Article I, Section 10 of the Constitution contains a clause that is about to be quoted a great deal. “No State shall, without the Consent of Congress … enter into any Agreement or Compact with another State, or with a foreign Power.”
On 23 January, in Switzerland, the governor of California met the director general of the World Health Organization and announced that California would join the Global Outbreak Alert and Response Network, becoming the first American state to do so. The announcement came one day after American withdrawal from the WHO took formal effect, and roughly two weeks after the federal government announced it was leaving dozens of other United Nations bodies.
The argument here is that the Compact Clause is the objection everyone will reach for, that it is the weakest of the three available, and that the strongest one has nothing to do with the text above.
The first test: what was actually joined
Start with the network, because a great deal follows from what kind of thing it is.
GOARN is not a treaty organization and it has no member states in the sense that the WHO has member states. It is a technical network, coordinated by the WHO, whose partners are institutions: hundreds of public health bodies, university departments, reference laboratories, response organizations and some national ministries. Its function is to move expertise and information quickly when an outbreak appears somewhere, and its partners contribute staff and analysis rather than money or sovereignty.
A state department of public health joining that is joining alongside laboratories and universities, on the same footing, under an institutional partnership. It is closer to a hospital system joining a research consortium than to a government acceding to an instrument.
That is not a technicality invented for the occasion. It is what determines whether the clause above is even engaged, because the clause speaks to agreements and compacts with a foreign power, and a WHO coordinated network of laboratories is not a foreign power.
The second test: what the clause has been held to mean
Read literally, the Compact Clause would void a very large amount of ordinary state activity, including reciprocal driver licensing arrangements with Canadian provinces and half the sister city agreements in the country. It has not been read literally for well over a century.
In Virginia v. Tennessee, decided in 1893, the Supreme Court held that the clause reaches agreements directed to the formation of any combination tending to increase the political power of the states, such as may encroach upon or interfere with the just supremacy of the United States. Agreements that do not do that fall outside it. The Court reaffirmed that approach in 1978 in United States Steel Corporation v. Multistate Tax Commission, declining to require congressional consent for an interstate body that had no power to compel anyone.
Apply that test. California has agreed to share epidemiological information and to make personnel available to a technical network. It gains no power over any other state, acquires no authority it did not have, and takes on no obligation it could not walk away from on a week’s notice. Whatever else that is, it is not a combination increasing the political power of a state at federal expense.
The Compact Clause argument is available and someone will make it. It is not a strong one.
The third test: the one that actually bites
The serious objection is not in Article I at all. It is the doctrine that a state may not act in a way that conflicts with the federal government’s conduct of foreign relations, even where Congress has passed no statute saying so.
The clearest case is American Insurance Association v. Garamendi, from 2003, in which the Supreme Court struck down a California statute requiring insurers to disclose policies issued in Europe before the war. The state’s law was not preempted by any act of Congress. It was preempted because it cut across an approach the executive had adopted in negotiations with foreign governments, and the Court held that the executive’s conduct of foreign policy was enough.
Garamendi is the precedent that should worry Sacramento, because the shape matches. The federal government has taken a deliberate position on the WHO, expressed by withdrawing from it. A state has then attached itself to a body the WHO coordinates. An argument that this interferes with a settled federal foreign policy is not frivolous.
It is also not obviously right, and the reason is the institutional point from the first test. Garamendi involved a state statute imposing binding obligations on private companies in a field where the President was actively negotiating. California has not legislated, has not bound anyone, and has not asserted a position on the merits of American membership. A health department sharing outbreak data with a laboratory network is a thinner thing to preempt, and the thinner it is the harder the argument gets.
The test that has already been run
There is one more reason to think the constitutional objections are weaker than they will sound, which is that they have been tried, against this state, and they lost.
California linked its cap and trade market to the one run by the Canadian province of Quebec, allowing allowances issued in one jurisdiction to be used in the other. In 2019 the Department of Justice sued, and it did not pick one theory. It argued the arrangement violated the Treaty Clause, the Compact Clause, the foreign affairs doctrine and the Foreign Commerce Clause, which is very nearly the complete list of things anyone could say about a state dealing with a foreign government.
On 17 July 2020 the district court ruled for California on the claims that remained, and the federal government subsequently agreed to dismiss its own appeal.
That litigation concerned a market linkage with binding effects on companies trading real assets across a border, which is a considerably heavier arrangement than sharing outbreak data with a laboratory network. If the heavier one survived all four theories, the lighter one starts from a comfortable position.
The caveat is that a district court decision binds nobody outside its district and was never tested on appeal, because the appeal was dropped rather than decided. It is a strong indication, not a settled rule.
The verdict, and what will actually decide it
On the law as it stands, California is probably within its rights, and the case against it is a foreign affairs preemption case rather than a Compact Clause case, which means it would turn on how directly this cuts across federal policy rather than on the words quoted at the top.
But the law is unlikely to be what settles this, and that is the more useful thing to watch.
Three practical questions will decide whether the arrangement means anything twelve months from now, and none of them is constitutional. The first is data: whether California’s outbreak reporting runs through systems the federal government operates, because a state that has to route its own surveillance data through a federal pipeline to get it does not have much to contribute independently. The second is money, since GOARN participation costs staff time and the state legislature will have to appropriate it in a budget cycle that has other problems. The third is whether other states follow, because one state is a gesture and eight states is a parallel structure, and the federal response to those two things would not be the same.
Watch the second one first. Everything in state government that survives its announcement survives because somebody put it in a budget, and the budget is due in the spring.
