The Administrative Record Is the Whole Case

Coverage of a court striking down a regulation tends to describe it as a ruling against the policy. Usually it is not. It is a ruling that the agency failed to explain its reasoning adequately at the time it acted.

This is the core of arbitrary and capricious review. A court asks whether the agency considered the relevant factors, addressed the significant comments it received, and articulated a rational connection between the facts it found and the choice it made.

Agencies lose these cases most often on the third point, and specifically on the failure to grapple with an alternative that a commenter proposed. The remedy is to send it back, whereupon the agency frequently adopts the same policy with a better explanation, and wins.

This makes the comment process far more consequential than its reputation suggests. A well constructed comment does not persuade an agency. It creates an obligation to respond, and it builds the record on which the eventual case will be decided.

The distinction matters for how these rulings are read, and almost every account of them gets it backwards. A court that vacates a rule has usually said the explanation was inadequate, not that the policy was unlawful.

The standard of review is the source of the confusion. It asks whether agency action was arbitrary and capricious, which sounds like a judgement about substance and operates as a judgement about reasoning.

An agency that reaches the same conclusion with a better explanation generally wins. That is the tell. If the policy itself were the problem, a fuller preamble would not cure it.

So the case is decided on the record, and the record is assembled during the rulemaking rather than during the litigation. By the time a challenge is filed, the material a court will consider is closed.

This is why the comment file carries far more weight than its reputation suggests. A comment that raises a specific alternative, with support, creates something the agency must address in writing or risk losing later.

The obligation is not to agree. It is to engage, and a preamble that ignores a substantial comment is the most common way a rule fails.

The rulemaking section sets the procedure and says remarkably little about the depth of response required, which has been filled in by courts case by case over decades.

The practical result is an arms race in length. Preambles have grown enormous because every paragraph is insurance against a later argument that something went unconsidered.

That has a cost nobody chose. A document written to survive review is not written to be understood, and the published preamble is now frequently longer than the rule it explains and read by almost nobody outside the litigation.

It also advantages the parties who can afford to file sophisticated comments, because a sophisticated comment is what generates the obligation. A member of the public writing in plain language creates far less of one.

None of which argues against the doctrine, which does real work. An agency that must explain itself in advance behaves differently from one that need not, and the discipline is visible in the quality of the reasoning.

The narrower observation is about where the contest happens. It is not in court and not at the announcement. It is in a comment file, months earlier, read by people who knew that was the moment.

There is a fair criticism that this doctrine lets courts substitute their judgement while claiming not to. A judge who finds an explanation inadequate has, in practice, decided something about the policy without saying so.

The answer is that the alternative is worse in both directions. A court that reviewed the merits directly would be legislating, and one that reviewed nothing would leave the most consequential decisions in government entirely unexamined.