On Tuesday the Court hears Montgomery v. Louisiana, a case concerning whether Miller v. Alabama, a decision in which the Court held sentencing juveniles to life without the possibility of parole was generally unconstitutional, applies retroactively. The Louisiana Supreme Court, in a collateral challenge to a final life sentence, held Miller doesn't apply retroactively, choosing to adopt as its state-law retroactivity standard Teague v. Lane, a case on retroactivity for purposes of federal habeas review. Since the Court has held that that Teague, at least, does not set a ceiling on the retroactive effect state courts may give the Court's decisions (though it has left open whether Teague interpreted the Due Process Clause, sub silentio, to require a retroactivity floor), there's an obvious doubt as to whether the Court has jurisdiction over Louisiana's application of Teague at all. The Court, recognizing that doubt, ordered the parties, and appointed an amicus, to brief the issue. The parties and the SG's office, which filed an amicus brief, support jurisdiction; the amicus was appointed to oppose it.
The Court only has jurisdiction, as far as I, the parties, the Court-appointed amicus, the SG's office, and commentators can tell, if one of two things are true: (1) the part of Teague that the Louisiana Supreme Court may have gotten wrong, namely Teague's exceptions to its general rule of non-retroactivity, is constitutionally mandated by the Due Process Clause and not just an interpretation of a federal habeas statute, or (2), simply by choosing to adopt a purely federal-habeas retroactivity standard for purposes of deciding state-law retroactivity, the Louisiana Supreme Court has injected a federal question into what would otherwise be a purely state-law case. Finally, by way of prologue, Michigan v. Long, the Court's seminal case on how to decide whether a state court decision rests on an adequate and independent state ground, is a red herring here, as the SG all but concedes, because Michigan only comes into play where a state-court case actually involved some federal claim in addition to a state-law claim, or at the very least some possibility that the state court believed federal law required the state court to, in interpreting its state constitution, reach a certain federal floor.
I don't have any interest in whether Teague's exceptions are constitutionally mandated in whole or part; there's a respectable argument that they are and respectable arguments that they're not. What I do find interesting, and disturbing, is the argument, made by the parties, the SG, several amici, and the commentators linked to above, that even if Teague isn't constitutionally mandated, whether Louisiana's voluntary application of it for state-law retroactivity purposes was correct is a federal question which the Court has jurisdiction to review. I'm going to offer a fairly detailed flyspecking of the arguments for this remarkable proposition, but first, consider these hypotheticals on which the Court would have jurisdiction if the parties, SG, et al. are right.
Suppose a state amends its state-constitutional analogue to the Equal Protection Clause to protect transgendered persons on a par with racial minorities. The state supreme court, interpreting this amendment, then holds that in reviewing any discrimination against transgendered persons, it will apply the Supreme Court's strict scrutiny test that the Court uses to review discrimination against certain protected classes under the Equal Protection Clause. It then produces a learned opinion citing every major case in which the Court has applied strict scrutiny since Korematsu, and deploying these precedents to shape and explain its understanding of what compelling interests are and what narrow tailoring requires.
I wouldn't have thought, until I read the Montgomery briefs, that anyone would say the state's application of strict scrutiny, federal though its origins may be, would present a federal question that the Court could decide. Yet I'm reasonably certain that if the Court adopts the standards set forth in those briefs, it would have jurisdiction in my hypothetical. For example, the SG says that "the Court has the power to review a state court's interpretation of federal law, even when the state court did not have to apply that law in administering state causes of action or in interpreting state statutes." Louisiana says the Court has jurisdiction because its state supreme court chose to rely "solely on the Teague framework in determining that Miller is non-retroactive [as a matter of state law]." Just so in my hypothetical; the hypothetical state chose to rely solely on the federal strict-scrutiny framework, citing lots of federal cases, in determining whether an instance of transgender discrimination violated the state constitution. And, satisfying the SG's test, the state court interpreted federal law, though the state court didn't have to apply it in the first place.
Some people might argue that Montgomery is different, perhaps because there isn't really any federal law that, as of yet, says discrimination against transgendered persons is subject to strict scrutiny, while there is federal law that applies Teague to determine Miller's retroactive effect. I don't, however, understand that distinction because - by hypothesis - there isn't federal law that applies Teague to determine Miller's retroactive effect in state-court collateral proceedings. There's just federal law that applies Teague for a different, though very similar, purpose (retroactivity in federal habeas), just as there's federal law that applies strict scrutiny for different, though fairly similar, purposes than deciding whether laws that discriminate against transgendered persons violate state constitutions. Others, specifically the SG, might say that Montgomery is different because there is no "independent state element" in Louisiana's retroactivity analysis (a limitation to his proposed jurisdictional rule the SG offers) - there's just Teague. On the other hand, in my hypothetical there is an independent state element in the state court's analysis, namely the provision of the state constitution that prohibits discrimination against transgendered persons in the first place. I don't see why this should matter, and nor does the SG, really, as every case he cites in support of jurisdiction here is chock-full of independent state elements, as we'll see below. But even if the distinction had any value, I don't understand it on these facts, because in Montgomery, there is an independent state element - the state's choice to give retroactive effect to certain decisions on collateral review, and its further choice to use Teague as its method of selecting which decisions those are.
Though I think any distinctions between Montgomery and my hypothetical are minor differences of degree, another hypothetical should collapse even those distinctions. The Supreme Court occasionally announces rules pursuant to its supervisory power over the lower federal courts; for example, the Court held in Young v. U.S. ex rel. Vuitton et Fils SA, purely as a matter of supervisory power, that the lower federal courts must refer criminal contempt matters to the local U.S. Attorney and appoint an independent prosecutor if the U.S. Attorney doesn't accept the referrals. A state supreme court could hold that its inferior courts must do, in criminal contempt matters, just what the Supreme Court held its inferior courts must do. It could then reach some controversial, cert-worthy result about the meaning of Young's independent-prosecutor requirement. Would the Court have jurisdiction there? There, the only "independent state element" is the decision to apply Young, as in Montgomery the only independent state element is the decision to apply Teague, and the federal law being applied by the state to its contempt prosecutions is federal law that applies in a precisely analogous federal context, just as federal law applies Teague in a precisely analogous context to that in which Louisiana applies Teague. Yet how could the Court possibly have jurisdiction to supervise a state supreme court's purely state-law supervision of its inferior courts, however much that supervision might borrow from federal law?
The parties and the SG argue, for the most part, that the Court has held before that it had jurisdiction to review a state court's interpretation of federal law that the state court or legislature chose to incorporate for state-law purposes, and thus must hold it again. That's both an accurate and completely misleading description of the Court's precedents. The simple distinction between Montgomery or my hypotheticals and the handful of precedents the supporters of jurisdiction in Montgomery cite is this: in that handful of precedents, the federal-law element a state voluntarily chose to incorporate into its state law was a real, not hypothetical, federal-law element. Federal law actually said something about what was happening in those state-court cases. It did not say something about what was happening only on the counterfactual hypothesis that those cases were federal cases, as Teague only says something as a matter of federal law about the retroactive effect of Miller in Louisiana's collateral proceedings on the counterfactual what-if scenario that Montgomery was a federal habeas case and not a Louisiana collateral proceeding. (Assuming, anyway, that Teague is not a constitutional rule, which one has to assume for the component of the jurisdictional question I'm discussing to matter.)
Consider the SG's examples. First, the SG points to Standard Oil Co. v. Johnson, a case in which the Court decided whether or not a military post exchange was an arm of the federal government. In Standard Oil, California had a gasoline tax that exempted sales to the government of the United States. California, we'll assume, wasn't constitutionally required to make that exemption; it just chose to. Once it did, however, it embedded a federal-law question into its gas tax as to whether certain entities were part of the United States government. Note that that question wasn't in the nature of a hypothetical, e.g., "if this were a case about applying a federal gas tax, what would the federal law be about the status of post exchanges?" The post exchanges California taxed either were or weren't part of the federal government as a matter of federal law, and the California Supreme Court could (and did) get controlling federal law on that question wrong.
Next the SG relies on Merrell Dow, where the Court held that the lower federal courts lacked statutory federal-question jurisdiction over state-law negligence per se claims founded on federal misbranding law, but said in dicta that the Court had jurisdiction under its broader jurisdictional grant to decide the federal issues in such an action. Though dicta, that's clearly correct. The Merrell Dow plaintiff said Merrell Dow was negligent per se under state law because Merrell Dow's labeling violated the Food, Drug and Cosmetic Act, which Ohio negligence law picks up in negligence per se actions. Ohio didn't have to care about FDCA violations. Nevertheless, the federal question Ohio chose to incorporate in its negligence law was a genuine, non-hypothetical federal question that Ohio courts could get wrong or right as a matter of federal law: did Merrell Dow violate the FDCA? The question wasn't, "hypothetically, if the FDCA applied in a context to which it does not apply, would there be a violation?"
Next, in Three Affiliated Tribes, the North Dakota Supreme Court appeared to hold, as a matter of state law, that Indian tribes couldn't sue non-Indians in certain circumstances in state court. That holding would create obvious constitutional questions, which the Court avoided deciding by reading the North Dakota decision to have construed a North Dakota state jurisdictional statute against the background of an incorrect understanding of federal law, on which federal law barred jurisdiction over the tribes' claims. The Court held it had jurisdiction to correct that misunderstanding and remanded the case for the North Dakota Supreme Court to construe its state statute afresh. Again, Three Affiliated Tribes involved a real federal question, not a what-if-federal-law-applied-here question: did, in fact, federal law bar jurisdiction over the tribes' claims? North Dakota was wrong about what federal law dictated, and the Court held it had jurisdiction to correct their error.
Finally, the SG relies on Ohio v. Reiner, a per curiam decision in which the Court reversed the Ohio Supreme Court's holding that a witness who denies culpability lacks a Fifth Amendment privilege against self-incrimination. This federal-law holding was embedded in a decision that a witness was improperly granted transactional immunity under an Ohio statute, which provides that in certain circumstances, an Ohio trial court must give transactional immunity to a witness, and compel him to testify, if he validly invokes his privilege against incrimination. The Ohio Supreme Court held Reiner had no privilege and therefore shouldn't have gotten transactional immunity. Ohio didn't have to write a transactional-immunity statute like the one it did, but once it did, it embedded a real federal question into its transactional-immunity state law: do witnesses have valid claims of privilege against self-incrimination under the Fifth Amendment? Reiner actually had one under the Fifth Amendment, the Ohio Supreme Court held she actually did not, and the Court corrected this error of actually applicable federal law and remanded the case for the Ohio Supreme Court to decide the remaining state-law issues absent the federal-law error.*
None of these cases are anything like Montgomery. In Montgomery (and in my hypotheticals), there is no real federal-law question, and no federal law for the Louisiana Supreme Court to be wrong about. Louisiana's test for retroactivity in collateral proceedings, as a matter of state law, has one element: Teague. Teague, however says nothing (by hypothesis) about the retroactive effect of Miller in Louisiana's collateral proceedings; Louisiana merely uses Teague by analogy to decide what it wants to say about the retroactive effect of Miller in its collateral proceedings. Likewise, in my hypotheticals, federal law doesn't really say anything about appointing prosecutors in state-court criminal-contempt proceedings, and federal law doesn't speak to whether transgender discrimination violates a state constitution.
Because federal law doesn't really speak to retroactivity in Louisiana collateral proceedings (by hypothesis), nor to any element or aspect of Louisiana's retroactivity test, Louisiana can't be wrong about what Teague says about Miller's Louisiana-law retroactivity, like Ohio could be (and was) wrong about whether Reiner had a Fifth Amendment privilege, like North Dakota could be (and was) wrong about whether federal law barred jurisdiction over tribal claims, like Ohio could be wrong about what the FDCA required Merrell Dow to put on a label, or like California could be (and was) wrong about whether federal law made military post exchanges parts of the federal government. Louisiana can only be wrong about what Teague says about Miller's retroactive effect in Louisiana on the imaginary scenario that Montgomery is a federal habeas petitioner and Montgomery is a federal case - or, the alternative counterfactual scenario that Teague governs Louisiana collateral proceedings when (by hypothesis) it doesn't.
A wildly misguided analogy from an amicus brief in support of jurisdiction illustrates the distinction I'm drawing. This amicus writes that teenagers aren't required to get a driver's license, but once they do they are required to follow state laws that would not otherwise apply to them. Quite so, of course, but the amicus then says that "the same is true with respect to state courts employing the Teague test"; they don't have to opt in to Teague, but once they do they have to "apply [it] correctly." Not so. A teenager who chooses to get a driver's license subjects herself to an actually applicable legal regime. On the other hand, a teenager who chooses to live by the Code of Judicial Ethics is not bound, at least by the Code of Judicial Ethics, to accurately follow the Code. The teenager may be bound by her own dictate, but the question of whether the teenager is accurately following the Code is a question of teenager-dictate rules, not a question of judicial ethics rules. One could sensibly reprimand such a teenager for "not living up to your moral code," by doing something which would be a violation of the Code were she a federal judge; one could not sensibly charge her with a violation of the Code. If a court existed with jurisdiction to decide issues of judicial ethics, to just the same expansive degree the Supreme Court has to decide issues of federal law that arise as elements of or defenses to non-federal claims, that court would obviously lack jurisdiction over the teenager. The same is true with state courts that choose to apply Teague. When they get Teague "wrong," they're getting state law wrong; they're not getting federal law wrong, because federal law is, by hypothesis, silent on the subject of the state courts' decisions.
As for the theory behind this frankly zany assertion of jurisdiction, it seems to be simply a confused notion that if state courts choose to borrow federal law for state-law purposes, they need to get federal law right and the Court needs to be there to correct them if they don't. The trouble with that, again, is that no one's getting federal law wrong in these scenarios. They're getting fantasy-land federal law wrong, where fantasy-land federal law is defined as an imaginary version of federal law where federal law says things and applies to things it doesn't say and doesn't apply to -- just as the teenager who can't understand the Code of Judicial Ethics is getting fantasy-land judicial ethics rules wrong. Why is it important, as a matter of federal law, for states to get fantasy-land federal law right? It's not.
* One other case that Louisiana cites, Oregon v. Guzek, is an even weaker case in this mold (a state statute permitted defendants to introduce evidence at death sentencing hearings that, and only that, the Federal Constitution requires defendants be allowed to introduce at death sentencing hearings), while the other cases parties cite are classic Michigan v. Long cases with state-law and federal-law claims where the Court held the state courts' judgments rested on the federal-law claims.