<< News Scan | Main | SCOTUS Today and Next Week >>

Extending versus applying "clearly established Federal law"

In the case of White v. Woodall, to be argued next Wednesday, December 11, the U.S. Supreme Court is finally squarely faced with a situation that it has only mentioned in passing remarks (obiter dicta, in legal Latin) until now.  How does a federal court apply the so-called "deference" standard of 28 USC §2254(d)(1) when the "clearly established federal law" that the habeas petitioner seeks to invoke was established in a different context?  Can he stitch together pieces of cases into a legal quilt to extend the rule he needs to new territory and ask the federal court to declare the state court's refusal to do so "unreasonable," so as to satisfy Congress's standard for an override of the state decision in federal court?

I think the answer is "no," and despite some unfortunate dicta I think it is quite clearly "no."  Allowing this "extension" analysis would undermine the very core of the purpose of §2254(d), as the Supreme Court recognized in Yarborough v. Alvarado.

CJLF's brief is here.  The full set of briefs is on SCOTUSblog's case page.  Some further thoughts on the arguments of the defendant and NACDL follow the break.
    In the main merits brief, the State [the certiorari petitioner, in the Supreme Court] argued that if existing law needs to be extended to new terrain to apply to the present case, that extension is necessarily not "clearly established"and cannot form a ground for relief under §2254(d)(1).  See Pet. Brief 41-42.  Amicus curiae Criminal Justice Legal Foundation made a more extended version of essentially the same argument, focusing particularly on the extension language from Green v. French, 143 F.3d 865, 869-870 (CA4 1998), discussed but not endorsed in Williams v. Taylor, 362 U.S. 362, 407-408 (2000).  See Brief for the Criminal Justice Legal Foundation as Amicus Curiae 10-16 ("CJLF Brief").  Defendant [the certiorari respondent, in the Supreme Court] invokes the straw-man fallacy to attack this argument, claiming that CJLF has argued that a case directly on point is required, see Resp. Brief 22, when the CJLF brief expressly says the opposite.  See CJLF Brief 14.

    Amicus National Association of Criminal Defense Lawyers engages in a more extreme straw-man fallacy, railing against an "asymmetrical" interpretation of §2254(d)(1).  See NACDL Brief 9.   No such interpretation has been advanced in this case.  CJLF's proposal is for the unadorned version of the rule in Williams, at 412-413, and NACDL fails to identify anything "asymmetrical" about this statement of the rule.

    Defendant contends that it is a "well-established principle" that a state court decision declining to extend existing precedent can be set aside under the "unreasonable application" prong of §2254(d)(1), citing a dozen cases, Resp. Brief 21-22 & n. 2, yet he fails to cite even a single case where a state court decision actually was set aside on that basis.  The plurality opinion in Ramdass v. Angelone, 530 U.S. 156, 166 (2000), repeats the language from Williams, but the actual holding is that the state court's refusal to extend was reasonable with a strong implication it was correct, see id., at 169, so there was no need to give close scrutiny to the question at issue here.  Ramdass's repetition of the extension language is dictum, not holding, and in any event it is a plurality opinion, not the opinion of the Court.

    Every one of the circuit decisions Defendant cites is also a case where the extension language ultimately makes no difference in the outcome of the case.  See Kibbe v. DuBois, 269 F.3d 26, 38 (1st Cir. 2001) (credible interpretation of federal precedent); Bierenbaum v. Graham, 607 F.3d 36, 49-50 (2d Cir. 2010) (reasonable application to facts of general standard in effect at the time); id., at 59 (state court correctly applied ineffective assistance standard); Williams v. Price, 343 F.3d 223, 229-230 (3d Cir. 2003) (precedent cited by petitioner involved different issues);  Booth-El v. Nuth, 288 F.3d 571, 578 (4th Cir. 2002) (fourth ex post facto category was not clearly established at the time of the state court decision);  Penry v. Johnson, 215 F.3d 504, 508-509 (5th Cir. 2000), aff'd in part, rev'd in part, 532 U.S. 782 (2001) (state court was correct in Court of Appeals' view); Campbell v. Coyle, 260 F.3d 531, 545-546 (6th Cir. 2001) (fact-bound case, reasonable application of existing standard to facts); Armstrong v. Bertrand, 336 F.3d 620, 625 (7th Cir. 2003) (cited precedent provides little support for argument); id., at 626 (defense unsupported by evidence); id., at 628 (cited precedents too far removed to be applicable); Carter v. Kemna, 255 F.3d 589, 592 (8th Cir. 2001) (state court's resolution of unresolved question was reasonable); Parker v. Scott, 394 F.3d 1302, 1311 (10th Cir. 2005) (application of broad due process standard to particular facts); id., at 1314 (same); id., at 1317 (same); id., at 1319 (state court prerequisite of claim missing); id., at 1320-1321 (application of correctly selected Strickland rule to particular facts); Kimbrough v. Secretary, DOC, 565 F.3d 796, 799 (11th Cir. 2009) (same).  In Miller v. Blackletter, 525 F.3d 890, 895-896 (9th Cir. 2008), the court says it is going to apply the extension language, but what is actually does is apply a broad, multiprong standard, see id., at 895, to the facts of the case, see id., at 896-897, and hold that the state court's decision was reasonable.  See id., at 898.  That is identification of the governing principle and application to the facts, see Williams, 529 U.S., at 413, in its pure form, making the "extension" language superfluous.

    Whether the "unreasonable application" clause of §2254(d)(1) has any "extension" aspect to it is therefore an open question.  It is most certainly not true, as Defendant contends, that a negative answer would leave the clause without meaning.  See Resp. Brief 11.  It would leave the clause with its primary meaning, the meaning it has had in nearly every case where it has been invoked to grant relief, the meaning stated in Williams, 529 U.S., at 413.  It applies when the state court has correctly identified the clearly established rule applicable to the situation from this Court's precedents but the petitioner claims that the state court unreasonably applied that rule to the particular facts of his case.

    The distinction between application of an existing rule to particular facts and the extension of a rule to new terrain has been probed more thoroughly in the retroactivity cases under Teague v. Lane, 489 U.S. 288 (1989), where the distinction more clearly controls the outcome.  Teague cases are pertinent to §2254(d)(1) cases, and vice versa, because Teague's "new rule" is essentially the inverse of §2254(d)'s "clearly established."  See Williams, 529 U.S., at 412.  Just last term, the Supreme Court invoked this inverse relationship to consider §2254(d) precedents in a Teague case.  See Chaidez v. United States, 133 S.Ct. 1103, 1107-1108 & n. 4 (2013).

    Teague cases rejecting efforts to extend existing law or refusing efforts to apply retroactively an extension made in another case are numerous.  For example, in Caspari v. Bohlen, 510 U.S. 383, 392-393 (1994) the Court rejected an attempt to extend two double jeopardy precedents from capital sentencing into noncapital sentencing.  "While our cases may not have foreclosed the application of the Double Jeopardy Clause to noncapital sentencing, neither did any of them apply the Clause in that context."  Id., at 393.  The extension would therefore have been a new rule.  

    Chaidez, supra, is the most recent example.  Strickland v. Washington, 466 U.S. 668 (1984) is the primary example of a broadly-worded rule applicable to "a myriad of factual contexts."  Id., at 1107 (quoting Wright v. West, 505 U.S. 277 (1992) (Kennedy, J., concurring in the judgment).  In almost all ineffective assistance cases, Strickland provides a rule that is "clearly established" for §2254(d)(1) and not "new" for TeagueChaidez was different, though, because the precedent involved, Padilla v. Kentucky, 559 U.S. 356 (2010) was an extension rather than an application of Strickland.
"In other words, prior to asking how the Strickland test applied ('Did this attorney act unreasonably?'), Padilla asked whether the Strickland test applied ('Should we even evaluate if this attorney acted unreasonably?'). And as we will describe, that preliminary question about Strickland's ambit came to the Padilla Court unsettled--so that the Court's answer ('Yes, Strickland governs here') required a new rule."  Chaidez, 133 S.Ct., at 1108 (emphasis in original).
    The Teague analysis ends with the finding that Padilla was a new rule, id., at 1113, equivalent to finding that it was not "clearly established."  What if a state prisoner had raised the same claim in federal habeas corpus, and the state had asserted §2254(d)(1) in defense but not Teague?  Would the habeas court be required to take the additional analytical step of deciding whether failure to extend was unreasonable?  That would make no sense.  Congress surely did not intend to make §2254(d) less restrictive than Teague.  See Williams, 529 U.S., at 410-412.*  If it were interpreted in that way, states could just slap on a Teague defense in addition to §2254(d) every time the habeas petitioner asserts an extension of clearly established law, and the end result would be the same.

    Defendant has demonstrated no convincing reason for retaining the additional extension analysis step from Green v. French.  The straightforward analysis described in Williams, at 412-413, is sufficient without adornment and correctly reflects the intent and language of the statute.
* On rare occasions, the Teague rule is more restrictive in practice because of timing differences in the rules.  See Beard v. Banks, 542 U.S. 406, 409-410 (2004).


When Musladin was decided, I'd hoped this extension issue would be put to rest. Musladin's holding--that a petitioner needs an earlier Supreme Court holding to prevail--seemed pretty strong.

Is it fair to say that the AEDPA relitigation bar is a limitation on the otherwise broad common-law powers of the federal appellate courts?

In law school, for example, they teach everyone that the way the common law works is for judges to (1) identify all factually similar holdings to the case at bar (2) distill some principle of law from those holdings, and (3) apply that new holding to the litigants. But AEDPA, even more than the rule of Teague v Lane, stops this common-law reasoning process before it starts (yes?).

If so, it must be difficult for an appellate judge to stop short like that on an issue. It would take self control to recognize that the habeas case is over when the State identifies one or two holdings from reported decisions, even from outside the jurisdiction, that go against the inmate. See also Price v. Vincent at n.2. It's likely even harder for a clerk who's only got a year or two to help make some new law.

Yes, use of habeas corpus to further extend the law is supposed to be shut down by AEDPA. The purpose of habeas under AEDPA is to correct clear errors by the state courts under current law, period.

In my view, §2254(d) is properly viewed as a modified rule of res judicata, not "deference." The defendant has had his bite at the apple in state court, and the general rule is that a litigant only gets one bite. For incarcerated prisoners, we make a partial exception to the rule in the case that the decision on the previous bite is so clearly wrong as to be outside the bounds of reasonable debate. As an error of this magnitude rarely happens and is easily recognizable when it does, federal habeas corpus cases should take very little time when the issues have already been decided in state court.

Leave a comment

Monthly Archives