The legal principle that “No person may be a judge in his own cause” can be traced back centuries. But is it a mere shibboleth, given lip service long after it has ceased to play an important role in our law? In Williams v. Pennsylvania, the Supreme Court affirmed its continued vitality, holding that a prosecutor-turned-judge was constitutionally required to recuse himself from a case involving a man for whom he (as prosecutor) had once recommended the death penalty.

The events that gave rise to Williams took place decades ago. In 1986, Terrance Williams stood trial for the brutal murder of Amos Norwood. Then-District Attorney of Philadelphia Ronald Castille approved the trial prosecutor’s request to seek the death penalty for Williams. Williams was convicted and sentenced to death. Over the next 26 years, Williams’s conviction and sentence were upheld on direct appeal, state postconviction review and federal habeas corpus review. In 2012, Williams filed a successive petition pursuant to Pennsylvania’s Post Conviction Relief Act, arguing that the trial prosecutor had obtained false testimony from his codefendant and suppressed material, exculpatory evidence in violation of the rule articulated by the Supreme Court in Brady v. Maryland.

The Philadelphia Court of Common Pleas agreed, granted Williams a stay of execution, and ordered a new sentencing hearing. The government asked the Pennsylvania Supreme Court to vacate the stay. The chief justice of the Pennsylvania Supreme Court was none other than Ronald Castille. Williams filed a response, along with a motion asking Chief Justice Castille to recuse himself or (if he declined to do so) refer the recusal motion to the full court for decision, on the grounds that Castille had a significant personal involvement in the case. Without explanation, Castille denied both the motion and the request. Castille then joined an opinion vacating the stay and reinstating Williams’s death sentence. Two weeks later, Castille retired from the bench. Williams sought review from the U.S. Supreme Court, arguing that Castille deprived him of due process of law by acting as both accuser and judge.

The Due Process of Law Clauses of the Fifth and Fourteenth Amendments have long been understood by the Supreme Court to guarantee (among other things) impartial adjudication—adjudication free from bias or even the probability of bias. No mechanical formula can be used to determine whether bias is actually or probably present, and the Court has not attempted to articulate one. It has rejected exclusive reliance upon common law rules requiring judicial recusal, instead looking to “all the circumstances of [the] case” to determine whether circumstances “offer a possible temptation to the average man as a judge . . . not to hold the balance, nice, clear and true.”

In a 5-3 decision, the Court ruled in Williams’s favor. Justice Anthony Kennedy, writing for the majority, emphasized the significance of the decision that Castille had made in seeking the death penalty for Williams. He explained that “whether to ask a jury to end the defendant’s life is one of the most serious discretionary decisions a prosecutor can be called upon to make” and that “without [Castille’s] express authorization, the Commonwealth would not have been able to pursue a death sentence against Williams.” Responding to the government’s argument that Castille’s decision “amounted to a brief administrative act limited to ‘the time it takes to read a one-and-a-half-page memo,’” Kennedy pointedly stated that the Court  “[would] not assume that then-District Attorney Castille treated so major a decision as a perfunctory task requiring little time, judgment, or reflection on his part.” Kennedy further noted that Castille had during his election campaign for chief justice touted his role in seeking the death penalty as a district attorney, claiming that he “sent 45 people to death rows.” Accordingly, the Court concluded that the possibility that Castille would be “psychologically wedded” to his previous position as a prosecutor or be unduly influenced by his own impressions of the case presented “an unacceptable risk of actual bias.” Although Castille’s vote to vacate the stay and reinstate Williams’s death sentence was not decisive, the Court determined that Williams was nonetheless entitled to relief, given the impossibility of determining the extent to which Castille influenced his peers’ votes and the importance of not only actual impartiality but the appearance of impartiality “to the public legitimacy of judicial pronouncements and thus to the rule of law itself.”

The majority opinion was met with two strong dissents, one written by Chief Justice John Roberts (joined by Justice Samuel Alito) and one written by Justice Clarence Thomas. The dissenters pointed out that Castille did not judge precisely the same case in which he had been involved as prosecutor. Williams had long since been convicted and sentenced to death, and a postconviction proceeding is not actually part of a criminal proceeding at all. Chief Justice Roberts noted that Williams did not allege that Castille had “any previous knowledge of the contested facts at issue in the habeas petition, or that he had previously made any decision on the questions raised by that petition.” Justice Thomas made the further argument that the Court’s approach was unmoored from history: “The early American conception of judicial disqualification was in keeping with the ‘clear and simple’ common-law rule—‘a judge was disqualified for direct pecuniary interest and for nothing else.’”

Even if one grants the validity of the dissenters’ distinction between a criminal proceeding and a postconviction proceeding, the majority's argument that Castille was likely to be psychologically wedded to his initial decision to seek the death penalty for Williams in determining whether to vacate a stay of Williams’s execution went unanswered. Further, “No person may be a judge in his own cause” expresses a broad principle prohibiting biased adjudication, not a mere procedural requirement or a narrow prohibition against judges having a certain kind of interest in an outcome. The common law rule requiring recusal in cases where judges had direct pecuniary interests did not fall from the sky—it was grounded in the recognition that such an interest was highly likely to produce a biased judgment, even if in fact a judge was capable of doing his duty and impartially adjudicating the case. There is no reason in principle why circumstances that could similarly bias judicial decision-making in favor of one party should not be treated similarly. While determining whether the requisite potential for bias is present in a given case may be difficult, the difficulty of an inquiry is no argument against pursuing it when the stakes are sufficiently high. Here, the stakes could not be higher. A biased judge, by definition, is incapable of performing the kind of impartial judicial engagement that is necessary to ensure that we are not deprived of what is rightfully ours arbitrarily, that is, on the basis of mere will rather than the principles of reason contained in our law. The difference between impartial and biased adjudication is thus the difference between the rule of law and the rule of man.  

The Court’s careful probing of individual judicial bias in Williams was thus proper. But it is striking that the Court has largely turned a blind eye to the systemic judicial bias that its own doctrines of judicial deference require in entire areas of our law. The Court has stated that judges applying the “rational-basis test,” the default standard of review in constitutional cases, must uphold the government’s actions unless challengers can perform the logically impossible feat of demonstrating that there is no conceivable basis for those actions. The Court has held that reviewing courts must defer to federal executive agencies’ interpretations of statutes when those statutes are “silent or ambiguous” as to whether the agencies may take a particular action, so long as those interpretations reflect “a reasonable choice within a gap left by Congress,” and that courts must accept agencies’ interpretations of regulations that those agencies issue, so long as they are not “plainly erroneous or inconsistent with the regulation.” Because the rational-basis test is used to evaluate all governmental burdens on our freedom that do not implicate one of a handful of rights deemed “fundamental” by the Supreme Court—that is to say, most governmental burdens on our freedom—and because the administrative power that federal agencies exercise touches nearly every aspect of our lives in some fashion, government officials often act as judges in their own cause. As a consequence, the principle honored by the Court in Williams is more honored in the breach than the observance.  

Nonetheless, one must start somewhere, and Williams is a most welcome decision. Impartial adjudication is not merely desirable—it is essential to our freedom from arbitrary power. The Court’s opinion in Williams is an affirmation of a principle that is not only deeply rooted in our nation’s history and tradition but critical to ensuring that people are not deprived of what is rightfully theirs here and now. It is also a model of the kind of adjudication that is required to vindicate that principle—and, indeed, the rule of law itself.