Wednesday, February 12, 2014
The "trial penalty" is a concept widely accepted by all the major actors in the criminal justice system: defendants, prosecutors, defense attorneys, court employees, and judges. The notion is that defendants receive longer sentences at trial than they would have through plea bargain, often substantially longer. The concept is intuitive: longer sentences are necessary in order to induce settlements and without a high settlement rate it would be impossible for courts as currently structured to sustain their immense caseload. While intuitively appealing, this view of the trial penalty is completely at odds with economic prediction. Since both prosecutors and defendants have the ability to reject unappealing settlements, sentences at trial should be nearly the same as those arrived at through pleas. This is a straightforward application of the "shadow of the law" concept articulated by Mnookin and Kornhauser, as well as others. This article attempts to answer two questions relating to the trial penalty. Why is belief in its existence so widespread, given that it is at odds with basic economic theory? Which theory does empirical analysis of actual sentencing data support? I argue that there is a fundamental misunderstanding that is largely responsible for the belief in the trial penalty: the failure to distinguish between conditional and unconditional expected values. I also provide a brief overview of an empirical study that attempts to distinguish between the two theories. The results of that study are surprising and support neither theory-not only is there no evidence for a trial penalty, there appears to be a trial discount!