Ending the Reign of Slot Machine Justice

Document Type

Article

Publication Title

New York University Annual Survey of American Law

Abstract

Federal appellate courts employ a random assignment system to select the circuit judges who will serve on any particular three-judge panel. The premise of this system is that random assignment is the fairest way of selecting judges, in light of resource constraints that make the use of full-circuit panels for every case impracticable.This paper contends that the random assignment system, while perhaps efficient, is certainly not fair. Winning a case in the U.S. courts of appeals hinges too much on luck, and not enough on the merits. This system produces slot machine justice, in which the outcome crucially turns on the three judges selected to hear the case. Aside from problems of basic fairness, the random selection of judges creates other serious problems for the judiciary as well. These include the decreased predictability of the law and extreme ideological results. As the circuit courts' caseload has increased, these problems have become even worse. Prominent jurists have noted that "there is an increasing likelihood that aberrant precedents will be influential" and that courts have become "less predictable and more quirky." These problems, to a large extent, can be traced to the random assignment system.

For this reason, the random assignment system should be replaced with a system in which the judges are assigned on the basis of the parties' preferences. Such a system will fundamentally improve the justice of the federal judiciary's decisions, as every federal appellate decision will reflect, as closely as possible, what the "average" three-judge panel on that federal circuit would decide. It will also mean that there is less variation between decisions made by the same circuit, enabling each court of appeals to develop more coherent and consistent, as well as less ideologically driven, bodies of jurisprudence.

Publication Date

2000

Link to Full Text

Share

COinS