Recently, we’ve received a new justification for the major questions doctrine (“MQD”) in the form of Ilan Wurman’s Importance and Interpretive Questions, which aims to recast the doctrine as a linguistic canon. Wurman’s project ask whether the MQD can find a defensible justification as an insight into either (1) how Congress drafts statutes or (2) how ordinary language is used when people in conversation delegate authority. You get the sense from the paper that this project was born of Ilan’s pessimism on the MQD’s pedigree as a substantive/constitutional canon. He says that several version of the MQD are not defensible or are only defensible if the doctrine doesn’t—as most scholars think it does—require clarity over and above plain meaning. What follows is an eclectic collection of evidence that gives Wurman’s piece a kitchen-sink feeling. I don’t mean that as a criticism; Wurman collects evidence from many sources to try to see if anything sticks. I think it’s a creative work that will give everyone a lot to deal with in the literature. But at the same time, it can be dizzying how Wurman toggles back and forth between different sorts of evidence; from (what I think of as) intentionalist evidence about congressional staffers’ views on drafting, to a theoretical work on language. So you’ve got to come into this one with some flexibility.
The evidence includes the following:
- Evidence of drafting practices—Wurman relies on the results from a question in Bressman and Gluck’s survey (“B-G Survey”) of congressional drafters from a decade ago.
- Theoretical work on ordinary language—Wurman relies on the work of Ryan Doerfler, whose paper, High-Stakes Interpretation, discussed the role that importance can play in ordinary conversation.
- The mischief rule—in a move that should be familiar to readers of the blog, Wurman argues that the MQD might serve a similar function to the mischief rule.
- Historical evidence—Wurman points to the Necessary and Proper Clause, agency law, contract law, and the historical work of Louis Capozzi.
Personally, most of this evidence is not for me because of various methodological commitments and my own assessment of the evidence. For starters, it’s unclear whether the B-G Survey is either relevant or particularly strong evidence. You can get a sense for what I mean here and here. You should also check out Chad Squitieri’s work, with its emphasis on what kind of evidence about Congress textualists ought to look to, here. Textualists had a lukewarm take on the study because it’s unclear that it adheres to textualists’ methodological commitments. (I doubt a lot of academically refined textualists will take this to heart, but that’s a conversation I’m not part of.) What’s more, even if you accept that the study is relevant evidence, and not just a kind of intentionalism that’s verboten, it’s not clear that it’s particularly strong evidence. The B-G Survey included the following question, which was posed to congressional drafters:
Q55. What kinds of statutory ambiguities or gaps do drafters intend for the agency to fill? (mark one)
a – Ambiguities/gaps relating to the details of implementation
b – Ambiguities/gaps relating to major policy questions
c – Ambiguities/gaps implicating questions of major economic significance
d – Ambiguities/gaps implicating questions of major political significance
e – Ambiguities/gaps relating ot the preemption of state law
f – Ambiguities/gaps relating to the division of labor between state and federal agencies when both are given implementation roles
g – Ambiguities/gaps relating to omissions in the statute
h – Ambiguities/gaps relating to the agency’s area of expertise
i – Other (explain)
To state the obvious, answers “b,” “c,” and “d” above are all modeled against the MQD. According to Bressman and Gluck, 60% of respondents chose the answers that aligned with MQDs premise. 28% of respondents said that “drafters intend for agencies to fill ambiguities or gaps relating to major policy question.” 38% of respondents said that “drafters intend for agencies to fill ambiguities or gaps relating to questions of major economic significance.” And 33% said that “drafters intend for agencies to fill ambiguities or gaps relating to questions of major political significance.” For starters, 40-60 split in the responses is nothing to scoff at. It’s hardly a runaway for the MQD-esque answer choices. But also, it’s important to remember that the congressional drafters in the B-G Study exhibited a general ignorance of many basic linguistic conventions and doctrines. With a few exceptions like Chevron deference, the staffers flunked on recognizing basic canons with long pedigrees like the rule of lenity. With that background in mind, it’s unclear to me whether a congressional drafter might have merely picked one of the non-MQD answers with a more anodyne instinct in mind—that Congress establishes the general outlines of federal policy that federal agencies will fill. Few would dispute—and this is a point relevant to my scholarship—that Congress routinely uses underdetermined language when “speaking” on socially and economically salient questions as a descriptive matter. While it might seem true in the abstract that drafters only “intend” on having agencies fill in the details, that “intention” may not match up with the realities of actual cases as they arise. What’s more, congressional drafters may not perceive major questions or agree with the courts’ characterizations of majorness. The problems with relying on this question are pretty endless. I have a lot of trouble crediting the results—at least for doctrinal purposes—of a single question posed to congressional staffers who seem to be opining on intent in the abstract and who lack the authority by themselves to make law that’s binding on the public.
A lot of the other evidence rehashes points that have already been discussed on this blog. Wurman points to the mischief rule to argue that the MQD might be understood as really working in a similar way. If that’s true, I don’t know what independent work it’s doing or the Court’s distinct modes of application that do not match the mischief rule proper. I offered my own take on the connection between the mischief rule and the MQD here. I do think the mischief rule operates kind of like the MQD if you squint, and that this observation could be important for textualists in an academic sense. The MQD has long been justified through a descriptive claim about how congresses (plural) “speak” in statutes in the abstract—they “speak” clearly when addressing major questions. The mischief rule provides a similar context to statutory interpretation with the descriptive claim that congresses (plural) legislative against mischiefs. That’s an important perspective but it’s all for not because the comparison does not actually provide any evidence that Congress “speaks” clearly when addressing major questions, a claim that any good student of Congress will find either hysterical or maddening. If the MQD is reframed in a similar vein as the mischief rule, it nonetheless fails because the context it provides is obviously wrong. (more on that in a piece I’m working on) And, finally, I’m not going to relitigate the Capozzi work here. You can read about it here, here, and here.
The rub for me is the portion of Wurman’s paper dealing with the work of Ryan Doerfler. Doerfler explained that “to say that the meaning of a statute is ‘clear’ or ‘plain’ is, in effect, to say that one knows what the statute means.” He continues, “[a]s numerous philosophers have observed, . . . ordinary speaks attribute ‘knowledge’—and, in turn, ‘clarity’—more freely or less freely depending upon the practical stakes.” “In low-stakes situations speaks are willing to concede that a person ‘knows’ this or that given only a moderate level of justification.” But if the stakes are high, “speakers require greater justification before allowing that someone ‘knows’ that same thing, holding constant that person’s evidence.”
For Wurman, Doerfler’s work connects nicely to the MQD:
The application to some of the major questions cases is intuitive, at least as to the threshold question of ambiguity. The meaning of an “occupational health and safety standard” may seem straightforward in an ordinary, relatively low-stakes regulation of the workplace. We might “know” that the statute permits such regulations, or find the statute is “clear” in this regard. But when dealing with a regulation that imposes a requirement on millions of individuals, that persists beyond the workplace itself, and which requirement is itself hugely controversial, it is intuitive to think that ordinary speakers would in fact demand more epistemic confidence before concluding that the statute in fact authorizes such a requirement. In other words, ordinary readers and speakers are more likely to find the statute ambiguous in that context than in a relatively lower-stakes context.
This is an interesting attempt to find a plausible justification for the MQD. Of course, you’d want to back up the intuitive appeal behind Doerfler’s work. (Word on the street is that folks are already on that project—more to see over the next few years on the empirical end of things.) Do all linguistic canons have empirical backing? Of course not. But many were adopted before we had the means for testing their premises. We wouldn’t want to make doctrine turn on linguistic insights that have not been sufficiently tested for the same basic reasons that we don’t drown people to test if they’re witches. We’ve come a long way!
On a more basic level, I’m skeptical of this approach, or more specifically, that it is anything like a proper linguistic canon or gives the MQD an acceptable pedigree. As Doerfler’s work suggests, and as Wurman seems to realize on page 45 of his draft, this linguistic insight cuts both ways. Doerfler’s work suggests that if an agency will foist a “hugely controversial” requirement on a ton of folks, the agency ought to seek out clear authority from Congress. But agencies aren’t the only folks who are in conversation with Congress. So are judges. We could reframe the question from the vaccine-or-test case as follows: should federal judges strike down an agency action if it poses significant consequences (e.g., the death of thousands of people). The upshot of Doerfler’s work is that judges should not strike down agency actions if they will be causing tremendous consequences—like, you know, exacerbating the devastation caused by a once-in-a-century pandemic. Wurman sees this problem and offers up the following:
Fortunately, the legal system already contingently addresses this question of framing differently: because agencies are creatures of statute, they must demonstrate authority for their actions. Thus, as a matter of constitutional structure, the agencies are the asserters of the legal claim and bear the burden of proof. Even if one does not buy this distribution of proof burdens, it is enough to say that the question addressed here is the meaning of the statute, which is not necessarily the same question as whether the agency has acted unlawfully; and on that former question, the insights about high-stakes interpretation militate in favor of a major questions canon of some sort.
Do I find this convincing? No, not really. I take as a given that when Congress gives an agency authority X, it’s really giving a federal agency the authority to do a range of things from agency actions A, B, C, and D. That’s just a view that turns on the hopeless indeterminacy and underdeteminacy of language. When a judge has to interpret a federal statute de novo, A, B, and C might be plainly within the authority granted to the agency. The remaining question is whether D, the agency action taken, is outside the realm of the permissible. We call that ultra vires. But in answering that question about agency action D that entails terrible consequences—death, destruction, yada yada—there’s no good reason I see that Doerfler’s linguistic insight does not counsel in favor of a kind of deference. Indeed, this kind of sounds like how many judges approached judicial review during a time in which we had a healthier separation of powers dynamic. They would strike down government actions only if those actions were plainly impermissible. That approach has fallen out of favor, but that’s the result of a social-legal movement that has cast deference as judicial abnegation. It’s all contingent. There are no right answers.
I think we’re all used to the observation that different canons can be in conflict and lead to different results. But it’s rarer that a linguistic insight seems to augur in favor of both deference and tighter judicial review simultaneously. I think this poses a question that should be on our minds; maybe not every insight into ordinary language lends itself to the construction of workable judicial doctrine. At the very least, we should not fool ourselves that the MQD as linguistic canon has some kind of clean pedigree. If the insight works both ways—in favor of both deference and stricter judicial review—then the choice between those options is a normative one. We’re back to where we started, the MQD’s feeble case as a substantive or constitutional canon. In the end, I’m trying to keep an open mind but am incredibly skeptical that this ordinary meaning provides a working pedigree for the MQD. At the same time, I would urge folks to monitor the empirical/theoretical work on this subject that’s bound to come out over the next few years. And I urge all the blog readers to grapple with the Wurman paper themselves. Although we depart on the merits of recasting the MQD as a linguistic canon, I think it’s a thought-provoking work that we should all engage with.