Breaking News

Argument preview: Journey to the Twilight Zone: When does a state’s obligation to consider “clearly established” law end?

Rory Little is a Professor of Law at U.C. Hastings College of the Law in San Francisco, a former clerk to Justices Stewart (ret.) and Brennan, and author of the American Bar Association’s “Annual Review of the Supreme Court’s Term, Criminal Cases.”

Rod Serling would appreciate the intricate issue presented in Greene v. Fisher, set for the afternoon argument on Tuesday, October 11.  (See the end of this post for a link to the theme music.)  What happens on federal habeas if the U.S. Supreme Court issues an opinion settling an open question of law during the “twilight zone” period between the last state-court merits decision on the question and the time the state conviction becomes “final” on direct appeal?  An intermediate state appellate court is often the last state court to issue a merits decision in criminal cases.  Yet “finality” may not occur for months or even years, when a state supreme court denies discretionary review and the time to seek certiorari in the U.S. Supreme Court expires (or, alternatively, cert. is denied).  It is settled (thanks to Teague v. Lane (1989) and Griffith v. Kentucky (1987)) that a new U.S. Supreme Court decision cannot normally be applied to decisions that are “final.” But is application of such a new decision further barred on federal habeas to a state criminal defendant, if the decision is issued before direct-review finality but in the “twilight zone” period after the last state-court merits decision?

Such accidents of timing are rare, but Greene’s case clearly presents the situation.  An intervening U.S. Supreme Court decision that was issued after a state court initially affirmed his conviction, but before the state supreme court denied review, appears to plainly implicate the validity of his conviction.  Teague says such an intervening decision must be applied, but does a Section of the Antiterrorism and Effective Death Penalty Act (AEDPA), 28 U.S.C. 2254(d)(1), require otherwise?  By a vote of two-to-one, the Third Circuit here said yes, while the First and Ninth Circuits have said no.  Thus Greene is a significant federal habeas case, and if decided broadly, it has significant implications for statutory construction in general.

Facts:  Timing is everything

Like most federal habeas cases, the procedural facts are lengthy.  Fortunately, they are also straightforward and largely undisputed.  But you have to play close attention to the timing of various courts’ decisions.

Eric Greene (known for unexplained reasons as “Trice” in his Pennsylvania proceedings) was convicted of second-degree murder in Pennsylvania at a joint trial with four codefendants.  In a robbery gone wrong, Greene was undisputedly not the shooter, but was identified by some of his codefendants as participating in the robbery in various ways (their confessions were in conflict as to precisely how).  Greene’s defense was mistaken identity.  He sought severance from his codefendants due to their statements that implicated him, relying on the Court’s 1968 decision in Bruton v. United States, which held that the Sixth Amendment’s Confrontation Clause prohibits the admission of a non-testifying codefendant’s confession against a defendant whom the confession also implicates.

The trial judge, however, denied severance and ordered instead – and not inconsistently with the law at the time – that the codefendants’ statements be “redacted” to omit any mention of Greene by name.  The government complied but – and again, not inconsistently with the law at that time – replaced Greene’s name at certain points with the word “blank” or similar symbols, making it clear that redaction of someone’s name had occurred.  (The state argues that Greene never clearly presented any objection to this manner of redaction, and in fact explicitly said “agreed” when it was proposed.  The Third Circuit found, however, that Greene “fairly presented” his Confrontation Clause claim to the state courts, and that the Superior Court “addressed the merits.”  This “waiver” issue seems not to be presented in the questions on which the Court granted cert., although some Justices generally opposed to federal habeas review of state criminal convictions might disagree.)
In any case, at Greene’s 1996 trial the co-defendants’ redacted statements were introduced.  The codefendants did not testify (and thus could not be cross-examined about their statements), and Greene was convicted and sentenced to life imprisonment.

Those familiar with the progeny of Bruton can already imagine the rest of the story.  On appeal, the Pennsylvania Superior Court rejected Greene’s Bruton argument, noting that “the substitution of the letter ‘X’ for a defendant’s name does not violate that defendant’s Bruton rights.”   Greene then timely petitioned for discretionary review to the Pennsylvania Supreme Court.  While that petition was pending, the U.S. Supreme Court decided Gray v. Maryland (1998), ruling that “as a class, redactions that replace a proper name with an obvious blank, … or similarly notify the jury that a name has been deleted” require the “same legal results” as a direct Bruton violation.  It is at least a substantial question whether Gray would have required a different result in Greene’s case.

Perhaps unsurprisingly, the Pennsylvania Supreme Court granted Greene’s petition for review.  Greene then filed a brief expressly invoking Gray.  But the government responded that Greene had never fairly presented the redaction issue below.  A few months later the Pennsylvania Supreme Court dismissed Greene’s petition as “improvidently granted” without explanation.  Greene’s appointed attorney then notified Greene that his representation of Greene was done, and in July 1999 the time for filing a petition for certiorari with the U.S. Supreme Court expired, rendering Greene’s state court conviction “final.”

Greene then filed, pro se, a Pennsylvania state collateral action, arguing that severance should have been granted but not raising his specific Bruton (and now Gray) “redaction” issue.  The parties now dispute the effect of this – Greene argues that Pennsylvania law prevented him from presenting an issue already raised and rejected on direct appeal, while the state argues that this was a further waiver of the issue.  (As noted above, the Third Circuit rejected these procedural-bar arguments.)  In any case, Greene’s state collateral petition was denied, and on appeal the Pennsylvania Superior Court affirmed, noting that his severance claim had been finally litigated on direct appeal and so could not provide a ground for collateral relief.  The Pennsylvania Supreme Court affirmed without opinion.

Greene then filed a timely federal habeas corpus petition.  The 1996 AEDPA amendments to the federal habeas statutes set a high standard of review for such petitions when they seek to overturn a state court criminal judgment.  The relevant portion of the provision at issue, 28 U.S.C. § 2254(d), directs that (emphasis added) such a petition “shall not be granted unless the adjudication of the claim (1) resulted in a decision that was contrary to . . . clearly established Federal law, as determined by the Supreme Court of the United States.”

The federal magistrate judge handling Greene’s petition determined that although Gray was “clearly established federal law,” it could not be applied to Greene’s state court judgment, because it had been rendered after the relevant merits decision in Pennsylvania – the decision of the Pennsylvania Superior Court.  The judge found that a federal court on habeas should consider only the law as it existed at the time the state court ruled on the merits of the claim.   Even though Gray had been issued before the Pennsylvania Supreme Court had disposed of Greene’s direct appeal, that court’s grant-then-DIG disposition was not a “precedential” decision under state law, and thus Gray was not “established” at the time of the last merits decision of the Pennsylvania courts.  Finally, under Bruton without Gray, the Pennsylvania court’s decision regarding redactions was not “an unreasonable application” of clearly established federal law.

The magistrate judge’s recommendation was adopted by the district court, and the Third Circuit, in a lengthy two-to-one decision, affirmed.  The Third Circuit noted that various U.S. Supreme Court opinions (and various separately writing Justices) have issued “conflicting guidance” on the question.  Moreover, the First Circuit had reached the opposite decision on this exact question (and the Ninth Circuit has agreed with the First).  Thus, although relatively rare, the situation arises often enough that a grant of certiorari was appropriate.


If the correct aphoristic reference is that the Supreme Court usually grants certiorari to reverse, then Greene should receive the benefit of Gray.  However, the Court’s various statements on the issue do not lend themselves to confident prediction without doubt.  Given Justice Scalia’s talent, in particular, for insightfully recasting issues once certiorari has been granted, this is a case in which oral argument may be particularly juicy.  We may miss Justice Stevens, who often served as Justice Scalia’s quick-witted repartee “foil” in such cases at oral argument, although the new Justices (Sotomayor and Kagan) have not yet fully settled into a predictable oral argument style.

Interestingly, statements from two recently retired Justices in the very same case appear to drive the current confusion.  In William v. Taylor (2000), Justice O’Connor wrote for the Court that “the statutory phrase [‘clearly established Federal law’] refers to the holdings, as opposed to the dicta, of this Court’s decisions at the time of the relevant state-court decision” (emphasis added).   Yet Justice O’Connor also wrote – in the same paragraph in Williams – that “whatever would qualify as an old rule under our Teague jurisprudence will constitute ‘clearly established Federal law …’ under § 2254(d)(1)” (emphasis added again).  There is no doubt that an “old rule” under Teague is one announced in a U.S. Supreme Court decision issued before a state court judgment is “final” on direct review – just as Gray was here.  Finally in Williams, Justice O’Connor also joined the part of Justice Stevens’s opinion for the Court in which he wrote that “the threshold question under AEDPA is whether [the petitioner] seeks to apply a rule of law that was clearly established at the time his state-court conviction became final” (emphasis added one more time).  (Yes, Williams is one of those great modern Supreme Court decisions where there are two different opinions for the Court, written by two different Justices, for different majorities on different aspects of the case.)  Significantly, the question now presented in Greene was decidedly not before the Court in Williams, thus rendering all the statements on the question pure dicta.  Yet dicta becomes doctrine when five Justices adopt it.

Greene’s Arguments

Greene (now represented by Jeff Fisher and Pam Karlan of Stanford’s Supreme Court Litigation Clinic) presents two arguments.  His primary argument is that there is no evidence that Congress, in enacting Section 2254(d)(1), sought to change existing “retroactivity” law, and that it is clear that the Supreme Court in Teague and Griffith has said that new decisions of the U.S. Supreme Court should apply to all criminal cases not yet “final” on direct review.   A ruling that Section 2254(d)(1) pushes back the date for application of new decisions, to whatever time a lower state court rules on the merits, would be a “sub silentio” legislative amendment of Teague, something that is strongly disfavored as a general matter of statutory construction.

Greene argues secondarily that “even if” § 2254(d)(1) contains a silent amendment of the retroactivity law of Teague, a decision of a state supreme court to deny discretionary review should be the “decision” to which § 2254(d)(1) refers; and that such a decision “resulted” from misapplication of clearly established federal law whenever the decision to deny review is announced after a law-changing U.S. Supreme Court decision has issued.

Greene argues a number of subsidiary points.  He invokes “rudimentary notions of fairness” to object to an interpretation of the statute that would prevent a defendant from ever receiving the benefit of a “new rule” issued by the Supreme Court during the “twilight” period before finality but after a lower court merits decision.  He points out that such a ruling would place additional workload pressures on the U.S. Supreme Court itself, in that the Court would have to become alert to certiorari petitions filed from direct state court criminal denials of review, in order to detect “twilight zone” questions.  Green argues that the Court would then have to use its GVR authority (“grant, vacate and remand”) to ensure that state courts apply new decisions consistently with Teague.  And – in what appears to be an aggressive forward-looking argument designed to force the Court to consider the long-range implications of an adverse ruling – Greene argues that the Court would have to extend its “right to counsel” doctrine to petitions for certiorari in such cases, because (Greene says) the State’s argument would make such petitions “the first, and likely the only, direct review” that the new decision would receive (quoting Halbert v. Michigan (2011).  This argument is likely to “scare” some of the Justices, as the Court has long held that there is no right to counsel for certiorari petitions.

Pennsylvania’s arguments

The state here argues that it is only common sense to rule that state courts need consider only the “established law as it existed at the time the state court ruled on the [defendant’s] claim.”   This is consistent with Congress’s apparent intent, in enacting Section  2254(d), to “prevent the use of habeas review to blindside state court judges with evidence and rules that were not before them at the time” they ruled.  Indeed, just last Term, in Cullen v. Pinholster, the Court ruled that a federal habeas petition must be reviewed on the evidentiary record that was before the state court at the time it ruled, rather than allow it to be supplemented with new evidence in the federal proceeding.

The state argues that Section 2254(d)(1) is a “modified form of res judicata,” and as such, naturally may result in applying a cut-off of claims beyond the doctrine of retroactivity.  The state argues that Congress need not leave a “trail of statutory history crumbs” to achieve this result – the meaning is plain from the statute.  And Teague will still apply where the state courts have rendered “no adjudication on the merits” (emphasis added).  Congress need not specify every doctrine arguably affected when it enacts a major statutory revision.

The state argues that “clearly unreasonable” in the statute naturally presumes that state court decisions will be evaluated based on the law in place at the time they rule.  The state points out that, in the ineffective assistance context, the Court has consistently said that an attorney’s conduct must be evaluated under “the norms prevailing when the representation took place.”

Moreover, “adjudication on the merits” as used in § 2254(d) necessarily precludes reliance on a denial of discretionary review that is not “on the merits.”  (Your reporter, however, believes that this argument misplaces the phrase within the statute.  The statute says that it is the “decision” that must be evaluated under the “contrary to or clearly unreasonable” standard of Section 2254(d)(1), not the adjudication.  The state, however, argues that Congress did not intend a difference, and appropriately quotes four prior habeas decisions in which the Court used the word “adjudication” rather than “decision.”)  It would be a denial of “comity,” the State argues, to deny state supreme courts the same non-merits discretionary power to deny review that the U.S. Supreme Court routinely exercises.

Indeed, the State closes by deriding Greene’s brief as mischaracterizing state courts as “inferior,” “evil,” and “ignorant.”  But this is at worst an outdated, and more likely unfair, description of state supreme courts.  Indeed, the Pennsylvania Supreme Court acted perfectly fairly and reasonably here.  After it granted review, it learned from the state’s opposing brief that Greene had not actually presented his Gray claim to the lower state courts.  Indeed, Greene had said “agreed” when the trial judge explained her redaction ruling, and called it a “brilliant” solution.  And then Greene did not object to the specific redactions on appeal.  Thus (the state argues), the Pennsylvania Supreme Court acted appropriately in ultimately denying its discretionary review.  Greene’s “aspersions” cast on state courts are unwarranted.

Ultimately, the state argues, all rules cutting off review will, in some cases, appear to yield disparate results.  In fact, of the five codefendant convictions in Greene’s case, two were final before Gray, while two others were adjudicated on the merits after Gray.  Only Greene’s case was presented in the middle period.  This demonstrates that “disparity is inevitable” – it is simply in the nature of time cutoffs.  There is nothing exceptionally unfair about following Congress’s plain language to cut off review in the very small number of cases like Greene’s, particularly since state supreme courts can be counted on to enforce new rules when the defendant has actually presented a relevant issue below.

Final Word

Greene ultimately presents this case as a major one in the general field of statutory construction.  The Court has said many times that Congress will not be construed to have significantly altered federal law absent “specific” and “clear” indications.  If Section 2254(d)(1) is found to have silently altered the Teague retroactivity rule, fifteen years after the statute was enacted, it will be a major break with this general doctrine.

The state depicts the case completely differently: Section 2254(d)’s language is “clear” and Congress surely intended the commonsense result that state courts that rule on the merits will not later be found to have misapplied “federal law” that was not established at the time they ruled.  This is not a conflict with, but rather a supplement to, Teague, and it creates no particular unfairness.  Meanwhile, both comity and finality require the result.

Now is the time to cue the “Twilight Zone” theme music. One only wishes that Justice O’Connor were here to explain what she really meant in Williams.  But legal doctrine (as well as dictum) has a way of outliving its authors, and even though Justice O’Connor is still very much alive, it is the current Justices who must determine what Congress meant.  The Court’s previous statements point in both directions, and none of those statements were “holdings” requiring stare decisis effect.

Indeed, perhaps most relevant is a prior statement made by Justice O’Connor in a very different context, the “takings” case of Lingle v. Chevron (2005) (to which Greene’s brief refers at page 34):  “On occasion, a would-be doctrinal rule or test finds its way into our case law through simple repetition….”   In Lingle, the Court referred to such a test as “regrettably imprecise” and ruled that it was not “appropriate” once different and precise facts were presented.  Next Tuesday’s Court will have to decide whether a very different Court a decade ago was “imprecise,” or correct, in its Williams dicta.

[Disclosure:  Goldstein & Russell, P.C. serves as co-counsel to the petitioner in this case.  Some of the attorneys in that firm work for this blog in various capacities, but the author of this post is not associated with either the firm or any parties or amici to this case.]

Recommended Citation: Rory Little, Argument preview: Journey to the Twilight Zone: When does a state’s obligation to consider “clearly established” law end?, SCOTUSblog (Oct. 8, 2011, 8:00 AM),