How Often Does A Recusal Result In An Equally Divided Court?

UPDATE, 3/7 at 2:45 pm: Ryan Black of Washington University in St. Louis (and a previous “Ask the Author” guest) let us know about a paper he co-authored with Lee Epstein of Northwestern University called “Recusals and the ‘Problem’ of an Equally Divided Supreme Court,” available here. Their paper focuses on a slightly different question than the one addressed in this post – namely, it looks only at the effect of “discretionary” recusals – but provides an abundance of additional details and figures that may be of interest.

With the Court’s recent ruling in Warner-Lambert affirming the judgment by an equally divided vote, it marked the second time this term that a recusal – of Chief Justice Roberts in this case – had prevented the Justices from issuing an opinion on the merits. (The first instance was the 4-4 split in Board of Education of New York v. Tom F, following the recusal of Justice Kennedy.) That sent our crack statistical team scrambling into action: when was the last time this happened twice in one term? And how often does it happen in general?

It turns out we didn’t have to go all that far back to find a term when the Court divided equally more than once: it happened in OT02, when Borden Ranch Partnership v. U.S. Army Corps of Engineers was affirmed without opinion and Justice Kennedy recused, and then again in Dow Chemical Co. v. Stephenson, with Justice Stevens recused. Before this Term, Dow Chemical was the last time any case was affirmed by an equally divided Court.

In fact, it’s odd that the Court has doubled up on such affirmances twice in the last five terms, because the Justices have mostly been able to avoid this circumstance of late: including the two most recent cases, it’s only happened 10 times in the last 18 years. Further, according to a search of U.S. Reports, it’s happened 45 times since 1970, which is around 1% of all cases. It occurred more frequently in the 1970s and ‘80s (17 and 18 times respectively) than it did in the ‘90s and ‘00s (6 and 4 times), but the Court also heard about twice as many cases per term back then. Despite some people’s concerns about what they perceive as increasingly frequent recusals by the Justices, on a percentage basis, the effect such recusals have had on the Court’s decisionmaking has been essentially consistent for the last 35-plus years. If anything, the issue occurs less often than it once did.

Stepping back, a few words about judicial recusal by way of background. There are two main reasons for recusal: personal (or familial) stake in a case, or previous experience with a case. The first is the most common, and this occurs when a Justice has stock in a company or when an immediate family member works for a company or law firm involved in a case. The second situation is confined (with rare exceptions) to the first few years of a Justice’s tenure, and this is where a recusal occurs because a justice has dealt with the case before in some capacity. It could be (as has happened in Chief Justice Roberts’s case) that his old law firm was involved in the matter and he was at the firm and involved directly or indirectly, or it could be that a Justice has seen a case already on the lower court he was elevated from and doesn’t believe he should get two bites at the apple, so to speak. This is the reason Chief Justice Roberts was recused in the case of Hamdan v. Rumsfeld, which he previously heard as a member of the US Court of Appeals for the District of Columbia Circuit.

Perhaps not surprisingly, it has historically been the recusal of a “swing” Justice, rather than that of a reliably liberal or conservative member of the Court, that has left the Court equally divided. For instance, the Court has never been evenly divided as a result of the recusal of Justice Scalia or Thomas, despite their nearly forty years of combined service. By contrast, the recusal of Justice O’Connor has left the Court equally divided six times; of Justice Kennedy three times; and of Justice Powell – the swing Justice of the Burger Court – a whopping thirteen times in only fifteen years on the bench.

UPDATE 3/7, 3:00 pm: Commenter Scott Nelson makes an excellent point below: “Eight of the affirmances by equally divided Court that resulted from the non-participation of Justice Powell came in the October 1984 Term, when Justice Powell was undergoing treatment for cancer and was unable to participate in cases heard during (I believe) the January, February and March sittings of the Court.” Indeed, this article from Time Magazine in 1985 explains that Justice Powell missed 56 arguments that term due to his illness. As Time notes, it’s not entirely clear why some of those split decisions were reargued so they could be decided with Justice Powell on the Court while others were simply disposed of with a non-precedential affirmance.

Of course, the retirement of a Justice may also lead to equally divided votes when the Court is forced to operate with only eight members, but the Justices have done well to avoid this circumstance. For instance, with the Court deadlocked upon the departure of Justice O’Connor in OT05, the Court reargued three cases with Justice Alito on the Court rather than affirm the decisions by an equally divided vote (those cases were Garcetti v. Ceballos, Kansas v. Marsh, and Hudson v. Michigan; see here for more on those cases). Thus, a retirement has not resulted in an equally divided Court since the decision in Hartigan v. Zbaraz, a case argued and decided after the retirement of Justice Powell in the summer of 1987, but before Justice Kennedy was able to replace him in February of 1988.