For at least two and a half centuries, since the time of Cesare Beccaria, criminologists have recognized that the effectiveness of punishment depends more on its swiftness and certainty than on its severity. Yet, in the United States today, our criminal-justice system has it exactly backwards, relying on what Mark Kleiman and Kelsey Hollander justifiably call “a system of randomly Draconian punishments.” In an insightful new article, Kleiman and Hollander describe in clear, simple terms the systemic reforms that ought to be adopted in order to implement the basic principle that certainty matters more than severity. See “Reducing Crime by Shrinking the Prison Headcount,” 9 Ohio St. J. Crim. L. 89 (2011). They even have an actual, working model of this system to demonstrate its effectiveness: Honolulu’s Project HOPE, which, using a system of summary sanctioning, has achieved dramatic reductions in recidvism among probationers. Klieman and Hollander argue that the HOPE model can and should be expanded to a number of other populations, including parolees.
I find much that is appealing in Kleiman and Hollander’s reform program. But I also think there is something important that is missing in their analysis. As I’ve pointed out in my criticisms of the “evidence-based decision making” movement, there are real dangers when we think about criminal justice purely in terms of the efficient management of social risk. The most fundamental problem with this model is that it assumes that we are capable of treating the people we label as “dangerous” in a restrained, carefully cost-benefit-justified way, when experience teaches that this sort of marking tends to unleash powerful emotional responses like disgust and vindictiveness. James Whitman has written eloquently, and I think persuasively, of these important tendencies.
Indeed, we should ask why it is that the certainty-based approach of Kleiman and Hollander is not the norm. After all, the central premises of their model date back to Beccaria.