Administrative Law: Rulemaking, Adjudication, and Judicial Review

Summary. How agencies make rules, decide cases, and get reviewed — and what has changed about deference.

Part I: Where most law actually lives

Congress passes a few hundred laws a year. Agencies issue thousands of rules, and decide millions of cases. If you have applied for a benefit, held a license, built something, run a business, or received a notice from a government office with an acronym, you have been in an administrative proceeding — whether or not anyone used the phrase.

The governing framework is the Administrative Procedure Act, enacted in 1946 and remarkably durable. It defines the vocabulary at 5 U.S.C. § 551, prescribes how rules are made at 5 U.S.C. § 553, how cases are decided at 5 U.S.C. § 554 and § 556, and how courts review both at 5 U.S.C. § 706.

Two categories, and everything follows from which one you are in:

Rulemaking is legislative in character: prospective, general, aimed at the future conduct of a class.
Adjudication is judicial in character: retrospective, particular, aimed at the rights of identified parties.

States have their own administrative procedure acts, most modeled on the federal one, and most of what follows has a state analogue. But the details differ, and a state proceeding runs on state rules.

Part II: Rulemaking

The ordinary path: notice and comment

5 U.S.C. § 553 prescribes three steps that look modest and generate enormous consequence:

Notice of proposed rulemaking, published in the Federal Register, stating the time, place, and nature of the proceedings, the legal authority, and either the terms of the proposed rule or a description of the subjects and issues involved.
Opportunity for interested persons to comment — to submit written data, views, or arguments.
A concise general statement of basis and purpose accompanying the final rule.

That third item is where the real work happens. "Concise" has proven aspirational: preambles routinely run hundreds of pages, because the agency must show it considered the significant comments and responded to them. A rule whose preamble ignores a serious objection is a rule at risk.

Exemptions. Section 553 excepts interpretative rules, general statements of policy, and rules of agency organization, procedure, or practice; matters relating to military or foreign affairs functions; and matters relating to agency management, personnel, public property, loans, grants, benefits, or contracts. It also permits skipping notice and comment for good cause where it is impracticable, unnecessary, or contrary to the public interest — an exception that is invoked frequently and policed skeptically.

The legislative/interpretative line is one of the most litigated distinctions in the field, because an agency that can characterize a document as "guidance" avoids the entire process. The general test asks whether the document has binding legal effect or merely explains what the agency already thinks the law requires.

Two doctrines that decide rulemaking challenges

Logical outgrowth. A final rule must be a logical outgrowth of the proposal. If the agency proposes one thing and finalizes something materially different, commenters had no fair opportunity to address it, and the rule fails for inadequate notice.

No judicially imposed extra procedures. In Vermont Yankee Nuclear Power Corp. v. NRDC, the Supreme Court held that courts may not require agencies to use procedures beyond those the APA, the organic statute, or the Constitution demand. Reviewing courts police the substance of the reasoning, not the procedural generosity of the process.

Arbitrary and capricious review, and the hard look

Under 5 U.S.C. § 706, a court sets aside agency action that is arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.

The canonical formulation comes from Motor Vehicle Manufacturers Association v. State Farm: 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.

Two corollaries that decide cases:

Post hoc rationalizations do not count. The agency is judged on the reasons it gave when it acted, not the reasons its lawyers construct later. Citizens to Preserve Overton Park v. Volpe established that review is based on the full administrative record before the decision-maker at the time.
Changing course requires an explanation. An agency may change its policy, but it must acknowledge that it is changing, and give a reasoned explanation — with more required where the prior policy engendered serious reliance interests or rested on factual findings that contradict the new ones.

The record is the case. In administrative litigation there is generally no discovery, no live testimony, and no new evidence. There is a certified administrative record and briefs about whether it supports what the agency did. Everything a challenger wants a court to see must be put into the record during the agency proceeding. This single fact should reorganize how anyone participates in a rulemaking.

Part III: Adjudication

Formal adjudication under 5 U.S.C. §§ 554, 556, and 557 applies where the organic statute requires a hearing "on the record." It looks like a trial: notice, an impartial presiding officer, the right to present evidence and cross-examine, a decision on the record, and a prohibition on ex parte communications and on the combination of prosecutorial and adjudicative functions in the same person.

Informal adjudication is everything else — and it is the overwhelming majority of what agencies do. A benefits determination, a permit decision, a licensing action. The APA supplies little procedure directly; the requirements come from the organic statute, the agency's regulations, and due process.

The due process framework is Mathews v. Eldridge: the private interest affected; the risk of erroneous deprivation through the procedures used and the value of additional safeguards; and the government's interest, including fiscal and administrative burdens. Goldberg v. Kelly had earlier required a pre-termination evidentiary hearing for welfare benefits, and the two cases together mark the range: sometimes a hearing must come first, sometimes a prompt one afterward suffices.

Whether due process applies at all depends on a property or liberty interest. Board of Regents v. Roth held that a property interest requires a legitimate claim of entitlement, not a unilateral expectation, created by an independent source such as state law — and Perry v. Sindermann held that such an entitlement can arise from an implied understanding as well as an express contract.

Bias and the combination of functions. Withrow v. Larkin held that combining investigative and adjudicative functions in one agency does not, without more, violate due process — a presumption of honesty and integrity attaches to administrative adjudicators. But Gibson v. Berryhill found a constitutional violation where board members had a pecuniary interest in the outcome, and that principle is very much alive in professional licensing.

Substantial evidence is the standard for review of findings in formal proceedings. Universal Camera Corp. v. NLRB instructs that the reviewing court must consider the whole record, including evidence detracting from the agency's conclusion — deferential, but not a rubber stamp.

Part IV: Deference — what changed

For forty years, the central question in administrative law was how much a court must defer to an agency's reading of the statute it administers.

Chevron U.S.A. Inc. v. NRDC supplied the answer: if the statute was silent or ambiguous on the precise question, a court deferred to any permissible agency construction. It was cited tens of thousands of times.

Loper Bright Enterprises v. Raimondo overruled it. Courts must exercise independent judgment in deciding whether an agency has acted within its statutory authority, and may not defer merely because a statute is ambiguous. The APA's instruction that the reviewing court "shall decide all relevant questions of law" means what it says.

What survives:

Skidmore respect — an agency's interpretation may be persuasive to the extent of its thoroughness, consistency, and validity of reasoning. Persuasion, not deference.
Express delegations. Where Congress explicitly gives an agency authority to define a term or fill a gap, the court's role is to police the boundaries of that delegated discretion, not to substitute its own view within it.
Prior holdings. Loper Bright stated that it did not call into question prior cases that relied on the Chevron framework; those holdings remain subject to statutory stare decisis.

Deference to an agency's reading of its own regulation is a separate doctrine. Auer v. Robbins established it; Kisor v. Wilkie preserved it while imposing serious limits: the regulation must be genuinely ambiguous after exhausting all the traditional tools of construction; the reading must be reasonable; and it must be the agency's authoritative position, implicate its substantive expertise, and reflect its fair and considered judgment rather than a convenient litigating position or an unfair surprise.

And the major questions doctrine. West Virginia v. EPA crystallized the principle that in extraordinary cases — where an agency claims authority of vast economic and political significance — courts will hesitate before reading a general or ancillary statutory provision to confer it, and will demand clear congressional authorization. It is not a canon of ambiguity resolution; it is a threshold skepticism about the claim of power itself.

Part V: Getting into court

Four threshold barriers stop more challenges than the merits ever do.

Standing. Lujan v. Defenders of Wildlife requires an injury in fact that is concrete and particularized and actual or imminent; causation traceable to the challenged action; and redressability. A generalized grievance about government following the law is not enough.

Final agency action. 5 U.S.C. § 704 makes final agency action reviewable, and Bennett v. Spear supplies the two-part test: the action must mark the consummation of the agency's decision-making process, and must be one by which rights or obligations have been determined or from which legal consequences will flow.

Ripeness. Abbott Laboratories v. Gardner asks whether the issues are fit for judicial decision and what hardship withholding review would cause the parties. A purely legal challenge to a regulation with immediate compliance consequences is generally ripe.

Exhaustion. Where the statute or a regulation requires it, administrative remedies must be exhausted before suit. Where the requirement is judge-made, exceptions exist — futility, irreparable harm, a challenge to the agency's jurisdiction. When in doubt, exhaust. More claims die on this than on any doctrine in the field.

The right of review comes from 5 U.S.C. § 702, and 5 U.S.C. § 701 carves out actions committed to agency discretion by law — a narrow exception applying where there is no law to apply. Relief pending review is available under 5 U.S.C. § 705.

And a recent change to timing. Corner Post, Inc. v. Board of Governors held that an APA claim accrues when the plaintiff is injured by final agency action, not when the rule was promulgated — meaning a party that did not exist when a rule issued may still challenge it. This materially expanded the window for facial challenges to old rules.

Part VI: Structural constitutional pressure

Three lines of cases are reshaping agencies from the outside.

Removal. Free Enterprise Fund v. PCAOB held that two layers of for-cause removal protection unconstitutionally insulated officers from presidential control. Seila Law LLC v. CFPB held that a single-director independent agency head removable only for cause violated the separation of powers — while severing the removal provision and leaving the agency standing.

Agency adjudication and the jury right. SEC v. Jarkesy held that where the SEC seeks civil penalties for fraud, the Seventh Amendment entitles the defendant to a jury trial in an Article III court, because the action is legal in nature and does not fall within the public rights exception. The implications for in-house enforcement adjudication across agencies are still being worked out, and the argument is now standard in any agency penalty proceeding.

Nondelegation and the major questions doctrine operate as companion constraints on how much authority a general statute can be read to confer.

The practical upshot for a practitioner: structural arguments that were exotic a decade ago are now routine, and preserving them costs little. Raise them early and specifically, because they are frequently waived by silence.

Part VII: The information statutes

The Freedom of Information Act, at 5 U.S.C. § 552, requires agencies to publish certain materials, make others available, and release records on request unless a statutory exemption applies — nine of them, covering classified material, internal personnel rules, information exempt under other statutes, trade secrets and confidential commercial information, privileged inter- and intra-agency memoranda, personal privacy, law enforcement records, financial institution examinations, and geological well data.

FOIA is a practical tool, not just a transparency principle. It produces the agency's own analysis, its correspondence, its guidance to staff, and the complaints of others — material that frequently reframes a dispute. Deadlines apply, fee waivers exist for requests in the public interest, and denials are administratively appealable and then reviewable in court de novo, with the burden on the agency.

Related statutes: the Privacy Act governs records about individuals; the Government in the Sunshine Act opens certain meetings; the Federal Advisory Committee Act governs outside advisory bodies; the Congressional Review Act permits legislative disapproval of rules; and the Regulatory Flexibility Act requires analysis of impacts on small entities. States have parallel open records and open meetings laws, often broader and faster than the federal ones.

Part VIII: Four encounters with an agency

Doctrine becomes concrete when a person is standing in it. Here are four common encounters and what the framework above actually requires.

A small manufacturer facing a new rule

A proposed rule would require emissions controls that would cost this company more than its annual profit. The comment period is sixty days.

What matters is that the comment period is the entire case. After the rule is final, review will be on the record the agency compiled — and a court will ask whether the agency considered an important aspect of the problem, in the language of State Farm. An aspect nobody raised is very hard to call important later.

So the comment should:

Submit data. Actual cost figures, engineering assessments, and the compliance timeline this facility would need. Agencies must respond to significant comments, and significance is measured in evidence.
Address the alternatives — including one the agency did not propose, with its own supporting analysis. An agency that fails to consider a reasonable alternative presented to it has a problem.
Raise every legal objection that might later be litigated: statutory authority, the adequacy of the notice, the Regulatory Flexibility Act analysis for small entities, and — if the claim of power is sweeping — the major questions point from West Virginia v. EPA.
Argue the statute independently. After Loper Bright, the agency's reading is not entitled to deference merely because the statute is ambiguous; say what the statute means and why.
Be filed on time. Late comments need not be considered at all.

And use FOIA early. A request under 5 U.S.C. § 552 for the technical support documents, the underlying studies, and communications with outside parties frequently produces the analytical weakness that a comment can exploit.

A licensee facing revocation

A state board notifies a professional of proceedings to revoke a license.

Start with the interest. A license is a property interest under Board of Regents v. Roth because state law creates a legitimate claim of entitlement to keep it absent cause. Due process therefore applies, and Mathews v. Eldridge supplies the calculus for how much.

Then read the agency's own procedural regulations, which will supply far more specific rights than the constitutional floor — discovery, subpoenas, prehearing conferences, the standard of proof, and the deadline to request a hearing.

Then consider the composition of the board. Withrow v. Larkin means that a board investigating and then deciding is not, by itself, a due process violation. But Gibson v. Berryhill means that board members with a pecuniary interest in the outcome — competitors, in the classic case — present a real constitutional problem. Ask who sits on the board and what they do for a living.

And a summary suspension before any hearing is governed by Barry v. Barchi, which upheld an interim suspension but required a prompt post-suspension hearing. If the suspension has been in place for months without a hearing, that is an argument.

A benefits claimant on the fourth level of appeal

A disability claimant has been denied, reconsidered, heard by an administrative law judge, and denied again.

Two things govern the next step.

First, the standard. Judicial review of the factual findings is for substantial evidence, and Universal Camera requires the reviewing court to weigh the whole record, including evidence that detracts from the agency's conclusion. That is deferential but not empty: an ALJ who credited one report and never mentioned three contrary ones has a substantial evidence problem.

Second, the record. New evidence is generally unavailable on review. Everything must have gone in below — which is why the counsel-at-the-hearing question matters so much in benefits systems. See Social Security Disability and Veterans Benefits.

And exhaustion. Where the statute makes administrative exhaustion a condition of jurisdiction, skipping a level ends the case regardless of its merits.

A company facing an in-house enforcement proceeding

An agency brings an administrative enforcement action seeking civil penalties, to be tried before the agency's own administrative law judge.

Three arguments now belong in the answer, and all are cheap to preserve:

The jury right. SEC v. Jarkesy held that an action for civil penalties resembling common law fraud is legal in nature and, absent a public rights exception, must be tried to a jury in an Article III court.
Removal protections. Free Enterprise Fund and Seila Law frame challenges to the insulation of the adjudicator and the agency head.
Appointments. Whether the ALJ was properly appointed under the Appointments Clause.

Preserve them explicitly and early. These arguments are routinely forfeited by silence, and they are worth little on appeal if raised for the first time there.

Part IX: State administrative law, which is where most people actually are

Nearly everything above has a state analogue, and most people's actual encounters with government are state or local: a driver's license, a professional board, a zoning board, a benefits agency, a school district, a licensing authority, a utility commission.

The structure is usually familiar — most states adopted some version of the Model State Administrative Procedure Act — but four differences matter enormously in practice.

1. Review is often broader. Many states permit a trial de novo or something close to it in certain proceedings, or allow the reviewing court to take additional evidence in defined circumstances. Some apply a "substantial evidence on the whole record" standard with noticeably more bite than federal practice.

2. Deadlines are shorter and often jurisdictional. A thirty-day window to petition for review, running from the mailing of the decision, is common — and in many states missing it cannot be cured. Find the deadline in the organic statute before doing anything else.

3. Local boards are staffed by part-time citizens. A zoning board, a licensing board, or a school board may be composed of volunteers applying rules they have not read closely, with a lawyer present only sometimes. This makes the record even more important, because it may be thin, and it makes courteous, well-documented advocacy unusually effective.

4. Open records and open meetings laws are often broader and faster than the federal ones, with shorter response deadlines and stronger enforcement. They are the fastest route to the agency's own analysis. Ask for the staff report, the correspondence, the prior decisions in comparable cases, and the minutes.

One more state-specific note: many states have a catch-all writ — certiorari, mandamus, or a statutory petition for review — that reaches decisions no specific review statute covers. If a decision seems unreviewable, ask about the writ before concluding there is no remedy.

Part X: The economics of an agency fight

Administrative litigation has a different cost structure from ordinary civil litigation, and it favors preparation over combat.

There is usually no discovery. No depositions, no interrogatories, no document requests against the agency. The record is compiled by the agency and certified. This makes the case cheaper — and it makes the agency-stage work vastly more important, because it is the only chance to create facts.

FOIA and state open records requests are the substitute for discovery, and they should be filed early, since the response can take months.

Fee-shifting exists. The Equal Access to Justice Act permits a prevailing party against the United States to recover fees where the government's position was not substantially justified, subject to net worth eligibility limits and rate caps. Many states have analogues. This meaningfully changes the calculus for small businesses and individuals, and it is frequently overlooked.

Interim relief matters more than usual, because agency processes are slow. 5 U.S.C. § 705 authorizes a reviewing court to postpone the effective date of agency action or preserve status quo pending review — and agencies themselves can stay their own actions. Ask the agency for a stay first; it is faster, cheaper, and sometimes granted.

And the realistic remedy is usually a remand, not a reversal. A successful arbitrary-and-capricious challenge typically produces a vacatur and a remand for further explanation — after which the agency may reach the same result with better reasoning. That is not nothing: remand costs time, changes political circumstances, and sometimes produces a different outcome. But a client who expects a court to substitute its judgment for the agency's has misunderstood the remedy.

Part XI: How to actually participate

In a rulemaking:

Comment substantively and on time. Late comments need not be considered.
Raise every objection you might later litigate. Issues not raised are generally forfeited.
Submit data, not adjectives. The agency must respond to significant comments, and significance is measured by evidentiary weight.
Address the alternatives the agency considered and the ones it did not.
Build the record, because the record is the case.

In an adjudication:

Read the agency's own procedural regulations, which supply most of the rights the APA does not.
Get everything into the record. New evidence is generally unavailable on review.
Object contemporaneously and specifically.
Watch the deadlines, which are short and jurisdictional in many schemes.
Exhaust, unless a recognized exception plainly applies.

On review:

Confirm final agency action, standing, ripeness, and exhaustion before anything else.
Find the venue and the deadline — many statutes route review directly to a court of appeals on a short clock.
Frame the argument around the reasoning: what the agency failed to consider, what it failed to explain, what it changed without acknowledging.
Argue the statute independently. After Loper Bright, the question is what the law means, not whether the agency's reading was permissible.

Part XII: What the shift in deference actually changed

The overruling of Chevron was reported as an earthquake. Whether it is one depends on where you are standing, and it is worth being precise about what moved.

What did not change. Review of agency fact-finding is unchanged — substantial evidence in formal proceedings, and the whole-record approach of Universal Camera. Review of agency policy judgments is unchanged — the arbitrary and capricious standard of State Farm, which was always the more frequently dispositive doctrine. Most agency actions were never decided on Chevron grounds, and those cases look the same today.

What changed. Where a case genuinely turns on what a statute means, the court now decides. An agency that previously prevailed because its reading was "permissible" must now show it is correct, or at least persuasive under Skidmore. Three practical consequences follow:

1. Statutory argument is worth more. Text, structure, history, and the traditional canons now do the work that a deference argument used to short-circuit. Briefs look different: longer on the statute, shorter on the standard of review.

2. Agency consistency matters more. Skidmore respect turns on thoroughness, consistency over time, and the validity of the reasoning. An agency that has said three different things about the same provision has less to offer a court than one that has been consistent for forty years — the reverse of the old regime, where a new reading could displace an old one so long as both were permissible.

3. Old rules are more vulnerable, and for longer. Corner Post held that an APA claim accrues when the plaintiff is injured, not when the rule issued — so a business formed last year may challenge a rule from decades ago. Combined with independent judgment on statutory meaning, this expands the field of contestable regulation considerably.

What is uncertain. How much Skidmore respect actually amounts to in practice; how far the express-delegation carve-out reaches; how courts will treat the many prior decisions that upheld agency readings under Chevron, which retain stare decisis effect but rest on a repudiated framework; and how the major questions doctrine interacts with independent judgment now that ambiguity no longer routes to the agency by default.

And a note on where deference still lives. Kisor v. Wilkie preserved deference to an agency's reading of its own regulation, but with conditions that are genuinely demanding — genuine ambiguity after exhausting the interpretive tools, reasonableness, an authoritative position implicating substantive expertise, and fair and considered judgment. Work through the Kisor steps explicitly when an agency invokes its own regulation; the doctrine now has more off-ramps than applications.

Part XIII: A defense of the whole enterprise, and an honest account of its costs

It is fashionable to describe the administrative state either as a usurpation or as the only thing standing between the public and catastrophe. Both accounts miss what is actually going on, and a practitioner is better served by the unromantic version.

Why agencies exist. Congress cannot write the rules for drug approval, aircraft airworthiness, water quality, bank capital, and food safety, and it cannot adjudicate millions of benefits claims. It writes statutes at a level of generality that requires elaboration, and delegates the elaboration to bodies with technical staff and continuous attention. The alternative is not a world of legislative specificity; it is a world in which the rules are made by nobody and enforced by no one.

What that costs. Four things, and they are real.

Diffuse accountability. A rule that affects millions is written by people no one elected, subject to review by a court applying a deferential standard. That is a genuine democratic cost, and the recent structural decisions — Seila Law, West Virginia v. EPA, Loper Bright — are responses to it, whatever one thinks of their reasoning.

Asymmetric participation. Notice and comment is open to everyone and used effectively by almost no one. Sophisticated, well-resourced parties file thousand-page comments with retained experts; the people most affected by a benefits rule generally file nothing. The procedure is neutral; the capacity to use it is not.

Combined functions. An agency that writes the rule, investigates the violation, prosecutes the case, and decides it — before an adjudicator it employs — presents a structural problem that Withrow v. Larkin tolerated and that SEC v. Jarkesy has begun to reopen.

Time. Agency processes are slow, and delay is itself a decision. A permit that takes six years, a benefits appeal that takes three, a rule that takes a decade — each of these allocates outcomes as surely as a ruling does.

What follows for someone in the system. Not cynicism, and not deference. The APA's procedures are genuinely available and genuinely effective when used properly: a well-supported comment does change rules; a well-built record does win appeals; a FOIA request does produce documents that reframe disputes. The people who do badly in administrative proceedings are overwhelmingly the ones who treated the agency stage as a formality on the way to a court that would look at everything fresh. That court does not exist. The agency stage is the case, and acting on that single insight is worth more than any doctrine in this article.

Part XIV: A vocabulary, decoded

Administrative practice runs on terms that mean something specific and are used loosely everywhere else. A short glossary, because misusing them costs credibility and sometimes rights.

Organic statute — the law creating the agency and granting its authority. Read it before the regulations; it is the source of every power and every limit, and after Loper Bright it is the battlefield.

Legislative rule — a rule with binding legal effect, requiring notice and comment. Interpretative rule — an agency's statement of what it thinks an existing requirement already means; no notice and comment required, and no independent binding force. General statement of policy — a statement of how the agency intends to exercise discretion, which by definition cannot bind. The labels are not controlling; legal effect is.

Final agency action — the consummation of the decision-making process from which rights, obligations, or legal consequences flow, per Bennett v. Spear. A preliminary determination, an inspection report, or a warning letter usually is not.

Administrative record — everything the decision-maker considered, certified by the agency. Not everything in the agency's files, and not everything the agency should have considered. Completing the record — supplementing it with material improperly omitted — is its own litigated question.

Remand without vacatur — a court finds a defect but leaves the rule in place while the agency fixes it. Common, and a reason not to promise a client that winning means the rule goes away.

Ex parte communication — a communication with the decision-maker about the merits, off the record, without notice to other parties. Prohibited in formal adjudication; more permissive in informal rulemaking, with disclosure obligations in some contexts.

Exhaustion vs. finality vs. ripeness — three distinct doctrines routinely conflated. Exhaustion asks whether you used the agency's process; finality asks whether the agency is done; ripeness asks whether the issue is fit for a court now.

Standing — your own injury, traceable and redressable, per Lujan. Not a synonym for "affected" or "interested."

Substantial evidence vs. arbitrary and capricious — the first reviews facts in formal proceedings; the second reviews reasoning and policy choices. In practice the two converge, but using the wrong term in a brief signals that you have not read 5 U.S.C. § 706.

Committed to agency discretion by law — the narrow exception in 5 U.S.C. § 701, applying only where there is no law to apply. Agencies claim it far more often than it applies.

Part XV: Five habits of people who win agency cases

They treat the agency stage as the trial. Every fact, every document, every objection goes in below, because the reviewing court will see the record and nothing else.
They read the agency's own regulations before the statute's generalities and before the case law. Most procedural rights in an administrative proceeding come from the agency's rules, and most are waived by not asking.
They file information requests early. 5 U.S.C. § 552 and state open records laws are the substitute for discovery, and the response takes months, so the request goes out in week one.
They preserve everything, specifically. Statutory authority, notice adequacy, the alternatives not considered, the reliance interests ignored, the constitutional structure of the adjudicator. Preservation is nearly free; waiver is total.
They ask what remedy they actually want. A vacatur, a remand, a stay under 5 U.S.C. § 705, a different license condition, or simply an explanation on the record. The answer shapes the argument, and a party that has not decided will not get any of them.

Frequently asked questions

What is the difference between a rule and a guidance document? Legal effect. A legislative rule binds and requires notice and comment; genuine guidance explains and does not.

Is Chevron still good law? No. Loper Bright overruled it. Courts now exercise independent judgment on statutory meaning, with Skidmore respect for persuasive agency reasoning.

Can an agency change its mind? Yes, but it must acknowledge the change and explain it — more thoroughly where there are serious reliance interests.

Can I introduce new evidence in court? Almost never. Review is on the administrative record. Build it below.

What is the major questions doctrine? In extraordinary cases involving vast economic and political significance, courts require clear congressional authorization before reading a general provision to confer the claimed power.

Do I have to exhaust administrative remedies? Where a statute or regulation requires it, yes. Where the requirement is judge-made, exceptions exist — but exhaust unless one plainly applies.