Anita Krishnakumar has posted this draft on SSRN.  Here is the abstract:

How much deference — or what kind — should courts give to longstanding agency interpretations of statutes? Surprisingly, courts and scholars lack a coherent answer to this question. Some legal scholars have assumed that longstanding agency statutory interpretations are treated with heightened deference upon judicial review, and some federal courts have made statements suggesting that this is the governing legal rule. But in practice, federal court review of longstanding agency interpretations — at both the court of appeals and the U.S. Supreme Court — turns out to be surprisingly erratic. Reviewing courts sometimes note the longevity of an agency’s statutory interpretation as a plus factor in their deference analysis, but at other times completely ignore or dismiss an agency interpretation’s longevity. Moreover, judicial rhetoric about the relevance of longevity in the review of agency statutory interpretations is inconsistent from case to case.

What makes this doctrinal incoherence particularly remarkable is that courts usually care much more about the predictability of statutory interpretations and about upsetting settled institutional practices. In fact, in two analogous contexts — judicial interpretations of statutes and historical executive branch practice in the constitutional arena — courts accord strong precedential effect, or a presumption of correctness, to established legal rules. This Article compares federal courts’ chaotic treatment of longstanding agency statutory interpretations with the precedential effect they give to longstanding judicial interpretations of statutes and the historical “gloss” effect that courts give to past executive practice in constitutional interpretation. It argues that longstanding agency interpretations of statutes are equally deserving of heightened judicial deference and that, at the very least, federal courts’ disparate treatment of such interpretations — without acknowledging or justifying the distinction — is troubling. The Article advocates that longstanding agency interpretations should be entitled to precedential effect by reviewing courts and outlines how such an approach might work.