Skepticism on the Supreme Court toward administrative authority has evolved into open hostility over the course of the past year in two cases related to the COVID-19 pandemic. The legal vehicle was not, as widely expected, rejection of Chevron’s deference rule or a reanimation of the nondelegation doctrine. Instead, it was formal elevation of the “major questions doctrine” into a substantive canon of construction. This new canon significantly curtails not only executive power (via agencies) but Congress’s legislative authority—and, ultimately, democratic control of policy. It adds a new veto point to the American political system, licensing judges to reject any delegation of power they deem economically or politically significant with little regard for statutory text. The only remedy is a super-clear statement in legislation, similarly subject to judicial discretion. For such major cases, the Court has shifted from deference to antideference, actively antagonistic to delegated power. By its architects’ own admission, this canon is simply the nondelegation doctrine in disguise. It threatens to cripple the administrative state, particularly in emergencies and in areas of evolving science, such as pandemics and climate change.
Over roughly the last two decades, it has become increasingly clear that a majority on the Supreme Court aims to reduce the power and reach of the administrative state in the American constitutional order. Most observers of this trend have focused on two potential changes in doctrine: an end to the practice of deferring to agency interpretations of ambiguous statutes (i.e., Chevron deference) and a revival of the principle that some Congressional delegations of power to agencies are so broad as to violate the Constitution (i.e., nondelegation). These two judicial forbearance doctrines have remained stable pillars of the administrative state for decades. A shift in either would reallocate authority over substantial parts of American law and American life from agencies, the President, and Congress to the Court. Despite strong rumblings, neither of these doctrinal changes has happened—at least not yet and not officially. Chevron remains good law (albeit severely weakened at the Supreme Court level), and the nondelegation doctrine’s slumber was narrowly preserved in 2019’s Gundy v. United States.
But in two recent cases striking down agency actions related to the COVID-19 pandemic—a CDC eviction moratorium and an OSHA vaccine-or-test mandate for large employers—the Court made an equally significant but almost completely unheralded anti-administrative doctrinal change. In so doing, it has arrogated to itself broad discretionary power to reject delegations of authority to administrative agencies without openly altering any doctrinal principle. Instead, the anti-administrativists have stolen a march via expansion of the “major questions doctrine” into a substantive canon of statutory construction.
The traditional major questions doctrine emerged in the 1990s, imposing different rules for statutory interpretation in “major” cases, i.e., those that rise above some level of political or economic significance. Until recently, it operated to deny deference to certain agency interpretations of law—that is, courts would not defer to interpretations of ambiguous statutory terms in “major” cases to which they might have deferred in lower-stakes cases. Chevron and its early progeny had shifted interpretive authority from courts to agencies—it was a “counter-Marbury for the administrative state,” as Cass Sunstein famously called it. The major questions doctrine, among other doctrines and practices reducing the scope of Chevron, clawed some of that power back.
The doctrine has been widely criticized for its indeterminacy, counter-democratic allocation of power from agencies to judges, and other alleged failings, though I have previously argued that it might, paradoxically, have benefited agency authority insofar as it protected Chevron in “normal” cases. Whatever its effects, the doctrine’s influence was limited, largely because it appeared only rarely.
But in recent cases—beginning with Utility Air Regulatory Group v. EPA in 2014 and maturing in the COVID cases—the major questions doctrine escaped the confines of Chevron to operate as an independent, substantive canon of statutory construction. The Court now requires Congress to “speak clearly when authorizing an agency to exercise powers of vast economic and political significance.” Cass Sunstein and Lisa Heinzerling have previously noted this nascent doctrinal shift, but the Court did not openly adopt it until the recent COVID cases. In both cases, the Court rejected agency authority on the grounds that Congress had failed to speak sufficiently clearly. In neither case did it even cite Chevron.
This shift from major questions doctrine to canon is subtle but powerful. More than a further pullback from Chevron deference, it is a reversal of it. Chevron gives agencies some range of interpretive authority when statutes are ambiguous. The major questions doctrine discards that deference, allowing courts to engage directly with statutes (and, therefore, with Congress). But the major questions canon is actively hostile to agency assertions of authority, allowing courts to reject agency interpretations in “major” cases of statutes that are insufficiently unambiguous. The major questions canon is thus a super-Marbury for the administrative state. Where the earlier major questions doctrine shifted a reviewing court from a deference regime to one of rough neutrality, the new canon further shifts from neutrality to antideference.
Nor did the Court announce or acknowledge the shift from doctrine to canon. Instead, the Justices act as if the canon is a long-settled part of the Court’s administrative law doctrine. In one sense this cannot be true—severance of the major questions inquiry from Chevron is a recent innovation. But it is right in another sense: the major questions canon is in fact simply the nondelegation doctrine masquerading as a principle of statutory interpretation. The traditional major questions doctrine was a nondelegation avoidance doctrine; now elevated to substantive canon, that separation has collapsed.
The major questions canon is therefore not (or at least not just) an assertion of judicial power over a modern administrative state. Instead, it asserts power over Congress—and, by extension, over popular rule and representative government. The normal legislative process is no longer adequate for “major” delegations. The new canon is a purely judicial creation, with indistinct and arbitrary boundaries that appear to shift to match the policy preferences of the judges applying it. And it is powerful, trumping statutory text and the Court’s standards for granting preliminary relief.
It is also a new, extra-constitutional veto point in an American political system already crippled by a surfeit of them. Whether that system can effectively respond to short-term emergencies like pandemics or longer-term challenges like climate change hangs in the balance. In 2022 the Court will again consider the EPA’s authority to regulate greenhouse gas emissions in West Virginia v. EPA, a case I and others have highlighted as a potential vehicle for further erosion of Chevron or reinvigoration of the non-delegation doctrine. Armed with the major questions canon, neither is necessary for the Court to impose its veto on the administrative state.
The major questions doctrine emerged relatively recently, in a pair of late-1990s Supreme Court cases challenging agency regulations: MCI Telecommunications Corp. v. AT&T Co. and FDA v. Brown & Williamson Tobacco Corp. By the time these cases were decided, the Court had long since established a doctrine of deference to agency interpretations of law, first as a standard, then as a rule in Chevron v. NRDC. Chevron’s domain was never universal, however, and over time, the Court reduced its scope. What came to be known as the major questions doctrine was one such carve-out.
Deference’s foundation is implied delegation—the assumption that Congress intends for agencies to fill statutory gaps. Just two years after Chevron, then-Judge Stephen Breyer suggested that this assumption should be discarded in “major” cases. A decade later, the Court would adopt this principle (though Breyer himself would not).
In MCI, Justice Scalia rejected a change in telecommunications rate policy by the FCC, denying deference to the agency in part because the regulation was, he claimed, a “radical or fundamental change” to the statutory scheme. The case augured a doctrinal change but did not explicitly announce one; it can be interpreted as a straightforward Chevron case, with the agency’s interpretation simply deemed “unreasonable.”
Six years later, in Brown & Williamson, the Court considered a challenge to the agency’s attempt to regulate tobacco products as “drugs.” In rejecting the agency’s statutory interpretation, Justice O’Connor characterized the case as “extraordinary” because “the FDA has now asserted jurisdiction to regulate an industry constituting a significant portion of the American economy.” Subsequent legislation was also interpreted by the Court to imply that Congress did not intend to grant the FDA authority over tobacco. Justice O’Connor’s opinion included the first clear articulation of the major questions doctrine:
Deference under Chevron to an agency’s construction of a statute that it administers is premised on the theory that a statute’s ambiguity constitutes an implicit delegation from Congress to the agency to fill in the statutory gaps. In extraordinary cases, however, there may be reason to hesitate before concluding that Congress has intended such an implicit delegation.
After Brown & Williamson, if a case was sufficiently important to qualify as “extraordinary,” courts would have sole authority to interpret the statute in question, without deference to any agency view.
After Brown & Williamson formalized the major questions doctrine, the Court seemed to forget about it—it was notably absent from 2007’s Massachusetts v. EPA, leading one scholar to declare it dead. But in King v. Burwell in 2015, the Court confirmed that it was alive and well. In King, the Court considered whether the Affordable Care Act could be interpreted to allow tax credits to be granted to customers of federal insurance exchanges, despite language in the statute that apparently limited such credits to users of state exchanges. The IRS said it could. In a majority opinion by Chief Justice Roberts, the Court declined to defer to the agency reading of the statute for multiple reasons, among them the major questions doctrine:
[Chevron] is premised on the theory that a statute’s ambiguity constitutes an implicit delegation from Congress to the agency to fill in the statutory gaps. . . . In extraordinary cases, however, there may be reason to hesitate before concluding that Congress has intended such an implicit delegation. . . . Whether . . . credits are available on Federal Exchanges is . . . a question of deep “economic and political significance” . . . [H]ad Congress wished to assign that question to an agency, it surely would have done so expressly.
Having denied deference to the agency’s reading, the Court proceeded to its own statutory analysis, but it ultimately confirmed the agency’s reading, based on traditional textual analysis. This makes King a perfect illustration of the traditional major questions doctrine: it operates within Chevron or, perhaps more accurately, as a threshold question before reaching Chevron—a Chevron Step Zero. But whether a case is “major” has no effect on the Court’s textual analysis.
The traditional, Chevron-focused major questions doctrine would not hold for long, however—five years later, the COVID cases would elevate it to a substantive canon. But this change was less sudden than it appears. Its roots lie in a separate line of cases dating back to the 2000s—or perhaps the 1980s. By the time King was decided, the Court had already begun moving toward a major questions canon, albeit under a different name.
Twice in complex non-delegation cases, the Court has come close to adopting a major questions canon, but it would not stick. In Industrial Union Department, AFL-CIO v. American Petroleum Institute (The Benzene Case) in 1980, Justice Stevens’ plurality opinion includes a passage that looks much like the major questions canon: “In the absence of a clear mandate in the Act, it is unreasonable to assume that Congress intended to give the Secretary the unprecedented power over American industry that would result from the Government’s view . . . .” Sunstein identifies this as the canon’s doctrinal root. But if so, the Court itself did not acknowledge it: Neither MCI, Brown & Williamson, nor King cite the Benzene Case.
Two decades later, in Whitman v. American Trucking Ass’ns, Justice Scalia’s majority opinion adopted a similar principle: “Congress . . . does not alter the fundamental details of a regulatory scheme in vague terms or ancillary provisions—it does not, one might say, hide elephants in mouseholes.” Scalia cited MCI and Brown & Williamson, but this is a significant and unacknowledged doctrinal shift to a clear statement rule, perhaps identical to the major questions canon. But Whitman, too, was not followed by progeny adopting a major questions canon. The Court would occasionally cite the “elephants in mouseholes” principle, but only once in a recognizable major questions case involving a delegation to an agency.
In both Whitman and the Benzene Case, the Court also considered disinterring the nondelegation doctrine, only to ultimately decline. But the threat of doing so, and the clear statement rules the cases appear to articulate, were threats to agency authority (and to Congress’ power to delegate)—Chekov’s guns placed silently on the wall.
We are therefore left with something of a puzzle. Whitman and the Benzene Case both offer statements that look like the major questions canon and suggest a connection to nondelegation, but in neither case does the Court acknowledge any shift in doctrine, nor is either case followed by progeny that apply such a canon—at least not until recently. Whitman and the Benzene Case are therefore (at most) important precursors.
B. Utility Air
For Sunstein and Heinzerling, canonization came in Utility Air Regulatory Group v. EPA in 2014. Justice Scalia’s majority opinion applies Chevron’s two-step process, identifying statutory ambiguity but refusing to defer because the agency’s interpretation was deemed unreasonable—making Utility Air an unusual Chevron step two case, but not doctrinally innovative, at least superficially. But Scalia’s reasoning was remarkably aggressive:
When an agency claims to discover in a long-extant statute an unheralded power to regulate “a significant portion of the American economy,” we typically greet its announcement with a measure of skepticism. We expect Congress to speak clearly if it wishes to assign to an agency decisions of vast “economic and political significance.”
If the roughly contemporaneous King v. Burwell is a clear articulation of the traditional major questions doctrine, this looks like an equally clear articulation of the major questions canon. Scalia cites both the Brown & Williamson and the Benzene Case, pulling two of the canon’s historical threads together.
But (contra Sunstein and Heinzerling), in my view the canon had not yet fully arrived. Although the passage above appears to adopt a major questions canon when read in isolation, Utility Air still operates within Chevron’s deference framework. In practice this distinction makes little or no difference to case outcomes: it is inconceivable that the Court would conclude an agency interpretation is unreasonable on major questions grounds yet still accept it. But the major questions doctrine in Utility Air is not enough to resolve the statutory interpretation inquiry. Justice Scalia’s opinion engages in substantial further textual analysis before rejecting the agency’s reading. The fact that Utility Air was followed a year later by King confirms that it did not shift—or was not yet understood to have shifted—the Court from major questions doctrine to canon.
But even if Utility Air did not canonize major questions, it opened the door. Jody Freeman called the passage above a “rhetorical flourish” and a “‘red meat’ reference to potential government overreach that some Justices toss to their conservative audiences.” But there was more to it than rhetoric, she suggested, calling the case “full of troubling hints and clues as to the Court’s skeptical mood” and armed with “legal improvised explosive devices” Utility Air was Scalia’s second such doctrinal IED—just as in Whitman, he advanced a new legal principle, profoundly dangerous to agency authority, in a case that gave the agency its preferred substantive result.
The Court did not immediately go through the door it had opened in Utility Air. The case was followed by an increase in the rate at which the Court cited the “elephants in mouseholes” principle, but not by cases adopting the major questions canon. Utility Air was ahead of its time. But by 2021, the Court’s personnel had changed: Justices Scalia, Kennedy, and Ginsburg had been replaced with Gorsuch, Kavanaugh, and Barrett. Of these, Gorsuch would become the major questions canon’s leading advocate.
And Utility Air did inspire Kavanaugh to adopt the canon—in US Telecom Ass’n v. FCC (2017), then Judge Kavanaugh invoked what he called a “major rules” doctrine. Because the net neutrality rule at issue was, in Judge Kavanaugh’s view, “one of the most consequential regulations ever issued,” it required “clear congressional authorization.” What Kavanaugh called the “major rules” doctrine is nothing more than the major questions canon. Once elevated to the Supreme Court, Kavanaugh confirmed his view that major questions was a canon, “closely related” to nondelegation.
The Court’s formal adoption of the major questions canon came in two recent decisions staying emergency agency actions arising from the COVID-19 pandemic.
In Alabama Ass’n of Realtors v. Department of Health and Human Services in 2021, the Court considered a nationwide moratorium on evictions issued by the Centers for Disease Control (CDC). In deciding six-to-three that a stay was warranted, the Court applied the standard framework, which requires the stay applicant to make “a strong showing that he is likely to succeed on the merits” and show that equitable factors such as irreparable injury and the public interest weigh in favor of a stay. The Court’s consideration of these equitable factors was brief, almost entirely subsumed into the merits analysis.
The merits, in the Court’s view, were resolved by application of two substantive canons: the longstanding federalism canon and the new major questions canon. The latter was triggered because, in the Court’s view, the moratorium asserted agency powers of “vast ‘economic and political significance’” and “a breathtaking amount of authority.”
Having concluded this was a major case, the Court considered whether the statute in question, the Public Health Service Act, was sufficiently clear to authorize the CDC moratorium. In relevant part, it authorized the CDC to “make and enforce such regulations as in [its] judgment are necessary to prevent the introduction, transmission, or spread of communicable diseases. . . .” In the Court’s view, this statutory language was far from adequate, partly because the relatively pedestrian examples of regulatory actions given in the statute make it “a stretch” to read the superficially broad language to authorize halting evictions.
Reasonable minds might differ over the degree to which the examples given in the statute limit the scope of agency authority in a pandemic or the degree to which an eviction moratorium is outside that scope. But the majority’s trump card was the major questions canon. Ambiguity is not enough, says the canon. This is not a Chevron case. Instead, Congress must speak clearly to delegate significant authority. And in the Court’s view, the statute lacked the required clarity or robustness: it was “a wafer-thin reed on which to rest such sweeping power.”
The opinion gives no guidance on what more Congress needed to have done in the statute, beyond warning that the agency’s reading would leave “no limit” on its powers (suggesting a nondelegation problem). The opinion does suggest two additional factors were significant. One is the absence of past regulations based on the same statutory provision that (in the Court’s view) “beg[i]n to approach the size or scope of the eviction moratorium.” This suggests a “use it or lose it” element of delegated authority—if an agency receives a broad grant but construes it narrowly, or lacks a reason to use it fully, those powers may be lost.
The Court also points to lack of post-enactment legislation as evidence of statutory intent and, by extension, requisite clarity. Because Congress had imposed a temporary eviction moratorium by statute, then allowed it to lapse, Congress (the Court reasoned) implicitly denied the CDC power to re-impose it under preexisting law. This echoes and extends Brown & Williamson—in both cases, post-enactment Congressional behavior informs interpretation of the relevant statute, but Alabama Realtors applies that principle to post-enactment inaction.
The complete absence of Chevron from the Alabama Realtors opinion confirms that it is a major questions canon case. Whether to defer to the agency’s reading is never in question—the Court simply goes about interpreting the statute de novo, with analysis dominated by the major questions canon’s clear statement rule.
A few months later, the Court removed any remaining doubt that the major questions doctrine had been elevated to a substantive canon. In National Federation of Independent Business v. OSHA (The Vaccine Case) in early 2022, the Court considered an OSHA emergency rule requiring large employers to either “ensure their workforces are fully vaccinated or show a negative test at least once a week.” Challengers alleged the rule exceeded OSHA’s statutory authority to issue workplace standards “reasonably necessary or appropriate to provide safe or healthful employment . . . .” and requested a stay.
In a per curiam opinion similar to Alabama Realtors, the same 6-3 majority of the Court granted the requested stay, relying exclusively on the major questions canon. Like Alabama Realtors, the Vaccine Case opinion does not cite Chevron. Instead, it applies the new two-step major questions canon analysis, considering first whether the regulation is sufficiently significant: “This is no ‘everyday exercise of federal power.’ It is instead a significant encroachment into the lives—and health—of a vast number of employees. ‘We expect Congress to speak clearly when authorizing an agency to exercise powers of vast economic and political significance.’” The regulation having qualified as a major question, the Court then considered whether the statute “plainly authorized” it. The Court concluded it did not, with a single paragraph of cursory statutory analysis:
The Act empowers the Secretary to set workplace safety standards, not broad public health measures. See 29 U.S.C. §655(b) (directing the Secretary to set “occupational safety and health standards”. . . . Confirming the point, the Act’s provisions typically speak to hazards that employees face at work. And no provision of the Act addresses public health more generally, which falls outside of OSHA’s sphere of expertise.
The Court further distinguished COVID from “work related dangers” within OSHA’s purview, instead likening it to “day-to-day dangers that all face from crime, air pollution, or any number of communicable diseases” and concluding the vaccine-or-test requirement was “strikingly unlike” past agency practice. This analysis is profoundly unpersuasive; OSHA has long regulated general risks that appear in the workplace—one need look no further than the Benzene Case itself to find an OSHA regulation of air pollution.
The opinion is strikingly similar to Alabama Realtors. It suggests that past agency practice with old statutes informs current bounds of authority, and relies on implied repeal by congressional inaction, this time extending to a non-binding Senate vote disapproving of the regulation. Where the Court had considered equitable factors only briefly in Alabama Realtors, it simply refused to do that analysis at all in the Vaccine Case, declaring “[i]t is not our role to weigh such tradeoffs.” Resolution of the case therefore collapsed entirely into the merits analysis.
Justice Gorsuch’s concurrence, joined by Justices Alito and Thomas, discusses the canon in more detail. For Gorsuch, the lack of statutory clarity is driven not by text, but by its age (50 years) and by Congressional inaction. But the concurrence goes deeper into the canon’s roots and rationale:
Not only must the federal government properly invoke a constitutionally enumerated source of authority to regulate . . . . It must also act consistently with the Constitution’s separation of powers. And when it comes to that obligation, this Court has established at least one firm rule: “We expect Congress to speak clearly” if it wishes to assign to an executive agency decisions “of vast economic and political significance.” We sometimes call this the major questions doctrine.
For this, Gorsuch cites Alabama Realtors and his own dissent in Gundy, in which he would have reanimated the nondelegation doctrine by discarding the “intelligible principle” standard in favor of stricter review. Completing the doctrinal loop, the Gundy dissent cites the major questions doctrine as an illustration of the Court’s attempts to rein in agency overreach. Though what Gorsuch describes there is the traditional major questions doctrine, in neither Gundy nor the Vaccine Case does he acknowledge or explain the shift from doctrine to canon. But in both, Gorsuch grounds the rationale for major questions in nondelegation. Indeed, in his Vaccine Case concurrence, Gorsuch writes that were it not for the major questions canon, the vaccine-or-test mandate would be unconstitutional on nondelegation grounds. The canon, Gorsuch says, is just another way to get at the same separation of powers problem: “The nondelegation doctrine ensures democratic accountability by preventing Congress from intentionally delegating its legislative powers to unelected officials. . . . The major questions doctrine serves a similar function by guarding against unintentional, oblique, or otherwise unlikely delegations of the legislative power.” This suggests the canon is a mere Congressional error correction measure. But Gorsuch continues, clarifying that agencies are the canon’s real target:
Later, the agency may seek to exploit some gap, ambiguity, or doubtful expression in Congress’s statutes to assume responsibilities far beyond its initial assignment. The major questions doctrine guards against this possibility by recognizing that Congress does not usually “hide elephants in mouseholes.” In this way, the doctrine is “a vital check on expansive and aggressive assertions of executive authority.”
Gorsuch thereby merges the Whitman “elephants in mouseholes” line of cases with the major questions mainstream. The “vital check” language comes from then-Judge Kavanaugh’s US Telecom dissent, in which he advanced his “major rules” doctrine. For the same sentence, Justice Gorsuch bizarrely cites my 2016 paper in which I defend major questions as protective of Chevron. At no point in that paper (and certainly not in the cited portion) do I argue that the doctrine is justified as a means to restrain agencies, much less that they are “expansive and aggressive” extralegal actors.
Justice Gorsuch’s opinion is only a concurrence, joined by two other justices. But it and the majority opinion apply the major questions canon in the same way, with Justice Gorsuch merely going into more depth. And there is no doubt that it now operates as a canon, leaving its Chevron constraints behind—none of the opinions in the Vaccine Case or Alabama Realtors even mention Chevron. In this switch to a canon, the major questions doctrine has subsumed the “elephants in mouseholes” line of cases and emerged, it appears, as the nondelegation doctrine in other clothes.
The COVID cases established the new canon. Two further climate-related cases in 2022 may further illustrate its significance.
Just weeks after the Vaccine Case, Judge James Cain of the Western District of Louisiana issued an order enjoining use of federal government social cost of carbon estimates. As in the COVID cases, Judge Cain’s opinion relies heavily on the major questions canon to conclude challengers are likely to succeed on the merits, with only the briefest consideration of the equities. Judge Cain concludes that estimation of the social cost of carbon is a “major” action which lacks clear authorization from Congress. Deference is never on the table. Chevron is not cited. The opinion cites virtually all of the major questions precedents, all the way back to Breyer’s 1986 article. Judge Cain’s order was struck down on appeal on standing grounds; its final fate remains unclear—but the case illustrates that the major questions canon is not restricted to the Supreme Court.
The Court will have another opportunity to apply the canon in West Virginia, a third challenge to EPA’s authority to regulate carbon emissions under the Clean Air Act. The case stems from the agency’s attempts to regulate fossil-fueled power plants, beginning with the Obama EPA’s Clean Power Plan, That rule was stayed by the Court, then repealed by the Trump EPA and replaced with the weaker Affordable Clean Energy (“ACE”) rule. Environmental groups and states challenged the ACE rule, and the D.C. Circuit vacated it in early 2021. In its ruling, the D.C. Circuit denied Chevron deference to the EPA and rejected arguments from the agency that the major questions doctrine compelled its narrow reading of the statute. Subsequently, the Biden EPA declined to defend the ACE rule, and indicated it had begun working on a replacement. Surprisingly, given the lack of an actual rule in place backed by the government, the Court granted certiorari.
West Virginia’s framing of the question presented invites a major questions ruling:
[Whether, . . . i]n . . . an ancillary provision of the Clean Air Act, . . . Congress constitutionally authorize[d] the Environmental Protection Agency to issue significant rules—including those capable of reshaping the nation’s electricity grids and unilaterally decarbonizing virtually any sector of the economy—without any limits on what the agency can require so long as it considers cost, nonair impacts, and energy requirements[.]
So framed, this presupposes that climate rules for power plants are “significant” because of their economic effects, triggering the major questions canon. And by calling the relevant provision “ancillary,” it also presupposes that the statute lacks the requisite clarity. Furthermore, it also raises the specter of constitutional limits—i.e., nondelegation.
West Virginia’s brief refers to “the major questions canon of construction,” defined as the requirement that “Congress must delegate with unmissable clarity if it intends to give an agency economy-transforming abilities to decide major questions. . . .” The brief treats the canon as a constant doctrinal principle dating back to the Benzene Case, rooted in nondelegation, and a response to “the danger posed by the growing power of the administrative state.” It does not mention Chevron deference. Another petitioner’s brief goes further, suggesting that the major questions doctrine requires rejecting any statutory interpretation that would convey “vast power to decide matters of great economic or political significance”—if accepted, this would openly merge major questions with the non-delegation doctrine. The respondents’ briefs argue that the doctrine should not apply, but none question the canon.
Perhaps the Court will dismiss West Virginia v. EPA on standing or other threshold grounds, as observers ranging from environmental groups to center-right law professor Jonathan Adler have encouraged. If not, the major questions canon is highly likely to play a role: major questions was mentioned more than forty times in oral arguments, by every Justice except Gorsuch. Judging by the COVID cases, the likely result is a decision by the Court crippling the ability of the federal government to act on climate change.
The Court has moved in a sharply anti-administrative direction in the last decade. Some Justices appear to view agencies (or at least agencies advancing policies they do not like) as rogue actors of questionable constitutionality. So far, this shift on the Court has primarily occurred via sharp decline in Chevron deference. There is some appetite on the Court for overturning Chevron and for reanimating the nondelegation doctrine.
In this environment, a shift from major questions doctrine to major questions canon might seem like small potatoes. But it is at least as great of a constraint on the administrative state and, ultimately, on Congress and popular rule. These dangers were apparent under the traditional major questions doctrine, but canonization has radically exacerbated them.
The Supreme Court has allowed the administrative state to function by ceding at least some interpretive authority to agencies since the New Deal (and probably since the founding). This forbearance empowered not just agencies but Congress, allowing it to legislate without constantly being second-guessed by the courts. Chevron crystallized this deference regime into a rule in the 1980s. The various exceptions to Chevron that emerged in the decades that followed, including the major questions doctrine, eroded that rule, but only partially altered the interbranch balance of power because they did not challenge the basic ability of Congress to delegate authority to agencies. Even if deference to an agency reading was not due in “major” cases, Congress remained in charge. The Court’s role in both major and “normal” cases, derived from the APA, was merely to interpret statutes as written, blocking agency actions that exceeded their delegated authority. The major questions doctrine therefore moved the court from a position of deference to one of neutrality with respect to the agency. In so doing, it undermined Chevron, both directly by excluding major cases, and indirectly by drawing into question the assumptions of agency competence and implied delegation on which Chevron relies. If you think Chevron is good, this is grounds for criticizing the doctrine, though an alternative interpretation I have suggested is that it protected Chevron in lower-stakes cases.
Canonization ends any debate over the relationship to Chevron. As illustrated by the COVID cases, Chevron disappears entirely. More than that, Chevron’s deference rule is reversed—agency interpretations of statutes that trigger “major” questions aren’t just denied deference, they are actively suspect. The Court has moved from neutrality to antideference. For the same reason, the major questions canon cannot fulfill the Chevron-shielding role I have earlier suggested was its sole redeeming feature.
The most prominent critique of the major questions doctrine has been that its boundaries are unclear, unpredictable, and arbitrary. The Court never says what makes a case “major” or “extraordinary,” other than a general reference to “economic and political significance.” Whether the regulatory action at issue is a break with past agency practice seems to be another factor. But all of these criteria are woefully indeterminate. Even pedestrian cases can be described as politically controversial—“[e]lephants and mouseholes are in the eye of the beholder.” In practice, whether a case qualifies as “major” is a thin line with “no metric . . . for making the necessary distinctions.” Then-judge Kavanaugh recognized as much in his US Telecom dissent, admitting that “determining whether a rule constitutes a major rule sometimes has a bit of a ‘know it when you see it’ quality.”
The major questions canon cases have further muddied boundaries, adding to the list of factors making a case “major” while providing only perfunctory analysis of why factors new and old are met by the regulation in question. In Alabama Realtors, the Court highlights the economic impact of the eviction moratorium, but its analysis of that impact is paper-thin—the $50 billion cost the Court cites is not an estimate of the moratorium’s impact, but an at best tangentially related figure: the total rent relief funding already provided by Congress during the pandemic. The Vaccine Case Court’s “analysis” of the regulation’s significance is even more perfunctory, little more than a bare assertion that “[t]here can be little doubt that OSHA’s mandate qualifies.” The only supporting fact is that an estimated 84 million Americans would have to get vaccinated or test regularly, which the Court characterizes as “no ‘everyday exercise of federal power.’” This suggests (but does not say) that regulations affecting individual autonomy, perhaps especially medical autonomy, will be more readily deemed “major” questions. In neither of the COVID cases is the political salience of the pandemic and policy responses discussed, though it must surely be a factor.
The COVID cases are not unique—the Court’s analysis of economic impacts in major questions cases is often slipshod and simplistic. As Richard Revesz notes, the Court often focuses on “decontextualized” regulatory costs. This creates perverse incentives for agencies, encouraging them to choose regulations with lower cost even if their preferred option has higher net benefits, or to fragment regulatory actions to avoid their being characterized as “major.” Revesz also argues that reliance on political salience “place[s] the courts in the uncomfortable (and untenable) position of determining what quantum of public attention is sufficient to divest an agency of a previously held power.”
After decades of major questions cases, the Court has failed to give any clear or consistent guidance on its boundaries. As Revesz suggests, the Court’s criteria for determining majorness “fail to offer ‘limited and precise standards that are clear, manageable, and politically neutral.’” As Justice Kagan asked at oral argument in West Virginia, “how big does a question have to be?” One is left with the distinct impression that a major question is nothing more than a challenge to a regulation that is personally unpalatable to at least five Justices. Under the major questions canon, it is not just statutory interpretation, but the standard of review that comes under judges’ full control.
The indeterminate scope of the major questions doctrine sharply undercuts the claim that it promotes democratic legitimacy. To be sure, it is far from the first substantive canon or clear statement rule. But many other substantive canons have clear (or at least clearer) boundaries. They are also well-established, and Congress can and does therefore legislate in their shadow. Congress is well aware (for example) of the federalism canon, can recognize when it is altering the federal/state balance of authority, and can therefore legislate more explicitly as the canon requires. On the contrary, it is hard or impossible to predict what will become a major question in the future. In many contexts, there is no way for Congress to know when delegated authority may be used, how consistently it will be interpreted, and when it will become politically controversial—and therefore which delegations demand additional clarity to satisfy the major questions doctrine.
The Court has given similarly scant guidance on what Congress must do to satisfy the doctrine’s clear statement rule. One might expect the answer to lie in the statutory text, analysis of which the traditional version of the doctrine purports to compel. Direct engagement with text (rather than agency views) is among the canon’s professed virtues.
But the COVID cases show that, in practice, the canon licenses remarkably atextual statutory analysis. Even apparently broad grants of authority may not be upheld. Old statutes are suspect, even those consistently in use, if the agency is changing its past practice or relying on an allegedly “ancillary” provision to do something new. Past agency practice informs the scope of authority—delegated authority appears to be “use it or lose it.” For example, in Alabama Realtors, the statute’s age and an alleged lack of similar past moratoria is relevant. That emergency CDC pandemic powers should be expected to be used only rarely is not. Implied repeal by later legislation, normally disfavored, is also common in major questions cases. In both COVID cases, even post-enactment inaction by Congress effected implied repeal, somehow reducing clarity of previous delegations. Justice Gorsuch’s Vaccine Case concurrence would revive a selectively-imposed one-house veto.
As Anita Krishnakumar observes, such statutory analysis in the COVID cases is “decidedly atextual.” Reliance on substantive canons is not new, of course, but “[u]sually, when the Justices invoke a substantive canon, they also at least attempt to analyze the statute’s text—even if only to conclude that the text is ambiguous, thereby (conveniently) necessitating recourse to a substantive canon.” Not so in major questions cases, where the allegedly extraordinary nature of the regulations at issue trumps any need to seriously engage with statutory text.
As Krishnakumar further notes, this atextuality is particularly surprising coming from the Court’s professed textualists. Justice Scalia often warned of the mischief enabled by substantive canons, exceptions to Chevron’s framework, and the temptations of the nondelegation doctrine. Freed from the chains of doctrine, he feared, judges would be free to impose their preferences, hidden by ostensibly neutral principles.
But whatever Scalia said about guarding against the temptations of judicial policymaking, he engaged in his share in major questions cases. He accepted Brown & Williamson’s reliance on post-enactment implied repeal, introduced the at best tenuously textual “elephants in mouseholes” principle in Whitman, and ushered in the major questions canon with his Utility Air opinion. Judicial forbearance was just fine for other people. Scalia more than any other Justice was the major questions canon’s architect.
In hindsight, it is possible to reevaluate Justice Scalia’s professed dislike of doctrinal innovations permitting judicial aggrandizement. Serving on a divided Court, such tools could be used by both sides. Not so for Scalia’s successors today. With a six-to-three majority, the Court’s anti-administrativists need fear no turnabout. Atextual opinions like Justice Gorsuch’s Vaccine Case concurrence cannot now be weaponized by the other side of the bench. Justice Kagan, a professed textualist, thinks this has gone too far, expressing frustration in recent oral arguments:
[W]e’re going to be thinking about the supposed major questions canon. There are other canons.
. . . Some of them help the government. Some of them hurt the government. . . . Maybe we should just toss them all out . . . .
. . . .
. . . I think kind of we should, honestly. Like, what are we doing here?”
But at least the atextualism of the major questions canon lets us see it for what it is: a license for judicial aggrandizement, in the hands of a profoundly anti-administrative Court. We have been down this road before: Chevron itself requires courts to determine statutory clarity. If the Court’s Chevron jurisprudence is any guide, the Justices are unable to do so consistently—Chevron’s statutory clarity standard has relentlessly grown, swallowing the deference rule. There is little reason to think the major questions canon’s clear statement rule will be applied any more consistently. None other than Justice Kavanaugh has claimed that judges “cannot make that initial clarity versus ambiguity decision in a settled, principled, or evenhanded way” and that judges instead decide cases by “selectively picking among a wealth of canons of construction.”
In forsaking text, judges applying the major questions canon have wrested control. Congress is no longer in charge of its own statutes. Because what makes a case major and what makes a statute sufficiently clear are entirely within the discretion of judges, there are no meaningful limits to the canon’s reach.
Structurally, the major questions doctrine creates a new policy veto point. Political polarization and the rise of the filibuster have made legislating difficult. The rise of the major questions canon means legislation that has navigated all the other constitutional and political veto points may then be rejected by the courts—not because it is explicitly unconstitutional, nor because an agency has gone further than the text of the statute allows, but because at least five justices have deemed it “major” legislation that is not sufficiently clear. And as Judge Cain’s social cost of carbon decision shows, that judicial veto can be imposed by a single district court judge.
Even worse, this new veto point makes it harder to navigate the existing ones. Surviving the canon (if it is possible at all) requires explicit delegation. But it is much harder to get legislative consensus behind explicit language. Congress may delegate to agencies not only because they have greater expertise, but also to avoid deciding a politically difficult point, or to delay doing so—that is normal, not illegitimate. If Congress can never delegate an allegedly “major” question but instead must answer it explicitly, the result may be that legislation cannot pass.
Even attempting to overcome the canon’s veto can undercut Congress’s power. New legislation will likely be required to satisfy the clear statement rule. But if politics, procedural barriers, veto points, or sheer complexity make it difficult or impossible for Congress to re-authorize or expand authority, then that can be weaponized under the major questions canon as indicative of Congressional intent not to do so—retroactive repeal by inaction. This is especially ironic because the difficulty of passing new legislation is often what inspires agencies to look to preexisting authority in the first place. But even if Congress does manage to pass new legislation, it may not be enough. The Court could still rule that it is insufficiently clear to grant authority, permitting only incremental regulation today but nothing more innovative or expansive in the future, or that it delegates more power than the Constitution allows.
Blake Emerson (among others) argues that the traditional major questions doctrine undermined “democratic-constitutional values . . . by failing to respect the deliberative capacities of administrative agencies.” It also reallocated interpretive authority from agencies controlled by Congress and the President to unelected and life-tenured judges, making government less representative and responsive.
The doctrine’s architects allege that it is necessary to restrain agencies that, in their view, are a threat to democratic accountability and constitutionally guaranteed freedoms, ever ready to expand their reach beyond the powers Congress has granted. For reasons that are never explained, Congress is asleep at the wheel and unable to restrain agencies. Thus, the task is left to judges. But if, instead, you view administrative government as democratically legitimate, with its authority flowing from the people through Congress and the President, the doctrine is a threat to those virtues, concentrated in those cases with the greatest political salience.
Canonization further increases judicial power. The traditional version of the doctrine could perhaps be defended on the grounds that it was a judicially created exception to implied delegation, itself arguably a judicial creation. But under the major questions canon, the agency exits the statutory interpretation picture, leaving the courts to deal directly with the statute and inviting judges to substitute their views not just for the agency’s, but for Congress’s. The very democratic and separation of powers principles frequently cited by the Court to justify the doctrine—that the elected legislature, not unelected bureaucrats must make the laws—are violated when judges with even greater removal from the electorate exercise a legislative veto. For David Driesen, the major questions canon is nothing more than “juristocracy”:
In important cases, the Court has abandoned the role that the Administrative Procedure Act assigns it—checking the executive branch when it contravenes the policies that Congress and the President have approved. Instead, it has assumed the role of constraining the faithful execution of the law based on unpredictable judicial fiats.
If the anti-administrativists want to constrain or roll back agency power, they should propose doing so openly and contest elections on that basis, not give courts a veto over policy.
The major questions canon also encodes a status quo bias, potentially crippling the ability of the federal government to deal with the most important public policy problems. Pandemics and other emergencies require flexible authority that can be deployed quickly and at scale. Congress cannot anticipate every policy measure that might be needed, and while it might be ideal if it authorizes them with specific new legislation, that is difficult in normal times and likely impossible in a crisis. Broad delegations of authority to the executive and/or to agencies (with Congressional and judicial oversight) are the only available response. Similarly, long-term problems in areas of evolving scientific understanding like climate change require regulatory durability and flexibility. The success of the Clean Air Act depends on those features, and broad delegations of authority that evolve with new information are at the core of modern administrative government.
But the major questions canon makes all these delegations suspect. Congress must separately and explicitly authorize every “major” delegation. If the Court means to reshape the post-New Deal order by making all such delegations illegitimate on nondelegation grounds, it should say so and face the political consequences, not hide behind an ostensibly neutral canon of statutory interpretation.
Moreover, the doctrine is further biased because it is only triggered when agencies assert authority to regulate in some new way, never when they decide not to regulate. As Lisa Heinzerling puts it,
The major questions doctrine quietly embeds [a] preference [for agency inaction] in the Court’s approach to statutory interpretation.
. . . [This] renders the doctrine not only political, but nonsensical. . . .
. . . [W]hether an agency is deciding not to act on an important problem, or deciding to act on that problem, it is deciding the very same question, with the same degree of economic and political significance. Only the direction, not the magnitude, of these decisions is different.
The major questions canon veto, unlike a Presidential veto, may be impossible for Congress to override with new legislation. A threat of judicial veto via the nondelegation doctrine remains, sometimes implicit and sometimes (as in Gorsuch’s Vaccine Case concurrence) explicit. The Court has sometimes acknowledged and scholars have long identified a connection between the major questions doctrine and nondelegation. The APA grounds judicial review of agency action in statutes. As Driesen argues, by creating a novel and atextual major questions canon, the Court has overstepped that authority. If the canon is not a pure judicial creation, it must therefore be grounded in the Constitution. The only plausible basis is nondelegation. The canon is no longer merely a related principle or an avoidance doctrine, it is the nondelegation doctrine, without speaking its name. Critics of administrative power have celebrated Gorsuch’s Vaccine Case concurrence as a “novel, unified theory of separation of powers,” making major questions and nondelegation “two distinct sides of the same coin.” This also helps explain the canon’s atextualism. As Justice Barrett has argued, if the source is the Constitution, fidelity to statutory text is secondary.
In fact, the canon’s indeterminacy makes it even broader than nondelegation. To reject a delegation of authority to an agency on nondelegation grounds requires the Court to say why it is too broad to survive constitutional scrutiny. Admittedly this is an imprecise exercise, but at least it’s something. To reject a delegation under the major questions canon, a Court need only say that it is meets a fuzzy majorness standard and fails to meet an even murkier clarity standard.
The major questions canon takes an entire class of cases not only out of Chevron’s deference regime, but out of any meaningful textual or contextual analysis. Instead of avoiding the difficulties of applying the nondelegation doctrine, the major questions canon achieves the same purpose sub rosa. Control over the bounds of the principle is entirely in the hands of judges, with little clarity and no limiting principle. In short, it licenses judicial policymaking while professing to protect Congress and the people from agency overreach. The impacts on democratic accountability and the effectiveness of administrative government are likely to be profoundly negative.
The major questions canon purports to be a matter of principle. It is in reality a matter of power, an assertion of unbounded judicial supremacy in the most important administrative law cases. The danger of major questions juristocracy is that judges—specific people, with lifetime tenure—are empowered to enact their political preferences. Gillian Metzger warned of a “1930s Redux”, a boldly anti-administrative Court relitigating interbranch power struggles thought resolved in the New Deal Era. But the major questions canon gives the Court powers that its 1930s counterparts never dreamt of.