Justice Stevens, Capital Punishment, and Path Dependence

By Mike Dorf

NPR has released two excerpts of an interview that Nina Totenberg conducted with Justice Stevens over the summer, shortly after his retirement.  In the first segment, Totenberg talks to Justice Stevens about, among other things, the death penalty.  Stevens says that voting to uphold the death penalty was his one regret in his career on the Court.  More on that in a moment, but first an acknowledgment of the obvious: Even at 90, Stevens could clearly continue to serve on the Court if he wanted to.  Indeed, as I'll note in my FindLaw column tomorrow, he was the one who suggested to Sen. Leahy that retired Justices be permitted to serve by designation when a colleague is recused.

Now a comment on the Stevens take on the death penalty.  Justice Stevens had not yet joined the Court in 1972 when a majority invalidated the then-existing death penalty in Furman v. Georgia.  In 1976, Potter Stewart,  Lewis Powell, and Stevens jointly authored the plurality opinion in Gregg v. Georgia that upheld the death penalty as reconstituted by Georgia and other states to require states to narrow the category of execution-worthy defendants using aggravating and mitigating factors.   In the interview with Totenberg, Stevens says that he expected that the sorts of factors upheld in Gregg would mean that only the worst of the worst would be executed, so that a death sentence would not be the sort of random lightning strike that Furman condemned.  But, Stevens goes on to tell Totenberg, in the ensuing years the Court expanded the number of people eligible for the death penalty and otherwise changed the procedural law so that the assumptions underlying Gregg no longer held.

Here I want to bracket two issues: (1) It's not entirely clear that Stevens has this right.  The pro-death-penalty wing of the Court has argued that the lightning-strike character of the death penalty is due to the self-contradictory nature of the Court's death-penalty jurisprudence: Furman and its progeny require that the sentencer's discretion be constrained by aggravating factors, but Lockett v. Ohio, Eddings v. Oklahoma, and their progeny require that the sentencer have discretion not to impose the death penalty based on non-enumerated mitigating factors.  Stevens was in the majority in both Lockett and Eddings, and if those decisions are the source of the failure of the Gregg assumptions, then he shouldn't be blaming the rest of the Court.  But it's also not clear that the tension between Furman and Lockett/Eddings is responsible for most of the mess, and as I said, I'd like to put that issue aside.

(2) I also want to put aside the possibility that Justice Stevens was saying that he now thinks Gregg was wrong, full stop.  It's hard to tell from the editing of the interview, but I think he was saying that Gregg was right when decided but undercut by subsequent jurisprudential moves from which he dissented, so that if he knew then what would follow, he would have voted differently.  At least I'll assume that's what he was saying, because it raises some interesting questions.

Here's the general form of the problem.  Suppose you think that the law on question 1 should be A, so long as the law on question 2 is X, but that if the law on question 2 were Y, then you would think that the law on question 1 should be B.  (There are many examples one could give of different legal questions being related in this sort of way.  Perhaps 1 is a question about the scope of jurisdiction and 2 is a question about the merits--so that wider jurisdiction is okay if there's a narrower merits rule, but that a broader merits rule would open the floodgates if the jurisdiction rule were too broad.)  Anyway, you're faced with a case that presents question 1 at a time when the Court hasn't yet decided question 2.  What do you do?

One possibility would be to take your best guess as to how the Court will decide question 2, and then decide question 1 accordingly.  You might even make this point explicitly: "I'm deciding that the answer to today's question is A on the assumption that if we were to decide question 2, we would say X."  But this can be dicta at most, and it's quite possible you won't be able to get colleagues to go along.  You could include the point as a separate concurrence but again, that doesn't really help in the future when none of your colleagues will feel bound by what you wrote.

Could you make your point even more strongly?  Could you perhaps make your vote on 1 contingent on how the Court later decides 2?  I don't see any mechanism for doing so, other than by saying that you'll no longer support the outcome in 1 should 2 come out the "wrong" way.  The full Court would still need to reconsider 1 in a later case, and by then the argument would have to be made not just that 1 should be decided differently as an initial matter but that stare decisis should be overcome.

In short, there is a pretty deep path-dependence in common-law-style systems of decision making, and there may be no sure-fire way to game one's way around it.