Flipping a coin to decide a case is among the most serious forms of judicial misconduct. Yet judges react quite differently to other types of lotteries. Judges tend to tolerate or encourage deliberately random decisions in nonjudicial settings ranging from military drafts to experimental welfare requirements. Equally striking, most adjudicators now embrace randomization within their own institutions: they commonly use lotteries to assign incoming cases to each other. This practice creates a remarkable tension. Because adjudicators vary in competence and ideology, randomizing their case assignments will effectively randomize outcomes in a subset of merits decisions. We might then ask whether the typical adjudicative system is sometimes at war with itself, condemning courtroom coin flips while operating backroom lotteries with similar effects.
This Article attempts to defend the judicial treatment of randomized decisions. The Article begins by investigating the concept of randomization and the leading justifications for randomizing social decisions. It then offers a consequentialist defense for the pattern of judicial reactions to official lotteries. This defense admits that case-assignment lotteries have merits-randomizing effects, and it accepts that a merits-randomization ban might be the best rule for fallible judges facing public relations problems. Even so, random case assignment can be justified based on three consequences, aside from the convenience of judges: (1) fairly allocating to litigants the tragically scarce and indivisible resource of judicial excellence, (2) roughly honoring the politics of the judicial appointments process, and (3) continuing a natural experiment on the determinants of judicial behavior. These arguments cannot explain why adjudicative institutions developed as they did. But they can exploit various social benefits that the system has produced, in a sense, randomly.