The Supreme Court often must step in when one branch of government tries to take on the role of another—but what happens if a branch actually tries to give away its own power? It’s a problem that turns on its head the Framers’ assumption that every branch will jealously guard its own power, an assumption which undergirds our constitutional checks and balances. Enter the nondelegation doctrine: the principle that Congress can’t delegate its own constitutionally-granted Article I power to another branch. This type of delegation has become a common problem as it allows legislators to pass off many tough decisions and sausage-factory compromises to the administrative state. They can take credit for passing legislation to achieve nice bottom-line goals like “clean air” without having to get into the nitty gritty of the tradeoffs necessary to bring that into effect—and without suffering electoral consequences for those key decisions of governing. Of course, electoral accountability is a key feature of our constitutional system, hence the importance of the nondelegation doctrine and the need for the Supreme Court to give it teeth.
Seven years ago, in Gundy v. United States, a fragmented Court narrowly rejected a nondelegation challenge to a federal statute, with no opinion commanding a majority. At the time, I pointed out how four justices (with a fifth, Justice Kavanaugh, not participating) were on record as prepared to breathe new life into the doctrine in an appropriate case. Last term’s FCC v. Consumers’ Research (2025), while initially appearing to be such a case, turned out to be a missed opportunity—albeit a speed bump rather than a roadblock to the cause of restoring a proper separation of powers.
Slated for the Supreme Court’s conference tomorrow is a cert petition that could finally bring about something not witnessed at the high court in over 90 years: a ruling grounded in the nondelegation doctrine. The case is RMS of Georgia, LLC v. Environmental Protection Agency, and it raises the following question: Did Congress violate Article I’s Vesting Clause when it handed the EPA free rein to decide which private companies get to participate in a multibillion-dollar market—and which get shut out—without supplying so much as a hint of how to make that call?
In 2020, Congress passed the American Innovation and Manufacturing (AIM) Act to phase down hydrofluorocarbons (HFCs)—the refrigerants in your air conditioner—by 85% by 2036. Fine; that’s a real policy choice, made by elected legislators, just as Article I contemplates. The Act even gave the EPA some direction on a tiny slice of the program: For the first five years, the EPA could allocate special allowances for a handful of statutorily defined “essential uses” if it made certain enumerated findings, and it had to set aside allowances for a short list of specialized applications—a category the EPA itself calculated at just 2% of the total. But on the question that actually matters—who gets the remaining 98% of allowances, the ones that determine whether a company can keep producing or importing HFCs at all—Congress said nothing. Not “fair and equitable,” not “reasonably necessary or appropriate,” not in the “public interest”—nothing. As the petition puts it, “[t]here is not one word in the AIM Act addressing the all-important question of how to decide who should get them.”
The EPA noticed the void and filled it with whatever it wanted. The agency openly touted its “considerable” and “significant discretion,” floated allocating allowances based on whether applicants were “minority- and woman-owned small businesses” facing “systemic racism” or “other challenges,” set aside allowances “for new market entrants,” and ultimately built its allocation scheme around data from an unrelated greenhouse-gas reporting program that was never designed to measure market share. Choice Refrigerants, a small Georgia company that holds patents on its own HFC blends, paid the price: It lost roughly 30% of the allowances it would have received under genuine market-share allocation, while a Chinese-backed competitor that was copying Choice’s patented products walked away with allowances of its own.
Existing Supreme Court precedent permits federal statutes to delegate power to the executive branch as long as Congress sets out an “intelligible principle” to guide its exercise of authority. Justice Gorsuch’s Gundy and Consumers’ Research dissents have criticized the Court for watering down that standard into what he called in the latter case the “milquetoast version of the intelligible principle test.” Notably in the pending case, when the EPA was challenged in the D.C. Circuit, the court did not dare say the AIM Act clears even the lax version of that test on its own terms—because it doesn’t. Instead, the panel manufactured a principle out of another law, declaring that Congress must have implicitly intended the AIM Act to borrow a market-share formula from Title VI of the Clean Air Act. Its principal evidence: a few stray remarks at a subcommittee hearing and the fact that both statutes happen to use “baseline years” to set their caps. Never mind that the AIM Act expressly incorporates other Title VI provisions by name and conspicuously leaves this one out—an omission courts are not supposed to wave away as accidental. And never mind that the EPA itself did not consider itself bound by (and did not follow) Title VI’s market-share formula.
A court that invents the guardrails Congress declined to write does not solve an Article I problem. It just commits a breach of its own. As the cert petition puts it, to do that is to “layer an Article III transgression on top of the Article I transgression. . . . Two separation-of-powers wrongs do not make a constitutional right.” Why would an agency ever bother articulating real constraints on its own discretion if a sympathetic court will invent some for it after the fact?
The nondelegation doctrine presents an important federalism consideration, articulated in the amicus brief of 21 states led by West Virginia: “[A]n amorphous ‘intelligible principles’ standard strips States of their power to hold lawmakers accountable and robs them of their experimental character, too.” For 98% of what it regulates, the AIM Act provides no standard at all. It would not pass muster under even the prevailing watered-down notion of an “intelligible principle.” Having protected so many other aspects of the separation of powers, the Court should grant the petition in RMS of Georgia to establish that the nondelegation doctrine deserves the same vigilance.