Even More Thoughts on State Court (Non)Obligation to Follow Federal Appeals Court Precedents (Wherein I Respond to Professor Frost)

by Michael Dorf

My post last week on Roy Moore's efforts to forestall same-sex marriage in Alabama prompted a number of interesting comments both on and off the blog. Some of those comments related to a point I have made before--namely, Moore's view, with which I agree, that the rulings of a lower federal court do not bind state courts as a matter of precedent in cases with different parties. Here I want to clarify the limited nature of my agreement with that proposition. I'll use a recent blog post and forthcoming law review article by Professor Amanda Frost to frame my remarks.

Professor Frost correctly notes that although the view expressed by Moore is the conventional wisdom, the Supreme Court has never definitively decided the question whether state courts are obligated to follow federal precedents of lower federal courts, and that at least some courts have taken the position that they are. She describes what I'll cheekily call the Dorf/Moore position as the "conventional wisdom," and she offers a challenge to it. Here I want to respond to that challenge, although I also want to concede a fair bit of ground to Professor Frost.

I'll begin with three concessions: (1) I agree with Professor Frost that Congress has the power to obligate state courts to give federal court rulings on federal law effect as binding precedents. (2) I also agree with her that there are good policy arguments for Congress to exercise this power, especially given the tendency of the current Supreme Court to allow state/federal splits to linger for years, causing legal uncertainty. (3) And finally, I agree that the SCOTUS could, as a matter of federal common law, rule that state courts are bound by decisions of lower federal courts--although, as I'll explain below, federal appeals courts lack this power due to a kind of Godelian paradox.

But before coming to that point, I want to explain why I also continue to think that the conventional wisdom is correct. Although Congress could tell state courts that they must abide by lower federal court precedents, I don't think that Congress ever did tell state courts to do so. Certainly no federal statute expressly states such an obligation.

Might an obligation of state courts to follow federal court precedents on federal law follow from Congress's decision in 1875 to vest original federal question jurisdiction in the lower federal courts (subject to a much-later-repealed amount-in-controversy minimum)?  It would be extremely odd to infer such an obligation from the naked grant of jurisdiction to federal district courts. Consider that the decisions of a federal district court judge do not even bind other district court judges (or even that same district court judge in a subsequent case with different parties), and so it would be very strange to give it greater effect in the state courts.

Thus, any state court obligation to follow lower federal court precedent would necessarily be at most an obligation to follow precedents of the federal appeals court that encompasses the state. And there doesn’t appear to be anything in the statutes conferring jurisdiction on those courts that would entail an obligation on state courts to follow their federal precedents. (As a prudential matter, state executive officials must abide by such precedents because they are subject to suit in federal court. Professor Vik Amar's Verdict column last week explains why this prudential obligation may even be conceived as a legal duty.)

The case for the conventional wisdom is buttressed by the history of the creation and adjustment of federal court federal question jurisdiction. When Congress first conferred federal question jurisdiction on federal trial courts in 1875, it conferred it on the circuit courts. At that time, circuit courts were appeals courts in some respects but also trial courts. I suppose it could be argued that in conferring jurisdiction on circuit courts, Congress intended for those circuit courts' rulings on federal law to be binding on state courts within those circuits, but again, there's nothing in the Act that appears to say that. Rather, the mechanism that gets the most print in the Act is removal. Plaintiffs or defendants who are unhappy with the way they expect a federal case to go in state court can remove to federal court.

So for the argument that the conventional wisdom is wrong to work, one has to assume that when Congress vested jurisdiction of federal question cases in the circuit courts, it meant their rulings on federal law to bind the state courts (although it never said so), but that when Congress later reorganized the courts and transferred federal question jurisdiction to the district courts, the binding impact of rulings of circuit courts (later renamed appeals courts) remained in those circuit courts, even though Congress never said anything about that either. These suppositions seem rather far-fetched.

So far I've shown (to my own satisfaction at least) that Congress never enacted a statute obligating state courts to treat federal appeals court decisions on federal law as binding. It's also clear that the SCOTUS has not created such an obligation as a matter of federal common law. But, it might be objected against the conventional wisdom, the SCOTUS also hasn't rejected such an obligation as a matter of federal common law either. And if one believes that federal common law--like common law more generally--can be discovered rather than simply made by courts, then one might think that there is already a federal common law obligation of state courts to follow the precedents of federal appeals courts on questions of federal law.

I think that view would be mistaken on at least two grounds. First, I think that at least since Erie Ry. v. Tompkins, the better view of common law (whether of the state or federal variety) is that it is judge-made law rather than judge-discovered law.

Second, even if one takes the view that judges discover common law, they would need to discover it somewhere. Federal common law is typically interstitial. It serves to implement and complement existing federal policies. As I have said, there are good policy reasons why Congress or perhaps even the Supreme Court ought to instruct state courts to follow federal appeals court precedents on federal law, but I don't think that one can fairly say that those policy reasons implement or radiate from or in some other way directly advance the existing policy choices reflected in Article III, 28 USC sec 1331, the Supremacy Clause, or any other existing federal policy. True, the decision to fashion a federal common law rule would serve some of the same policies as those other provisions serve--especially uniformity and supremacy for federal law--but that factor would only be in play if one assumes that federal courts make federal common law, and the premise of this paragraph is that one wants to know whether a federal judge committed to the pre-legal realist idea that judges discover common law could discover a federal common law principle binding state courts to federal appeals court rulings on federal law. I don't think the existing materials can be fairly said to support such a discovery (assuming that it is even coherent to talk about judges discovering law).

So the conventional wisdom is right. Neither Congress nor the Supreme Court has made the federal law pronouncements of federal appeals courts binding on state courts. As I've conceded, though, perhaps one or the other should.

But I've left out a set of actors. The Supreme Court is not the only court that can make federal common law. In principle, any court faced with a request to make federal common law can oblige. After the SCOTUS, the state courts themselves or the federal appeals courts are the most logical actors to make federal common law binding the state courts to the federal appeals court rulings on federal law. In an earlier post, I explained why a state court could not take this obligation on as a matter of state law, but I want to clarify here that I think a state court could make federal common law thus binding itself, subject to being overruled on that point by the SCOTUS.

What about federal appeals courts? As Professor Frost notes in her article, both the courts of appeals for the Eighth and Ninth Circuits have indicated that they think that their federal law rulings bind state courts--although the Eighth Circuit decision saying that was (as she acknowledges in a footnote) reversed on other grounds by the Supreme Court, and the Ninth Circuit statement to that effect was only posed as speculation. (Fun fact: back in 1991, when I was a law clerk to Judge Reinhardt, I wrote the first draft of the language from the Ninth Circuit case that Professor Frost cites --which doesn't mean I'm persuaded today.)

Okay, let's suppose that a federal appeals court wants to hold that state courts in its geographical coverage area are bound by federal appeals court rulings on federal law. How, exactly, will it do so? It can say so in some case, and that will be res judicata in a later case involving the same parties. But in a state court case involving different parties, if the state court believes that it is not bound by federal appeals court rulings on federal law, then the fact that the federal appeals court says otherwise will not itself be binding. The federal appeals court cannot bootstrap itself binding effect on state courts that deny that effect. Thus, only the state courts and ultimately the SCOTUS can really make the federal common law that would tell state courts to follow federal appeals court rulings on questions of federal law.