Some Broad Thoughts on Judicial Recusal

I’ll be testifying this afternoon before the House Subcommittee on Courts and Competition Policy hearing on “Examining the State of Judicial Recusals After Caperton v. A.T. Massey Coal Co.,” and I thought I’d post my written statement here.

It’s not terribly original, I’m sure, because it’s aimed just to give some broad thoughts on why judicial recusal is a difficult question, and why general standards such as “recuse when participating would create the appearance of potential bias,” or “recuse when your impartiality might reasonably be questioned,” are not very helpful. (Such standards might be useful in creating more precise rules, if one considers worry about possible bias or appearance of possible bias alongside other policies. But by themselves, the standards aren’t likely to do a good job.)

Dear Members of the Committee:

I was asked to testify about judicial recusal law following Caperton v. A.T. Massey Coal Co. I was one of the lawyers in that case, but I am not testifying here in that capacity; I’m speaking as a legal academic, representing only my own views. Moreover, since the Supreme Court decision, the Caperton litigation has been proceeding in state court and not federal court, and is governed by the Supreme Court’s constitutional rule, not a federal statute. Nothing I say here would therefore bear on that litigation.

On then to recusal law and federal judges. All agree that judicial impartiality is extremely important. At the same time, judges, like all human beings, have their own histories which have led them to their own attitudes and preconceptions, including towards litigants in their courtrooms.

A judge might have voted for or against a politician who ends up as a litigant before the judge. A politician might have publicly voted for or against the judge. A judge’s family member might have been murdered, which might affect a judge’s attitude towards people accused of murder, or of a particular kind of murder. Cf. Strickler v. Pruett, 1998 WL 340420 (4th Cir. 1998) (Luttig, J.) (responding to a motion for recusal in such a case). A constitutional case may affect the rights of the judge alongside the rights of millions of others, for instance if a female judge is ruling in a case involving the Equal Protection Clause and sex classification — or if a male judge is ruling in the same case. A constitutional case may affect the legal rights or financial standing of a group, such as a religious denomination, that the judge has voluntarily joined.

Moreover, judges usually become judges because they have had a successful career as lawyers, and because they have made political connections as a result of that career. Before ascending the bench, they might have gotten involved in political campaigns, perhaps in campaigns to elect the Senator who later urges the judge’s appointment, or the President who ultimately appoints the judge.

Or they might have gotten involved in state politics, perhaps helping elect a state governor who then appoints them to state judgeships, which becomes stepping-stones to federal judgeships. In the process, they would have gotten involved with the state party apparatus, would have been endorsed or opposed by state newspapers and state politicians, would have made friends in the state or federal Administration, and so on. Just to give two examples with which I’m most familiar, both involving highly respected jurists, my former boss Justice Sandra Day O’Connor was an Arizona Senate majority leader before being appointed to state judgeships and then eventually to the U.S. Supreme Court, and my former boss Judge Alex Kozinski (now the Chief Judge of the Ninth Circuit) was involved in President Reagan’s election campaign, and was then a political appointee in the administration before being appointed to the Court of Federal Claims and then to the Court of Appeals.

As a result of their careers as lawyers, and as judges, judges also make many friends and close acquaintances among lawyers — their classmates, coworkers, political allies, law clerks, and the like. In many smaller towns, where the number of lawyers is likewise small, judges may have social relationships with almost all the leading local lawyers.

Even in D.C., no small town, many of the lawyers who practice before the Supreme Court are former law clerks for Supreme Court Justices, friends of the Justices, or both (since many law clerks end up enjoying very close relationships with their former bosses). To give just three extremely well-regarded examples, the former Solicitor General was once a law clerk for Justice Scalia, the current Principal Deputy Solicitor General was once a law clerk for Justice Breyer, and the private lawyer who represented the University of Michigan in the Supreme Court’s landmark Grutter and Gratz affirmative action cases was once a law clerk for Justice Rehnquist, who was then still alive and on the Court.

As a result of their careers as lawyers, judges also acquire assets; and they may have assets through family connections as well. They may be partners in family businesses, or they may have real estate investments, or they may own stocks. Being a federal judge is a full-time job, so judges do not work in outside businesses. But they are not required to sell all their property and put it in a bank account (or even in a blind trust), something that might be financially quite burdensome, and might drive away many excellent candidates for the federal bench.

Judges also acquire spouses and families. For instance, a judge’s wife may be a businesswoman, and the judge will end up profiting from her business (and may even be a co-owner of her business, under state community property laws). No-one should ask the businesswoman to retire when her husband is appointed to the bench. Likewise, judges may have children or siblings who are businesspeople or lawyers, and who have friends or enemies of their own, who get into trouble with the law, who run for political office, and so on. Of course those judges will assiduously recuse themselves from cases where a party is a relative or a close friend, or where the judge or the judge’s family has a direct business interest in one of the parties. But many cases that come before the judges will inevitably set precedents that foreseeably benefit the judges’ family members, and perhaps even the judges’ own investments.

And judges acquire enemies and hostilities as well as acquiring friends. Lawyers and businesspeople may spend money trying to get a judge voted out of office, or even impeached or recalled. Newspapers may write editorials that condemn a judge’s decisions or ethics, and speak against the judge’s election or reelection. Litigants may say nasty things, both in court and out of it, about judges who rule against them or who seem likely to rule against them.

Recusal rules are in large measure aimed at preventing improper influences that stem from all these connections — but the rules are also necessarily limited by the ubiquity of such connections. Thus, for instance, a judge should recuse himself from a case when he owns even a little bit of stock in one of the litigants. But a judge does not have to recuse himself just because the rule of law announced in a case (for instance, federal preemption of some state regulation of a class of businesses) may affect the business fortunes of the companies in which he or his family members own stock.

A judge should recuse himself in a case in which a friend is a party. But a judge does not have to recuse himself just because he is friendly with one of the lawyers in the case. Likewise, a Catholic judge need not recuse himself in a case involving the Catholic Church (such as City of Boerne v. Flores, in which the petitioner was an Archbishop), no matter how devoted he might be to his Catholicism and to the Church.

Likewise, the Supreme Court held, in Caperton v. A.T. Massey Coal Co., 129 S. Ct. 2252 (2009), that a state Supreme Court Justice must recuse himself in a case where one party’s officer spent a good deal of money advocating for the defeat the judge’s election opponent, because that advocacy had “had a significant and disproportionate influence on the electoral outcome” in the judge’s favor. There was, the Court held, a “sufficiently substantial” “risk that [this] influence engendered actual bias” on the judge’s part in favor of one party. But it’s not clear that a U.S. Supreme Court Justice needs to recuse himself in a case where one of the parties is the President who appointed him, or a Senator who backed his appointment.

In fact, in Clinton v. Jones, 520 U.S. 681 (1997), two of the Justices who heard the case had been appointed by one of the parties, President Clinton. The President had had an even more “significant and disproportionate influence” in placing the Justices on the Court than the influence involved in Caperton: In Caperton, A.T. Massey’s CEO helped persuade the public to elect someone to a judicial seat, but in Clinton, the President personally made the decision to appoint two of the Justices to their seats. Yet that was not seen a reason for the Justices to have to recuse themselves, both in Clinton and in other past cases.

Likewise, a judge may well recuse himself if he feels sufficient personal hostility to a particular litigant or lawyer who is appearing before him. Yet if that were a binding rule, parties and lawyers could judge-shop simply by publicly condemning judges whom they don’t want deciding their cases, or by prominently opposing the judges in election campaigns or confirmation battles. If all it takes to force the recusal of a judge you dislike is to spend, say, $20,000 running sufficiently harsh ads against him — on the theory that the judge will now be hostile to you — then many litigants might be happy to do that. That’s even more so if publicly insulting a judge would suffice to get the judge recused.

Is there a risk that a judge will be biased against people who insult him? Of course; judges are human like the rest of us. But mandating recusal in such situations would be unacceptable, because of the danger of strategic judge-shopping.

If our legal system’s only goal were to try to minimize improper influences on judicial decisionmaking, then perhaps it would mandate recusal in all these situations. But this is not our system’s only goal. Rather, the recusal standards have to balance a wide range of goals. For instance, we want to be able to get final decisions without recusals that leave a Supreme Court split 4-4, or lacking a quorum. We want to be able to appoint high-quality judges who come from the place in which they are to sit, and who are vetted by elected officials in which they are to sit. We want judges’ families not to be unduly handicapped in their own professional lives. We want judges to be free to continue their social lives, and in large measure their financial lives.

We want judges to get even more involved with their communities and the legal profession by participating in various bar events and in community education events, even though this may increase the number of the judges’ social and professional contacts that could potentially lead to some possibility of bias. We don’t want litigants to be able to judge-shop by saying or doing things that force the judge’s recusal.

All this also illustrates the weaknesses of formulating rules based on the “appearance of potential bias,” or on whether a judge’s “impartiality might reasonably be questioned.” (For more on this, see Prof. Ronald Rotunda’s excellent article, Judicial Ethics, the Appearance of Impropriety, and the Proposed New ABA Judicial Code, 34 Hofstra L. Rev. 1337 (2006).) If the rules are attempts to provide an empirical test — would a typical citizen reasonably worry that the judge might be biased? — then they will yield far more recusals that can be justified. A typical reasonable citizen, for instance, might well question the impartiality of a judge in a case where one of the litigants had earlier publicly lambasted the judge, perhaps calling the judge a racist or corrupt or incompetent. But as was discussed above, there are important practical reasons not to require recusals in such cases.

Likewise, a typical reasonable citizen might well worry that a judge would be biased in favor of a lawyer who had once been a law clerk for the judge, or whom the judge knows socially. But applying such a test would mean that judges in small towns would either have to be hermits or outsiders, and the most prominent lawyers in D.C. would be unable to argue before the most prominent courts in D.C.

In fact, what has happened with such “appearance of potential bias” rules — such as the federal recusal statute, which requires recusal whenever a judge’s “impartiality might reasonably be questioned” — is that courts have turned this phrase from a mostly empirical question (might a reasonable person reasonably question the judge’s impartiality?) into a legal label. See, e.g., Liteky v. United States, 510 U.S. 540 (1994) (discussing the “extrajudicial source rule,” one such legal gloss on the standard). Things that many reasonable citizens would see as potentially bias-inducing become permissible, because that’s what the legal precedents say.

There are good reasons for this result, as the discussion above suggested. But the consequence is that the legal system seems to promise the public one thing and deliver another. The result might therefore be less public confidence in the judiciary rather than more: People who expect that a judge will recuse himself — not because of actual bias, but because of an appearance of potential bias — may become disappointed or even outraged when it turns out that the “appearance of potential bias” standard, as interpreted by judges, doesn’t actually call for recusal.

So what should Congress do in this situation? First, it should recognize that judicial recusal rules must try to reconcile many different public interests, and that no formula such as “appearance of potential bias” can capture them all.

Second, it should tread cautiously, and not act unless there seems to be a serious problem. I haven’t heard much evidence that there is indeed a serious problem in the federal judiciary with bias towards or against various parties. Certainly people have made these arguments, whether or not correctly, as to elected state judiciaries. And certainly people have argued that federal judges are unduly biased in favor or against particular legal conclusions (e.g., abortion rights) or interpretive mechanisms (e.g., a living Constitution, or the acceptance of foreign influence on American constitutional law). But neither of those problems, if they are problems, can be dealt with through federal recusal rules.

Third, even if there is a serious problem, the trick is finding a sound solution that does more good than harm. It’s hard to evaluate any particular solutions unless they are laid on the table, in specific terms; the devil is in the details in such matters. All I can say is that I haven’t seen any particular proposal that seems likely to make a substantial improvement here.

Fourth, and now focusing much more narrowly on the Caperton case, it’s not clear to me what, if anything, the Congress needs to do to respond to that case. To be sure, the logic of Caperton isn’t limited to cases where a party or a party’s official spent money in a judicial election, which would by definition involve state judges and not federal judges. Caperton’s rationale could also apply, as I suggested above, to cases where a party played an important role in placing a federal judge on the bench, or in trying to keep the judge off the bench — for instance if a party was (1) the President, (2) the Senator from the judge’s home state, (3) the political party which had backed the judge, (4) an influential newspaper that editorialized for or against the judge when he was nominated, (5) an influential advocacy group that publicly called for or against the judge’s confirmation, or even (6) an important witness at the judge’s confirmation hearings. As Chief Justice Roberts suggested in his Caperton dissent, it’s not obvious whether under the majority’s logic “a debt of gratitude for endorsements by newspapers, interest groups, politicians, or celebrities [would] also give rise to a constitutionally unacceptable probability of bias.” Still, given that the Court’s decision is indeed unclear on this, it’s probably better for Congress to wait for the courts to elaborate on this question.

I hope that these thoughts have been helpful; please let me know if you’d like me to elaborate them on further, or to answer any questions you might have about them.