Agency Interpretations, Bell Curves, and Skidmore ooomph under Loper Bright (2/12/25)

Professor Christopher Walker, here, a frequent commentator on administrative law and on deference under Chevron and Loper Bright, has offered an interview on the current scene under Loper Bright.  Judicial Constraints on Agency Action (The Regulation Review 2/9/25), here. In that article, he states, based on his study of a large set of Chevron opinions, that “there was nearly a 25 percentage-point difference in agency-win rates when the courts of appeals applied Chevron deference than when they did not.”

I did a similar study of two smaller sets of opinions but enough to feel comfortable that it was a reasonably fair sample set. My conclusion was different from Walker’s and more in line with Second Circuit Judge Jon Newman’s conclusion that courts often invoke Chevron but do what they want to anyway—that is, interpret as they think is right (the best interpretation). Jon O. Newman, On Reasonableness: The Many Meanings of Law’s Most Ubiquitous Concept, 21 J. App. Prac. & Process 1, 83 (2021) (emphasis supplied), here. If that is right, Chevron was not ever as outcome determinative as people imagined from the rhetoric or the apparent win rate such as Professor Walker posits.

The conceptual model I posit is that, when courts defaulted to what appeared to be an agency win because the interpretation was “reasonable,” many of those cases really involved the courts’ determinations or hunches that the agency interpretations were the best interpretations. That’s the observation Judge Newman made. If that observation is true (I think it is), there should be a higher win rate because the indicated 25% difference in win rates in the Chevron era meant that, often, even usually, an agency win was not that just that the interpretation was reasonable but that the court thought it was best. Stated another way, Chevron was only outcome determinative when an agency not best interpretation was approved under Chevron. Judge Newman (and I) conclude that that was likely significantly less than 50% of the time when courts noised about Chevron.

If one were to posit that Chevron’s range of reasonableness for ambiguous statutes was a normal bell curve, that would mean that (i) roughly 50% would be reasonable and the best interpretations and (ii) 50% would be reasonable but not the best interpretations. That would also mean that, in the Loper Bright era when a court must make the call of best interpretations and not best interpretations, the agency should win roughly half of the cases that would have applied Chevron under the Chevron regime. And, if Judge Newman's observation and my observations from my datasets are correct, the bell curve would not be the normal bell curve.  But assume that it were a normal bell curve. Then, if you factor in some type of Skidmore ooomph for agency interpretations, I suspect that the curve for agency interpretations will shift materially in favor of agency interpretations. So, maybe, agency interpretations prevail as the best interpretations in 60% of the times agency interpretations would have been subject to the Chevron regime. (I don’t know where exactly to calibrate the Skidmore ooomph, but it will likely be something; I chose 60% just 10% higher than 50% to illustrate that that at 10% the agency outcome is 20% as compared to Walker’s agency outcome of 25% under Chevron; that is fairly close and others might calibrate the Skidmore ooomph at higher than 10%.)

That is the basic concept. Many cases that appeared to “apply” Chevron simply applied an agency interpretation the court thought was best. I guess we’ll have to await a statistically significant set of opinions under the Loper Bright best interpretation regime to draw better conclusions. But I would not be surprised if agency win rates are consistent with Professor Walker's win rate statistics in the Chevron era.