Conservative legal thought has always been attracted to formalisms, and there is nothing inherently wrong with that. Formalism can be a useful and appropriate tool. Formalism can have a logic and even an elegance to it, as in Chief Justice Burger’s insistence in Chadha on fidelity to the constitution’s “finely wrought and exhaustively considered procedures” for legislating.

Lately, however, conservative legal thought and jurisprudence has taken a turn toward facile formalism. Facile formalism is not anchored in constitutional text, like the legislative procedures Chief Justice Burger rhapsodizes. Facile formalism is based on abstractions not grounded in law or logic. Facile formalism is a thin veil to cloak the exercise of raw power to impose normative preferences. Collins v. Mnuchin underscores the pitfalls of facile formalisms and why they ultimately fail.

Facile formalism has a long pedigree, dating at least to the Court’s ill-fated attempts to curb Congress’ power to enact economic regulation under the Commerce Clause by maintaining a “zone of interests” that excluded from the definition of commerce patently commercial activities like production, manufacturing, and mining. The Court drew this line to advance its own substantive vision of federalism, which preserved a regulatory sandbox for states to play in, safe from incursion by the federal government. But, of course, this doctrinal barrier crumbled in the face of the incontrovertible reality of an interconnected, national economy.

Facile formalism has returned in recent years as a device to curb contemporary efforts by the federal government to regulate the national economy. Its latest guise: simple counting. Basic arithmetic is now the weapon of choice to eviscerate the administrative state. The Court previewed the counting pretense in Free Enterprise Fund v. PCAOB, where it found a dual layer of removal protection to unduly interfere with the President’s ability to take care that the laws be faithfully executed. The Court insisted that two layers were too many. Yet, it proceeded to impose a remedy that had no functional effect on the President’s ability to control PCAOB Members. This was sufficient, according to the Court, because one is less than two.

Since Free Enterprise Fund, facile formalism has seeped into the law and policy of regulatory review as a tool to curb agencies in their exercise of statutorily delegated power. President Trump’s “2-for-1” Executive Order and OMB implementing guidance require agencies to repeal two regulations for every new one they promulgate. This Order summarily discarded utility maximization as the lodestar of regulatory decision-making and replaced it with crude regulation counting, which is riddled with pathologies that I detail in this article. The upshot is that agencies may now forgo painstaking calculation of the relative costs and benefits of their policies as long as they know how to subtract two.

Seila Law v. Consumer Financial Protection Bureau elevated counting to the status of constitutional first principles. There, the Court found unconstitutional the good cause removal provision protecting the Director of the Consumer Financial Protection Bureau (CFPB) from being fired at the whim of the President because one CFPB Director is less than five Federal Trade Commissioners (whose identical good cause removal protections were upheld in Humphrey’s Executor). Literally, the decision rests on a formalism that facile. In dissent, Justice Kagan chides her colleagues for their Schoolhouse Rock vision of the separation of powers, but their analysis is more Sesame Street than Schoolhouse Rock.

The Court employs this counting device to advance its vision of the unitary executive. The Court invokes the “Take Care” clause, the Article II vesting clause, and some anodyne quotes from Madison on the separation of powers to argue that the Founders intended to give the President unfettered removal power over executive branch officials even though they declined to say so in the Constitution. Strong tea from those inclined to ridicule others for finding constitutional rights “hiding” in or “emanating” from the Constitution. The Seila Law majority then go on to characterize cases that cut against its newly minted “general rule” as narrow exceptions that must be strictly confined to their facts. Thus, although the CFPB director is subject to the exact same statutory removal restriction as the FTC Commissioners protected in Humphrey’s, that restriction is unconstitutional in Seila Law because there is only one of her and there are five of them.

Others have thoroughly documented the perverse results wrought by unfettered presidential removal power—how it is more likely to lead to incompetence, corruption, and cronyism than to the accountability imagined by unitary executive theorists—so I will make only two brief points on this score here. First, one need look no further than the rising COVID body count and the inept and hyper-politicized federal response to the disease to doubt the wisdom of further politicizing independent, expert agencies statutorily charged with protecting the public health and welfare. Second, the dream of political accountability fades when we wake to a president who has cynically sought to manipulate executive agency decision-making to hide his own wrongdoing and his administration’s policy failures from the electorate. It is hard to fathom that this is what Madison had in mind.

To be fair, these are functional objections to the Court’s formalist turn, and there can be good reasons for maintaining formalist distinctions. Perhaps the best justification for formalism is that it provides clear, predictable doctrinal application. But the recent argument in Collins v. Mnuchin reveals that the Court’s facile formalism does not even deliver that. So, while we’re counting, here’s another number: three. That is the number of different tests applied by the justices and the litigants during oral argument in Collins. One would think that Seila Law’s general rule + narrow exceptions test would easily decide Collins. After all, there is only one director of the Federal Housing Finance Authority (FHFA), and that director is protected by a good cause removal restriction like the one struck down in Seila Law. One equals one, therefore the removal restriction must go.

But wait! Does one really equal one? To figure this out, litigants and justices reached back to employ the functional executive interference test from Morrison v. Olsen, debating just how much control the President actually has over the FHFA director and whether that amount of control is sufficient to allow the President to carry out his Article II duties. Many reached even further back to the categorical test that prevailed before Morrison, asking whether the FHFA director was exercising executive power—or some other type of power—when he placed Fannie Mae and Freddie Mac into conservatorship. Not only were there three different tests in play, but justices mixed formalism and functionalism with reckless abandon in applying these tests. For instance, Justice Sotomayor pointed out that, as a formal matter, conservatorship is a judicial function. And Justice Alito could not resist the functionalist retort that the FHFA must be exercising executive power because of the significant effects its actions have on the economy.

The Court will have its hands full untangling all of this. However, much of the argument suggested that the justices might be looking for ways to skirt hard constitutional questions of their own making—perhaps deciding the case on narrower statutory grounds. That won’t make the hard questions go away. It will simply create a pall of uncertainty over regulatory decision-making that taxes agencies and creates litigating opportunities for members of the elite Supreme Court bar and their disappointed corporate clients. Just as the Founders intended.

Jodi L. Short is the Associate Dean for Research and the Honorable Roger J. Traynor Professor of Law at UC Hastings College of the Law.