With few exceptions, the federal courts of appeals have jurisdiction over—and only over—“final decisions” of the district courts. Yet there is a little-known but highly consequential rule, known as the “administrative remand rule,” that an order remanding an administrative agency’s decision under the Administrative Procedure Act is final only as to the agency. That is, only the agency may appeal a remand, and if the agency does not appeal, neither the plaintiff nor any intervenor may do so.

The administrative remand rule thus acts as an appellate gatekeeper in almost all cases involving remands to federal agencies. It accordingly shapes much of the administrative law coming out of the federal courts of appeals. The rule also provides insight into how the courts of appeals conceive of finality and deference to agency decisionmaking. Despite all this, the academic literature discusses the administrative remand rule only generally and only in the context of larger concerns about finality and federal jurisdiction. Academics and practitioners alike must ask themselves whether the rule, as it currently exists, makes sense.

This Article describes the administrative remand rule’s origins, formulations, and purposes. The rule emerged in the 1970s and 80s to promote judicial economy by avoiding piecemeal appeals, even at the cost of injustice in particular cases. Today, the rule still largely serves that wise purpose, but it is not without problems. The federal circuits have all adopted the rule—but very different versions of it. A nonagency party’s ability to appeal a remand thus varies significantly from circuit to circuit and, in some cases, even within the same circuit. The rule is problematic in certain categories of cases—most notably where a remand is likely to be meaningless or where it facilitates ongoing harm to parties or resources. And the rule underscores larger concerns in the fields of federal jurisdiction and administrative law, including geographic forum shopping, temporal forum shopping, politicized agency decisionmaking, inconsistent judicial review, and reduced access to the courts. This Article argues for modest adjustments that would make the administrative remand rule more consistent across the circuits, better balance judicial efficiency and justice, and promote better agency decisionmaking.

* Assistant Professor of Law and Co-Director, Environmental Law Clinic, Stanford Law School. All errors are my own. Nothing worth doing is done alone, and this Article is no exception. I am grateful to Anne Joseph O’Connell, David Freeman-Engstrom, and Daniel Ho for their expertise and guidance; to my colleague Deborah Sivas for her wisdom; to Aaron Avila, Norm Dupont, Jared Fish, and Robbie Beahrs for their close review and helpful edits; and to my family, friends, and partner for their faith. Aaron, my friend and former colleague, bears special mention; we had originally intended to co-author this Article, but he graciously agreed to let me pursue this project on my own. Finally, I am immeasurably grateful to the Stanford Law Review board and editors for selecting this Article for publication, suggesting insightful edits, and bringing it across the finish line. Their care and dedication are unmatched.