[This series of posts is based on Richard Re’s forthcoming Harvard Law Review article, The Due Process Exclusionary Rule.]

In my first post, I explained that exclusionary doctrine is in flux. Canonical cases like Mapp v. Ohio suggested that “all” unconstitutionally obtained evidence must be suppressed, but Herring v. United States and other recent cases counsel against suppression when police act in “good faith.” So, which principle should be the exception, and which the rule?

Current doctrine approaches this question based on an evaluation of social costs and benefits. This post will criticize that prevailing approach. The basic problem is that the Court’s cost/benefit analysis founders for want of empirics.

According to the Court, its holdings are primarily guided by two metrics: The effect of exclusion on police behavior and on crime. But it is very hard to obtain reliable data regarding either of these variables, and the most famous effort to do so (by Dallin Oaks in 1970) could not reach conclusive results. Of course, the justices are brilliant attorneys, and they are fully equipped to marshal plausible arguments based on intuition and anecdote. Without hard data, however, these contentions amount to little more than intelligent conjecture.

Take Herring, the 5-4 Roberts Court decision on the exclusionary rule that I discussed in my first post. In that case, police arrested and searched an individual based on an electronic database indicating that the individual was the target of an outstanding arrest warrant. But the record was wrong: there actually was no such warrant. In this context, would suppression “pay its way”? The Court answered no based on the costs of suppression, even as it acknowledged that the police engaged in deterrable negligence when maintaining their database. Meanwhile, Justice Ginsburg’s dissent reached the opposite conclusion, even though the police error could have been no more than an isolated accident in that jurisdiction. In truth, neither side could be confident as to the consequences of ruling one way or the other.

This empirical void highlights a serious legitimacy problem. The Supreme Court is not the branch of government meant to conduct fact-finding on the sociology of criminal justice. It hears dozens of cases a year, issues decisions in a matter of months, and is not staffed by persons chosen primarily for their capacity for social science research. Trial courts and courts of appeals, which are generally deluged with thinly argued criminal cases, are even more hopelessly unable to engage in empirical rigor. Yet Herring’s cost/benefit analysis presumes that courts should resolve what look like paradigmatically legislative questions.

Finally, the Court’s cost/benefit analysis doesn’t actually support a “good faith” rule. For example, the Court’s analysis should be interested not just in whether the police exhibited good faith, but also in the nature of the defendant’s alleged crime. This suggests that the Court’s willingness to apply the exclusionary rule should vary depending on the underlying offense. The good-faith rule might also be contingent on local conditions. Where the police are particularly prone to error, for instance, perhaps good faith should be no excuse.

In sum, cost/benefit analysis is not a promising way to craft exclusionary jurisprudence. In my next post, I’ll sketch an alternative framework.

The views set forth in these posts are the personal views of the author and do not necessarily reflect those of the law firm with which he is associated.