 |
|||||
|
|
|
|
|
||
|
|
|||||
|
|
|
|
|
||
In 28 U. S. C. §2244(b)(2)(A), Congress made an exception to the abuse-of-the-writ rule for new rules "made retroactive to cases on collateral review by the Supreme Court . . . ." Petitioner asks this Court, in essence, to rewrite this clause to say "made retroactive . . . by the district or circuit court in accordance with principles established by the Supreme Court." See Brief for Petitioner 9.
The plain meaning of "a new rule . . . made retroactive . . . by the Supreme Court" is to require action by this Court on the specific rule in question and specifically addressing retroactivity, rather than some other issue. Implicitly recognizing this, petitioner asks the Court to deviate from the plain meaning by quoting a disparaging remark about the draftsmanship of the statute. See Brief for Petitioner 14, n. 5 (quoting Lindh v. Murphy, 521 U. S. 320, 336 (1997)). This comment in Lindh dealt with an issue where the meaning was not plain. The difficulties which attended the drafting of this statute require, even more than normal, that the meaning be respected where it is plain.
Artistry in drafting, as in any endeavor, requires a single, coherent vision. The Mona Lisa was not painted by a committee. The statute at issue was "hammered out over the course of a great many negotiations." 141 Cong. Rec. 15,020, col. 1 (1995) (statement of Sen. Specter). Hammering leaves dents. Respect for the democratic process requires that the product of legislative compromise be left where the legislature put it and not moved somewhere else by tortured construction. In West v. Vaughn, 204 F. 3d 53, 59 (2000), the Third Circuit sought to avoid the plain meaning of "made" by claiming that Congress could have used other words to make that meaning even plainer. This approach demands more of the legislative process than is realistic. While negotiating statutory language, once a point is settled and plainly expressed, negotiators move on to other points. They do not typically reopen an agreed point to alter subtle nuances of language. The Third Circuit's approach effectively demands perfection in drafting. Respect for the compromise actually reached by the democratic process requires a more realistic interpretation.
The habeas reforms of the AEDPA were the result of a profound dissatisfaction with the way habeas cases were being handled in the district courts and courts of appeals. Despite such landmark cases as Teague v. Lane, 489 U. S. 288 (1989) and McCleskey v. Zant, 499 U. S. 467 (1991), cases continued to drag on interminably. The problem was particularly acute in capital cases, where habeas corpus was effectively nullifying capital punishment. See 141 Cong. Rec. 15,037-15,038 (1995) (statement of Sen. Nickles).
The debacle surrounding the execution of Robert Alton Harris was repeatedly cited by supporters as illustrating the need for change. See, e.g., 141 Cong. Rec. 15,013, col. 2 (1995) (statement of Sen. Specter); id., at 4107-4108 (statement of Mr. Cunningham); id., at 4111-4112 (statement of Mr. Cox). Harris's case became "final" on direct appeal in October, 1981. Harris v. California, 454 U. S. 882 (1981). State collateral proceedings and two overlapping federal petitions took another nine years. See Harris v. Pulley, 852 F. 2d 1546 (CA9 1988), opinion amended, 885 F. 2d 1354 (1989), cert. denied, 493 U. S. 1051 (1990). A third federal petition was denied, largely on Teague and McCleskey grounds, but that took another two years. See Harris v. Vasquez, 949 F. 2d 1497, 1528-1529 (CA9 1990), cert. denied, 503 U. S. 910 (1992). One would expect that the resolution of the third petition in a case with no serious question of guilt would be the end, but instead there followed a series of evasions of the McCleskey limitation, see Gomez v. United States Dist. Court for the Northern Dist. of Cal., 503 U. S. 653, 653 (1992) (per curiam), eventually culminating in an unusual, possibly unique, order from this Court prohibiting any more federal stays. Vasquez v. Harris, 503 U. S. 1000 (1992). The complete 141-step chronology is given in Lungren and Krotoski, Public Policy Lessons from the Robert Alton Harris Case, 40 UCLA L. Rev. 295, 315-326 (1992). While the Harris case was unique in the shamelessness of the evasion of McCleskey, it was not unusual in the length of delay in the district and circuit courts. Supporters of stronger limits had horror stories from their own states. See, e.g., 141 Cong. Rec. 4098-4101 (1995) (statement of Mr. Heineman); id., at 15,062, col. 2 (statement of Sen. Hatch).
An interpretation of the statute to do nothing more than codify the case law is at war with its obvious and well-known purpose. Congress understood that reforms implemented by this Court were being frustrated by foot-dragging and evasion at lower levels and took steps to change that. At several points, the mechanism Congress chose was to elevate the decision to a higher level of the federal judiciary.
Some courts of appeals had held that their own precedents could be "old rules" for Teague,(2) but Congress limited 28 U. S. C. §2254(d)(1) to "clearly established federal law, as determined by the Supreme Court." The initial decision to allow a successive petition to be considered was removed from the district court and placed in the hands of a three-judge panel of the Court of Appeals. See 28 U. S. C. §2244(b)(3). The "new rule" exceptions refer to the Supreme Court, as discussed further below. See §§2244(b)(2)(A) and (d)(1)(C), 2254(e)(2)(A)(i), 2255, ¶6, subd. (3) and ¶8, subd. (2). These level-raising mechanisms indicate a lack of confidence in the lower federal courts. Congress believed, with justification, that the Teague and McCleskey reforms were not being faithfully implemented, and moving the decisions to higher levels was one of its mechanisms for dealing with that problem. In particular, Congress limited the retroactivity exception for successive petitions to new rules made retroactive by this Court.
Comparison of the three retroactivity provisions applicable to state prisoners is instructive. (The federal counterparts in §2255 are the same.) The three create parallel exceptions to three related doctrines: abuse of the writ, procedural default, and delayed petitions. The first two are identical: "the claim relies on a new rule of constitutional law, made retroactive to cases on collateral review by the Supreme Court, that was previously unavailable . . . ." §§2244(b)(2)(A), 2254(e)(2)(A)(i) (emphasis added). The delayed petition language, i.e., the new statute of limitations, is different: "if the right has been newly recognized by the Supreme Court and made retroactively applicable to cases on collateral review . . . ." §2244(d)(1)(C).
Significantly, the latter language moves the "by the Supreme Court" requirement out of the retroactivity clause into the substantive clause. Petitioner contends that "by the Supreme Court" in the retroactivity clause of subdivision (b)(2)(A) means conformity to this Court's retroactivity jurisprudence, but the words would be superfluous if that were all they meant. Their absence from the corresponding clause of subdivision (d)(1)(C) does not authorize the courts of appeals to develop their own body of independent retroactivity law. They must still follow Teague. The difference in language means that, for the statute of limitations, state prisoners can file within one year of this Court's recognition of a new right, even if this Court has not resolved the status of the right's retroactivity on habeas corpus, as it typically would not if the new rule is made in a direct review case. It makes sense that Congress would be more generous with the statute of limitations, as that relatively harsh rule cuts off prisoners who have never had a habeas petition, while the other two limits apply only to prisoners who did have a prior state or federal petition.
In Lindh v. Murphy, 521 U. S. 320, 326-329 (1997), this Court found significance in the AEDPA's use of different language in two different chapters of Title 28. Here we have different language in two subsections of the same section. The inference is much stronger here. Congress put the "by the Supreme Court" requirement in the retroactivity clause of one subsection and not the other because it intended different meanings.
Congress knew what it was doing, and the Seventh Circuit correctly recognized the meaning. "[R]etroactive application must be declared by the Supreme Court itself." Talbott v. Indiana, 226 F. 3d 866, 868 (2000). As discussed in part III A, infra, Sullivan v. Louisiana, 508 U. S. 275 (1993) decided a question distinct from retroactivity. That decision could not and did not make Cage retroactive.
Petitioner contends that the Court's order in Adams v. Evatt, 511 U. S. 1001 (1994) makes Cage v. Louisiana, 498 U. S. 39 (1990) (per curiam), retroactive on habeas corpus. See Brief for Petitioner 28. The order was to grant certiorari, vacate the judgment, and remand, often referred to as a "GVR order." See Lawrence v. Chater, 516 U. S. 163, 166 (1996) (per curiam). This contention is completely without merit. A GVR order decides nothing. "It is precisely because of uncertainty that we GVR." Id., at 172.
The typical GVR involves an intervening development in the law which creates a "reasonable probability that the decision below rests upon a premise that the lower court would reject if given the opportunity for further consideration . . . ." Id., at 167. "Reasonable probability" is a modest threshold. It is less than preponderance. See Kyles v. Whitley, 514 U. S. 419, 434 (1995). Sometimes the intervening development is clearly controlling and the outcome on remand is nearly certain. See, e.g., Thomas v. American Home Products, Inc., 519 U. S. 913, 914 (1996) (Scalia, J., concurring) (federal court's interpretation of state law now clearly incorrect in light of subsequent state supreme court decision). Such certainty is not required, or even the norm. A GVR may follow a new agency interpretation of a statute, which is entitled to deference but not unquestioning acceptance. See Lawrence, 516 U. S., at 171. GVRs may also follow "plausible confessions of error without determining their merits." Ibid. (emphasis added).
Particularly pertinent to this present issue, the Court may GVR a case for consideration in light of a new decision by this Court without deciding whether the new decision applies to the GVR'd case. A month after Teague, this Court GVR'd a capital case in which the Court of Appeals had applied a new rule, and the state had not raised retroactivity. Zant v. Moore, 489 U. S. 836 (1989) (per curiam); id., at 836 (Brennan, J., concurring) (waiver issue should be considered on remand); id., at 837 (Blackmun, J., dissenting) (defense should be deemed waived). Surely the GVR in Zant v. Moore did not decide whether Teague applied to the penalty phase of capital cases. That issue was reserved in Teague and decided at the end of the term in Penry v. Lynaugh, 492 U. S. 302, 313-314 (1989). The issue of when nonretroactivity is waived was not decided until years later. See Schiro v. Farley, 510 U. S. 222, 228-229 (1994).
The Court has said often and emphatically that denial of certiorari is not precedent. See Teague, 489 U. S., at 296. A similar unequivocal statement is in order for GVRs. If GVRs were precedent, the Court would have to decide the merits before issuing them, cf. ibid., and that would defeat their purpose. See Lawrence, 516 U. S., at 172-173. The GVR in Adams v. Evatt indicated that the issue was doubtful enough to require another look. That is all it meant. The retroactivity of Cage has not yet been decided by the Court.
|
|
||||
|
|
|
|
||
2. Amicus believes that these cases were wrongly decided. See Scheidegger, Habeas Corpus, Relitigation, and the Legislative Power, 98 Colum. L. Rev. 888, 948 (1998).