Go to:	Appendix	Case Menu	Briefs Menu

INTEREST OF AMICUS CURIAE

The Criminal Justice Legal Foundation (CJLF)⁽¹⁾ was formed to promote the interests of victims of crime and the law-abiding public in the criminal justice system. Allowing appeals on procedural points by petitioners with no substantial claims on the merits would further delay justice and consume resources needed for more worthy cases, contrary to the interests CJLF was formed to protect.

SUMMARY OF ARGUMENT

Hohn v. United States has already held that amended §2253 applies to cases filed before AEDPA but appealed afterward, by applying the statute to Hohn's case despite his Lindh v. Murphy argument. The only basis for this conclusion in a federal prisoner case was a rejection of sweeping language in Lindh purporting to apply its holding beyond §2254(d) to all of AEDPA's chapter 153 amendments.

Lindh was wrongly decided as an original matter. In addition, it is inconsistent with the later case of Martin v. Hadix, which rejected a very similar argument. To clear up the confusion, "negative implication" of pure prospectivity of one provision from an express statement of retroactivity for another should be categorically rejected.

The amended statute allows appeals by those petitioners who have substantial constitutional claims on the merits. An appeal, once allowed, permits review of the procedural prerequisites of the arguably meritorious claims.

ARGUMENT

I. Hohn implicitly held that amended §2253 applies;
the only question is why.

The first question is whether the amendment to 28 U. S. C. §2253 in the Antiterrorism and Effective Death Penalty Act ("AEDPA"), reprinted in Appendix A, applies to a case filed in district court before the effective date of the act but appealed after. The principal, perhaps only, argument that it does not is the statement in Lindh v. Murphy, 521 U. S. 320, 327 (1997) that the amendments to chapter 153 of title 28 apply only to a case filed after the effective date of AEDPA.

Precisely that argument has already been made to this Court and been rejected. Before AEDPA, Arnold Hohn, a federal prisoner, filed a motion under 28 U. S. C. §2255 to vacate his conviction. Hohn v. United States, 524 U. S. 236, 239 (1998). He filed his notice of appeal three months after the effective date, and the court of appeals treated his notice as an application for a certificate of appealability. Ibid. Before AEDPA, §2253 did not apply to federal prisoners.

Hohn argued that, under Lindh, the amendment did not apply to his case. Brief for Petitioner in Hohn v. United States, No. 96-8986, pp. 41-42. The Court's opinion in Hohn does not mention this argument, but proceeds to decide that this Court has statutory certiorari jurisdiction to review denial of the certificate by the court of appeals. Hohn, 524 U. S., at 253.

It is well established that "[q]uestions which merely lurk in the record, neither brought to the attention of the court nor ruled upon, are not to be considered as having been so decided as to constitute precedents." Webster v. Fall, 266 U. S. 507, 511 (1925) (emphasis added). When the point has been briefed, however, the rule is different. Clemons v. Mississippi, 494 U. S. 738, 747-748, n. 3 (1990), held that when a point is argued either by a party or by the dissenting opinion and not addressed by the Court, the refusal to address the argument is an implicit rejection of it. See also James B. Beam Distilling Co. v. Georgia, 501 U. S. 529, 539, n. 2 (1991) (opinion of Souter, J.).

The Court may decline to reach some arguments for reasons other than the merits. See, e.g., Yee v. City of Escondido, 503 U. S. 519, 535 (1992) (not fairly included in question presented in certiorari petition); Califano v. Yamaski, 442 U. S. 682, 693 (1979) (constitutional argument need not be addressed when party asserting it prevails on other grounds). No such reason appears for ignoring Hohn's Lindh argument. The conclusion that §2253 applied to Hohn's case was an essential prerequisite to the question of whether a certificate application is a "[c]ase[] in the court[] of appeals" within the meaning of 28 U. S. C. §1254. Cf. Caspari v. Bohlen, 510 U. S. 383, 389-390 (1994) (necessary predicate question is fairly included). If §2253 did not apply, then Hohn's original notice of appeal, see 524 U. S., at 240, would have been valid. If so, then he indisputably would have had a case in the court of appeals. The difficult question of whether to overrule House v. Mayo, 324 U. S. 42, 44 (1945) (per curiam) would not have been presented, and it would have been improper to decide it. We should not assume that this Court would reach out to overrule a precedent in a case where it was totally unnecessary. Under the rule of Clemons, supra, the Hohn Court's refusal to address petitioner's Lindh argument was a rejection of it.

There are two possible bases for Hohn's rejection of this argument. One possibility is the expansion of the concept of the certificate application as a "case." Under this approach, the application is a new case, separate from the original case and filed after the effective date. Hence, the new statute would apply.

As applied to a state prisoner such as Slack, this would not be too much of a stretch beyond Hohn. Under either AEDPA or pre-AEDPA law, Slack had to file an application, creating a "case." For Hohn, though, this theory would depend on circular logic. Under pre-AEDPA law Hohn had an appeal as of right. His notice of appeal was treated as an application for a certificate only on the assumption that §2253 applied. See 524 U. S., at 240. Without that assumption, Hohn's holding that an application for a certificate is a "case" would have to be stretched to hold that a notice of appeal is also a "case." A notice of appeal, however, does not typically present issues for briefing and decision, cf. id., at 245, and that would be stretching Hohn too far.

There is an alternative basis for rejection of the Lindh argument. Notwithstanding some expansive language, the actual question presented in Lindh was limited to the temporal scope of a single, new subdivision: 28 U. S. C. §2254(d). See Lindh, supra, 521 U. S., at 324. Hohn's application of the §2253 amendments may be read as an unwillingness to expand Lindh beyond the specific holding of that case on the narrow question presented.

II. Lindh should not be extended beyond §2254(d).

A. Landgraf and "Express."

In Landgraf v. USI Film Products, 511 U. S. 244 (1994), this Court clearly stated that in the absence of an express statement by Congress, the court must ask if application of the new statute would be retroactive. Id., at 280. This statement must, of course, be considered in the context of the point being discussed. Notwithstanding the contrary statement in Lindh, supra, 521 U. S., at 325, the context indicated that the "express command" requirement worked both ways, i.e., that the "judicial default rules" applied whenever Congress had not made an express statement to the contrary.

The statement is the conclusion of the section which resolves the tension between two canons of statutory construction. Landgraf, 511 U. S., at 263. "The first is the rule that 'a court is to apply the law in effect at the time it renders its decision . . . .' " Id., at 264 (quoting Bradley v. School Bd. of Richmond, 416 U. S. 696, 711 (1974)). "The second is the axiom that '[r]etroactivity is not favored in the law . . . .' " Ibid. (quoting Bowen v. Georgetown Univ. Hospital, 488 U. S. 204, 208 (1988)). The resolution was that the Bradley presumption would yield to the Bowen presumption when they were in conflict. Id., at 277. With that exception, however, nothing in Landgraf indicates any backing off from the Bradley rule. The reason the court must inquire into retroactivity, it would seem, is because in the absence of retroactive effect the Bradley rule controls.

A two-way express-statement rule would "giv[e] legislators a predictable background rule against which to legislate." See id., at 273. When these two rules cover the field, Congress can reasonably predict what rule will apply when it makes no express statement. Such a rule would most often implement the intent of Congress. When Congress acts to change a rule of law, it does so because it believes the new rule balances the competing interests better than the old rule. In the case of "retroactive" application, though, that balance may be altered by a party's justified reliance on the old rule. Hence, the Bowen presumption prevents application in cases where the balance is most likely to differ from the one weighed by Congress. Where application would not be retroactive, the balance of interests is not much different from purely prospective application, and application of the new rule conforms most closely to Congress's assessment of the appropriate balance. The judicial default rules will often yield a better and fairer result than any simple rule Congress could write into the statute. Each provision of an act would be evaluated separately, "in light of ordinary judicial principles concerning the application of new rules to pending cases and preenactment conduct." Id., at 280.

B. Lindh v. Murphy.

At issue in Lindh was the question of whether the new subsection (d) of 28 U. S. C. §2254, added by AEDPA, applied in a noncapital case. The state naturally relied on the seemingly clear "express command" language of Landgraf. The Lindh majority answered this argument with the surprising assertion that the argument "ignores context." 521 U. S., at 325. As discussed supra, at 4-5, the argument is fully consistent with the context. Lindh then announces, in effect, that the seemingly clear two-step analysis of Landgraf is really a three-step analysis. Absent an express statement from Congress, the court must examine the statute for implied statements that the statute operates to some degree more prospectively than it would under the default rule. Only if no such implication were found would the court proceed to Landgraf's second step, now the third step, of asking whether application would be truly retroactive. See id., at 326.

The remainder of the Lindh majority's arguments consists of two steps: (1) Congress specified the temporal reach of the new Chapter 154 but not of the amendments to existing chapter 153 and therefore must have intended that they be different; (2) the intended difference must be pure prospectivity, i.e., inapplicability to all cases in which the petition was filed before April 24, 1996, regardless of whether application would be retroactive within the meaning of Landgraf. Both steps are non sequiturs.

Taking the second step first, assuming arguendo that Congress did have a different intent regarding chapter 153 than it had regarding chapter 154, it does not by any means follow that pure prospectivity of the entire chapter was that intent. The "inference that chapter 153 was not meant to apply to pending cases" is by no means "straightforward," see Lindh, 521 U. S., at 332, if that inference is meant to cover every application of every change made by §§101-106 of the AEDPA.

These amendments make many different kinds of changes. Application of some of those changes to pending cases present none of the fairness concerns that underlie the disfavored status of retroactive application, see, e.g., 28 U. S. C. §2254(b)(2) (denial of unexhausted claim on the merits); Landgraf, 511 U. S., at 265-267, and there is no conceivable reason why Congress would not want to apply such uncontroversial reforms to pending cases. Conversely, it would be wasteful and inefficient to apply the certificate of appealability requirement with total retroactivity, so as to require new certificates in cases already briefed on appeal under the old procedure.

Landgraf noted that Congress may deliberately decide not to decide retroactivity, with the intent that the courts will decide on the basis of judicially created default rules. Id., at 260-261. That is, from the premise that Congress did not intend total retroactivity of every change made to chapter 153, it does not follow that Congress intended that all of those changes be inapplicable to pending cases. The thesis "that all those other provisions must be treated uniformly for purposes of their application to pending cases . . . is by no means an inevitable one." Ibid. The contrary was "highly probable" in Landgraf, id., at 261, and it is highly probable here. That contrary thesis is that Congress intended that the courts would examine each provision and apply it to pending cases or not, based on whether its application would be unfairly "retroactive" under Landgraf. That different intent would fully "explain[] the different treatment," cf. Lindh, 521 U. S., at 329, be fully consistent with Landgraf, and be fully consistent with the value judgment of Congress that the amended statutes balance the competing interests better than the earlier ones.

More fundamental than the question of what intent §107(c) expressed about §§101-106 is the question of whether it expressed any intent at all. The lynchpin of this holding is the statement that nothing but a different intent explains the absence of a parallel provision in the chapter 153 amendments. Lindh, 521 U. S., at 329. Not so.

Chapter 154 is a fundamentally different kind of enactment from the amendments to chapter 153. The chapter 153 amendments are significant, but none depends for its operation on any of the others. Given Landgraf's apparent assurance that the various provisions of an act would be considered individually, the need for an express statement as to all the chapter 153 amendments is diminished. Chapter 154, in contrast, can only be applied to a case as a coherent whole. It adopts a quid pro quo arrangement in which the state receives certain benefits in return for extending assistance to the prisoner above the constitutional minimum. See L. Powell, et al., Judicial Conference of the United States Ad Hoc Committee on Federal Habeas Corpus in Capital Cases, Committee Report and Proposal 6 (1989), reprinted in 135 Cong. Rec. 24,694, 24,695, col. 2 (1989) ("Powell Committee Report"). An express statement on its application was therefore much more necessary than it was for chapter 153.

Even with regard to the individual provisions, Lindh overstates the similarity of the two chapters. Both chapters supposedly contain changes that "govern[] standards affecting entitlement to relief." Lindh, 521 U. S., at 329. But the only such provision cited for chapter 154 is its "incorporation" of the new §2254(d) from chapter 153. See id., at 327. This is the well-known fallacy of assuming the conclusion. If the "rival" interpretation of §2264(b) is correct, see id., at 339-340 (Rehnquist, C.J., dissenting), then chapter 154 does not make the new §2254(d) applicable to any cases to which it would not otherwise apply, and therefore it does not make any change to the standards affecting entitlement to relief. The supposed similarity of the two chapters is an extremely thin reed on which to rest a negative implication.

The most clearly fallacious of the Lindh majority's arguments is the last one. The argument goes that the sole purpose of mentioning subdivisions (a), (d), and (e) of §2254 in §2264(b) is to make applicable to pending cases changes in those subdivisions that would otherwise be prospective only. The simple, obvious, and conclusive refutation to that interpretation is that §2254(a) was not amended, and hence making changes to it retroactive could not possibly have been the Congressional purpose. Far from being a "loose end," cf. Lindh, 521 U. S., at 336, this is a knockout punch.

The "rival" analysis of the dissent is far more coherent than the one adopted by the Lindh majority. See id., at 336; id., at 339-340 (dissent). It accounts for and gives meaning to every word of the statute, rather than leaving a "loose end." Space does not permit a full discussion here, so we will simply refer to the dissent and to our brief in Lindh. See Brief for Criminal Justice Legal Foundation as Amicus Curiae in Lindh v. Murphy, No. 96-6298, pp. 19-21. We will add only one additional comment. The majority expresses doubt that Congress intended to relax the exhaustion rule and puzzlement as to why it would. 521 U. S., at 333-334, n. 7. The intent and reason for it are stated with crystal clarity in the Powell Committee Report, supra, at 22, 135 Cong. Rec., at 24,698, col. 2, and quoted in the dissent. 521 U. S., at 341, n. 2. Similarly, the reason for the "subject to" clause of §2264(b) is easily seen by comparing the final language with the draft of the Powell Committee Report, at 22, 135 Cong. Rec., at 24,698, col. 2, which could have been interpreted to eliminate all other prerequisites to reaching the merits. The majority opinion's lack of awareness of this well-known, definitive legislative history casts further doubt on its already shaky analysis.

In a world of silk purses and pigs' ears, Lindh v. Murphy is not a silk purse of the art of statutory interpretation. Cf. 521 U. S., at 336. The narrow holding of Lindh, that §2254(d) does not apply to cases filed before its enactment, is not at issue in the present case. A broader interpretation, that the chapter 153 amendments in their entirety do not apply, should be rejected in favor of applying the default rule to each remaining provision. There is no need in this case to overrule Lindh, but amicus submits that the Court should decline to go beyond that decision, by even a fraction of an inch. Cf. Silverman v. United States, 365 U. S. 505, 512 (1961).

C. Martin v. Hadix, Stare Decisis, and Hairline Distinctions.

Last term in Martin v. Hadix, 527 U. S. __, 144 L. Ed. 2d 347, 119 S. Ct. 1998 (1999), this Court rejected a "negative implication" argument very similar to the one accepted in Lindh. Martin involved the cap on attorneys' fees enacted in the Prison Litigation Reform Act, Pub. L. No. 104-134, §803(d)(3), 110 Stat. 1321-66; 42 U. S. C. § 1997e(d)(3). Following the successful path of the Lindh petitioner, Hadix and the other prisoners argued that Congress's omission of a provision for application to pending cases in §803, while including one in §802, implied that §803 was completely inapplicable to any case pending on the date of enactment, even when application would not be "retroactive" within the meaning of Landgraf. See Martin, 144 L. Ed. 2d, at 358-359, 119 S. Ct., at 2004-2005.

The Martin Court noted that in Lindh, "[t]his argument carried special weight because both chapters addressed similar issues," i.e., they both established "new standards for review . . . ." Id., at 359, 119 S. Ct., at 2005. The "same negative inference [did] not arise" regarding §§802 and 803, however, because they "address wholly distinct subject matters." Ibid. This "similar" versus "distinct" subject matter distinction is nowhere near as clear from the face of the statutes as one might think from reading the Martin opinion. As noted earlier, the Lindh majority's conclusion that chapter 154 changed the substantive standards for review rested entirely on the fallacy of assuming the conclusion. See supra, at 7.

Nor are the subjects of §§802 (18 U. S. C. §3626) and 803 (42 U. S. C. §1997e) really wholly distinct. The subject of §802 is "Appropriate remedies with respect to prison conditions." 18 U. S. C. §3626, caption. In §803, we also find a limitation on a remedy for prisoner: no damages for mental or emotional injury without physical injury. 42 U. S. C. §1997e(e). Section 802 addresses procedure in prison conditions litigation. §3626(e). So does §803. §1997e(f). The specific provision at issue in Martin capped attorneys' fees with reference to the rate payable to appointed counsel for criminal defendants. §1997e(d)(3). Section 802 contains a strikingly similar cap on fees payable to special masters. §3626(f)(4). The line between "similar" and "distinct" subject matters is less than clear. Martin may well be correct that Lindh is distinguishable and that the statute at issue falls on the other side of whatever line Lindh draws, but that conclusion, however correct, offers minimal guidance for deciding the next case.

The tension between Martin and Lindh does not end with this hairline distinction, though. In reaching its holding regarding fees earned after the effective date, Martin implicitly rejects a central feature of the Lindh decision. The inferences of pure prospectivity that the prisoners sought to draw from the statute and its history were, indeed, "weak." See Martin, 144 L. Ed. 2d, at 360, 119 S. Ct., at 2006. They were not, however, zero, and Martin cites no inferences whatever on the other side of the scale. If a mere preponderance were sufficient, anything at all would beat nothing at all. Yet Martin rejects the weak inferences and applies the new law in those circumstances where its application would not be retroactive. Id., at 361-362, 119 S. Ct., at 2007.

The presumption in favor of applying current law whenever it would not be "retroactive" does carry some weight in interpretation of the statute, after all. That is, by enacting the statute, Congress has declared that the new law is better than the old one and presumably wants that new, better law to apply whenever its application would not cause a "manifest injustice." See id., at 362, 119 S. Ct., at 2007. At its bottom line, Martin holds that the absence of an express command plus the absence of a retroactive effect suffice to make the new law applicable. Ibid. This holding contradicts the holding of Lindh that a mere preponderance of one interpretation over another was sufficient to delay the effect of Congress's reforms well beyond the point of nonretroactivity. Cf. Lindh, 521 U. S., at 336.

Before a rule of law can be respected and obeyed, it must be discernable. See Douglas v. Jeannette, 319 U. S. 157, 182 (1943) (Jackson, J., concurring in the judgment). A lower court trying to do its duty to obey the precedents in both Lindh and Martin will find the task nearly impossible. The goals of stability and predictability that underlie stare decisis, see Hilton v. South Carolina Public Railways Comm'n, 502 U. S. 197, 202 (1991), would be better served by frankly rejecting Lindh's negative inference thesis and embracing a predictable background rule. See United States v. Dixon, 509 U. S. 688, 712 (1993) (overruling case was preferable to pretending it survived when it did not).

The clearest, simplest, and, amicus submits, best rule is simply to apply the "express command" requirement both ways. In the absence of an express command, the anti-retroactivity presumption blocks any application of the new statute that would have retroactive effect, and in all other circumstances the presumption in favor of applying current law controls.

A lesser step, sufficient to resolve the present case, would be a categorical rejection of the negative implication argument. Congress should be able to state the temporal scope of one provision without fear of the unintended consequence of limiting the temporal scope of another provision. Statutes are not enacted by a single individual with a single intent. They are products of negotiation and compromise. Negotiation may focus on one provision at a time, with little regard for other sections as language is hammered out. That difficult process is made more difficult if language must be scrutinized not only for what it says and does but also for any conceivable effect it may have on other sections.

The amendments to §2253 are purely procedural, do not affect primary conduct, and do not impact any legitimate reliance interests. They should apply, as Hohn applied them, to all cases appealed after April 24, 1996.

III. A certificate of appealability requires a substantial claim on the merits, not procedure.

Section 102 of the AEDPA amended 28 U. S. C. §2253 as set forth in Appendix A. In the first two paragraphs and the header language of the third, Congress made stylistic changes, extended the requirement to cover federal prisoners, and changed the name from "certificate of probable cause" to "certificate of appealability" ("COA"). New subdivision (c)(2) largely codifies the standard of Barefoot v. Estelle, 463 U. S. 880, 892, n. 3 (1983). It also limits the substantive grounds to denial of constitutional rights, thus focusing the inquiry on the actual claims for relief, not the procedural prerequisites. New subdivision (c)(3) requires the certificate to specify the issue which satisfies the paragraph (2) requirement. "The whole point of the new procedure is to limit appeals to issues on which a substantial showing of the denial of a federal constitutional right has been made." Burns v. Gammon, 173 F. 3d 1089, 1090, n. 2 (CA8 1999).

As applied to an appeal from the district court's denial of a constitutional claim on the merits, the statute is straightforward. Petitioner must meet the Barefoot standard for each claim. Only issues meeting this standard are cognizable on appeal. See ibid. Applying this statute to issues dismissed on procedural grounds seems to have raised groundless fears that review of issues would be blocked if the statute were applied as written. Amicus CJLF suggests that a simple, straightforward application that respects both the statutory language and the intent of the act is available.

The most consistent theme throughout the AEDPA reforms is to cut down on the number of different proceedings that must be held before a case is truly final. We see this in the severe restriction on successive petitions. 28 U. S. C. §2244(b). We see it in the limitation on evidentiary hearings. 28 U. S. C. §2254(e)(2). Most pertinent to the present question, we see it in the authorization to dismiss meritless, unexhausted claims in 28 U. S. C. §2254(b)(2). Congress has expressed its judgment that cases should not be bounced around between courts on preliminary procedural questions when the underlying claim is obviously without merit.

On its face, §2253(c)(2) requires the judge considering a certificate application to look at the merits of the underlying claims and not the procedural prerequisites to granting relief. If there is a substantial claim on the merits, the appeal should go forward. If there is not, the case should end with the district court's denial.

The question of whether an appeal can proceed is distinct from the questions of what issues the appellate court can or must consider and what relief is available if it finds the district court erred. If the underlying claim has substance, the petitioner/appellant can appeal any adverse rulings on the procedural prerequisites. The requirement of a showing of a substantial claim on the merits does not depend on whether the district court ruled on the merits. The statute unambiguously requires the showing for all appeals. Conversely, the statute does not require any showing at all on the procedural prerequisites. Petitioner gets the certificate even if his arguably meritorious claim is unarguably defaulted. He may, however, simply get a summary affirmance with his certificate, so long as he has an opportunity to brief his case. See Barefoot, supra, 463 U. S., at 893-894 (expedited procedure may be appropriate, despite issuance of certificate).

The procedures devised by the courts of appeals are less consistent with the language and purpose of the statute and ultimately less efficient. In Murphy v. Johnson, 110 F. 3d 10, 10 (CA5 1997), the district court had dismissed the petition for nonexhaustion. Relying on pre-AEDPA precedent, Murphy held that the applicant for a COA must make "a credible showing of exhaustion," before the court would consider whether his underlying claim met the Barefoot standard. Id., at 11. The new statute quite simply does not say that. Whether a constitutional right has been denied and whether state remedies have been exhausted are entirely separate questions.

From the standpoint of avoiding unjust results, it makes sense to apply the certificate filter to the merits and not the procedural ground. If a meritorious claim has been denied on procedural grounds, that procedural ruling deserves careful scrutiny, as a miscarriage of justice is possible. On the other hand, if the underlying claim is clearly meritless, it really does not matter if the district court erred on the procedure; no harm has been caused.

In Whitehead v. Johnson, 157 F. 3d 384 (CA5 1998), the Fifth Circuit got the statute completely backwards. The district court had dismissed Whitehead's petition for nonexhaustion. Id., at 386. The court of appeals applied the first part of its Murphy test holding that Whitehead not only had a "credible" case on exhaustion, but that he prevailed on the point. Id., at 387. The court granted the COA, vacated the judgment, and remanded for a decision on the merits without applying the statutory standard to the merits. Id., at 388. The reason given for this action was the requirement that the district court address each issue first. Ibid.

There are two problems with this justification: one legal and one practical. The legal problem is that the requirement that the district judge consider every issue before the court of appeals is not a Congressional command. FRAP Rule 22(b) does contemplate that the district judge should consider the application as a whole first, but, if the district judge should err, the remedy expressly provided is a new application to the circuit judges. On the other hand, §2253(c) is a Congressional command. That statute forbids the court of appeals from disturbing the district court's decision until a certificate has issued, and it further forbids the issuance of the certificate until a judge has determined that the petitioner has a substantial claim on the merits.

The Whitehead case illustrates the practical wisdom of Congress's choice and the wastefulness of the Fifth Circuit approach. Whitehead's claim involved the calculation of "time credits" for a person with multiple sentences. See 157 F. 3d, at 385-386. Whatever merit this claim might have under state law, the allegation that the calculation violates the Constitution appears quite far-fetched. If the court of appeals had simply applied the Barefoot standard to the merits as its first step, it likely would have ended the litigation right there. Instead it decided the exhaustion issue and sent the case back to district court, requiring a new proceeding and decision there, another application for a certificate, and, if it is denied, another application to the court of appeals. This is the kind of multiplication of proceedings that Congress sought to eliminate.

Another line of cases effects a partial judicial repeal of the statute by conjuring up the "preclude all review" bogeyman. Nichols v. Bowersox, 172 F. 3d 1068, 1070 (CA8 1999) (en banc) involved the purely procedural question of timely filing and the "mailbox rule." The court construed §2253(c) as being completely inapplicable to such procedural appeals. "Otherwise, a final order entered by a district court based upon a question antecedent to the merits, if adverse to the petitioner, could never be reviewed on appeal." Id., at 1070, n. 2.

The problem, again, is that this holding flies in the face of an unambiguous Congressional command. Section 2253(c) does not say "the final order, if decided on the merits." It just says "the final order." As Judge Arnold noted in dissent, there is a simple solution to the perceived problem that respects this clear language: "A requirement that the prisoner make some kind of abbreviated showing on the merits (perhaps in an appropriate case accompanied by an offer of proof) before he or she can take an appeal, even on a matter unconnected with the merits, is a perfectly rational (if rather cumbersome) one." Id., at 1078 (dissenting opinion). Preliminary procedural questions are thus not precluded from review, but review is limited to petitioners with some realistic chance of prevailing on the merits.

A COA should issue when, and only when, the habeas petitioner has made a substantial showing of merit on one or more of his substantive claims. As no such showing has been made in the present case, no certificate should issue.

CONCLUSION

The decision of the Court of Appeals for the Ninth Circuit denying a certificate of appealability should be affirmed.

Respectfully submitted,

Kent S. Scheidegger

	Beginning of this file	Appendix	Briefs Menu

1. No counsel for a party authored this brief in whole or in part, and no one other than CJLF made any contribution to its preparation or submission. Both parties have consented in writing to its filing.