The Role of Fungibility in the Spending Clause Challenge to Restrictions on State Aid in the American Rescue Plan Act
by Michael C. Dorf
Section 9901 of the American Rescue Plan Act of 2021 appropriates approximately $220 billion to state, local, territorial, and tribal governments. Most of those funds (over $195 billion) go to states, raising the question whether the conditions placed on how states use those funds are constitutional. Because of a highly dubious precedent of the Supreme Court, there is a substantial possibility that a key condition could be invalidated. Here I'll explain why one of the arguments for the invalidity of the condition is especially bad.
Article I, Section 8 of the Constitution empowers Congress "to lay and collect taxes, duties, imposts and excises, to pay the debts and provide for the common defense and general welfare of the United States." Congress provides for the general welfare by spending the money it raises through taxes, borrowing, and other means. On the face of the constitutional provision, there is no limit to how Congress spends money. Nonetheless, the Court has articulated limits on the Spending power.
Where do those limits come from? The supposedly textualist and originalist conservative Justices of the Supreme Court have discovered within the Constitution's silences an unenumerated prohibition on Congress issuing directives to the legislative and executive branches of state governments. The majority opinion in the second case just linked, authored by self-professed originalist and textualist Justice Scalia, begins its analysis as follows: "Because there is no constitutional text speaking to this precise question, the answer to the [question presented] must be sought in historical understanding and practice, in the structure of the Constitution, and in the jurisprudence of this Court." That's right. The same Justice who would snarkily chastise his colleagues for "inventing" a Constitution when they found individual rights, was perfectly happy to invent states' rights, including today's topic, the so-called anti-commandeering doctrine.
But I digress. What has the anti-commandeering doctrine got to do with the Spending Clause? Well, when Congress spends money through an intermediary such as a state government or private actor, it often attaches conditions to ensure that the money is spent on the projects for which it is intended or to ensure that it does not subsidize bad acts such as illicit discrimination. Thus, for example, Title VI forbids recipients of federal funds from engaging in race discrimination. The Supreme Court has placed limits on the kinds of conditions Congress may attach to federal funds to states. If conditional spending violates those limits, then the conditions become simple commands to the states in violation of the anti-commandeering principle. Thus, although the conditional spending limits were articulated by the Supreme Court before it officially invented the anti-commandeering rule, they effectively pre-suppose such a rule.
The canonical formulation of the limits on conditional spending was articulated in South Dakota v. Dole, where the Court upheld a requirement that states receiving federal highway funds forbid alcohol purchases and public consumption by persons under 21 or else lose a small percentage (in the particular case 5%) of the funds otherwise allocable. The Court there found that the conditions satisfied a four-part test: (1) an essentially toothless requirement that the spending be for the "general welfare"; (2) a clear statement of the condition so that states can decide whether taking the money justifies compliance; (3) a fairly loose requirement that the conditions be germane to the funds' purpose (the looseness illustrated by Dole itself, where the only tie was that both highway funds and the drinking age have some connection to automobile accident prevention); and (4) compliance with any other constitutional provisions (such as the 21st Amendment in Dole). Although not listed in Dole as a separate factor, the Court there also stated a requirement (ostensibly drawn from prior cases) that we might label (5) the financial inducement offered by Congress must not be so great as to pass the point at which "pressure turns into compulsion" or coercion.
Prior to 2012, no major post-Dole case found a violation of any of the foregoing limits, except for the clear-statement requirement, but such decisions can be seen as simple statutory construction. Things changed in the Obamacare case, NFIB v. Sebelius, where a seven-justice majority found a violation of limit (5): the federal government was threatening to withhold so much Medicaid funding that states had effectively no choice but to accept the money and thus comply with the conditions. The Court also thought that there was something problematic about Congress attaching "new" conditions to "old" funds, even though, as Justice Ginsburg powerfully argued in dissent, the prior statute expressly warned of this possibility and Congress could have simply terminated the "old" program and re-enacted it as part of a whole "new" one. Meanwhile, the majority gave no clear account of how big an offer of funds must be to make it too big to refuse (and thus coercive) or even how one determines what denominator to use in deciding that the funds that could be withheld are too big a percentage, i.e., a percentage of what?
Let's put aside the problems with pre-existing doctrine for now. Ohio has sued the Treasury Department, seeking to invalidate a condition on the funds it is slated to receive. In its brief in support of its motion for preliminary injunction, Ohio gestures at an "it's too big" argument, but that seems uncertain. Ohio says that the American Rescue Plan Act would provide 7% of its total annual budget. That's admittedly a lot of money, but not quite as much as the 10% at issue in NFIB, and crucially, a one-time infusion, as opposed to the ongoing payments at issue in NFIB. Accordingly, Ohio places principal reliance on a different argument.
Ohio challenges the provision of American Rescue that it calls the "Tax Mandate." It says that a state may not use American Rescue funds "to either directly or indirectly offset a reduction in the net tax revenue . . . resulting from a change in law, regulation, or administrative interpretation during the covered period . . . ." But, says Ohio, because money is fungible, "any money that a State receives through the Act will necessarily offset, either directly or indirectly, every tax deduction that the State might pursue." And because setting tax rates is a state sovereign prerogative, the "Tax Mandate" is an impermissible condition. Hence, Ohio argues, the condition is invalid as conditional spending and the "Tax Mandate" is impermissible commandeering.
Although I am no fan of the Court's application of the limits on conditional spending in NFIB, I acknowledge that there is something to the notion that Congress should not be permitted to leverage its spending power too far. Cases presenting the problem of so-called "unconstitutional conditions" involve similar issues. For example, most liberals do not like the idea that a legislature might provide funds for public housing but only on condition that residents waive their Fourth Amendment rights against unreasonable searches and seizures; likewise, we don't like the notion that Congress funds medical care for indigent people but only if the doctors forgo the right to give sound medical advice about abortion; etc. The question of where to draw the line between permissible and impermissible conditions is admittedly difficult.
How should the line be drawn? It is generally accepted that the one thing conditional spending limits should not restrict is the ability of the legislature to tell the funding recipient (whether a State or an individual) how to spend the money. Accordingly, one of the most restrictive views of conditional spending in the extant literature was the proposal by Prof Lynn Baker of the University of Texas Law School that Congress should not be permitted to engage in what she called "regulatory spending"--imposing conditions on states it could not impose directly--but should of course be permitted to engage in what she called "reimbursement spending"--i.e., imposing conditions on how states spend the money they receive.
But here's the thing. When Congress tells states that they should use the American Rescue funds to cover pandemic response costs and/or to restore services that were cut due to the financial impact of the pandemic but not simply to pass along the money to taxpayers via tax cuts, that's core reimbursement spending. And the fact that money is fungible is precisely what makes it so. A state should not be permitted to say "hey, we took the money and spent it to administer vaccines, admittedly at the same under-funded level as before we took the money, and we also gave rich people some tax cuts, but those tax cuts involved other money." That line of reasoning doesn't work precisely because money is fungible.
Put differently, Ohio's argument is exactly backwards. It invokes a feature of the "Tax Mandate"--its response to the fungibility of money--as though that feature makes it invalid, whereas in fact fungibility shows that the condition satisfies even the strictest limit on conditional spending.
I'll conclude with a couple of caveats: (1) Numerous other red-state Attorneys General have also threatened to sue the federal government over the condition barring tax cuts funded by American Rescue. Perhaps they will advance other, more persuasive, arguments. (2) As always, the fact that I find an argument to be unpersuasive or even terrible is no guarantee that the Trump-packed federal courts will see it the same way. Indeed, in my darker moments, I sometimes think it suggests the opposite.