A Conceptual Analysis of Chevron Footnote 9’s Approach to (Possibly) Mitigating Chevron Deference (10/15/23; 2/6/24)

I am working on a paper addressing the issue of whether APA § 10(e) of the original Administrative Procedure Act in 1946 (now codified at 5 USC § 706, here). My principal contribution is to bring the tax authorities into the discussion. Tax authorities are important to the discussion but have been overlooked or misunderstood by those writing on the subject.

Today’s blog addresses the commotion about whether rigorous statutory interpretation is a cure, in whole or in part, to so-called "reflexive deference." This topic was originally in the drafts of the paper, but I took it out to slim the paper down and now offer the discussion here.

The cure championed by some (e.g., Justice Kavanaugh) is to deploy rigorous statutory interpretation at Chevron Step One to determine the best interpretation without ambiguity. (Remember that only at Chevron Step Two after determining ambiguity does Chevron deference apply.) This approach is the so-called Chevron Footnote 9 approach based on Chevron’s footnote 9 (Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 843 n9  (1984) (case citations omitted)):

   n9  The judiciary is the final authority on issues of statutory construction and must reject administrative constructions which are contrary to clear congressional intent. If a court, employing traditional tools of statutory construction, ascertains that Congress had an intention on the precise question at issue, that intention is the law and must be given effect.

The notion is that more rigorous interpretation will shrink the scope of the ambiguity and will shrink (or tame) the scope of deference. Deference deniers view this as a positive good to at least partially emasculate Chevron deference.

This blog is inspired by a tax phenomenon at the heart of abusive tax shelters sold with “opinions” written by prominent lawyers and accountants that the key legal edifice (and components) supporting a bullshit tax shelter opinion would “more likely than not” prevail. That type of legal opinion was inspired by the fact-finding concept that preponderance of the evidence meant a finding that, on the evidence, the fact was more likely than not. The fact-finding concept was sometimes conceptualized as a finding that the fact was more than 50% likely based on the evidence. A 50% or lesser likelihood meant the party bearing the burden of persuasion lost on that fact issue. So, in theory, if the fact-finder found that the fact was 51% likely the party bearing the burden of persuasion on that issue wins. I hope at this point you have spotted the problem—what exactly is the difference between 50% and 51% likelihood? Can a fact-finder really perceive that fine a difference in a way meaningful to make a rational fact-finding? Isn’t this a context where there is a range rather than a finite percentage. See in a similar context in fact-finding, John A. Townsend, Burden of Proof in Tax Cases: Valuation and Ranges—An Update, 73 Tax Lawyer 389 (2020), here.

Applying that theoretical concept to “law-finding,” what is the difference between a 51% and a 50% likelihood for a legal opinion. Can any rational or responsible law-finder—whether a judge in a case or a lawyer rendering a legal opinion--make that fine an analysis? Specifically, in the current context, is a judge’s or a lawyer’s belief that the likelihood of being the correct interpretation is 51% (proponent wins) or 50% (proponent loses) meaningful? Is that sliver of difference of 1% (or with finer tuning, .000001%) meaningful to anyone? See e.g., Daniel J. Hemel and Aaron L. Nielson, Chevron Step One-and-a-Half, 84 U. Chi. L. Rev. 757, 781-782 (2017) (using a similar spectrum analysis)

Or is there, as in fact-finding, a range somewhere on either side of 50% in law interpretation that falls in the reasonable range of ambiguity? Now we are back to Chevron and deference that preceded it. Using the percentage likelihood, will not honest assessments of close legal interpretation have a range say from 55% to 45% where a responsible interpreter cannot make a finite legal interpretation of either 51% or 50% and must instead say there is a range of ambiguity?

Imagine a spectrum of confidence levels as to interpretation, where the two ends of a spectrum are 100% and 0%. In the context where deference is in issue, the confidence level is toward the middle of the spectrum because deference requires ambiguity. Say, without a rigorous interpretation, a court is willing to hypothesize the ambiguity range is from 40% and 60%, respectively, with the likelihood for either interpretation in that range (the ambiguity zone). Any way one views ambiguity, there is a range of ambiguity, whether the range is, for example, 30% (65%-35%), 20% (60%-40% or, say with really rigorous interpretation, 10% (55%-45%). (I ask what is the difference in the uncertain enterprise of law interpretation between a 55% likelihood and a 45% likelihood? Even if there is only a 1% spread (which I doubt is possible), there is still a range that honest interpreters will know is there. (Note the key condition that the interpreter be honest; honesty requires a certain level of awareness of interpretive limitations without some arbitrary tie-breaker, say going with the gut (policy preferences) or flipping a coin.)

I suppose that we should deal with a finite determination at 50%. Fact-finders and observers of fact-finders call that equipoise, although it is really a broader percentage range where the fact-finder is unable to find the fact with confidence. Certainly, if deference were limited to exact equipoise (it is not) as to whether the Chevron-qualified interpretation can be finitely determined, deference may not be that objectionable, conceptually at least. Honest interpreters know you can’t slice the bologna that thin. There is going to be a range of ambiguity where, for example, a law-finder may assess the interpretation at, say, 53% but not so certain that the assessment of likelihood cannot be 47%. (If the law-finder is able to do that, I suggest that the law-finder is drawing on powers that mortals do not possess or using something arbitrary, like their own policy preferences or even a flip of a coin, to make the determination.)

The point I hope I have made is that, while recognizing the value of rigorous statutory interpretation at Chevron Step One, it is not the panacea that Chevron Footnote 9 enthusiasts like to imagine in their fantasies of being greater law-finders than they really are. Moreover, I think even with more rigorous interpretation, they will find that the agency interpretation is likely to be the best interpretation which does not require any deference. All of which means that the Chevron applications with non-rigorous statutory interpretations are really masking the courts’ belief that the agency interpretation is best (no deference) but stopping at finding reasonableness (best interpretations are reasonable too!).

Added 2/7/24 10am:

On ranges in confidence levels in statutory interpretation, the following is excerpted from Brett M. Kavanaugh, Book Review, Fixing Statutory Interpretation, 129 Harv L Rev 2118, 2137-2138 (2016) (footnotes omitted):

          First, judges must decide how much clarity is needed to call a statute clear. If the statute is 60-40 in one direction, is that enough to call it clear? How about 80-20? Who knows?

          Second, let's imagine that we could agree on an 80-20 clarity threshold. In other words, suppose that judges may call a text "clear" only if it is 80-20 or more clear in one direction. Even if we say that 80-20 is the necessary level of clear, how do we then apply that 80-20 formula to particular statutory text? Again, who knows? Determining the level of ambiguity in a given piece of statutory language is often not possible in any rational way. One judge's clarity is another judge's ambiguity. It is difficult for judges (or anyone else) to perform that kind of task in a neutral, impartial, and predictable fashion.

          I tend to be a judge who finds clarity more readily than some of my colleagues but perhaps a little less readily than others. In practice, I probably apply something approaching a 65-35 rule. In other words, if the interpretation is at least 65-35 clear, then I will call it clear and reject reliance on ambiguity-dependent canons. I think a few of my colleagues apply more of a 90-10 rule, at least in certain cases. Only if the proffered interpretation is at least 90-10 clear will they call it clear. By contrast, I have other colleagues who appear to apply a 55-45  [*2138]  rule. If the statute is at least 55-45 clear, that's good enough to call it clear.

          Who is right in that debate? Who knows? No case or canon of interpretation says that my 65-35 approach or my colleagues' 90-10 or 55-45 approach is the correct one (or even a better one). Of course, even if my colleagues and I could agree on 65-35, for example, as the appropriate trigger, we would still have to figure out whether the text in question surmounts that 65-35 threshold. And that itself is a difficult task for different judges to conduct neutrally, impartially, and predictably.

          The simple and troubling truth is that no definitive guide exists for determining whether statutory language is clear or ambiguous. In a considerable understatement, the Supreme Court itself has admitted that "there is no errorless test for identifying or recognizing 'plain' or 'unambiguous' language." Professor Ward Farnsworth has elaborated persuasively on that point, arguing that "[t]here are no rules or clear agreements among judges about just how to decide whether a text is ambiguous."

 See also William Baude & Ryan D. Doerfler, The (Not So) Plain Meaning Rule, 84 U. Chi. L. Rev. 539, 559-560 (2017) (discussing and citing Kavanaugh’s article).