Go to:	Parts IV and V	Case Menu	Briefs Menu

III. "Contrary to" governs rule selection, "unreasonable application" governs rule application, and  "clearly established" qualifies both.

"Thus, as we read the statute, a court evaluating a habeas petition under § 2254(d)(1) must engage in a three-step process: First, the court must 'survey the legal landscape,' using an inquiry similar to that under Teague, to ascertain the federal law applicable to the petitioner's claim that is 'clearly established' by the Supreme Court at the time of the state court's adjudication. Second, the court must determine whether the state court adjudication was contrary to the clearly established Supreme Court case law, either because the state court failed to apply the proper Supreme Court precedent, or because the state court reached a different conclusion on substantially similar facts. If the state court's decision is not contrary to law, the reviewing court must then determine whether the state court unreasonably applied the relevant Supreme Court authority. The state court decision must stand unless it is not debatable among reasonable jurists that the result of which the petitioner complains is incorrect."

This statement of the rule is consistent with the language, history, and intent of the statute, and it provides a straightforward framework for analyzing habeas claims under the statute. With a few qualifications and explanations, amicus submits it should be adopted.

Section 2254(d)(1) contains two branches which share a common trunk. The trunk is the requirement that the rule in question be "clearly established" by Supreme Court case law. The term "clearly established" appears in Teague case law in Sawyer v. Smith, 497 U. S. 227 (1990), which borrows it from the qualified immunity cases. Id., at 236 (quoting Anderson v. Creighton, 483 U. S. 635, 639 (1987)).

Although the new statute is not a rule of retroactivity, it does share this element with the Teague line. "Clearly established" requires a certain degree of specificity of the rule of law to be applied. The principle of reliability in capital sentencing, for example, is not specific enough to be an established rule. See ibid.

If the determination of whether there is a clearly established Supreme Court precedent governing the case is a Teague-like inquiry, as the borrowing of the term from Sawyer strongly implies, then it makes sense that this determination involves the same step of the decision-making process as Teague. There are "three distinct functions" in this process: "law declaration, fact identification, and law application." See Monaghan, Constitutional Fact Review, 85 Colum. L. Rev. 229, 234 (1985). Teague, being a doctrine of retroactivity, necessarily addresses rules, i.e., law declaration rather than law application. See Wright v. West, 505 U. S. 277, 307 (1992) (Kennedy, J., concurring in the judgment).

While "contrary to law" review in some contexts encompasses both law declaration and law application, see Brief for Petitioner 34, n. 17, in this case Congress has created a separate clause for application. Amicus ACLU claims support from the fact that the Senate's modification of the House language lacks the prepositional phrase "to the facts" after the word "application." ACLU Brief 29. This, they would have the Court believe, worked a complete transformation of the subject matter of the second branch, changing the subject from application of law to fact into an unrelated inquiry about deriving new rules from existing ones. See also Liebman & Ryan, supra, 98 Colum. L. Rev., at 871-872.

There are several answers to this contention. First, the attachment of the prepositional phrase "to the facts" is not essential to the meaning of "application" in this context. Justice Kennedy's opinion in Wright v. West, supra, 505 U. S., at 308, refers to "specific applications" without adding "to the facts." Professor Monaghan's influential article, see supra, at 20, likewise refers to "law application" without appending "to the facts." Indeed, Professor Monaghan himself interpreted this language as applying to mixed questions, and Senator Cohen read his letter to that effect on the floor of the Senate. 141 Cong. Rec. S7839 (daily ed. June 7, 1995).

Second, none of the people involved in the transformation from the House to the Senate version saw the deletion of "to the facts" as transforming "application" from mixed questions to something else. The language emerged from "very extended negotiations" between Senator Specter and Senator Hatch, id., at S7803, both of whom confirmed that the bill applied a reasonableness standard to the state court's application of the law to the facts. 142 Cong. Rec. 7772-7773 (1996) (statement of Sen. Hatch); id., at 7799 (statement of Sen. Specter). While "application" can have a different meaning, the people who made the change from the House to the Senate version did not intend one.

Third, a rule that lower federal courts could overturn state decisions on mere disagreement with their application of established law, while the zone of disagreement on law declaration is protected from such interference, would be incompatible with the basic nature of §2254(d) as a rule of prior adjudication. The underlying premise is that one bite at the apple, with review up the appellate chain, is normally all a litigant gets, with exceptions to prevent injustice when the normal mechanism seriously malfunctions.

This premise is consistent with this Court's "long history of distinguishing between collateral and direct review [citation] and confining collateral relief to cases that involve fundamental defects or omissions inconsistent with the rudimentary, demands of fair procedure [citations]." Brecht v. Abrahamson, 507 U. S. 619, 640 (1993) (Stevens, J., concurring) (emphasis added). This was precisely Senator Hatch's point. "After all, Federal habeas review exists to correct fundamental defects in the law. After the State court has reasonably applied Federal law, it is hard to say that a fundamental defect exists." 142 Cong. Rec. 7772 (1996). Collateral attack on a final judgment is a drastic remedy, reserved for egregious errors. Simple disagreement on close questions regarding the application of settled law to varying facts does not rise to that level.

Unlike retroactivity, nothing in the principle of respect for prior adjudications by coordinate courts excludes that principle from the law application portion of the decision. "The comity interest [in Teague] is not, however, in saying that since the question is close the state-court decision ought to be deemed correct because [the lower federal courts]⁽⁴⁾ are in no better position to judge. That would be the real thrust of a principle based on deference." Wright v. West, supra, 505 U. S., at 308 (Kennedy, J., concurring in the judgment). That is indeed the real thrust of §2254(d). See supra, at 4, 8.

The First Circuit in O'Brien v. Dubois, 145 F. 3d 16, 22 (CA1 1998), objected that reading this branch to refer to mixed questions was "embroidery of the statute's text." In reality, it is right there in the word "application." Furthermore, this interpretation does not scavenge language discarded by the Senate from the House version, because the Senate kept the key operative word, "application," discarding only a prepositional phrase the negotiators evidently considered surplusage.

As the Third Circuit noted, the Fourth Circuit's "catalogue" of various situations warranting "contrary to" or "unreasonable application" treatment would be difficult to apply in practice. Matteo v. Superintendent, 171 F. 3d 877, 888 (CA3 1999) (en banc). Yet the First/Third Circuit approach has a similar difficulty. The line between rules specific enough for "contrary to" treatment versus those that warrant "unreasonable application" is not well defined. See supra, at 13-16.

Amicus CJLF submits that the simple approach is the best one. "Contrary to" applies to law declaration, as Professor Monaghan called it, or "pristine legal standard[s]," in the words of Miller v. Fenton, 474 U. S. 104, 114 (1985). "Unreasonable application" applies to "law application," also known as "mixed questions of law and fact." Identifying the rule of law and then applying the law to the facts is a familiar two-step process, which all members of the profession have been doing since law school. Analyzing habeas claims in this way will present the fewest difficulties.

Some decisions have indicated that this bifurcation approach applies a radically different level of scrutiny to the two aspects of the state decision. See Matteo, 171 F. 3d, at 887. Properly applied, it does not. "Contrary to" must be read in conjunction with the "clearly established" requirement. Because this is a Teague-like inquiry, it incorporates the principle of Butler v. McKellar, 494 U. S. 407, 415 (1990) that susceptibility to debate among reasonable minds is enough to render a rule "new" for Teague and not "clearly established" for §2254(d)(1). Thus, the basic test of reasonableness applies to both branches of §2254(d)(1).

"Contrary to" review is not de novo review of the underlying claim. The federal court does exercise independent judgment, but it exercises that judgment on whether the prerequisite for an exception to the prior adjudication bar has been met, i.e., whether the state court decision is contrary to clearly established Supreme Court precedent. See Scheidegger, supra, 98 Colum. L. Rev., at 959.

In making this determination, the precedents of the federal circuits are not irrelevant, cf. Furman v. Wood, 169 F. 3d 1230, 1232 (CA9 1999), but neither is the circuit's own precedent controlling, or even particularly weighty. Instead, as under Teague when properly applied, a survey of the entire legal landscape is called for; state as well as federal precedents must be considered. Caspari v. Bohlen, 510 U. S. 383, 395 (1994).⁽⁵⁾ The mere existence of conflicting precedent does not per se disqualify a rule from "clearly established" status, because some opinions fail to give careful thought to their holdings. However, a substantial split with reasoned opinions on both sides would establish that neither side is "clearly established." See Vieux v. Pepe, No. 98-1864, n. 4, and accompanying text (CA1 July 19, 1999) (applying similar reasoning, albeit under "unreasonable application" rather than "clearly established"). Conversely, unanimity among numerous reasoned opinions that a Supreme Court precedent does apply to a particular situation could render a perfunctory holding applying a different rule "contrary to . . . clearly established federal law."

Some complaints have been made that virtually no habeas petitions will be granted under such a standard. See, e.g., ACLU Brief 9 ("virtually null set"). On the eve of the twenty-first century, cases that warrant relitigation of claims already addressed should be rare. Federal habeas is for "fundamental defects," Brecht v. Abrahamson, supra, 507 U. S., at 640 (Stevens, J., concurring), and our state courts and state corrective mechanisms have evolved to the point that few fundamental defects go uncorrected. Federal habeas is a safety net for situations that should never occur, but do on occasion. Both state and federal courts have been known to fail to apply clearly established Supreme Court precedent. See, e.g., Pennsylvania v. Bruder, 488 U. S. 9, 11 (1988) (per curiam).

Petitioner asks, "What review is required, though, when at the time the state court acted, there was no rule for resolving the claim to which its 'decision' could be 'contrary?' " Brief for Petitioner 36-37. He follows this question with language from Lambrix v. Singletary, 520 U. S. 518, 529, n. 3 (1997) describing a state of the law in which Lambrix held that the rule was new and the claim was Teague-barred. Id., at 538. The answer is not "unreasonable application," as petitioner suggests; the answer is "none." In such cases the rule in question is not "clearly established." This trunk is common to both the "contrary to" and "unreasonable application" branches. If there is no clearly established rule to which the state decision could be "contrary," there is also no rule of which it could be an "unreasonable application."

The final question is how to judge "unreasonable application." As noted supra, at 15, amicus CJLF doubts that the various verbal formulations for the "unreasonable application" branch will really matter in practice. For colorful expression, it would be hard to beat the definition of "clearly erroneous" in Parts and Elec. Motors, Inc. v. Sterling Elec., Inc., 866 F. 2d 228, 233 (CA7 1988): "To be clearly erroneous, a decision must strike us as more than just maybe or probably wrong; it must . . . strike us as wrong with the force of a five-week old, unrefrigerated dead fish." However, Congress used the word "unreasonable" and the best source for the meaning of that word would seem to be the reasonableness concept in existing habeas case law. That is the standard of Butler, supra, 494 U. S., at 415, "susceptible to debate among reasonable minds." See also Saffle v. Parks, 494 U. S. 484, 490 (1990); O'Dell v. Netherland, 521 U. S. 151, 164 (1997) ("reasonable jurist," post-AEDPA but applying same standard).

In summary, then, the habeas court should proceed in three steps, as the Eleventh Circuit said in Neelley. See supra, at 19.

1) Conduct a Teague-like survey of Supreme Court opinions first, then the rest of the "legal landscape," to determine if the rule petitioner asserts and its applicability to the present case are "clearly established," which is equivalent to "dictated by precedent."

a) No: the paragraph (1) exception to the general rule of preclusion of §2254(d) is inapplicable. Unless paragraph (2) applies, the claim is barred.

b) Yes: proceed to step 2.

2) Did the state court apply the clearly established rule of law?

a) No: the prior adjudication does not bar the claim. Proceed to the "merits" under pre-AEDPA law, i.e., 28 U. S. C. §2254(a) as previously construed.

b) Yes: proceed to step 3.

3) Was the state court's application of that rule to the facts of the case reasonable, i.e., within the limits "susceptible to debate among reasonable minds"?

a) Yes: the claim is barred on the ground it has previously been adjudicated, unless §2254(d)(2) applies.

b) No: the claim is not barred on this ground. Proceed to the merits, if there is anything left to decide.

This straightforward approach, faithfully followed, will substantially advance Congress's goal of speeding up habeas review, by ending the second-guessing of state courts on close questions and reserving collateral relief for cases of fundamental error. See supra, at 21-22; see Brown v. Allen, 344 U. S. 443, 537 (1953) (Jackson, J., concurring in the result) (needles and haystacks). In Appendix A, we reprint the Tenth Circuit's unpublished disposition of an appeal to illustrate the proper application in a typical habeas case, where petitioner is simply judge-shopping his marginal claim past a second set of judges. The federal court reviews the state decision, sees that it applied the correct standard for the claim, and checks that the decision on the facts is a reasonable one. Case dismissed.

	Beginning of this file	Part IV and V	Case Menu

4. The opinion actually says "we" here, but this Court's plenary review authority on certiorari to state courts remains unimpaired. This statute is about the relative roles of the state and lower federal courts.

5. There are Teague cases in the circuits granting habeas relief based solely on the circuit's own rule as an "old" rule, rather than surveying the entire legal landscape. Caspari makes clear these cases are wrongly decided, see Scheidegger, supra, 98 Colum L. Rev., at 948, and n. 433, and Congress has simply confirmed that.    
 
August 1999