Waiving Chevron

Introduction

The Supreme Court’s decision in Chevron, U.S.A., Inc v Natural Resources Defense Council, Inc1 has been a boon for federal agencies. The decision created “Chevron deference,” which gives agencies considerable leeway to interpret statutes. Under Chevron’s two-step formula, courts will defer to an agency interpretation if (Step One) the statute is ambiguous and (Step Two) the interpretation is “based on a permissible construction of the statute.”2 So long as the agency selects a meaning that could reasonably be taken from (frequently opaque) statutory language, the agency’s interpretation will stand. In effect, Chevron adds the interpretive power—the power “to say what the law is”—to the litany of other agency powers.3

Yet from time to time, agencies may want to shed their interpretive power and not seek Chevron deference, a move I call “Chevron waiver.” This can be done by stipulating to the court that the agency no longer wishes to defend its interpretation or does not claim Chevron deference. An agency may choose to waive Chevron for any number of reasons. Perhaps new information has come to light. Maybe the agency is under pressure from an engaged and intense lobby. Or maybe, after a presidential transition, new agency leadership wants to reverse the previous administration’s policy. The latter occurred in 2017. Global Tel*Link v Federal Communications Commission4 concerned caps on the rates charged by prison inmate–calling services imposed by the Federal Communications Commission (FCC) during the Obama administration. In the wake of President Donald Trump’s inauguration, the FCC declined to defend its interpretation and seek Chevron deference. A panel of the DC Circuit respected the agency’s position and did not apply the familiar two-step framework despite the presence of an intervenor arguing for deference.5 Six weeks later, however, the panel had second thoughts and amended its opinion. The revised opinion held that the Obama FCC’s interpretation was unreasonable all along and would therefore fail Chevron’s second step, so the court did not need to reach the issue of what to do when the agency does not seek Chevron deference.6

This case illustrates that not all Chevron questions have been answered. In particular, judges have not grappled with Chevron waiver—a question that goes to the heart of why Chevron deference exists. Each of the three judges on the panel wrote separately about Chevron waiver, but not one cited an authority. Instead, the judges resorted to conclusory statements that Chevron deference “would make no sense”7 or “would be inappropriate” when the agency does not seek deference.8 The panel retreated to intuition rather than relying on authority because, in fact, there was no authority to cite.

If permitted, there is good reason to think that agencies will make periodic use of Chevron waiver, especially in an age in which administrative powers are increasingly used as partisan weapons.9 But should agencies be allowed to waive Chevron? This Comment explores the law behind the DC Circuit’s intuitions and rejects the court’s initial conclusion that waiver should be allowed.

This Comment proceeds as follows. Part I describes the unique circumstances of Global Tel*Link and generalizes to describe how an agency can waive Chevron and the set of circumstances that could lead to Chevron waiver. Chevron waiver would act as another element of Chevron’s Step Zero, the inquiry into whether the Chevron framework should be applied. This Comment therefore evaluates Chevron waiver as a Step Zero question. Part II looks to the case law formulating Chevron Step Zero, but the standard doctrinal formulations for Step Zero address only when an agency is entitled to deference and shed little light on whether that deference is mandatory. Hence, Part III examines the Chevron waiver question not from Step Zero’s doctrinal formulations but from the principles working beneath the surface of the Step Zero cases. Specifically, the Supreme Court’s Step Zero cases consider why a hypothetical reasonable Congress intended its statute’s ambiguities to be resolved by the agency. The reasons for Chevron deference—agency expertise, accountability, and rigorous agency procedures—all help determine when Chevron applies. As Part III discusses, the particular risk of Chevron waiver is that it allows the agency to circumvent the rigorous and accountability-enhancing, policy-formulating procedures set up by Congress. Given that overriding concern, this Comment urges a prohibition on Chevron waiver.

I. The Anatomy of Chevron Waiver

The Chevron doctrine increases the power of agencies.10 Under Chevron, courts defer to the agency that administers the statute. The obvious question is why an agency would ever want a reviewing court to give greater scrutiny to its interpretations, lessening its power. This Part argues that there are common scenarios in which an agency can be expected to want to waive Chevron and that, if courts recognize the power to waive Chevron, agencies will add Chevron waiver to their policymaking toolkit. First, this Part describes Global Tel*Link, in which the issue of Chevron waiver came before the DC Circuit as a matter of first impression. Part I.B then discusses how an agency can go about waiving Chevron. Finally, Part I.C describes the common conditions under which an agency will want to waive Chevron and why, if Chevron waiver is permitted, it could become a feature of administrative practice.

A. Global Tel*Link: Chevron Waiver in a Case of First Impression

Section 276 of the Communications Act11 empowers the FCC to “promote the widespread deployment of payphone services to the benefit of the general public.”12 The statutory definition of payphone services includes what the industry terms “inmate telephone service[s]” offered in correctional institutions.13 The Act directs the FCC to “take all actions necessary . . . to prescribe regulations that establish a per call compensation plan to ensure that all payphone service providers are fairly compensated for each and every completed . . . interstate call.”14 Rates for inmate calling services were historically left unregulated by the FCC.15

Inmates face significantly higher calling rates than they would at a public payphone. Before the first federal rate caps were introduced, inmates in some states paid as much as $17.30 for a fifteen-minute long-distance collect call.16 One reason rates are so high is that service providers compete for contracts at correctional facilities, often by bidding up site commissions, the shares of revenue that go to the correctional facility.17 Rates are also higher inside prisons and jails because of security concerns. Texas, for instance, requires providers to integrate voice biometric screening.18 Of course, higher calling rates limit an inmate’s ability to contact family members and attorneys.19

After lobbying from the families of inmates, the FCC decided to regulate inmate calling services. Following a 2012 notice of proposed rulemaking and comments from interested parties, the FCC adopted interim rate caps for inmates’ interstate calls in 2013.20 Several parties, including inmate calling–service providers, challenged the 2013 order in the DC Circuit, which granted a partial stay of the regulation and subsequently held the case in abeyance as the FCC began a second round of rulemaking on inmate calling services.21

The new order—promulgated in 2015 after traditional notice-and-comment rulemaking and challenged in Global Tel*Link—adopted tiered rate caps, with more stringent caps reflecting the lower cost associated with providing service at larger correctional facilities.22 The rate caps were formulated without factoring in the cost of site commissions, in the hope that the regulation would trigger change-of-law clauses in inmate calling–service contracts; renegotiations between providers and facilities; and a new, site commission–free market norm.23 Most dramatically, the FCC claimed authority over intrastate inmate calls, applying the rate caps to intrastate calls.24 The regulation was promulgated with all three Democratic appointees voting in favor and the two Republican-appointed commissioners dissenting. Then-Commissioner Ajit Pai challenged the majority’s reading of the organic statute’s requirement that providers be “fairly compensated.”25 The language comes from the Telecommunications Act of 1996,26 which was Congress’s response to the monopoly that dominated the telecom industry. Pai reasoned that, given this context, Congress was concerned with service rates that, because of subsidies or discriminatory conduct, were “too low to ensure fair compensation,” not too high.27 In addition, Pai explained the source of the FCC’s plenary rate-setting authority was not the specific payphone provision in § 276 but rather § 201 of the Communications Act, which limited the authority to interstate and international service.28

Global Tel*Link originated in several providers’ petitions for judicial review of the 2015 order. The providers challenged the 2015 order on a number of grounds, including that the rate caps were arbitrary and capricious and that the FCC had exceeded its statutory authority.29

Presidential politics transformed a run-of-the-mill case about agency authority and allegedly heavy-handed regulations into a case about whether an agency could waive its considerable discretion under Chevron. Global Tel*Link was fully briefed in November 2016, just days after Donald Trump was elected president.30 The FCC’s five-member board is appointed by the president, with, customarily, three members representing the president’s party and two representing the other party.31 The vote on inmate calling–service rate caps followed party lines, so a shift in the partisan makeup of the FCC signaled a likely shift in the Commission’s position on the rate caps.

Argument was scheduled for February 6, 2017. Anticipating a shift in the Commission’s position, on January 11, 2017, the DC Circuit asked the parties to “show cause” for not holding the case in abeyance as the Commission turned over.32 Commissioner Jessica Rosenworcel’s term expired on January 3 (she has since been renominated and confirmed33 ), and Chairman Tom Wheeler, following custom, was to resign effective January 20.34 By January 21, the two dissenters from the 2015 order would constitute a majority of the three active commissioners. Yet the DC Circuit panel determined that arguments should go ahead as scheduled.35 The Obama administration argued that relief for inmate families had been delayed long enough.36 The administration was probably also keen to keep the case before the assembled panel, which included two Democratic appointees and only one Republican, Judge Laurence Silberman. Silberman dissented from the court’s decision to hold arguments as scheduled, preferring to give the FCC a sixty-day window to determine whether it would revisit the 2015 order.37 He also noted that, eight years earlier, a similar luxury had been afforded to the incoming Environmental Protection Agency (EPA), giving the Obama administration time to determine if it wanted to rescind a Bush-era rule.38

Three days after Inauguration Day, Pai was designated FCC Chairman. The majority of the Commission now held the position that the FCC did not possess statutory authority to promulgate intrastate rate caps for inmates.39 The majority soon directed the FCC’s counsel to abandon the argument contained in the first section of its brief: that the FCC had authority to cap intrastate rates.40 In addition, the FCC abandoned the argument that it lawfully considered industry-wide average costs when it set the rate caps, a practice the industry lawyers had argued was contrary to the statute.41 Because of the change in its argument, the FCC ceded ten minutes of its time to intervenors representing families of inmates.42 The intervenors were permitted to argue for the 2015 order in its entirety.43

Each retracted section of the brief concerned the agency’s interpretation of the text of the Communications Act. The section concerning intrastate rates explicitly cited Chevron, while the section concerning average costs cited a case that applied Chevron to support deference to the agency’s line drawing.44 The intervenors’ brief largely left these Chevron arguments to the agency. Their oral advocate began his argument by pointing to a footnote in the intervenors’ brief that he said incorporated the Chevron arguments.45 His brief only touched on Chevron, the advocate said, because the FCC had briefed the issue and he was following instructions “not to file repetitive briefs.”46 None of the judges immediately pushed back against the intervenors’ claim that its briefing of the Chevron issue was “more than adequate.”47 Later in the appearance, Silberman pressed the intervenors’ advocate about whether Chevron arguments had been abandoned, and the advocate insisted that the agency had to promulgate its new policy through the usual means.48

Judge Harry T. Edwards, writing for the majority, interpreted the agency’s abandonment of the statutory interpretation arguments as waiving Chevron deference altogether. Though agency promulgations carrying the force of law are “presumptively” subject to Chevron review, the presumption was overcome by the agency’s decision to abandon its interpretation.49 It would simply “make no sense” to apply Chevron to a position the agency has abandoned.50 Without Chevron deference, the court invalidated both the intrastate rates and the use of industry-wide averages.51

Each of the other judges on the panel offered his or her own answer to whether an agency abandoning its statutory interpretation argument constituted a waiver of Chevron deference. Silberman joined Edwards’s opinion and wrote separately to say that, even granting Chevron deference, the order exceeded the agency’s powers.52 Because Silberman would have refused to approve the order even under a deferential review, he did not need to comment on whether Chevron had been waived. Yet Silberman explicitly endorsed Edwards’s view of Chevron waiver, writing, “I especially agree that Chevron deference would be inappropriate in these unusual circumstances.”53 Judge Nina Pillard dissented.54 In her view, the court should look at the agency’s interpretation of the statute when it promulgated the rule and take no heed of a postpromulgation attempt to waive Chevron.55 She then instructed the new FCC majority that, if it wanted to change the rule, the statute gave it the “latitude to do so” through notice-and-comment rulemaking.56 Not only did none of the judges cite an authority for their conclusions about what to do when an agency abandons its statutory interpretation argument,57 none of them engaged with the extensive academic literature and case law on what Chevron is, how it is justified, and when it applies. Edwards cited the conclusion in United States v Mead Corp58 that Chevron applies to agency determinations having the force of law but then posited—without reasoning—a Chevron waiver exception.59 Pillard hinted at a concern that the agency was doing an end run around notice and comment but gave no reason why such a concern should be decisive.60

Seven weeks later, the panel retreated from its apparent holding that the FCC could waive Chevron deference.61 Edwards and Silberman joined an order amending and purportedly clarifying the decision. Edwards now embraced Silberman’s view that the FCC’s extension of authority over intrastate rates would fail Chevron’s second step and clarified that the use of industry-wide averages was arbitrary and capricious (rather than outside the bounds of the statute).62 Because each agency action was impermissible even assuming Chevron deference, the majority did not need to reach the question of whether the agency waived Chevron and whether that would be permissible.63

The “Clarification and Amendment of the Majority Opinion” was more amendment than clarification. The original opinion quite clearly endorsed the possibility of Chevron waiver, but the majority, perhaps spooked by the intervenors’ en banc petition, which charged the court with creating a “dangerous loophole [for agencies] to evade judicial review,”64 attempted to read its endorsement of Chevron waiver out of the opinion.65 Despite the evident shift in its approach to Chevron waiver, the majority claimed it was simply clarifying “any confusion” and recovering a “point” that may have been “lost.”66 If we take the panel at its word, it never held that Chevron waiver was either permissible or impermissible. It remains an open question whether the original decision’s straightforward allowance of Chevron waiver will have any precedential value.

B. Defining Chevron Waiver

Even if the initial holding in Global Tel*Link was superseded by the Clarification and Amendment of the Majority Opinion, it is still worth examining that holding. The majority reasoned that “it would make no sense for this court to determine whether the disputed agency positions advanced in the Order warrant Chevron deference when the agency has abandoned those positions.”67 That holding poses the question of why it would “make no sense.” To understand the rationale, we need to know whether Chevron arguments were sufficiently presented. If not, it would be possible to understand the opinion as rooted in the norm against judges raising issues the parties have not presented.68 But if the arguments were presented, then Global Tel*Link is best read as adding a threshold question before the Chevron analysis.69 Under that reading, courts would need to ask whether the agency seeks deference. Only if the agency seeks deference can deference be given.

From the facts of Global Tel*Link, there are reasons to think the court rested its decision on some grounds other than the norm against judicial issue creation. The industry parties who opposed the order certainly briefed the statutory interpretation issue.70 The FCC retracted its counterarguments, but the intervenors may have incorporated them in a footnote that referenced similar arguments in a previous FCC order.71 Moreover, at oral argument, the intervenors offered reasons for thinking that the Chevron issue had been presented to the court. In its opinion, the majority did not attempt to rebut the intervenors’ arguments. If the intervenors failed to sufficiently present Chevron arguments, the panel did not say so.

If the majority decided Global Tel*Link on judicial issue-creation grounds, the decision as a whole would not say anything remarkable about Chevron. Courts frequently rely on parties’ choices not to brief Chevron arguments.72 However, aside from Global Tel*Link, the cases in which courts have relied on an agency’s choice not to brief Chevron arguments are all cases in which the agency merely opted not to waste pages on unpersuasive or spurious arguments.73 Perhaps the norm against judicial issue creation should operate differently when the agency’s litigating position reflects a policy choice and not a mere decision to prioritize its best arguments.74 But that is only a suggestion. The rules and customs surrounding whether judges will raise issues sua sponte are a vast and confounding topic that lies adjacent to this Comment.

Instead, this Comment looks at whether an agency must seek Chevron deference in order to receive it.75 It is fair to read the initial Global Tel*Link opinion as resting on a theory that, to apply the Chevron framework in suits in which the agency is a party, an agency must have sought Chevron deference. In particular, the court refused to grant Chevron deference to positions advanced in an order because “the agency ha[d] abandoned those positions,”76 not because none of the parties presented those positions.

The failure to seek such deference or the intentional abandonment of the arguments for deference is what this Comment terms “Chevron waiver.” Whether an agency’s Chevron waiver can prevent the application of the Chevron framework is, like other threshold questions, a matter for Chevron Step Zero.77 This Comment therefore evaluates Chevron waiver at Step Zero.78

C. The Mechanics of Chevron Waiver

To waive Chevron, the agency’s litigators must make clear that they do not seek Chevron deference. They can do this in a number of ways. They could send a letter to the court intentionally abandoning an argument in their brief or, at oral argument, they could state that they do not seek Chevron deference.79 They could also simply fail to raise Chevron arguments before a lower court. This latter strategy could be termed “Chevron forfeiture.” When a party fails to press an argument, it is forfeited; but when a party voluntarily relinquishes an argument, it is waived.80 It is usually easier for courts to know an issue has been waived than that it has been forfeited because it is easier to notice an act than an omission. But Chevron forfeiture will be nearly as easy for a court to police as Chevron waiver. Chevron is generally only applicable to rules and orders that were promulgated through certain Chevron deference–signaling procedures, so, by observing the procedure, the court will know that Chevron could apply.81 Moreover, these procedures force the agency to give reasons for its statutory interpretation in the public record. So even when the agency’s litigators do not seek Chevron deference at the beginning of litigation, the court will have access to a lengthy, published paper trail that includes the reasoned statutory interpretation. A court that wishes to prevent Chevron waiver would be able to close the forfeiture loophole by deferring to the previously published interpretation.

Because Chevron must be waived in a court document, the office with the authority to waive will depend on how Congress has allocated litigating authority. Only the government actor with litigating authority can waive Chevron. As a general matter, the Department of Justice (DOJ) and the attorney general are empowered to conduct litigation in which a federal agency is a party.82 For cases before the Supreme Court, the attorney general has delegated litigating authority to the solicitor general.83 Yet the general grant of litigating authority is rife with exceptions. Many agencies—both independent and executive—have statutory grants of independent litigating authority, at least at certain levels of litigation or over certain matters.84 Independent litigating authority is at its apex for the Federal Election Commission (FEC), which represents itself in all proceedings except before the Supreme Court,85 and at its nadir for the National Transportation Safety Board, which has no independent litigating authority.86 Many of the allocations of litigating authority are quite complex. The Consumer Product Safety Commission, for instance, has independent litigating authority over injunctive actions in the district courts but must have the consent of the attorney general to conduct its litigation in other cases, except at the Supreme Court, for which it never has litigating authority.87 In Global Tel*Link, the FCC was empowered to waive because it has litigating authority over direct appeals from its orders.88 For a different agency or a different matter, a DOJ lawyer may have the sole authority to select the litigating position.

Waiver could also occur at each step of the Chevron analysis. In Global Tel*Link, the FCC retracted its entire argument that it had authority to regulate intrastate calls, all contained in Section I of its brief.89 That section contained arguments relevant to each step of the Chevron analysis. The agency abandoned its Step Zero argument that its interpretations of its own jurisdiction merited Chevron deference.90 It abandoned its Step One argument that the statute was ambiguous.91 And it abandoned an elaborate Step Two argument that its interpretation was reasonable.92 The DC Circuit treated the abandonment of all three arguments at the logically primary step, Step Zero, holding that, because the agency failed to seek deference, the whole of Chevron analysis would not be applied.

Yet not all Chevron waiver includes all three steps. At each Chevron step, the agency could withdraw all arguments or stipulate that it disagrees with the deference-friendly view. For example, at Step Zero the agency could make no claims about whether Chevron deference is merited or it could explicitly state that deference is not merited. Either example would act as a Chevron waiver. That is, either one would implicate a rule that an agency must seek deference in order to get it.

Waiver gets more complicated if it inheres at Chevron’s later steps—and some of the circumstances are beyond the scope of this Comment. An agency that does not want its regulation to receive deference could “waive” at Step One rather than Step Zero. It would agree that the regulation is of the sort that merits the Chevron framework but either withdraw its argument that the statute is ambiguous or, more emphatically, argue that the statute unambiguously forecloses the regulation.93 At Step Two, the agency could argue that the interpretation is unreasonable or make no argument about reasonableness. One common way in which an agency “waives” Step Two is by asserting that Chevron analysis concludes (in the agency’s favor) at Step One. The agency argues the statute unambiguously demands its interpretation and thus the agency does not submit Step Two arguments about the reasonableness of its interpretation. Sometimes, in a move called Chevron Step One-and-a-Half, the court will decide that the statute is ambiguous and remand to the agency rather than apply Step Two.94

D. The Varieties of Chevron Waiver

This Section describes the circumstances under which we might expect an agency to try to waive Chevron.95 Because Chevron increases an agency’s interpretive powers, the more interesting question is when an agency might be motivated to waive Chevron, but it’s worth first spelling out the elementary conditions of Chevron waiver. For one, waiver is a midlitigation action, so there must be ongoing litigation. Not only that, the agency action being contested must turn, to some extent, on a question of statutory interpretation. It would make no sense to waive Chevron in an arbitrary-and-capricious challenge that would never be subject to deferential review under Chevron. Lastly, because Chevron waiver weakens the agency’s position, the agency will waive only when it desires to reverse the action for which it has been sued.

Yet Chevron waiver will not be used every time the agency wants to reverse its actions. Some agency actions can be undone more easily and predictably through means other than Chevron waiver. An agency that uses Chevron waiver must still inform the court of its waiver, show up for oral argument, and leave the fate of the agency action in the hands of a judicial panel. It is not particularly onerous, but it is more onerous and more uncertain than the simple revocation of, say, a guidance document.96 Chevron waiver is, however, always less onerous—if more uncertain—than undertaking the multimonth (or longer) procedures necessary to issue a new rule,97 revoke an old one,98 or, if applicable, conduct a new adjudication.99

In addition to the different costs to the agency of promulgation and Chevron waiver, there may also be a different reward. An interpretation that goes through notice and comment generally cannot be rescinded except through notice and comment.100 But a judicial interpretation after Chevron has been waived may be even “stickier.”101 A judicial decision at Step One rather than Step Two binds future administrations.102 That is, even after notice and comment, future administrations cannot promulgate an interpretation that a court has said is unambiguously foreclosed. For instance, the administration cannot replace an interpretation that a court has said is the unambiguous command of the statute. The rule applies whether a court has explicitly resolved the case at Chevron Step One or determines that the statute is unambiguous even without applying Chevron analysis.103 An agency considering whether to waive Chevron must consider whether the court is more likely to bind future administrations if it is analyzing the statute under Chevron’s two steps or in their absence.104 It then must consider how confident it is that the court will read the statute in the way that it now prefers. The best-case scenario for an agency that wants to reverse a policy is for the court to read the newly favored interpretation as the unambiguous command of the statute, binding future administrations. The worst-case scenario is for the court to bind the agency to an interpretation it disfavors. If Chevron waiver increases the chances of a binding judicial interpretation, then waiver is a high-risk, high-reward choice. Even if Chevron waiver decreases the chances of a binding interpretation, the agency takes a gamble when it waives because it relies on the court to reverse a policy it could reverse on its own, albeit more burdensomely.

No matter whether Chevron waiver increases or decreases the chances that future administrations will be bound, it is clear that Chevron waiver arises when the agency tries to reverse a policy that was promulgated in a binding manner. We can classify six broad categories of circumstances that would lead an agency to reverse such a policy. The first two come after a presidential transition. The next four all involve the same administration seeking to undo its own policy, distinguishing political from technical motivations and agency-initiated reversals from White House–initiated ones.

The first scenario is what happened in Global Tel*Link: a presidential transition changed agency policy while litigation was ongoing.105 Courts can always avoid the possibility of Chevron waiver in these instances by holding the case in abeyance until the new administration determines whether it wants to retract its rule. But there might be good reasons not to delay argument, such as the desire for relief for inmate families that the court found persuasive in Global Tel*Link.106

The presidential transition is especially ripe for Chevron waiver. Modern congressional gridlock means that the postelection political capital of a new administration must be spent largely through administrative actions. Such capital has a short half-life, and so it must be spent quickly. New agency officials, moreover, will seek to change many binding rules but only have so much latitude to do so, especially if they encounter bureaucratic resistance.107 Chevron waiver is a means of reversing agency policy that generally imposes fewer costs on the agency than the alternatives.

A second scenario also follows a presidential transition. In this scenario, however, litigation begins after the new agency leadership has taken their role, perhaps tacitly invited by agency leadership. This is similar to the “sue and settle” practice, in which an agency invites a sympathetic nongovernmental organization to sue the agency for failing to undertake a regulatory action, which is then mandated by a consent decree. Like Chevron waiver, “sue and settle” is a means of averting the rigors of notice and comment or interagency review, the White House’s somewhat prolonged process for subjecting certain agency actions to cost-benefit analysis and otherwise bringing them in line with the administration’s priorities.108

The other scenarios do not involve a presidential transition. Instead, the same administration has changed its policy. We can divide these scenarios along two axes: the motivation for the change and the actor initiating it.

Figure 1:  Examples of Intra-administration Policy Reversal

	Politically Motivated	Technically Motivated
White House Initiated	Issue not subject to interagency review becomes politically controversial.	Change to cost-benefit formula.
Agency Initiated	Agency is heavily lobbied post-promulgation; priorities of DOJ run contrary to that of the promulgating agency.	New information relevant to action changes agency’s mind; DOJ has litigating authority and is concerned with implications from agency’s construction of statute.

Agencies are given more leeway to reverse their policies than are, say, courts following stare decisis because agencies are designed to be able to expertly respond to a changing world and to be (somewhat) responsive to popular opinion.109 Individual estimations of agencies’ capacity for genuine expertise or genuine responsiveness will differ. Because one might be more solicitous of agency claims to expertise and less solicitous of claims to responsiveness, or vice versa,110 it’s worth separating out these two factors, at least conceptually. Hence, we should make a distinction between agency reversals rooted in technical expertise and agency reversals rooted in policy concerns. In the former, the agency responds to new evidence coming to light. In the latter, the agency responds to pressures from the president, Congress, various interest groups, or even to nontechnical judgments made by the agency staff itself.

Though we may be able to conceptually separate technical agency motivations from political agency motivations, in practice courts are likely to struggle with the distinction. Take, for example, Food and Drug Administration v Brown & Williamson Tobacco Corp,111 which concerned the agency’s decision to reverse decades of agency policy and regulate tobacco as a drug. Was this a technical decision motivated by the accumulation of evidence of tobacco’s harmful effects or a political decision under pressure from both the president and specific lobbies? The answer in Brown & Williamson is undoubtedly both. The difficulty of separating the political from the technical is not limited to issues of vast public consequence.112 One reason is that social scientific techniques often rest on a series of subjective commitments or observations.113 What’s more, political pressures on the agencies have only grown, even on seemingly narrow issues.114 Recognizing the reality of congressional gridlock, industry and activist groups have devoted much more attention to influencing administrative outcomes, including through more oblique means, such as public relations campaigns, lobbying Congress, and ex parte communications.115 Such pressures not only mix the political with the technical, but they can overpower administrative norms, including, relevant for our purposes, the norm that agencies adopt litigation positions that defend their chosen policy.116

Then there is the question of who initiates the policy change. Frequently, this will be the president, who intervenes and cajoles the litigator into changing its position, which the agency initially developed largely without presidential input. President-initiated waiver is more plausible for low-impact actions that evaded interagency review when they were developed117 but that, upon implementation, attract public attention. Actions that merit the ire of more sophisticated regulated parties are more likely to attract presidential attention during the comment phase of the initial promulgation, so this scenario is probably limited to actions that affect less sophisticated parties.

The litigator could also initiate the change. Here the question of litigating authority becomes paramount.118 Chevron could be waived by an agency with independent litigating authority when it discovers new evidence, seeks to quietly undermine a presidential priority, or perhaps when the agency’s counsel disagrees with its technical staff and the agency head tolerates the reversal. Chevron waiver when the DOJ has litigating authority, by contrast, could reflect a difference in policy priorities between the expert agency and the attorney general.119 One could even imagine a situation in which the dispute between the attorney general and the agency is technical. For example, the agency’s interpretation of a statute it administers might have implications for how other statutes are read, compromising administrative priorities. Likewise, in cases before the Supreme Court, the solicitor general might be torn between his role as a lawyer for the government and his role as the “Tenth Justice,” the Court’s expert partner in interpreting the law.120 He might drop Chevron arguments he thinks are meritless or that would create an undesirable precedent. Alternatively, he might see the president as his client and undermine the aims of the independent agency.121

With the current sample size of one, it is impossible to know which actors and which motivations will most often lead to Chevron waiver or if Chevron waiver is largely a post-transition phenomenon. Returning to Figure 1, we do not know if Chevron waiver is most likely in the political/White House–initiated, political/agency-initiated, technical/White House–initiated, or technical/agency-initiated circumstance. However, we should expect the motivation and initiating-actor axes to interact. Because the president is generally considered to be more of a political actor than the agencies, we expect political reversals to be disproportionately provoked by the president and technical reversals by the agency. Yet that expectation does not exclude the possibility of the opposite scenarios. The president can also reverse policy for technical reasons—for instance, the White House’s Office of Information and Regulatory Affairs (OIRA) can adjust its cost-benefit assumptions—and agency motivations may be political too. Though we cannot know which motivations and which actions will most often spark a policy reversal, the combinations of possibilities are common enough that, if permitted, agencies will from time to time seek to use Chevron waiver.

II. Chevron Waiver and the Step Zero Doctrine

A hypothetical doctrine prohibiting Chevron waiver—or alternatively, specifying that an agency must seek Chevron deference in order to be given Chevron deference—would apply at Chevron’s “Step Zero,” which asks whether a given agency action is even subject to the Chevron inquiry.122 This Part describes the major Step Zero cases, which have established a murky doctrine, and then applies that doctrine, by its own terms, to a situation in which the agency has abandoned its interpretation in litigation. As Part II.B explains, the doctrinal formulations of Step Zero do not resolve the question of Chevron waiver. But building on the Step Zero cases Part II.A lays out, Part III applies Step Zero’s deeper intuitions about Chevron and the reasonable Congress to conclude Chevron deference should not be waivable.

A. Defining Step Zero

Before Chevron, courts relied on a number of factors to calibrate the scrutiny they accorded agency actions.123 The Chevron opinion, however, appeared to suggest that deference does not vary with the agency’s persuasiveness.124 Instead, it identified a category of agency actions that are accorded deference and a two-step test for applying that deferential review.125 And Chevron introduced a new justification for deferring to agency interpretations: a construction of congressional understanding.126 Courts should defer to agency interpretations of statutes because Congress allocated interpretive authority to the agency.127 Some allocations provide explicit permission for gap filling, which therefore merits deference unless “manifestly contrary to the statute.”128 More commonly, the allocation of interpretive authority is implicit. Even then, the court cannot substitute its reading of the statute for the agency’s reasonable interpretation.129

But Chevron offered little guidance for determining which category of agency actions merited deference.130 It was not clear whether Chevron deference was tied to the specific procedure that generated the interpretation or if it depended on whether the interpretation was legally binding.131 Actions taken after more rigorous agency procedures and actions that have the force of law are not wholly coextensive. Many binding regulations avoid notice and comment, whether through the good cause exception132 or by going through other means of promulgation, such as negotiated rulemaking.133

Mead seemingly resolved the question of what triggers Chevron while adding complications of its own. Chevron analysis is prompted, the Court said, “when it appears that Congress delegated authority to the agency generally to make rules carrying the force of law, and that the agency interpretation claiming deference was promulgated in the exercise of that authority.”134 Chevron’s Step Zero therefore begins from a construction of the enacting Congress’s intent. That Congress required “relatively formal” procedures like notice and comment for agency action is not dispositive, but it is a “very good indicator” of an intention to endow the agency with the authority to interpret ambiguities in the statute with the force of law.135 The force of law inquiry was presumably emphasized because, when an agency acts with the same force over private conduct as Congress is capable of when it exercises its statutory authority, gaps and ambiguities in the statute could be said to deputize the agency as a kind of junior legislator.136

Mead did not elaborate much on which clues beyond “relatively formal” procedures suggest congressional intent to endow the agency with the power to act with the force of law.137 A few considerations are implicit in Mead’s reasoning. Chevron deference is more plausible when the action (1) has precedential value, (2) is binding on third parties, (3) is promulgated high in the agency hierarchy, or when (4) the volume of actions of the same type is not too high.138 Justice Antonin Scalia, who had celebrated the categorical nature of Chevron, denigrated this approach as “th’ol’ ‘totality of the circumstances’ test.”139

The analysis quickly became even less rule-based. A year after Mead, Justice Stephen Breyer, who had long argued deference should vary with the circumstances, wrote for the Court in Barnhart v Walton.140 Barnhart described a five-factor Step Zero test in which Chevron’s application depends on some combination of whether (1) the interpretation is “interstitial,” (2) the agency acts within its expertise, (3) the interpretive question is important “to administration of the statute,” (4) administration is complex, or (5) the agency has given the question “careful consideration . . . over a long period of time.”141 Five years after that, Breyer, again writing for the Court, laid out a different set of factors.142 As in Mead, each multifactor test evaluated whether Congress intended to allocate interpretive authority to the agency.143

Though Mead pushed the Step Zero inquiry in a more holistic direction, with the ultimate focus on whether the agency acted with the force of law, agency procedure still plays an essential role in Step Zero. In practice, the lower courts generally extend Chevron analysis only to the products of notice-and-comment rulemakings and formal adjudications.144 The Supreme Court, in particular, pays close attention to procedure. “[P]ost-Mead,” Professor Kristin Hickman observed that “the Court has never actually extended Chevron deference to interpretations lacking with notice-and-comment rulemaking or relatively formal adjudication procedures.”145 Moreover, when Scalia had the opportunity to write for the Court in 2013’s City of Arlington, Texas v Federal Communications Commission,146 he eschewed any multifactor inquiry and, at least in dicta, treated notice-and-comment rulemaking or formal adjudication as the triggers for Chevron analysis.147

Cases in which the agency has proceeded by notice-and-comment rulemaking or formal adjudication are the easy Step Zero cases. The agency has proceeded relatively formally and acts with the force of law, so Chevron analysis is applied. In harder cases, however, the lower courts often supplement Mead’s force-of-law question by looking to a broader range of factors, merging or selecting between those factors articulated in Barnhart and those factors implicit in Mead.148

B. Applying Step Zero

Chevron waiver is most likely when the agency has no less onerous means of reversing policy. By and large, agency policies without the force of law can be reversed relatively straightforwardly and agency policies with the force of law can only be reversed through means more onerous than Chevron waiver.149 Agencies will thus try to waive deference only when an interpretation has the force of law—the same circumstance in which the interpretation, under Mead, is entitled to Chevron deference.150 But that an interpretation is entitled to deference does not mean the agency cannot waive that entitlement. Mead is silent on whether Chevron is mandatory for such interpretations.151

Chevron waiver is one of those harder cases in which a court could turn to the factors latent in Mead or enumerated in Barnhart. But most of the factors bear no relation to the Chevron-waiver situation. All we know about a theoretical case of Chevron waiver is that the agency wants to abandon its previous interpretation, which would usually be entitled to Chevron deference. An interpretation that the agency now wants to reverse may be interstitial, or it may not be. The fact that the agency wants to reverse it tells us next to nothing about its interstitiality, its complexity, its foundation in expertise (after all, reversals can have technical or political motivations), its bindingness, its importance to the administration of the statute, or the volume of actions of a similar type. Likewise, though the Mead court was more comfortable with deference if the interpretation came from a higher-ranking official, it is not evident that Chevron waiver empowers or disempowers higher-ranking officials. A president may urge Chevron waiver to undermine the interpretation of an agency staffer, but a lawyer in the Justice Department may also use Chevron waiver to undermine the interpretation of the agency.

Barnhart’s “careful consideration . . . over a long period of time”152 factor applies to Chevron waiver but still gives little guidance. While the initial interpretation was likely developed through a procedure like notice and comment that ensured some care was taken, the agency’s newly preferred interpretation has no such assurances. Hence, Barnhart may suggest that Chevron cannot be waived. But all this factor does is suggest that the original interpretation merited deference. When the agency waives, it is not asking for deference for any alternative interpretation that may not have been carefully promulgated; it is merely inviting the court to determine the best interpretation, without deference. Again we reach the question of whether the Mead and Barnhart factors entitle an interpretation to deference or make deference mandatory—a question those cases do not address. Instead of applying Step Zero as a doctrinal test, we need to look to the principles that constructed that doctrinal test in the first place.

III. Chevron Waiver and the Reasonable Congress

The murky tests of Step Zero are a function of its conceptual ambition. Namely, the Step Zero cases—from Chevron to Mead to Barnhart—all try to grant Chevron deference when Congress would intend to allocate interpretive authority to the agency and withhold Chevron deference when Congress would not.153 Step Zero focuses on congressional intent because Chevron deference is justified by a construction of congressional intent.154

And not just any Congress. Step Zero is an inquiry into how a hypothetical reasonable Congress would want to allocate interpretive authority.155 The justices have been remarkably frank that Chevron deference rests on a “legal fiction” of congressional intent.156 Chevron itself was agnostic on whether Congress intended to grant the agency policymaking discretion when it left a gap in the statute, or if the issue slipped Congress’s mind, or if Congress deadlocked.157 To the Chevron Court, it did not matter. Because Chevron did not hinge on what the enacting Congress actually wanted, courts apply Step Zero by making presumptions about congressional priorities.158

With Step Zero’s doctrinal formulations unable to resolve the question of whether Chevron is waivable or mandatory, it is necessary to look to the deeper intuitions that inform Step Zero—namely, the reasons we have Chevron at all. This Part describes the hypotheses about Congress that have explicitly and implicitly justified Chevron deference and shaped the Step Zero doctrine, examining whether Chevron waiver vindicates or undermines the values at stake. After an aside on whether Congress would prefer a rule- or standard-like Step Zero, it focuses on expertise and accountability, the two traditional justifications made explicit in Chevron. However, this Part finds that they give only weak, conflicting guidance for Chevron waiver.159 Stronger guidance and persuasive normative arguments are found in a hypothesis that is largely derived from the scholarly literature: that Chevron is justified as a reward for agency thoroughness.160 Because Chevron waiver undercuts the goods provided by rigorous ex ante agency procedures, this Comment urges its prohibition, a procedure-protecting rule of law that has echoes in canonical administrative law doctrine.

A. An Aside on Step Zero Rules and Standards

The aim of Part III is to recommend a doctrine to govern Chevron waiver. Before delving into how the principles underlying Step Zero inform a view of Chevron waiver, it’s worth asking whether, all else equal, a reasonable Congress would prefer a more rule-like doctrine or a more flexible standard.

Chevron’s foundations suggest a preference for clearer rules. Later interpreters have understood the Chevron decision as concretizing the then-existing deference regime in a more easily administrable across-the-board presumption.161 It makes sense for Congress to prefer a clear presumption. That way, when Congress sets out to legislate, it can better anticipate how its legislative product will be applied and better discharge its constitutional duty.162 Of course, the desire for clarity that may have motivated Chevron did not carry over to the Mead and Barnhart Courts.

The clearest lines would be to allow all Chevron waiver or to blanketly prohibit it. All other concerns aside, blanketly prohibiting Chevron waiver is more easily administrable. If Chevron waiver is permissible, judges (and legislators) will need to confront a range of thorny questions about who can waive, how, and when.163

But there are other imaginable lines of relative clarity. Waivers by one administration in order to undo the interpretations of the last administration are identifiable and an exception could be carved out for them, albeit with some difficulty.164 Other more intricate lines are less defensible. It is hard to distinguish between waivers instigated by one official or another because the lawyer with litigating authority may waive under instructions from the president or from various agency officials. Likewise, distinguishing technical rationales from political ones will not be easy because rationales need not be given when omitting an argument from a brief. Moreover, even if a technical rationale were given, it could be mere pretext for political motives.

B. Agency Expertise

The Chevron opinion offered two main hypotheses for why Congress left a gap for the agency to fill. The first of these is that it wanted the more expert agency to fill in the details of the statutory scheme.165 On this account, statutory ambiguities result because Congress has reached the limits of its factfinding and technical sophistication. Congress thereby anoints the agency an expert, gap-filling junior legislature. The concern with expertise may have informed Mead’s “force of law” test. Whether the agency is acting with the force of law is a decent proxy for when Congress accepts that the more sophisticated agency is acting as a junior legislature. Likewise, the extent to which the agency’s interpretation is a product of its expertise was one of the Step Zero factors in Barnhart and was implicit in other factors, such as complexity and the interstitial nature of the question.166

Some Chevron waivers will be motivated by technical expertise, and some will not. But acknowledging that a court cannot easily separate out technical waivers from nontechnical waivers means that we must consider whether a blanket rule against waiver or a rule allowing waiver is more in keeping with the interest in expertise. On the one hand, waiver may be expertise enhancing because it allows the agency to more easily achieve its most recent view, and it is assumed that technical knowledge is cumulative—that we are getting more expert all the time. But on balance, the interest in expertise probably cuts against Chevron waiver. An interpretation born of notice-and-comment rulemaking or formal adjudication is an interpretation that has engaged with technical arguments, whether presented in comments or cross-examination.167 A reviewing court cannot be assured that that interpretation is the more expert interpretation, but it can be assured that it was developed through a process that at least attempts to force sober analysis of relevant evidence.168 To allow an agency to waive the deference usually given to that interpretation is to subordinate the fruits of an expertise-channeling process to interpretations that could be arrived at in any number of possible ways, including nontransparent political bidding.

C. Accountability

The Chevron Court also hypothesized that Congress would choose to let another “political branch” resolve ambiguities when Congress deadlocked.169 Congress would want courts to grant deference to the agency’s interpretation because it would make that interpretive process more accountable in two senses. First, Chevron enhances electoral accountability. The agency, as part of the executive branch, is responsive to the same group—the electorate—as Congress and thereby engages in the same kind of interest-group balancing, approximating how Congress would have interpreted the statute. The Chevron opinion considered electoral accountability a feature of its deference regime. The Court admitted that whoever interpreted these ambiguous statutes would wind up making “policy choices,” and it thought it more reasonable to vest that power in the agency, which can “rely upon the incumbent administration’s views of wise policy,” than to vest it in judges, who are less equipped to “reconcile competing political interests.”170 Second, Chevron enhances accountability to Congress. Congress might prefer to allocate interpretive authority to the agency because Congress maintains oversight powers and powers of the purse that can check the agency in ways it cannot check courts.171 Congress might also think it is better positioned to use its legislative power to reverse an agency decision than to reverse a court ruling.172 A doctrine like Chevron, which steers contested policy disputes from courts to agencies, will magnify congressional influence over policy and thereby be appealing to the reasonable legislator.173

Each sense of accountability finds some grist in Mead. When Mead tied deference to the rank of the agency official rendering the interpretation, it granted more deference to the agency actors most responsive to the democratically elected president.174 And by broadly linking deference with agency proceedings that are more public in nature, Mead encouraged agencies to put Congress on notice of its actions.175 The most visible agency actors and agency actions are the most accountable to the electorate and to Congress.176

As with the expertise factor, accountability gives support for each side of the Chevron waiver question. Here a reviewing court must separate out the varieties of Chevron waiver. Waiver following a presidential transition enhances electoral accountability because it enables the newly elected coalition to more easily pursue its aims.177 Likewise, post-transition waiver enhances accountability to Congress because Congress can only oversee the new administration and not the old one. Outside the presidential transition context, Chevron waiver probably undermines accountability.178 The notice-and-comment process generates opportunities for control by the political branches. It alerts Congress and the regulated parties to the pending action and often requires coordination with the president and OIRA.179 There is therefore some assurance that the interpretation Chevron waiver undermines—the interpretation generated by notice and comment—is the product of consultations with elected officials and their closest advisors.180 There is no such assurance that the decision to waive Chevron, which can be made by a general counsel for an agency with independent litigating authority, is made in coordination with elected officials.181

Congress’s interest in agencies abiding by procedures that alert Congress to pending actions presumes that Congress expects to be able to conduct meaningful oversight. But what if Congress has a lower estimation of its own powers? Congress may not expect it will be able to meaningfully oversee the agencies, given the sophistication of the agency’s actions and the collective action problems inherent in a legislative body.182 In that case, a reasonable Congress would oppose Chevron waiver—even after a presidential transition. If Congress cannot meaningfully oversee an agency, Congress may prefer procedures that force the agency to submit to comments from the interested parties, who will approximate the sort of pressures a less hamstrung Congress could bring to bear.

On net, therefore, a reasonable Congress interested in accountability would oppose Chevron waiver. If there is an exception, it is for the post-transition waiver, but the case for the post-transition waiver relies on a few suspect premises: that the people gave a mandate to pursue the new interpretation,183 that Congress believes it can meaningfully oversee the agencies, and that a post-transition exception is easily administrable.184 The next factor further weakens the case for the post-transition carve out.

D. Rewarding Rigorous Agency Procedures

Step Zero links the agency’s chosen procedure and the deference the agency is accorded, though that relationship may not be one-to-one. Asking about agency procedure is frequently the first step in the reviewing court’s Step Zero analysis,185 but actions that do not follow relatively formal procedures can still be accorded Chevron deference.186 Put another way, though the procedure the agency used cannot predict the whole of Chevron’s domain, the procedure can predict much of it.

There are static and dynamic reasons why a hypothetical Congress might want Chevron deference to be the reward for rigorous agency procedures. Statically, the Congress might think that agency output is superior on an array of criteria when the agency follows more formal procedures.187 Dynamically, Congress might want to incent agencies to take on these procedures, in effect awarding an agency’s ex ante scrupulousness with ex post deference.188

An emphasis on ex ante procedures can be justified through fuller conceptions of the Chevron rationales discussed above.189 For one, the procedures may enhance expertise. A reviewing court may be more confident that the agency acted according to its expertise because it went through the discipline of responding to the comments of interested parties or the arguments presented in the adversarial adjudicatory process. An appeal to the expert reputation of the agency is not as persuasive as an appeal to the demonstrated thoughtfulness of the agency. Likewise, a fuller picture of accountability can justify an emphasis on ex ante procedures. If accountability is about responsiveness to stakeholder voices, not just those expressed through elections, then administrative procedures can be accountability enhancing. Professor Jud Mathews has argued that giving those impacted by a decision input in the decision-making is a means of ensuring “non-domination,” a foundational democratic principle.190

Congress might also favor more formal agency procedures because agencies who engage in these procedures act more “legislatively.” Notice-and-comment rulemaking and adjudication simulate multiple functions of federal lawmaking. For instance, notice of the potential for regulation is given to impacted parties well before the regulation is finalized, a procedure akin to the legislative requirement that bills be introduced in committee.191 Moreover, the regulation itself is shaped slowly, giving space for legislative ideals of deliberation and careful consideration.192 In this vein, Mead hailed notice-and-comment rulemaking and formal adjudication for fostering “fairness and deliberation.”193

If deference is the reward for agency procedural rigor, courts should be especially suspicious of Chevron waiver. A midlitigation policy reversal does not require that the agency genuinely respond to the technical concerns of affected parties as it did when it initially formulated the interpretation. Likewise, the failure to include those parties in the midlitigation reversal subverts the way notice-and-comment rulemaking and formal adjudication give affected parties the chance to be heard, ensure that they will consent to the outcome because they respect the integrity of the process, and encourage the agency to act in ways that resemble congressional deliberation. Chevron waiver breaks the desirable link between procedure and deference, allowing the agency to bypass the hurdles Congress set in the path of a rule’s rescission. The concern is the same whether the agency seeks to change policy after an election or not. Chevron waiver thus runs contrary to one of the significant intuitions underlying Chevron, an intuition that has helped shape much of the Step Zero doctrine.

When we step back from the doctrinal tests, which were constructed by courts not anticipating the possibility of Chevron waiver, and look to the principles that led to those tests, it becomes apparent that Chevron waiver should not be permitted. Step Zero is informed and shaped by a series of principles: expertise, accountability, and procedural values. Neither expertise nor accountability gives firm guidance as to whether Chevron deference is mandatory or optional, but the interest in rigorous agency procedures does. If one weighs these principles evenly, the interest in rigorous procedures—by virtue of being the only principle that strongly comes down either for or against Chevron waiver—should control the issue.194 But the case can be made even more strongly.

Not only is the interest in rigorous procedures the tiebreaking factor; it should be thought of as the foremost factor. For one, the procedural rationale for Chevron secures the other rationales. It is through ex ante procedures like notice and comment that observers can be more sure agencies are bringing their expertise to bear, are coordinating with the political branches, and are alerting individuals who have something at stake in the pending action. Congress’s interest in these procedures subsumes many of the other reasons for Chevron. It also provides justifications of its own. In particular, these procedures vindicate an idea of democratic legitimacy that is more comprehensive than electoral accountability or congressional oversight.195

The assertion that Chevron deference is so deeply tied up with rigorous agency procedures that, once an interpretation has gone through those procedures, deference should be mandatory may strike the careful reader as an odd assertion. After all, Mead came on the heels of Christensen v Harris County,196 which suggested a rule tying deference to procedure and another rule tying deference to “force of law.”197 Mead explicitly chose “force of law.”198 The contention in this Comment is not that an agency-procedure test is a perfect substitute for the many Step Zero tests but rather that the reasonable Congress’s interest in agency procedures explains a good deal of the shape of those Step Zero tests and, more to the point, helps explain the question those tests don’t address: whether deference is mandatory.

It would not be unprecedented to require deference in certain circumstances, whether the agency seeks it or not. Administrative law is willing to constrain an agency’s choice of litigating positions for the sake of the constitutional structure. Most notably, the rule of Securities and Exchange Commission v Chenery Corp199 forbids a court from reviewing an agency’s post hoc rationalizations.200 Put another way, judicial review is limited to the grounds articulated when the action was taken. Chenery I has a Chevron corollary, Bowen v Georgetown University Hospital,201 which clarified that Chevron deference is given only to statutory interpretations articulated when the agency took the action, not to subsequent “convenient litigating position[s].”202 There are many principled explanations for Chenery I, but Professor Kevin Stack has persuasively rooted Chenery I in functional constitutional concerns.203 He contends that Chenery I ensures many of the aims of the reasonable Congress. By forcing the agency to give its reasons upfront lest they be disregarded by a reviewing court, Chenery I ensures that the agency has exercised its expertise and that politically accountable officials have accepted the agency’s rationale.204

Applying Chenery I to the question of Chevron waiver poses the question of just what Chenery I held. If it merely forbade the review of post hoc rationalizations, Chenery I does not help us think through Chevron waiver. Chevron waiver is a post hoc removal of the initial reasoning, not a new rationalization. But under Stack’s justification of Chenery I, the rule is not simply about post hoc rationalizations. Rather, one could restate Chenery I as the proposition that a court must limit its review to the reasons the agency offered when the action was taken.205 If we apply this broader construction of Chenery I to the question of Chevron waiver, then Chevron waiver is prohibited. A reviewing court would limit itself to the agency’s original reasoning—including its expected deference—and not yield to clever, post hoc litigating positions. The prohibition on Chevron waiver could thus be justified as an application of the long-standing Chenery I rule, or, at the very least, is rooted in the same concern that agencies not evade the requirement to rigorously give their reasons before acting.206

Conclusion

In an age of high-stakes, high-pressure administration, Chevron deference is front and center. Rooted in a recognition that Congress generally prefers that the agency, rather than the court, have the authority to interpret ambiguities and fill gaps in a statute, Chevron has only raised the stakes of agency action. The Global Tel*Link case floated a new possibility for Chevron deference: whether the court can apply the Chevron framework when an agency does not seek it. If Chevron waiver is permissible, there are many circumstances in which an agency will seek to use this new tool: before or after presidential transitions, under pressure from the president or on its own accord, for technical or political reasons. In all cases, an agency will waive Chevron because it wants to speedily reverse a position arrived at through some legally binding process.

As a possible new threshold inquiry before the Chevron framework is applied, Chevron waiver would inhere at Step Zero. Step Zero is an imprecise body of doctrine, and its doctrinal formulations neither prescribe, imply, nor prohibit the possibility of Chevron waiver. The doctrinal confusion results from the Court’s underlying aim. Namely, the Court seeks to apply deference in a way that vindicates Chevron’s purposes, particularly those purposes the Court can ascribe to a hypothetical reasonable Congress that had a reason for leaving ambiguity in the statute for the agency to resolve.

The reasons for Chevron deference caution against allowing Chevron waiver. If Chevron waiver is tolerated, agencies can more easily undo policies arrived at through relatively formal procedures. But we have Chevron deference in large part because courts assume a reasonable Congress wants more expert, accountable agencies to fill in statutory ambiguities while following certain procedures. These procedures—which Chevron waiver would directly undermine—help ensure that the agency actually wrestles with technical arguments, more fully deliberates, alerts Congress and individuals to a pending action, works with elected officials, and provides a basic opportunity for individual participation in the decision-making. Chevron waiver erodes the practical reasons for having Chevron deference at all.

• 1467 US 837 (1984).
• 2Id at 843.
• 3See Cass R. Sunstein, Chevron Step Zero, 92 Va L Rev 187, 188–89 (2006). Professor Cass Sunstein calls Chevron a “counter-Marbury for the administrative state.” Id at 189. See Marbury v Madison, 5 US (1 Cranch) 137, 177 (1803) (“It is emphatically the province and duty of the judicial department to say what the law is.”).
• 4866 F3d 397 (DC Cir 2017).
• 5Id at 402, 407–08.
• 6Id at 417.
• 7Id at 408.
• 8Global Tel*Link, 866 F3d at 418 (Silberman concurring).
• 9See generally Thomas O. McGarity, Administrative Law as Blood Sport: Policy Erosion in a Highly Partisan Age, 61 Duke L J 1671 (2012).
• 10See Kent Barnett and Christopher J. Walker, Chevron in the Circuit Courts, 116 Mich L Rev 1, 30–31 (2017) (observing that agencies won 77 percent of the circuit court cases in which the panel applied Chevron deference but less than 54 percent of the cases in which it did not).
• 11Relevant sections added in the Telecommunications Act of 1996, Pub L No 104-104, 110 Stat 56, codified in various sections of Title 47.
• 1247 USC § 276(b)(1).
• 1347 USC § 276(d).
• 1447 USC § 276(b)(1).
• 15See Report and Order, In the Matter of Policies and Rules Concerning Operator Service Providers, 6 FCC Rec 2744, 2752 (1991).
• 16Report and Order and Further Notice of Proposed Rulemaking, In the Matter of Rates for Interstate Inmate Calling Services, 28 FCC Rec 14107, 14126 (2013) (“2013 ICS Order”).
• 17Id at 14129.
• 18See Joint Brief for ICS Carrier Petitioners, Global Tel*Link v Federal Communications Commission, No 15-1461, *5 (DC Cir filed Nov 14, 2016) (“Providers’ Brief”).
• 19See 2013 ICS Order at 14130 (cited in note 16).
• 20See id at 14147.
• 21See Order, Securus Technologies, Inc v Federal Communications Commission, No 13-1280 (DC Cir filed Jan 13, 2014); Order, Securus Technologies, Inc v Federal Communications Commission, No 13-1280 (DC Cir filed Dec 6, 2014).
• 22Rates for Interstate Inmate Calling Services, 80 Fed Reg 79136, 79137 (2015) (“2015 Order”).
• 23See id at 79154 (“[W]e conclude that our actions in this Order constitute changes in law and/or instances of force majeure that are likely to alter or trigger the renegotiation of many ICS contracts.”).
• 24Id at 79139.
• 25Second Report and Order and Third Further Notice of Proposed Rulemaking, In the Matter of Rates for Interstate Inmate Calling Services, 30 FCC Rec 12763, 12961–65 (2015).
• 26Pub L No 104-104, 110 Stat 56, codified as amended in various sections of Title 47.
• 27In the Matter of Rates for Interstate Inmate Calling Services, 30 FCC Rec at 12961–62.
• 28Id at 12960.
• 29See Providers’ Brief at *17–18 (cited in note 18).
• 30The final reply brief was filed on November 14, 2016, six days after the presidential election. See generally Joint Reply Brief for the ICS Carrier Petitioners, Global Tel*Link v Federal Communications Commission, No 15-1461 (DC Cir filed Nov 14, 2016).
• 31The statute prevents the president from appointing more than three commissioners from his own party. See 47 USC § 154(b)(5). Customarily, the chairman resigns at the beginning of each new administration, allowing the president to appoint a member of his own party as chairman and giving the president’s party a majority. See Margaret Harding McGill and Alex Byers, FCC Chairman Tom Wheeler to Resign (Politico, Dec 15, 2016), archived at http://perma.cc/B5VW-JF2X.
• 32See Order, Global Tel*Link v Federal Communications Commission, No 15-1461, *1 (DC Cir filed Jan 11, 2017).
• 33See Jessica Rosenworcel: Commissioner (Federal Communications Commission), archived at http://perma.cc/NJH7-8GMK.
• 34See Letter, Global Tel*Link v Federal Communications Commission, No 15-1461, *1 (DC Cir filed Jan 31, 2017) (“Jan 31 Letter”).
• 35Order, Global Tel*Link v Federal Communications Commission, No 15-1461, *1 (DC Cir filed Jan 18, 2017) (“Jan 18 Order”).
• 36Response of the Respondents to Show Cause Why These Cases Should Not Be Placed in Abeyance, Global Tel*Link v Federal Communications Commission, No 15-1461, *3 (DC Cir filed Jan 17, 2017). As so often happens, the two sides switched positions on procedural issues as the political upper hand shifted. The government had successfully argued for an abeyance of the litigation over the 2013 ICS order as the FCC considered comments that eventually led to the 2015 order. See Joint Response of Petitioners Global Tel*Link Corp. and Centurylink Public Communications, Inc. to Order to Show Cause, Global Tel*Link v Federal Communications Commission, No 15-1461, *2–3 (DC Cir filed Jan 17, 2017).
• 37Jan 18 Order at *2 (cited in note 35) (Silberman dissenting).
• 38Id, citing California v Environmental Protection Agency, No 08-1178 (DC Cir, Feb 25, 2009).
• 39Jan 31 Letter at *1 (cited in note 34).
• 40Id.
• 41Id at *1–2. See Brief for Respondents, Global Tel*Link v Federal Communications Commission, No 15-1461, *48 (DC Cir filed Nov 10, 2016) (“FCC Brief”).
• 42Jan 31 Letter at *2 (cited in note 34).
• 43Id.
• 44FCC Brief at *32, 39, 49 (cited in note 41).
• 45Oral Argument, Global Tel*Link v Federal Communications Commission, No 15-1461, 1:00:56–1:01:51 (DC Cir Feb 6, 2017), archived at http://perma.cc/4SZW-SS5H, referencing Final Brief of Intervenors Ulandis Forte, et al (the “Wright Petitioners”), Global Tel*Link, Inc v Federal Communications Commission, No 15-1461, *23 n 106 (DC Cir filed Nov 11, 2016). The footnote cited by the intervenors did not reference Chevron by name but did reference statutory interpretation arguments in the FCC’s 2013 order.
• 46Id.
• 47Id.
• 48Id at 1:15:38–1:16:25.
• 49Global Tel*Link, 866 F3d at 407.
• 50Id at 408.
• 51Id at 407–08, 412.
• 52Id at 418 (Silberman concurring).
• 53Global Tel*Link, 866 F3d at 418 (Silberman concurring).
• 54Id at 419 (Pillard dissenting).
• 55See id at 420 (asserting that “the question for us is whether the FCC’s view when it promulgated the challenged rule . . . was a permissible construction of the statute”) (emphasis added) (quotation marks omitted).
• 56Id.
• 57See note 73 for examples of cases in which an agency did not pursue Chevron arguments but, unlike in Global Tel*Link, had no intent to change policy by doing so.
• 58533 US 218 (2001).
• 59Global Tel*Link, 866 F3d at 407–08.
• 60Id at 425 (Pillard dissenting).
• 61Id at 416 (Clarification and Amendment of the Majority Opinion).
• 62Id at 417.
• 63Global Tel*Link, 866 F3d at 417 (Clarification and Amendment of the Majority Opinion) (“We need not and do not decide whether we were required to follow Chevron . . . even though the agency declined to defend its position before the court.”).
• 64Petition for Rehearing En Banc, Global Tel*Link, Inc v Federal Communications Commission, No 15-1461, *8 (DC Cir filed July 28, 2017). Note that the intervenors’ objection is nonsensical. Chevron waiver necessarily gives more opportunity for judicial review because, with Chevron waived, judges need not defer to agency interpretations. Chevron waiver is, therefore, not a loophole allowing the agency to evade judicial review. If it is a “dangerous loophole” to anything, Chevron waiver is a means of evading formal agency decision-making procedures. See Part III.D.
• 65See Global Tel*Link, 866 F3d at 417 (Clarification and Amendment of the Majority Opinion).
• 66Id.
• 67Id at 408.
• 68For a survey of judicial practice when arguments are not raised, see Amanda Frost, The Limits of Advocacy, 59 Duke L J 447, 462–64 (2009) (observing a general norm against courts raising issues sua sponte, with many significant exceptions for issues of subject matter jurisdiction, standing, federalism and comity constraints, preclusion, abstention, sovereign immunity, and what she calls “exceptional” “merits issues”).
• 69Professor Aaron Nielson has interpreted the DC Circuit as adopting this view. See Aaron Nielson, D.C. Circuit Review—Reviewed: A New Step for Chevron? (Yale J Reg: Notice & Comment, June 16, 2017), archived at http://perma.cc/S683-46K2.
• 70See Providers’ Brief at *40–47 (cited in note 18).
• 71See notes 45–48 and accompanying text.
• 72For examples of cases in which a private party failed to brief or intentionally abandoned Chevron, see Lubow v US Department of State, 783 F3d 877, 884 (DC Cir 2015) (finding it unnecessary to rule on Chevron’s first step “because the plaintiffs make no argument that the statute unambiguously compels their interpretation”); Albanil v Coast 2 Coast, Inc, 444 Fed Appx 788, 796 (5th Cir 2011) (withholding Chevron deference because a plaintiff failed to raise it before the district court).
• 73For examples of cases in which the agency did not seek deference, see SSC Mystic Operating Co, LLC v National Labor Relations Board, 801 F3d 302, 316–18 (DC Cir 2015) (Srinivasan concurring) (speculating that deference was not sought in Laurel Baye Healthcare of Lake Lanier, Inc v National Labor Relations Board, 564 F3d 469 (DC Cir 2009), because the Board plausibly did not believe it had the authority to undertake jurisdictional interpretations); MBIA Insurance Corp v Federal Deposit Insurance Corp, 708 F3d 234, 240 (DC Cir 2013) (noting that the FDIC was not seeking deference, presumably because it had not promulgated a regulation interpreting the statute); Lawson v FMR LLC, 670 F3d 61, 82 (1st Cir 2012) (explaining that the agency correctly stipulated that no deference was merited); Oklahoma Natural Gas Co, a Division of ONEOK, Inc v Federal Energy Regulatory Commission, 940 F2d 699, 704 (DC Cir 1991) (mentioning the agency’s failure to seek deference for a jurisdictional interpretation, which the Supreme Court had not yet determined to merit deference).

  For a case in which the agency failed to raise Chevron arguments at the lower court, see Commodity Futures Trading Commission v Erskine, 512 F3d 309, 314 (6th Cir 2008) (noting, in addition to the forfeiture of Chevron arguments, that the agency’s interpretation was not promulgated formally enough to be accorded deference even if Chevron had been briefed). Sometimes an appellate court will still entertain a deference argument not raised at the lower court, only to determine that deference did not apply. See Colorado v Sunoco, Inc, 337 F3d 1233, 1243 n 4 (10th Cir 2003) (entertaining newly raised deference arguments because those arguments, first, were not outcome determinative and, second, could affect proceedings on remand).

• 74The law of judicial issue creation is notoriously undefined and discretionary. As Professor Amanda Frost has put it, some exceptions to the norm are “so broadly worded as to essentially give courts carte blanche to raise new issues at any time.” Frost, 59 Duke L J at 461 (cited in note 68). Courts may have separation of powers reasons not to mechanically apply the norm against judicial issue creation when the agency abandons an issue. When the norm is applied to agencies, it allows agencies to dramatically constrain the grounds upon which a court can rule. Such a practice would give the agency tremendous influence over the interpretation of the law while also retaining its executive functions. The joining of those functions in one administrative body incents the agency to promulgate vague rules that maximize its discretion when it faces litigation. See id at 484, citing John F. Manning, Constitutional Structure and Judicial Deference to Agency Interpretation of Agency Rules, 96 Colum L Rev 612, 645–54 (1996). For additional normative reasons courts should bring up Chevron sua sponte, see Part III.D.
• 75The possibility of Chevron waiver presumes that deference is customarily accorded to agencies. In other words, it presumes that Chevron is still good law. Yet Chevron is at risk for erosion or even elimination. Two Supreme Court justices have already signaled their opposition to Chevron. See Gutierrez-Brizuela v Lynch, 834 F3d 1142, 1149–58 (10th Cir 2016) (Gorsuch concurring); Michigan v Environmental Protection Agency, 135 S Ct 2699, 2712–14 (2015) (Thomas concurring). The intellectual criticism of Chevron is diverse. See, for example, Philip Hamburger, Chevron Bias, 84 Geo Wash L Rev 1187, 1189 (2016) (regarding Chevron as an abdication of the constitutionally vested judicial power); Gutierrez-Brizuela, 834 F3d at 1153 (Gorsuch concurring) (describing Chevron as conflicting with the Administrative Procedure Act); Aditya Bamzai, The Origins of Judicial Deference to Executive Interpretation, 126 Yale L J 908, 999 (2017) (disputing the idea that Chevron is rooted in centuries-old judicial practices); Jack M. Beermann, End the Failed Chevron Experiment Now: How Chevron Has Failed and Why It Can and Should Be Overruled, 42 Conn L Rev 779, 784 (2010) (declaring that Chevron has increased inefficiency and confusion).
• 76Global Tel*Link, 866 F3d at 408 (emphasis added).
• 77See Sunstein, 92 Va L Rev at 191 (cited in note 3) (defining Step Zero as “the initial inquiry into whether the Chevron framework applies at all”).
• 78See Part II.
• 79Agencies also sometimes file amicus briefs in suits that center on statutory provisions the agency administers and for which the agency’s interpretations are usually given Chevron deference. The case that Chevron deference should not be granted if an agency fails to assert Chevron deference in an amicus brief is much weaker than when an agency is a party. Amici generally have a more attenuated stake in a case than a party. To hold that their briefs (and their briefs’ omissions) should narrow the grounds on which a court can rule would undermine the American legal system’s principled, longstanding reliance on party presentation of issues. So long as one of the parties asserts that Chevron deference is merited—no matter what the agency does—Chevron arguments are not waived. Indeed, the Supreme Court has considered a case in which the federal agency’s amicus brief was silent on Chevron deference but one of the parties asserted that Chevron deference should be accorded. See PUD No 1 of Jefferson County v Washington Department of Ecology, 511 US 700, 728–29 (1994) (Thomas dissenting). Despite the omission, the Court analyzed the statute under the deferential two-step Chevron standard of review. Id at 712. That case concerned a state agency’s interpretation of the Clean Water Act’s certification requirements. Rather than make a Chevron case for its own interpretation (which favored the state regulator), the EPA’s amicus brief argued that, even under the petitioners’ interpretation of the statute, the state’s regulation was permissible. Id at 729 n 1 (Thomas dissenting). The Court nonetheless applied Chevron and deferred to the EPA’s interpretation. Id at 712.
• 80See United States v Olano, 507 US 725, 733 (1993) (“Whereas forfeiture is the failure to make the timely assertion of a right, waiver is the ‘intentional relinquishment or abandonment of a known right.’”), quoting Johnson v Zerbst, 304 US 458, 464 (1938).
• 81Courts determine whether Chevron applies by looking to the procedure the agency used in developing the interpretation. See text accompanying notes 137–45. Though there is some uncertainty about whether less common procedures trigger Chevron deference, the two most common “relatively formal” procedures—notice-and-comment rulemaking and formal adjudication—do trigger Chevron. In fact, the Supreme Court has only ever extended Chevron to interpretations arising under notice-and-comment rulemaking and formal adjudication. Kristin E. Hickman, The Three Phases of Mead, 83 Fordham L Rev 527, 548 (2014).
• 8228 USC § 516.
• 83See 28 CFR § 0.20(a).
• 84For a helpful chart distinguishing agencies without independent litigating authority from those independent and executive agencies with “[f]ull or [p]artial” independent litigating authority, see Kirti Datla and Richard L. Revesz, Deconstructing Independent Agencies (and Executive Agencies), 98 Cornell L Rev 769, 800 (2013).
• 85See 26 USC §§ 9010(a), 9040(a); Federal Election Commission v NRA Political Victory Fund, 513 US 88, 96 (1994). The FEC may represent itself before the Court in the narrow set of cases involving the presidential election funds. Id at 94.
• 86See Neal Devins, Unitariness and Independence: Solicitor General Control over Independent Agency Litigation, 82 Cal L Rev 255, 278 (1994).
• 8715 USC §§ 2071(a), 2076(b)(7).
• 88See Devins, 82 Cal L Rev at 279 (cited in note 86). The FCC has litigating authority over appeals from its orders (except before the Supreme Court), but the DOJ handles “actions launched in district court.” Id. FCC lawyers and DOJ lawyers frequently collaborate. DOJ lawyers were listed on the FCC’s Global*Tel Link brief but not on its letter retracting Chevron arguments. FCC Brief at *66 (cited in note 41); Jan 31 Letter at *2 (cited in note 34).
• 89See Jan 31 Letter at *1 (cited in note 34).
• 90See FCC Brief at *27 (cited in note 41).
• 91Id at *28.
• 92Id at *28–38.
• 93An agency cannot, however, argue that the court should defer to its new reasoning. See Bowen v Georgetown University Hospital, 488 US 204, 212–13 (1988).
• 94Chevron Step-One-and-a-Half doctrine is largely a creation of the DC Circuit. For a comprehensive discussion of how it operates and when agencies might take advantage of it, see generally Daniel J. Hemel and Aaron L. Nielson, Chevron Step One-and-a-Half, 84 U Chi L Rev 757 (2017).
• 95Throughout this Section, I will speak of the “agency” as the actor who waives Chevron. Yet as already noted, a litigator must waive Chevron. That litigator could be an employee of the agency that promulgated the rule or an employee of an agency within the DOJ, such as the Office of the Solicitor General or the Civil Division. If a DOJ attorney waives, it could be at the behest of the promulgating agency or another official within the executive branch. “Agency” provides a useful simplification, encompassing waiver by the agency’s counsel and by DOJ lawyers at the behest of the promulgating agency. It also follows from the Global Tel*Link panel’s emphasis on the agency’s abandonment of the pro-Chevron position. See Global Tel*Link, 866 F3d at 408. For the complexities introduced by the division of litigating authority, see notes 118–21 and accompanying text.
• 96That is, unless the guidance document is a “significant” guidance document, which has been subject to interagency review at least since the George W. Bush administration. See generally Office of Management and Budget, Final Bulletin for Agency Good Guidance Practices, 72 Fed Reg 3432 (2007). A guidance document is significant if it has an effect on the economy of at least $100 million or “adversely affect[s] in a material way the economy, a sector of the economy, productivity, competition, jobs, the environment, public health or safety, or State, local, or tribal governments or communities”; creates an inconsistency with another agency; “[m]aterially alter[s] the budgetary impact of entitlements, grants, user fees, or loan programs or the rights and obligations of recipients thereof”; or “[r]aise[s] novel legal or policy issues.” Id at 3439.
• 97See 5 USC § 553.
• 98See Federal Communications Commission v Fox Television Stations, Inc, 556 US 502, 515–16 (2009).
• 99See 5 USC §§ 554, 556–57.
• 100See Perez v Mortgage Bankers Association, 135 S Ct 1199, 1206 (2015) (explaining that the Administrative Procedure Act requires that “agencies use the same procedures when they amend or repeal a rule as they used to issue the rule in the first instance”), citing Fox Television, 556 US at 515.
• 101For a discussion of why agencies would desire “stickier” interpretations rather than simple flexibility, see Aaron L. Nielson, Sticky Regulations, 85 U Chi L Rev 85, 117–25 (2018) (“[T]he ability to make credible commitments matters because agencies have more long-term options if regulated parties trust that agency policies are durable.”).
• 102See National Cable & Telecommunications Association v Brand X Internet Services, 545 US 967, 982–83 (2005).
• 103Id at 984–85 (holding that AT&T Corp v City of Portland, 216 F3d 871 (9th Cir 2000), which did not analyze the regulation under Chevron, did not bind the agency because it described its reading of the statute as the best one, not the only permissible one, implying that its interpretation would have been binding had it said its interpretation was unambiguously commanded by the statute). See also UC Health v National Labor Relations Board, 803 F3d 669, 688 (DC Cir 2015) (Silberman dissenting):

  The first [question Brand X addresses is] how [ ] reviewing courts deal with a pre-Chevron judicial decision if the agency subsequently disagrees. The Supreme Court explained that if a prior judicial decision announced the only acceptable interpretation of a statute that opinion governed, but if the earlier judicial opinion—properly read—only relied on a better interpretation, the agency was free to adopt a different reasonable construction.

• 104We do not know whether Chevron waiver makes a court any more likely to resolve the issue at Step One in favor of the agency’s newfound preferences; it’s an empirical question with a very small data set.
• 105See generally Jan 31 Letter (cited in note 34).
• 106See generally Jan 18 Order (cited in note 35).
• 107The previous administration will have tried to promulgate rules through notice and comment, knowing they are hard to rescind. See Nina A. Mendelson, Agency Burrowing: Entrenching Policies and Personnel before a New President Arrives, 78 NYU L Rev 557, 592–93 (2003).
• 108See generally William L. Kovacs, Keith W. Holman, and Jonathan A. Jackson, Sue and Settle: Regulating behind Closed Doors (US Chamber of Commerce, May 2013), archived at http://perma.cc/5MWP-UA4N.
• 109See Randy J. Kozel and Jeffrey A. Pojanowski, Administrative Change, 59 UCLA L Rev 112, 137 (2011). In their article, Professors Randy Kozel and Jeffrey Pojanowski chronicle the various ways in which administrative policy change is and is not subject to higher standards of review. Id at 138–59.
• 110See Jeremy Rabkin, Judicial Compulsions: How Public Law Distorts Public Policy 95–105 (Basic Books 1989) (contrasting a Weberian confidence in technically proficient bureaucracy with the view of The Federalist, which thought that delegating executive responsibility to popular government was the best means to good social policy).
• 111529 US 120 (2000).
• 112See, for example, Wendy E. Wagner, The Science Charade in Toxic Risk Regulation, 95 Colum L Rev 1613, 1617 (1995) (referencing the phenomenon of “camouflaging controversial policy decisions as science”).
• 113See F.A. Hayek, The Counter-revolution of Science: Studies on the Abuse of Reason 41–60 (Liberty Fund 1979).
• 114See McGarity, 61 Duke L J at 1704–10 (cited in note 9). Professor Thomas McGarity largely laments this trend and credits much of it to right-of-center groups who have more foundational disagreements with the regulatory state. Id at 1709–10. But administrative law is unlikely to successfully counter the expansion of the range of viewpoints in our politics. Administrative law must instead adapt to the breakdown of any technocratic consensus.
• 115Id.
• 116One possible answer to the question of why agencies have not tried to waive Chevron before is that the increased pressures on the agencies have eroded the norms that kept past agencies from so obviously undermining their promulgated policy.
• 117Agencies frequently act to insulate themselves from presidential review by opting to proceed by adjudication, crafting rules that fall just beneath the significance threshold for interagency review, burying the Office of Information and Regulatory Affairs (OIRA) in technical substance, forcing OIRA to review their products under tight statutory deadlines, and seeking out coalition partners among the many staffers who together make up the White House. See Jennifer Nou, Agency Self-Insulation under Presidential Review, 126 Harv L Rev 1755, 1782–1803 (2013).
• 118See Elizabeth Magill and Adrian Vermeule, Allocating Power within Agencies, 120 Yale L J 1032, 1060 (2011) (hypothesizing about the incentives created by the inconsistent allocation of independent litigating authority).
• 119Id at 1060–61.
• 120See Michael W. McConnell, The Rule of Law and the Role of the Solicitor General, 21 Loyola LA L Rev 1105, 1106 (1988) (describing different approaches to the solicitor general’s responsibilities).
• 121But see Devins, 82 Cal L Rev at 290 (cited in note 86) (“While the Solicitor General may seek to moderate the independent agency’s position because of competing agency or executive interests, he rarely tosses aside the independent agencies’ views.”).
• 122Note that even “waivers” at Step One or Step Two would apply at Step Zero. An agency that refuses to argue the statute is ambiguous (Step One Waiver) and an agency that refuses to argue the interpretation of an ambiguous statute is reasonable (Step Two Waiver) would take those positions because it does not seek deference. Under a rule that the Chevron framework cannot be applied when the agency does not seek deference, such Step One and Two Waivers would bar the application of the Chevron framework—just like any other agency action that fails Step Zero.
• 123See Skidmore v Swift & Co, 323 US 134, 140 (1944) (according deference to agency judgments depending “upon the thoroughness evident in its consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade, if lacking power to control”).
• 124Chevron, 467 US at 842–43.
• 125See Antonin Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke L J 511, 516. See also Adrian Vermeule, Introduction: Mead in the Trenches, 71 Geo Wash L Rev 347, 348 (2003) (“The Chevron opinion itself is best read as an attempt to simplify and clarify the preexisting, and notoriously muddled, law of deference to agency interpretations.”). Yet not all have accepted the view that Chevron replaced the Skidmore deference standard with a Chevron deference rule. As both justice and judge, Stephen Breyer has taken the opposite view, arguing that Chevron merely added a factor—hypothetical congressional intent—to the multifactor analysis in Skidmore. See Christensen v Harris County, 529 US 576, 596 (Breyer dissenting). See also Stephen Breyer, Judicial Review of Questions of Law and Policy, 38 Admin L Rev 363, 380–81 (1986). As a consequence of this interpretation, Breyer has favored a more standard-like conception of Chevron. See SAS Institute Inc v Iancu, 138 S Ct 1348, 1364 (2018) (Breyer dissenting) (describing Chevron as “rule of thumb” in the only paragraph of his dissent which Justice Elena Kagan did not sign). The court has oscillated between these two views. See text accompanying notes 134–47.
• 126See Smiley v Citibank (South Dakota), NA, 517 US 735, 740–41 (1996) (characterizing Chevron as justifying deference in “a presumption that Congress . . . understood that the ambiguity would be resolved . . . by the agency, and desired the agency (rather than the courts) to possess whatever degree of discretion the ambiguity allows”).
• 127Chevron, 467 US at 843–44. See also Scialabba v Cuellar de Osorio, 134 S Ct 2191, 2214 (2014) (Roberts concurring) (“Courts defer to an agency’s reasonable construction of an ambiguous statute because we presume that Congress intended to assign responsibility to resolve the ambiguity to the agency.”).
• 128Chevron, 467 US at 843–44.
• 129Id at 844.
• 130Chevron’s one clear limiting principle is that it applies only to an agency’s “construction of the statute which it administers.” Id at 842 (emphasis added). Later courts have read Chevron to limit deference to agency interpretations of statutes that Congress charged the particular agency with administering. See Smiley, 517 US at 739. Hence, statutes that are administered by no single agency, like the Administrative Procedure Act, Pub L No 79-404, 60 Stat 237 (1946), codified at 5 USC § 500 et seq, or the Religious Freedom Restoration Act, Pub L No 103-141, 107 Stat 1488 (1993), codified at 42 USC § 2000bb–2000bb-4, are not thought to grant agency interpretations Chevron deference. See Thomas W. Merrill, Step Zero after City of Arlington, 83 Fordham L Rev 753, 759–60 (2014).
• 131See Christensen, 529 US at 587:

  Here, however, we confront an interpretation contained in an opinion letter, not one arrived at after, for example, a formal adjudication or notice-and-comment rulemaking. Interpretations such as those in opinion letters—like interpretations contained in policy statements, agency manuals, and enforcement guidelines, all of which lack the force of law—do not warrant Chevron-style deference.

  Christensen, decided a year before Mead, did not specify whether these latter interpretations did not warrant Chevron deference because they were promulgated under less rigorous procedures, or because they lacked the force of law, or both, or neither.

• 132See 5 USC § 553(b)(B).
• 133See 5 USC §§ 561–70.
• 134Mead, 533 US at 226–27. Scholars contest whether Mead posited that “force of law” is an independent criterion for which agency procedures like notice and comment are evidence. The opinion is something of a mess, having finessed several views among the justices to obtain eight votes. See Merrill, 83 Fordham L Rev at 766–67 (cited in note 130).
• 135Mead, 533 US at 229–30.
• 136See John F. Duffy, Administrative Common Law in Judicial Review, 77 Tex L Rev 113, 199–202 (1998); Thomas W. Merrill and Kristin E. Hickman, Chevron’s Domain, 89 Georgetown L J 833, 876–79 (2001). But see Mark Seidenfeld, Chevron’s Foundation, 86 Notre Dame L Rev 273, 288 (2011):

  One can read authority to act with the force of law as implying a designation of interpretive primacy only if one engages in the semantic sleight of hand of equating creating policy within the bounds of the statute with resolving gaps and ambiguities in the statute. But they are not the same. Filling in gaps and clarifying ambiguities means resolving issues about what the statute requires and prohibits, while creating policy-based rules means adding legal requirements that neither permit what the statute prohibits nor prohibit what the statute requires.

• 137See Hickman, 83 Fordham L Rev at 533 (cited in note 81).
• 138See Mead, 533 US at 232–34. See also Lisa Schultz Bressman, How Mead Has Muddled Judicial Review of Agency Action, 58 Vand L Rev 1443, 1452–53 (2005).
• 139Mead, 533 US at 241 (Scalia dissenting). Scalia proposed instead that all “authoritative” agency interpretations trigger Chevron analysis. See id; Christensen, 529 US at 591 (Scalia concurring). Chevron waiver would be relatively straightforward under the authoritativeness test. Interpretations that “represent the judgment of central agency management, approved at the highest levels” are authoritative, including the opinion of the general counsel for the agency. Mead, 533 US at 258 n 6 (Scalia dissenting). A general counsel’s declaration that the agency had abandoned the interpretation would probably mean it was no longer authoritative and thus not subject to Chevron deference. By the logic of Scalia’s Mead and Christensen opinions, Chevron waiver is likely permissible.
• 140535 US 212 (2002).
• 141Id at 222.
• 142See Long Island Care at Home, Ltd v Coke, 551 US 158, 173–74 (2007). Breyer looked to whether the agency rule “sets forth important individual rights and duties,” the agency “focuses fully and directly upon the issue,” the agency “uses full notice-and-comment procedures,” “the resulting rule falls within the statutory grant of authority,” and “the rule itself is reasonable.” Id.
• 143See Barnhart, 535 US at 225 (“[The factors] lead us to read the statute as delegating to the Agency considerable authority to fill in, through interpretation, matters of detail related to its administration.”); Long Island Care at Home, 551 US at 173 (“[T]he ultimate question is whether Congress would have intended, and expected, courts to treat an agency’s rule, regulation, application of a statute, or other agency action as within, or outside, its delegation to the agency of ‘gap-filling’ authority.”).
• 144See Hickman, 83 Fordham L Rev at 551 (cited in note 81).
• 145Id at 548.
• 146569 US 290 (2013).
• 147See id at 307. See also Patrick J. Smith, Chevron Step Zero after City of Arlington, 140 Tax Notes 713, 718 (2013) (suggesting that City of Arlington tacitly replaced Mead’s multifactor inquiry).
• 148For descriptions of how circuit courts have examined the harder Step Zero cases, see Hickman, 83 Fordham L Rev at 551 (cited in note 81); Bressman, 58 Vand L Rev at 1459–64 (cited in note 138). Though many circuits fight their way through the thicket of Mead and Barnhart factors when faced with an agency action that is neither a legislative rule nor a formal adjudication, others find ways to avoid this complex inquiry altogether. See Daniel S. Brookins, Essay, Confusion in the Circuit Courts: How the Circuit Courts Are Solving the Mead-Puzzle by Avoiding It Altogether, 85 Geo Wash L Rev 1484, 1496–1500 (2017) (describing Step Zero avoidance as the “norm” in the hard cases). One way of avoiding hard Step Zero cases is to determine that the action fails Step Two and so the court need not reach the Step Zero question. Id at 1497, citing California Department of Social Services v Thompson, 321 F3d 835 (9th Cir 2003), and Cook v Food & Drug Administration, 733 F3d 1 (DC Cir 2013). Eventually Global Tel*Link took this tack. See Global Tel*Link, 866 F3d at 417 (Clarification and Amendment of the Majority Opinion).
• 149See text accompanying notes 96–99.
• 150With a few independently justified exceptions, such as the major-question doctrine. See note 153.
• 151A fine-toothed reading of Mead might suggest an answer. “We have recognized a very good indicator of delegation meriting Chevron treatment,” Mead says, “in express congressional authorizations to engage in the process of rulemaking or adjudication that produces regulations or rulings for which deference is claimed.” Mead, 533 US at 229 (emphasis added). But it would be too lawyerly to treat this sentence as pronouncing that a claim of deference is a necessary element of a grant of deference. The Court mentions that deference was claimed in order to specify that the Court is not granting general deference to the agency. Rather, congressional authorization to rulemake grants deference for the resulting rules and congressional authorization to adjudicate grants deference for the resulting orders.
• 152Barnhart, 535 US at 222.
• 153This includes the “major questions” exception to Chevron’s domain that runs parallel to the Mead doctrine. For a concise description of the major-questions doctrine as a Chevron Step Zero issue, see United States Telecom Association v Federal Communications Commission, 855 F3d 381, 418–23 (DC Cir 2017) (Kavanaugh dissenting) (“For an agency to issue a major rule, Congress must clearly authorize the agency to do so. If a statute only ambiguously supplies authority for the major rule, the rule is unlawful.”). Notably, in his United States Telecom Association dissent, then-Judge Brett Kavanaugh observes that the Supreme Court has rooted the major-questions exception in both a nondelegation canon and in “a presumption that Congress intends to make major policy decisions itself, not leave those decisions to agencies.” Id at 419 (Kavanaugh dissenting). See, for example, King v Burwell, 135 S Ct 2480, 2488–89 (2015) (applying the major-questions exception at Step Zero and withholding Chevron deference in “extraordinary cases” in which it is implausible “that Congress has intended such an implicit delegation”); Utility Air Regulatory Group v Environmental Protection Agency, 134 S Ct 2427, 2444 (2014) (“We expect Congress to speak clearly if it wishes to assign to an agency decisions of vast ‘economic and political significance.’”); Gonzales v Oregon, 546 US 243, 267 (2006) (“The idea that Congress gave the Attorney General such broad and unusual authority through an implicit delegation in the CSA’s registration provision is not sustainable.”). See also Breyer, 38 Admin L Rev at 370 (cited in note 125) (“Congress is more likely to have focused upon, and answered, major questions, while leaving interstitial matters to answer themselves in the course of the statute’s daily administration.”).
• 154See Chevron, 467 US at 865; John F. Manning, Chevron and the Reasonable Legislator, 128 Harv L Rev 457, 464 (2014) (explaining that Chevron relies as much on the “reasonable legislator construct” as the Skidmore and Mead doctrines, which explicitly rely on the construct).
• 155See Manning, 128 Harv L Rev at 458 (cited in note 154) (“Every framework used by the Court for determining the availability of deference has rested on a legal fiction about presumed legislative intent.”); SAS Institute, 138 S Ct at 1364 (Breyer dissenting) (stating that courts determine how much interpretive authority Congress intended to give the agency “by using a canon-like, judicially created construct, the hypothetical reasonable legislator, and asking what such legislators would likely have intended had Congress considered the question”). For a normative defense of the reasonable legislator framework, see Merrill and Hickman, 89 Georgetown L J at 872 (cited in note 136). In some recent opinion judges have focused, at least rhetorically, on the will of the actual Congress rather than that of the hypothetical reasonable Congress. See, for example, City of Arlington, 569 US at 296, 307 (stating that there is no need to wade into the “murky waters” of congressional intent because Congress legislated against a background rule of agency interpretive power).
• 156Breyer, 38 Admin L Rev at 370 (cited in note 125). See also Scalia, 1989 Duke L J at 517 (cited in note 125) (“[A]ny rule adopted in this field represents merely a fictional, presumed intent.”). For an empirical defense of the Court’s legal fiction, see Lisa Schultz Bressman, Essay, Reclaiming the Legal Fiction of Congressional Delegation, 97 Va L Rev 2009, 2041 (2011).
• 157Chevron, 467 US at 865:

  Perhaps that body consciously desired the Administrator to strike the balance at this level, thinking that those with great expertise and charged with responsibility for administering the provision would be in a better position to do so; perhaps it simply did not consider the question at this level; and perhaps Congress was unable to forge a coalition on either side of the question, and those on each side decided to take their chances with the scheme devised by the agency. For judicial purposes, it matters not which of these things occurred.

• 158Sometimes, though, a court has more to work with than just a presumption of congressional priorities. The Dodd-Frank Wall Street Reform and Consumer Protection Act of 2011 codified that the Office of the Comptroller of the Currency’s preemption determinations were to be accorded Skidmore deference rather than the Chevron deference that would be accorded under the usual application of Step Zero. 12 USC § 25b(b)(5)(A). The legislative history shows that Congress understood the Skidmore-Chevron background norm when drafting Dodd-Frank. See Kent H. Barnett, Codifying Chevmore, 90 NYU L Rev 1, 34 (2015). When Congress explicitly codifies the standard of review for agency actions, courts are not free to select an alternative standard of review. It may be inferred from the fact that Congress knows how to codify an unwaivable standard of review that Congress’s failure to so codify is a permission slip for Chevron waiver. There are two problems with that inference. First, codifying Chevron or Skidmore is a relatively new (and not yet replicated) legislative tool. Previous drafters may not have imagined that they would have any use for this tool. Second, the decision to codify a standard of review is a departure from the background norm. Unless Congress understands the background norm to license Chevron waiver, the failure to depart from the background norm is not an endorsement of Chevron waiver. Dodd-Frank, for instance, codified a Skidmore standard in order to depart from the Chevron norm, not to avert Skidmore waiver.
• 159See Parts III.B, III.C.
• 160See Part III.D.
• 161Most notably, see Scalia, 1989 Duke L J at 516 (cited in note 125). It is true that Justice Breyer does not share this view. For Breyer, a step-by-step Chevron framework is subordinated to a holistic inquiry into “whether Congress would want a reviewing court to defer to the agency interpretation at issue.” Nicholas R. Bednar and Kristin E. Hickman, Chevron’s Inevitability, 85 Geo Wash L Rev 1392, 1438 (2017), quoting Hickman, 83 Fordham L Rev at 541 (cited in note 81). Breyer does not explicitly address the question of whether the reasonable Congress would prefer a more rule-like or a more standard-like Step Zero because, for Breyer, the standard-like task of imagining the will of the reasonable Congress is the Chevron analysis.
• 162See Scalia, 1989 Duke L J at 517 (cited in note 125) (describing Chevron as creating a “background rule of law”). See also Sunstein, 92 Va L Rev at 203 (cited in note 3).
• 163The “who waives” question relates to the question posed in the scholarly community about whose interpretations merit Chevron deference. See, for example, David J. Barron and Elena Kagan, Chevron’s Nondelegation Doctrine, 2001 S Ct Rev 201, 204 (arguing that Chevron deference should extend only to those to whom Congress delegates the power to administer the given statute and only when that delegatee “takes personal responsibility for the decision”); Mead, 533 US at 258 & n 6 (Scalia dissenting) (linking Chevron to “authoritative” agency interpretations, which come from “central agency management,” including—relevant to Chevron waiver—the general counsel for the agency). The “how” question is also difficult: A brief? A letter? Must there be some review by specific agency heads? And when? Only before oral argument?
• 164It is not readily apparent what counts as “post-transition”: Are acting directors and officials held over from the previous administration counted as part of the new administration and therefore permitted to waive or are they not? Should there be some sort of statute of limitations, prohibiting Chevron waiver some number of days after Inauguration Day?
• 165Chevron, 467 US at 865. See also Pension Benefit Guaranty Corp v LTV Corp, 496 US 633, 651–52 (1990) (“[P]ractical agency expertise is one of the principal justifications behind Chevron deference.”). Of course, expertise does not explain the full shape of Chevron jurisprudence. A Step Zero largely rooted in expertise would limit Chevron to interstitial, technical questions rather than more general questions of statutory interpretation on which we might think the court is expert. This has not been the case.
• 166See Barnhart, 535 US at 222.
• 167For a broader discussion of the connection between Chevron’s domain and agency procedure, see Part III.D.
• 168See Jerry L. Mashaw, Small Things Like Reasons Are Put in a Jar: Reason and Legitimacy in the Administrative State, 70 Fordham L Rev 17, 26 (2001) (“‘Expertise’ is no longer a protective shield to be worn like a sacred vestment. It is a competence to be demonstrated by cogent reason-giving.”).
• 169Chevron, 467 US at 865.
• 170Id. See also Manning, 96 Colum L Rev at 626 (cited in note 74) (“Chevron makes sense of original constitutional commitments to electoral accountability by presuming that Congress has selected agencies rather than courts to resolve serious ambiguities in agency-administered statutes.”).
• 171See Douglas W. Kmiec, Judicial Deference to Executive Agencies and the Decline of the Nondelegation Doctrine, 2 Admin L J 269, 282 (1988); Jack M. Beermann, Congressional Administration, 43 San Diego L Rev 61, 151–53 (2006).
• 172See Kmiec, 2 Admin L J at 281–82 (cited in note 171). But see Merrill and Hickman, 89 Georgetown L J at 865–67 (cited in note 136) (expressing skepticism that Chevron promotes accountability to Congress).
• 173All this assumes that the reasonable legislator is motivated to maintain his or her authority over policy choices. But the reasonable legislator may not be a civic republican. Modern legislators often (rationally) prefer to punt the many lose-lose choices legislators face. Congress’s determination to avoid politically risky choices and to please all comers may in fact be the root of much of the statutory ambiguity Chevron addresses. See Morris P. Fiorina, Legislative Choice of Regulatory Forms: Legal Process or Administrative Process?, 39 Pub Choice 33, 46–52 (1982).
• 174See Mead, 533 US at 233–34.
• 175See Beermann, 43 San Diego L Rev at 152–53 (cited in note 171) (supporting the Chevron framework “when there are likely to be good channels of communication between Congress and the agency,” opposing Chevron “when there are not,” and showing how recent Supreme Court decisions cohere with this dichotomy).
• 176Like expertise, accountability cannot fully explain Chevron’s domain. Article III courts are less democratically and congressionally accountable than agencies, judges are harder to remove than agency heads, and all deference shifts interpretive authority from the courts to the more accountable agency. A solely accountability-based Step Zero would therefore maximize deference. Step Zero may not take this shape in part because, as an absolute matter, neither courts nor agencies are very accountable. Each was designed to be insulated from certain popular pressures, as Chief Justice John Roberts elegantly argued in his City of Arlington dissent. See City of Arlington, 569 US at 313–14 (Roberts dissenting).
• 177There are a few problems with appealing to a popular mandate to justify waiving Chevron. For one, administrative law generally does not grant procedural exceptions to new administrations on account of their newness. See Motor Vehicle Manufacturers Association of the United States, Inc v State Farm Mutual Automobile Insurance Co, 463 US 29, 57 (1983). In addition, one wonders whether there will be much of a popular mandate to pursue a given interpretation of a statute, especially if the major-questions limitation on Chevron is otherwise in place. See note 153. The question of which policy proposals are given a popular mandate is almost always unanswerable, but low-salience ones that did not come up in the campaign—like reversing rate caps for inmate calling services—seem especially unlikely to have secured popular approval.
• 178And perhaps also within the post-transition context. Congress set up a procedure by which a new administration, along with Congress, may repeal some of the last work of the past administration. See Congressional Review Act, Pub L No 104-121, 110 Stat 868 (1996), codified at 5 USC § 801 et seq. Chevron waiver may supplement the Congressional Review Act in ways that run contrary to the limited reversal power authorized by the Act.
• 179See McNollgast, Political Control of the Bureaucracy, in Peter Newman, ed, The New Palgrave Dictionary of Economics and the Law 50, 54 (Macmillan 1998) (describing the APA’s notice-and-comment procedures as performing a “fire alarm” function).
• 180For a broader discussion of the connection between Chevron’s domain and agency procedure, see Part III.D.
• 181There may be a way to ensure that Chevron can be waived only when done in consultation with some elected or high-ranking officials. Litigators who waive Chevron could describe the directive to waive, and courts could be more solicitous of waivers directed by a presidential appointee or the president himself. That said, Congress might still be wary that this means of changing policy does not ensure consultation with the right elected officials. In particular, unlike notice-and-comment rulemaking, neither a directive from a high-ranking official nor a description of that directive submitted to the court would alert Congress to the change in policy.
• 182Professors David Epstein and Sharyn O’Halloran have developed a transaction-costs model for congressional delegation. As a general matter, Congress delegates issues to the agencies when it is unable to overcome its inherent collective action problems. The result is that agencies are given authority over the precise issues that Congress is least able to act on. Epstein and O’Halloran conclude that we should also have low expectations for Congress’s ability to collectively oversee these agencies, given that the agencies handle matters over “which congressional policymaking is most prone to failure.” David Epstein and Sharyn O’Halloran, Delegating Powers: A Transaction Cost Politics Approach to Policy Making under Separate Powers 74 (Cambridge 1999). But see Brian D. Feinstein, Congress in the Administrative State, 95 Wash U L Rev 1187, 1232–37 (2018) (demonstrating that oversight hearings are powerful tools for checking administrative agencies).
• 183See note 205.
• 184For some concerns about the administrability of a presidential-transition exception, see note 164.
• 185See Hickman, 83 Fordham L Rev at 550 (cited in note 81) (“[M]any circuit courts in practice seem quite simply to extend Chevron review to the notice-and-comment regulations and formal adjudications mentioned in Christensen and Mead.”). Step Zero may be even more closely tied to agency procedure post–City of Arlington. See text accompanying note 147.
• 186See, for example, Mylan Laboratories, Inc v Thompson, 389 F3d 1272, 1279–80 (DC Cir 2004); Fournier v Sebelius, 718 F3d 1110, 1118 (9th Cir 2013).
• 187See text accompanying notes 167–68 and 179.
• 188Sunstein, 92 Va L Rev at 225–26 (2006) (cited in note 3) (describing the results of Mead as giving the “pay me now or pay me later” choice—that is, choose the constraints of procedure or those of more searching judicial review). See also Perez v Mortgage Bankers Association, 135 S Ct 1199, 1211–12 (2015) (Scalia concurring) (noting that the APA excluded certain actions from notice-and-comment rulemaking under the assumption that they would be subject to de novo judicial review). But see generally Mark Seidenfeld, Playing Games with the Timing of Judicial Review: An Evaluation of Proposals to Restrict Pre-enforcement Review of Agency Rules, 58 Ohio St L J 85 (1997) (defining some circumstances under which postenforcement review is superior to ex ante notice-and-comment review).
• 189See notes 167–68 and 178–81 and accompanying text.
• 190See Jud Mathews, Minimally Democratic Administrative Law, 68 Admin L Rev 605, 637–40 (2016). The consequence of Mathews’s emphasis on “non-domination” is that the intensity of the review should vary with the adequacy of the agency procedure. Deference is accorded when the agency “adequately took into account the interests of those affected by its decision,” but more scrutiny is applied when the agency did not. Id at 645–46.
• 191See Mark Seidenfeld, A Civic Republican Justification for the Bureaucratic State, 105 Harv L Rev 1511, 1559–60 (1992).
• 192See id at 1541–62. A preference for slow and deliberate lawmaking is deeply rooted in American constitutional theory. See Joseph M. Bessette, Deliberative Democracy: The Majority Principle in Republican Government, in Robert A. Goldwin and William A. Schambra, eds, How Democratic Is the Constitution? 102, 105–11 (AEI 1980). Slow and deliberate lawmaking may enable expertise by facilitating a genuine contest of ideas and setting a stumbling block before knee-jerk proposals, but it has other virtues too. In particular, slow, deliberative government yields consent, legitimacy, stability, and social peace. See Alexander M. Bickel, The Morality of Consent 15–18 (Yale 1975).
• 193Mead, 533 US at 230. But a reasonable Congress might not be so sanguine about these procedures, which may give little guarantee of efficient results, solicitous regulators, or genuine deliberation. See E. Donald Elliott, Re-inventing Rulemaking, 41 Duke L J 1490, 1492–93 (1992):

  No administrator in Washington turns to full-scale notice-and-comment rulemaking when she is genuinely interested in obtaining input from interested parties. Notice-and-comment rulemaking is to public participation as Japanese Kabuki theater is to human passions—a highly stylized process for displaying in a formal way the essence of something which in real life takes place in other venues.

  See also Aaron L. Nielson, In Defense of Formal Rulemaking, 75 Ohio St L J 237, 269 (2014) (chronicling some of the defects in notice-and-comment rulemaking that reduce legitimacy and efficiency). Even if these procedures do not fully guarantee a given set of ends, they offer more guarantees than a procedure-free alternative like Chevron waiver. Elliott and Nielson argued for more rigorous procedures, not less.

• 194Note that Chevron Step One-and-a-Half, though a choice to not present a Chevron argument, does not face these normative problems. Whereas Chevron waiver is always a departure from the reasoning at promulgation, litigating positions that trigger Step-One-and-a-Half generally follow from the reasoning at promulgation. See note 94; Hemel and Nielson, 84 U Chi L Rev at 765–71, 779–81 (cited in note 94).
• 195See text accompanying notes 189–94.
• 196529 US 576 (2000).
• 197See note 131.
• 198See notes 134–36 and accompanying text.
• 199318 US 80 (1943) (“Chenery I”).
• 200Id at 94. See also Citizens to Preserve Overton Park, Inc v Volpe, 401 US 402, 419 (1971) (“‘[P]ost hoc’ rationalizations . . . have traditionally been found to be an inadequate basis for review.”); State Farm, 463 US at 50 (“[A]n agency’s action must be upheld, if at all, on the basis articulated by the agency itself.”).
• 201488 US 204 (1988).
• 202Id at 213.
• 203See generally Kevin M. Stack, The Constitutional Foundations of Chenery, 116 Yale L J 952 (2007).
• 204See id at 993–98.
• 205It is not surprising that courts have focused on “post hoc rationalizations.” Agencies more often find creative new ways to defend their policy, not creative new ways to undermine it, so courts may have trouble imagining evasions of the agency’s initial rationale that are not post hoc rationalizations.
• 206Late in this Comment’s publication process, the DC Circuit issued an opinion that affirmed the reasoning presented here. In a copyright opinion written by Judge Sri Srinivasan and joined by Judges Judith W. Rogers and Thomas B. Griffith, the court held that an agency’s failure to invoke Chevron in its briefing does not forfeit the right to Chevron deference. SoundExchange, Inc v Copyright Royalty Board, 2018 WL 4440299, *8–9 (DC Cir). Chevron analysis instead depends on whether the “agency manifests its engagement in the kind of interpretive exercise to which review under Chevron generally applies—that is, interpreting a statute it is charged with administering in a manner (and through a process) evincing an exercise of its lawmaking authority.” Id at 9. Not only can Chevron analysis be applied when the agency fails to raise Chevron at litigation, the agency is not even required to cite to Chevron during its decision-making. Id. What matters is the process the agency used.

  The court’s rule seems to have been shaped by a fear that too much would depend upon the quality of the agency’s lawyers. A rule requiring agency lawyers to brief Chevron arguments would tend toward a cumbersome “magic words” requirement. Id. And “[a]fter all, ‘it is the expertise of the agency, not its lawyers,’ that ultimately matters.” Id. But why should a court be more concerned about the potential for poor agency lawyering to alter regulatory outcomes in the Chevron context than in any other? This Comment offers one rationale: that Chevron, like Chenery I, is a judicial rule meant to privilege and incent a rigorous and complete agency decision-making process.