Introduction: Judicial Review of Agency Actions

To this point, we’ve learned about agencies’ procedural obligations under the APA and their authorizing statutes, and how courts review potential violations of procedural requirements. In this part of our course, we learn about how courts review agency actions to determine whether they are substantively unlawful. You will learn how courts review agency actions where an agency’s decision is arbitrary and capricious, or where an agency misinterpreted a statute.

Judicial Review of Statutory Interpretations

We will first learn about how courts review agencies’ statutory interpretations. This review often arises when the meaning of constitutional, statutory, or regulatory provisions are in question. Because federal agencies get their authority from Congressional statutes, much of agencies’ decisionmaking includes interpreting statutes. Agencies interpret statutes to determine the scope of their rulemaking and adjudicatory authority, and to hammer out the details of their rules, orders, and guidance.

APA Section 706(2)(A) says that courts can hold unlawful and set aside agency actions that are “not in accordance with the law,” and APA Section 706(2)(C) says that courts can hold unlawful and set aside agency actions that are “in excess of statutory jurisdiction, authority, or limitations, or short of statutory right.”

Courts give agency decisions a lot of “deference.” Deference is defined as “humble submission and respect,” or “giving agencies the benefit of the doubt.” Courts give agency decisions a lot of leeway for several reasons: 

Courts are supposed to be politically neutral, while agencies are acting under authority delegated to them by elected Congresspeople and the President. Courts strive to respect the division of powers between the three branches of government.

Agencies are staffed with experts on the things they regulate: the EPA is teeming with environmental scientists and experts, and the Department of Health and Human Services (“HHS”) hires doctors and public health experts to carry out its Congressionally delegated mandates.

In Skidmore, a case decided in 1944 (long before Chevron) lists the factors courts consider when they extend “lesser deference” to agencies’ interpretations of statutes when those interpretations lack the “force of law.”

In Chevron v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984), the Supreme Court provides the test that courts use to determine how much deference to give agencies’ statutory interpretations.

 5 U.S. Code § 706. Scope of review

To the extent necessary to decision and when presented, the reviewing court shall decide all relevant questions of law, interpret constitutional and statutory provisions, and determine the meaning or applicability of the terms of an agency action. 

The reviewing court shall—

(1) compel agency action unlawfully withheld or unreasonably delayed; and

(2)hold unlawful and set aside agency action, findings, and conclusions found to be—

(A) arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law;

(B) contrary to constitutional right, power, privilege, or immunity;

(C) in excess of statutory jurisdiction, authority, or limitations, or short of statutory right;

(D) without observance of procedure required by law;

(E) unsupported by substantial evidence in a case subject to sections 556 and 557 of this title or otherwise reviewed on the record of anagency hearing provided by statute; or

(F) unwarranted by the facts to the extent that the facts are subject to trial de novo by the reviewing court.

In making the foregoing determinations, the court shall review the whole record or those parts of it cited by a party, and due account shall be taken of the rule of prejudicial error.

In making the foregoing determinations, the court shall review the whole record or those parts of it cited by a party, and due account shall be taken of the rule of prejudicial error.

Skidmore v. Swift & Co.

323 U.S. 134 (1944)

1. JUSTICE JACKSON delivered the opinion of the Court.

Seven employees of the Swift and Company packing plant at Fort Worth, Texas, brought an action under the Fair Labor Standards Act to recover overtime. [The employees worked at the packing plant during the day and were paid weekly salaries for that work. However, the same employees also stayed at the fire hall (or “within hailing distance”) three and a half to four nights a week to answer alarms in case of a fire. For this work, employees would be paid fifty to sixty-four cents for every alarm they answered, but would not be paid hourly overtime wages. The employees seek that overtime pay.]

The trial court found the evidentiary facts as stipulated; it made no findings of fact, as such, as to whether, under the arrangement of the parties and the circumstances of this case . . . the fire hall duty or any part thereof constituted working time. The trial court found [...] as a “conclusion of law” that “the time plaintiffs spent in the fire hall subject to call to answer fire alarms does not constitute hours worked, for which overtime compensation is due them under the Fair Labor Standards Act, as interpreted by the Administrator and the Courts,” and in its opinion observed, “of course we know pursuing such pleasurable occupations or performing such personal chores, does not constitute work.” The Circuit Court of Appeals affirmed.

[W]e hold that no principle of law found either in the statute or in Court decisions precludes waiting time from also being working time. We have not attempted to, and we cannot, lay down a legal formula to resolve cases so varied in their facts as are the many situations in which employment involves waiting time. Whether, in a concrete case, such time falls within or without the Act is a question of fact to be resolved by appropriate findings of the trial court. [...] We do not minimize the difficulty of such an inquiry where the arrangements of the parties have not contemplated the problem posed by the statute. But it does not differ in nature or in the standards to guide judgment from that which frequently confronts courts where they must find retrospectively the effect of contracts as to matters which the parties failed to anticipate or explicitly to provide for.

Congress did not utilize the services of an administrative agency to find facts and to determine in the first instance whether particular cases fall within or without the Act. Instead, it put this responsibility on the courts. But it did create the office of Administrator, impose upon him a variety of duties, endow him with powers to inform himself of conditions in industries and employments subject to the Act, and put on him the duties of bringing injunction actions to restrain violations. Pursuit of his duties has accumulated a considerable experience in the problems of ascertaining working time in employments involving periods of inactivity and a knowledge of the customs prevailing in reference to their solution. From these he is obliged to reach conclusions as to conduct without the law, so that he should seek injunctions to stop it, and that within the law, so that he has no call to interfere. He has set forth his views of the application of the Act under different circumstances in an interpretative bulletin and in informal rulings. They provide a practical guide to employers and employees as to how the office representing the public interest in its enforcement will seek to apply it. 

The Administrator thinks the problems presented by inactive duty require a flexible solution, rather than the all-in or all-out rules respectively urged by the parties in this case, and his Bulletin endeavors to suggest standards and examples to guide in particular situations. In some occupations, it says, periods of inactivity are not properly counted as working time even though the employee is subject to call. Examples are an operator of a small telephone exchange where the switchboard is in her home and she ordinarily gets several hours of uninterrupted sleep each night; or a pumper of a stripper well or watchman of a lumber camp during the off season, who may be on duty twenty-four hours a day but ordinarily “has a normal night’s sleep, has ample time in which to eat his meals, and has a certain amount of time for relaxation and entirely private pursuits.” [...] “Hours worked are not limited to the time spent in active labor but include time given by the employee to the employer. . . .”

The facts of this case do not fall within any of the specific examples given, but the conclusion of the Administrator, as expressed in the brief amicus curiae, is that the general tests which he has suggested point to the exclusion of sleeping and eating time of these employees from the workweek and the inclusion of all other on-call time: although the employees were required to remain on the premises during the entire time, the evidence shows that they were very rarely interrupted in their normal sleeping and eating time, and these are pursuits of a purely private nature which would presumably occupy the employees’ time whether they were on duty or not and which apparently could be pursued adequately and comfortably in the required circumstances; the rest of the time is different because there is nothing in the record to suggest that, even though pleasurably spent, it was spent in the ways the men would have chosen had they been free to do so.

There is no statutory provision as to what, if any, deference courts should pay to the Administrator’s conclusions. And, while we have given them notice, we have had no occasion to try to prescribe their influence. The rulings of this Administrator are not reached as a result of hearing adversary proceedings in which he finds facts from evidence and reaches conclusions of law from findings of fact. They are not, of course, conclusive, even in the cases with which they directly deal, much less in those to which they apply only by analogy. They do not constitute an interpretation of the Act or a standard for judging factual situations which binds a district court’s processes, as an authoritative pronouncement of a higher court might do. But the Administrator’s policies are made in pursuance of official duty, based upon more specialized experience and broader investigations and information than is likely to come to a judge in a particular case. They do determine the policy which will guide applications for enforcement by injunction on behalf of the Government. Good administration of the Act and good judicial administration alike require that the standards of public enforcement and those for determining private rights shall be at variance only where justified by very good reasons. The fact that the Administrator’s policies and standards are not reached by trial in adversary form does not mean that they are not entitled to respect. This Court has long given considerable and in some cases decisive weight to Treasury Decisions and to interpretative regulations of the Treasury and of other bodies that were not of adversary origin.

We consider that the rulings, interpretations and opinions of the Administrator under this Act, while not controlling upon the courts by reason of their authority, do constitute a body of experience and informed judgment to which courts and litigants may properly resort for guidance. The weight of such a judgment in a particular case will depend 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.

[In] this case, although the District Court referred to the Administrator’s Bulletin, its evaluation and inquiry were apparently restricted by its notion that waiting time may not be work, an understanding of the law which we hold to be erroneous. Accordingly, the judgment is reversed and the cause remanded for further proceedings consistent herewith.

Chevron v. Natural Resources Defense Council, Inc. (“NRDC”)

467 U.S. 837 (1984)

JUSTICE STEVENS delivered the opinion of the Court.

In the Clean Air Act Amendments of 1977, Congress enacted certain requirements applicable to States that had not achieved the national air quality standards established by the Environmental Protection Agency (EPA). [The 1977 Amendments required permits for “new or modified stationary sources” of air pollution. The EPA promulgated a rule that interpreted the phrase “stationary source” to include a “bubble policy.” The bubble policy treats industrial plants with many pollution-emitting devices (“smokestacks,” etc.) as a single “stationary source.” This interpretation of the statute lets plants with multiple air-polluting devices install or modify one piece of equipment without obtaining a permit, so long as the alteration does not increase the plant’s total emissions. NRDC argued that the EPA misinterpreted the phrase “stationary source,” arguing that Congress did not intend it to include a “bubble policy.” Instead, the NRDC argued that Congress intended that a plant would have to obtain a permit any time it created a new source of pollution or modified an existing source, if the effect is to increase pollution, even if the increased pollution is offset by decreasing pollution from other sources. The Court of Appeals agreed with NRDC, because it believed the NRDC’s interpretation better served the spirit of the 1977 Clean Air Act amendments.]

The basic legal error of the Court of Appeals was to adopt a static judicial definition of the term “stationary source” when it had decided that Congress itself had not commanded that definition [...] 

When a court reviews an agency’s construction of the statute which it administers, it is confronted with two questions. First, always, is the question whether Congress has directly spoken to the precise question at issue. If the intent of Congress is clear, that is the end of the matter; for the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress. If, however, the court determines Congress has not directly addressed the precise question at issue, the court does not simply impose its own construction on the statute, as would be necessary in the absence of an administrative interpretation. Rather, if the statute is silent or ambiguous with respect to the specific issue, the question for the court is whether the agency’s answer is based on a permissible construction of the statute.

“The power of an administrative agency to administer a congressionally created . . . program necessarily requires the formulation of policy and the making of rules to fill any gap left, implicitly or explicitly, by Congress.” Morton v. Ruiz, 415 U. S. 199, 231 (1974). If Congress has explicitly left a gap for the agency to fill, there is an express delegation of authority to the agency to elucidate a specific provision of the statute by regulation. Such legislative regulations are given controlling weight unless they are arbitrary, capricious, or manifestly contrary to the statute. Sometimes the legislative delegation to an agency on a particular question is implicit rather than explicit. In such a case, a court may not substitute its own construction of a statutory provision for a reasonable interpretation made by the administrator of an agency.

We have long recognized that considerable weight should be accorded to an executive department's construction of a statutory scheme it is entrusted to administer, and the principle of deference to administrative interpretations has been consistently followed by this Court whenever decision as to the meaning or reach of a statute has involved reconciling conflicting policies, and a full understanding of the force of the statutory policy in the given situation has depended upon more than ordinary knowledge respecting the matters subjected to agency regulations. 

“. . . If this choice represents a reasonable accommodation of conflicting policies that were committed to the agency’s care by the statute, we should not disturb it unless it appears from the statute or its legislative history that the accommodation is not one that Congress would have sanctioned.” United States v. Shimer, 367 U. S. 374, 382, 383 (1961).

In light of these well-settled principles it is clear that the Court of Appeals misconceived the nature of its role in reviewing the regulations at issue. Once it determined, after its own examination of the legislation, that Congress did not actually have an intent regarding the applicability of the bubble concept to the permit program, the question before it was not whether in its view the concept is “inappropriate” in the general context of a program designed to improve air quality, but whether the Administrator’s view that it is appropriate in the context of this particular program is a reasonable one. Based on the examination of the legislation and its history which follows, we agree with the Court of Appeals that Congress did not have a specific intention on the applicability of the bubble concept in these cases, and conclude that the EPA’s use of that concept here is a reasonable policy choice for the agency to make.

[The Court reviews the legislative history of the 1977 Clean Air Act amendments, and analyzes the language and Congressional intent of those amendments. While the amendments do not specifically reference the “bubble concept,” the EPA explained, in depth, its rationale for considering a bubble exemption in its proposed rule. In contrast, the legislative history does not mention the “bubble” concept, nor does it speak directly to the “stationary source” interpretation at hand. The Court concluded, “We find that the legislative history as a whole is silent on the precise issue before us. It is, however, consistent with the view that the EPA should have broad discretion in implementing the policies of the 1977 Amendments.”]

The arguments over policy that are advanced in the parties’ briefs create the impression that respondents are now waging in a judicial forum a specific policy battle which they ultimately lost in the agency and in the 32 jurisdictions opting for the “bubble concept,” but one which was never waged in the Congress. Such policy arguments are more properly addressed to legislators or administrators, not to judges.

In these cases the Administrator’s interpretation represents a reasonable accommodation of manifestly competing interests and is entitled to deference: the regulatory scheme is technical and complex, the agency considered the matter in a detailed and reasoned fashion, and the decision involves reconciling conflicting policies. Congress intended to accommodate both interests, but did not do so itself on the level of specificity presented by these cases. 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.

Judges are not experts in the field, and are not part of either political branch of the Government. Courts must, in some cases, reconcile competing political interests, but not on the basis of the judges’ personal policy preferences. In contrast, an agency to which Congress has delegated policymaking responsibilities may, within the limits of that delegation, properly rely upon the incumbent administration’s views of wise policy to inform its judgments. While agencies are not directly accountable to the people, the Chief Executive is, and it is entirely appropriate for this political branch of the Government to make such policy choices — resolving the competing interests which Congress itself either inadvertently did not resolve, or intentionally left to be resolved by the agency charged with the administration of the statute in light of everyday realities.

When a challenge to an agency construction of a statutory provision, fairly conceptualized, really centers on the wisdom of the agency's policy, rather than whether it is a reasonable choice within a gap left open by Congress, the challenge must fail. In such a case, federal judges — who have no constituency — have a duty to respect legitimate policy choices made by those who do. The responsibilities for assessing the wisdom of such policy choices and resolving the struggle between competing views of the public interest are not judicial ones: “Our Constitution vests such responsibilities in the political branches.” TVA v. Hill, 437 U. S. 153, 195 (1978).

We hold that the EPA’s definition of the term “source” is a permissible construction of the statute which seeks to accommodate progress in reducing air pollution with economic growth. “The Regulations which the Administrator has adopted provide what the agency could allowably view as . . . [an] effective reconciliation of these twofold ends . . . .” United States v. Shimer, 367 U. S., at 383.

The judgment of the Court of Appeals is reversed.

The Chevron doctrine gives agencies a lot of deference to interpret statutes where Congress has not spoken directly to the issue. However, there are several limits to that deference. This lesson will describe the boundaries of Chevron deference, namely:

• Chevron deference only applies when agencies’ interpretations have the “force of law.”

• Chevron deference does not apply when judges decide that the agency interpretation concerns a “major question” with “vast economic or political significance.”

Limits on Chevron Deference: Force of Law Requirement

After issuing the Chevron opinion, the Supreme Court clarified that Chevron deference only applies to agency interpretations that have the force of law. The benchmark case for this limitation on Chevron deference is United States v. Mead Corporation.

Justice Souter references two cases throughout the Mead decision: Skidmore and Christensen. Skidmore is a case from 1944 (long before Chevron or Mead) that lists the factors courts consider when they extend “lesser deference” to agencies’ interpretations of statutes when those interpretations lack the “force of law.” Christensen, which was decided only a year before Mead, explains why courts should extend some deference to agencies’ statutory interpretations even when they do not have the force of law

Christensen v. Harris County

529 U.S. 576 (2000)

Justice Thomas, delivered the opinion of the Court.

Under the Fair Labor Standards Act of 1938 (FLSA), States and their political subdivisions may compensate their employees for overtime by granting them compensatory time or “comp time,” which entitles them to take time off work with full pay. If the employees do not use their accumulated compensatory time, the employer is obligated to pay cash compensation under certain circumstances. Fearing the fiscal consequences of having to pay for accrued compensatory time, Harris County adopted a policy requiring its employees to schedule time off in order to reduce the amount of accrued compensatory time. Employees of the Harris County Sheriff's Department sued, claiming that the FLSA prohibits such a policy. The Court of Appeals rejected their claim. Finding that nothing in the FLSA or its implementing regulations prohibits an employer from compelling the use of compensatory time, we affirm [...]

I

Petitioners are 127 deputy sheriffs employed by respondents Harris County, Texas, and its sheriff, Tommy B. Thomas (collectively, Harris County). It is undisputed that each of the petitioners individually agreed to accept compensatory time, in lieu of cash, as compensation for overtime.

As petitioners accumulated compensatory time, Harris County became concerned that it lacked the resources to pay monetary compensation to employees who worked overtime after reaching the statutory cap on compensatory time accrual and to employees who left their jobs with sizable reserves of accrued time. As a result, the county began looking for a way to reduce accumulated compensatory time. It wrote to the United States Department of Labor’s Wage and Hour Division, asking “whether the Sheriff may schedule non-exempt employees to use or take compensatory time.” The Acting Administrator of the Division replied:

“[I]t is our position that a public employer may schedule its nonexempt employees to use their accrued FLSA compensatory time as directed if the prior agreement specifically provides such a provision . . . .

“Absent such an agreement, it is our position that neither the statute nor the regulations permit an employer to require an employee to use accrued compensatory time." Opinion Letter from Dept. of Labor, Wage and Hour Div. (Sept. 14, 1992).

After receiving the letter, Harris County implemented a policy under which the employees’ supervisor sets a maximum number of compensatory hours that may be accumulated. When an employee's stock of hours approaches that maximum, the employee is advised of the maximum and is asked to take steps to reduce accumulated compensatory time. If the employee does not do so voluntarily, a supervisor may order the employee to use his compensatory time at specified times.

Petitioners sued, claiming that the county’s policy violates the FLSA because § 207(o)(5)—which requires that an employer reasonably accommodate employee requests to use compensatory time—provides the exclusive means of utilizing accrued time in the absence of an agreement or understanding permitting some other method. The District Court agreed, granting summary judgment for petitioners and entering a declaratory judgment that the county’s policy violated the FLSA. The Court of Appeals for the Fifth Circuit reversed, holding that the FLSA did not speak to the issue and thus did not prohibit the county from implementing its compensatory time policy [...] We granted certiorari because the Courts of Appeals are divided on the issue [...]

At bottom, we think the better reading of § 207(o)(5) is that it imposes a restriction upon an employer’s efforts to prohibit the use of compensatory time when employees request to do so; that provision says nothing about restricting an employer’s efforts to require employees to use compensatory time. Because the statute is silent on this issue and because Harris County’s policy is entirely compatible with § 207(o)(5), petitioners cannot, as they are required to do by 29 U. S. C. § 216(b), prove that Harris County has violated § 207 [...]

In an attempt to avoid the conclusion that the FLSA does not prohibit compelled use of compensatory time, petitioners and the United States contend that we should defer to the Department of Labor’s opinion letter, which takes the position that an employer may compel the use of compensatory time only if the employee has agreed in advance to such a practice. Specifically, they argue that the agency opinion letter is entitled to deference under our decision in Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U. S. 837 (1984). In Chevron, we held that a court must give effect to an agency’s regulation containing a reasonable interpretation of an ambiguous statute.

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. Instead, interpretations contained in formats such as opinion letters are “entitled to respect” under our decision in Skidmore v. Swift & Co., 323 U. S. 134, 140 (1944), but only to the extent that those interpretations have the “power to persuade,” As explained above, we find unpersuasive the agency’s interpretation of the statute at issue in this case.

Of course, the framework of deference set forth in Chevron does apply to an agency interpretation contained in a regulation. But in this case the Department of Labor’s regulation does not address the issue of compelled compensatory time. The regulation provides only that “[t]he agreement or understanding [between the employer and employee] may include other provisions governing the preservation, use, or cashing out of compensatory time so long as these provisions are consistent with [§ 207(o )].” 29 CFR § 553.23(a)(2) (1999) (emphasis added). Nothing in the regulation even arguably requires that an employer’s compelled use policy must be included in an agreement. The text of the regulation itself indicates that its command is permissive, not mandatory.

Seeking to overcome the regulation’s obvious meaning, the United States asserts that the agency’s opinion letter interpreting the regulation should be given deference under our decision in Auer v. Robbins, 519 U. S. 452 (1997). In Auer, we held that an agency’s interpretation of its own regulation is entitled to deference. But Auer deference is warranted only when the language of the regulation is ambiguous. The regulation in this case, however, is not ambiguous—it is plainly permissive. To defer to the agency’s position would be to permit the agency, under the guise of interpreting a regulation, to create de facto a new regulation. Because the regulation is not ambiguous on the issue of compelled compensatory time, Auer deference is unwarranted.

* * *

As we have noted, no relevant statutory provision expressly or implicitly prohibits Harris County from pursuing its policy of forcing employees to utilize their compensatory time. In its opinion letter siding with the petitioners, the Department of Labor opined that “it is our position that neither the statute nor the regulations permit an employer to require an employee to use accrued compensatory time.” Opinion Letter (emphasis added). But this view is exactly backwards. Unless the FLSA prohibits respondents from adopting its policy, petitioners cannot show that Harris County has violated the FLSA. And the FLSA contains no such prohibition. The judgment of the Court of Appeals is affirmed.

United States v. Mead Corp.

533 U.S. 218 (2001)

Justice Souter, delivered the opinion of the Court.

The question is whether a tariff classification ruling by the United States Customs Service deserves judicial deference. The Federal Circuit rejected Customs's invocation of Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U. S. 837 (1984), in support of such a ruling, to which it gave no deference. We agree that a tariff classification has no claim to judicial deference under Chevron, there being no indication that Congress intended such a ruling to carry the force of law, but we hold that under Skidmore v. Swift & Co., 323 U. S. 134 (1944), the ruling is eligible to claim respect according to its persuasiveness.

I

Imports are taxed under the Harmonized Tariff Schedule of the United States (HTSUS), 19 U.S.C. § 1202. [The HTSUS authorizes Customs to fix the final classification and rate of duty applicable to merchandise, and to disseminate guidance about how to secure “a just, impartial, and uniform appraisement of imported merchandise.” The agency fulfills this statutory obligation through “ruling letters” that set tariff classifications for particular imports. According to 19 C.F.R. § 177.9, a ruling letter:

“represents the official position of the Customs Service with respect to the particular transaction or issue described therein and is binding on all Customs Service personnel in accordance with the provisions of this section until modified or revoked. In the absence of a change of practice or other modification or revocation which affects the principle of the ruling set forth in the ruling letter, that principle may be cited as authority in the disposition of transactions involving the same circumstances.”

A ruling letter is binding on all Customs Service personnel until it is modified or revoked, but the letter can be modified or revoked without notice to any person beyond the importer to whom the letter was addressed. The regulations also say that “no other person should rely on the ruling letter or assume that the principles of that ruling will be applied in connection with any transaction other than the one described in the letter.”] Since ruling letters respond to transactions of the moment, they are not subject to notice and comment before being issued, may be published but need only be made “available for public inspection.” [...]

Respondent, the Mead Corporation, imports “day planners,” three-ring binders with pages having room for notes of daily schedules and phone numbers and addresses, together with a calendar and suchlike. The tariff schedule on point falls under the HTSUS heading for “[r]egisters, account books, notebooks, order books, receipt books, letter pads, memorandum pads, diaries and similar articles,” which comprises two subcategories. Items in the first, “[d]iaries, notebooks and address books, bound; memorandum pads, letter pads and similar articles,” were subject to a tariff of 4.0% at the time in controversy. Objects in the second, covering “[o]ther” items, were free of duty. 

Between 1989 and 1993, Customs repeatedly treated day planners under the “other” HTSUS subheading. In January 1993, however, Customs changed its position, and issued a ruling letter classifying Mead’s day planners as “Diaries . . . , bound” subject to [the 4.0% tariff.] [Mead challenged the re-classification in the Court of International Trade (“CIT”), which granted the Government’s motion for summary judgment. Mead appealed the decision before the Federal Circuit, where the Government argued that classification rulings, like Customs regulations, deserve Chevron deference. The Federal Court reversed the CIT decision, deciding that classification rulings had a weaker Chevron claim because they “do not carry the force of law and are not, like regulations, intended to clarify the rights and obligations of importers beyond the specific case under review.”] 

We granted certiorari in order to consider the limits of Chevron deference owed to administrative practice in applying a statute. We hold that administrative implementation of a particular statutory provision qualifies for Chevron deference 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. Delegation of such authority may be shown in a variety of ways, as by an agency’s power to engage in adjudication or notice-and-comment rulemaking, or by some other indication of a comparable congressional intent. The Customs ruling at issue here fails to qualify, although the possibility that it deserves some deference under Skidmore leads us to vacate and remand.

II

When Congress has “explicitly left a gap for an agency to fill, there is an express delegation of authority to the agency to elucidate a specific provision of the statute by regulation,” Chevron, 467 U. S., and any ensuing regulation is binding in the courts unless procedurally defective, arbitrary or capricious in substance, or manifestly contrary to the statute. See [...] APA, 5 U.S.C. §§ 706(2)(A), (D). But whether or not they enjoy any express delegation of authority on a particular question, agencies charged with applying a statute necessarily make all sorts of interpretive choices, and while not all of those choices bind judges to follow them, they certainly may influence courts facing questions the agencies have already answered. “[T]he well-reasoned views of the agencies implementing a statute ‘constitute a body of experience and informed judgment to which courts and litigants may properly resort for guidance,’” Bragdon v. Abbott, 524 U.S. (1998) (quoting Skidmore, 323 U.S.), and “[w]e have long recognized that considerable weight should be accorded to an executive department’s construction of a statutory scheme it is entrusted to administer . . . .” Chevron. The fair measure of deference to an agency administering its own statute has been understood to vary with circumstances, and courts have looked to the degree of the agency’s care, its consistency, formality, and relative expertness, and to the persuasiveness of the agency’s position. The approach has produced a spectrum of judicial responses, from great respect at one end, to near indifference at the other. Justice Jackson summed things up in Skidmore v. Swift & Co.:

“The weight [accorded to an administrative] judgment in a particular case will depend 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.”

Since 1984, we have identified a category of interpretive choices distinguished by an additional reason for judicial deference. This Court in Chevron recognized that Congress not only engages in express delegation of specific interpretive authority, but that “[s]ometimes the legislative delegation to an agency on a particular question is implicit.” Congress, that is, may not have expressly delegated authority or responsibility to implement a particular provision or fill a particular gap. Yet it can still be apparent from the agency’s generally conferred authority and other statutory circumstances that Congress would expect the agency to be able to speak with the force of law when it addresses ambiguity in the statute or fills a space in the enacted law, even one about which “Congress did not actually have an intent” as to a particular result. When circumstances implying such an expectation exist, a reviewing court has no business rejecting an agency’s exercise of its generally conferred authority to resolve a particular statutory ambiguity simply because the agency's chosen resolution seems unwise, but is obliged to accept the agency’s position if Congress has not previously spoken to the point at issue and the agency's interpretation is reasonable, see [...] 5 U.S.C. § 706(2) (a reviewing court shall set aside agency action, findings, and conclusions found to be “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law”).

We have recognized a very good indicator of delegation meriting Chevron treatment in express congressional authorizations to engage in the process of rulemaking or adjudication that produces regulations or rulings for which deference is claimed. It is fair to assume generally that Congress contemplates administrative action with the effect of law when it provides for a relatively formal administrative procedure tending to foster the fairness and deliberation that should underlie a pronouncement of such force. Thus, the overwhelming number of our cases applying Chevron deference have reviewed the fruits of notice-and-comment rulemaking or formal adjudication. That said, and as significant as notice-and-comment is in pointing to Chevron authority, the want of that procedure here does not decide the case, for we have sometimes found reasons for Chevron deference even when no such administrative formality was required and none was afforded. The fact that the tariff classification here was not a product of such formal process does not alone, therefore, bar the application of Chevron.

There are, nonetheless, ample reasons to deny Chevron deference here. The authorization for classification rulings, and Customs’s practice in making them, present a case far removed not only from notice-and-comment process, but from any other circumstances reasonably suggesting that Congress ever thought of classification rulings as deserving the deference claimed for them here.

No matter which angle we choose for viewing the Customs ruling letter in this case, it fails to qualify under Chevron. On the face of the statute, to begin with, the terms of the congressional delegation give no indication that Congress meant to delegate authority to Customs to issue classification rulings with the force of law. We are not, of course, here making any global statement about Customs’s authority, for it is true that the general rulemaking power conferred on Customs authorizes some regulation with the force of law [...] 

It is difficult, in fact, to see in the agency practice itself any indication that Customs ever set out with a lawmaking pretense in mind when it undertook to make classifications like these. Customs does not generally engage in noticeand-comment practice when issuing them, and their treatment by the agency makes it clear that a letter’s binding character as a ruling stops short of third parties; Customs has regarded a classification as conclusive only as between itself and the importer to whom it was issued, and even then only until Customs has given advance notice of intended change. Other importers are in fact warned against assuming any right of detrimental reliance. 

Indeed, to claim that classifications have legal force is to ignore the reality that 46 different Customs offices issue 10,000 to 15,000 of them each year. Any suggestion that rulings intended to have the force of law are being churned out at a rate of 10,000 a year at an agency’s 46 scattered offices is simply self-refuting. Although the circumstances are less startling here, with a Headquarters letter in issue, none of the relevant statutes recognizes this category of rulings as separate or different from others; there is thus no indication that a more potent delegation might have been understood as going to Headquarters even when Headquarters provides developed reasoning, as it did in this instance [...]

In sum, classification rulings are best treated like “interpretations contained in policy statements, agency manuals, and enforcement guidelines.” Christensen, 529 U. S., at 587. They are beyond the Chevron pale.

To agree with the Court of Appeals that Customs ruling letters do not fall within Chevron is not, however, to place them outside the pale of any deference whatever. Chevron did nothing to eliminate Skidmore’s holding that an agency’s interpretation may merit some deference whatever its form, given the “specialized experience and broader investigations and information” available to the agency, and given the value of uniformity in its administrative and judicial understandings of what a national law requires. 

There is room at least to raise a Skidmore claim here, where the regulatory scheme is highly detailed, and Customs can bring the benefit of specialized experience to bear on the subtle questions in this case: whether the daily planner with room for brief daily entries falls under “diaries,” when diaries are grouped with “notebooks and address books, bound; memorandum pads, letter pads and similar articles,” and whether a planner with a ring binding should qualify as “bound,” when a binding may be typified by a book, but also may have “reinforcements or fittings of metal, plastics, etc.” A classification ruling in this situation may therefore at least seek a respect proportional to its “power to persuade,” Skidmore, supra; see also Christensen. Such a ruling may surely claim the merit of its writer’s thoroughness, logic, and expertness, its fit with prior interpretations, and any other sources of weight  [...]

Since the Skidmore assessment called for here ought to be made in the first instance by the Court of Appeals for the Federal Circuit or the CIT, we go no further than to vacate the judgment and remand the case for further proceedings consistent with this opinion.

B.D. PARKER, Circuit Judge:

Title IV of the HEA directs the Secretary of the Department of Education to implement various federal student financial aid programs. The Perkins Loan Program is one such program, designed to assist institutions of higher education in financing low-interest loans to financially needy students.

Under the program, the DOE provides federal monies to participating institutions. …In other words, Perkins Loans are "campus-based": The schools independently determine eligibility, advance funds, collect payments and make decisions concerning loan forgiveness. …

In 1985, … Congress amended the statute to encourage graduates to work in various areas of public service, such as teaching and the Peace Corps. This encouragement took the form of partial or total Perkins Loan cancellation. …  The section of the statute pivotal to this appeal … provides that "loans shall be canceled . . . for service": "(I) as a full-time employee of a public or private nonprofit child or family service agency who is providing, or supervising the provision of, services to high-risk children who are from low-income communities and the families of such children." [For loan cancellation, the] borrower applies to the lending school, which, rather than the DOE, bears the responsibility for determining the applicant's eligibility for loan cancellation.

[Since 1992 DOE has given] guidance about eligibility for loan cancellation for child or family service. The extent to which we are required to defer to these efforts is the critical issue on this appeal.

In 1995, the DOE enacted a regulation purportedly implementing the child or family service cancellation provision. In doing so, the DOE did not add institutional gloss or agency wisdom but rather incorporated verbatim the statute, into its own regulation. …

In addition … each year the Federal Student Aid Office of the DOE issues a Student Financial Aid Handbook to participating institutions to assist them in responding to loan cancellation requests. The Handbooks introduced a new qualification … not found in the statute, requiring that the services be extended "only" to high-risk children …

Along with these sources of assistance, participating institutions may obtain guidance on loan cancellation requests by contacting the Policy Development Division of the DOE's Office of Post-secondary Education. …

De La Mota

De La Mota applied for loan cancellation in 2000 through three academic institutions: City University of New York ("CUNY") and Manhattanville College where she did her undergraduate work, and New York Law School ("NYLS"). In the ACS Child Support Litigation Unit, she litigates paternity actions and prosecutes child support cases. For two years her loans were forgiven, but in the third year, one school, NYLS, rejected her application and demanded back payment for the previous two years. CUNY and Manhattanville continued to forgive her loans.

The DOE advised NYLS not to cancel her loans, because the services she provided to children were neither "direct" nor "only to high-risk children." The word "only" originated in the Handbooks, not the … statute or … regulation. The requirement of "direct" services  … first appeared in an April 12, 2001 informal e-mail from a DOE Program Specialist to NYLS: "The borrower must be providing services directly to the high-risk children. In this case, the borrower is providing services to the City of New York as an attorney, she is not providing services directly to high-risk children." …Upon advice from the Program Specialist that De La Mota's services and supervision of the provision of services were not "directly" nor "only" targeted to low-income, high-risk children, NYLS rejected her application, informing her that the DOE had determined that her position at ACS did not qualify for loan cancellation, reversed her previous cancellations, and directed her to make back payments. …

I.

As an initial matter, we note that the Appellants appear comfortably to meet the statute's textual qualifications for loan forgiveness. They are "full-time employee[s] of a public . . . nonprofit child or family service agency" who are "providing . . . services to high-risk children who are from low-income communities and the families of such children." … Congress thought cancellation subsidies were necessary because, traditionally, child and family service work tended to be low-paying, and the subsidies would encourage qualified graduates to enter the field.

II.

Notwithstanding the text of the statute and its intended purposes,… the DOE inserted "only" as a qualification in the 1996-1997 and 2001-2002 Handbooks…. On the basis of this qualification, the DOE recommended denial of loan forgiveness if an applicant performed any service, however sporadic or minimal, for anyone not at high-risk and not from a low-income community. Later, a DOE Program Specialist concluded and informally opined that loan cancellation was available only to applicants who provided services "directly" and "exclusively" to high-risk children and their families. This requirement effectively disqualified any public interest attorney litigating on behalf of poor, high-risk children; according to the DOE, an ACS attorney's client is New York City and an attorney provides services directly only to her client. … Significantly, the DOE did not use its regulation, to interpret the statute, as the regulation simply repeated the text of the statute verbatim, without the refinement or explication suggestive of any perceived ambiguity in the meaning of "providing services." …

Chevron deference is clearly inapplicable to the DOE's interpretation . . . of the term "providing services" to permit them to be provided "only," "directly" and "exclusively" to highrisk children. Under Chevron, 467 U.S. at 865, courts accord deference to an interpretation of a statute adopted by the agency that has been charged by Congress with responsibility for administering the provision. This deference, according to the DOE, is required here because of the "presumption that Congress, when it left ambiguity in a statute meant for implementation by an agency, understood that the ambiguity would be resolved, first and foremost, by the agency, and desired the agency (rather than the courts) to possess whatever degree of discretion the ambiguity allows." Smiley v. Citibank (South Dakota), N.A., 517 U.S. 735, 740-41 (1996). These are necessary but hardly sufficient requirements for Chevron deference.

The Supreme Court gave further attention to this issue in United States v. Mead Corp., 533 U.S. 218 (2001). It observed that an agency lacks a "lawmaking pretense in mind" when it makes pronouncements not binding on third parties and not in a notice-and-comment fashion. Mead's framework has been understood to distinguish "between formal and informal procedures" and "between generality and particularity in administrative decision making," where the former in each category warrant Chevron deference, but the latter do not. Here the "directly" and "exclusively" qualifications did not emerge from any formal rule-making procedures. As far as we can determine, they are ad hoc, previously unwritten rules, supplied by a DOE staff employee when determining appellants' eligibility. The requirement of "only" serving high-risk children, which first surfaced in the Handbook, fares no better, as its provenance is equally informal. Christensen v. Harris County, 529 U.S. 576, 587 (2000), made clear that "interpretations contained in policy statements, agency manuals and enforcement guidelines, all of which lack the force of law — do not warrant Chevron style deference."

Appellants contend that although a Skidmore analysis is appropriate, we ought not defer to the DOE's various informal or ad hoc interpretations. The DOE urges deference under Skidmore, because when denying loan cancellation, it contends that it "bring[s] the benefit of specialized experience to bear." We disagree. The DOE's critical narrowing requirements — "exclusively" and "directly" — hardly rise to the level of an agency interpretation since, as we have seen, they were pronounced initially by a DOE staff member. By contrast, the official narrowing of the statute with the addition of the qualification "only" originated in two DOE Handbooks; therefore the weight we accord the DOE's interpretation of this statute is determined through Skidmore analysis.

III.

Skidmore respect … depends on [the interpretation’s] "thoroughness," "validity," "consistency," and "power to persuade.".

A. Thoroughness

The DOE's 1996-1997 and 2001-2002 Handbooks interpret the eligibility requirement as applicable to those "providing services only to high risk children." In application, the DOE Program Specialist did not rely solely on "only," but grafted on "exclusively" and "directly." These latter requirements are neither synonymous with "only" nor equivalent in their narrowing force. For us to find "thoroughness evident in its consideration," the DOE would have had, at minimum, to adhere in practice to its own Handbook language and meaning..

Furthermore, thoroughness is impossible for an agency staff member to demonstrate when the staff member does not report to the Secretary, bears no law-making authority, and is unconstrained by political accountability. Thorough consideration requires a macro perspective that a staff member, acting alone, lacks. … We have shown deference to the opinions of agency officials who, though not an agency secretary or commissioner, hold substantial responsibility. … The more critical point here, of course, is the immateriality of one staff member's interpretation, when Congress expressly delegated rule-making authority to the Secretary. … The DOE has not convinced us that its Program Specialist exhibits either this authority or expertise.

Moreover, it is insufficient for the DOE as an agency merely to adopt the staff member's interpretation after the fact, or during the course of subsequent litigation. "[A] position adopted in the course of litigation lacks the indicia of expertise, regularity, rigorous consideration, and public scrutiny that justify Chevron deference." …Such endorsements also lack the thoroughness required for Skidmore respect.

B. Validity

The "validity" element of Skidmore analysis draws our attention to whether an agency pronouncement is well-reasoned, substantiated, and logical. …

The DOE contends that, in practice, it needs to narrow § 1087ee(a)(2)(I) in order to distinguish, for instance, between the support staff (administrative assistants, janitors, fundraisers) at a child or family service agency and those properly deserving loan cancellation. However, the addition of "only" does not help to reach this end. …

The "directly" requirement, when interpreted to exclude public interest attorneys because the government is their "client," further erodes § 1087ee(a)(2)(I). All qualifying "full-time employee[s] of a public or private nonprofit child or family service agency" serve multiple beneficiaries — the children at risk, the community, the government, the employing agency, the donors and trustees of the employing agency, to name the most obvious. The mere fact that a public interest attorney represents a client does not necessarily make that client the exclusive beneficiary of their work.

The DOE also requires provision of services "exclusively" to high-risk and low-income children. … The statute does not require that the children be low-income, but simply that they be from low-income communities. Congress has defined "low-income communities" for this purpose as "communities in which there is a high concentration of children eligible to be counted under title I of the Elementary and Secondary Education Act of 1965."  In other words, the low-income requirement is by definition fluid, guided by concentration rather than fixed by rules. …

C. Power to Persuade

Perhaps most importantly, the DOE interpreted the statute in an advisory capacity. The DOE argued in its brief: "Plaintiffs, however, cannot impute to DOE the positions taken by CUNY Law School or Tulane University. . . . [A]s DOE has maintained throughout this litigation, it is the individual educational institutions that render decisions upon loan cancellation, not DOE." In proceedings below, the DOE conceded that it "does not make the determination of whether a borrower qualifies for a cancellation but upon request provides guidance to institutions that make the decisions." (emphasis added). We are especially disinclined to defer to an agency when it does not purport to speak authoritatively. In the past we have declined to show an agency deference when "there is no indication in the record of the process through which [the agency] arrived at its interpretation" and the agency itself "labels its interpretation as `tentative'" because "we cannot say with confidence that [the agency's] interpretation came about as the result of a reasoned process."

In sum, the DOE's interpretation — the addition of "only" in the Handbooks, coupled with the use of "directly" and "exclusively" in an advisory role — lacks the power to persuade, and the DOE fails to show thoroughness in its consideration or validity in its reasoning. Thus, even assuming arguendo that the agency offered consistent guidance, we are not bound to defer to its construction of the statute.

* * *

Limits on Chevron Deference: Major Questions Doctrine

In the last few decades, the Supreme Court has placed another limitation on the Chevron Doctrine’s scope. The “major questions doctrine” holds that courts should not defer to agency statutory interpretations that concern questions of “vast economic or political significance.” The Supreme Court justifies this limitation with the non-delegation doctrine. According to the Supreme Court, courts are supposed to interpret “major” legal questions, not administrative bureaucrats. 

Some legal scholars argue that this reading of the non-delegation doctrine mistakenly takes political decisions out of the hands of agency officials appointed by elected leaders and puts the decisions into the hands of courts that lack political accountability. They argue that the major questions doctrine lets judges view statutes de novo, without deferring to agency interpretations, contradicting the Chevron principles. The major questions doctrine runs counter to the idea that important political decisions should be resolved by Congress. 

King v. Burwell is one of a constellation of cases where the Supreme Court relies on the major questions doctrine to override Chevron deference and assert its own statutory interpretation rather than respecting the agency’s interpretation. King v. Burwell, is a case about the Affordable Care Act (“ACA”), one of the most politically controversial laws (if not the most controversial) passed in the Obama era. The Supreme Court did not defer to the Internal Revenue Service (“IRS”) interpretation of the ACA when it decided whether insurance purchased on a Federal or State Exchange qualified for tax subsidies.

King v. Burwell

135 S. Ct. 475 (2015)

CHIEF JUSTICE ROBERTS, delivered the opinion of the Court.

The Patient Protection and Affordable Care Act adopts a series of interlocking reforms designed to expand coverage in the individual health insurance market. First, the Act bars insurers from taking a person’s health into account when deciding whether to sell health insurance or how much to charge. Second, the Act generally requires each person to maintain insurance coverage or make a payment to the Internal Revenue Service. And third, the Act gives tax credits to certain people to make insurance more affordable.

In addition to those reforms, the Act requires the creation of an “Exchange” in each State—basically, a marketplace that allows people to compare and purchase insurance plans. The Act gives each State the opportunity to establish its own Exchange, but provides that the Federal Government will establish the Exchange if the State does not.

This case is about whether the Act’s interlocking reforms apply equally in each State no matter who establishes the State’s Exchange. Specifically, the question presented is whether the Act’s tax credits are available in States that have a Federal Exchange.

I

The Patient Protection and Affordable Care Act grew out of a long history of failed health insurance reform. In the 1990s, several States began experimenting with ways to expand people’s access to coverage [One common State approach was to bar insurers from denying coverage or increasing premiums to people because of pre-existing health conditions.] Those requirements were designed to ensure that anyone who wanted to buy health insurance could do so.

The [requirements] had an unintended consequence: They encouraged people to wait until they got sick to buy insurance. Why buy insurance coverage when you are healthy, if you can buy the same coverage for the same price when you become ill? This consequence—known as “adverse selection”—led to a second: Insurers were forced to increase premiums to account for the fact that, more and more, it was the sick rather than the healthy who were buying insurance. And that consequence fed back into the first: As the cost of insurance rose, even more people waited until they became ill to buy it.

This led to an economic “death spiral.” As premiums rose higher and higher, and the number of people buying insurance sank lower and lower, insurers began to leave the market entirely. As a result, the number of people without insurance increased dramatically [...]

In 1996, Massachusetts adopted the [protections for people with pre-existing conditions]. But in 2006, Massachusetts added two more reforms: The Commonwealth required individuals to buy insurance or pay a penalty, and it gave tax credits to certain individuals to ensure that they could afford the insurance they were required to buy. The combination of these three reforms—insurance market regulations, a coverage mandate, and tax credits—reduced the uninsured rate in Massachusetts to 2.6 percent, by far the lowest in the Nation.

The Affordable Care Act adopts a version of the three key reforms that made the Massachusetts system successful. First, the Act adopts the guaranteed issue and community rating requirements. The Act provides that “each health insurance issuer that offers health insurance coverage in the individual . . . market in a State must accept every . . . individual in the State that applies for such coverage.” The Act also bars insurers from charging higher premiums on the basis of a person’s health. 

Second, the Act generally requires individuals to maintain health insurance coverage or make a payment to the IRS. Congress recognized that, without an incentive, “many individuals would wait to purchase health insurance until they needed care.” So Congress adopted a coverage requirement to “minimize this adverse selection and broaden the health insurance risk pool to include healthy individuals, which will lower health insurance premiums.” In Congress’s view, that coverage requirement was “essential to creating effective health insurance markets.” Congress also provided an exemption from the coverage requirement for anyone who has to spend more than eight percent of his income on health insurance [...] 

Third, the Act seeks to make insurance more affordable by giving refundable tax credits to individuals with household incomes between 100 percent and 400 percent of the federal poverty line. Individuals who meet the Act’s requirements may purchase insurance with the tax credits, which are provided in advance directly to the individual's insurer. 

These three reforms are closely intertwined. As noted, Congress found that the guaranteed issue and community rating requirements would not work without the coverage requirement. And the coverage requirement would not work without the tax credits. The reason is that, without the tax credits, the cost of buying insurance would exceed eight percent of income for a large number of individuals, which would exempt them from the coverage requirement [...] 

In addition to those three reforms, the Act requires the creation of an “Exchange” in each State where people can shop for insurance, usually online. An Exchange may be created in one of two ways. First, the Act provides that “[e]ach State shall . . . establish an American Health Benefit Exchange . . . for the State.” Second, if a State nonetheless chooses not to establish its own Exchange, the Act provides that the Secretary of Health and Human Services “shall . . . establish and operate such Exchange within the State.”

The issue in this case is whether the Act’s tax credits are available in States that have a Federal Exchange rather than a State Exchange. The Act initially provides that tax credits “shall be allowed” for any “applicable taxpayer.” 26 U.S.C. §36B(a). The Act then provides that the amount of the tax credit depends in part on whether the taxpayer has enrolled in an insurance plan through “an Exchange established by the State under section 1311 of the Patient Protection and Affordable Care Act." 26 U.S.C. §§36B(b)-(c) (emphasis added).

The IRS addressed the availability of tax credits by promulgating a rule that made them available on both State and Federal Exchanges. As relevant here, the IRS Rule provides that a taxpayer is eligible for a tax credit if he enrolled in an insurance plan through “an Exchange,” which is defined as “an Exchange serving the individual market . . . regardless of whether the Exchange is established and operated by a State . . . or by HHS.” At this point, 16 States and the District of Columbia have established their own Exchanges; the other 34 States have elected to have HHS do so.

Petitioners are four individuals who live in Virginia, which has a Federal Exchange. They do not wish to purchase health insurance. In their view, Virginia’s Exchange does not qualify as “an Exchange established by the State,” so they should not receive any tax credits. That would make the cost of buying insurance more than eight percent of their income, which would exempt them from the Act’s coverage requirement. 

Under the IRS Rule, however, Virginia’s Exchange would qualify as “an Exchange established by the State,” so petitioners would receive tax credits. That would make the cost of buying insurance less than eight percent of petitioners’ income, which would subject them to the Act’s coverage requirement. The IRS Rule therefore requires petitioners to either buy health insurance they do not want, or make a payment to the IRS.

Petitioners challenged the IRS Rule in Federal District Court. The District Court dismissed the suit, holding that the Act unambiguously made tax credits available to individuals enrolled through a Federal Exchange. The Court of Appeals for the Fourth Circuit affirmed. The Fourth Circuit viewed the Act as “ambiguous and subject to at least two different interpretations.” The court therefore deferred to the IRS's interpretation under Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U. S. 837 (1984). 

II

[...] When analyzing an agency’s interpretation of a statute, we often apply the two-step framework announced in Chevron, 467 U. S. 837. Under that framework, we ask whether the statute is ambiguous and, if so, whether the agency’s interpretation is reasonable. This approach “is premised on the theory that a statute’s ambiguity constitutes an implicit delegation from Congress to the agency to fill in the statutory gaps.” FDA v. Brown & Williamson Tobacco Corp., 529 U. S. 120, 159 (2000). “In extraordinary cases, however, there may be reason to hesitate before concluding that Congress has intended such an implicit delegation.”

This is one of those cases. The tax credits are among the Act’s key reforms, involving billions of dollars in spending each year and affecting the price of health insurance for millions of people. Whether those credits are available on Federal Exchanges is thus a question of deep “economic and political significance” that is central to this statutory scheme; had Congress wished to assign that question to an agency, it surely would have done so expressly. It is especially unlikely that Congress would have delegated this decision to the IRS, which has no expertise in crafting health insurance policy of this sort. This is not a case for the IRS.

It is instead our task to determine the correct reading of [ACA Section 36B]

[...] We have held that Congress “does not alter the fundamental details of a regulatory scheme in vague terms or ancillary provisions.” Whitman v. American Trucking Assns., Inc., 531 U. S. 457, 468 (2001). But in petitioners’ view, Congress made the viability of the entire Affordable Care Act turn on the ultimate ancillary provision: a sub-sub-sub section of the Tax Code. We doubt that is what Congress meant to do. Had Congress meant to limit tax credits to State Exchanges, it likely would have done so in the definition of “applicable taxpayer” or in some other prominent manner. It would not have used such a winding path of connect-the-dots provisions about the amount of the credit.

Petitioners’ arguments about the plain meaning of Section 36B are strong. But while the meaning of the phrase "an Exchange established by the State” may seem plain “when viewed in isolation,” such a reading turns out to be “untenable in light of [the statute] as a whole.” In this instance, the context and structure of the Act compel us to depart from what would otherwise be the most natural reading of the pertinent statutory phrase.

Reliance on context and structure in statutory interpretation is a “subtle business, calling for great wariness lest what professes to be mere rendering becomes creation and attempted interpretation of legislation becomes legislation itself.” For the reasons we have given, however, such reliance is appropriate in this case, and leads us to conclude that Section 36B allows tax credits for insurance purchased on any Exchange created under the Act. Those credits are necessary for the Federal Exchanges to function like their State Exchange counterparts, and to avoid the type of calamitous result that Congress plainly meant to avoid.

In a democracy, the power to make the law rests with those chosen by the people. Our role is more confined—“to say what the law is.” Marbury v. Madison, 1 Cranch 137, 177 (1803). That is easier in some cases than in others. But in every case we must respect the role of the Legislature, and take care not to undo what it has done. A fair reading of legislation demands a fair understanding of the legislative plan.

Congress passed the Affordable Care Act to improve health insurance markets, not to destroy them. If at all possible, we must interpret the Act in a way that is consistent with the former, and avoids the latter. Section 36B can fairly be read consistent with what we see as Congress’s plan, and that is the reading we adopt.

The judgment of the United States Court of Appeals for the Fourth Circuit is Affirmed.

We know now that courts apply a very deferential standard of review when they evaluate agencies’ interpretations of statutes. Now we will learn about the standards of review courts use to evaluate all of the other decisions agencies make. That standards of review are established in the APA.

Under APA Section 706(2)(E)

The reviewing court shall—

(E) unsupported by substantial evidence in a case subject to sections 556and 557 of this title or otherwise reviewed on the record of an agency hearing provided by statute[.]

What is The Substantial Evidence Standard?

The APA sets a different standard of review when courts are reviewing formal rulemaking or adjudication procedures. Section 706(2)(E) says that reviewing courts can hold unlawful agency action, findings, and conclusions that are "unsupported by substantial evidence in a case subject to sections 556 and 557 of this title or otherwise reviewed on the record of an agency hearing provided by statute..." This "substantial evidence" standard assumes that there is a detailed record of agency decisions and rationales that was created in formal rulemaking or adjudication procedures.

In practice, courts have applied the 706(2)(E) substantial evidence standard similarly to the arbitrary and capricious standard. While the Supreme Court has said that the “substantial evidence” standard is even more deferential than the arbitrary and capricious standard of review, the standards are so similar and so deferential that they are functionally the same. In assessing both formal and informal adjudication, reviewing courts generally avoid second-guessing or revisiting agency fact-finding in formal adjudication, only doing so in rare cases when the agency decision is “clearly erroneous” or when there is definite and firm evidence that the agency acted in error. Dickinson v. Zurko, 527 U.S. 150 (1999) (describing the “substantial evidence” standard of review).

The APA sets a different standard of review when courts are reviewing formal rulemaking or adjudication procedures. Section 706(2)(E) says that reviewing courts can hold unlawful agency action, findings, and conclusions that are "unsupported by substantial evidence in a case subject to sections 556 and 557 of this title or otherwise reviewed on the record of an agency hearing provided by statute..." This "substantial evidence" standard assumes that there is a detailed record of agency decisions and rationales that was created in formal rulemaking or adjudication procedures.

In practice, courts have applied the 706(2)(E) substantial evidence standard similarly to the arbitrary and capricious standard. While the Supreme Court has said that the “substantial evidence” standard is even more deferential than the arbitrary and capricious standard of review, the standards are so similar and so deferential that they are functionally the same. In assessing both formal and informal adjudication, reviewing courts generally avoid second-guessing or revisiting agency fact-finding in formal adjudication, only doing so in rare cases when the agency decision is “clearly erroneous” or when there is definite and firm evidence that the agency acted in error. Dickinson v. Zurko, 527 U.S. 150 (1999) (describing the “substantial evidence” standard of review).

What is Arbitrary and Capricious Review? 

The arbitrary and capricious standard of review is most commonly used to assess the factual basis of agency informal, notice-and-comment, rulemaking, but it applies to all agency actions according to the APA. 

Under the “arbitrary and capricious” standard of review, courts look to see whether agencies have taken a “hard look” at the underlying questions of policy and fact upon which their decisions are based. According to the hard look doctrine, agencies have to justify their decisions with adequate reasoning to pass muster under the arbitrary and capricious standard. In practice, the arbitrary and capricious standard and review is very similar, if not identical, to step two of the Chevron test. 

Citizens to Preserve Overton Park v. Volpe

401 U.S. 402 (1971)

Opinion of the Court by MR. JUSTICE MARSHALL, announced by MR. JUSTICE STEWART.

The growing public concern about the quality of our natural environment has prompted Congress in recent years to enact legislation designed to curb the accelerating destruction of our country’s natural beauty. We are concerned in this case with § 4(f) of the Department of Transportation Act of 1966 and § 18(a) of the Federal-Aid Highway Act of 1968. These statutes prohibit the Secretary of Transportation from authorizing the use of federal funds to finance the construction of highways through public parks if a “feasible and prudent” alternative route exists. If no such route is available, the statutes allow him to approve construction through parks only if there has been “all possible planning to minimize harm” to the park.

Petitioners, private citizens as well as local and national conservation organizations, contend that the Secretary has violated these statutes by authorizing the expenditure of federal funds for the construction of a six-lane interstate highway through a public park in Memphis, Tennessee. Their claim was rejected by the District Court, which granted the Secretary’s motion for summary judgment, and the Court of Appeals for the Sixth Circuit affirmed. After oral argument, this Court granted a stay that halted construction and, treating the application for the stay as a petition for certiorari, granted review. We now reverse the judgment below and remand for further proceedings in the District Court.

Overton Park is a 342-acre city park located near the center of Memphis. The park contains a zoo, a nine-hole municipal golf course, an outdoor theater, nature trails, a bridle path, an art academy, picnic areas, and 170 acres of forest. The proposed highway, which is to be a six-lane, high-speed, expressway, will sever the zoo from the rest of the park [...] 

In April 1968, the Secretary announced that he concurred in the judgment of local officials that I-40 should be built through the park. And in September 1969 the State acquired the right-of-way inside Overton Park from the city. Final approval for the project—the route as well as the design— was not announced until November 1969 [...] Neither announcement approving the route and design of I-40 was accompanied by a statement of the Secretary’s factual findings. He did not indicate why he believed there were no feasible and prudent alternative routes or why design changes could not be made to reduce the harm to the park.

Petitioners contend that the Secretary’s action is invalid without such formal findings and that the Secretary did not make an independent determination but merely relied on the judgment of the Memphis City Council. They also contend that it would be “feasible and prudent” to route I-40 around Overton Park either to the north or to the south. And they argue that if these alternative routes are not “feasible and prudent,” the present plan does not include “all possible” methods for reducing harm to the park. Petitioners claim that I-40 could be built under the park by using either of two possible tunneling methods [...]

[The standard of review in] § 706 of the Administrative Procedure Act provides that a “reviewing court shall . . . hold unlawful and set aside agency action, findings, and conclusions found” not to meet six separate standards. In all cases agency action must be set aside if the action was “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law” or if the action failed to meet statutory, procedural, or constitutional requirements. In certain narrow, specifically limited situations, the agency action is to be set aside if the action was not supported by “substantial evidence.” And in other equally narrow circumstances the reviewing court is to engage in a de novo review of the action and set it aside if it was “unwarranted by the facts.”

Petitioners argue that the Secretary's approval of the construction of I-40 through Overton Park is subject to one or the other of these latter two standards of limited applicability. First, they contend that the "substantial evidence" standard of § 706(2)(E) must be applied. In the alternative, they claim that § 706(2)(F) applies and that there must be a de novo review to determine if the Secretary’s action was “unwarranted by the facts.” Neither of these standards is, however, applicable.

Review under the substantial-evidence test is authorized only when the agency action is taken pursuant to a rulemaking provision of the Administrative Procedure Act itself, 5 U.S.C. § 553, or when the agency action is based on a public adjudicatory hearing. See 5 U.S.C. §§ 556, 557. The Secretary’s decision to allow the expenditure of federal funds to build I-40 through Overton Park was plainly not an exercise of a rulemaking function. And the only hearing that is required by either the Administrative Procedure Act or the statutes regulating the distribution of federal funds for highway construction is a public hearing conducted by local officials for the purpose of informing the community about the proposed project and eliciting community views on the design and route. The hearing is nonadjudicatory, quasi-legislative in nature. It is not designed to produce a record that is to be the basis of agency action—the basic requirement for substantial-evidence review. 

Petitioners’ alternative argument also fails. De novo review of whether the Secretary’s decision was “unwarranted by the facts” is authorized by § 706(2)(F) in only two circumstances. First, such de novo review is authorized when the action is adjudicatory in nature and the agency fact finding procedures are inadequate. And, there may be independent judicial fact finding when issues that were not before the agency are raised in a proceeding to enforce nonadjudicatory agency action. Neither situation exists here.

Even though there is no de novo review in this case and the Secretary’s approval of the route of I-40 does not ultimately have to meet the substantial-evidence test, the generally applicable standards of § 706 require the reviewing court to engage in a substantial inquiry [...] 

The court is first required to decide whether the Secretary acted within the scope of his authority. This determination naturally begins with a delineation of the scope of the Secretary’s authority and discretion [...] As has been shown, Congress has specified only a small range of choices that the Secretary can make. Also involved in this initial inquiry is a determination of whether on the facts the Secretary’s decision can reasonably be said to be within that range. The reviewing court must consider whether the Secretary properly construed his authority to approve the use of parkland as limited to situations where there are no feasible alternative routes or where feasible alternative routes involve uniquely difficult problems. And the reviewing court must be able to find that the Secretary could have reasonably believed that in this case there are no feasible alternatives or that alternatives do involve unique problems.

Scrutiny of the facts does not end, however, with the determination that the Secretary has acted within the scope of his statutory authority. Section 706(2)(A) requires a finding that the actual choice made was not “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” To make this finding the court must consider whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment. Although this inquiry into the facts is to be searching and careful, the ultimate standard of review is a narrow one. The court is not empowered to substitute its judgment for that of the agency.

The final inquiry is whether the Secretary’s action followed the necessary procedural requirements. Here the only procedural error alleged is the failure of the Secretary to make formal findings and state his reason for allowing the highway to be built through the park.

Undoubtedly, review of the Secretary’s action is hampered by his failure to make such findings, but the absence of formal findings does not necessarily require that the case be remanded to the Secretary. Neither the Department of Transportation Act nor the Federal-Aid Highway Act requires such formal findings. Moreover, the Administrative Procedure Act requirements that there be formal findings in certain rulemaking and adjudicatory proceedings do not apply to the Secretary’s action here [...] 

Moreover, there is an administrative record that allows the full, prompt review of the Secretary’s action that is sought without additional delay which would result from having a remand to the Secretary.

That administrative record is not, however, before us. The lower courts based their review on the litigation affidavits that were presented. These affidavits were merely “post hoc” rationalizations which have traditionally been found to be an inadequate basis for review. And they clearly do not constitute the “whole record” compiled by the agency: the basis for review required by § 706 of the Administrative Procedure Act. 

Thus it is necessary to remand this case to the District Court for plenary review of the Secretary’s decision. That review is to be based on the full administrative record that was before the Secretary at the time he made his decision. But since the bare record may not disclose the factors that were considered or the Secretary's construction of the evidence it may be necessary for the District Court to require some explanation in order to determine if the Secretary acted within the scope of his authority and if the Secretary's action was justifiable under the applicable standard.

The court may require the administrative officials who participated in the decision to give testimony explaining their action. Of course, such inquiry into the mental processes of administrative decisionmakers is usually to be avoided. And where there are administrative findings that were made at the same time as the decision, there must be a strong showing of bad faith or improper behavior before such inquiry may be made. But here there are no such formal findings and it may be that the only way there can be effective judicial review is by examining the decisionmakers themselves. 

The District Court is not, however, required to make such an inquiry. It may be that the Secretary can prepare formal findings [...] that will provide an adequate explanation for his action. Such an explanation will, to some extent, be a “post hoc rationalization” and thus must be viewed critically. If the District Court decides that additional explanation is necessary, that court should consider which method will prove the most expeditious so that full review may be had as soon as possible.

Reversed and remanded.

Motor Vehicle Manufacturers Association v. State Farm Mutual Automobile Insurance

463 U.S. 29 (1983)

JUSTICE WHITE delivered the opinion of the Court.

The development of the automobile gave Americans unprecedented freedom to travel, but exacted a high price for enhanced mobility. Since 1929, motor vehicles have been the leading cause of accidental deaths and injuries in the United States. In 1982, 46,300 Americans died in motor vehicle accidents and hundreds of thousands more were maimed and injured. While a consensus exists that the current loss of life on our highways is unacceptably high, improving safety does not admit to easy solution. In 1966, Congress decided that at least part of the answer lies in improving the design and safety features of the vehicle itself. But much of the technology for building safer cars was undeveloped or untested.

Before changes in automobile design could be mandated, the effectiveness of these changes had to be studied, their costs examined, and public acceptance considered. This task called for considerable expertise and Congress responded by enacting the National Traffic and Motor Vehicle Safety Act of 1966. The Act, created for the purpose of “reduc[ing] traffic accidents and deaths and injuries to persons resulting from traffic accidents,” directs the Secretary of Transportation or his delegate to issue motor vehicle safety standards that “shall be practicable, shall meet the need for motor vehicle safety, and shall be stated in objective terms.”

In issuing these standards, the Secretary is directed to consider “relevant available motor vehicle safety data,” whether the proposed standard “is reasonable, practicable and appropriate” for the particular type of motor vehicle, and the “extent to which such standards will contribute to carrying out the purposes” of the Act. 

[...] We review today whether [the National Highway Traffic Safety Administration (“NHTSA”)] acted arbitrarily and capriciously in revoking the requirement in Motor Vehicle Safety Standard 208 that new motor vehicles produced after September 1982 be equipped with passive restraints [otherwise known as automatic seatbelts and airbags] to protect the safety of the occupants of the vehicle in the event of a collision. Briefly summarized, we hold that the agency failed to present an adequate basis and explanation for rescinding the passive restraint requirement and that the agency must either consider the matter further or adhere to or amend Standard 208 along lines which its analysis supports.

I

The regulation whose rescission is at issue bears a complex and convoluted history. Over the course of approximately 60 rulemaking notices, the requirement has been imposed, amended, rescinded, reimposed, and now rescinded again.

As originally issued by the Department of Transportation in 1967, Standard 208 simply required the installation of seatbelts in all automobiles. It soon became apparent that the level of seatbelt use was too low to reduce traffic injuries to an acceptable level. The Department therefore began consideration of “passive occupant restraint systems” — devices that do not depend for their effectiveness upon any action taken by the occupant except that necessary to operate the vehicle. Two types of automatic crash protection emerged: automatic seatbelts and airbags. [No rulemaking was successful for a decade, until President Carter appointed Brock Adams as the U.S. Secretary of Transportation.]

Within months of assuming office, Secretary Brock Adams [...] issued a new mandatory passive restraint regulation, known as Modified Standard 208. The Modified Standard mandated the phasing in of passive restraints beginning with large cars in model year 1982 and extending to all cars by model year 1984. The two principal systems that would satisfy the Standard were airbags and passive belts; the choice of which system to install was left to the manufacturers [...] 

In February 1981, however, Secretary of Transportation Andrew Lewis reopened the rulemaking due to changed economic circumstances and, in particular, the difficulties of the automobile industry. Two months later, the agency ordered a one-year delay in the application of the Standard to large cars, extending the deadline to September 1982, and at the same time, proposed the possible rescission of the entire Standard. After receiving written comments and holding public hearings, NHTSA issued a final rule (Notice 25) that rescinded the passive restraint requirement contained in Modified Standard 208.

II

In a statement explaining the rescission, NHTSA maintained that it was no longer able to find, as it had in 1977, that the automatic restraint requirement would produce significant safety benefits. This judgment reflected not a change of opinion on the effectiveness of the technology, but a change in plans by the automobile industry. In 1977, the agency had assumed that airbags would be installed in 60% of all new cars and automatic seatbelts in 40%. By 1981 it became apparent that automobile manufacturers planned to install the automatic seatbelts in approximately 99% of the new cars. For this reason, the lifesaving potential of airbags would not be realized.

Moreover, it now appeared that the overwhelming majority of passive belts planned to be installed by manufacturers could be detached easily and left that way permanently. Passive belts, once detached, then required “the same type of affirmative action that is the stumbling block to obtaining high usage levels of manual belts.” For this reason, the agency concluded that there was no longer a basis for reliably predicting that the Standard would lead to any significant increased usage of restraints at all [...]

State Farm Mutual Automobile Insurance Co. and the National Association of Independent Insurers filed petitions for review of NHTSA’s rescission of the passive restraint Standard. The United States Court of Appeals for the District of Columbia Circuit held that the agency’s rescission of the passive restraint requirement was arbitrary and capricious. [The D.C. Court of Appeals found] that the rescission of Standard 208 was arbitrary and capricious for three reasons. First, the court found insufficient as a basis for rescission NHTSA’s conclusion that it could not reliably predict an increase in belt usage under the Standard.

The court held that there was insufficient evidence in the record to sustain NHTSA’s position on this issue, and that, “only a well justified refusal to seek more evidence could render rescission non-arbitrary.” Second, a majority of the panel concluded that NHTSA inadequately considered the possibility of requiring manufacturers to install nondetachable rather than detachable passive belts. Third, the majority found that the agency acted arbitrarily and capriciously by failing to give any consideration whatever to requiring compliance with Modified Standard 208 by the installation of airbags [...]

III

Both the Act and the 1974 Amendments concerning occupant crash protection standards indicate that motor vehicle safety standards are to be promulgated under the informal rulemaking procedures of the Administrative Procedure Act. 5 U.S.C. § 553. The agency’s action in promulgating such standards therefore may be set aside if found to be “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. §706(2)(A). We believe that the rescission or modification of an occupant-protection standard is subject to the same test [...] 

Petitioner Motor Vehicle Manufacturers Association (MVMA) disagrees, contending that the rescission of an agency rule should be judged by the same standard a court would use to judge an agency's refusal to promulgate a rule in the first place — a standard petitioner believes considerably narrower than the traditional arbitrary-and-capricious test. We reject this view. The [Motor Vehicle Safety Act] expressly equates orders “revoking” and “establishing” safety standards; neither that Act nor the APA suggests that revocations are to be treated as refusals to promulgate standards [...]

The Department of Transportation accepts the applicability of the “arbitrary and capricious” standard. It argues that under this standard, a reviewing court may not set aside an agency rule that is rational, based on consideration of the relevant factors, and within the scope of the authority delegated to the agency by the statute. We do not disagree with this formulation. The scope of review under the “arbitrary and capricious” standard is narrow and a court is not to substitute its judgment for that of the agency.

Nevertheless, the agency must examine the relevant data and articulate a satisfactory explanation for its action including a “rational connection between the facts found and the choice made.” In reviewing that explanation, we must “consider whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment.” Normally, an agency rule would be arbitrary and capricious if the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.

The reviewing court should not attempt itself to make up for such deficiencies; we may not supply a reasoned basis for the agency’s action that the agency itself has not given. SEC v. Chenery Corp., 332 U. S. 194, 196 (1947). We will, however, “uphold a decision of less than ideal clarity if the agency’s path may reasonably be discerned.” [...]

V

The ultimate question before us is whether NHTSA’s rescission of the passive restraint requirement of Standard 208 was arbitrary and capricious. We conclude, as did the Court of Appeals, that it was. We also conclude [...] that further consideration of the issue by the agency is therefore required. We deal separately with the rescission as it applies to airbags and as it applies to seatbelts.

A

The first and most obvious reason for finding the rescission arbitrary and capricious is that NHTSA apparently gave no consideration whatever to modifying the Standard to require that airbag technology be utilized Standard 208 sought to achieve automatic crash protection by requiring automobile manufacturers to install either of two passive restraint devices: airbags or automatic seatbelts. There was no suggestion in the long rulemaking process that led to Standard 208 that if only one of these options were feasible, no passive restraint standard should be promulgated. Indeed, the agency’s original proposed standard contemplated the installation of inflatable restraints in all cars.

Automatic belts were added as a means of complying with the standard because they were believed to be as effective as airbags in achieving the goal of occupant crash protection. At that time, the passive belt approved by the agency could not be detached. Only later, at a manufacturer’s behest, did the agency approve of the detachability feature—and only after assurances that the feature would not compromise the safety benefits of the restraint. Although it was then foreseen that 60% of the new cars would contain airbags and 40% would have automatic seatbelts, the ratio between the two was not significant as long as the passive belt would also assure greater passenger safety.

The agency has now determined that the detachable automatic belts will not attain anticipated safety benefits because so many individuals will detach the mechanism. Even if this conclusion were acceptable in its entirety, standing alone it would not justify any more than an amendment of Standard 208 to disallow compliance by means of the one technology which will not provide effective passenger protection. It does not cast doubt on the need for a passive restraint standard or upon the efficacy of airbag technology.

In its most recent rule-making, the agency again acknowledged the life-saving potential of the airbag:

“The agency has no basis at this time for changing its earlier conclusions in 1976 and 1977 that basic airbag technology is sound and has been sufficiently demonstrated to be effective in those vehicles in current use. . . .” NHTSA Final Regulatory Impact Analysis XI-4 (Oct. 1981). 

Given the effectiveness ascribed to airbag technology by the agency, the mandate of the [Safety] Act to achieve traffic safety would suggest that the logical response to the faults of detachable seatbelts would be to require the installation of airbags. At the very least this alternative way of achieving the objectives of the Act should have been addressed and adequate reasons given for its abandonment. But the agency not only did not require compliance through airbags, it did not even consider the possibility in its 1981 rulemaking. Not one sentence of its rulemaking statement discusses the airbags-only option. 

[...]

What we said in Burlington Truck Lines, Inc. v. United States, 371 U. S., at 167, is apropos here:

“There are no findings and no analysis here to justify the choice made, no indication of the basis on which the [agency] exercised its expert discretion. We are not prepared to and the Administrative Procedure Act will not permit us to accept such . . . practice. . . . Expert discretion is the lifeblood of the administrative process, but ‘unless we make the requirements for administrative action strict and demanding, expertise, the strength of modern government, can become a monster which rules with no practical limits on its discretion.

We have frequently reiterated that an agency must cogently explain why it has exercised its discretion in a given manner, and we reaffirm this principle again today. The automobile industry has opted for the passive belt over the airbag, but surely it is not enough that the regulated industry has eschewed a given safety device. For nearly a decade, the automobile industry waged the regulatory equivalent of war against the airbag and lost -- the inflatable restraint was proved sufficiently effective. Now the automobile industry has decided to employ a seatbelt system which will not meet the safety objectives of Standard 208. This hardly constitutes cause to revoke the Standard itself. Indeed, the Act was necessary because the industry was not sufficiently responsive to safety concerns.

[...]

Petitioners also invoke our decision in Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, Inc.,435 U. S. 519 (1978), as though it were a talisman under which any agency decision is by definition unimpeachable. Specifically, it is submitted that to require an agency to consider an airbags-only alternative is, in essence, to dictate to the agency the procedures it is to follow. [Petitioners misread Vermont Yankee.] In Vermont Yankee, we held that a court may not impose additional procedural requirements upon an agency.

We do not require today any specific procedures which NHTSA must follow. Nor do we broadly require an agency to consider all policy alternatives in reaching decision. It is true that rulemaking “cannot be found wanting simply because the agency failed to include every alternative device and thought conceivable by the mind of man . . . regardless of how uncommon or unknown that alternative may have been . . . .”

But the airbag is more than a policy alternative to the passive restraint Standard; it is a technological alternative within the ambit of the existing Standard. We hold only that given the judgment made in 1977 that airbags are an effective and cost-beneficial life-saving technology, the mandatory passive restraint rule may not be abandoned without any consideration whatsoever of an airbags-only requirement.

B

Although the issue is closer, we also find that the agency was too quick to dismiss the safety benefits of automatic seatbelts. NHTSA’s critical finding was that, in light of the industry’s plans to install readily detachable passive belts, it could not reliably predict “even a 5 percentage point increase as the minimum level of expected usage increase.” The Court of Appeals rejected this finding because there is “not one iota” of evidence that Modified Standard 208 will fail to increase nationwide seatbelt use by at least 13 percentage points, the level of increased usage necessary for the Standard to justify its cost. Given the lack of probative evidence, the court held that “only a well justified refusal to seek more evidence could render rescission non-arbitrary.” 

Petitioners object to this conclusion. In their view, “substantial uncertainty” that a regulation will accomplish its intended purpose is sufficient reason, without more, to rescind a regulation. We agree with petitioners that just as an agency reasonably may decline to issue a safety standard if it is uncertain about its efficacy, an agency may also revoke a standard on the basis of serious uncertainties if supported by the record and reasonably explained. Rescission of the passive restraint requirement would not be arbitrary and capricious simply because there was no evidence in direct support of the agency’s conclusion. It is not infrequent that the available data do not settle a regulatory issue, and the agency must then exercise its judgment in moving from the facts and probabilities on the record to a policy conclusion.

Recognizing that policymaking in a complex society must account for uncertainty, however, does not imply that it is sufficient for an agency to merely recite the terms “substantial uncertainty” as a justification for its actions. As previously noted, the agency must explain the evidence which is available, and must offer a “rational connection between the facts found and the choice made.” Generally, one aspect of that explanation would be a justification for rescinding the regulation before engaging in a search for further evidence.

In these cases, the agency’s explanation for rescission of the passive restraint requirement is not sufficient to enable us to conclude that the rescission was the product of reasoned decisionmaking. To reach this conclusion, we do not upset the agency’s view of the facts, but we do appreciate the limitations of this record in supporting the agency’s decision. We start with the accepted ground that if used, seatbelts unquestionably would save many thousands of lives and would prevent tens of thousands of crippling injuries [...] We move next to the fact that there is no direct evidence in support of the agency’s finding that detachable automatic belts cannot be predicted to yield a substantial increase in usage.

The empirical evidence on the record, consisting of surveys of drivers of automobiles equipped with passive belts, reveals more than a doubling of the usage rate experienced with manual belts. Much of the agency’s rulemaking statement — and much of the controversy in these cases — centers on the conclusions that should be drawn from these studies. The agency maintained that the doubling of seatbelt usage in these studies could not be extrapolated to an across-the-board mandatory standard because the passive seatbelts were guarded by ignition interlocks and purchasers of the tested cars are somewhat atypical.

Respondents insist these studies demonstrate that Modified Standard 208 will substantially increase seatbelt usage. We believe that it is within the agency's discretion to pass upon the generalizability of these field studies. This is precisely the type of issue which rests within the expertise of NHTSA, and upon which a reviewing court must be most hesitant to intrude.

But accepting the agency’s view of the field tests on passive restraints indicates only that there is no reliable real-world experience that usage rates will substantially increase. To be sure, NHTSA opines that “it cannot reliably predict even a 5 percentage point increase as the minimum level of expected increased usage.” But this and other statements that passive belts will not yield substantial increases in seatbelt usage apparently take no account of the critical difference between detachable automatic belts and current manual belts. A detached passive belt does require an affirmative act to reconnect it, but — unlike a manual seatbelt — the passive belt, once reattached, will continue to function automatically unless again disconnected.

Thus, inertia — a factor which the agency’s own studies have found significant in explaining the current low usage rates for seatbelts — works in favor of, not against, use of the protective device. Since 20% to 50% of motorists currently wear seatbelts on some occasions, there would seem to be grounds to believe that seatbelt use by occasional users will be substantially increased by the detachable passive belts. Whether this is in fact the case is a matter for the agency to decide, but it must bring its expertise to bear on the question [...]

The agency also failed to articulate a basis for not requiring nondetachable belts under Standard 208. It is argued that the concern of the agency with the easy detachability of the currently favored design would be readily solved by a continuous passive belt, which allows the occupant to “spool out” the belt and create the necessary slack for easy extrication from the vehicle. The agency did not separately consider the continuous belt option, but treated it together with the ignition interlock device in a category it titled “Option of Adopting Use-Compelling Features.” The agency was concerned that use-compelling devices would “complicate the extrication of [an] occupant from his or her car.”

“[T]o require that passive belts contain use-compelling features,” the agency observed, “could be counterproductive [, given] . . . widespread, latent and irrational fear in many members of the public that they could be trapped by the seat belt after a crash.” In addition, based on the experience with the ignition interlock, the agency feared that use-compelling features might trigger adverse public reaction.

By failing to analyze the continuous seatbelts option in its own right, the agency has failed to offer the rational connection between facts and judgment required to pass muster under the arbitrary-and-capricious standard. We agree with the Court of Appeals that NHTSA did not suggest that the emergency release mechanisms used in nondetachable belts are any less effective for emergency egress than the buckle release system used in detachable belts.

In 1978, when General Motors obtained the agency’s approval to install a continuous passive belt, it assured the agency that nondetachable belts with spool releases were as safe as detachable belts with buckle releases. NHTSA was satisfied that this belt design assured easy extricability: “[t]he agency does not believe that the use of [such] release mechanisms will cause serious occupant egress problems . . . .” While the agency is entitled to change its view on the acceptability of continuous passive belts, it is obligated to explain its reasons for doing so [...]

JUSTICE REHNQUIST, with whom THE CHIEF JUSTICE, JUSTICE POWELL, and JUSTICE O'CONNOR join, concurring in part and dissenting in part.

I join Parts I, II, III, IV, and V-A of the Court's opinion. In particular, I agree that, since the airbag and continuous spool automatic seatbelt were explicitly approved in the Standard the agency was rescinding, the agency should explain why it declined to leave those requirements intact. In this case, the agency gave no explanation at all. Of course, if the agency can provide a rational explanation, it may adhere to its decision to rescind the entire Standard.

I do not believe, however, that NHTSA’s view of detachable automatic seatbelts was arbitrary and capricious. The agency adequately explained its decision to rescind the Standard insofar as it was satisfied by detachable belts [...]

The agency’s changed view of the standard seems to be related to the election of a new President of a different political party. It is readily apparent that the responsible members of one administration may consider public resistance and uncertainties to be more important than do their counterparts in a previous administration. A change in administration brought about by the people casting their votes is a perfectly reasonable basis for an executive agency’s reappraisal of the costs and benefits of its programs and regulations. As long as the agency remains within the bounds established by Congress, it is entitled to assess administrative records and evaluate priorities in light of the philosophy of the administration.

Department of Homeland Security v. Regents of the University of Calif., 
140 S.Ct. 1891(2020)

CHIEF JUSTICE ROBERTS delivered the opinion of the Court, except as to Part IV.

In the summer of 2012, the Department of Homeland Security (DHS) announced an immigration program known as Deferred Action for Childhood Arrivals, or DACA. That program allows certain unauthorized aliens who entered the United States as children to apply for a two-year forbearance of removal. Those granted such relief are also eligible for work authorization and various federal benefits. Some 700,000 aliens have availed themselves of this opportunity.

Five years later, the Attorney General advised DHS to rescind DACA, based on his conclusion that it was unlawful. The Department's Acting Secretary issued a memorandum terminating the program on that basis. The termination was challenged by affected individuals and third parties who alleged, among other things, that the Acting Secretary had violated the Administrative Procedure Act (APA) by failing to adequately address important factors bearing on her decision. For the reasons that follow, we conclude that the Acting Secretary did violate the APA, and that the rescission must be vacated.

I

A

In June 2012, the Secretary of Homeland Security issued a memorandum announcing an immigration relief program for "certain young people who were brought to this country as children." Known as DACA, the program applies to childhood arrivals who were under age 31 in 2012; have continuously resided here since 2007; are current students, have completed high school, or are honorably discharged veterans; have not been convicted of any serious crimes; and do not threaten national security or public safety. DHS concluded that individuals who meet these criteria warrant favorable treatment under the immigration laws because they "lacked the intent to violate the law," are "productive" contributors to our society, and "know only this country as home."

"[T]o prevent [these] low priority individuals from being removed from the United States," the DACA Memorandum instructs Immigration and Customs Enforcement to "exercise prosecutorial discretion[ ] on an individual basis . . . by deferring action for a period of two years, subject to renewal." In addition, it directs U. S. Citizenship and Immigration Services (USCIS) to "accept applications to determine whether these individuals qualify for work authorization during this period of deferred action," as permitted under regulations long predating DACA's creation, see 8 CFR § 274a.12(c)(14) (2012) (permitting work authorization for deferred action recipients who establish "economic necessity"); 46 Fed. Reg. 25080-25081 (1981) (similar). Pursuant to other regulations, deferred action recipients are considered "lawfully present" for purposes of, and therefore eligible to receive, Social Security and Medicare benefits. See 8 CFR § 1.3(a)(4)(vi); 42 CFR § 417.422(h) (2012).

In November 2014, two years after DACA was promulgated, DHS issued a memorandum announcing that it would expand DACA eligibility by removing the age cap, shifting the date-of-entry requirement from 2007 to 2010, and extending the deferred action and work authorization period to three years. In the same memorandum, DHS created a new, related program known as Deferred Action for Parents of Americans and Lawful Permanent Residents, or DAPA. That program would have authorized deferred action for up to 4.3 million parents whose children were U. S. citizens or lawful permanent residents. These parents were to enjoy the same forbearance, work eligibility, and other benefits as DACA recipients.

Before the DAPA Memorandum was implemented, 26 States, led by Texas, filed suit in the Southern District of Texas. The States contended that DAPA and the DACA expansion violated the APA's notice and comment requirement, the Immigration and Nationality Act (INA), and the Executive's duty under the Take Care Clause of the Constitution. The District Court found that the States were likely to succeed on the merits of at least one of their claims and entered a nationwide preliminary injunction barring implementation of both DAPA and the DACA expansion. See Texas v. United States, 86 F. Supp. 3d 591, 677-678 (2015).

A divided panel of the Court of Appeals for the Fifth Circuit affirmed the preliminary injunction. Texas v. United States, 809 F. 3d 134, 188 (2015). In opposing the injunction, the Government argued that the DAPA Memorandum reflected an unreviewable exercise of the Government's enforcement discretion. The Fifth Circuit majority disagreed. It reasoned that the deferred action described in the DAPA Memorandum was "much more than nonenforcement: It would affirmatively confer `lawful presence' and associated benefits on a class of unlawfully present aliens." From this, the majority concluded that the creation of the DAPA program was not an unreviewable action "committed to agency discretion by law." (quoting 5 U. S. C. § 701(a)(2)).

The majority then upheld the injunction on two grounds. It first concluded the States were likely to succeed on their procedural claim that the DAPA Memorandum was a substantive rule that was required to undergo notice and comment. It then held that the APA required DAPA to be set aside because the program was "manifestly contrary" to the INA, which "expressly and carefully provides legal designations allowing defined classes" to "receive the benefits" associated with "lawful presence" and to qualify for work authorization.

This Court affirmed the Fifth Circuit's judgment by an equally divided vote, which meant that no opinion was issued. United States v. Texas, 579 U. S. ___ (2016) (per curiam). For the next year, litigation over DAPA and the DACA expansion continued in the Southern District of Texas, while implementation of those policies remained enjoined.

Then, in June 2017, following a change in Presidential administrations, DHS rescinded the DAPA Memorandum. In explaining that decision, DHS cited the preliminary injunction and ongoing litigation in Texas, the fact that DAPA had never taken effect, and the new administration's immigration enforcement priorities.

Three months later, in September 2017, Attorney General Jefferson B. Sessions III sent a letter to Acting Secretary of Homeland Security Elaine C. Duke, "advis[ing]" that DHS "should rescind" DACA as well. Citing the Fifth Circuit's opinion and this Court's equally divided affirmance, the Attorney General concluded that DACA shared the "same legal . . . defects that the courts recognized as to DAPA" and was "likely" to meet a similar fate. Id., at 878. "In light of the costs and burdens" that a rescission would "impose[ ] on DHS," the Attorney General urged DHS to "consider an orderly and efficient wind-down process."

The next day, Duke acted on the Attorney General's advice. In her decision memorandum, Duke summarized the history of the DACA and DAPA programs, the Fifth Circuit opinion and ensuing affirmance, and the contents of the Attorney General's letter. "Taking into consideration the Supreme Court's and the Fifth Circuit's rulings" and the "letter from the Attorney General," she concluded that the "DACA program should be terminated."

Duke then detailed how the program would be wound down: No new applications would be accepted, but DHS would entertain applications for two-year renewals from DACA recipients whose benefits were set to expire within six months. For all other DACA recipients, previously issued grants of deferred action and work authorization would not be revoked but would expire on their own terms, with no prospect for renewal.

B

Within days of Acting Secretary Duke's rescission announcement, multiple groups of plaintiffs ranging from individual DACA recipients and States to the Regents of the University of California and the National Association for the Advancement of Colored People challenged her decision in the U. S. District Courts for the Northern District of California [. . .], the Eastern District of New York […], and the District of Columbia. The relevant claims are that the rescission was arbitrary and capricious in violation of the APA and that it infringed the equal protection guarantee of the Fifth Amendment's Due Process Clause.

All three District Courts ruled for the plaintiffs. ... In doing so, each court rejected the Government's threshold arguments that the claims were unreviewable under the APA and that the INA deprived the court of jurisdiction.

[...]

In NAACP, the D. C. District Court … granted partial summary judgment to the plaintiffs on their APA claim, holding that Acting Secretary Duke's "conclusory statements were insufficient to explain the change in [the agency's] view of DACA's lawfulness." … The District Court stayed its order for 90 days to permit DHS to "reissue a memorandum rescinding DACA, this time providing a fuller explanation for the determination that the program lacks statutory and constitutional authority."

Two months later, Duke's successor, Secretary Kirstjen M. Nielsen, responded via memorandum.  She explained that, "[h]aving considered the Duke memorandum," she "decline[d] to disturb" the rescission. Secretary Nielsen went on to articulate her "understanding" of Duke's memorandum, identifying three reasons why, in Nielsen's estimation, "the decision to rescind the DACA policy was, and remains, sound."

First, she reiterated that, "as the Attorney General concluded, the DACA policy was contrary to law."

Second, she added that, regardless, the agency had "serious doubts about [DACA's] legality" and, for law enforcement reasons, wanted to avoid "legally questionable" policies.

Third, she identified multiple policy reasons for rescinding DACA, including (1) the belief that any class-based immigration relief should come from Congress, not through executive non-enforcement; (2) DHS's preference for exercising prosecutorial discretion on "a truly individualized, case-by-case basis"; and (3) the importance of "project[ing] a message" that immigration laws would be enforced against all classes and categories of aliens. In her final paragraph, Secretary Nielsen acknowledged the "asserted reliance interests" in DACA's continuation but concluded that they did not "outweigh the questionable legality of the DACA policy and the other reasons" for the rescission discussed in her memorandum.

The Government asked the D. C. District Court to revise its prior order in light of the reasons provided by Secretary Nielsen, but the court declined. In the court's view, the new memorandum, which "fail[ed] to elaborate meaningfully" on the agency's illegality rationale, still did not provide an adequate explanation for the September 2017 rescission.

[. . . ]

The issues raised here are . . .  whether the rescission [of DACA] was arbitrary and capricious in violation of the APA [...]

III

[...]

B

We turn, finally, to whether DHS's decision to rescind DACA was arbitrary and capricious. As noted earlier, Acting Secretary Duke's justification for the rescission was succinct: "Taking into consideration" the Fifth Circuit's conclusion that DAPA was unlawful because it conferred benefits in violation of the INA, and the Attorney General's conclusion that DACA was unlawful for the same reason, she concluded—without elaboration—that the "DACA program should be terminated."

Respondents maintain that this explanation is deficient for three reasons. Their first and second arguments work in tandem, claiming that the Duke Memorandum does not adequately explain the conclusion that DACA is unlawful, and that this conclusion is, in any event, wrong. While those arguments carried the day in the lower courts, in our view they overlook an important constraint on Acting Secretary Duke's decisionmaking authority—she was boundby the Attorney General's legal determination.

The same statutory provision that establishes the Secretary of Homeland Security's authority to administer and enforce immigration laws limits that authority, specifying that, with respect to "all questions of law," the determinations of the Attorney General "shall be controlling." 8 U. S. C. § 1103(a)(1). Respondents are aware of this constraint. Indeed they emphasized the point in the reviewability sections of their briefs. But in their merits arguments, respondents never addressed whether or how this unique statutory provision might affect our review. They did not discuss whether Duke was required to explain a legal conclusion that was not hers to make. Nor did they discuss whether the current suits challenging Duke's rescission decision, which everyone agrees was within her legal authority under the INA, are proper vehicles for attacking the Attorney General's legal conclusion.

Because of these gaps in respondents' briefing, we do not evaluate the claims challenging the explanation and correctness of the illegality conclusion. Instead we focus our attention on respondents' third argument—that Acting Secretary Duke "failed to consider . . . important aspect[s] of the problem" before her. Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co., 463 U. S. 29, 43 (1983).

Whether DACA is illegal is, of course, a legal determination, and therefore a question for the Attorney General. But deciding how best to address a finding of illegality moving forward can involve important policy choices, especially when the finding concerns a program with the breadth of DACA. Those policy choices are for DHS.

Acting Secretary Duke plainly exercised such discretionary authority in winding down the program. Among other things, she specified that those DACA recipients whose benefits were set to expire within six months were eligible for two-year renewals.

But Duke did not appear to appreciate the full scope of her discretion, which picked up where the Attorney General's legal reasoning left off. The Attorney General concluded that "the DACA policy has the same legal . . . defects that the courts recognized as to DAPA." So, to understand those defects, we look to the Fifth Circuit, the highest court to offer a reasoned opinion on the legality of DAPA. That court described the "core" issue before it as the "Secretary's decision" to grant "eligibility for benefits"— including work authorization, Social Security, and Medicare—to unauthorized aliens on "a class-wide basis." The Fifth Circuit's focus on these benefits was central to every stage of its analysis. …And the Court ultimately held that DAPA was "manifestly contrary to the INA" precisely because it "would make 4.3 million otherwise removable aliens" eligible for work authorization and public benefits.

But there is more to DAPA (and DACA) than such benefits. The defining feature of deferred action is the decision to defer removal (and to notify the affected alien of that decision). And the Fifth Circuit was careful to distinguish that forbearance component from eligibility for benefits. As it explained, the "challenged portion of DAPA's deferred-action program" was the decision to make DAPA recipients eligible for benefits. The other "[p]art of DAPA," the court noted, "involve[d] the Secretary's decision—at least temporarily—not to enforce the immigration laws as to a class of what he deem[ed] to be low-priority illegal aliens." Borrowing from this Court's prior description of deferred action, the Fifth Circuit observed that "the states do not challenge the Secretary's decision to `decline to institute proceedings, terminate proceedings, or decline to execute a final order of deportation. And the Fifth Circuit underscored that nothing in its decision or the preliminary injunction "requires the Secretary to remove any alien or to alter" the Secretary's class-based "enforcement priorities." In other words, the Secretary's forbearance authority was unimpaired.

Acting Secretary Duke recognized that the Fifth Circuit's holding addressed the benefits associated with DAPA. In her memorandum she explained that the Fifth Circuit concluded that DAPA "conflicted with the discretion authorized by Congress" because the INA "`flatly does not permit the reclassification of millions of illegal aliens as lawfully present and thereby make them newly eligible for a host of federal and state benefits, including work authorization.'" Duke did not characterize the opinion as one about forbearance.

In short, the Attorney General neither addressed the forbearance policy at the heart of DACA nor compelled DHS to abandon that policy. Thus, removing benefits eligibility while continuing forbearance remained squarely within the discretion of Acting Secretary Duke, who was responsible for "[e]stablishing national immigration enforcement policies and priorities." But Duke's memo offers no reason for terminating forbearance. She instead treated the Attorney General's conclusion regarding the illegality of benefits as sufficient to rescind both benefits and forbearance, without explanation.

That reasoning repeated the error we identified in one of our leading modern administrative law cases, Motor Vehicle Manufacturers Association of the United States, Inc. v. State Farm Mutual Automobile Insurance Co. There, the National Highway Traffic Safety Administration (NHTSA) promulgated a requirement that motor vehicles produced after 1982 be equipped with one of two passive restraints: airbags or automatic seatbelts. Four years later, before the requirement went into effect, NHTSA concluded that automatic seatbelts, the restraint of choice for most manufacturers, would not provide effective protection. Based on that premise, NHTSA rescinded the passive restraint requirement in full.

We concluded that the total rescission was arbitrary and capricious. As we explained, NHTSA's justification supported only "disallow[ing] compliance by means of" automatic seatbelts. It did "not cast doubt" on the "efficacy of airbag technology" or upon "the need for a passive restraint standard." Given NHTSA's prior judgment that "airbags are an effective and cost-beneficial lifesaving technology," we held that "the mandatory passive restraint rule [could] not be abandoned without any consideration whatsoever of an airbags-only requirement."

While the factual setting is different here, the error is the same. Even if it is illegal for DHS to extend work authorization and other benefits to DACA recipients, that conclusion supported only "disallow[ing]" benefits. It did "not cast doubt" on the legality of forbearance or upon DHS's original reasons for extending forbearance to childhood arrivals. Thus, given DHS's earlier judgment that forbearance is "especially justified" for "productive young people" who were brought here as children and "know only this country as home," the DACA Memorandum could not be rescinded in full "without any consideration whatsoever" of a forbearance-only policy.

The Government acknowledges that "[d]eferred action coupled with the associated benefits are the two legs upon which the DACA policy stands." It insists, however, that "DHS was not required to consider whether DACA's illegality could be addressed by separating" the two.  According to the Government, "It was not arbitrary and capricious for DHS to view deferred action and its collateral benefits as importantly linked." Perhaps. But that response misses the point. The fact that there may be a valid reason not to separate deferred action from benefits does not establish that DHS considered that option or that such consideration was unnecessary.

The lead dissent acknowledges that forbearance and benefits are legally distinct and can be decoupled. (opinion of THOMAS, J). It contends, however, that we should not "dissect" agency action "piece by piece." The dissent instead rests on the Attorney General's legal determination—which considered only benefits—"to supply the `reasoned analysis'" to support rescission of both benefits and forbearance. But State Farm teaches that when an agency rescinds a prior policy its reasoned analysis must consider the "alternative[s]" that are "within the ambit of the existing [policy]." Here forbearance was not simply "within the ambit of the existing [policy]," it was the centerpiece of the policy: DACA, after all, stands for "Deferred Action for Childhood Arrivals." But the rescission memorandum contains no discussion of forbearance or the option of retaining forbearance without benefits. Duke "entirely failed to consider [that] important aspect of the problem." State Farm.

That omission alone renders Acting Secretary Duke's decision arbitrary and capricious. But it is not the only defect. Duke also failed to address whether there was "legitimate reliance" on the DACA Memorandum. When an agency changes course, as DHS did here, it must "be cognizant that longstanding policies may have `engendered serious reliance interests that must be taken into account.' "It would be arbitrary and capricious to ignore such matters." Yet that is what the Duke Memorandum did.

For its part, the Government does not contend that Duke considered potential reliance interests; it counters that she did not need to. In the Government's view, shared by the lead dissent, DACA recipients have no "legally cognizable reliance interests" because the DACA Memorandum stated that the program "conferred no substantive rights" and provided benefits only in two-year increments. But neither the Government nor the lead dissent cites any legal authority establishing that such features automatically preclude reliance interests, and we are not aware of any. These disclaimers are surely pertinent in considering the strength of any reliance interests, but that consideration must be undertaken by the agency in the first instance, subject to normal APA review. There was no such consideration in the Duke Memorandum.

Respondents and their amici assert that there was much for DHS to consider. They stress that, since 2012, DACA recipients have "enrolled in degree programs, embarked on careers, started businesses, purchased homes, and even married and had children, all in reliance" on the DACA program. The consequences of the rescission, respondents emphasize, would "radiate outward" to DACA recipients' families, including their 200,000 U. S.-citizen children, to the schools where DACA recipients study and teach, and to the employers who have invested time and money in training them. In addition, excluding DACA recipients from the lawful labor force may, they tell us, result in the loss of $215 billion in economic activity and an associated $60 billion in federal tax revenue over the next ten years. Meanwhile, States and local governments could lose $1.25 billion in tax revenue each year.

These are certainly noteworthy concerns, but they are not necessarily dispositive. To the Government and lead dissent's point, DHS could respond that reliance on forbearance and benefits was unjustified in light of the express limitations in the DACA Memorandum. Or it might conclude that reliance interests in benefits that it views as unlawful are entitled to no or diminished weight. And, even if DHS ultimately concludes that the reliance interests rank as serious, they are but one factor to consider. DHS may determine, in the particular context before it, that other interests and policy concerns outweigh any reliance interests. Making that difficult decision was the agency's job, but the agency failed to do it.

DHS has considerable flexibility in carrying out its responsibility. The wind-down here is a good example of the kind of options available. Acting Secretary Duke authorized DHS to process two-year renewals for those DACA recipients whose benefits were set to expire within six months. But Duke's consideration was solely for the purpose of assisting the agency in dealing with "administrative complexities." She should have considered whether she had similar flexibility in addressing any reliance interests of DACA recipients. The lead dissent contends that accommodating such interests would be "another exercise of unlawful power," but the Government does not make that argument and DHS has already extended benefits for purposes other than reliance, following consultation with the Office of the Attorney General.

Had Duke considered reliance interests, she might, for example, have considered a broader renewal period based on the need for DACA recipients to reorder their affairs. Alternatively, Duke might have considered more accommodating termination dates for recipients caught in the middle of a time-bounded commitment, to allow them to, say, graduate from their course of study, complete their military service, or finish a medical treatment regimen. Or she might have instructed immigration officials to give salient weight to any reliance interests engendered by DACA when exercising individualized enforcement discretion.

To be clear, DHS was not required to do any of this or to "consider all policy alternatives in reaching [its] decision." Agencies are not compelled to explore "every alternative device and thought conceivable by the mind of man." Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, Inc., 435 U. S. 519, 551 (1978). But, because DHS was "not writing on a blank slate," it was required to assess whether there were reliance interests, determine whether they were significant, and weigh any such interests against competing policy concerns.

The lead dissent sees all the foregoing differently. In its view, DACA is illegal, so any actions under DACA are themselves illegal. Such actions, it argues, must cease immediately and the APA should not be construed to impede that result.

The dissent is correct that DACA was rescinded because of the Attorney General's illegality determination. But nothing about that determination foreclosed or even addressed the options of retaining forbearance or accommodating particular reliance interests. Acting Secretary Duke should have considered those matters but did not. That failure was arbitrary and capricious in violation of the APA.

* * *

We do not decide whether DACA or its rescission are sound policies. "The wisdom" of those decisions "is none of our concern." We address only whether the agency complied with the procedural requirement that it provide a reasoned explanation for its action. Here the agency failed to consider the conspicuous issues of whether to retain forbearance and what if anything to do about the hardship to DACA recipients. That dual failure raises doubts about whether the agency appreciated the scope of its discretion or exercised that discretion in a reasonable manner. The appropriate recourse is therefore to remand to DHS so that it may consider the problem anew.

The judgment in NAACP is affirmed. The judgment in Regents is vacated in part and reversed in part. And in Batalla Vidal the February 13, 2018 order granting respondents' motion for a preliminary injunction is vacated, the November 9, 2017 order partially denying the Government's motion to dismiss is affirmed in part, and the March 29, 2018 order partially denying the balance of the Government's motion to dismiss is reversed in part. All three cases are remanded for further proceedings consistent with this opinion.

It is so ordered.