When the Supreme Court creates a new rule of federal constitutional law, to what extent is the retroactivity of that rule a federal question? On March 20, the Supreme Court asked for further briefing on that issue in its consideration of the pending certiorari petition in Danforth v. Minnesota, No. 06-8273. At SCOTUSblog, Lyle Denniston has this post and the state's brief.

Many rules in this area are already established. Rules governing substantive criminal law such as what acts can be made criminal and what punishments can be imposed for which crimes and on which defendants are fully retroactive. Thus, cases such as Lawrence v. Texas, 539 U.S. 558 (2003) and Atkins v. Virginia, 536 U.S. 304 (2002) have no retroactivity limitation. New rules of federal constitutional criminal procedure, on the other hand, will apply retroactively on direct review, Griffith v. Kentucky, 479 U.S. 314 (1987), but under the rule of Teague v. Lane, 489 U.S. 288 (1989), they do not apply on habeas corpus to cases that were already final on appeal when the new rule was created.

In theory, there is an exception to Teague for new rules that have the "the primacy and centrality of the rule adopted in Gideon..." [v. Wainwright, 372 U.S. 335 (1963)]. Saffle v. Parks, 494 U.S. 484 (1990); Whorton v. Bockting, slip op. at 14 (2007). How many new rules of Gideon magnitude have been created since 1989, and how many remain to be created? Zero and almost certainly zero.  This exception was born comatose, and the Court has come tantalizingly close to declaring it brain-dead, see Bockting, slip op. at 10, but it hasn't quite yet.

On both direct and state collateral review, the Supreme Court has reversed state courts for not giving the defendant the benefit of a rule that federal law requires be applied retroactively. Griffith was a direct review case, and Yates v. Aiken, 484 U.S. 211 (1988) reversed a denial of state collateral relief on nonretroactivity grounds. The state has no constitutional obligation to provide collateral review at all, but if it does (and all states do), it must give federal rules the retroactive application required by federal law, at least for the benefit of the defendant.

But if a state court decides to apply a new federal rule retroactively where Teague does not require it, is that decision reviewable by the Supreme Court as a federal question? Minnesota relies on the pre-Teague case of Michigan v. Payne, 412 U.S. 47 (1973) and on sweeping language from a civil retroactivity case, American Trucking Associations, Inc. v. Smith, 496 U.S. 167, 177 (1990).

The contrary view would likely be based on the idea that federal review of state decisions is for the purpose of protecting the citizens of the state from the state government, not for the purpose of protecting the state itself from its own courts. Justice Stevens has long asserted this view. See, e.g., his dissent in Michigan v. Long, 463 U.S. 1032 (1983). Congress originally gave the Supreme Court federal question jurisdiction in state cases only where the state court decision was against the federal right asserted, but that limitation is long gone. Federal question jurisdiction under 28 U.S.C. § 1257 is now symmetrical, giving the Court as much authority to correct state decisions that give overly expansive interpretations of federal rights as it has to correct overly restrictive interpretations. Justice Stevens did not prevail in Long or in the many cases since where the Court has exercised that authority in criminal cases. If the Court does take the case, I think Minnesota is likely to prevail on the merits.