Appeal from the Department of Energy Agency Nos. 89 Fed. Reg. 11,434, 89 Fed. Reg. 65,520

Before Haynes, Ho, and Oldham, Circuit Judges .

Andrew S. Oldham, Circuit Judge :

The Biden Administration proposed numerous regulations to make home appliances more expensive and less useful. See, e.g. , Louisiana v. DOE , 90 F.4th 461 (5th Cir. 2024). As part of that effort, one official even proposed banning gas stoves. See Ari Natter, US Safety Agency to Consider Ban on Gas Stoves Amid Health Fears , Bloomberg (Jan. 9, 2023), https://perma.cc/7E2V-DFN5 (quoting statement of Consumer Product Page 2Safety Commission official Richard Trumpka, Jr.). These efforts generated significant controversy and public backlash.

Rather than confront that controversy, the previous Administration attempted to dodge it. In this case, the Department of Energy created new regulations on home cooking appliances, including gas stoves, through a “Direct Final Rule.” What does that mean? It means the Biden Administration promulgated a rule without giving the public advance notice or the opportunity to comment on it. Petitioners are several States that are aggrieved by the Direct Final Rule. Their objections are well taken, so we grant the petition for review.

In February 2024 the Department of Energy (' DOE ' or 'the Department”) published a Direct Final Rule (“the Rule” or “the DFR ”). The DFR 'adopted new and amended energy conservation standards for' consumer-grade stoves and cooktops. Energy Conservation Standards for Consumer Conventional Cooking Products, 89 Fed. Reg. 11434 (Feb. 14, 2024) (' DFR ') (codified at 10 C.F.R. § 430). We ( A ) outline the Department’s authority to issue the DFR , ( B ) discuss the DFR at issue, and ( C ) recount the States’ challenge.

The Department's authority over household stoves and cooktops comes from the Energy Policy and Conservation Act. See Pub. L. No. 94-163, 89 Stat. 871 (1975) (' EPCA ' or 'the Act'). Under the EPCA , the Department may promulgate minimum efficiency standards for a variety of products, including air conditioners, water heaters, lamps, and cooking appliances. See 42 U.S.C. § 6295. The EPCA requires that any new efficiency standard “be designed to achieve the maximum improvement in energy efficiency” that is both “technologically feasible and economically Page 3justified.” Id. § 6295(o)(2)(A). A new standard is 'economically justified' if “the benefits of the standard exceed its burdens.” Id. § 6295(o)(2)(B)(i). The EPCA lays out seven factors to guide the cost-benefit analysis.¹ If DOE weighs these factors and finds a new standard justified, it has two options.

First, DOE can engage in the traditional notice-and-comment rulemaking. That works just like every other notice-and-comment process. The Department can publish a notice of proposed rulemaking, solicit comments from “interested persons,” respond to those comments in nonarbitrary and non-capricious ways, and then promulgate a final rule. See id. § 6295(p)(1)–(3).

Second, and in the alternative, the EPCA allows the Department to promulgate a “Direct Final Rule.” Id. § 6295(p)(4). As the name suggests, direct final rules can be promulgated without advance notice of the proposed rule or a pre-decisional comment period. Instead, the Department need only obtain a statement recommending an energy conservation standard that is “submitted jointly by interested persons . . . fairly representative of relevant points of view.' Id. § 6295(p)(4)(A). The Department then considers whether the proposed standard “is in accordance” with the seven energyefficiency factors. If it is, DOE “may issue a final rule that establishes” the recommended standard. Id. § 6295(p)(4)(A)(i).

Page 4Because DFR s dispense with foundational guardrails on rulemaking, the EPCA places important limits on them. After issuing a DFR , the Department “shall solicit public comment for a period of at least 110 days with respect to” that DFR . Id. § 6295(p)(4)(B). And “[n]ot later than 120 days after the date on which a' DFR is issued the Department 'shall withdraw the [ DFR ] if . . . the [Department] receives 1 or more adverse public comments relating to the [ DFR ]” and “the [Department] determines that such adverse public comments . . . may provide a reasonable basis for withdrawing the [ DFR ] under' the seven energy-efficiency factors. Id. § 6295(p)(4)(B), (C)(i)(I)–(II). And after withdrawing the DFR , the Department must publish the reasons for withdrawing the DFR in the Federal Register and then proceed with ordinary notice-and-comment rulemaking like virtually every other administrative agency in Washington, D.C. Id. § 6295(p)(4)(C)(ii)(I)–(II).

In this case, the Department initially tried to implement new efficiency standards for consumer cooking appliances through notice-andcomment rulemaking. See Energy Conservation Standards for Consumer Conventional Cooking Products, 88 Fed. Reg. 6818, 6820 (Feb. 1, 2023). The Department also proposed efficiency requirements that would have effectively banned gas stoves and ranges. See DFR , 89 Fed. Reg. at 11484 (noting comments on the 2023 proposal predicting “market elimination of gas cooking products'). But the Department never got past notice-andcomment. After a veritable avalanche of adverse comments from a variety of industry groups, the Department abandoned both initiatives. See ibid. (summarizing comments on the 2023 rule). This also was the first time since the EPCA 's enactment that the Department had proposed efficiency standards for consumer kitchen appliances.

Page 5Having failed at notice and comment, the Department then decided to bypass notice and comment altogether. It instead promulgated a DFR that would have the force and effect of law before any comments could derail it. In the DFR , the Department set many of the same energy-efficiency requirements it had attempted earlier. See id. at 11435 . It limited these requirements to 'consumer conventional cooking products,' like ovens, stoves, and cooktops. Ibid . While the new efficiency standards were multifaceted, they boiled down to a single metric: the maximum “Integrated Annual Energy Consumption,” or IAEC , standard. Ibid. Roughly speaking, the IAEC reflects how much energy an appliance may consume in a given year when used by the average consumer. Id. at 11436. The lower the IAEC , the less energy an appliance may consume in daily operation and, as a result, the more efficient the manufacturer’s design must be.

The DFR also announced a ban on the use of a certain type of power supply, called a linear power supply. Ibid. Power supplies convert alternating current ( AC ) into direct current ( DC ) to power the appliance. In short, linear power supplies involve a single step to convert from AC to DC , while switching power supplies involve a two-step process. Because of their twostep process, the latter require more components than the former and thus are more mechanically complex. But switching power supplies consume less power than linear power supplies. So to promote energy efficiency, the Department decided to ban linear power supplies.

Before promulgating the DFR , the Department published a joint statement from purported stakeholders. 42 U.S.C. § 6295(p)(4)(A). The statement was signed by representatives of the Department, several industry groups, and environmental activists. Relevant here, it proposed a range of Page 6new standards for covered appliances and recounted the stakeholders' agreement with those standards.

The Department issued the DFR on February 14, 2024. In accordance with the EPCA , the Department noted in the DFR that it would observe a 110-day comment period; withdraw the DFR if the comments gave a reasonable basis to believe the Rule did not comply with the EPCA ; and, if necessary, begin the notice-and-comment process instead. DFR , 89 Fed. Reg. at 11434. The 110-day comment period would run from February 14 to June 3, and the DFR would become effective on June 13, 2024. Ibid. In the event of withdrawal, the Department promised to post a “timely withdrawal of this rule” in the Federal Register consistent with the statutorily mandated 120-day period. Ibid. ; see 42 U.S.C. § 6295(p)(4)(C)(i).

Petitioners are six States that filed two timely comments on the last day of the 110-day comment period, June 3, 2024.² Nebraska's Attorney General authored one comment and was joined by 22 other States. See ARID.187. Utah and Montana added another comment. See ARID.190. The Nebraska comment attacked the rule on several fronts, including that the Department failed to account for criticism leveled against the exact same standards in the failed notice-and-comment rulemaking attempt; that the signers of the Joint Statement were not fairly representative; and that the DFR 's standards did not satisfy the seven energy-efficiency factors in § 6295(o)(2)(B)(i). The comment from Utah and Montana raised additional arguments: that the Department erroneously assumed its new efficiency Page 7standards would not reduce product lifespan; that requiring switching power supplies over linear power supplies would reduce product lifespan; and that none of the DFR 's emissions calculations accounted for the reduction in lifespan, since a shorter lifespan would require consumers to buy more products in the long run and increase emissions.

The Department did not respond until August 12, 2024. On that day, the Department filed a Confirmation Notice on the DFR ’s docket, stating that it found none of the States’ comments provided a reasonable basis for withdrawing the rule. See Energy Conservation Program: Energy Conservation Standards for Consumer Conventional Cooking Products, 89 Fed. Reg. 65520 (Aug. 12, 2024) ('Confirmation Notice'). The States responded by filing a petition for review in this court on October 10, fiftynine days later. The Department then moved to dismiss the petition for lack of jurisdiction. We carried that motion with the petition.

Historical context is essential to understanding the parties’ arguments. We therefore begin with ( A ) the longstanding presumption of judicial review of agency action and ( B ) the history of direct final rulemaking.

As long as there has been an administrative state, there has been judicial review of agency action. Congress created the 'first modern regulatory agency,' the Interstate Commerce Commission (' ICC '), in 1887. West Virginia v. EPA , 597 U.S. 697, 740 (2022) ( Gorsuch, J. , concurring) (quoting Susan E. Dudley, Milestones in the Evolution of the Administrative State , 3 Daedalus (Nov. 2020)). While the ICC was initially limited to the role of supervising railroad rates, Congress expanded its authority in the early twentieth century. Paul Stephen Dempsey, The Rise and Fall of the Interstate Commerce Commission: The Tortuous Path from Page 8Regulation to Deregulation of America’s Infrastructure , 95 Marquette L. Rev. 1151, 1163-65 (2012). For Progressives looking to enlarge the administrative state, the ICC became “the model” agency. Id. at 1163. The ICC was Congress’s model for the Food and Drug Administration (1906), the Federal Trade Commission (1914), and the Tariff Commission (1916). Ronald A. Cass, Rulemaking Then and Now: From Management to Lawmaking , 28 Geo. Mason L. Rev. 683, 691 (2021). Rule-by-expert became all the rage. See Cochran v. SEC , 20 F.4th 194, 214–15 (5th Cir. 2021) ( Oldham, J. , concurring); Joseph Postell, Bureaucracy in America 174– 77 (2017).

From the very beginning, however, even the progressive Congresses did not allow administrative agencies to reign unchecked. The Hepburn Act of 1906 formalized judicial review of agency action in federal court. And in a series of cases applying the Hepburn Act, the Court adopted the now-familiar standard of appellate review for challenges to agency lawmaking. The reviewing court was limited to the agency’s evidentiary record, deferred to agency determinations of fact supported by substantial evidence, and decided questions of law de novo. See Fred Halbhuber, The State-Law Origins of the Appellate Review Model , 135 Yale L.J. 2674, 2686-88 (2026). These Hepburn Act precedents were the foundation for the Administrative Procedure Act (“ APA ”), which Congress adopted in 1946.

In these early days, however, Congress did not affirmatively authorize judicial review of all agencies and all agency actions. For much of the early twentieth century, Congress authorized judicial review on a patchwork, agency-by-agency basis. See, e.g. , Roni A. Elias, The Legislative History of the Administrative Procedure Act , 27 Fordham Env. L. Rev. 207, 207-11 (2015). What if Congress was silent about a particular agency and a particular petition for review? As early as 1902, the Supreme Court suggested that judicial review presumptively applies-unless and until Congress clearly Page 9abrogates it: “Otherwise, the individual is left to the absolutely uncontrolled and arbitrary action of a public and administrative officer” even where that action “is unauthorized by any law and is in violation of the rights of the individual.” Am. Sch. of Magnetic Healing v. McAnnulty , 187 U.S. 94, 110 (1902).

Subsequent cases fleshed out this presumption. In case after case, the Court rejected any contention that judicial review could be cut off by congressional silence. See Philadelphia Co. v. Stimson , 223 U.S. 605, 620 (1912) (declaring judicial review 'applicable to a Federal officer acting in excess of his authority or under an authority not validly conferred”); Shields v. Utah Idaho Cent. R.R. Co. , 305 U.S. 177, 184 (1938) ('[W]e think respondent was entitled to resort to equity in order to obtain a judicial review of the questions of the validity and effect of the Commission’s determination purporting to fix its status.”); Columbia Broad. Sys. v. United States , 316 U.S. 407, 422 (1942) (“The Commission’s contention that the regulations are no more reviewable than a press release is hardly reconcilable with its own recognition that the regulations afford legal basis for cancellation of the license of a station.'); Stark v. Wickard , 321 U.S. 288, 310 (1944) (“Congress established courts to adjudicate cases and controversies as to claims of infringement of individual rights whether by unlawful action of private persons or by the exertion of unauthorized administrative power.”); Bd. of Governors of Fed. Rsrv. Sys. v. Agnew , 329 U.S. 441, 444 (1947) (rejecting the Board’s argument that the “absence of an express right of review” means that orders of the Board were not subject to review).

Importantly, this presumption both predated and informed the APA . The APA 'was but one of a long sequence of laws in which Congress authorized courts to review agency action.” Mila Sohoni, The Past and Future of Universal Vacatur , 133 Yale L.J. 2304, 2326 (2024); see also Abbott Laby’s v. Gardner , 387 U.S. 136, 140 (1967) (emphasizing that the APA 'embodies Page 10the basic presumption of judicial review”); Citizens to Preserve Overton Park, Inc. v. Volpe , 401 U.S. 402, 410 (1971); Heckler v. Chaney , 470 U.S. 821, 826 (1985). And while preAPA cases grounded the presumption in first principles, see, e.g. , McAnnulty , 187 U.S. at 110, Congress cemented the presumption in the APA ’s text.

The APA provides only two circumstances when agency action is not subject to judicial review: when '(1) statutes preclude judicial review; or (2) agency action is committed to agency discretion by law.' 5 U.S.C. § 701(a). Both these exceptions are extraordinarily narrow. Courts have read statutes to preclude judicial review only upon “clear and convincing evidence of” legislative command. Bowen v. Mich. Acad. of Fam. Physicians , 476 U.S. 667, 672 (1986). And courts have read the agency-discretion exception to apply only in 'rare . . . decisions traditionally left to agency discretion.' Dep't of Homeland Sec. v. Regents of the Univ. of Calif. , 591 U.S. 1, 16-17 (2020).

In sum, Article III courts have been reviewing administrative regulations since the birth of the administrative state. From the very beginning, courts have presumed that such review was available even in the absence of statutory text. And this presumption was confirmed in the text of the APA . As a result, both text and history require us to be skeptical of any argument that Congress intended to cut off the opportunity for judicial review of agency action.

Now, to the history of direct final rulemaking.

At the beginning, agencies primarily crafted new rules through adjudicating individual cases. See SEC v. Chenery Corp. , 332 U.S. 194, 201– 02 (1947). But adjudication was inherently 'retrospective, unpredictable, and ar[ose] in fact-bound situations,” so it was difficult for regulated parties Page 11to anticipate. Trader Joe's v. NLRB , 167 F.4th 766, 795 (5th Cir. 2026) ( Oldham, J. , dissenting). At its apex in the New Deal, this approach led Roscoe Pound to opine that 'agencies were acting without considered judgment, without due process, without sufficient consideration of the issues, and without granting parties the right to be heard or procedures for relief.' Elias, supra at 210.

Congress responded by formalizing the agency rulemaking process in the APA . 5 U.S.C. § 553. Following the APA , a court 'assumes that an agency's exercise of rulemaking must abide by [ APA ] procedures unless certain exceptions apply.” City of Billings v. Transp. Sec. Admin. , 153 F.4th 46, 51 (D.C. Cir. 2025).

One such exception allows an agency to avoid notice and comment “for good cause.” 5 U.S.C. § 553(b)(4)(B). Over the years, agencies have attempted to exploit the good-cause loophole in various ways. See Philip Hamburger , Is Administrative Law Unlawful? 111 (2014) (referring to these as a 'cascade of evasions'). First, agencies began proposing interim final rules, or IFRs . These rules were immediately effective but subject to alteration in response to comments. See generally Michael Asimow, Interim-Final Rules: Making Haste Slowly , 51 Admin. L. Rev. 703 (1999). When IFR s proved difficult, agencies invented direct final rules, or DFR s. Like IFR s, DFR s were promulgated without advance notice or the opportunity for members of the public to object or comment. Unlike IFR s, however, DFR s could not be altered in response to post-promulgation comments. Rather, DFR s were all or nothing. The agency simply published a rule with a statement that the rule would become effective as written unless the agency received an adverse comment. See generally Ronald M. Levin, Direct Final Rulemaking , 64 Geo. Wash. L. Rev.¹ (1995).

Page 12If an agency can unilaterally dispense with advance notice and comment by announcing “good cause,” you might reasonably wonder why any agency would ever promulgate rules with notice and comment. The answer is that agencies traditionally limited DFR s to consensus -based regulations. The EPA popularized this approach when it developed state implementation plans under the Clean Air Act in the 1980s. Id. at 4. Many of these plans were 'straightforward' and received zero comments. Ibid. If the proposal would receive zero comments and enjoy universal support, the reasoning went, then waiting for a proposed rule to go into effect after a meaningless comment period was just delay for delay’s sake. Id. at 2; see also Administrative Conference of the United States, Recommendation 95-4, Procedures for Noncontroversial and Expedited Rulemaking , 60 Fed. Reg. 43108, 43110 (June 15, 1995). So just as the United States Senate can avoid a roll-call vote by unanimous consent, agencies arrogated to themselves the power to skip notice and comment by unanimous consent.

What happened if someone—anyone—objected? Just as one objection defeats unanimous consent in the Senate, so too did it defeat the “good cause” for issuing a DFR . The EPA invariably withdrew a DFR that proved more controversial than expected-even when it received only a single adverse comment. Levin, supra , at 5; see also id. at 23 (“[A]n agency should resort to [a DFR ] only in developing rules that are truly likely to be noncontroversial.”). Other agencies followed this practice. Id. at 5. Alongside DFR s, agencies also published notices of proposed rulemaking for notice-and-comment procedure. That way, if anyone pushed back on the DFR , the agency could simply withdraw the DFR and proceed with notice and comment the old-fashioned way. See id. at 3–6. In this way, the DFR process harmonized the agency's interest in prompt promulgation of uncontroversial rules and the regulated community’s interest in procedural regularity. This procedure explains why there are virtually no judicial Page 13precedents on DFR s: If anyone objected to a DFR , the agency withdrew it. Indeed, after about a decade of using DFR s, the EPA reported in 1999 that a DFR had never faced a legal challenge. Id. at 10.

Congress drew on this practice when it added a DFR provision to the EPCA in 2007. Responding to the Department's complaint that ordinary rulemaking had “hindered the expeditious promulgation of rules . . . even in cases where there is broad agreement,' Blue Br. at 7, Congress provided explicit DFR authority in § 6295(p)(4) via the Energy Independence and Security Act of 2007. Congress could hardly have tracked prior DFR practice more closely. For one, DFR s were permitted only upon a joint statement 'by interested persons' that showed consensus among parties “fairly representative of relevant points of view,” such as “manufacturers of covered products, States, and efficiency advocates.” 42 U.S.C. § 6295(p)(4)(A). Further, DFR s would be published simultaneously with an 'identical' notice of proposed rulemaking. Id. § 6295(p)(4)(A)(i). The Department was expected to withdraw DFR s on receipt of '1 or more adverse public comments,' assuming those comments ' may provide a reasonable basis” for doubting the DFR 's legality. Id. § 6295(p)(4)(C)(i)(I)– (II) (emphasis added). And the Secretary had to “proceed with the notice of proposed rulemaking” if the DFR were withdrawn. Id. § 6295(p)(4)(C)(ii)(I).

In sum, § 6295(p)(4) reflects Congress's intent to formalize preexisting DFR practice. That practice was reserved for noncontroversial rules that were withdrawn upon minimal pushback from regulated parties.

We now turn to ( A ) the Department's attack on our authority to review the States’ petition and ( B ) the merits of the petition itself.

Page 14We have authority to review the States’ challenge. We explain ( 1 ) why the text and history of the EPCA confirm that authority and ( 2 ) why the Department’s counterarguments are unpersuasive.

1

Start as always with the statutory text. As to the timeliness of any petition for review, several provisions of the EPCA are relevant. After promulgating a DFR , the EPCA requires the Department to accept adverse comments for 'at least 110 days.' 42 U.S.C. § 6295(p)(4). Then the Department has ten more days—so 120 days after the promulgation of the DFR —to announce whether it will withdraw the DFR . Id. § 6295(p)(4)(B). If the Department withdraws the DFR , it 'shall not be considered a final rule.” Id. § 6295(p)(4)(C)(iii). The EPCA also provides a timing provision for all petitions for review: “[A]ny person who will be adversely affected by a rule prescribed under” the Act may petition for review “within 60 days after the date on which the rule is prescribed.” Id. § 6306(b)(1).

The question presented is when a DFR is 'prescribed' within the meaning of § 6306(b)(1)? The ordinary meaning of “prescribe” is “[t]o lay down authoritatively as a guide, direction, or rule of action; to impose as a peremptory order; to dictate; direct; ordain.' Prescribe , Webster’s Second New International Dictionary 1954 (1934; 1951 ed.). That is, “prescribe” connotes a final, binding, authoritative rule.

The Department contends that the DFR was 'prescribed' on the date it first appeared in the Federal Register, February 14, 2024. After all, the Department reasons, the DFR was final and authoritative on the day it was prescribed—that’s why it’s called a direct final rule. So that means the 60day window for filing any petition began on February 14 and concluded on April 15, 2024. Because the States did not petition until August 12 of that Page 15year, the Department claims that we lack authority to adjudicate their petition.

We agree with the Department that finality matters, but we disagree that it matters in a way that helps the Department. In the context of the APA , an administrative action is “final” and reviewable “if: (1) the agency’s action is the consummation of the agency's decisionmaking process; and (2) the action must be one by which rights or obligations have been determined, or from which legal consequences will flow.” Texas v. Equal Emp. Opportunity Comm’n , 827 F.3d 372, 380 (5th Cir.) (quotations omitted), opinion withdrawn on reh’g , 838 F.3d 511 (5th Cir. 2016). But a DFR ’s publication is not the consummation of the Department’s decisionmaking. Far from it. The Department must “solicit public comment” for at least 110 days, withdraw the DFR where comments 'may provide a reasonable basis' for doing so and, after withdrawal, proceed with the ordinary notice-and-comment process and “publish” an explanation for withdrawal in the Federal Register. 42 U.S.C. § 6295(p)(4)(B)–(C). Under the statute’s plain language, then, the Department’s work is not done until it receives public comments and adheres to the DFR . And the Department cannot close up shop until the comment period ends and no one objects to the DFR : That’s because the Department “ shall withdraw the direct final rule' upon receipt of one or more adverse comments. Id. § 6295(p)(4)(C)(i) (emphasis added). Thus, DFR s are not “prescribed” until the Department adheres to the challenged standard.

The EPCA 's finality provision reinforces this conclusion. In the section entitled “[t]reatment of withdrawn final rules,” the EPCA states that a withdrawn DFR 'shall not be considered to be a final rule.' Id. § 6295(p)(4)(C)(iii). If a withdrawn DFR is not final, then a DFR cannot be final until the Department adheres to it. Under the Department’s contrary reading, the DFR is “final” on day 1 (when it issues) and then non-final on day 120 (when the Department withdraws it in the face of public opposition).

Page 16Going from dead to alive is the great promise of Holy Scripture, but the Department cites no authority for its relevance to administrative law.

Similarly, if DFR s are “prescribed” when published on day 1, then Congress's comment period is meaningless surplusage. Congress gave interested members of the public 110 days to comment; but in the Department’s view, the last 50 of those are meaningless. Why? Because any petition for review must be filed on day 60, the Department says—50 days before the end of the comment period. So if you wait until, say, day 61 to file your negative comment—something that Congress expressly authorized you to do—then by the time you provide your comment, your right to judicial review has already expired. Thus, by hypothesis, the Department could receive zero negative comments for the first 60 days, then 10,000 blisteringly negative comments in the last 50. Imagine further that all 10,000 of those comments raise myriad well-taken objections to the DFR , including that it’s directly contrary to the laws and Constitution of the United States. According to the Department, it retains absolute, unbridled power and discretion to ignore all those comments and retain the (by hypothesis) flagrantly illegal rule. Because by the time the first negative comment was timely filed, the period of judicial review was already closed. In our view, that makes a mindless hash of Congress’s statutory scheme. That’s particularly true when a much more reasonable reading of the statute is readily apparent: The DFR is reviewable, as virtually all agency action is, when the Department’s work is done, i.e., upon adherence.

*

In sum, statutory text and the presumption of reviewability all point in the same direction: DFR s are “prescribed” under § 6306(b)(1) when the rulemaking process concludes—not when it starts. When the Department receives adverse comments potentially requiring withdrawal, the process Page 17concludes when the Department rejects those comments and adheres to the DFR .

Here, the DFR became final on August 12. That is when the Department formally notified the States that it did not believe their negative comments required withdrawal of the DFR . See Confirmation Notice, Fed. Reg. at 65520. The States filed their petition for review 59 days later, on October 10. Because the petition was filed within § 6306(b)(1)'s 60-day deadline, our review is timely.³

2

None of the Department’s counterarguments is persuasive.

First , the Department denies that its position forecloses judicial review. The Department asserts that the States could have enjoyed review by simultaneously filing their adverse comments and petition for review. That is not at all obvious. In any other regulatory context, a petition for review that precedes a final decision would be unripe. Bennett v. Spear , 520 U.S. 154, 177 (1997). And if a party timely provides comments urging the agency to reconsider an otherwise-final rule, the doctrine of incurable prematurity makes the rule nonfinal as to that party until the agency rejects its comments. See, e.g. , Clifton Power Corp. v. FERC , 294 F.3d 108, 110–11 (D.C. Cir. 2002). Thus, we are unaware of any legal rule that directs parties to presume that an agency will unlawfully ignore its well-taken comment. Rather, the incurableprematurity doctrine directs parties to comment; wait until that comment is Page 18rejected; and then petition for judicial review. The Department offers no reason to think Congress meant to abrogate the incurable-prematurity doctrine when it enacted § 6295(p)(4). Moreover, even if we could stomach the Department’s comment-and-simultaneously-contest approach, that would allow the Department to shorten the comment period from 110 days to 60. Congress gave members of the public 110 days to comment—and to seek judicial review when the Department unlawfully ignores valid comments. The Department cannot overrule that choice and force American citizens to file their comments and petitions in only 60 days.

Second , the Department compares this case to Milice v. Consumer Product Safety Commission , 2 F.4th 994 (D.C. Cir. 2021). In Milice , the court held that a DFR was 'promulgat[ed]' on the day it was published, that a 60day deadline to petition began running on that date, and that a later-filed petition was untimely. Id. at 1000–02. The Department urges us to follow the same approach here.

To understand why Milice is not persuasive, you first must understand the unconventional scheme at issue in that case. Milice involved the Consumer Product Safety Improvement Act (' CPSIA '). The CPSIA allowed the Consumer Product Safety Commission (“ CPSC”) to adopt a safety standard from a private organization called ASTM International. When ASTM International updated its standard, CPSC was allowed to publish a notice in the Federal Register; give the public a chance to comment on the update; and then, barring meaningful objection, update the federal safety standard to reflect the new ASTM International standard. On September 20, 2019, the CPSC published a notice of intent to update its safety standard for infant bath seats, implementing an update by ASTM International. The notice gave the public 30 days to comment on the update; absent negative feedback, the update would go into effect on December 22, 2019. Milice petitioned for review on February 20, 2020, without ever filing Page 19a comment. The D.C. Circuit held her petition untimely because it was filed more than 60 days after the September 20 notice. See 15 U.S.C. § 2060(g)(2) (requiring any petition for review to be filed within 60 days of “promulgation”).

Milice is distinguishable in numerous ways. First, the CPSC did not shorten a comment period provided by Congress. The Commission gave the public 30 days to comment, and no act of Congress provided a different comment period. Second, no part of the agency's conduct in Milice was designed to avoid normal notice-and-comment rulemaking. In fact, the CPSC was disabled from publishing the new safety standard for notice and comment by a combination of the CPSIA and the fact that it was incorporating by reference a private organization’s revised safety standard. See Milice , 2 F.4th at 997; 5 U.S.C. § 552(a)(1) (incorporation-by-reference rules); Administrative Conference of the United States, Recommendation 2011-5, Incorporation by Reference , 77 Fed. Reg. 2257 (Dec. 8, 2011). Third, the CPSIA required any would-be petitioner to file within 60 days of the “promulgation” of a new standard. 15 U.S.C. § 2060(g)(2). It’s one thing to say a rule is 'promulgated' on the day it's originally published; it's an altogether different thing to say a rule is “prescribed” when it is published. 42 U.S.C. § 6306(b)(1). Fourth and finally, the D.C. Circuit suggested that Milice's petition could have been timely if she had commented as the petitioners did here; if Milice had done that, she could have (and indeed must have) waited until the agency rejected her comment before any petition would have been ripe. Milice , 2 F.4th at 1000-01. So it's unclear that applying Milice could even help the Department here, since all agree the States filed timely comments opposing the rule.

Third , the Department defends reading 'prescribed' to mean 'published' by pointing to the Second Circuit's decisions in Natural Resources Defense Council v. Abraham , 355 F.3d 179 (2d Cir. 2004) and Page 20Natural Resources Defense Council v. NHTSA , 894 F.3d 95 (2d Cir. 2018). True, Abraham read 'prescribed' in § 6306(b)(1) to refer to final publication. 355 F.3d at 196. And the court in NHTSA cited this reading favorably when interpreting another part of the EPCA . 894 F.3d at 105. But both Abraham and NHTSA concerned an ordinary rule, preceded by notice and comment. Both decisions relied on this detail to hold that, because the respective rules had cleared all procedural hurdles to finality, they should be treated as “prescribed” when published. See Abraham , 355 F.3d at 195–96; NHTSA , 894 F.3d at 105–06. Applying the same logic to DFR s means that they are not prescribed when published-with a DFR, publication is the beginning of the rulemaking process, not “the culminating event.” Abraham , 355 F.3d at 196. The Department’s citations confirm, rather than contravene, our reading.

Fourth , the Department argues that even if the DFR were “prescribed” when the agency decided not to withdraw it, Mississippi still missed the 60-day deadline. Because the EPCA requires that the Secretary make his withdrawal determination “not later than 120 days… [of the rule being] published ,” 42 U.S.C. § 6295(p)(4)(C)(i) (emphasis added), the States should have assumed that their comments were rejected on the 120th day— even before the agency did anything to accept or reject them. That's preposterous. Recall that the DFR at issue here provided nothing but a deadline for comments and an effective date. The States could only know that the agency had considered and rejected their adverse comments when the agency affirmatively told them so. Because that information came with the Confirmation Notice, that’s when the clock began running.

Consider the alternative: Parties would run to court on the 120th day and file petitions for review on the assumption that the Department would act unlawfully and ignore their comments. Again, the doctrine of incurable prematurity would bar any petition filed before the agency rejected the Page 21petitioner's comments. See Clifton Power , 294 F.3d at 112. And moreover, the Federal Reporter teems with decisions rejecting petitions for review and mandamus applications that are premised on agencies missing statutory deadlines like the 120-day provision in § 6295(p)(4)(C)(i). See, e.g. , Telecomm. Rsch. & Action Ctr. v. FCC , 750 F.2d 70, 80 (D.C. Cir. 1984) (creating a multi-factor balancing test for whether an agency violating a statutory deadline justifies intervention); In re Barr Laby’s., Inc. , 930 F.2d 72, 75 (D.C. Cir. 1991) ('[Violation of a statutory deadline] does not, alone, justify judicial intervention.”). And in all those cases, the federal courts say: Just give the agency a reasonable period of time, over and above the statutory deadline, to do the right thing. See ibid. ; see also 32 Wright & Miller’s Federal Practice & Procedure § 8168 (2d ed.) (“Courts, bowing to practical reality and their general disinclination to allocate scarce agency resources, commonly will allow an agency more time to complete a rulemaking.'). Moreover, if the Department could avoid judicial review through inaction, it would allow the governmental wrongdoer to benefit from nonfeasance. The Department offers no support for its guessing-game approach to DFR challenges.

Fifth , and finally, the Department claims that the States lack standing to bring this challenge. Specifically, it claims that if the States' challenge turns on the date the Department confirmed its adherence to the DFR in the Confirmation Notice, the challenge is really to the Confirmation Notice, not the DFR . And because the Confirmation Notice was “without legal effect,” it could not have injured the States within the meaning of Article III . See Red Br. at 12. But the Department is wrong to assume that using the Confirmation Notice as the relevant date for finality means the States are challenging the Notice instead of the DFR . The Department cites nothing to support this interpretive sophistry. For good reason: The States challenged the merits of Page 22the DFR , first by comment and later by petition. That petition is a challenge to the DFR , not to the notice confirming the rule’s finality.

*

In sum, the Department invites this court to interpret a statute to prevent a party from seeking judicial review of agency action—no matter how contrary to law that action really is. Because the Department’s argument runs contrary to statutory text, the longstanding presumption of judicial review, and common sense, we reject the invitation.

With our authority to hear this challenge settled, we turn to the merits. We ( 1 ) explain that the EPCA required the Department to withdraw the challenged DFR and ( 2 ) show why the Department errs in arguing otherwise.

The States’ comments identified several defects in the DFR , but here they largely focus on two of those purported errors. First, the States highlight the Nebraska comment's claim that the joint 'statement' preceding the DFR failed to include 'interested persons that are fairly representative of relevant points of view” as required by law. Blue Br. at 22. Second, they point to Utah and Montana's allegation that the DFR 's ban on linear power supplies failed to account for decreased product lifespan and was not economically justified. Thus, goes the argument, the Department erred in refusing to withdraw the DFR after these comments provided 'a reasonable basis' for doubting the DFR 's legality. We agree and grant the States' petition.

Start with the joint statement. This requirement is no formality—the Department’s authority to propose a DFR is triggered only upon receipt of Page 23a statement from “persons that are fairly representative” of stakeholders. 42 U.S.C . § 6295(p)(4)(A). Here, the Department did not receive a statement from representative stakeholders. Instead, it polled a select group of NGO s and industry representatives, even though the Department was on notice that numerous States had opposed similar standards for stoves and cooktops in the previous notice-and-comment rulemaking. Rather than including all stakeholders, the Department appears to have particularly excluded States it knew would oppose the DFR . What's worse, the EPCA specifically mentions States as stakeholders in describing the fair representation requirement. Ibid. The Department's failure to include States beggars belief. That alone should have triggered withdrawal of the DFR .

Next, take the DFR 's ban on linear power supplies. As Utah and Montana noted in their comment, switching power supplies (the Department’s preference) are more mechanically complex than linear ones. It is a generally accepted rule of mechanical engineering that the more complex the system, the more prone it is to failure. See ARID.190 (citing engineering sources). Yet in banning linear supplies and effectively mandating switching supplies, the Department did not account for any decrease in reliability. See DFR , 89 Fed. Reg. at 11477-78. Indeed, the Department seems to have assumed that appliances using linear supplies would have the same maintenance costs and lifespans as those using switching supplies. But that is not true. As the States pointed out, more frequent repair and replacement of appliances both reduce utility and increase energy consumption at the manufacturing and disposal stages. The Department should have accounted for these potential effects: Section 6295(o)(2)(B)(i) of the EPCA specifically lists lifetime maintenance, utility, and performance as relevant factors that must be weighed when determining whether a standard is economically justified. See 42 U.S.C. § 6295(o)(2)(B)(i)(II)–(IV) (obligating the Department to Page 24consider, among other things, “the savings in operating costs throughout the estimated average life of the covered product . . . compared to any increase in . . . maintenance expenses,” “the total projected amount of energy . . . savings,” and “any lessening of the utility or the performance of the covered products likely to result from the imposition of the standard”). These factors have teeth: Section 6295(o)(2)(A) makes clear that the Department may only prescribe a standard it “determines” is “economically justified” under the factors. Yet the Department simply did not account for whether increased complexity required by its standard would reduce reliability or increase associated costs, let alone weigh that effect as required by § 6295(o). See DFR , 89 Fed. Reg. at 11477–78.

Further, Congress did not erect a high bar to trigger the Secretary’s withdrawal duty. Comments need only reveal a possibility of unlawfulness: The Department ' shall withdraw' a rule upon a showing that adverse comments ' may provide a reasonable basis” for doing so. See id. § 6295(p)(4)(C)(i)(II) (emphases added). Congress’s use of “may” merely requires “probabilit[y], not certaint[y].” United States v. ALCOA , 377 U.S. 271, 280 (1964); see also United States v. Lexington Mill & Elevator Co. , 232 U.S. 399, 411 (1914). And Congress was equally clear in imposing a mandatory duty to withdraw on the Department. See Maine Cmty. Health Options v. United States , 590 U.S. 296, 310 (2020) ('[T]he word shall usually connotes a requirement.”). The States’ comments met this low threshold.

The Department was free, of course, to proceed the way rulemaking agencies normally do—by proposing its standards in an NPRM , responding in non-arbitrary and non-capricious ways to public comments, and then promulgating a final rule that comports with the EPCA . What the Department could not do was gloss over public outcry, allow the public to file after-the-fact comments, and then ignore those comments in violation of Congress’s plain dictates. See 42 U.S.C. § 6295(p)(4)(C)(i)(II).

The Department’s counterarguments are unpersuasive.

First , the Department claims that the joint statement fairly represented stakeholders because it was 'submitted jointly' by industry groups, efficiency advocates, and the States of California, Massachusetts, and New York. But it is most accurate to say that the statement here was not “submitted jointly” by any States at all. While the Department cites three state signatories, the Department actually is referring to comments that were filed after the joint statement was received, purporting to 'support' its standards. The text of § 6295(p)(4)(A) refers to a singular “statement,” not after-the-fact comments supporting a statement. Plus, the comments at issue were not submitted by officials acting in the sovereign capacity of California, Massachusetts, or New York, but by representatives from various state agencies. Yet § 6295(p)(4)(A) refers to “States,” even capitalizing the term to denote the sovereigns themselves. The Department cites no authority for the proposition that after-the-fact comments received by three handpicked state agency officials satisfies the reference to States as relevant stakeholders that must be fairly represented in § 6295(p)(4)(A).

Second , the Department appears to recognize that California, Massachusetts, and New York are not fairly representative of the United States—the understatement of the day—but it contends the EPCA does not require representation 'across the ideological spectrum.' Oral Arg. at 28:25–35. That is true but irrelevant. The EPCA says nothing about perfect “ideological” or partisan representation. But it does require “fair[] representati[on].” 42 U.S.C. § 6295(p)(4)(A). This requirement for broad consensus maps neatly onto the origins of DFR practice under the APA . See Part II . B , supra . And it cannot be contended with a straight face that California, Massachusetts, and New York “fairly represent” a cross-section Page 26of States. That is particularly so given numerous States-far more than three—opposed the 2023 notice-and-comment attempt to enact strikingly similar standards. Consider the absurd implications of the Department's position. ATF , for example, could issue a gun-related DFR after considering a 'fair representation' of views in Texas, Oklahoma, and Alabama. The EPA could issue an ethanol-fuel DFR after considering a “fair representation” of views in Iowa and Nebraska. That would turn DFR s— which were supposed to be instruments for efficient rulemaking in the face of consensus—into partisan weapons that can cut off judicial review in the most hotly contested issues of the day.

Third , the Department disputes that it was required to consider reductions in reliability and, alternatively, that it did consider reliability in promulgating the DFR . The Department seems to be of two minds on this point: On the one hand, it dismisses the States' reliability concerns as “theoretical conjecture.” Red Br. at 28. But, on the other, it claims to have carefully considered the issue and determined that Utah and Montana's comment 'failed to rebut the data before the agency.' Ibid. That data is summarized in a portion of the DFR ’s Confirmation Notice, which states that the Department read online reliability surveys and found that, at least in those surveys, there was “no indication that higher-efficiency products are less reliable . . . relative to baseline products.” 89 Fed. Reg. at 65526. The Department also points to comments from manufacturers who signed the joint statement and stated that they did not expect the DFR ’s standards to decrease reliability. Ibid.

But the Department misunderstands the relevant standard. The EPCA does not require commenters to 'rebut [the] data before the agency' and prove the DFR ’s unlawfulness in order to trigger withdrawal. Red Br. at 28. Rather, the States were only required to lodge adverse comments, contesting the rule in good faith. At that point, the question became whether Page 27those adverse comments provided a “reasonable basis” for withdrawing the rule. 42 U.S.C. § 6295(p)(4)(C)(i)(II).

The States' comments plainly did that. Utah and Montana's comment, for example, provided engineering sources to substantiate its concerns about switching power supplies decreasing reliability. See Confirmation Notice, 89 Fed. Reg. at 65526 (describing Utah and Montana’s comment). None of the Department’s arguments on reliability assuaged that concern. For example, the Department’s proffered surveys may well show that more efficient models are not always more costly to repair than less efficient models. Ibid. But even if that trend were true generally, it does not reveal whether requiring switching power supplies universally would reduce reliability across all models (which is what the Utah and Montana comment alleged). Nor do the comments from manufacturers move the needle—the bald assertion that manufacturers did not expect repair costs to increase is no better than the Department’s bald assertion of the same. Ibid. In short, the States provided reasonable grounds to show that the DFR was not premised on consensus; that real problems needed additional study; and that the Department could not regulate first and ask questions never. Those concerns triggered withdrawal of the DFR under the EPCA 's lenient 'reasonable basis” standard, and the Department violated the statute when it refused to withdraw the DFR . See Texas v. EPA , 156 F.4th 523, 540 (5th Cir. 2025) (noting the 'established principle that a reviewing court must determine whether the agency complied with the procedures mandated by the relevant statutes” (quotation omitted)). Accordingly, we grant the States’ petition to “set aside” the invalidly retained rule. 5 U.S.C. § 706(2)(D).

* * *

The previous administration's gas-stove measures were among its most controversial energy proposals. The idea that all that controversy could Page 28be elided using the DFR —a mechanism designed for consensus rulemaking—is untenable. And it’s all the more untenable in the face of the EPCA ’s comment and review provisions, which the Department would have us rewrite. We reject the invitation.

The Department’s motion to dismiss the petition for review for lack of jurisdiction is DENIED . The petition for review is GRANTED and the matter is REMANDED to the Department for further proceedings consistent with this opinion.

Page 29Haynes , Circuit Judge , concurring in part:

Regarding jurisdiction, while I think the Department of Energy makes some good points, I concur with the conclusion of the prescribed timing in the majority opinion. Accordingly, on this point, I concur solely in the ultimate conclusion that we have jurisdiction. Turning to the merits, my view is that we should simply remand to the Department to reconsider its decisions in this situation in light of the information set forth in this appeal.

Accordingly, I agree with remanding but only on that basis.