Agency Structure Does Not Tell You What Authority a Decision Had

Influence maps.

Budget dashboards.

Organograms with enough dotted lines to suggest a municipal squid has entered the chat.

A comparative-government consultant will rank the Department of Miscellaneous Vital Matters above the Bureau of Slightly Smaller Matters.

Then explain that it has a broad mandate, a cabinet secretary, an impressive headcount, and an annual report with a photograph of a bridge.

Then a letter arrives from the Deputy Assistant Director for Seasonal Permits.

It denies the Mayor of Larchmont’s application to install a ceremonial foghorn near a wetland.

The decision has a docket number, a signature, a date, some attachments, and an opinion about its own importance.

Your organization chart is now standing there in a cardigan, unable to say which provision lets this particular office deny this particular application.

An agency profile answers institutional questions. It doesn’t identify the power asserted for a decision.

The minimum identity block has seven fields.

Canonical name.

Government level.

Legal form.

Parent.

Creation authority.

Current official URL.

Last-verified date.

Governance adds six slots.

Head or board.

Seats.

Appointment authority.

Confirmation requirement.

Term or tenure.

Removal rule.

With quorum or acting-official authority as an optional seventh.

Those facts tell you where a body lives, who nominally steers it, and which drawer holds the family photographs.

They don’t tell you what happened to the foghorn.

The league table arrives early

The powerful-agency ranking is a very handsome useless object.

It combines mandate breadth, political direction, finance, public visibility and decision output into one large number. Then asks it to speak legal Latin.

Agency comparison deserves a more modest six-row matrix.

Scope.

Control.

Decision mode.

Finance.

Publication trail.

Change cost.

Each row needs its own unit, period and source.

“Power” is where those labels go to disappear.

Start with scope, if you actually need a comparison.

It has two ends.

A broad portfolio with several components, and a narrow mission around a defined programme.

GAO’s 2009 inventory counted 15 departments and 88 other executive entities.

That’s a dated classification count.

Not an authority meter.

A broad department can coordinate.

A specialist office can concentrate expertise.

Neither fact tells you whether the Seasonal Permits desk can operate the foghorn provision. Somebody still has to open the statute.

Political structure gets its own filing cabinet.

Compare three levers.

Who appoints leaders.

Who can remove them.

Where the body sits in the executive hierarchy.

A multi-member commission with tenure rules differs structurally from a department headed by a presidential appointee.

Humphrey’s Executor concerns a particular arrangement for an FTC commissioner. Not a stamp that turns every commission into “independent” in the legally interesting sense.

An adjective is not a delegation. It’s what a delegation wears to lunch.

Finance has the same problem with better charts.

The relevant labels are appropriated resources, obligations, outlays and award spending. Four different comparisons, not a single pile called spending.

USAspending reports more than 400 data elements, and its account and award data answer different questions.

An obligation is a binding commitment.

An outlay is a payment.

Neither is a citation to authority.

Your foghorn may consume a tasteful amount of money. An outlay is not a power source, any more than a catering invoice is a zoning ordinance.

A decision is not a small department

Your useful first question isn’t how much agency you can see from space.

It’s what kind of action happened.

Agency output gives you two primary procedural forms. Rulemaking, which creates generally applicable future-facing rules. Adjudication, which resolves individual cases, applications or disputes.

An enforcement action and a refusal to enforce bring related but different review questions. An internal management act may be none of these.

Name the action type before the criticism.

Otherwise you make the ceremonial mistake of calling a rule “a case” because it upset somebody.

Rulemaking has a public procedural shape you can check.

The APA’s default notice-and-comment provision, 5 U.S.C. 553, directs notice, an opportunity for interested people to submit written data, views or arguments, and a concise general statement of basis and purpose with the final rule.

Your working sequence has three stages.

Identify authority.

Give the required notice and participation opportunity. Explain the final rule against the material received.

A comment page full of aggrieved dockside landlords is not your issue.

The question is which required step, source, or explanation is claimed to be absent.

Adjudication hands you a different luggage tag.

Your rule record has four fields.

Source authority.

Docket or regulatory identifier.

Publication citation.

Codified destination.

Your adjudication record has four others.

Case or docket number.

Parties or applicant.

Decision date.

Final disposition.

The Federal Register can add four identifier classes.

Issuing agency.

CFR title and part.

Agency docket or internal file number.

Regulatory Information Number.

The CFR gives you 50 titles, updated through Federal Register amendments. The FederalRegister.gov XML is informational rather than the controlling legal edition.

A web page can be a very efficient way to find a document it cannot finally prove.

And formal procedure doesn’t appear because the foghorn dispute acquired a binder.

The governing statute and the APA decide whether formal process applies.

5 U.S.C. 554 has particular formal-adjudication rules only where the enabling statute triggers them.

For an individual decision, a fair process usually needs four answers.

Proposed action.

Supporting facts or rules.

Response route.

The person with authority to decide after the response.

Mathews v. Eldridge treats adequacy as context-sensitive.

Not a universal pre-deprivation hearing with tiny sandwiches.

Two separate witnesses

Your agency record has five mandate fields.

Statutory purpose.

Affected population or territory.

Delegated functions.

Authorized instruments.

Explicit exclusions or limits.

The enacted text or a reliable codification supports those claims.

A current organization page supports a different proposition for you.

Present programmes, components, operating officials.

A Government Manual-style description may offer purpose, role, legislative or executive authority, programmes, major operating officials and history.

At least five sourceable categories.

None of its pleasant currentness amends the enabling law.

This is where your perfect agency profile becomes dangerous.

A page says the Bureau “administers wetlands permits.” Fine. A current page may identify the Deputy Assistant Director.

Also fine.

Your authority card still asks whether the cited source assigns this decision to the bureau.

Whether a delegation places it with that component. Whether the asserted instrument reaches the decision.

The profile gives you institutional placement. The statute, regulation, delegation or other claimed source gives you the proposed legal bridge.

Don’t use a current web navigation menu to build a bridge over a statutory gap.

It’ll have a lovely hover state and no load rating.

Your currentness needs its own dates too.

Agency history gives you four event types.

Establishment.

Rename.

Transfer.

Termination or merger.

The 2002 DHS transition involved 22 agencies and programmes, and required documentation of financial, logistical and human resources for transfers.

That’s evidence that components, records, assets and personnel can move under you.

It doesn’t silently update the source of authority for a decision the component made after the move.

Your card records the current parent and component, then obtains the relevant creation, transfer or delegation instrument.

The record that survives the meeting

The action-and-authority card starts with one decision. It needs no graph database in a sensible hat.

Retrieve the enabling source, the current organization page, the action type, the responsible component, the action record materials, your review route.

Keep your mandate citation separate from your current organization source.

Then run three passes of your own.

Identity, action, review.

Identity asks whether your legal body, your parent, your component and your current page belong together.

Action asks what was done, and what authority is asserted.

Review asks whether there’s finality, a forum, and a stated ground.

IDENTITY
agency | legal form | parent | creation / transfer source | current component source

ACTION
rulemaking / adjudication / enforcement / non-enforcement | action date
responsible component | asserted statute / regulation / delegation
docket or case ID | publication or decision | record materials

REVIEW
finality | review route | stated ground | source of review rule | limitation

That review line is not a grab bag marked “bad government.”

Your review petition or brief ordinarily separates five questions.

Reviewability.

Finality.

Source of law.

Standard of review.

Requested relief.

Under 5 U.S.C. 706(2), the federal APA lists six set-aside grounds.

Contrary to law.

Contrary to constitutional right.

Beyond statutory authority.

Without required procedure.

Unsupported by substantial evidence where applicable. Unwarranted by facts where trial de novo is authorized.

Your card names one.

“Arbitrary” cannot be a universal condiment.

Your record line is equally literal.

Overton Park requires review to engage the administrative record, rather than post hoc litigation rationalizations.

In covered formal proceedings, 5 U.S.C. 556 places testimony, exhibits and papers on which findings rest in the exclusive record for decision.

Your reasoned-decision criticism can identify four common failures.

An impermissible factor.

An unconsidered important aspect.

A premise contrary to the record.

An implausible explanation.

Four distinct claims.

They don’t get stronger by arriving in a group chat.

Non-enforcement deserves its own box on your card, because it refuses to behave like a failed sanction.

A party subject to enforcement can contest authority, procedure or arbitrariness in its own case.

A complainant seeking enforcement against somebody else may face the Heckler v. Chaney presumption that non-enforcement is unreviewable, where the statute supplies no meaningful judicial standard.

A mandatory command, a binding rule, discriminatory treatment, or a mistaken belief about jurisdiction can change your analysis.

“They did nothing” is a description.

Not yet a review ground.

An authority problem also isn’t your whole lawsuit in a hat.

Under current US federal statutory review, Loper Bright directs a court to exercise independent judgment about the best reading of a statute. Agency technical reasoning may still have persuasive weight.

Kisor gives you a distinct constrained inquiry for an agency’s interpretation of its own regulation.

Your card records which text is being interpreted, before it names any doctrine. Statute, agency regulation, guidance, or order.

Generic deference has a pleasant roomy sound because it has failed to unpack.

And if you establish a defect, the ordinary judicial result has two steps. The action loses its legal support. The matter goes back for lawful reconsideration.

The court normally does not write the agency’s policy on a cocktail napkin.

A renewed decision can consider the missing factor, use the correct interpretation, provide required process, or explain itself using the proper record. And still reach the same ultimate result.

Which is why your card stops before validity, reasonableness, political importance, or every possible limit.

It has established a threshold question.

Not won a parade.

Where the chart stops

The sophisticated ranking can go back to rendering its circles now.

It’s useful for comparing a stated structural axis. Your organization chart can identify the component whose signature sits on a decision.

Neither can tell you what authority that signature had.

For that, take the small action record. Connect your institutional placement to the asserted source.

Identify the decision type and the review route. Leave the conclusion where your record can reach it.

Call it the action-and-authority card.