In contrast, the practices governing the determination of the probability of conviction reinforce uncertainty.

The paper explores ways of reconciling the different treatments of certainty in the legal system. The overriding normative principle that governs the determination of certainty can be stated as follows. The criminal law system aims at providing maximum deterrence at minimal costs. The costs of the criminal law system are determined by the costs of sentencing and by the costs of the detection and conviction systems. Since the cost of sentencing depends on the average sanction, the first question to answer is what is worse from the potential criminal's point of view: a sentencing scheme which guarantees certainty with respect to the size of the sanction, or a sentencing scheme which does not guarantee such a certainty. Similarly since the costs of detection and conviction depend on the average probability of detection and conviction, the first question is what is worse from the potential criminal's point of view: a detection and conviction scheme which guarantees certainty with respect to the probability of detection and conviction, or a scheme which does not guarantee such a certainty.

Once we discover which scheme is worse from the perspective of the criminal, the State should adopt this scheme since it has larger deterrent effects. Thus, if one finds out that potential criminals prefer a sentencing scheme that does not guarantee certainty, the State should adopt a sentencing scheme that is certain. Similarly, if one finds out that criminals prefer a sentencing scheme, which guarantees certainty with respect to the probability of detection and conviction, the State should adopt a sentencing scheme that guarantees uncertainty.

Our paper invokes insights from psychological experiments to examine the attitudes of criminals towards uncertainty and illustrates that criminals are likely to prefer a scheme in which the size of the sentence is uncertain while the probability of detection and conviction is certain. Consequently, we argue that the choice to increase certainty 1with respect to the size of the sentence and to decrease certainty with respect to the probability of detection and conviction can ultimately be justified on efficiency-based grounds.