Tuesday, August 30, 2011
Parole boards evaluate the public safety risk posed by parole-eligible prisoners to determine whether they should be released. In this Essay, I argue that this process, at least as it operates in California, is fundamentally flawed because it asks the wrong question. Rather than ask whether an inmate poses any public safety risk, parole board officials should instead ask whether this risk is worth taking.
One way to answer this question would be to make our calculations more inclusive of all the costs and benefits of release and comparing them with the costs and benefits of retention. Elementary as this might seem, there is no analysis of costs and benefits in California beyond the requirement that any risk not be “unreasonable.”
But even if we could figure out costs and benefits with a greater degree of precision than is currently possible, quantification of the costs and benefits still does not tell us whether the risk is worth taking. This Essay proposes that our conversations about risk are not merely confined to bloodless, actuarial issues: they need to involve normative issues as well.
I propose, as a means of examining these issues, two contradictory mechanisms for pushing the normative issues to the fore, one systemic, the other individualistic. The systemic mechanism would answer the question by establishing hard numbers for release – such as population or percentage targets. This would push the system as a whole away from individual assessments towards decisions more in line with social costs and benefits. The individual mechanism, on the other hand, would account for the normative elements of parole release, leaving the decision to a body used to fact-intensive inquiries that require moral legitimacy: the jury.
Ultimately, these two proposals highlight the hybrid nature of parole – its mixture of risk and desert. By exploring these two proposals, I hope to deepen the conversation about ways in which the goals and objectives of parole determine the procedures and mechanisms of release.