Chevron deference, which requires courts to defer to reasonable agency interpretations of statutes that the agency administers, is a central doctrine of administrative law. Despite its importance, Chevron rests on uneasy ground. Critics have raised various legal objections to the doctrine. One is that requiring courts to defer to agency interpretations of statutes violates the Article III power to say what the law is; another is that permitting agencies to render binding interpretations of statutes violates Article I; a third is that the justification for Chevron— that Congress has implicitly delegated interpretive power to agencies—rests on a fiction and violates the APA’s command that delegations of that sort be explicit.

When doctrine conflicts with law, two general choices are available. One is to come up with a new rationale for the doctrine that avoids the legal objections. Backfilling of this sort has a long history; Holmes gave ancient examples in The Common Law. Chevron deference itself an example. The original Chevron decision based deference largely on agency expertise and political accountability, but later decisions have based Chevron deference on the conclusion that, in various organic statutes, Congress has implicitly delegated to agencies the power to issue binding interpretations of statutes. That post-hoc rationale doesn’t fare very well because it raises all the objections mentioned above.

The other way to address legal problems with doctrine is to change the doctrine to avoid the legal shortcomings. That is the approach I’ve proposed in a new paper. It argues that we should modify Chevron from a doctrine of statutory interpretation to a doctrine of remedies. Instead of requiring courts to defer to agency interpretations of statutes, this Remedial Chevron would limit the power of the court to vacate agency action. A court could vacate agency action based on an erroneous legal interpretation only if the interpretation was unreasonable.

Modifying Chevron in this way would retain a good deal of Chevron’s benefits. It would retain a major role for agencies in policymaking based on statutory interpretations. And it would avoid a lot of the legal objections to Chevron. Remedial Chevron would not run into the Article III problem of judicial deference to interpretations. Courts would not defer to interpretations; they would simply lack the power to vacate based on wrong, but reasonable, interpretations. It would avoid the Article I objection, because it does not rest on the theory that agencies have the power to render binding interpretations. And it raises fewer APA problems than the current doctrine. It would also obviate difficult Chevron step zero questions, and it would provide a clean divide between where Skidmore and Chevron apply.

For those who are interested, the paper is available here.

As always, I’d be most grateful for any comments.