The Supreme Court recently granted certiorari in Montgomery v. Louisiana[1] to determine whether the Court’s holding in Miller v. Alabama, that “the Eighth Amendment forbids a sentencing scheme that mandates life in prison without possibility of parole for juvenile offenders,”[2] applies retroactively to cases on collateral review.[3] That question is important in its own right, as we have previously discussed.[4] But the Court also ordered argument on an additional, threshold question—one that, although perhaps less “sexy” than the merits question, may have profound implications for the scope of the Due Process Clause and retroactivity jurisprudence: Does the Supreme Court have jurisdiction over the case at all?[5] That is, does Montgomery’s claim, which was nominally rejected on state law grounds by the Louisiana Supreme Court,[6] even raise a federal question?

Before turning to that question, some background is necessary. The Court’s decision in Teague v. Lane[7] provides the modern framework governing retroactivity—that is, whether a decision announcing a “new” rule of constitutional law applies to defendants who were convicted before the rule’s articulation. Under Teague, new rules apply on direct review, but not on collateral review; thus, a case announcing a new rule applies only to those defendants whose convictions were not final when the rule was announced.[8] This rule of non-retroactivity on collateral review has two exceptions. Under Teague’s first exception, new substantive rules of criminal law—decisions that “narrow the scope of a criminal statute by interpreting its terms” or “that place particular conduct or persons covered by the statute beyond the State’s power to punish”—apply retroactively.[9] Importantly for our purposes, substantive rules include those that “plac[e] a certain class of individuals beyond the State’s power to punish by death,” because the Court has found such rules “analogous to . . . rule[s] placing certain conduct beyond the State’s power to punish at all.”[10] Under Teague’s second exception, “watershed” rules of criminal procedure apply retroactively.[11] Montgomery concerns Teague’s first exception.[12]

The next piece of the puzzle is the Court’s 2008 decision in Danforth v. Minnesota, which held that Teague’s background rule of nonretroactivity is not binding on the states because Teague merely construed the federal habeas statute.[13] Thus under Danforth, state courts are free to determine retroactivity using more generous standards than Teague’s, although the Danforth Court was careful to leave open whether states are constitutionally required to apply Teague’s two exceptions.[14]

Finally, a word about the reviewability of state court decisions. The U.S. Supreme Court possesses jurisdiction to review only those state court decisions that present a dispositive federal question. Put slightly differently, the Court lacks jurisdiction to review state court decisions that rest on “adequate and independent state grounds.”[15] Were the rule otherwise, the Court would issue an advisory opinion, because “the same judgment would be rendered by the state court after [the U.S. Supreme Court] corrected [the state court’s] views of federal laws.”[16] Not all state court decisions, though, are clear as to whether they are based on state or federal law. In those circumstances, the Court applies the Michigan v. Long presumption: if “a state court decision fairly appears to rest primarily on federal law, or to be interwoven with the federal law,” the Court presumes that the state court decision is based on federal law.[17] This presumption is overcome only by an explicit statement to the contrary.[18]

I. The Jurisdictional Issue in Montgomery and its Due Process Implications

This brings us to Montgomery v. Louisiana, which began when the petitioner, Henry Montgomery, filed a motion in Louisiana state court arguing that Miller was retroactive and thus entitled him to resentencing.[19] The state trial court denied his motion, and the Louisiana Supreme Court denied review, citing its decision in State v. Tate,[20] in which it held that Miller was not retroactive under Teague.[21] Tate, in turn, cited the Louisiana Supreme Court’s seminal retroactivity decision, State ex rel. Taylor v. Whitley, which adopted the federal Teague standards for “all cases on collateral review in [Louisiana] state courts.”[22] While “recogniz[ing] that [it was] not bound to adopt the Teague standards,” the Whitley court determined that Teague’s approach was desirable because it promoted clarity and respect for finality in criminal proceedings.[23]

This procedural background frames the jurisdictional issue. The Louisiana Supreme Court’s rejection of Montgomery’s postconviction motion rested on Tate and Whitley, two Louisiana state court decisions. Those decisions applied Teague, but the Whitley court explicitly stated that Louisiana was “not bound” by federal retroactivity standards,[24] raising the possibility that the Louisiana courts apply their own retroactivity law (which is permissible under Danforth) and look for federal law only for persuasive guidance.[25] If that is true, then arguably the Louisiana Supreme Court’s rejection of Montgomery’s motion does not present a federal question.[26]

This issue is closely intertwined with the issue left open in Danforth—whether Teague’s exceptions are binding on the states. As a matter of federal constitutional law, all courts are required to resolve the claims before them in accordance with the Due Process Clause.[27] Thus, if the Due Process Clause requires retroactivity for substantive rules, then the Louisiana Supreme Court’s allegedly erroneous failure to apply Miller retroactively in Montgomery presents a federal question. If, however, the Due Process Clause does not require the retroactivity of substantive rules, then Louisiana’s decision not to apply Miller retroactively was arguably a matter of state law and is thus unreviewable by the Supreme Court. While the Court could punt on the constitutional question and find jurisdiction under Long, the best course for the Court both doctrinally and jurisprudentially is to find federal jurisdiction on the grounds that Teague’s first exception is constitutionally required.

In our view, it is clear that the Due Process Clause requires the retroactivity of substantive rules on collateral review, and so Montgomery raises a federal question. As the Court noted in Foucha v. Louisiana, “[f]reedom from bodily restraint has always been at the core of the liberty protected by the Due Process Clause from arbitrary governmental action.”[28] Teague’s first exception provides for retroactivity where certain conduct may no longer be punished or a certain sentence may no longer constitutionally be imposed on a given class of individuals.[29] Because the continued imprisonment of an individual who cannot constitutionally be imprisoned would violate the Due Process Clause’s prohibition on arbitrary and unjustified governmental restraint, Teague’s substantive-rule exception must be constitutionally required. As the Seventh Circuit put it, “[i]f it would be unconstitutional to punish a person for an act that cannot be subject to criminal penalties it is no less unconstitutional to keep a person in prison for committing the same act.”[30] Or, as Justice Brennan explained, “a decision holding certain conduct beyond the power of government to sanction or prohibit must be applied to prevent the continuing imposition of sanctions for conduct engaged in before the date of that decision.”[31] Indeed, Justice Harlan, the father of the modern retroactivity doctrine, could not have been clearer that the first Teague exception applies to “‘substantive due process’ rules.”[32]

The Court’s amicus offers two arguments against this inescapable conclusion. Both, however, fail to account for the requirements of the Due Process Clause.

The first argument is that, under Danforth, the Teague decision was an exercise in statutory construction, and so Teague’s exceptions must not be constitutionally mandated.[33] This grossly overstates Danforth’s reasoning. Danforth held that states were not bound by Teague’s general rule of non-retroactivity; the Court was careful not to conflate that general rule with Teague’s exceptions.[34] Nowhere did the Danforth majority suggest that Teague’s exceptions were statutorily grounded. In fact, the Court stated that “[f]ederal law simply ‘sets certain minimum requirements that States must meet but may exceed in providing appropriate relief.’”[35] Thus, Danforth explicitly declined to resolve the question of whether states are bound by Teague’s exceptions and, if anything, suggested that the Constitution provides a “minimum” level of retroactive relief that is binding in all adjudications.

What is more, Teague’s first exception is different in kind from Teague’s background rule, given that substantive rules are not “exceptions” to Teague at all.[36] They “are simply ‘not subject to the bar’—that is, they apply to all convictions, period, no matter when the conviction became final.”[37] Even if Danforth had suggested that the Teague exceptions derive from the federal habeas statute, such a suggestion would still not support the amicus’s argument.

The amicus’s second argument is that the availability of federal habeas relief eliminates any constitutional problem with a state’s failure to allow the retroactive application of substantive rules. Because federal courts can grant retroactive relief, the argument goes, state courts are not required to do so.[38] But this argument confuses the availability of a forum for a claim with the substance of the claim itself. In constitutional terms, it addresses an argument under the Suspension Clause, rather than one under the Due Process Clause.

It is likely true that the Suspension Clause would still be satisfied if a state court refused to apply substantive rules retroactively. Under Boumediene v. Bush, the Suspension Clause does not insist on any particular vehicle for relief so long as “an adequate substitute for the writ of habeas corpus” exists.[39] Federal courts are surely adequate substitutes. Indeed, as the amicus observes, “[s]tate collateral proceedings are not constitutionally required as an adjunct to the state criminal proceedings.”[40]

But unlike the Suspension Clause, which apparently requires only a forum, the Due Process Clause applies in all fora, whether on direct review or collateral, in state court or federal. If it violates the Due Process Clause to continue to imprison an individual who has a valid claim under a retroactive rule, state courts are obligated to grant release.[41] States simply have no authority not to issue the relief required by the Due Process Clause, regardless of the availability of another forum. For this reason, the amicus’s greater-includes-the-lesser argument (because state habeas relief is not generally required, it need not be required for a particular type of claim) is a nonstarter. This is not how the Due Process Clause—or federal law more generally—works. A state is not required to hear habeas cases—but if it does, the Due Process Clause applies:

Even if a State has no constitutional obligation to grant criminal defendants a right to appeal, when it does establish appellate courts, the procedures employed by those courts must satisfy the Due Process Clause. Likewise, even if a State has no duty to authorize parole or probation, if it does exercise its discretion to grant conditional liberty to convicted felons, any decision to deprive a parolee or a probationer of such conditional liberty must accord that person due process. Similarly, if a State establishes postconviction proceedings, these proceedings must comport with due process.[42]

II. Michigan v. Long: A Flawed Way to Avoid the Due Process Clause Issue

Michigan v. Long[43] provides an alternate route to the merits that avoids these constitutional questions. Louis    iana has chosen to apply Teague, even though, under Danforth, it is not required to do so. The Louisiana Supreme Court’s determination that Miller is not retroactive thus rests exclusively on its interpretation of federal law. To the extent that the Long presumption is even needed, it compels a finding of jurisdiction because there is no indication in the decisions below, much less the required plain statement, that they rested on state law.[44]

The Court’s amicus again contends otherwise. The amicus argues that the Louisiana Supreme Court’s decision was not interwoven with federal law because the court “appl[ied] state law to retroactivity and us[ed] non-binding federal cases as persuasive authority.”[45] This argument rests on a false premise. Louisiana did not choose to apply its own state-law retroactivity standards and use federal cases as mere “persuasive authority.” Rather, it adopted the federal standard and applied that federal standard as a matter of state law.[46] Since the Louisiana Supreme Court made that choice, it must apply Teague correctly—just as the Michigan court in Long was required to apply federal precedents correctly, given its choice to rest its decision on federal, rather than state, search and seizure law.[47]

While this may be the easiest way to dispose of the jurisdictional question in Montgomery, it is not the best. Presumably, the Court granted certiorari to resolve a deep split among lower courts about Miller’s retroactivity. However, a holding that the Court has jurisdiction under Long would do little to resolve that split since, as the law currently stands (recall that it is uncertain whether Teague’s first exception binds the states), any state court that had misconstrued Miller’s retroactivity under Teague would not be bound by the Court’s decision. Indeed, any holding on the Miller question would not even be binding in Montgomery’s case, since the Louisiana Supreme Court could articulate a different retroactivity rule on remand. Furthermore, a holding predicated on jurisdiction under Long would provide no guidance to states that simply choose not to apply the Teague framework.[48]

The Court’s amicus recognizes that a finding of jurisdiction under Long would not be binding on state courts, but mistakenly argues that this result means that any interpretation of Teague issued by the Court in Montgomery amounts to an advisory opinion.[49] Not so. An opinion vacating a decision below is not an advisory opinion just because the lower court eventually reaches the same result on alternative grounds.[50] If the Supreme Court were to hold that Montgomery was entitled to relief under Teague’s first exception, a state court decision denying relief under a different retroactivity standard would be entirely consistent with the Court’s ruling unless and until the Court holds that the first exception is binding on the states.

Thus, although Michigan v. Long leads to the right result, it would be along the wrong path.


The Court clearly has jurisdiction in Montgomery. What is less clear, however, is the path the Court will take to reach the merits. It has two options: a broad holding resting (perhaps implicitly) on the Due Process Clause, or a narrow holding resting on Louisiana’s voluntary decision to apply Teague. The Court should choose the former and definitively resolve the split of authority on Miller’s retroactivity while also eliminating any misconceptions about the applicability of substantive rules to cases on collateral review in the state courts.

Jason M. Zarrow is an associate at O’Melveny & Myers, LLP. William H. Milliken is an associate at Kellogg, Huber, Hansen, Todd, Evans & Figel, PLLC (beginning November 2015).