Toolkit · Administrative Law

The Administrative Law Toolkit

5 U.S.C. §§ 551

Type
Toolkit
Jurisdiction
National
Date
2026-09-08
Topic
Administrative Law

This toolkit is the front door to administrative law and agency practice on UpLaw. It maps the field in eleven pillars — the constitutional structure of the administrative state (delegation, appointments, removal), the Administrative Procedure Act's basic division between rulemaking and adjudication, how to participate in notice-and-comment and build a comment record that survives review, guidance and subregulatory action, agency adjudication and the jury right after SEC v. Jarkesy, investigations and enforcement, judicial review under the post-Loper Bright framework, the major questions doctrine, justiciability and timing, remedies, and the agency-specific regimes covered elsewhere in the library. Chevron deference is gone: Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024), restored de novo statutory interpretation, and every pillar below reflects that. A recurring worked example follows one fictional regulated manufacturer, Calder Instruments, as it collides with five different agencies. Use it as a research map — skim the landscape, jump to the pillar you need, and follow the links down into the detail.

The Administrative Law Toolkit

Administrative Law > Agency Practice | Published 8 September 2026 - Updated 6 October 2026 | UpLaw - uplaw.us

Most federal law that touches a business is not made by Congress and not applied by a court. It is written by an agency, enforced by an agency, and first adjudicated by an agency. Administrative law is the discipline of holding that machinery to its own rules — and after 2024, courts are no longer required to take the agency's word for what the law means.

Keywords: administrative law, Administrative Procedure Act, notice-and-comment rulemaking, agency adjudication, judicial review, Loper Bright, Skidmore respect, arbitrary and capricious, major questions doctrine, Jarkesy, Corner Post, final agency action, exhaustion, vacatur, guidance documents

How to read this toolkit

Administrative law feels abstract until you notice that it is the operating system underneath almost every other practice area in this library. The SEC's enforcement program, the EPA's permits, the FTC's competition rules, the IRS's regulations, the Department of Labor's overtime thresholds, USCIS's adjudications, and BIS's export licenses are all instances of one repeating pattern: an agency exercising delegated power, through a procedure, subject to review. Learn the pattern once and you can practice in any of them.

This toolkit walks that pattern in eleven pillars. Each pillar gives you the plain-English doctrine, the anchor authority, the traps that catch practitioners, and a link into UpLaw's deeper guide, checklist, or article where the real detail lives. It is a research map, not a treatise. The treatise is Administrative Law: A Comprehensive Guide to Agency Action and Judicial Review; the working how-to for a challenge is Challenging Agency Action Under the APA: A Practical Checklist. Keep both open alongside this page.

Use it three ways. Read it end to end once to build the mental model of how power flows from statute to rule to enforcement to court. Jump to a pillar when you have a live question — "is this guidance or a legislative rule?", "is the action final?", "did we waive that argument?" — and follow the deep-dive links. Or use the consolidated gated checklist near the end as an audit of any agency matter, whether you are trying to shape a rule, survive an enforcement action, or vacate a regulation.

One currency warning that colors everything below. Chevron deference is no longer the law. Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024), overruled Chevron U.S.A. Inc. v. Natural Resources Defense Council and held that courts must exercise independent judgment in deciding whether an agency has acted within its statutory authority. Briefs, memos, and CLE decks that still recite "Step One / Step Two" are not merely dated — they are wrong. Every pillar in this toolkit reflects the post-Loper Bright framework.

Meet the running example. Calder Instruments is a fictional mid-cap manufacturer of industrial and medical sensors. It is publicly traded, runs a coating line that discharges to a river and vents to the air, sells dual-use sensors abroad, sponsors visas for its engineers, and is subject to a new agency rule capping a chemical it depends on. Over the next eleven pillars, Calder will be regulated, investigated, prosecuted in an in-house tribunal, and will ultimately sue. Its problems are the spine of this map.

The administrative landscape at a glance

The eleven pillars, in the order this toolkit walks them:

  1. Constitutional structure — Where agency power comes from and what limits it: nondelegation, appointments, removal. (Art. II; Whitman; Seila Law)
  2. The APA's basic architecture — Rulemaking vs. adjudication; formal vs. informal. (5 U.S.C. §§ 551–559)
  3. Notice-and-comment rulemaking — How a legislative rule is made, and how to participate so your objection survives. (5 U.S.C. § 553)
  4. Guidance and subregulatory action — Interpretive rules, policy statements, FAQs, and "regulation by press release." (§ 553(b)(A); Perez)
  5. Adjudication, ALJs, and the jury right — In-house tribunals after SEC v. Jarkesy, 603 U.S. 109 (2024). (§§ 554, 556–557)
  6. Investigations and enforcement — Subpoenas, CIDs, and the presumption against reviewing nonenforcement. (Morton Salt; Heckler v. Chaney)
  7. Judicial review after Loper Bright — De novo statutory interpretation, Skidmore respect, arbitrary-and-capricious review. (§ 706; State Farm)
  8. The major questions doctrine — When "vast economic and political significance" demands clear congressional authorization. (West Virginia v. EPA)
  9. Justiciability and timing — Standing, final agency action, ripeness, exhaustion, and the limitations clock after Corner Post. (Bennett; Lujan; Darby)
  10. Remedies — Vacatur, remand without vacatur, § 705 stays, and interim relief.
  11. The agency-specific regimes — Where each regulator lives in this library.

Pillar 1 — Constitutional structure: delegation, appointments, removal

Agencies have no inherent power. Every valid agency act traces back to a statute, and the statute must itself be constitutional. Three structural doctrines police that chain.

Nondelegation. Congress may hand an agency broad authority so long as it supplies an "intelligible principle" to guide the exercise of discretion — the test announced in J.W. Hampton, Jr., & Co. v. United States, 276 U.S. 394 (1928), and applied with famous permissiveness ever since (Mistretta v. United States, 488 U.S. 361 (1989); Whitman v. American Trucking Ass'ns, 531 U.S. 457 (2001)). The Supreme Court has not invalidated a statute on nondelegation grounds since 1935, but several Justices have signaled interest in a stricter test (see the separate writings in Gundy v. United States (2019)). In practice, the doctrine's energy has migrated into the major questions doctrine (Pillar 8), which does much of the same work through interpretation rather than invalidation.

Appointments. Officers of the United States must be appointed under the Appointments Clause; "inferior officers" may be appointed by the President, courts, or heads of departments. Lucia v. SEC (2018) held that SEC administrative law judges are inferior officers who must be so appointed — a holding that voided proceedings before improperly appointed ALJs and rippled through every agency with an in-house bench.

Removal. Free Enterprise Fund v. PCAOB, 561 U.S. 477 (2010), struck down two layers of for-cause removal protection. Seila Law LLC v. CFPB (2020) held that a single-director agency wielding substantial executive power cannot be insulated from at-will presidential removal, narrowing Humphrey's Executor v. United States, 295 U.S. 602 (1935), toward multimember bodies exercising quasi-legislative functions.

The practice point that matters. Structural constitutional objections used to be trapped inside the agency's own process. Axon Enterprise, Inc. v. FTC (2023) changed that: a party may bring a here-and-now constitutional challenge to an agency's structure directly in federal district court without first grinding through the administrative proceeding. When Calder receives its first administrative complaint, structural defenses are on the table immediately — not years later on petition for review.

Pillar 2 — The APA: rulemaking versus adjudication

The Administrative Procedure Act, 5 U.S.C. §§ 551–559 and 701–706, is a short statute that organizes almost everything. Its foundational move is a definitional one: agency action is either a rule ("of general applicability and future effect," § 551(4)) or an order resulting from adjudication (§ 551(6)–(7)). Rulemaking is legislative in character and prospective; adjudication is judicial in character and typically retrospective and party-specific.

The constitutional intuition behind the divide is older than the APA. Bi-Metallic Investment Co. v. State Board of Equalization, 239 U.S. 441 (1915), held that when a rule affects a large class, the remedy is political — you get a comment period, not a hearing. Londoner v. City and County of Denver, 210 U.S. 373 (1908), had held the opposite for individualized determinations: those get process. That is the whole architecture in two cases.

Each track then splits into formal and informal versions. Formal rulemaking (§§ 556–557, triggered only when a statute requires rules be made "on the record after opportunity for an agency hearing") is nearly extinct. Informal — "notice-and-comment" — rulemaking under § 553 is the workhorse. Formal adjudication under §§ 554, 556–557 is alive and well (SEC, NLRB, EOIR-adjacent proceedings), while vast quantities of informal adjudication (a visa denial, a permit issuance, a benefits determination) proceed under nothing but the agency's organic statute and the Due Process Clause.

Trap. Agencies sometimes choose adjudication to make policy that looks like a rule — announcing a new standard in a single enforcement order and applying it to everyone thereafter. That is generally permitted, but it interacts badly with retroactivity and fair-notice principles, and it is a productive line of attack. Conversely, agencies sometimes label a legislative rule "guidance" to skip § 553 (Pillar 4). Characterization is contested ground, and it is where cases are won.

Pillar 3 — Notice-and-comment rulemaking and how to participate effectively

Section 553 imposes a deceptively simple sequence: publish a notice of proposed rulemaking in the Federal Register; give interested persons an opportunity to comment; and, when adopting the final rule, include a concise general statement of its basis and purpose. Layered on top are the Regulatory Flexibility Act (5 U.S.C. §§ 601–612, small-entity analysis), OIRA review under Executive Order 12866 for significant rules, and the Congressional Review Act (5 U.S.C. §§ 801–808), which gives Congress a window — measured in session and legislative days, not calendar days, so compute it carefully — to disapprove a rule by joint resolution.

Judicial gloss has done the real work.

  • Logical outgrowth. A final rule must be a logical outgrowth of the proposal; if the agency changes course so sharply that commenters had no fair chance to address the final version, the rule fails for inadequate notice. See Long Island Care at Home, Ltd. v. Coke (2007).
  • Disclose the data. An agency must reveal the technical studies and data on which it relies, so that comments can engage with them — the rule of Portland Cement Ass'n v. Ruckelshaus (D.C. Cir. 1973).
  • Respond to significant comments. The basis-and-purpose statement must engage the material objections. Silence is a State Farm problem.
  • But courts cannot invent procedure. Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519 (1978), forbids courts from imposing procedural requirements beyond those the APA or the organic statute impose.

Why the comment record is the whole ballgame. Judicial review is review on the administrative record. Arguments not raised in comments are generally forfeited; evidence not in the record generally cannot be added later. The comment you write in month three determines the brief you can write in year two.

Calder's trade association files a comment on the proposed chemical cap. A weak comment says the rule is "burdensome and unnecessary." A strong comment does four things: it identifies the specific statutory provision the agency lacks authority to stretch (setting up Pillar 7); it submits the cost data and engineering alternatives the agency failed to consider (setting up State Farm); it flags the small-entity impacts the RFA analysis missed; and it proposes a concrete, less burdensome alternative that the agency must now address or explain away. Calder's counsel treats the comment as the opening brief of the eventual lawsuit, because that is what it is.

For the mechanics of finding the proposal, the docket, and the authorities that govern it, start with Finding the Law: A Beginner's Guide to Legal Research. The FTC's ill-fated non-compete rule is a case study in the entire lifecycle — proposal, record, and a district court setting it aside — traced in The Future of Non-Competes: FTC Rulemaking and the State-Law Patchwork.

Pillar 4 — Guidance documents and subregulatory action

Section 553(b)(A) exempts interpretive rules, general statements of policy, and rules of agency organization, procedure, or practice from notice and comment. That exemption is the single most heavily litigated seam in administrative law, because agencies constantly issue documents — guidance, FAQs, Dear Colleague letters, compliance manuals, no-action positions — that function like binding rules but were never proposed for comment.

The doctrinal line: a legislative rule creates new rights or duties and has the force of law; an interpretive rule merely explains what the agency thinks an existing statute or rule already requires. If a document is legislative in substance, calling it "guidance" does not save it, and it must be vacated for failure to observe § 553 procedure (§ 706(2)(D)). Conversely, Perez v. Mortgage Bankers Ass'n, 575 U.S. 92 (2015), held that an agency may amend or repeal a genuinely interpretive rule without notice and comment — burying the D.C. Circuit's contrary Paralyzed Veterans doctrine.

How agencies' own regulations get read. Auer v. Robbins, 519 U.S. 452 (1997), gave agencies deference in interpreting their own ambiguous regulations. Kisor v. Wilkie, 588 U.S. 558 (2019), left Auer standing but caged it: the regulation must be genuinely ambiguous after exhausting all the ordinary tools of construction; the reading must be reasonable; it must be the agency's authoritative and considered position, implicate its substantive expertise, and reflect its fair and considered judgment — not a convenient litigating position or an unfair surprise. In practice, a lawyer who takes Kisor's steps seriously usually finds the regulation is not ambiguous at all.

Calder's problem is a familiar one: an agency FAQ, posted without process, declares that a class of Calder's sensors falls inside a controlled category. No rule was amended. No comment was taken. Inspectors nonetheless treat the FAQ as law. Whether Calder can sue depends on finality (Pillar 9); whether it wins depends on whether the FAQ is legislative in substance (this pillar) and whether the underlying statute bears the agency's reading (Pillar 7). Compliance programs live and die on this seam — see The Corporate Compliance Toolkit and, for a regime that runs almost entirely on subregulatory interpretation, Healthcare Compliance: A Comprehensive Guide to HIPAA, Stark, and the Anti-Kickback Statute.

Pillar 5 — Agency adjudication, ALJs, and the jury right after Jarkesy

Formal adjudication (§§ 554, 556–557) is a trial without an Article III judge: an ALJ presides, the agency's enforcement staff prosecutes, the ALJ issues an initial decision, and the agency heads — the same commissioners who authorized the case — review it. The obvious structural discomfort of that arrangement is the engine of modern doctrine.

The 2024 earthquake. In SEC v. Jarkesy, 603 U.S. 109 (2024), the Supreme Court held that when the SEC seeks civil penalties for securities fraud, the defendant has a Seventh Amendment right to a jury trial in an Article III court; the "public rights" exception did not cover a claim that replicates common-law fraud. The practical consequence: agencies seeking punitive monetary penalties for common-law-analogue claims must generally go to federal court. The consequence lawyers must not overstate: Jarkesy did not abolish agency adjudication. Benefits determinations, licensing, immigration proceedings, and public-rights matters remain in-house. The question in any given case is whether the claim and the remedy look like a common-law action for damages.

Due process floor. Even in informal adjudication, the Constitution supplies a floor. Goldberg v. Kelly, 397 U.S. 254 (1970), and the balancing test of Mathews v. Eldridge, 424 U.S. 319 (1976) — private interest, risk of erroneous deprivation and value of added procedure, government interest — govern how much process is due.

Where to see it in practice. The SEC's administrative and federal-court tracks are mapped in SEC Enforcement: A Comprehensive Guide to Investigations, Wells Notices, and Remedies and The Securities Litigation and Enforcement Toolkit. The tax system's parallel universe — examination, Appeals, and a specialized Article I court — is walked in Handling an IRS Audit and Tax Controversy: A Comprehensive Guide. And the largest adjudicatory system in the federal government, immigration court, gets its own map in The Removal Defense and Humanitarian Relief Toolkit and The Immigration and Citizenship Toolkit.

Pillar 6 — Agency investigations and enforcement

Before there is an order, there is an investigation. Agency subpoenas and civil investigative demands are broadly enforceable: under United States v. Morton Salt Co., 338 U.S. 632 (1950), an agency may investigate "merely on suspicion that the law is being violated, or even just because it wants assurance that it is not," so long as the inquiry is within the agency's authority, the demand is not indefinite, and the information sought is reasonably relevant. The realistic objections are relevance, burden, privilege, and scope — not "you have no basis."

The mirror-image doctrine is prosecutorial discretion. Heckler v. Chaney, 470 U.S. 821 (1985), holds that an agency's decision not to enforce is presumptively committed to agency discretion by law and therefore unreviewable under § 701(a)(2) — a presumption reinforced on standing grounds in United States v. Texas (2023). The corollary is § 706(1): a court may compel agency action "unlawfully withheld or unreasonably delayed," but Norton v. Southern Utah Wilderness Alliance, 542 U.S. 55 (2004), confines that to discrete actions the agency is legally required to take. You cannot use § 706(1) to run an agency.

When the subpoena lands at Calder's front desk, the first hour matters more than the next six months: preserve documents, scope the hold, run privilege, and decide who speaks. Work it with Responding to a Government Subpoena or Investigation: A Practical Checklist.

Pillar 7 — Judicial review: the standards after Loper Bright

Section 706 is the heart of the APA. It directs the reviewing court to decide all relevant questions of law and to set aside agency action that is (A) arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law; (B) contrary to constitutional right; (C) in excess of statutory authority; (D) without observance of procedure required by law; (E) unsupported by substantial evidence in formal proceedings; or (F) unwarranted by the facts on de novo review.

Questions of law: independent judgment. Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024), overruled Chevron and held that § 706's command to "decide all relevant questions of law" means what it says. Courts do not defer to an agency's interpretation of an ambiguous statute. Ambiguity is not a delegation; it is a problem for the court to solve using the traditional tools of construction. Two important qualifiers survive. First, where a statute expressly delegates discretionary authority to the agency (for example, to set a standard that is "appropriate" or to fill a gap the statute identifies), the court's role is to police the boundaries of that delegated discretion — not to substitute its own policy. Second, an agency's longstanding, contemporaneous, and well-reasoned interpretation may still command respect under Skidmore v. Swift & Co., 323 U.S. 134 (1944), in proportion to "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." Skidmore is persuasion, not deference. Do not brief it as a softer Chevron.

Questions of policy: arbitrary and capricious. Motor Vehicle Manufacturers Ass'n v. State Farm Mutual Automobile Insurance Co., 463 U.S. 29 (1983), remains the standard and, post-Loper Bright, is where most cases are actually decided. An agency must examine the relevant data and articulate a satisfactory explanation, including a rational connection between the facts found and the choice made. A rule is arbitrary if the agency relied on factors Congress did not intend it to consider, entirely failed to consider an important aspect of the problem, offered an explanation that runs counter to the evidence, or is so implausible that it could not be ascribed to a difference in view or agency expertise. Related lines: a change of policy is permitted but the agency must acknowledge that it is changing course and give good reasons (FCC v. Fox Television Stations, 556 U.S. 502 (2009)), and must account for serious reliance interests (Encino Motorcars, LLC v. Navarro (2016); DHS v. Regents of the University of California (2020)). A pretextual explanation — a stated reason that is not the real reason — will not do (Department of Commerce v. New York (2019)). And in Ohio v. EPA (2024) — a stay ruling, so read it for its reasoning rather than as a final merits holding — the Court treated EPA's failure to address a significant comment as a likely reasoned-decisionmaking failure rather than a harmless one.

Questions of fact. In formal proceedings, the standard is substantial evidence on the record considered as a whole — evidence a reasonable mind might accept as adequate, evaluated against the contrary evidence too (Universal Camera Corp. v. NLRB, 340 U.S. 474 (1951)). Review is confined to the administrative record the agency actually had before it (Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402 (1971)); extra-record evidence and post hoc rationalizations by counsel are out.

The practical shift. Loper Bright did not make agencies easy to beat. It moved the fight. The best challenges now pair a textual argument (the statute simply does not authorize this) with a record argument (and in any event the agency did not reason its way there). Calder's brief against the chemical cap runs on both rails at once. To convert this doctrine into a filing, use Challenging Agency Action Under the APA: A Practical Checklist; to place the case within federal procedure generally, see Federal Civil Litigation: A Roadmap and Resource Guide and, because APA cases are often resolved on cross-motions on the record, Moving for Summary Judgment Under Rule 56: A Practical Checklist. For petitions for review taken straight to the courts of appeals under an organic statute, Federal Civil Appeals: A Comprehensive Guide is the procedural companion. And because this area is moving fast, verify every proposition against current law using Case Law Research: Finding, Reading, and Verifying Judicial Decisions.

Pillar 8 — The major questions doctrine

Some assertions of agency power are so consequential that ordinary interpretation is not enough. Under the major questions doctrine, when an agency claims authority over a matter of "vast economic and political significance," courts require clear congressional authorization — and will not find it in a long-extant but rarely used statutory provision. West Virginia v. EPA, 597 U.S. 697 (2022), invalidating the Clean Power Plan's generation-shifting approach, is the doctrine's fullest statement. Its antecedent is the "elephants in mouseholes" principle of Whitman: Congress does not hide fundamental changes in ancillary provisions.

The doctrine has real teeth: National Federation of Independent Business v. OSHA (2022) stayed OSHA's vaccine-or-test emergency temporary standard; Alabama Ass'n of Realtors v. HHS (2021) rejected the CDC eviction moratorium; Biden v. Nebraska (2023) rejected mass student-loan cancellation under the HEROES Act.

How to use it — and how not to. The doctrine is not a general "this rule is big" objection. It bites hardest when several markers converge: an unheralded or transformative reading of an old statute; a subject matter outside the agency's traditional expertise; a claim of power Congress has repeatedly considered and declined to grant; and enormous economic stakes. Note also the relationship to Loper Bright: with Chevron gone, the doctrine is less a deference-override and more a clear-statement canon operating inside ordinary de novo interpretation. Calder's chemical cap is expensive, but expensive is not "major" — the agency has regulated this pollutant under this provision for thirty years. Calder's better argument is textual and State Farm, not major questions, and a lawyer who leads with a weak major-questions claim spends credibility she needs elsewhere.

Pillar 9 — Justiciability: standing, finality, ripeness, exhaustion, limitations

More APA cases die on timing and gatekeeping than on the merits. Five gates:

Standing. Lujan v. Defenders of Wildlife, 504 U.S. 555 (1992): injury in fact that is concrete, particularized, and actual or imminent; causation; redressability. Regulated parties usually have easy standing. Regulatory beneficiaries — competitors, neighbors, associations — must work for it, and the associational-standing and procedural-injury doctrines matter enormously.

Final agency action. Section 704 permits review of "final agency action." Bennett v. Spear, 520 U.S. 154 (1997), supplies the two-part test: the action must (1) mark the consummation of the agency's decisionmaking process, not be tentative or interlocutory, and (2) be one from which legal consequences will flow or rights or obligations are determined. This is the doctrine that decides whether Calder can sue over the FAQ from Pillar 4. If the FAQ is genuinely advisory, it is unreviewable; if the agency treats it as binding and inspectors enforce it, Calder has a strong finality argument.

Ripeness. Abbott Laboratories v. Gardner, 387 U.S. 136 (1967): weigh the fitness of the issues for judicial decision against the hardship to the parties of withholding review. A purely legal challenge to a rule that forces an immediate and costly compliance choice is ripe on the day the rule issues.

Exhaustion. Do not assume you must exhaust. Darby v. Cisneros, 509 U.S. 137 (1993), holds that under the APA, exhaustion of intra-agency appeals is required only when the statute or an agency rule expressly requires it and makes the action inoperative pending appeal. Many organic statutes do impose exhaustion and issue-exhaustion requirements — but check, do not guess. Relatedly, arguments not presented to the agency during the comment period or the hearing are typically forfeited on review; this is the second reason the comment record from Pillar 3 is worth doing properly.

Limitations. The default federal limitations period for civil actions against the United States is six years under 28 U.S.C. § 2401(a). Corner Post, Inc. v. Board of Governors of the Federal Reserve System, 603 U.S. 799 (2024), held that an APA claim does not accrue until the plaintiff is injured by the final agency action — not when the rule was promulgated. A company formed in 2024 may therefore challenge a 1990s rule. Combined with Loper Bright, Corner Post has meaningfully reopened old regulations to fresh attack. Be alert to organic statutes with their own, far shorter, review windows (many petition-for-review provisions run 60 days), which override the default.

Pillar 10 — Remedies: vacatur, remand, and interim relief

Section 706(2) instructs the reviewing court to "hold unlawful and set aside" unlawful agency action. The conventional reading is that a successful APA challenge produces vacatur of the rule or order — a remedy that runs to the rule itself, not merely to the parties. That reading is contested: some Justices and commentators argue § 706 authorizes only party-specific relief, and the Court's decision limiting universal injunctions in Trump v. CASA, Inc. (2025) sharpened the debate without squarely resolving the status of APA vacatur. Brief the remedy, do not assume it. Ask for vacatur, and plead a fallback.

Remand without vacatur. Where the agency's error may be curable and vacatur would be seriously disruptive, courts — following the D.C. Circuit's Allied-Signal, Inc. v. U.S. Nuclear Regulatory Commission, 988 F.2d 146 (D.C. Cir. 1993), framework — may remand while leaving the rule in place, weighing the seriousness of the deficiency against the consequences of vacating. Expect the agency to ask for this. Expect to have to argue against it on both prongs.

Interim relief. Section 705 authorizes the agency or the reviewing court to postpone the effective date of agency action pending review to prevent irreparable injury. In parallel, ordinary preliminary-injunction standards apply. And when the agency action has already caused loss, the remedies question becomes one of sovereign immunity and statutory waiver — the terrain of Suing the Government: A Guide to Damages in Administrative and Public-Entity Litigation. Note the APA's own limit: § 702 waives sovereign immunity for suits seeking relief other than money damages. The APA gets you the rule vacated; it does not get you paid.

Pillar 11 — The agency-specific regimes across this library

Administrative law is a grammar. Here is the vocabulary, agency by agency, with the map for each.

Agency / regime Core statutory hook Where to go in UpLaw
SEC Securities Act; Exchange Act; APA §§ 554–557; Jarkesy SEC Enforcement; The Securities Litigation and Enforcement Toolkit
FTC / DOJ Antitrust FTC Act § 5; Sherman & Clayton Acts The Antitrust and Competition Toolkit
EPA CAA, CWA, CERCLA, RCRA; permits and NOVs The Environmental Law Toolkit
IRS Internal Revenue Code; Treasury regulations; Tax Court Handling an IRS Audit and Tax Controversy
DOL FLSA; regulations at 29 C.F.R. Part 541 Wage and Hour Law Under the FLSA; The Employment Law Toolkit
USCIS / EOIR INA; immigration court adjudication The Removal Defense and Humanitarian Relief Toolkit; Employment-Based Immigration
CBP / BIS / OFAC Tariff Act; EAR; ITAR; IEEPA The International Trade and Sanctions Toolkit
CPSC Consumer Product Safety Act; recalls (federal) — but the underlying products liability is state tort law Managing a Product Recall: A Practical Checklist; The Tort and Products Liability Toolkit
HHS / OCR HIPAA; Stark; Anti-Kickback Statute Healthcare Compliance
CFPB / FTC (consumer) FDCPA, FCRA, TCPA, state UDAP Consumer Protection Law; The Consumer Protection and Credit Toolkit
State & local land use boards Zoning enabling acts; variance standards (state law) The Land Use and Development Toolkit
Privacy & AI regulators FTC Act § 5; state privacy statutes The Privacy and Data Protection Toolkit; AI Governance and Compliance

Two warnings before you use that table.

First — several of these regimes are not federal at all, and the table's federal grammar does not reach them. Tort and products liability, land use and zoning, and insurance are overwhelmingly state law, and they genuinely diverge from state to state. There is no nationwide rule of design defect (some states apply a consumer-expectations test, some a risk-utility test, some a hybrid); no nationwide rule of comparative versus contributory fault; no nationwide variance standard; and no nationwide standard of review for a zoning board's decision. The federal agencies in the table (CPSC, for example) sit alongside that state law, not on top of it: a CPSC recall is a federal administrative act, but the lawsuit that follows is a state-law tort claim tried under state substantive rules. Insurance is regulated primarily by the states under the McCarran-Ferguson Act, 15 U.S.C. §§ 1011–1015, and by state insurance commissioners applying state codes. Whenever this toolkit's federal framework meets one of those areas, stop and confirm the rule of the state you are actually in.

Second — several of these regimes run on numbers and lists that change constantly, and nothing in this toolkit should be relied on for a current figure. Hart-Scott-Rodino reporting thresholds and filing-fee tiers are adjusted annually; tariff rates, duty classifications, exclusions, and trade-remedy orders change with administrations, proclamations, and investigations, sometimes within weeks; and OFAC's Specially Designated Nationals list, BIS's Entity List, and the sanctions programs behind them are amended on a rolling basis, often without warning. Verify every threshold, rate, and designation against the agency's current published source on the day you rely on it. A stale HSR threshold, a stale tariff line, or a stale screening list is not a technicality — it is a filing failure, a customs penalty, or a sanctions violation. See The Antitrust and Competition Toolkit and The International Trade and Sanctions Toolkit for the framework, but never for the number.

A note on state administrative law. Most states have their own APA, many drawing on the Model State Administrative Procedure Act, and many have not followed Loper Bright — a number of state courts still apply deference doctrines of their own, some have narrowed them, and a few have abolished them by statute or constitutional amendment. Zoning boards, licensing boards, insurance commissioners, and state environmental agencies run on state administrative law. Never assume the federal framework transfers. Confirm your state's APA, its standard of review, its exhaustion rules, and its (often very short) appeal deadlines.

Worked example: Calder Instruments, start to finish

Pull the threads together. Here is how the eleven pillars actually sequence in one matter.

Month 1 — the proposal. An agency publishes an NPRM capping the concentration of a solvent Calder uses on its coating line. The economic analysis assumes a substitute chemical that Calder's engineers know fails at the temperatures its medical sensors require. Calder's counsel does not treat this as a compliance problem. She treats it as the first day of the lawsuit (Pillar 3).

Months 2–4 — building the record. Calder and its trade association file a comment that (1) argues the organic statute authorizes limits on emissions, not on input concentrations, and quotes the operative words; (2) submits bench data showing the assumed substitute's failure mode, with the underlying test protocol; (3) proposes a concrete alternative — a capture-and-destroy control at the vent — with costed engineering; and (4) identifies the small suppliers the Regulatory Flexibility Act analysis omitted. Every argument Calder may want in year two is in the comment in month three, because arguments not presented to the agency are generally forfeited (Pillars 3, 9).

Month 9 — the final rule. The agency finalizes. It never engages the bench data. Its basis-and-purpose statement waves at the proposed alternative in a footnote. It also quietly narrows an exemption in a way no one could have anticipated from the proposal.

Month 9, week 2 — characterize and choose the forum. The rule is a legislative rule (general applicability, future effect, force of law), so this is § 553 territory and § 706 review (Pillars 2, 7). Counsel checks the organic statute: does it provide a direct petition for review in the court of appeals on a short clock, or does Calder sue in district court under the APA? This question is answered before anything is drafted, because getting it wrong is usually fatal (Pillar 9). Assume here that the statute is silent and district court it is.

The claims. Calder pleads on three rails:

  • § 706(2)(C) — in excess of statutory authority. The statute reaches emissions; the rule reaches inputs. Under Loper Bright, the court decides this itself, de novo, using the ordinary tools of construction. No Chevron step two. The agency's long-held contrary view gets whatever Skidmore respect its reasoning earns — which, given that the agency changed positions in 2019, may be very little.
  • § 706(2)(D) — procedure. The narrowed exemption is not a logical outgrowth of the proposal; commenters had no fair chance at it. And the agency relied on a modeling study it never disclosed (Portland Cement).
  • § 706(2)(A) — arbitrary and capricious. Under State Farm, the agency failed to consider an important aspect of the problem (the substitute does not work), failed to address a significant alternative squarely presented, and offered an explanation that runs counter to the record evidence.

What Calder does not plead. Major questions. The agency has regulated this solvent under this provision for thirty years; the rule is expensive but not transformative, and no marker converges (Pillar 8). Leading with a weak major-questions claim would cost credibility on the three claims that can actually win.

Meanwhile — the FAQ. A different office of the same agency posts an FAQ declaring that Calder's sensor line falls inside a controlled category. Inspectors cite it. Counsel asks the two questions that decide it: is it final agency action under Bennett v. Spear — consummation plus legal consequences (Pillar 9) — and is it legislative in substance despite the label, such that it needed notice and comment (Pillar 4)? If inspectors are enforcing it, the finality argument is strong and the FAQ is vulnerable.

And the enforcement action. The agency's enforcement staff opens an investigation and issues a CID. Calder's first hour: hold, scope, custodians, privilege (Pillar 6). If the agency later seeks civil penalties on a claim that replicates a common-law action, Jarkesy is on the table (Pillar 5) — and Axon means Calder can take a structural challenge to district court now, not after two years in the agency's own tribunal (Pillar 1).

Remedy. Calder asks for vacatur of the rule under § 706(2), a § 705 stay of the compliance date (which bites in eleven months), and — anticipating the agency's request — briefs both Allied-Signal prongs against remand without vacatur (Pillar 10). It does not ask for damages: § 702 waives immunity only for relief other than money damages.

That is administrative law. One statute, one record, one set of gates, applied over and over.

The landscape as a working checklist

Run these gates on any agency matter. Each is gated: if the answer is wrong, stop and fix it before moving on.

Characterize the action

  • [ ] Is this a rule or an order? Rulemaking or adjudication? Formal or informal?
  • [ ] If the agency calls it "guidance," is it legislative in substance — does it create rights or duties, or bind the agency's own discretion?
  • [ ] What is the organic statute, and what does it actually authorize? Read it before you read the regulation.

Participate before you litigate

  • [ ] Did we comment during the NPRM window, and did the comment raise every argument we may later need?
  • [ ] Did we put our data, cost evidence, and less-burdensome alternatives into the record?
  • [ ] Did we preserve statutory-authority, procedural, and reliance-interest objections in terms?

Investigation and enforcement posture

  • [ ] Litigation hold issued; scope, custodians, and privilege screen defined.
  • [ ] Subpoena/CID assessed for authority, relevance, burden, and privilege — not "no basis."
  • [ ] Structural constitutional defenses (appointments, removal, jury right) identified — and Axon considered for an immediate district-court challenge.
  • [ ] If the agency seeks civil penalties for a common-law-analogue claim, is there a Jarkesy jury-right objection?

Gatekeeping

  • [ ] Standing under Lujan: injury, causation, redressability — pleaded with facts, not adjectives.
  • [ ] Final agency action under Bennett v. Spear: consummation + legal consequences.
  • [ ] Ripeness under Abbott Labs: fitness + hardship.
  • [ ] Exhaustion: required by statute or rule? (Darby) Issue-exhaustion satisfied?
  • [ ] Limitations: 28 U.S.C. § 2401(a) six years, accruing on injury (Corner Post) — unless an organic statute imposes a shorter petition-for-review window. Calendar it today.
  • [ ] Correct forum: district court under the APA, or a direct petition for review in the court of appeals?

Merits

  • [ ] Statutory authority (§ 706(2)(C)): independent judgment under Loper Bright; no Chevron. Is there an express delegation, and did the agency stay inside it?
  • [ ] Skidmore respect assessed honestly — consistency, contemporaneity, thoroughness, persuasiveness.
  • [ ] Major questions: only if the markers genuinely converge.
  • [ ] Procedure (§ 706(2)(D)): notice, logical outgrowth, data disclosure, response to significant comments.
  • [ ] Reasoned decisionmaking (§ 706(2)(A)) under State Farm: important aspects considered, alternatives addressed, policy change acknowledged (Fox), reliance interests weighed (Encino), no pretext.
  • [ ] Facts: substantial evidence on the whole record (Universal Camera); record-bound review (Overton Park); no post hoc rationalization.

Remedy

  • [ ] Ask for vacatur under § 706(2); brief it, do not assume it.
  • [ ] Anticipate a remand-without-vacatur request and argue both Allied-Signal prongs.
  • [ ] Seek a § 705 stay or preliminary injunction if the compliance date bites first.
  • [ ] Remember § 702: the APA waives immunity for relief other than money damages. Damages need a different statute.

State matters

  • [ ] Which state APA governs? What is the standard of review, and does the state still defer to agencies (many do; Loper Bright does not bind them)?
  • [ ] What is the appeal deadline? State review windows are often measured in weeks, and they vary widely — look it up, do not assume.
  • [ ] If the matter touches tort, products liability, land use/zoning, or insurance, identify the governing state rule. There is no nationwide answer.

Currency

  • [ ] Any threshold, rate, or list in play — HSR reporting thresholds, tariff rates and classifications, OFAC/BIS designations — verified against the agency's current published source today, not from a memo or a toolkit.
  • [ ] Every case checked for subsequent history. Chevron is overruled; make sure nothing in your brief still leans on it, directly or through a case that applied it.

Primary Authority

Cited by name and holding, without pinpoint reporter pages: Lucia v. SEC (2018); Seila Law LLC v. CFPB (2020); Axon Enterprise, Inc. v. FTC (2023); Gundy v. United States (2019); Department of Commerce v. New York (2019); DHS v. Regents of the University of California (2020); Encino Motorcars, LLC v. Navarro (2016); Long Island Care at Home, Ltd. v. Coke (2007); Ohio v. EPA (2024); Biden v. Nebraska (2023); NFIB v. OSHA (2022); Alabama Ass'n of Realtors v. HHS (2021); United States v. Texas (2023); Trump v. CASA, Inc. (2025); Portland Cement Ass'n v. Ruckelshaus (D.C. Cir. 1973); Allied-Signal, Inc. v. U.S. Nuclear Regulatory Commission (D.C. Cir. 1993). Verify current status and exact reporter pages before filing; this area is moving quickly, and several of the 2024–2025 decisions above are still being worked out in the lower courts.

Secondary Sources

  • Kenneth Culp Davis & Richard J. Pierce, Administrative Law Treatise — the standard reference.
  • Richard J. Pierce, Jr., Sidney A. Shapiro & Paul R. Verkuil, Administrative Law and Process.
  • Model State Administrative Procedure Act — the baseline for many state APAs (adoption and amendment vary widely by state).
  • Attorney General's Manual on the Administrative Procedure Act (1947) — the contemporaneous construction courts still cite.
  • The Federal Register and the Unified Agenda of Regulatory and Deregulatory Actions — where rules are proposed, finalized, and forecast.

Related UpLaw Resources

Related Guides & Articles

Related Checklists

Related Toolkits

This toolkit is general legal information from UpLaw (uplaw.us), not legal advice, and does not create an attorney-client relationship. Laws vary by state and change over time; verify current requirements and consult qualified counsel before acting.