More on the Death Penalty and Deterrence

[Here is a post by Doug Berman of Oio State, from his blog Sentencing Law and Policy (click here):]

March 24, 2005

Concerns about (and blogsphere buzz on) DP paper

The Sunstein and Vermeule article that I posted here last night, which contends that capital punishment may be morally obligatory if it saves lives through its deterrent effect, is already generating blogsphere buzz.  In addition to Eugene Volokh's initial post here, there are now thoughtful discussions of the paper at Crooked Timber here (with lots of comments) and at Crescat Senentia here and at Mirror of Justice here and here. Because, as I noted before, this topic intrigues me greatly, I wish to weigh in by highlighting two concerns I have about the paper, one empirical and one normative:

1.  An empirical concern: are the data sound?  As Karl Keyes notes in this comment and as others in the blogsphere spotlight, Sunstein and Vermeule rely heavily on data which are shaky at best.  The authors concede that their arguments depend entirely on evidence that capital punishment deters, and that evidence is hardly conclusive.  (Next month at Ohio State, as detailed here, Columbia Prof Jeff Fagan is scheduled to give the annual Reckless Lecture on this topic, and his telling title is "Science and the Illusion of Deterrence in the Death Penalty: Cold Fusion All Over Again.")

2.  A normative concern: doesn't the argument prove too much?  Sunstein and Vermeule are focused on murders in their discussion of "a life-life tradeoff," but their claims would seem readily extended to other kinds of killings.  In my class discussions and in my own thinking, I find the deterrence arguments especially challenging when we consider drunk driving fatalities.  Statistics show over 17,000 alcohol-related driving fatalities each year (data here), and I have to think we could significantly reduce that number by executing just a few drunk drivers.  (Drunk driving seems like a much more deterrable crime than some other killings, and recent history suggests laws and public awareness can have a significant impact on alcohol-related driving fatalities.)  Are Sunstein and Vermeule prepared to argue that execution of drunk drivers is morally obligatory (at least in states like California, Florida, and Texas that have a high number of alcohol-related driving fatalities)?

In short, I ultimately found the Sunstein and Vermeule paper unsatisfying because they duck what I consider to be the really hard questions.