Recusals and the 'problem' of an equally divided Supreme Court.

• Author: Black, Ryan

INTRODUCTION

For over five decades now we political scientists have been systematically studying the United States Supreme Court. We have examined how the justices attain their seats, (1) what factors explain the Court's decision to grant certiorari, (2) what impact oral arguments have, (3) whether the Chief Justice self-assigns particularly important cases, (4) why the justices vote the way they do, (5) what rationales they invoke to justify their decisions, (6) and in what ways those decisions affect social, legal, and economic policy. (7) In short, it seems as though no feature of the Court has escaped our attention--with one notable exception: recusal. A search of political science articles in J-STOR (8) on the term "recusal" yields exactly one article, which, in fact, was not at all about recusal. (9) We also ought note that broadening the J-STOR search (to include terms such as disqualif! w/10 Supreme Court or disqualif! w/10 judge or disqualif! w/10 justice) was equally unproductive, turning up no articles. By contrast, a Lexis search of law reviews on that same term produces 1,395 articles. (10)

Why the discrepancy exists is a good question, but we do not think the answer is a lack of interest on the part of political scientists. Quite the opposite. While a justice's decision to recuse "isn't entirely ... discretionary," (11) as Justice Ginsburg once said, (12) neither is it entirely lacking in discretion:

[F]or ... a court of appeals judge on a three-judge panel ... [i]f there were any doubt, that judge could step out and let one of her colleagues replace her. But on the Supreme Court, if one of us is out, that leaves eight, and the attendant risk that we will be unable to decide the case, that it will divide evenly.... When cases divide evenly, we affirm the decision below automatically. Because there's no substitute for a Supreme Court Justice, it is important that we not lightly recuse ourselves. (13) Ifill underscores Ginsburg's point when she notes that the justices

enjoy the unreviewable power to determine individually whether and when to disqualify themselves from cases in which their impartiality could reasonably be questioned. Historically this appears to have produced a highly idiosyncratic application of the [disqualification] standard. (14) By way of example, she points to Justice Marshall's decision to recuse himself from cases in which the NAACP (or NAACP Legal Defense Fund) appeared as counsel, (15) and compares it to then-Associate Justice Rehnquist's refusal to disqualify himself in Laird v. Tatum. (16)

That room for choice exists over the decision to recuse provides the makings of a political science problem, (17) and we can think of many possible solutions. It may be, for example, that a justice is less willing to recuse herself in cases she thinks will produce an equally divided Court--a possibility to which Justice Ginsburg alludes above. So too, it could be that certain kinds of justices, perhaps those who are pivotal in a particular area of the law or who have been on the Court for some period of time, may also be less inclined to remove themselves from particular cases.

Other explanations are readily apparent, and later we discuss several. But the larger point is that a dearth of possibilities hardly exists. Where there is a void--and what we think explains the lack of research on recusal--is in the data to assess those explanations, and the near insurmountable obstacles in collecting such data. Fundamentally, the problem boils down to this: While we can observe when justices recuse themselves, we cannot observe when they considered recusing themselves but did not. (18) We thus lack a "denominator" for testing hypotheses about the decision to recuse; only by assuming that a justice could choose against participating in each and every case could we develop one. (19) But that assumption is insufficiently realistic to put to use in serious research.

And we do not. What we do instead is take a more manageable cut at the problem of recusal and study whether it is in fact a problem. We raise this question in light of the abundance of commentary echoing Justice Ginsburg's concern: that Supreme Court justices ought not take recusal "lightly" because of the possibility of a four-to-four split. As Bleich and Klaus note,

[I]n a case where one justice is recused, an even number remains, which creates a risk of deadlock by an equally divided Court. Just last term, for example, Justice O'Connor was recused in a case concerning the scope of federal removal jurisdiction, because she owned stock in one of the parties. The court wound up deadlocking, 4-4, and as a result an important issue of federal procedural law was not decided and remains the subject of a split between the federal circuits today. (20) Justice Scalia apparently concurs with this general sentiment. In a controversial memorandum explaining his decision to participate in Cheney v. United States District Court, (21) he wrote:

Let me respond, at the outset, to Sierra Club's suggestion that I should "resolve any doubts in favor of recusal." That might be sound advice if I were sitting on a Court of Appeals.... On the Supreme Court, however, the consequence is different: The Court proceeds with eight Justices, raising the possibility that, by reason of a tie vote, it will find itself unable to resolve the significant legal issue presented by the case.... A rule that required members of this court to remove themselves from cases in which official actions of friends were at issue would be utterly disabling. (22) And, now, Chief Justice Rehnquist has taken this idea to new heights. Owing to his bout with thyroid cancer, he has said he will only participate in cases "when necessary to prevent a tie vote." (23) That Rehnquist would take this step is not altogether surprising. Any number of times in the past he has expressed concerns about the consequences of affirming judgments by an equally divided Court. (24)

What we examine is whether this is a legitimate concern on the part of the legal commentators and justices who have expressed it. Specifically, we explore the extent to which discretionary recusals produce an equally divided Court. As it turns out (and contrary to some commentary), in only a small fraction of cases in which one justice recuses him- or herself (forty-nine out of 599) does a tie result. (25)

We develop this finding in three steps. First, we provide a brief look at the "problem" of an equally divided Court; next we present our analysis of the results of discretionary recusals. Along these lines, we provide some descriptive data on recusals since the 1946 term, as well as demonstrate that recusals generally do not produce equally divided Courts. We end with a discussion on possible explanations for our results and with suggestions for future research on recusals.

1. "AFFIRMED BY AN EQUALLY DIVIDED COURT"

   While several contemporary justices have written about the "problem" of an equally divided Court, it is hardly a novel issue. Indeed, as far back as 1792, the Court divided three-to-three on a motion filed by the Attorney General--with the result being a denial of the motion. (26) Three decades later when the justices split in The Antelope, (27) the Court once again affirmed on the question that divided them "without much discussion." (28) Only in the following term, in Etting v. Bank of the United States, did Chief Justice Marshall explain that "the principles of law which have been argued cannot be settled; but the judgment is affirmed, the court being divided in opinion upon it." (29)

   Subsequent cases have fleshed out Chief Justice Marshall's words such that now, when the justices evenly divide, the resulting decision ("affirmed by an equally divided Court"): (30)

   * affirms the decision of the court below; (31)

   * binds the parties under the principle of res judicata; (32) and

   * carries no precedential weight. (33)

   Over the Court's use of these rules, little disagreement exists; (34) in fact, as Hartnett explains, the justices have been unwavering in their attachment to them. (35)

   Nor does much disagreement exist over the problems associated with equally divided Courts. (36) Aside from the sheer "embarrassment" of issuing such decisions, (37) they are inefficient: They take up precious space on the Court s plenary docket, waste the time and energy of the justices and their clerks, as well as tap the time, energy, and resources of lawyers, litigants, and any amici curiae. (38) Then there is the matter of the decision (or, rather, non-decision) itself. By "failing to provide definitive statements of the law," (39) thereby enabling lower courts at least occasionally to continue on a collision course, affirmances by an equally divided Court can leave a legal area murkier than before the justices entered it. (40)

   It is no wonder thus that when Justice Scalia recused himself from Elk Grove Unified School District v. Newdow, (41) commentators expressed concern over the possibility of an equally divided Court. (42) It also is no wonder that generations of justices have worked to minimize the possibility. We already have mentioned Chief Justice Rehnquist's decision to participate only in evenly divided disputes. This may be extreme, but it is far from the only example of the Court or one of its members taking steps to avoid the dreaded "affirmed by an equally divided" disposition. When Justice Stone became ill, his colleagues held over West Coast Hotel Co. v. Parrish (43) because "a decision by an evenly divided Court was thought an 'unfortunate outcome.'" (44) Likewise, after the Court heard arguments in NAACP v. Button, (45) and after Justice Frankfurter wrote a draft majority opinion (upholding the constitutionality of the Virginia law), Justice Whittaker resigned, leaving the Court (apparently) divided four to four. Rather than hand down a evenly divided judgment, rearguments were ordered--with the end...