| Syllabus | Opinion [ Ginsburg ] | Dissent [ Souter ] |
|---|---|---|
| HTML version PDF version | HTML version PDF version | HTML version PDF version |
DENEICE A. MAYLE, WARDEN, PETITIONER
v.
JACOBY LEE FELIX
ON WRIT OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT
[June 23, 2005]
Justice Souter, with whom Justice Stevens joins, dissenting.
This case requires the Court to decide how the relation back provision of Rule 15(c)(2) of the Federal Rules of Civil Procedure ought to apply in federal habeas corpus cases, when neither text nor precedent provides clear guidance. I see nothing in habeas law or practice that calls for the Court’s narrow construction of the rule, and good reasons to go the other way, including the unfortunate consequence that the Court’s view creates an unfair disparity between indigent habeas petitioners and those able to afford their own counsel. I respectfully dissent.
I
At the outset, there is need for care in understanding the narrow scope of the problem this case presents. A habeas petitioner’s opportunity to amend as a matter of course, without permission of the trial court, exists only before the responsive pleading is served, and even then only once. Rule 15(a). After one amendment, or after the government files the answer or other response, assuming one is even required, see Habeas Corpus Rule 4, the prisoner may not amend without the court’s leave or the government’s consent, Fed. Rule Civ. Proc. 15(a). While leave to amend “shall be freely given when justice so requires,” ibid., justice does, after all, have to require it, and the District Courts will presumably say no, for example, in the face of unjustifiable delay or threatened prejudice to the State. See Foman v. Davis, 371 U.S. 178, 182 (1962); see also Brief for Professor Arthur R. Miller et al. as Amici Curiae 20—21 (describing reasons courts regularly deny leave to amend and citing cases); 6 C. Wright, A. Miller, & M. Kane, Federal Practice and Procedure §§1487—1488 (2d ed. 1990) (hereinafter Wright & Miller) (discussing reasons leave to amend may be and often is denied, including delay and prejudice). The Court’s concern for “unconstrained” recourse to petition amendments, ante, at 16, is thus misplaced.
The limited opportunity to amend also supplies perspective on the claim that Felix’s reading of the relation back rule would undermine the 1-year limitation period of the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) and the statute’s concomitant concern for finality of judgments. See ante, at 15—16. In fact, AEDPA’s objectives bear little weight in the analysis, because the very point of every relation back rule is to qualify a statute of limitations, and Rule 15(c) “is based on the notion that once litigation involving particular conduct or a given transaction or occurrence has been instituted, the parties are not entitled to the protection of the statute of limitations against the later assertion by amendment of defenses or claims that arise out of the same conduct, transaction, or occurrence as set forth in the original pleading.” 6A Wright & Miller §1496, at 64. That AEDPA’s statute of limitations, like any other, may be trumped by relating back when the subject of the amendment arises out of the same conduct, transaction, or occurrence described in the original pleading, does not alone help us to figure out what conduct, transaction, or occurrence is the same.
II
Felix’s disputed right to amend with relation back effect turns entirely, as the Court says, ante, at 9, on how narrowly or how broadly the tripartite authorization for relation back ought to be construed: whether the relevant “conduct, transaction, or occurrence”1 to which a habeas petition refers includes the underlying trial (which resulted in the custody being challenged) or is limited to the set of facts underlying each trial ruling claimed to be constitutionally defective (in this case, the unconfronted videotaped testimony and the interrogation that produced the incriminating statement). If the former, a habeas petitioner will have the benefit of relation back for any amendment raising trial error, subject to the district judge’s discretion to deny leave except for the one amendment of right; if the latter, a petitioner is effectively precluded from making any amendment unless a single trial ruling amounts to distinct errors or an underlying fact is the subject of distinct rulings, notwithstanding Congress’s evident intent to provide relation back in habeas proceedings, see 28 U.S.C. § 2242; Fed. Rule Civ. Proc. 81(a)(2); Habeas Corpus Rule 11.
The text alone does not tell us the answer, for either the facts specific to the claim or the trial as a whole could be the relevant “conduct, transaction, or occurrence.” The Court assumes that the former approach is correct and then proceeds to explain, based on that assumption, the infirmity of a contrary approach. For example, the Court asserts that under Felix’s rule, “all manner of factually and temporally unrelated conduct may be raised after the statute of limitations has run … .” Ante, at 11, n. 6. But in saying this the Court presumes that the relevant transaction is what occurred outside the courtroom. Felix’s entire argument is that the proper transaction is instead what occurred in court, namely the imposition of the conviction that justifies the challenged custody. If he is right, then the Court’s assertion is incorrect, for what Felix seeks to add is a claim not about “factually and temporally unrelated conduct,” ibid., but about conduct that occurred at the same trial as the conduct addressed in the initial petition. That newly addressed conduct will hardly be “temporally unrelated” to what was previously targeted; it likely will have occurred on the same day of trial as the original conduct or within a few days. Nor will it be “factually … unrelated” to the previously raised in-court conduct, for it will almost certainly involve the same judge, the same parties and attorneys, the same courtroom, and the same jurors. Again, my point is just that much of the Court’s argument lacks force because it assumes that the proper transaction is what occurred outside the courtroom rather than inside, when that is the question we must answer.
The Court also cautions that “it would be anomalous to allow relation back under Rule 15(c)(2) based on a broader reading of the words ‘conduct, transaction, or occurrence’ in federal habeas proceedings than in ordinary civil litigation.” Ante, at 16. The cases the Court cites to establish the scope of civil relation back, however, see ante, at 10—13, simply stand for the proposition that an amendment relates back only if it deals with the same conduct, transaction, or occurrence. Felix does not purport to claim anything more.2
At first glance, an argument for the
narrow reading urged by petitioner Mayle inheres in the
distinctive pleading requirement for habeas petitions. Unlike
the generous notice-pleading standard for the benefit of
ordinary civil plaintiffs under Federal Rule of Civil Procedure
8(a), see Conley v. Gibson, 355 U.S. 41, 47
(1957), Habeas Corpus Rule 2(c) requires habeas petitioners to
“specify all the grounds for relief available,” and
to “state the facts supporting each ground.” The
Court implies that because pleading must be factually specific,
the “conduct, transaction, or occurrence” of Federal
Rule of Civil Procedure 15(c) must be specifically factual to a
parallel degree; as the Court puts it, a habeas petitioner will
plead claims “discretely,” ante, at 14, such that
each ground for relief “would delineate an
‘occurrence,’
Nor is there any policy underlying the particular habeas pleading rule that requires a more grudging relation back standard. As the Court concedes, ante, at 9, the purpose of the heightened pleading standard in habeas cases is to help a district court weed out frivolous petitions before calling upon the State to answer. See Advisory Committee’s Note on Habeas Corpus Rule 2, 28 U.S.C., p. 469; Advisory Committee Note on Rule 4, id., at 471 (“[I]t is the duty of the court to screen out frivolous applications and eliminate the burden that would be placed on the respondent by ordering an unnecessary answer”); 1 R. Hertz & J. Liebman, Federal Habeas Corpus Practice and Procedure §11.6, p. 573, n. 3 (4th ed. 2001) (hereinafter Hertz & Liebman) (“[F]act pleading, like other habeas corpus rules and practices, enables courts … to separate substantial petitions from insubstantial ones quickly and without need of adversary proceedings”); Note, Developments in the Law–Federal Habeas Corpus, 83 Harv. L. Rev. 1038, 1175 (1970) (“The justification for stringent pleading requirements in habeas corpus is thought to lie in the need to protect the courts from the burden of entertaining frivolous applications”). Identifying meritless claims has nothing to do with the effect of amendment to initial petitions for relief, except in the remote sense that an amendment will require a district judge to examine one more item. But there is no claim here that Federal Rule of Civil Procedure 15(c) has to be narrow to protect judges; the government is objecting because it wants fewer claims to defend, and that objection is unrelated to the habeas fact-pleading standard.3
While considerations based on habeas pleading fail to pan out with support for Mayle’s restricted reading of Rule 15(c), several reasons convince me that Felix’s reading is right. Most obvious is the fact that both of his claims can easily fit within the same “transaction or occurrence,” understood as a trial ending in conviction resulting in a single ultimate injury of unlawful custody. (“Conduct” sounds closer to underlying facts, perhaps, but Rule 15(c) turns on either conduct, transaction, or occurrence.) The Court acknowledges that Felix’s claims regarding his own interrogation and the videotaped testimony of witness Kenneth Williams are potentially actionable here only because the resulting incriminating statements were introduced at trial, ante, at 14, but argues that they nevertheless arise out of separate transactions or occurrences because they rest on distinct “essential predicate[s],” ibid., meaning pretrial acts. It is certainly true that the claims depend on those distinct pretrial acts, but the claims depend equally on the specified trial errors, without which there would be no habeas claim: without the introduction of each set of statements at trial, Felix would have no argument for habeas relief, regardless of what happened outside of court.4 The Court’s own opinion demonstrates this, as its descriptions of Felix’s two claims refer not only to what happened outside court but also to what happened at trial, and they specifically ground the alleged constitutional violations on the latter. See ante, at 1 (“He initially alleged … that the admission into evidence of videotaped testimony of a witness for the prosecution violated his rights under the Sixth Amendment’s Confrontation Clause”); ante, at 2 (“He asserted that, in the course of pretrial interrogation, the police used coercive tactics to obtain damaging statements from him, and that admission of those statements at trial violated his Fifth Amendment right against self-incrimination”).5 Moreover, habeas review will of course look at more of the underlying trial record than just the ruling admitting the disputed evidence, for Felix’s claims like a great many others will call for examining the trial record as a whole for signs of requisite prejudice or reversible error.6 Here, for example, if a court were to conclude that introducing Felix’s statements did violate the Fifth Amendment, relief would still turn on whether the error was harmless. This would call for a careful look at the other evidence admitted at trial, including the statements said to have come in contrary to the Confrontation Clause. In sum, Felix’s claims are not outside the text of Rule 15(c)(2).
Then there are a number of indications that Congress would not want the rule read narrowly, the first centering on the word “transaction.” That term not only goes to the breadth of relation back, but also to the scope of claim preclusion. E.g., Kremer v. Chemical Constr. Corp., 456 U.S. 461, 482, n. 22 (1982) (“Res judicata has recently been taken to bar claims arising from the same transaction even if brought under different statutes …”); accord, 1 Restatement (Second) of Judgments §24(1) (1980) (“[T]he claim extinguished includes all rights … with respect to all or any part of the transaction, or series of connected transactions, out of which the action arose”). For purposes of claim preclusion in habeas cases, the scope of “transaction” is crucial in applying AEDPA’s limitation on second or successive petitions: with very narrow exceptions, federal habeas limits a prisoner to only one petition challenging his conviction or sentence. See 28 U.S.C. § 2244(b)(1).7 The provisions limiting second or successive habeas petitions regard the relevant “transaction” for purposes of habeas claim preclusion as the trial that yielded the conviction or sentence under attack; once a challenge to that conviction or sentence has been rejected, other challenges are barred even if they raise different claims. By contrast, under the Court’s view of Rule 15(c) that the relevant “transaction” is the facts or conduct underlying each discrete claim, a prisoner should be allowed to file a second habeas petition so long as it is based on different underlying facts or conduct (i.e., on what the Court considers a separate “transaction”). The Court thus adopts, for purposes of relation back in habeas cases, a definition of “transaction” different from the one Congress apparently intended for purposes of claim preclusion in habeas cases. Judge Easterbrook explained this tension in Ellzey v. United States, 324 F.3d 521 (CA7 2003), and the Court offers no evidence that Congress would have decreed any such apparent anomaly within the body of habeas standards.8
There is, rather, a fair indication that Congress would have intended otherwise, in the fact that it has already placed limits on the right of some habeas petitioners to amend their petitions. In Chapter 154 of Title 28, providing special procedures for habeas cases brought by petitioners subject to capital sentences in certain States, Congress specifically prohibited amendment of the original habeas petitions after the filing of the answer, except on the grounds specified for second or successive petitions under 28 U.S.C. § 2244(b). See §2266(b)(3)(B). Congress’s intent to limit capital petitioners’ opportunity to amend (and thus to take advantage of relation back) makes sense owing to capital petitioners’ incentive for delay, but the provision it enacted also helps us make sense of Rule 15(c) in the usual habeas case where a prisoner has no incentive to string the process out. For Congress has shown not only that it knows how to limit amendment in habeas cases, but also that it specifically considered the subject of limiting amendment in such cases and chose not to limit amendment in the ordinary ones.
The final reason to view the trial as the relevant “transaction” in Rule 15(c)(2) lies in the real consequences of today’s decision, which will fall most heavily on the shoulders of indigent habeas petitioners who can afford no counsel without the assistance of the court. In practical terms, the significance of the right to amend arises from the fact that in the overwhelming majority of cases, the original petition is the work of a pro se petitioner. See Duncan v. Walker, 533 U.S. 167, 191 (2001) (Breyer, J., dissenting) (93% of habeas petitioners in study were pro se (citing U.S. Dept. of Justice, Office of Justice Programs, Bureau of Justice Statistics, Federal Habeas Corpus Review: Challenging State Court Criminal Convictions 14 (1995))); 1 Hertz & Liebman, §12.2, at 601 (“[N]early all” federal habeas petitioners commence proceedings either without legal assistance or with only the aid of a fellow inmate or a volunteer attorney). Unless required by statute, appointment of counsel is most often a matter of discretion on the part of the court. The district judge may well choose not to exercise that discretion unless and until a habeas proceeding advances to the stage of discovery or evidentiary hearing. See Habeas Corpus Rule 6(a) (requiring appointment of counsel for indigent petitioner “[i]f necessary for effective discovery”); Rule 8(c) (requiring appointment of counsel “[i]f an evidentiary hearing is warranted”). And the judge almost certainly will not appoint counsel until after the preliminary review of the petition to see whether it plainly warrants dismissal. See Rule 4. Where a petition (even in its pro se form) has survived this review by showing enough merit to justify appointing counsel, it makes no sense to say that counsel (appointed because of that apparent merit) should be precluded from exercising professional judgment when that judgment calls for adding a new ground for relief that would relate back to the filing of the original petition. For by hobbling counsel this way, the Court limits the capacity of appointed counsel to provide the professional service that a paid lawyer, hired at the outset, can give a client. The lawyer hired at the start of the proceeding will be able to draft an original petition containing all the claims revealed to his trained eye; if the same lawyer is appointed by the court only after the petitioner has demonstrated some merit in an original pro se filing, he and his prisoner client will have no right to state all claims by adding to the original petition, unless the lawyer happens to be appointed and able to get up to speed before the statute of limitations runs out. The rule the Court adopts today may not make much difference to prisoners with enough money to hire their own counsel; but it will matter a great deal to poor prisoners who need appointed counsel to see and plead facts showing a colorable basis for relief. 9
The Court of Appeals got it right, and I respectfully dissent.
Notes
1. There is a tendency toward the gestalt in reading the phrase, but the three items are distinct, and a party claiming the benefit of the rule need satisfy only one.
2. In any event, it is not clear why it is more “capacious,” ante, at 11, to regard a single trial lasting days or weeks as one transaction or occurrence than it is, for example, to view numerous separate protests filed with the Interstate Commerce Commission over a period of two years (each in response to a different proposed tariff amendment) as one transaction or occurrence, see Clipper Exxpress v. Rocky Mountain Motor Tariff Bureau, Inc., 690 F.2d 1240, 1260, n. 29 (CA9 1982) (“The protests involve a single transaction or occurrence” (emphasis deleted)), cited ante, at 11. The Court responds that in Clipper Exxpress the amendment was “tied to the same operative facts as those initially alleged. ” Ante, at 11, n. 5. But as just noted, those “operative facts” (i.e., the relevant transaction) consisted of a number of separate protests filed with the Interstate Commerce Commission over a period of two years, each in response to a different proposed tariff amendment. This is, to say the least, a rather expansive transaction, much more so in my view than a single trial involving (for all claims stemming from it) the same judge, the same parties, the same attorneys, the same jury, the same indictment, and so on.
3. Neither does the warning on the model habeas petition (that failure to set forth every ground for relief may preclude the presentation of additional grounds later) tell us anything about relation back. The Court implies that it does, ante, at 9, but the language on the form says nothing about relation back, and if the Court’s implication were correct then the warning would also bar amendments filed within the limitations period.
4. By contrast, use at trial of the fruits of the alleged police misconduct would not be a prerequisite to success in an action under Rev. Stat. §1979, 42 U.S.C. § 1983 because such an action would indeed be challenging the conduct itself rather than the custody obtained by use at trial of the fruits of that conduct. Cf. ante, at 12 (citing Jackson v. Suffolk County Homicide Bureau, 135 F.3d 254 (CA2 1998), where the Court of Appeals, in a §1983 case, concluded that two different instances of postarrest police conduct were not part of a single transaction or occurrence). The Court’s analysis thus lies in some tension with our understanding that the signal, defining feature setting habeas cases apart from other tort claims against the State is that they “necessarily demonstrat[e] the invalidity of the conviction,” Heck v. Humphrey, 512 U.S. 477, 481—482 (1994); see generally Wilkinson v. Dotson, 544 U.S. ___, ___ (2005) (slip op., at 3—7).
5. There are other examples of the Court’s describing Felix’s claims with reference to the trial. See ante, at 3 (“Felix’s Fifth Amendment claim rested on the prosecution’s introduction of statements Felix made during pretrial police interrogation… . His Sixth Amendment claim related to the admission of the videotaped statements prosecution witness Kenneth Williams made at a jailhouse interview”); ante, at 4 (“On direct appeal, Felix urged … that the admission of Williams’ videotaped statements violated Felix’s constitutional right to confront the witnesses against him. He did not, however, argue that admission of his own pretrial statements violated his right to protection against self-incrimination”).
6. See Neder v. United States, 527 U.S. 1, 18 (1999) (“The erroneous admission of evidence in violation of the Fifth Amendment’s guarantee against self-incrimination, and the erroneous exclusion of evidence in violation of the right to confront witnesses guaranteed by the Sixth Amendment are both subject to harmless-error analysis under our cases” (citations omitted)); Penry v. Johnson, 532 U.S. 782, 795 (2001) (success on Fifth Amendment self-incrimination claim in habeas case requires showing that the error had “substantial and injurious effect or influence in determining the jury’s verdict” (internal quotation marks omitted)); see also, e.g., Banks v. Dretke, 540 U.S. 668, 691 (2004) (elements of prosecutorial misconduct claim under Brady v. Maryland, 373 U.S. 83 (1963), include showing of prejudice); Donnelly v. DeChristoforo, 416 U.S. 637, 643 (1974) (improper prosecutorial comment not reversible error unless remarks “so infec[t] the trial with unfairness as to make the resulting conviction a denial of due process”); Strickland v. Washington, 466 U.S. 668, 695 (1984) (to find prejudice for purposes of ineffective assistance claim, court “must consider the totality of the evidence before the judge or jury”).
7. The Court asserts that my argument here “builds … on an apparent assumption that claim preclusion operates in habeas cases largely as it does in mine-run civil cases.” Ante, at 15, n. 6. In actuality, the argument rests only on a fact we have previously recognized: that AEDPA’s “restrictions on successive petitions constitute a modified res judicata rule … .” Felker v. Turpin, 518 U.S. 651, 664 (1996).
8. The Court is mistaken in stating that I “would read Rule 15(c)(2)’s words, ‘conduct, transaction, or occurrence,’ into … 28 U.S.C. § 2244(b) and 2255, ¶8 … .” Ante, at 15, n. 6. What I would do is adopt, for purposes of reconciling Rule 15(c)(2) with AEDPA’s one-year statute of limitations, a definition of “transaction” that is consistent with what other sections of AEDPA, those governing second or successive petitions, functionally regard as the relevant “transaction.”
9. It is not that I see the Court’s rule as constitutionally troubling. But this case requires us to apply text that is ambiguous, and the Court’s resolution of that ambiguity is based on the assumption that when Congress authorized the appointment of counsel in habeas cases, it would have intended the appointed lawyer to have one hand tied behind his back, as compared with an attorney hired by a prisoner with money. That is not in my view a sound assumption. (The Court also observes that in this case counsel had plenty of time to file an amended petition, but that fact cannot drive this decision, for the rule the Court adopts today will of course apply in cases other than this one.)