Prosecutorial Discretion, Desuetude, and a Postscript on Gravity

by Michael Dorf

My latest Verdict column argues that there is a link between the Obama immigration policy of deferred action and the non-indictment of (now-former) Ferguson police officer Darren Wilson for the killing of Michael Brown. In both cases, the best justification for the less-than-fully-zealous enforcement of the law rests on prosecutorial discretion. As I note in the column, however, that doesn't mean that either the Obama policy or the throw-all-the-evidence-possible-at-the-grand-jury approach of prosecutor Bob McCulloch is warranted, because prosecutorial discretion is itself problematic--albeit inevitable.

In an earlier post, I noted that the OLC memo in support of the Obama deferred action program sets out criteria to guide the exercise of prosecutorial discretion. These criteria recognize the primacy of the legislature and also indicate that limited resources are a key factor in justifying any particular act of non-enforcement as prosecutorial discretion rather than simply dereliction. I argued there that notwithstanding the OLC's laudable effort to cabin prosecutorial discretion, the Obama theory may be insufficient. My worry was that, taken seriously, the OLC criteria could undercut the Administration's decision to under-enforce federal marijuana laws in states that have legalized (medical or recreational) marijuana, making the criteria too strict to account for what I take to be a generally shared view that the marijuana policy ought to be permissible.

From the other direction, others have expressed the view that the OLC criteria are too weak: they would permit a future Republican Administration to under-enforce various provisions of the tax code, including those essential to enforcing the Affordable Care Act. The crucial ambiguity, in my view, concerns what counts as legislative acquiescence in a policy of non-enforcement and when that acquiescence is legitimate.

Let's begin with an easy case, based on Guido Calabresi's views about desuetude. A case like Griswold v. Connecticut, Calabresi argued in A Common Law for the Age of Statutesmight have been better justified on a desuetude theory than on the right-to-privacy gronds given in the actual SCOTUS opinion. The state law forbidding contraceptive use had long fallen into disuse, only producing a test case because the case was set up. Thus, Calabresi argued, the Court might have only provisionally invalidated the law and "remanded" to the legislature to see whether the law still reflected popular views.

Calabresi claimed that his notion of a constitutionally required legislative remand was rooted in the Supreme Court's own practice (and previously elaborated by Alex Bickel and Harry Wellington). However, it is hard to find any examples of US courts actually basing a constitutional remand on desuetude--with the exception of Calabresi's own concurring opinion in Quill v. Vacco. But if desuetude is not generally acknowledged as a justification for courts setting laws aside as unconstitutional, it nonetheless is usually seen as a legitimate basis for the exercise of prosecutorial discretion. Indeed, if a law has been on the books but not enforced for many years but then suddenly a prosecutor dusts it off and starts to enforce it, many people would say that the prosecutor has acted unfairly.

Why? I can think immediately of three reasons.

First, the long period of desuetude may have induced reliance. People may have assumed that they did not risk prison if they violated the long-ignored law. Thus there is at least a notice issue. Of course, the notice problem can be overcome by an executive announcement that henceforth the non-enforcement will (prospectively) end.

Second, desuetude can sap the strength of a repeal movement. If the government had been vigorously enforcing a law that had become unpopular, that enforcement could have generated a campaign to repeal the law. Thus, when a prosecutor dusts off an old law, he is, in a sense, bringing a prosecution under a law that does not reflect the current popular will.

Third, desuetude can be a sign of legislative acquiescence. If a law has been on the books but not enforced for some long period of time, yet the legislature has enacted no legislation instructing the executive to enforce the law vigorously, we might conclude that the legislature has acquiesced in non-enforcement.

This last point is, to my mind, the most difficult. The border between legislative acquiescence and shortcircuiting the legislative process may be in the eye of the beholder.

Suppose that a Republican president adopts a policy of "deferred action" with respect to non-payment of the taxes owed by people who fail to sign up for health insurance under the ACA (assuming it survives King v. Burwell). But suppose further that the Democrats retain sufficient strength in Congress (perhaps via the filibuster) to prevent the enactment of a law actually repealing the ACA. Under these circumstances, could Congress really be said to acquiesce in the non-enforcement of the ACA simply because Republicans in Congress (backed by the Republican president) can block any effort by Democrats to insist that the president enforce the ACA? Wouldn't it be more accurate to say that the ACA remains divisive in such a case, so that Congress has neither acquiesced in nor forbidden the exercise of prosecutorial discretion to refuse to enforce the ACA?

Permitting the legislature's mere failure to act (one way or the other) to count as acquiescence in executive non-enforcement would thus seem to undermine a core tenet of legislation: that absent a sunset provision, laws remain on the books and valid until repealed through the same legislative process that enacted them. But if that's so, it's hard to see how desuetude can ever get off the ground. Perhaps desuetude can only be a legitimate basis for the exercise of prosecutorial discretion where the desuetude arises originally out of some other, legitimate basis for the exercise of prosecutorial discretion, such as a resource constraint.

I will leave matters rather unsatisfactorily there. Both here and in the Verdict column I do not reach firm conclusions about either the general principles of prosecutorial discretion or how those principles ought to apply in particular controversial cases. Prosecutorial discretion turns out to be one of those unusual phenomena that is ubiquitous and seemingly straightforward but, upon inspection, proves deeply mysterious.

Postscript: With a hat tip to the late great Richard Feynman for this example, I would point to gravity as another such phenomenon. Small children wonder why objects fall down. Their parents tell them it's because of "gravity," which is a word, not an explanation. When the children grow bigger and take high school physics, they are taught Newton's Law, which is a description but still not an explanation, for it raises a question that troubled Newton himself: how does action at a distance occur? A college physics course provides a better model--massive objects curve space-time--but that's still only a model. Until physics discovers the graviton or some other means by which gravitational attraction actually occurs, we will only understand that gravity works, not how it works.