United States federal administrative law encompasses statutes, rules, judicial precedents, and executive orders, that together form administrative laws that define the extent of powers and responsibilities held by administrative agencies of the United States government, including executive departments and independent agencies. Because Congress, the president, and the federal courts have limited resources to address all issues, specialized powers are often delegated to a board, commission, office, or other agency. These administrative agencies oversee and monitor activities in complex areas, such as commercial aviation, medical device manufacturing, and securities markets.
Former Supreme Court Justice Stephen Breyer has defined the legal rules and principles of administrative law in four parts: (1) define the authority and structure of administrative agencies; (2) specify the procedural formalities employed by agencies; (3) determine the validity of agency decisions; and (4) define the role of reviewing courts and other governmental entities in relation to administrative agencies.[1]
U.S. federal agencies have the power to adjudicate, legislate, and enforce laws within their specific areas of delegated power. The United States does not use administrative courts.[2] Adjudication is carried out internally within agencies by administrative law judges.
The authority of federal administrative agencies stems from their organic statutes, and must be consistent with constitutional constraints and the scope of authority granted by statute.[3]
Federal administrative agencies, when granted the power to do so in a statutory grant of authority from Congress, may promulgate rules that have force of law. Agencies "legislate" through rulemaking—the power to promulgate (or issue) regulations. Such regulations are codified in the Code of Federal Regulations (CFR) and published in the Federal Register. Rules of lesser effect are published in a host of forms, including manuals for agency staff and for the public, circulars, bulletins, letter rulings, press releases, and the like.
Section 551 of the Administrative Procedure Act gives the following definitions:
The primary administrative law statutes and other laws that govern agency rule making include:[4]
Limits on the power of agencies to promulgate regulations include:
Agencies may not promulgate retroactive rules unless expressly granted such power by the agency's organic statute. Bowen v. Georgetown University Hospital,
There is no broad prohibition against an agency's regulation that does not serve the "public convenience, interest, or necessity." The law presumes that rulemaking conducted with procedural safeguards of the statutes and Executive Orders noted above reflect a rational balancing of interests by the agency, and a court will strike down a regulation only for violation of those procedures.
Agencies are permitted to rely on rules in reaching their decisions rather than adjudicate, where the promulgation of the rules is within the agency's statutory authority, and the rules themselves are not arbitrary or capricious. Heckler v. Campbell, .
Agencies must abide by their own rules and regulations. Accardi v. Shaughnessy, .
Courts must defer to administrative agency interpretations of the authority granted to them by Congress (1) where the intent of Congress was ambiguous and (2) where the interpretation was reasonable or permissible. Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., . Chevron is probably the most frequently cited case in American administrative law.[6] However, in June 2024, the Supreme Court overruled Chevron in Loper Bright Enterprises v. Raimondo; past administrative decisions made before Loper Bright under the Chevron deference remained in place, but future agency administration actions in interpreting Congressional language are more likely to be subject to judicial review.[7]
There are five levels of rulemaking procedure:
"Nonlegislative rules" include three main classes:[9]
A class called "guidance" includes all rules not promulgated by legislative procedure. Such rules may be published as guidance, guidelines, agency staff manuals, staff instructions, opinion letters, interpretive memoranda, policy statements, guidance manuals for the public, circulars, bulletins, advisories, press releases stating agency position, and the like. The class of "guidance" is almost, but not exactly, coextensive with the union of the sets of interpretative rules, general statements of policy, and housekeeping rules.
Every statute and regulation has some lingering ambiguity. Someone has to have authority to adopt some interpretation, and do so with a minimum of procedural delay. So the law grants every agency the authority to promulgate interpretative rules, and to do so with minimal procedural fuss. By default, most interpretations slot into the *interpretative rule" category of 5 U.S.C. § 553(b).
If an interpretation satisfies a long list of criteria, then the interpretation is binding on parties before the agency, courts, and the agency itself, under Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc. (for agency interpretations of statute)[11] or Auer v. Robbins (for agency interpretations of regulations).[12] (The inquiries under Chevron and Auer are slightly different. But the analytical similarities overshadow the differences. For this short article, we will gloss over the differences, and treat Chevron and Auer together.)
Any interpretation that fails any one of the Chevron/Auer eligibility bullets from the list below falls into the residual category of "interpretative rule".
Fundamentally, the § 553(b) "interpretative" exemption from notice and comment is a rule of necessity—essentially all laws have some ambiguity, that ambiguity has to be interpreted, and (for public-facing substantive rules) the agency is the party that can do so expeditiously and fairly. Deference follows to the degree the agency demonstrates fairness and diligence in developing its interpretation (under Chevron, Auer, or Skidmore v. Swift & Co., as appropriate).
The quid pro quo for an agency's choice to exercise the "interpretative" option, and forego the formalities required for legislative rulemaking or for Chevron or Auer deference, is that the agency has very little enhanced power to enforce its interpretation. If a party challenges the agency's interpretation, an agency's invocation of the "interpretative" exemption surrenders any claim to heightened Chevron or Auer deference,[13] and the interpretation falls into the residual category, under which a court gives Skidmore deference to an agency's informed position:[14]
As a practical matter, an agency operates under the agency's interpretative rules. The law permits parties before the agency to argue alternative interpretations, and under the law, agencies are supposed to respond to the arguments, and not foreclose alternatives suggested by parties. But as a practical matter, agencies seldom give anything more than short shrift consideration to alternatives. On judicial review, the practical reality is that a court is most likely to agree with the agency, under Skidmore deference. But Skidmore deference is only as strong as the quality of the agency's analysis, and courts regularly overturn "interpretative" rules.
Some agency interpretations are binding on parties and the courts, under Chevron deference:[15]
But an agency has to earn this deference; it is far from automatic. When an agency interprets its own organic statute (for Chevron) or a regulation that it promulgated (under Auer), and the interpretation meets all the following prerequisites, only then does the agency receive the high deference of Chevron or Auer.
In addition to the three classical steps, an agency must observe additional procedural formalities:
An "interpretative" rule sets out the agency's interpretation of a statute or rule, without altering rights or obligations. If the interpretation fails at least one of the Chevron/Auer criteria, then the interpretation falls into the category of "interpretative rule" which binds only the agency itself, and is entitled to at most Skidmore deference.
=
The line for permissible exercise of the § 553(b) "interpretative" exemption is blurry—courts and treatise writers uniformly complain about this.[22] The most basic requirement for the "interpretative" exemption is that the agency "interpret" a validly promulgated law (statute or regulation), by following a recognizable interpretative path originally set out by the statute or regulation. An agency may promulgate an "interpretative" rule "only if the agency's position can be characterized as an 'interpretation' of a statute or legislative regulation rather than as an exercise of independent policymaking authority."[23] Mere "consistency" (in the sense of "absence of clash") is insufficient.[24] Most "gap filling" is beyond the scope of *interpretative" authority. A valid interpretative rule merely explains, but does not add to or alter, the law that already exists in the form of a statute or legislative rule.[25] An "interpretative" rule cannot create a new requirement, carve-out, or exception from whole cloth. If the rule changes "individual rights and obligations" (rather than resolving ambiguity), the rule requires legislative procedure.
An agency may promulgate interpretative rules outside the scope of its rule making authority. Where an agency can only issue legislative rules pursuant to an express grant of authority from Congress, an agency may (and is encouraged to) issue advisory interpretations to guide the public.
If an agency elects the "interpretative" shortcut, there are almost no procedural requirements, beyond the publication required by 5 U.S.C. § 552 and § 552(d). The decision maker must ensure that there is indeed an ambiguity that is not resolved by any binding law, but if the ambiguity exists, the decision maker simply interprets as best he or she may. If the issue is outside the agency's scope of rule making authority,[26] the agency must follow the agency or courts that do have authority on that specific issue.
=
In return for the privilege of bypassing rule making procedure, the agency risks loss of binding effect for an interpretative rule. "An agency issuing an interpretative rule ... may well intend that its interpretation bind its own personnel and may expect compliance from regulated individuals or entities. Nonetheless, the agency cannot expect the interpretation to be binding in court; because it does not have the force of law, parties can challenge the interpretation."[27] Many courts have characterized interpretative rules as only "hortatory" and "lacking force of law".[28] [29]
In proceedings before the agency, a party may advance alternative positions or interpretations, and the agency must address them, without relying on an interpretative rule as the last word.[30] But as a practical matter, agencies tend to enforce their interpretative rules until forced to concede error, and parties simply acquiesce until the costs of the agency's interpretation exceed the cost of court litigation.[31]
Interpretative rules are binding on agency employees, including its administrative law judges (ALJs).[32] If an interpretative rule (say, a provision of an agency staff manual, or memorandum to agency staff, or interpretation in a Federal Register rulemaking notice) sets a "floor" under the rights of a party, individual employees have no discretion to back out of the agency's interpretation or create ad hoc exceptions adverse to the party.
On judicial review, the effect of an interpretative rule was explained by the Eighth Circuit, discussing "well-settled principles of administrative law":
"Statements of policy" are even weaker statements than "interpretative" rules. Agencies use them to express agency preferences, but with no binding effect. Policy statements are "tentative intentions", nonbinding rules of thumb, suggestions for conduct, and tentative indications of an agency's hopes. Policy statements have no binding effect. A policy statement "genuinely leaves the agency and its decisionmakers free to exercise discretion", and "a statement of policy may not have a present effect: a 'general statement of policy' is one that does not impose any rights and obligations".[33]
Agencies likewise use "policy statements" to offer a unilateral quid pro quo or set a floor for agency procedure ("If you the public do X, we the agency promise favorable outcome Y. If you don't do X, you can still convince us to do Y by arguing the controlling law.").
A near-certain indicator of a "statement of policy" is the word "should" rather than "must".
Agency policy statements are not binding on courts, and therefore receive no Chevron deference, not even weak Skidmore deference.[34]
"Guidance" is a residual category for any rule issued by an agency but not in a formally promulgated regulation. Most non-Chevron interpretative rules, and most general statements of policy, are issued as guidance.
Only three classes of law administered by agencies are binding against members of the public: statute (as interpreted by the courts), regulations, and common law. Against members of the public, the default rule, embodied in the Administrative Procedure Act, 5 U.S.C. § 553, and elaborated in the Good Guidance Bulletin, is that subregualtory guidance documents do not have force of law, and do not bind the public. Perez v. Mortgage Bankers Association, 575 U.S. 82 (2015); e.g., 15 C.F.R. § 29.2. Nothing in agency guidance documents is binding against members of the public, except—
An agency can change its guidance with very little procedure (unlike a regulation), but as long as the guidance reads as it does, parties are entitled to rely on it for the three classes of promises listed above. In the late 1990s and early 2000s, many agencies attempted to bypass the APA's requirements for rulemaking by tucking rules that went beyond interpretation of ambiguity into informal documents like agency staff manuals and the like. This is simply illegal. The Executive Office of the President stepped in to stop bootleg rulemaking in 2007, and forbade this practice.[36] Some agencies, for example, the U.S. Patent and Trademark Office, have nonetheless continued to defy the law, and state their formal refusal to implement the President's directive.
Executive Order 12866, which was issued in 1993, requires agencies (other than independent agencies) to submit proposed rules for reviews by the Office of Information and Regulatory Affairs if the rule meets certain criteria. Rules that are "economically significant" (meeting the criteria of "an annual effect on the economy of $100 million or more or adversely affect in a material way the economy") require a cost-benefit analysis.
The Congressional Review Act passed in 1996 created a category of major rules, which are those that OIRA determines result in either: (1) "an annual effect on the economy of $100,000,000", (2) "a major increase in costs or prices for consumers, individual industries, Federal, State, or local government agencies, or geographic regions", or (3) "significant adverse effects on competition, employment, investment, productivity, innovation, or on the ability of United States-based enterprises to compete with foreign-based enterprises in domestic and export markets". If the rule is major, an additional report must be provided to congressional committees.
About 2,500 to 4,000 rules are published per year.[37] Of those, in 2012, 68% were classified as Routine/Info/Other while the remainder were Significant/Substantive.[37] From 1997 to 2012, the number of major rules for Congressional Review Act purposes has ranged from 100 (2010) to 50 (2002).[37]
Section 551 of the Administrative Procedure Act gives the following definitions:
There are two ways that an individual can attain the right to a hearing in an adjudicative proceeding. First, the Due Process Clause of the 5th Amendment or 14th Amendment can require that a hearing be held if the interest that is being adjudicated is sufficiently important or if, without a hearing, there is a strong chance that the petitioner will be erroneously denied that interest.[38] A hearing can also be required if a statute somehow mandates the agency to hold formal hearings when adjudicating certain issues.
The adjudication will typically be completed with a written report containing findings of fact and conclusions of law, both at the state and federal level.[39] If the affected does not wish to contest the action, a consent order may be published allowing for the hearing to be bypassed.
See main article: Federal tribunals in the United States.
The Administrative Procedure Act puts "adjudication" and "rulemaking" on opposite sides of a "dichotomy".[40]
Agency actions are divided into two broad categories discussed above, rulemaking and adjudication. For agency decisions that have broad impact on a number of parties, including parties not specifically before the agency, the agency must use the procedures of rulemaking (see the bullet list in above). Because actions by rulemaking affect many parties, rulemaking procedures are designed to ensure public participation, and are therefore more cumbersome, except that the agency is permitted to seek comment by publication of notice, without soliciting the views of specific parties. For decisions that, at first instance, affect only a small number of parties that actually appear before the agency (see above), the agency may use procedures that are generally simpler, but that require the agency specifically solicit input from the directly affected parties before the agency applies the rule more broadly.
Adjudication decisions may become precedent that binds future parties, so the transition zone between the scope of issues requiring rulemaking and the scope of issues that may be determined by case-by-case adjudication is very hazy.
The classical test for the dividing line is seen in the contrast between two cases decided a few years apart, both involving taxes levied by the city of Denver Colorado. In 1908, in Londoner v. City and County of Denver, a tax levied on residents of a particular street was held to be an adjudication, and the Supreme Court ordered the city to do a "do over", because the residents of the street were not given sufficient opportunity to be heard. Then, in 1915, in Bi-Metallic Investment Co. v. State Board of Equalization, a tax levied on the entire city of Denver was held to be rulemaking, and the city's action of imposing the tax was affirmed.
Factors tending to make an act adjudicative in nature:
For most agencies, the choice of whether to promulgate rules or proceed by common law adjudicative decisions rests in the informed discretion of agencies. SEC v. Chenery Corp., (Dissenting opinion arguing that the decision permitted agencies to rule arbitrarily, without law). Agencies may also announce new policies in the course of such adjudications. Some agencies' organic statutes obligate the agency to use rulemaking, for example, the U.S. Patent and Trademark Office, 35 U.S.C. § 2(b)(2)(B).
A party aggrieved by an agency action (either rulemaking or adjudication) may seek judicial review (that is, sue) as provided by an agency's organic statute or by §§ 701-706 of the Administrative Procedure Act.
Studies of judicial review typically find that 70% of agency rules are upheld with the Supreme Court upholding 91% of rules; a 2011 empirical study of judicial review found that 76% were upheld,[41] although the D.C. Circuit, which hears many administrative law cases, has been found less deferential than other courts.[42]