The Ontology of Sovereign Immunity
by Michael Dorf
Last week's SCOTUS ruling in Upper Skagit Indian Tribe v. Lundgren resolved nearly nothing. The Court vacated and remanded a ruling of the Washington Supreme Court for reconsideration of the question whether the defendant American Indian tribe had sovereign immunity, without the impediment of a misunderstanding of prior SCOTUS precedent under which the state court had previously labored. Yet as I explain in my latest Verdict column, the case is potentially important because of arguments made in a concurring opinion by Chief Justice Roberts, joined by Justice Kennedy--arguments which, if taken seriously, undercut the basis for state sovereign immunity, as well as various other doctrines.
To oversimplify, the Chief Justice's concurrence strongly suggests that tribal sovereign immunity and its exceptions should not be constructed in a way that denies claimants whose substantive rights have been violated any effective remedy. That principle is sound, I argue, but contradicts much of the Court's own jurisprudence, including a more newsworthy case decided the same day as Upper Skagit--Epic Systems Corp. v. Lewis--in which the Court aggressively construed the federal Arbitration Act in a manner that effectively denies numerous plaintiffs effective remedies for violations of their substantive rights.
The column addresses that tension between the Court's tribal sovereign immunity doctrine and both other branches of sovereign immunity doctrine and other lines of cases. Here I want to suggest that there is another tension between how the court treats tribal sovereign immunity and how it treats state sovereign immunity--an ontological consideration. ("Ontology" is a fancy word for the nature of a thing.) State sovereign immunity has constitutional status, whereas tribal sovereign immunity is merely a matter of federal common law, thus overridable by Congress. Is that difference justified?
The Constitution nowhere expressly mentions sovereign immunity, but Court cases treat such immunity as an inherent incident of sovereignty, a point supposedly confirmed by the Eleventh Amendment--even though the Eleventh Amendment itself does not mention sovereign immunity and appears to be written in a way that simply disclaims a certain category of federal court jurisdiction. But building on the 1890 case of Hans v. Louisiana, a series of SCOTUS cases in the 1990s and 2000s establish that: States have immunity in state and federal court (as well as before at least some administrative agencies) against private lawsuits for damages or other "retrospective" relief, regardless of whether the plaintiff is a state resident and regardless of whether the plaintiff sues for relief under state or federal law; a state can waive its own sovereign immunity, but Congress may not abrogate a state's sovereign immunity unless it does so clearly and when acting pursuant to its power to enforce the Reconstruction Amendments, a power that is fairly narrowly construed.
Last week's SCOTUS ruling in Upper Skagit Indian Tribe v. Lundgren resolved nearly nothing. The Court vacated and remanded a ruling of the Washington Supreme Court for reconsideration of the question whether the defendant American Indian tribe had sovereign immunity, without the impediment of a misunderstanding of prior SCOTUS precedent under which the state court had previously labored. Yet as I explain in my latest Verdict column, the case is potentially important because of arguments made in a concurring opinion by Chief Justice Roberts, joined by Justice Kennedy--arguments which, if taken seriously, undercut the basis for state sovereign immunity, as well as various other doctrines.
To oversimplify, the Chief Justice's concurrence strongly suggests that tribal sovereign immunity and its exceptions should not be constructed in a way that denies claimants whose substantive rights have been violated any effective remedy. That principle is sound, I argue, but contradicts much of the Court's own jurisprudence, including a more newsworthy case decided the same day as Upper Skagit--Epic Systems Corp. v. Lewis--in which the Court aggressively construed the federal Arbitration Act in a manner that effectively denies numerous plaintiffs effective remedies for violations of their substantive rights.
The column addresses that tension between the Court's tribal sovereign immunity doctrine and both other branches of sovereign immunity doctrine and other lines of cases. Here I want to suggest that there is another tension between how the court treats tribal sovereign immunity and how it treats state sovereign immunity--an ontological consideration. ("Ontology" is a fancy word for the nature of a thing.) State sovereign immunity has constitutional status, whereas tribal sovereign immunity is merely a matter of federal common law, thus overridable by Congress. Is that difference justified?
The Constitution nowhere expressly mentions sovereign immunity, but Court cases treat such immunity as an inherent incident of sovereignty, a point supposedly confirmed by the Eleventh Amendment--even though the Eleventh Amendment itself does not mention sovereign immunity and appears to be written in a way that simply disclaims a certain category of federal court jurisdiction. But building on the 1890 case of Hans v. Louisiana, a series of SCOTUS cases in the 1990s and 2000s establish that: States have immunity in state and federal court (as well as before at least some administrative agencies) against private lawsuits for damages or other "retrospective" relief, regardless of whether the plaintiff is a state resident and regardless of whether the plaintiff sues for relief under state or federal law; a state can waive its own sovereign immunity, but Congress may not abrogate a state's sovereign immunity unless it does so clearly and when acting pursuant to its power to enforce the Reconstruction Amendments, a power that is fairly narrowly construed.
Almost every aspect of the state sovereign immunity doctrine is controversial. Here I am focusing on the last part: its ontology. If state sovereign immunity were a matter of federal common law--i.e., judge-made law that is not required by the Constitution--then Congress could change it. And in fact, one of the leading opponents of the modern state sovereign immunity doctrine, Justice Souter, argued for just that status for state sovereign immunity (here and here). Justice Souter thought that it was a mistake to treat US law as containing any principle of state sovereign immunity, but he would have made his peace with the mistake of Hans by treating it as merely a common-law doctrine. The Court in Hans, Souter said, should be understood to have said that when The People adopted the original Constitution and the Eleventh Amendment, they expected that the common law doctrine of state sovereign immunity would come along with it. But as a common law doctrine, he contended, it could be overridden by Congress.
Why did Justice Souter think that? Because otherwise sovereign immunity fits poorly with small-r republican government. Sovereign immunity originated in England where the sovereign was the king or queen, who, as the source of the law, could not be held accountable to it. Various justices in Chisholm v. Georgia took the position that this view is incompatible with republican government. In the wake of the Eleventh Amendment (which overruled Chisholm) and the expansion thereof in Hans, we must find a way to reconcile sovereign immunity with republican government.
Justice Holmes provided a reconciliation this way: "A sovereign is exempt from suit not because of any formal conception or obsolete theory, but on the logical and practical ground that there can be no legal right as against the authority that makes the law on which the right depends." It's not that the government stands above the People, but that the government makes the law, including the law on which the right to sue depends.
That's not a fully satisfactory explanation for why there should ever be sovereign immunity against constitutional claims, but it does imply that whatever sovereign immunity exists should be a matter of common law only -- i.e., it should be waivable by Congress. Congress, as the repository of sovereign lawmaking power, may choose not to assert its sovereign immunity. That explains federal sovereign immunity.
Building on this theory, Justice Souter argued in the modern cases that Congress should also be able to abrogate states' sovereign immunity against federal law claims, because the states are not "the authority that makes" federal law. Thus, Justice Souter dissented from all of the decisions disallowing congressional abrogation on grounds that Congress was not acting pursuant to its power to enforce the Reconstruction Amendments. If state sovereign immunity is just a matter of federal common law, then Congress should be able to abrogate it using any of its powers, he thought.
However, Justice Souter lost that battle. State sovereign immunity is treated as constitutional law, abrogable by Congress only when acting to enforce the Reconstruction Amendments.
Notably, the Court has made clear that tribal sovereign immunity and foreign sovereign immunity are both matters of common law, not constitutional law. The Court says as much about tribal sovereign immunity in Upper Skagit, citing earlier such statements. And Congress has, via the Foreign Sovereign Immunities Act, lifted some of the default foreign sovereign immunity that had been recognized by the Court, thus treating foreign sovereign immunity as a matter of common law as well. (Customary international law and/or treaties can obligate the US to retain some forms of foreign sovereign immunity, but if so, the duty would be under international law only; a federal statute completely eliminating foreign sovereign immunity might thus put the US in breach of its international law obligations but would be fully operative domestically.)
Accordingly, state sovereign immunity is an ontological anomaly. Unlike the other three forms of sovereign immunity--which are completely subject to the control of Congress--state sovereign immunity has a constitutional core that cannot be changed by Congress. Can that anomaly be justified?
The conventional answer since Hans has been to say that when the Constitution was ratified, everyone expected that state sovereign immunity would be preserved. But that hardly answers the question, because we have to ask next: "preserved as what?" Souter asks that question and answers that it was preserved, if at all, only as common law. A satisfactory solution to the puzzle must distinguish states from the federal government, foreign sovereigns, and Indian tribes.
One possibility is to say that the states were the parties to the deal that created the US. Foreign sovereigns were not. Indian tribes were recognized as having some status under the Constitution but were not parties to the deal. And the federal government wasn't party to the deal; it was the result of the deal.
The difficulty with that answer is that it contradicts what a majority said in US Term Limits v. Thornton, which is that The People as a whole, not the states in their respective sovereign capacities, created the United States. To be sure, four justices dissented on just that point in US Term Limits, and they were four of the five justices who made up the majority in the roughly contemporaneous state sovereign immunity cases. But Justice Kennedy, who was the fifth vote in those state sovereign immunity cases, joined the US Term Limits majority. So for him, at least, there is a problem with relying on some special status of the states as parties to the Constitution as a basis for distinguishing state sovereign immunity from other forms of sovereign immunity.
Where does all of this leave us? I do not imagine that the Court will overrule Hans or its expansions of Hans anytime soon. Still, comparison of the Court's treatment of state sovereign immunity doctrine with the three other kinds of sovereign immunity does create a special burden of justification for treating state sovereign immunity as not fully changeable by Congress. And I do not think that the burden has been satisfied.