In 2017, Congress used the Congressional Review Act to disapprove a sprawling Federal Communications Commission rule on broadband privacy. Seven years later, when the FCC updated its rules requiring telecom carriers to notify customers of data breaches, industry groups objected. Because one portion of the rule had similarly addressed data breaches, they argued, the 2017 CRA resolution disallowed the new rule.

Last week, the full U.S. Court of Appeals for the Sixth Circuit heard argument in that case, Ohio Telecom Association v. FCC, after a panel of the court ruled for the agency last year. If the en banc court agrees with petitioners’ claims that the 2017 resolution bars the 2024 rule, the consequences could reach far beyond the telecom industry.

The Congressional Review Act gives Congress a fast-track process to veto recent agency rules. Though the 1996 law was once obscure, Congress has used it expansively in recent years, with CRA resolutions comprising nearly 20% of the laws passed by the current Congress. To prevent agencies from simply reissuing what Congress rejected, the CRA bars an agency from adopting a new rule that is “substantially the same” as a disapproved rule. A key question in Ohio Telecom is how to apply the “substantially the same” test.

Until this case, just about everyone had applied that test the same way: by comparing the agency’s entire new rule to its entire disapproved rule. Agencies under both parties have done so, finding new rules permissible even when some provisions carried over from the disapproved rule. The only other court to touch the issue, the U.S. Court of Appeals for the Ninth Circuit, did the same. And Congress, too, has understood the CRA to apply to rules in their entirety, not to individual pieces of them.

That approach makes sense. Large agency rules, including the 2016 broadband-privacy rule whose disapproval is at issue in Ohio Telecom, often contain dozens or hundreds of provisions.   When Congress rejects a rule under the CRA, it is rejecting the entire package, not objecting to every line in it independently. An omnibus rule likely contains provisions that no member of Congress objected to, or perhaps even noticed. An amicus brief from Public Citizen laid this out nicely, explaining how the phrase “substantially the same” calls for a narrow test.

The Ohio Telecom petitioners’ reading would change all that. If every provision of a disapproved rule (and anything substantially the same) were permanently off limits, CRA votes would significantly amend the agency’s authority by preventing the agency from reissuing hundreds of provisions, from the controversial to the mundane—even though CRA resolutions do not amend the text of the U.S. Code or any enacted law.

The petitioners’ argument rests on an isolated reading of a few words. The Administrative Procedure Act defines a “rule” to include “the whole or a part” of a qualifying agency statement, and the CRA borrows that definition. So, petitioners say, every part of a disapproved rule is itself a disapproved “rule” that can never be reissued.

That reading appears clever, but it collapses in context, as an amicus brief that I co-authored on behalf of NYU Law’s Institute for Policy Integrity details. The full text of the “substantially the same” provision shows that it uses the disapproved rule alone as its benchmark. When the CRA refers to part of a rule, it says so expressly. And applying the petitioners’ definition consistently across the statute would produce absurd results: agencies would have to decide whether each individual provision is a “major rule,” and the Comptroller General would have to issue dozens of separate reports for a single regulation. No court, agency, or act of Congress has ever read the CRA that way.

The petitioners’ reading would also make it very difficult to understand an agency’s authority following a CRA resolution. If that reading were adopted, entities could no longer rely on the statute to find the limits on the agency’s power, but rather would need to look through hundreds of provisions from disapproved rules. That’s quite a conclusion to draw from a terse resolution that does not amend the text of the U.S. Code or any enacted law, and it would stand in severe tension with the Supreme Court’s instruction that Congress “does not alter the fundamental details of a regulatory scheme in vague terms.” If this interpretation were adopted, Congress could use the CRA to legislate in cryptic and unrecognizable ways.

The petitioners’ reading would also tie agencies in knots, limiting their authority in arcane and bizarre ways. Agencies would have to make significant changes to uncontroversial or housekeeping provisions from disapproved omnibus rules, including definitional provisions. For example, under the first Trump Administration, the Department of Labor expanded the list of occupations for which drug testing could be used to assess eligibility for unemployment benefits, after Congress disapproved the Obama Administration’s prior rule. Although some legislators said they disapproved the Obama-era rule because it was too limited, the petitioners’ reading would have barred the Department of Labor from relisting any of that regulation’s occupations when it issued a new rule—the opposite of legislators’ stated intent.

Nor would the effect of petitioners’ broad reading necessarily be deregulatory, contrary to what petitioners’ counsel indicated at last week’s oral argument. For instance, in 2021, Congress used the CRA to disapprove the Environmental Protection Agency’s 2020 rollback of methane standards, thereby restoring the agency’s 2016 standards. Under petitioners’ theory, EPA presumably could not remove any of the obligations or provisions of the 2016 rule that Congress’s disapproval restored, even if conditions significantly evolved.

The strongest argument supporting petitioners’ theory is that without it, petitioners argue, an agency can circumvent a CRA resolution by reissuing a single disapproved rule as multiple parts. But this is far easier said than done, because the parts would need individual justification and would themselves be subject to judicial review and the CRA. Moreover, such obvious circumvention could be policed without wholly reinterpreting the CRA and creating the host of other problems summarized above.

At argument last week, several judges on the Sixth Circuit seemed drawn to the petitioners’ theory. Some questions from the bench floated an even broader reading. For instance, one judge questioned whether a CRA resolution should bar the agency from readopting any statement in the preamble of a disapproved rule, such as uncontested background principles on the agency’s authority. Such a reading would result in agencies losing uncontroversial authorities that Congress, in a terse CRA resolution, never meant to strip. Other judges questioned how narrowly to slice a rulemaking “part,” positing that even a short string of words from the original rule could not be reissued.

The court can avoid these problems by reading the CRA as it has always been read: to bar reissuance only of the whole disapproved rule. As one question at oral argument suggested, that reading fits comfortably with the statutory definition of “rule,” which can be “the whole or a part” of an agency statement—as in either one, but not necessarily both in all cases. In the CRA’s reissuance bar, statutory context makes clear it means the whole. That is the best reading of the statute, and the en banc court should adopt it.

Max Sarinsky is the legal director at the Institute for Policy Integrity at New York University School of Law.