COVID-19 has (rightly) pushed from the headlines almost every political controversy lacking a coronavirus angle. Up until recently, the debate over so-called “sanctuary jurisdictions” was one such displaced topic. But then the President a few weeks ago, on April 28, identified a link between the two subjects: federal aid to state and local governments—to help replace tax revenues lost because of the pandemic—might be conditioned, the President suggested, on state and local willingness to abandon their “sanctuary” policies and assist the federal government in immigration enforcement by providing information and person power.
Such federal spending conditions, we expect, will never come to pass. The House of Representatives, which will have to agree to the terms of any relief law that attaches strings, would never assent to what the President suggested. Indeed, the House will likely go out of its way to make sure a funding bill cannot be construed so as to give the executive branch any delegated discretion to impose requirements related to sanctuary policies, or others the administration might come up with, as prerequisites to state or local receipt of federal monies.
Still, the President’s offhand remark, however unlikely to ripen into federal policy, does point up how many unanswered fundamental constitutional questions remain in the conditional spending arena. In the space below, we list and briefly describe some of the most important of them.
1. Does the Clear Notice Given to States Have to Emanate From Congress?
Over the past several decades the Supreme Court has consistently required that federal funding conditions be laid out unambiguously so that states are fully aware of the terms of any bargain that is being struck prior to the deal going into effect. But important questions linger about where that clear notice needs to be lodged.
In a prominent recent case (discussed in detail here) on the so-called Byrne-grant program administered by the U.S. Department of Justice (DOJ) to provide funding support for state and local law enforcement, the U.S. Court of Appeals for the Second Circuit said that the clear notice to which states are entitled can come from the Executive branch (prior to the states’ receipt of federal monies). In that case, because the DOJ explicitly told state and local applicants for federal funds the details of the requirements (involving sanctuary policies, interestingly enough) at the time of grant application, states and localities were put on adequate notice of their obligations prior to the making of any promises in exchange for money, and thus could not later complain that they had been misled or duped into the bargain.
But the Second Circuit brushed aside what could be an important countervailing consideration—the fact that while Congress might have indicated that states could be made subject to some conditions, Congress was not clear and unambiguous in its enactment about precisely what kinds of conditions recipient states must obey. The Second Circuit conceded that the notice given to states in the present instance “was provided by DOJ rather than Congress,” but thought that so long as states received clear notice—from a branch of the federal government–before they accepted the funding, the anti-deception reason for a clear-notice requirement was fully satisfied.
Perhaps this reasoning works, but perhaps instead federalism principles do require Congress—rather than the Executive branch pursuant to a broad delegation—to decide on the specific subject matters of the conditions attached to federal grants to states. In an analogous arena, many commentators have argued that executive agencies should not be given broad license to construe federal statutes so as to preempt state laws, because decisions about mandatory preemption should be made by Congress.
One rejoinder might be that because conditional funding (and conditional preemption) involves “deals” between the feds and the states—unlike ordinary preemption, in which states are given no choice in the matter—vagueness by Congress should be curable by the executive branch. In other vagueness contexts (e.g., involving constitutional due process or the First Amendment) judicial interpretations—especially by state courts, which are not encumbered by Article III limits on advisory opinions and the like—of a statute can put individuals on notice as to the meaning of a law and thus avoid the problem of vagueness in future disputes. But we query whether, in the federalism context, where other plain-statement rules, including that announced in 1991 in Gregory v. Ashcroft (which requires Congress to clearly state in the text of a regulatory statute that the statute applies to state and local government entities before states can be required to obey) seem designed to make sure Congress has carefully considered state interests, Congress itself has to be the one to lay out the basic terms of conditional funding deals. The Court hasn’t had occasion to discuss this issue in any depth in the conditional spending doctrine or (what should be treated identically) the conditional preemption arena. For that reason, the Second Circuit, even if it thought there were merit in requiring Congress to reflect on and lay down more precise conditions, might not have believed it had sufficient running room from the Court to fashion new doctrine.
The Supreme Court itself may soon have a chance to address this question (and related matters). Just two days after President Trump’s April 28 musings, the U.S. Court of Appeals for the Seventh Circuit rejected the Second Circuit’s analysis and outcome with regard to Byrne grant funding, thus setting up a conflict within the circuits that the Supreme Court may find necessary or helpful to resolve. If and when the Court takes the case, it should address the question of which federal institution must identify (with a high degree of clarity) the applicable conditions attached to a federal grant to the states. (This question may fit in nicely with the Court’s increasing interest in the so-called non-delegation doctrine; perhaps conditional spending is an arena in which a majority will seek to enforce that doctrine.)
Let us imagine (counterfactually) that the House (perhaps because state and local budget stress were so large and because the President and his party decided to play hardball) were to go along with the President’s desires and explicitly condition federal aid to states on the abandonment of sanctuary policies. There would still be the question whether the federal government as a whole could impose such terms on states. And that raises essential, unanswered questions about other aspects of conditional spending doctrine beyond the clear-notice requirement.
2. Do South Dakota v. Dole and subsequent cases require “germaneness”?
Most constitutional law professors probably teach South Dakota v. Dole (where Congress directed that a portion of otherwise available federal highway funds be withheld from states “in which the purchase or public possession . . . of any alcoholic beverage by a person who is less than twenty-one years of age is lawful”) as requiring that there be a reasonable relationship between the conditions the federal government insists upon, and the federal interest behind the federal spending in the first place. It turns out, Dole didn’t quite impose such a requirement. Instead, Chief Justice Rehnquist’s majority opinion observed only that “our [prior] cases have suggested (without significant elaboration) that conditions on federal grants might be illegitimate if they are unrelated ‘to the federal interest in particular national projects or programs.’”
“Might be illegitimate” is not the same as “are unconstitutional.” The Court in Dole found there to be a satisfactory relationship between highway funding and underage drinking in any event (based on the notion that both involved road safety), so the Court didn’t need to explain what, exactly, if anything, is constitutionally required. Indeed, even though the Court has since (mis)characterized Dole as enshrining a requirement that the condition be germane to the funding (as, for example, in the majority opinion in New York v. United States), the Court has not once invalidated a federal spending measure on the ground that the condition is not sufficiently related to the funding.
Because the Court hasn’t really employed the germaneness idea, we don’t know what its content is, or indeed whether it can have any meaningful content.
3. How Could the Court Operationalize a Requirement that Spending Conditions and Federal Spending Objectives Be Related?
Scholars have argued, quite powerfully, that a spending condition will always be germane to the funding if the purpose of the funding is understood at a sufficiently high level of generality to subsume the subject matter of the condition. So, for example, had the condition for highway funding in Dole been a state’s having made the possession of guns near schools illegal, such a condition doesn’t seem as germane as the drinking-age condition if the purpose of highway funding is highway safety. But if the purpose of the highway funding is thought of as safety more generally, then both conditions might be equally germane.
At present, the Court has not provided any meaningful guidance about how courts should discern the relevant purpose of the funding. Need it be the actual, sincere purpose Congress had in mind, supported by legislative history? Many federal spending programs simultaneously promote multiple federal interests: is one sufficient to meet the germaneness requirement? And should the Court revisit a move it made in the 1930s to allow Congress to spend money for the “general welfare” regardless of whether Congress’s purposes tie into any of the more specific Article I fonts of federal legislative power? These would be important questions for the Court to explore if the germaneness prong of the test is to have bite. More generally, the Court has never articulated a precise justification for any germaneness test. (In New York v. United States the Court said that without a germaneness requirement “the spending power could render academic the Constitution’s other grants and limits of federal authority,” but provided no elaboration.) Giving states notice and requiring Congress to specify conditions arguably promote federalism values, by facilitating state autonomy and federal deliberation. Whether and how germaneness does—beyond simply constraining federal power in some instances—is less evident.
Even if we know what the purpose of the spending is, and we know what the condition imposed is, how do we know whether the two are sufficiently connected? In her dissent in Dole, Justice O’Connor argued that Congress’s power under the spending clause is limited to tracking or “specif[ying] in some way how the money should be spent, so that Congress’ intent in making the grant will be effectuated.” To the extent that Congress’s conditions go beyond specifying how the money need be spent, Congress would be “regulating,” not spending, and its conditions would need to be justified by some Congressional power other than spending authority: “A requirement that is not such a specification is not a condition, but a regulation, which is valid only if it falls within one of Congress’ delegated regulatory powers.”
In Dole, Justice O’Connor thought that requiring states seeking highway funding to raise the drinking age was not a condition on how the federal monies were spent, nor something Congress could accomplish itself under the Commerce Clause. O’Connor therefore voted to invalidate the spending condition.
Justice O’Connor’s proposed approach is alluring, but it has potential problems. Let us look quickly at exactly what it means to “specify in some way how the money should be spent.” Imagine Congress had said, in the highway funding arena: “This money can be used to build interstate or intrastate roads, provided the speed limit on all roads facilitated by this money shall be 55 miles per hour under state law.” Presumably that would be a permissible specification of the way federal funds are spent, regardless of whether Congress would have power to directly and regulatorily mandate speed limits on intrastate roads. Now imagine this one: “This money can be used to build interstate and intrastate roads in the state, but no money shall be used to build any road that is connected to any road on which a person under 21 can lawfully walk, drive, or ride with any alcohol in her system.” We believe Justice O’Connor would likely find this specification too “attenuated” from the funding itself. But questions of attenuation are precisely those her approach seeks to avoid by means of what she calls “a clear place at which the Court can draw the line between permissible and impermissible conditions on federal grants”—the specification notion.
Nor is the second part of Justice O’Connor’s test—the suggestion that a “regulation [within a funding law that does not count as a specification. . . can still be] valid [but] only if it falls within one of Congress’ delegated regulatory powers”—easy to understand. In Dole, she argued that if Congress had the power to regulate the drinking age throughout the U.S. directly (which she did not believe the Commerce Clause permitted), then the requirement in the funding law that states do so would have been permissible. But as she herself would point out in the majority opinion in New York v. United States a few years later, federal power to regulate citizens directly is not the same as federal power to tell the states to use their regulatory power. And she understood the condition at issue in Dole to be that “a State will raise its drinking age to 21” (emphasis added). Yet under the anti-commandeering principle in New York that she would recognize and celebrate, the federal government doesn’t enjoy the delegated regulatory power to tell states what their drinking age under state law shall be; whether Congress could regulate alcohol purchase itself would be beside the point in that part of the New York analysis.
The inscrutability of Justice O’Connor’s approach is also apparent if we try to apply her insights to the COVID-19/sanctuary policy setting. On the one hand, it might seem that her approach would reject President Trump’s proposed linkage because state abandonment of sanctuary policies is not a specification of how federal funds are spent, and because Congress (under the anticommandeering rule in New York v. United States and its progeny) can’t simply mandate state cooperation in immigration. But if the question isn’t whether Congress can mandate use of state regulatory power, but whether Congress can enforce immigration laws itself (analogous to her suggestion that Congress might have been able to regulate alcohol access in Dole, but for the Twenty-First Amendment limiting Commerce Power), then the condition might be acceptable under her test.
In any event, Justice O’Connor’s approach was rejected in Dole, in favor of (to the extent Dole required anything here) an even more inscrutable assessment of fit between condition and spending purpose in the first place. How would that play out in the COVID-19/sanctuary realm? At first blush, the two topics seem unrelated. But consider the following argument by the feds: “We seek to replace losses in tax revenues caused by the COVID-19 pandemic, and (naturally) at the same time reduce the need for further assistance in future pandemics. Because immigration (legal and illegal) can be a major driver of pandemic spread, we want states and local governments to be on board with federal immigration enforcement policies to be able to maximally combat the spread of pandemics, and avoid problems of the kind we confront and are trying to ameliorate today.” What initially seemed like a tenuous fit looks more plausible, depending on how the purpose is identified, and how Congress might rationally think matters are connected.
4. And What About the “Coercion” Question in Sebelius?
Finally, what about the possible argument states might make (that was also invoked in the Byrne grant context) that sanctuary abandonment conditions would amount to unconstitutional coercion, inasmuch as the States have no real choice to decline the funding (and comply with the conditions) given the States’ economic realities. Certainly, federal COVID-19 relief funding is essential to the financial wellbeing of many if not all state and local governments (in part because, unlike the federal government, almost all state and local entities are legally constrained from running whatever kinds of budget deficits they’d like). And so the coercion argument may seem powerful.
Yet there is no modern Supreme Court case squarely holding that a deal offered by Congress to states is invalid simply because it is too adhesive. And, for now, that is a good thing.
To be sure, some lower courts (including the Second Circuit) and some commentators seem to read the Medicaid portion of Sebelius to mean that if a particular deal Congress offers in fact places too much pressure on state fiscs, then states needn’t live up to the agreement. The Court’s own mention in that case of how the Medicaid expansion placed a “gun” to the states’ heads may well contribute to that impression. But as one of us has explained in detail in academic commentary and in a previous column, this reading of Sebelius is unnecessary and dangerous.
It is unnecessary because a careful parsing of Sebelius indicates that a sufficient reason Obamacare was deemed by a majority of the Court to be coercive was that Congress was vitiating the terms of a preexisting deal. (Seeking to unilaterally alter important terms of an agreed-upon deal and impose new terms is inherently coercive regardless of the amount of money at stake.) That is, any federal coercion present was a function of a lack of meaningful notice to, and thus consent by, the states.
In this regard, Chief Justice Roberts’s majority opinion effectively characterized Medicaid as one continuing program in which the states had essentially been told they would receive money on an ongoing basis, subject to minor coverage and funding modifications, until the program was formally repealed. He thus identified the problem with Obamacare’s expansion of Medicaid as the federal government’s having “surpris[ed] participating States with post[-]acceptance or ‘retroactive’ conditions. . . . A State could hardly anticipate that Congress’s reservation [in the original Medicaid Act] of the right to ‘alter’ or ‘amend’ the Medicaid program included the power to transform it so dramatically.”
This notice-based reading of Sebelius is buttressed significantly by the fact that Chief Justice Roberts apparently conceded (dissenting) Justice Ginsburg’s claim that Congress could lawfully have repealed Medicaid and replaced it with Obamacare. Chief Justice Roberts replied to that assertion not by saying that repeal today would be unconstitutional, but only that it would be politically difficult. Yet repeal/replace would impose the same substantive duress on states—the same “gun” to their heads—as did Obamacare. The only legal difference between the two is that Congress in the original Medicaid Act reserved for itself the right to “repeal . . . any provision.” While the kind of “alter[ations]” and “amend[ments]” Congress and the states had in mind in 1965 may be open to debate, a right of “repeal” is more textually straightforward.
A broader reading of Sebelius is not only less lawyerly; it creates daunting slippery-slope problems about how big a hole in a state’s budget is too big for Congress to be able to blow, and substantial if not insurmountable institutional-legitimacy questions about why courts are the right institutions to be drawing such inherently political lines.
To the extent that the Sebelius Court itself opened the door to these (seductive but problematic) broader notions of substantive coercion, this is yet another aspect of conditional spending doctrine the Court should address and clarify.