A federal statute authorizes the various U.S. Courts of Appeals to consist of between six and twenty-eight judges in regular active service. Except when the federal appellate courts hear or rehear cases en banc, these courts consider and decide cases in panels of three judges.

Every U.S. Court of Appeals has adopted the policy that in the absence of a change in the governing law from a controlling authority, no later panel is authorized to overrule the decision of an earlier panel. Rather, if an earlier panel's decision happens to have reached the wrong result, either rehearing en banc or a poll of all active judges to see whether they are willing to sign-off on a change in governing precedent in the absence of rehearing en banc is necessary.

Notwithstanding the policy that a later three-judge panel is powerless to overrule the decision of an earlier three-judge panel in the absence of a change in the controlling law, experience reveals that sometimes two different three-judge panels of the same federal appellate court will issue irreconcilable decisions on a given question of law. Once that happens, what is the next three-judge panel from the circuit to do when confronted by the very same question that divided two earlier three-judge panels?

In the vast majority of federal appellate courts, the answer is clear: The earliest three-judge panel opinion to resolve a question controls, and any intervening conflicting decisions are to be ignored. The St. Louis, Missouri-based U.S. Court of Appeals for the Eighth Circuit does not follow this first-in-time rule, however. Rather, in the Eighth Circuit, a later panel faced with a conflict among earlier Eighth Circuit rulings is free to choose to follow whichever strand of authority it finds most persuasive. Kilmartin v. Dormire, 161 F.3d 1125, 1127 (8th Cir. 1998).

On August 1, 2003, a three-judge panel of the U.S. Court of Appeals for the Fourth Circuit confronted this very question in the case of McMellon v. United States, 338 F.3d 287 (4th Cir. 2003). The majority, in an opinion by Circuit Judge William B. Traxler, Jr., ruled that it was bound to follow the earliest Fourth Circuit case to have definitively resolved the question presented, even though later Fourth Circuit cases had reached a different result and most of the other circuits to have considered the question also disagreed with the Fourth Circuit's earliest ruling on the subject.

In a fascinating dissenting opinion, however, Circuit Judge Paul V. Niemeyer advanced the following argument. The rule that later panels should not disregard the controlling holdings of earlier panels is one of policy, not of federal judicial power. Later decisions that violate the policy are as much binding decisions of the federal court of appeals as the earliest decision to resolve an issue. Accordingly, a panel that is faced with conflicting earlier rulings from three-judge panels of its own court has no choice but to violate the policy that a later panel should not reach an outcome different from that reached by an earlier panel. Therefore, the panel faced with the question today should side with whichever of the earlier decisions it views as most correct, instead of adhering to the decision that was earliest in time, whether most correct or not.

On October 8, 2003, the Fourth Circuit granted the federal government's petition for rehearing en banc in the McMellon case. The en banc Fourth Circuit unquestionably will have the authority to follow whichever line of authority it deems to be most persuasive, without regard to which three-judge panel ruling issued earliest. I hope, however, that the en banc Fourth Circuit will also endeavor to address the approach to dealing with conflicting panel precedents that Judge Niemeyer raised. And, for the reasons that follow, the Fourth Circuit should reject both Judge Niemeyer's approach and the approach that the Eighth Circuit follows.

Federal appellate courts are of course free to adopt a policy that would allow any three-judge panel to resolve a question whichever way it believed best, without regard to whether a previous three-judge panel of the court has already decided that very same question. As explained at the outset, however, no federal appellate court has chosen to proceed in this manner, valuing the existence of fixed legal rules over the risk that it will be especially difficult and burdensome (through the en banc process) to repair the occasional errors that arise.

A policy that would allow a third three-judge panel to choose to follow whichever of two earlier conflicting three-judge panel rulings that it found preferable would necessarily give rise to the creation of more unresolved intra-circuit conflicts than would a policy that makes the first precedential three-judge panel ruling on an issue definitive in the absence of rehearing en banc. For if one three-judge panel disagreed with how another three-judge panel resolved an issue, that second panel could issue a conflicting ruling, thereby allowing later panels the option of ruling either way on the question. If the second panel's conflicting outcome reaches a result that would lack the votes to prevail before the court en banc, later panels that preferred the second panel's result would be able to decide cases in a manner that defied the views of a majority of the court's active judges. It would undoubtedly take a while before the issue in conflict was accepted for resolution before the court en banc. In the interim, there would exist uncertainty over what law governed.

On the other hand, under the approach followed by the vast majority of circuits, if a later panel believed that an earlier panel's on-point ruling had reached an incorrect result, the later panel would have no choice but to seek en banc resolution. Yet at all times there would be the ability to ascertain what the law actually was, as reflected in the earliest decision by federal appellate court in question.

In a perfect world, three-judge panels of a given federal appellate court would never issue conflicting decisions, and the very first three-judge panel ruling to resolve an issue of law would always reach the right result. In the real world, however, intra-circuit conflicts do arise, and the first three-judge panel ruling to decide a question does not always provide the right answer.

Nevertheless, it is when an intra-circuit conflict exists that the need for certainty concerning what rule is the law is at its greatest. Only by enforcing the policy that the first three-judge panel ruling to decide an issue provides a definitive resolution unless reversed by the court en banc can a federal appellate court satisfy the need for certainty. As a result, the en banc Fourth Circuit in the McMellon case should reject Judge Niemeyer's proposed approach, and the Eighth Circuit in an appropriate case should also abandon its minority view that a three-judge panel, when faced with conflicting Eighth Circuit authority, may follow the decision is believes best, instead of being bound by the decision that issued the earliest.

This article is reprinted with permission from the October 13, 2003, issue of The Legal Intelligencer © 2003 NLP IP Company.

Back