Lee A. Rand v. James Rowland Nadim Khoury, M.D., William Bunnell Roy Lee Johnson Leo R. Estes, 154 F.3d 952

Opinion by Judge TASHIMA; Concurrence by Judge REINHARDT; Concurrence by Judge THOMAS; Dissent by Judge KLEINFELD.

TASHIMA, Circuit Judge:

We address the continued viability and application of the “pro se prisoner fair notice” requirement of Rule 56 of the Federal Rules of Civil Procedure (“Rule 56”) adopted in Klingele v. Eikenberry, 849 F.2d 409 (9th Cir.1988), and modified in Arreola v. Mangaong, 65 F.3d 801 (9th Cir.1995), to require that the notice be given only by the district court itself. The three-judge panel (“panel”) held that remand was required because the district court failed to give the notice required by these cases to the pro se prisoner plaintiff, i.e., that notice given by the moving party was insufficient under Arreola. Rand v. Rowland, 113 F.3d 1520, 1522-24 (9th Cir.1997). On en bane review, we reaffirm the central holding of Klingele that a pro se prisoner is entitled to fair notice of the requirements of Rule 56. We further hold, however, that because Klingele’s central concern was that the prisoner receive fair notice of the requirements of Rule 56 and the consequences of such a motion, Klingele’s notice purpose is fulfilled even though the prisoner does not receive the notice from the district court, so long as notice is given and its contents are complete. We therefore overrule Arreola v. Mangaong, 65 F.3d 801.

We find, however, that the notice given by-defendants in this case was deficient in several respects. Accordingly, we withdraw the panel’s opinion concerning Klingele’s fair notice requirement, 1 and reverse the district court’s grant of summary judgment.

I.

FACTS AND PROCEDURAL BACKGROUND

Lee A. Rand (“Rand”) is a California state prison inmate. He filed a civil rights complaint under 42 U.S.C. § 1983 against several prison officials, most of whom are medical officers (“defendants”). Rand was confined to the prison infirmary of the California Correctional Institution at Tehachapi (“CCI”), after testing HIV-positive. He remained at CCI for six months pending his transfer to the California Medical Facility at Vacaville, which was better able to care for HIV-positive inmates.

Rand claims that his Eighth Amendment rights were violated while he was confined at CCI because he was allegedly denied access to exercise, proper medical care, clothing and heating, and personal hygiene items. He asserted further claims under the First Amendment (right to telephone, right to free exercise of religion) and the Fourteenth Amendment (due process, equal protection). Rand twice asked the district court to appoint counsel for him, but both requests were denied.

Defendants moved for summary judgment on all of Rand’s claims. Their motion contained a section entitled, “Notice of Rules Relating To Summary Judgment.” 2

In 1993, the district court adopted certain of the recommendations of the report of the magistrate judge to whom the matter had been referred (the “1993 Order”). The court granted summary judgment for defendant James Rowland, the former Director of the California Department of Corrections, on the ground that he had been sued only in his supervisory capacity. The 1993 Order also concluded that defendants had not adequately briefed the issues relating to the merits of five of Rand’s claims until they filed their Objections to the Magistrate’s Findings and Recommendations. Therefore, the district court declined to adopt the magistrate judge’s recommendation on these issues and referred these claims back to the magistrate judge.

In 1994, the magistrate judge issued his second report which recommended that defendants’ motion be denied on the issue of qualified immunity, that Rand’s claims for injunctive and declaratory relief be dismissed as moot, and that his claims regarding medical care, heating and clothing, out-of-cell exercise, access to clergy, and access to the law library be dismissed because Rand had not suffered any constitutional deprivations. On January 17, 1995, the district court adopted the magistrate judge’s second report and recommendation, and dismissed the action in its entirety (the “1995 Order”). Rand then appealed from the denials of his requests for counsel, the 1993 Order, and the 1995 Order.

The panel concluded that both the 1993 and the 1995 Orders must be vacated and the matter remanded because the district court failed to provide Rand with the Rule 56 notice required by Klingele, 849 F.2d 409. It held that the Klingele-type notice given by defendants was insufficient because Arreola, 65 F.3d 801, required that the notice be given by the court itself. See Rand, 113 F.3d at 1522-23. We then took the case en banc to review the validity and application of the Klingele rule. Rand v. Rowland, 122 F.3d 39 (9th Cir.1997).

II.

DISCUSSION

A. The Fair Notice Requirement Historical Background

In reviewing the fair notice requirement, we are cognizant that we are not writing on a clean slate. Rather, we reexamine a doctrine that we adopted a decade ago and which has been in existence for three decades. See Hudson v. Hardy, 412 F.2d 1091 (D.C.Cir.1968) (per curiam) (the seminal “fair notice” case which devised the rule that district courts should provide pro se prisoners with “fair notice of the requirements of the summary judgment rule”). Our province and obligation as an en banc court “is to review the current validity of challenged prior decisions.” United States v. Aguon, 851 F.2d 1158, 1167 n. 5 (9th Cir.1988) (en banc).

In reexamining its prior holdings, the Supreme Court considers “a series of prudential and pragmatic considerations designed to test the consistency of overruling a prior decision with the ideal of the rule of law, and to gauge the respective costs of reaffirming and overruling a prior case.” Planned Parenthood v. Casey, 505 U.S. 833, 854, 112 S.Ct. 2791, 120 L.Ed.2d 674 (1992) (joint opinion of O’Connor, Kennedy and Souter, JJ.). The Court will inquire whether the rule has proven to be intolerable simply in defying practical workability; whether the rule is subject to a kind of reliance that would lend a special hardship to the consequences of overruling and add inequity to the cost of repudiation; whether related principles of law have so far developed as to have left the old rule no more than a remnant of abandoned doctrine; or whether facts have so changed, or come to be seen so differently, as to have robbed the old rule of significant application or justification. Id. at 854-55, 112 S.Ct. 2791. Similar concerns permeate our analysis here. See also Aguon, 851 F.2d at 1175-76 (Reinhardt, J., concurring).

As we have noted,-, the fair notice rule originated in the District of Columbia Circuit thirty years ago. See Hudson, 412 F.2d 1091. Emphasizing the disabilities of incarceration, the court held that before granting summary judgment against a pro se prisoner, the district court was required to provide the prisoner with “fair notice” of the requirements of the summary judgment rule. Id. at 1094. Five other circuits, including ours, followed Hudson and adopted the fair notice requirement. See Klingele, 849 F.2d 409; Champion v. Artuz, 76 F.3d 483, 486 (2d Cir.1996) (holding that “summary judgment should not be entered by default against a pro se plaintiff who has not been given any notice that failure to respond will be deemed a default”); Griffith v. Wainwright, 772 F.2d 822, 825 (11th Cir.1985) (holding that “an adverse party must be given express, ten-day notice of the summary judgment rules, of his right to file affidavits or other material in opposition to the motion, and of the consequences of default”); Lewis v. Faulkner, 689 F.2d 100, 102 (7th Cir.1982) (requiring that “a prisoner who is a plaintiff in a civil case and is not represented by counsel is entitled to receive notice of the consequences of failing to respond with affidavits to a motion for summary judgment”); Roseboro v. Garrison, 528 F.2d 309, 310 (4th Cir.1975) (per curiam) (requiring “that the plaintiff be advised of his right to file counter-affidavits or other responsive material and alerted to the fact that his failure to so respond might result in the entry of summary, judgment against him”). Only one circuit has refused to adopt the fair notice requirement for pro se prisoners. See Williams v. Browman, 981 F.2d 901, 903 (6th Cir.1992) (“There is no authority in this Circuit for the proposition that a district court must advise a pro se prisoner of his right to file counter-affidavits or other responsive material or that he must be alerted to the fact that his failure to so respond with such material might result in entry of summary judgment against him.”). 3

We first approved of Hudson’s fair notice requirement in Jacobsen, 790 F.2d at 1364 n. 4, where we refused to apply the rule to pro se non-prisoners. Id. at 1364-67. We expressly adopted Hudson’s fair notice requirement in Klingele, where we held that “[district courts are obligated to advise prisoner pro per litigants of Rule, 56 requirements.” 849 F.2d at 411-12. In Klingele, we adopted a bright-line rule that district courts were required to give the required notice in all cases; we refused to be drawn into a “particularized analysis of each prisoner litigant’s sophistication.” Id. at 411. Several years later, we interpreted Klingele as requiring that the requisite notice could be given only by the district court itself. Arreola, 65 F.3d at 802 (reversing and remanding case where summary, judgment movant had provided pro se prisoner with notice of Rule 56’s requirements).

B. Need for the Fair Notice Rule

We begin our evaluation of the Klingele rule with an acknowledgment of the uniqueness of the summary judgment motion. Unknown at the common law, the summary judgment motion was not introduced in England until 1855 and was restricted to actions upon bills of exchange- and promissory notes. Charles E. Clark & Charles U. Samenow, The Summary Judgment, 38 Yale L.J. 423, 424 (1929). The promulgation of the Federal Rules of Civil Procedure in 1938 marked the first time in the United States that use of the summary judgment procedure was authorized in all civil actions. See 11 Moore’s Federal Practice § 56 App.l00[l] (3d ed.1997); William W Sehwarzer, Alan Hirsch & David J. Barrans, The Analysis and Decision of Summary Judgment Motions, 139 F.R.D. 441, 447 (1991).

However, summary judgment was resisted and disfavored by the courts for many years. 11 Moore’s Federal Practice § 56 App.l00[2]; see, e.g., Poller v. Columbia Broad. Sys., Inc., 368 U.S. 464, 473-75, 82 S.Ct. 486, 7 L.Ed.2d 458 (1962) (summary judgment unavailable in large complex antitrust litigation in which mental staté of defendant corporation’s officers was at issue, as this constituted disputed factual question requiring jury consideration rather than “trial by affidavit”); Sid & Marty Krofft Television Prods., Inc. v. McDonald’s Corp., 562 F.2d 1157, 1165 (9th Cir.1977) (summary judgment improper where slightest doubt remained about the facts). It took nearly fifty years for the Supreme Court to pave the way toward mainstream acceptance of the summary judgment procedure with its trilogy of summary judgment cases in the mid-1980s. See Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986); Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986); Celotex Corp. v. Catrett, 477 U.S. 317, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986).

Prior to the 1986 trilogy, we termed the impact of a summary judgment motion “drastic.” See, e.g., Mutual Fund Invs., Inc. v. Putnam Management Co., 553 F.2d 620, 624 (9th Cir.1977); Zephyr Cove Lodge, Inc. v. First Nat'l Bank, 478 F.2d 1121, 1127 (9th Cir.1973) (Ely, J., dissenting); Consolidated Elec. Co. v. United States ex rel. Gough Indus., Inc., 355 F.2d 437, 438 (9th Cir.1966); Hoffman v. Babbitt Bros. Trading Co., 203 F.2d 636, 637 n. 1 (9th Cir.1953). We long recognized that summary judgment affects substantive rights. See Hoffman, 203 F.2d at 637 n. 1. While a complaint that is sought to be dismissed under Rule 12(b)(6) can usually be amended, see, e.g., Foman v. Davis, 371 U.S. 178, 182, 83 S.Ct. 227, 9 L.Ed.2d 222 (1962); 10A Charles Alan Wright, Arthur R. Miller & Mary Kay Kane, Federal Practice and Procedure: Civil 2d § 2713 (2d ed.1983), a grant of summary judgment is a final adjudication on the merits. See Griffith, 772 F.2d at 825 n. 4; 18 Charles Alan Wright, Arthur R. Miller & Edward H. Cooper, Federal Practice and Procedure: Jurisdiction § 4444 (1981).

The summary judgment motion requires the pro se prisoner to confront a myriad of challenges. 4 Unschooled in the intricacies of civil procedure, the lay litigant’s intuition is that his or her claim will proceed to trial regardless of the outcome of a summary judgment motion.” See Lewis, 689 F.2d at 102. For pro se prisoners, this intuition is reinforced by their experience with the criminal justice system, which has no counterpart to the summary judgment motion. In sum, “it would not be realistic to impute to a prison inmate ... an instinctual awareness that the purpose of a motion for summary judgment is to head off a full-scale trial by conducting a trial in miniature, on affidavits, so that not submitting affidavits is the equivalent of not presenting any evidence at trial.” Jacobsen, 790 F.2d at 1364 n. 4.

The fair notice requirement naturally follows from our policy of liberal construction in favor of pro se litigants. Litigants have a statutory right to self-representation in civil matters. 28 U.S.C. § 1654 (1982). They must be ensured meaningful access to the courts. See Bounds v. Smith, 430 U.S. 817, 823, 97 S.Ct. 1491, 52 L.Ed.2d 72 (1977); Wolff v. McDonnell, 418 U.S. 539, 579, 94 S.Ct. 2963, 41 L.Ed.2d 935 (1974); Johnson v. Avery, 393 U.S. 483, 485, 89 S.Ct. 747, 21 L.Ed.2d 718 (1969); Hatfield v. Bailleaux, 290 F.2d 632, 637 (9th Cir.1961). Consequently, we tolerate informalities from civil pro se litigants. See Haines v. Kerner, 404 U.S. 519, 520-21, 92 S.Ct. 594, 30 L.Ed.2d 652 (1972) (pro se pleadings are held to less stringent standards than those prepared by attorneys). The concern for meaningful access for the pro se litigant provides a basis for the fair notice rule, but it is not sufficient in itself to justify its application. See Jacobsen, 790 F.2d at 1364-67 (refusing to extend the Hudson rule to non-prisoner pro se litigants).

Rather, we adopted the Hudson rule because of our concern for the pro se prisoner litigant who also faces the unique handicaps of incarceration. See Jacobsen, 790 F.2d at 1364 n. 4. In the intervening years, nothing has changed to make us conclude that the handicaps prisoners face in prosecuting their own cases have somehow diminished. The reality of incarceration is still that prisoners are not at liberty to seek out representation. See Jacobsen, 790 F.2d at 1364 n. 4 (citing Hudson, 412 F.2d at 1094) (stating that prisoners probably would be unable to retain counsel even if they had the financial means to do so). Prisoners are limited in their ability to interview witnesses and seek out evidence that they must submit in order to defeat a motion for summary judgment. See Jacobsen, 790 F.2d at 1364 n. 4; Hudson, 412 F.2d at 1095 (prisoners may lack sufficient access to sources of proof needed to demonstrate the existence of a question of material fact). Pro se prisoners are also limited in their access to legal materials. Jacobsen, 790 F.2d at 1364 n. 4 (citing Moore v. Florida, 703 F.2d 516, 520 (11th Cir.1983)). Finally, confinement makes compliance with procedural deadlines difficult because of restrictions on the prisoner’s ability to monitor the lawsuit’s progress. Cf. Houston v. Lack, 487 U.S. 266, 270-71, 108 S.Ct. 2379, 101 L.Ed.2d 245 (1988) (interpreting Fed. R.App. P. 4 to find that pro se prisoner’s notice of appeal is “filed” when delivered to prison officials).

The fair notice rule is consistent with the Supreme Court’s teaching that affirmative measures are sometimes required to ensure that a prisoner’s access to the courts is “adequate, effective, and meaningful.” Bounds, 430 U.S. at 822-23, 97 S.Ct. 1491 (prisoners’ right of access to courts requires states to assist prisoners in filing legal papers in all cases which in turn requires the provision of adequate law library or assistance from legally-trained personnel). For all of these reasons, we believe that the twin infirmities of imprisonment and proceeding without counsel, which provided the basis for the creation of the fair notice rule, remain solid grounds for the rule’s retention and its restriction to this class of litigants. See Jacobsen, 790 F.2d at 1364 n. 4; Hudson, 412 F.2d at 1095.

The dissent and the panel concurrence, Rand, 113 F.3d at 1526 (O’Scannlain, J., specially concurring), question whether the Klingele rule constitutes an improper amendment to Rule 56. In Jacobsen, we indicated that the proper route for extending the notice requirement to all pro se litigants was through the normal rule amendment process rather than'through adjudication. 790 F.2d at 1366. However, we explicitly excluded the fair notice rule of Hudson from our analysis. Id. at 1364 n. 4. We are confident that the institution of the fair notice rule in Klingele was and is a proper exercise of judicial authority.

The Seventh Circuit has held that the fair notice requirement is implicit in Rule 56 itself. See Lewis, 689 F.2d at 101-02. The Lewis court held that the implication that the judge should give the party opposing a summary judgment motion a reasonable opportunity to submit counter-affidavits was unmistakable, especially in light of subsection (f) of Rule 56. 5 See id. at 101. Then turning to the conversion of a Rule 12(b)(6) motion into a Rule 56 motion and the parties’ statutory right to a reasonable opportunity to present all materials pertinent to a Rule 56 motion, the court concluded, “[w]e cannot believe that the draftsmen of the Federal Rules intended a summary judgment opponent to have a reasonable opportunity to present factual materials only where the motion was captioned as a Rule 12(b)(6) motion.... ” Id.

We are in agreement with this interpretation of Rule 56 because it effectuates the purpose of the Federal Rules to eliminate “procedural booby traps” which could prevent “unsophisticated litigants from ever having their day in court.” Surowitz v. Hilton Hotels Corp., 383 U.S. 363, 373, 86 S.Ct. 845, 15 L.Ed.2d 807 (1966); see also Martin v. Reynolds Metals Corp., 297 F.2d 49, 56 (9th Cir.1961) (“One of the purposes of the Federal Rules of Civil Procedure was to take the sporting element out of litigation ... ”); Fed. R.Civ.P. 1 (purpose of Rules is “to secure the just, speedy, and inexpensive determination of every action”) (emphasis added).

Adoption of the fair notice requirement is also a proper exercise of our supervisory power over the courts within our circuit. United States v. Jacobs, 429 U.S. 909, 911, 97 S.Ct. 299, 50 L.Ed.2d 277 (1976) (Marshall, J., dissenting) (recognizing the “well-established supervisory power of the courts of appeals over the district courts in their respective jurisdictions”). This supervisory authority can be exercised when “necessary to proper judicial administration in the federal system.” La Buy v. Howes Leather Co., 352 U.S. 249, 259-60, 77 S.Ct. 309, 1 L.Ed.2d 290 (1957); see Burton v. United States, 483 F.2d 1182, 1187 (9th Cir.1973) (quoting La Buy) (finding that it was the court’s duty under its supervisory power to assure strictest compliance with Fed.R.Crim.P. 11). We may invoke our supervisory power to establish procedural rules that are not specifically mandated by statute or by the Constitution. See United States v. Rubio-Villareal, 967 F.2d 294, 297 (9th Cir.1992) (knowledge jury instruction); United States v. Seawell, 550 F.2d 1159, 1162-63 (9th Cir.1977) (Allen charge); United States v. Gutierrez-Espinosa, 516 F.2d 249, 250 (9th Cir.1975) (“presumption of truth” jury instruction); cf. Jacobs, 429 U.S. at 911, 97 S.Ct. 299 (Marshall, J., dissenting) (supervisory power is not limited to enforcing constitutional rights). Thus, we may require the district courts within our circuit “to follow procedures deemed desirable from the viewpoint of sound judicial practice although in no-wise commanded by statute or by the Constitution.” Cupp v. Naughten, 414 U.S. 141, 146, 94 S.Ct. 396, 38 L.Ed.2d 368 (1973). We believe that the retention of the Klingele rule is “sound judicial practice.” 6

Finally, we find it significant that the Klin-gele notice rule has proven practical and workable. At oral argument, the state’s at-' torney, who has litigated prisoner rights cases for many years, informed the court that Klingele notice is routinely and easily given in all of the district courts before which he practices. In fact, based on the notions of fairness we outlined above, he agreed with Rand in urging this court to retain the Klin-gele notice requirement for pro se prisoner litigants; the only disagreement between the parties in this regard is whether the notice may come from the moving party, as well as from the court. We now turn to that issue.

C. Source of the Notice

Having determined that we will retain the Klingele rule, we must now decide who should give that fair notice to the pro se prisoner. Rand argues that the district court is the only proper and legitimate source of such notice. See Arreola, 65 F.3d at 802. We cannot agree. Although Rand’s argument that an inmate may naturally be inclined to distrust any information conveyed' by his adversaries who may also be his jailers is not without force, it is not sufficiently compelling to adopt a rule which elevates source over substance of a properly worded notice.

Until now, we have never explicitly considered who must give the prisoner notice. Based on our language in Klingele, that “[district courts are obligated to advise prisoner pro per litigants of Rule 56 requirements,” 849 F.2d at 411-12, we held in Ar-reola that only the district court can provide the notice. 65 F.3d at 802. We conclude, however, that this terse per curiam decision failed fully to consider the issue, resulting in a rule which “so elevated form over substance” so as to subsume any purported justification for Klingele’s creation. Rand, 113 F.3d at 1525, 1527 (O’Scannlain, J., specially concurring).

The central purpose of the Klingele rule is to make the pro se prisoner aware of the requirements and consequences of Rule 56. See Klingele, 849 F.2d at 411; Hudson, 412 F.2d at 1094. Although it may be preferable for the district court to assume this task, we do not agree that it cannot be performed adequately by the movant for summary judgment. See Lewis, 689 F.2d at 102 (“[W]e trust that counsel for the defendants in prisoner civil rights cases in this circuit will lift this new burden from the judges’ shoulders, by henceforth including in any motion for summary"judgment in a case where the plaintiff is not assisted by counsel a short and plain statement that any factual assertion in the movant’s affidavits will be accepted as being true unless the plaintiff submits his own affidavits or other documentary evidence contradicting the assertion.”). We, thus, conclude that either the court or the moving party may provide the requisite notice. See id. at 103. We believe that the concerns identified by Rand are sufficiently met if the required notice from the moving party is a separate document, served with the moving papers, and states that the court has required that it be given. 7

In sum, Klingele’s requirement that “[district courts are obligated, to advise prisoner pro per litigants of Rule 56 requirements,” 849 F.2d at 411-12, may be met by the summary judgment movant providing the prisoner with notice. If the movant provides the notice, it must be in a separate form that the plaintiff will recognize as given pursuant to the court’s requirement. It may not be provided within the summary judgment motion or in the papers ordinarily filed in support of the motion. Of course, the ultimate responsibility of assuring that the prisoner receives fair notice remains with the district court.

D. The Contents of the Notice

In delineating the contents of the Klingele notice, we are aided by the panel’s comprehensive discussion of the issue. Rand, 113 F.3d at 1524-25. We believe that the starting point for the contents of the notice is to be found in the reasons for the creation of the Klingele rule — the complexity of the summary judgment rule combined with the lack of legal sophistication of the pro se prisoner. See Jacobsen, 790 F.2d at 1364 n. 4 (citing Lewis, 689 F.2d at 102). Thus, to begin, the notice must be phrased in ordinary, understandable language calculated to apprise an unsophisticated prisoner of his or her rights and obligations under Rule 56. See Moore, 703 F.2d at 521 (mandating that the “notice must be sufficiently clear to be understood by a pro se litigant and calculated to apprise him of what is required under Rule 56”) (quoting Davis v. Zahradnick, 600 F.2d 458, 460 (4th Cir.1979)); see also Timms v. Frank, 953 F.2d 281, 285 (7th Cir.1992) (requiring a “short and plain statement in ordinary English”). Second, we reaffirm the holdings of Jacobsen and Klingele requiring that the pro se prisoner be informed of his or her right to file counter-affidavits or other responsive evidentiary materials and be alerted to the fact that the failure to do so might result in the entry of summary judgment against the prisoner. Klingele, 849 F.2d at 411; Jacobsen, 790 F.2d at 1365 n. 8. Third, and most importantly, the pro se prisoner must be informed of the effect of losing on summary judgment. Any notice must clearly and conspicuously inform the pro se prisoner that if the motion for summary judgment is granted, the prisoner’s case will be over. The notice should also state that if the pro se prisoner fails to controvert the moving party with opposing counter-affidavits or other evidence, the moving party’s evidence might be taken as the truth, and final judgment -may be entered against the prisoner without a trial. See Champion, 76 F.3d at 486; Griffith, 772 F.2d at 825; Moore, 703 F.2d at 520-21 (citation omitted). Finally, if any local rule provides for additional requirements, such as a statement of genuine issues, the notice must include a short and plain statement of the local rule’s requirements. As noted earlier, if given by the moving party, the notice must indicate that it is required to be given by the court. 8

E. Harmless Error Inquiry

In Klingele, we adopted a bright-line rule that notice of Rule 56 must be given in every prisoner pro se case. We declined to assume, or to require district courts to assume, the burdensome task of assessing each particular litigant’s sophistication in legal matters, and held that a failure to give adequate notice was reversible error, without engaging in a harmless error analysis. Klingele, 849 F.2d at 411. We continue to believe that harmless error review is inappropriate in most eases. We will not assume the impossible task of attempting to divine whether an individual inmate understood what was at stake if he failed to put all of his evidence before the court. The record before this court demonstrates that, in the absence of instruction to the contrary, pro se prisoners are apt to hold back available relevant evidence at the summary judgment stage. Here, Rand may have failed to present available expert evidence and his own deposition testimony which would have been helpful in opposing defendants’ motion for summary, judgment.

We do not believe that Rule 61’s harmless error provision prevents us from eschewing a case-by-case harmless error inquiry in these circumstances. Rule 61 was never meant to be'applied in circumstances where the harmfulness of the error could not be fully assessed on the record. We thus presume prejudice and forego harmless error inquiry in other contexts. We do this when the district court converts a Rule 12(b)(6) motion into a Rule 56 motion without proper notice to' a pro se litigant. See Garaux v. Pulley, 739 F.2d 437, 439 (9th Cir.1984) (requiring stringent adherence to notice requirement upon conversion of motion to dismiss to motion for summary judgment). See also Waters v. Young, 100 F.3d 1437, 1442 (9th Cir.1996) (failure to give Rule 50(a) notice to pro se litigant constitutes per se reversible error without regard to prejudice). We see no reason to depart from this practice in the context of a fair notice violation, except in the unusual case.

We are supported in our position by the practice of the majority of the other circuits who have adopted the fair notice requirement. See, e.g., Graham v. Lewinski, 848 F.2d 342, 344 (2d Cir.1988) (vacating and remanding without engaging in harmless error analysis); Griffith, 772 F.2d at 825 (same); Ham v. Smith 653 F.2d 628, 631 (D.C.Cir.1981) (rejecting government’s position that court should affirm the district court’s judgment because prisoner’s suit was frivolous and declining to inquire into merits of the lawsuit before prisoner was given an opportunity to respond to summary judgment motion); Roseboro, 528 F.2d at 310 (vacating and remanding without engaging in harmless error analysis); but see Timms, 953 F.2d at 286 (engaging in harmless error analysis and holding that pro se litigant is not entitled to relief if pro se cannot prove any set of facts in support of claim entitling claimant to relief).

We recognize, however; that there may be the unusual ease where the harmlessness of the failure to give the required notice may be established on the record or by judicial notice. For example, judicial notice by the district court of its own records, either at the behest of the defendant or sua sponte, may disclose that the plaintiff had recently been served with a Klingele notice in prior litigation. Similarly, an objective examination of the record may disclose that the pro se prisoner litigant has a complete understanding of Rule 56’s requirements gained from some other source. One such example would be where the plaintiffs response to the defendant’s summary judgment motion itself cites this case and evidences a familiarity with Appendix “A” to this opinion.

Because Klingele establishes notice as a substantial right, any error in failing to provide that notice will necessarily “affeet[ ] the substantial rights of the parties,” 28 U.S.C. § 2111; Fed.R.Civ.P. 61, in most cases. And it would only be in the unusual case that the record will establish that failure to receive a Klingele notice is harmless. We refuse, however, to embark upon the type of subjective scrutiny of the prisoner’s pleadings in which the dissent engages in order to reach its conclusion that Rand must have been “well aware of the requirements of Rule 56.” We emphasize that this search for harmlessness in the exceptional case must be undertaken on an objective basis. 9

F. Adequacy of the Notice to Rand

We now turn to the notice provided by defendants in this case. A two-page notice entitled “Notice of Rules Relating to Summary Judgment” was included in the defendants’ Notice of Motion for Summary Judgment. See supra, n. 2. Although the notice was a commendable attempt to comply with Klingele, we are forced to find it wanting in several respects. First, the notice was not phrased in ordinary understandable language calculated to apprise an unsophisticated prisoner of his rights and obligations under Rule 56. It was not a “short and plain statement in ordinary English.” Timms, 953 F.2d at 285. Rather, peppered with citations to relevant legal authority and quotes from relevant caselaw, the notice appears calculated more to inform a legal practitioner (much like a continuing legal education summary) than a legal novice. Second, we note that the notice did meet our requirement that the pro se prisoner be informed of his or her right to file counter-affidavits or other responsive materials. It also alerted Rand to the fact that the failure to do so might result in the entry of summary judgment against him. The notice stated that Rand could not simply rely on the allegations of his complaint. Most importantly, however, the notice failed to inform Rand of the effect of losing the summary judgment motion. It nowhere stated that if the motion were granted, Rand would have no trial and his case would be over. Accordingly, we reverse the entry of summary judgment against Rand, and remand the ease to the district court. 10

III.

CONCLUSION

We reaffirm our decade-old rule under Klingele that a pro se prisoner must be given fair notice of the requirements of Rule 56. Because such notice may be given by the moving party, we overrule Arreola which held to the contrary and withdraw all of the panel’s opinion, except Part II.C. We find, however, that the notice provided by defendants to Rand failed to meet the fair notice requirements. We therefore reverse the district court’s grant of summary judgment for defendants and remand the case to the district court for further proceedings.

REVERSED and REMANDED.

APPENDIX “A”

NOTICE — WARNING

This Notice is Required to be Given to You by The Court

The defendants have made a motion for summary judgment by which they seek to have your case dismissed. A motion for summary judgment under Rule 56 of the Federal Rules of Civil Procedure will, if granted, end your case.

Rule 56 tells you what you must do in order to oppose a motion for summary judgment. Generally, summary judgment must be granted when there is no genuine issue of material fact — that is, if there is no real dispute about any fact that would affect the result of your case, the party who asked for summary judgment is entitled to judgment as a matter of law, which will end your case. When a party you are suing makes a motion for summary judgment that is properly supported by declarations (or other sworn testimony), you cannot simply rely on what your complaint says. Instead, you must set out specific facts in declarations, depositions, answers to interrogatories, or authenticated documents, as provided in Rule 56(e), that contradict the facts shown in the defendant’s declarations and documents and show that there is a genuine issue of material fact for trial. If you do not submit your own evidence in opposition, summary judgment, if appropriate, may be entered against you. If summary judgment is granted, your case will be dismissed and there will be no trial.

[Local Rule_of the District Court also requires, in addition, that you include as a part of your opposition to a motion for summary judgment_]

REINHARDT, Circuit Judge, concurring.

I join fully in Judge Tashima’s excellent opinion. I write separately only to make one point explicit. We consider here, as Judge Tashima points out, a doctrine that has been in existence for three decades, that only one circuit court has rejected, that has worked extremely well, and that both sides 'urge us to retain. Under these circumstances, the considerations employed by the Supreme Court in deciding whether it should overrule its own precedents are useful in guiding our decision as to whether to overrule the Klin-gele doctrine.

In other circumstances, however, our criteria will be considerably different, as the majority opinion indicates by its citation to my separate concurrence in United States v. Aguon, 851 F.2d 1158 (9th Cir.1988) (en banc) and by the holding it announces in Section C. The role of an en banc court in determining whether to overrule circuit precedent, and the factors it must consider, are in most cases not similar to the role played by the Supreme Court when deciding whether to overrule its precedents or the considerations the that Court must weigh in that process. I have explored these differences fully in my earlier Aguon concurrence and will not reiterate them here. I will only repeat that ordinarily a decision of the en banc court as to whether to overrule or modify circuit precedent will involve a variety of factors, a number of which are clearly not pertinent to the Supreme Court’s determinations as to whether to overrule its own prior law. One such factor, for instance, may be whether the other circuits have considéred our opinion and rejected it in a manner that persuades us that we erred, a factor that could hardly even arise in Supreme Court jurisprudence. 1

While we applied the Supreme Court’s Casey standard in order to preserve and affirm the historic Klingele doctrine, if we applied that standard in all circumstances, we would not be overruling Arreola. The overruling of Arreola, which relates only to a particular implementation of the basic doctrine, is fully consistent with the use of the factors set forth in my concurring opinion in Aguon, but clearly does not meet the criteria set forth in Justices O’Connor, Kennedy and Souter’s Casey opinion, at least as they are described in Judge Tashima’s majority opinion. I point this out not by way of criticism or disagreement with the majority opinion, but solely in order to make it clear that the Casey standard, while it may-be applicable to the question whether to overrule the historic Klingele doctrine, is neither the standard we employ in Section C of our opinion nor the standard that we use generally as an en bane court when determining whether to overrule our own prior precedent. After all, overruling precedent is an important function that is expressly reserved to the en banc court, and a panel is ordinarily not free to do so. See Spinelli v. Gaughan, 12 F.3d 853, 855 n. 1 (9th Cir.1993). The Supreme Court functions in an entirely different manner.

THOMAS, Circuit Judge, with whom JUDGE HAWKINS joins, concurring.

Were I writing on a clean slate, I would not adopt the Klingele rule. I find no de-fendable distinction to be made between prisoner pro se litigants and those whose economic circumstances prevent them from obtaining legal counsel. The reasons traditionally advanced for providing prisoners with procedural notices apply with equal force to non-prisoner pro se litigants, who have more or less successfully labored without a Klingele notice since Jacobsen v. Filler, 790 F.2d 1362 (9th Cir.1986). Indeed, many prisoners have greater access to law libraries and legal assistance than do those without financial means, providing inmates a greater ability to apprise themselves of procedural rules. Thus, at least a portion of the rationale which underlay Hudson v. Hardy, 412 F.2d 1091 (D.C.Cir.1968) has dissipated over the course of the last thirty years.

Additionally, in the years since Hudson was decided, we have developed other procedural protections for pro se litigants. See, e.g., Eldridge v. Block, 832 F.2d 1132, 1136 (9th Cir.1987) (reaffirming that trial courts have a duty to provide pro se litigant notice of complaint deficiencies with opportunity to correct unless amendment would be futile); Balistreri v. Pacifica Police Dept., 901 F.2d 696, 699 (9th Cir.1990) (noting that courts have a general duty “to ensure that pro se litigants do not lose their right to a hearing on the merits of their claim due to ignorance of technical procedural requirements”).

However, we are not confronted with an unmarked tablet, and stare decisis commands that we move with some degree of caution in overturning procedures which have governed our Circuit for a decade. This is particularly true when the litigants before us have not raised an objection to the continued vitality of Klingele; in fact, both parties endorse it and strongly urge its continuation. Thus, although I might be persuaded in an appropriate case to revisit Klingele in its entirety, that issue is not, in my view, properly presented by this appeal. Therefore, based on the majority opinion and this concurrence, the Klingele notice requirement is preserved.

The issues squarely before us are whether (1) Arreola v. Mangaong, 65 F.3d 801 (9th Cir.1995) should be overruled, and (2) whether failure to provide Klingele notice is reversible or harmless error. The majority opinion holds that failure to provide Klingele notice will be reversible error subject to a harmless error exception in limited circumstances. On these points, I readily agree for the reasons enunciated by the majority.

KLEINFELD, Circuit Judge, joined by TROTT, FERNANDEZ, RYMER, and T.G. NELSON, Circuit Judges, dissenting.

I respectfully dissent. We have mistakenly preserved a rule that affects pro se prisoner cases, about a fourth of our case load. In the course of doing so, we have gone beyond the bounds of our authority to make rules for district courts, and we have confusingly confined the law of harmless error.

Fortunately, we have today overruled Arreola v. Mangaong, 65 F.3d 801 (9th Cir.1995). In Arreola, we held that if the party moving for summary judgment sent the prisoner respondent exactly the same notice as the district court was supposed to send under Klingele v. Eikenberry, 849 F.2d 409 (9th Cir.1988), we would automatically reverse because of the district court’s failure to send the notice. No more. Now the movant can avoid a Klingele reversal by sending the respondent a copy of our form, Appendix A to the majority opinion, with the motion for summary judgment. Because it is now easy for frequent litigants with form banks, like state attorneys general offices, to avoid Klin-gele reversals of summary judgments, Klin-gele has been turned from an entirely arbitrary burden that falls on litigants when a district court deputy clerk makes a mistake, into a trap for the unwary that will affect lawyers and litigants in their first cases against prisoner pro ses.

Unfortunately, we have a retained the Klingele rule. It might be desirable if all litigants fully understood how summary judgment procedure works. There is no sound basis in the law for singling out pro se prisoner litigants for special education. Nor is the special education we require, mailing of a boilerplate form, likely to do them or anyone else much good. Nor is it especially fair. We do not require Klingele advice for pro se litigants not in prison, whose claims are far more likely to be meritorious, such as unpaid subcontractors on government construction jobs whose claims are too small to hire lawyers on contingent fee. If a prisoner pro se sues a defendant who is also pro se but not in prison, we give special assistance under Klin-gele to the litigant in prison, a senselessly unfair approach.

The greatest unfairness of Klingele is to victims of unmeritorious prisoners’ lawsuits. The case at bar has been pending for nine years! The prison employees Rand sued have had to live with it all that time, perhaps reporting it when they applied for credit or liability insurance, perhaps facing personal liability and lost sleep on account of being sued. Yet despite Rand’s failure after nine years to come up with evidence, it shall continue to go on and on because of our reversal.

The majority fortunately has reduced the harm caused by Klingele, by holding that failure to send a Klingele explanation may be harmless error. But the majority’s novel version of the harmless error rule, using the terms “objective,” “subjective,” “exceptional,” and “unusual,” is mistaken. The statutes and rales do not entitle us to put our thumbs on the scale with these adjectives. I am concerned that application of harmless error doctrine in other kinds of cases may be infected by our error in this one. So I am sorry to say that in this en banc rehearing, we have caused almost as much trouble as we have cured.

I. Klingele mistaken.

A. We lack authority to impose Klin-gele.

We invented, in Klingele v. Eikenberry, 849 F.2d 409 (9th Cir.1988), a rale that requires the district court to advise pro se prisoners about the requirements of Federal Rule of Civil Procedure 56 before their claims are dismissed on summary judgment. Our rale in Klingele had no basis in the United States Code or the Federal Rules of Civil Procedure. We lack authority to impose general rales for district courts.

Congress gave the authority to make procedural rales for the district courts to the Supreme Court and the district courts, not to the courts of appeal. We have rule-making authority for our own court, at 28 U.S.C. § 2071(a), but not for district courts. The law on who has the power to make general rales of procedure for the district courts is clear: “The Supreme Court shall have the power to prescribe general rales of practice and procedure .'.. for cases in the United States district courts_” 28 U.S.C. § 2072(a). The procedure is elaborate, and includes various bodies, but does not include the courts of appeal. See 28 U.S.C. §§ 2072(b), 2078.

Congress .has clearly provided that district courts may, by prescribed procedures, promulgate rules consistent with the Federal Rules of Civil Procedure and the United States Code for management of pro se litigation and summary judgment procedure: “all courts established by Act of Congress may from time to time prescribe rules for the conduct of their business.” 28 U.S.C. § 2071(a) (emphasis added); see also Fed. R.Civ.P. 83(a). The word “their” means that a court can make rules for itself, but not for other courts. Under this unambiguous scheme, the general rales are uniform nationally, made after consideration of views from many sources. Congress takes advantage of local knowledge by enabling courts to make rules not inconsistent with the national scheme for governance of their own affairs, about which their judges have direct knowledge. Circuit judges are in between those who have direct local knowledge and those who can make uniform national rules; we do not have much to contribute.

The Supreme Court has protected against issuance of rules by federal courts outside their statutory authority, in order to assure “exacting observance of the statutory procedures”:

The problem then is one which peculiarly calls for exacting observance of the statutory procedures surrounding the rule-making powers of the Court, see 28 U.S.C. § 331 (advisory function of Judicial Conference), 28 U.S.C. § 2073 (prior report of proposed rule to Congress), designed to insure that basic procedural innovations shall be introduced only after mature consideration of informed opinion from all relevant quarters, with all the opportunities for comprehensive and integrated treatment which such consideration affords.

Miner v. Atlass, 363 U.S. 641, 650, 80 S.Ct. 1300, 4 L.Ed.2d 1462 (1960). See also Charles Alan Wright, Arthur R. Miller & Richard L. Marcus, 12 Federal Practice & Procedure. § 3153 at 516. The majority takes the view that we can infer authority “to enact basic procedural innovations” from the need to protect prisoners’ access to the courts and from our general supervisory power.

We do not have “supervisory -power” over district courts so broad that we can exercise authority that Congress expressly gave only to other institutions. “The supervisory power is part of the common law, and no court has common law power to disregard a rule or statute that was within the authority of Congress to enact.” United States v. Widgery, 778 F.2d 325, 329 (7th Cir.1985).

The majority relies heavily on dicta in Jacobsen v. Filler, 790 F.2d 1362 (9th Cir.1986), to justify the exercise of rule making authority, But Jacobsen rejected extension of prisoner protections to non-prisoner pro se cases; it did not decide anything about the rights of pro se prisoners who file lawsuits. And Jacobsen justifies its refusal to expand pro se rights by noting that even if desirable, expansion would (as in the case at bar) exceed our rule-making authority:

Finally, even if a substantive notice requirement were desirable, it should be enacted through formal amendment rather than piecemeal adjudication. Rule 56’s separate notice provision (compare Rule 56(c) with Rule 6(d)) and description of summary judgment (compare Rule 56(e) with Rule 12(b)) indicate that the Supreme Court and- its Advisory Committee have considered the special problems raised by the summary judgment procedure and, by failing to require specific notice of the nature of summary judgment, have concluded that the present federal rules ... already apprise litigants of their summary judgment obligations. Requiring additional notice to pro se litigants would be an accretion onto Rule 56(c), not an interpretation of it; and as an ad hoc amendment it would not be standardized, codified, or subject to collective decision making.

Id. at 1366 (footnotes omitted).

B. It is time to reappraise Klingele.

The majority emphasizes how well and long-established Klingele is, as justification for keeping it. A ten year old rule in only a few circuits, rejected by others, is hardly Hadley v. Baxendale, 9 Exch. 341 (1854). Summary judgment, on the other hand, is treated as an upstart in the majority decision, despite being universal in federal civil procedure since 1938.

Of the thirteen federal circuits, seven have no Klingele notice requirements. A minority, six, have something along similar lines as Klingele. The Fifth Circuit has rejected a Klingele-type rule, holding that “[t]he notice afforded by the Rules of Civil Procedure and the local rules are, in our view, sufficient.” Martin v. Harrison County Jail, 975 F.2d 192, 193 (5th Cir.1992). The Eleventh Circuit rule provides for extra notice to all pro se litigants facing summary judgment motions; it does not give preferred treatment to prisoners as ours does. See Griffith v. Wainwright, 772 F.2d 822 (11th Cir.1985). The Seventh Circuit version allows for reversal only if the plaintiff can demonstrate prejudice from the lack of notice. See Sellers v. Henman, 41 F.3d 1100, 1102 (7th Cir.1994). Four circuits (including the Eleventh, discussed above) appear to require automatic reversal if the pro se prisoner in a civil suit is not given notice of what Federal Rule of Civfl Procedure 56 means. See Champion v. Artuz, 76 F.3d 483, 486 (2d Cir.1996); Neal v. Kelly, 963 F.2d 453 (D.C.Cir.1992); Griffith v. Wainwright, 772 F.2d 822 (11th Cir.1985); Roseboro v. Garrison, 528 F.2d 309 (4th Cir.1975).

About a quarter of all the appeals filed in our court are prisoner petitions, mostly pro se. See United States Courts-Ninth Circuit, 1995 Annual Report at 53 (of 8,367 appeals, 2,138 were prisoner petitions). Of course, some of these are extremely serious, such as death penalty habeas corpus petitions. But most are frivolous. People with serious disputes cannot get to the front of the line, cannot get their appeals decided faster, because so much of our time is absorbed with prisoners’ suits. As Justice Kennedy has said, “many of these suits involve our basic charter in support of claims which fall somewhere between the frivolous and the farcical and so foster disrespect for our laws.” Crawford-El v. Britton, 523 U.S. -, 118 S.Ct. 1584, 140 L.Ed.2d 759 (1998) (Kennedy, J., concurring). That is not to say Rand’s case is frivolous or farcical; but the mechanical Klingele rule we hold onto today applies to a class of cases most of which are.

Congress and the Supreme Court have both recently given us good reason to reevaluate our Klingele policy of special assistance for prisoners’ claims facing summary judgment. Congress, recognizing the burden of frivolous prisoners’ suits on the federal courts, and their diversion of courts’ time away from more worthy endeavors, amended the in forma pauperis statute in the 1996 Prison Litigation Reform Act to require dismissal at an early stage. Congress added a new requirement that courts “shall” dismiss a prisoner’s case “at any time” if the court determines that the action “is frivolous or malicious,” or “fails to state a claim on which relief may be granted,” or “seeks monetary relief against a defendant who is immune from such relief.” 28 U.S.C. § 1915(e)(2)(B). Under the new law, prisoners must pay filing feés based on a schedule calibrated to their ability. 28 U.S.C. § 1915(a)(3). Also, under the new statute, Congress prohibited in for-ma pauperis appeals “not taken in good faith.” Id.

The Supreme Court, in the course of explicating certain aspects of qualified immunity, recently reminded us of the desirability of summary judgment as a way to dispose of prisoners’ claims not frivolous on their face, but lacking a genuine issue of material fact or entitlement to relief as a matter of law. The Court said that “summary judgment serves as the ultimate screen to weed out truly insubstantial lawsuits prior to trial.” Crawford-El v. Britton, 523 U.S. -,-, 118 S.Ct. 1584, 1598, 140 L.Ed.2d 759 (1998). Rejecting a novel rule regarding immunity imposed on these cases by the District of Columbia Circuit, the Court said we ought to leave the handling of procedural management to the broad discretion of district judges instead of making up categorical rules:

It is the district judges rather than appellate judges like ourselves who have had the most experience in managing cases in which an official’s intent is an element. Given the wide variety of civil rights and “constitutional tort” claims that trial judges confront, broad discretion in the management of the factfinding process may be more useful and equitable to all the parties than the categorical rule imposed by the Court of Appeals.

Id. at -, 118 S.Ct. at 1598. This one-two punch, the Prison Litigation Reform Act of 1996 and the Crawford-El language in 1998, are sufficient reason for us to reexamine the consistency of Klingele with the United States Code, federal rules, Supreme Court, and fairness to litigants other than pro se prisoners.

I do not think we are doing anything worthwhile, even for prisoners, with our Klingele rule. There, is already a clear explanation of summary judgment procedures in plain English. It is the text of Rule 56. Prisoners can read it, certainly as well as they can read the federal or state criminal statutes they are in prison for violating. See Martin v. Harrison County Jail, 975 F.2d 192, 193 (5th Cir.1992). It is illogical and unfair to tell a man “everyone is charged with knowing the law,” and “ignorance of the law is no excuse,” when we send him to prison for violating a law he did not know about, perhaps as hard to understand as RICO or discharge in violation of an NPDES permit, and then when in prison he sues someone, to say he is not charged with know ing the law governing procedure in his own lawsuit until we send him a personal explanation.

Rule 56 is about as clear as the form the majority opinion appends, and considerably more complete. Our new form, at appendix A of the majority opinion, does not tell prisoners about partial summary judgments, as does Federal Rule of Civil Procedure 56(c). It fails to warn prisoners that affidavits must be “made on personal knowledge,” and that they must state “facts such as would be admissible in evidence,” and “show affirmatively that the affiant is competent to testify to the matters stated therein,” as Rule 56(e) does. It does not tell them about the possible availability of continuances for additional discovery, as Rule 56(f) does. It does not warn prisoners of the deadlines, usually requiring careful parsing of Federal Rules of Civil Procedure 56 and 6 and the Local Rules. Nor does our supposedly helpful explanation explain what it means by a “real dispute about any fact” or enlighten the prisoner on the hearsay rule, competency of witnesses, foundation, and all the other complexities that frequently require district courts to disregard affidavits. It does not even tell prisoners what its own terms of art, such as “declarations,” mean.

This is not to suggest that we should expand our form, or send prisoners to school to learn better how to sue people. We are not supposed to be advocates for a class of litigants, and it is hard to help pro ses very much without being unfair to their adversaries. Appendix A is no worse than any other boilerplate form we are likely to devise. The problem is that no such form is likely to do much good. Sending the prisoners copies of Rule 56 would be better. But if what we said was that district court must send prisoners copies of Rule 56 when summary judgment motions are made against their civil claims, that would beg the question of why we should reverse for failure to send what prisoners already have access to in prison law libraries (sometimes better access than other pro ses, who may live away from accessible law libraries). “It is not sensible for the court to tell laymen that they must file an ‘affidavit’ without at the same time explaining what an affidavit is; that, in turn impels a rudimentary outline of the rules of evidence.” Jacobsen v. Filler, 790 F.2d 1362, 1365 (9th Cir.1986).

There is not much reason to assume, as the majority does, that sending prisoners a form saying a few things about summary judgment procedure will make proceedings fairer. Those who cannot understand Rule 56 are not likely to understand the majority’s form either. We do not achieve more fairness by sending them a piece of paper that has no meaning to them. All we do is generate some arbitrary and capricious reversals in cases where deputy clerks and secretaries to assistant attorneys general neglect to send the meaningless paper.

If we were simply to leave district courts alone in this matter, respecting their own authority and the absence of ours in the controlling statutes, they would likely achieve more substantial fairness for pro ses than our boilerplate form. Sometimes an experienced district judge or magistrate reads a prisoner’s papers, thinks “if this is true, this fellow really may have been wronged,” and issues a sua sponte order directing the prisoner and the prison authorities to produce the papers likely to show whether the claim should go to trial. Individualized processes have practical utility in identifying and fairly adjudicating prisoners’ complaints, unlike Klingele boilerplate forms. Today’s decision may discourage useful district court initiatives to promote fairness, because after reading the California Attorney General’s form that we reject today as saying too much and not enough in not quite the right way, some district judges may think it is too dangerous to experiment.

The pleadings stage is more dangerous to good claims than the summary judgment stage, because pro se prisoner petitions are often hard to understand, yet the statute plainly commands dismissal of frivolous claims “before docketing.” 28 U.S.C. § 1915A. There are plenty of ways to lose a meritorious case at every stage, and it is capricious to single out summary judgment for a special paperwork requirement.

There is no justification for treating prisoners’ complaints with special solicitude that we do not give to other pro se complaints. It seems to be assumed by some that prisoners cannot hire lawyers, but no facts have been offered to support that assumption. An occasional prisoner may have a great deal of money, and many have some. And of those prisoners with good claims for substantial money damages, all have plenty of money, for purposes of having the ability to hire a lawyer.

The contingent fee and 42 U.S.C. § 1988 give every prisoner with a good claim for substantial damages the financial ability to assure a lawyer adequate recompense for prosecuting the prisoner’s case. We have approved § 1988 awards where the attorneys obtained many times as much money as they got for their clients. See, e.g., Morales v. City of San Rafael, 96 F.3d 359 (9th Cir.1996) ($139,783.25 for the lawyer, $17,500 for the client, for an arrest without probable cause). Millions of dollars have been awarded in a single case to prisoners’ lawyers. See Gates v. Deukmejian, 987 F.2d 1392 (9th Cir.1992).

That a person litigates pro se may indicate that the person simply did not have a meritorious and substantial enough claim to induce a lawyer to take the ease. That is among the many reasons why judges must avoid “undesirable, open-ended participation by the court” in eases with pro ses on one side. Rand v. Rowland, 113 F.3d 1520, 1526 (9th Cir.1997) (O’Scannlain, J., concurring).

II. Harmless error.

A. We must disregard harmless error.

The majority opinion contradicts itself on harmless error. First it says that in Klin-gele we “held that a failure to give adequate notice was reversible error, without engaging in harmless error analysis.... We continue to believe that harmless error review is inappropriate in most cases.” That means no affirmances, when the Klingele form has not been sent out, based on harmless error. The majority opinion says harmless error analysis is an “impossible task.” Then the majority opinion says we can affirm despite failure to send a Klingele form in “the unusual case where the harmlessness of the failure to give the required notice may be established on the record or by judicial notice.” But that means the analysis of harmless error in Klin-gele was not correct, and harmless error review is not an impossible task.

The majority suggests slipping the camel through the needle’s eye with the terms “unusual,” “exceptional,” and “objective.” Evidently we are to affirm despite failure of the district court or counsel to send the pro se a Klingele form in the. “unusual” and “exceptional” case where “objective” rather than “subjective” analysis .of fhe record shows harmlessness. This novel reformulation of harmless error doctrine conflicts with controlling authority and. does not make practical sense.

The reason that we do not have authority to invent a new kind of harmless error rule is that Congress and the federal rules have already laid down the law. We have to apply it, not substitute different rules some of us may prefer. Congress has commanded that in “any” case at all, which would include a pro se prisoner’s civil lawsuit, we must disregard errors “which do not affect the substantial rights of parties”:

§ 2111. Harmless error.

On the hearing of any appeal or writ of certiorari in any case, the court shall give judgment after an examination of the record without regard to errors or defects which do not affect the substantial rights of the parties.

28 U.S.C. § 2111 (emphasis added). Failure to send a Klingele form, even if we adhere to Klingele, affects a prisoner’s procedural right to receive the form, not necessarily any substantive right to relief on account of the wrong he sues for. The statute means that unless failure to send the Klingele form affects the substantive right for the violation of which the prisoner sues, we are prohibited from reversing, whether the case is “unusual” and “exceptional” or not. . .

Federal Rule of Civil Procedure 61 says that “no error or defect ... in anything ... omitted by the court or by any of the parties is ground for ... vacating ... or otherwise disturbing a judgment or order, unless refusal to take such action appears to the court inconsistent with substantial justice.” This means that even if the district court and opposing counsel omit the Klingele form from whatever papers they send the prison-, er, we are not permitted to vacate a summary judgment, unless the failure appears to the appellate panel inconsistent with “substantial” justice. “Substantial” justice means justice relating to substance rather than form. This rule does not allow us to flee from the exercise of judgment, as the majority’s “objective” test suggests. We “must disregard any error or defect in the proceeding which does not affect the substantial rights of the parties.”

The word “must” in the second sentence of the rule and the phrase “appears to the court inconsistent with substantial justice”,in the first sentence mean we are commanded to exercise judgment based on how the record looks to the judges on the panel. And our judgment has to focus on “substantial” justice, not on such “objective” procedural minutiae as whether the prisoner received a Klin-gele form in another lawsuit he filed. We appellate judges routinely do what Rule 61 says, as we must, on other errors. We read records and decide whether it appears to us that errors affected substantial justice. The rule requires us to do the same thing when the district court and counsel fail to send Klingele forms.

There is no way to reconcile today’s majority opinion, limiting harmless error to “exceptional” and “unusual” cases, with the command of the statute and rule that we disregard harmless error in “any” case. Likewise, there is no way to reconcile the majority’s restriction of harmless error affirmances to “objective” determinations involving no exercise of judgment (a deputy clerk or a well-programmed database could see whether there was a Klingele form in another file with the same prisoner as plaintiff) with the rule requiring us to decide whether the error “appears to the court inconsistent with substantial justice.” The flight from judgment required by today’s majority opinion violates Rule 61.

The reason why Congress commanded us to disregard harmless error is that it has been engaged in a century-long campaign to move courts away from their nineteenth century tendency to reverse for technicalities without substantive significance. “Congress enacted (or approved) harmless error rules to reverse” the approach that in the nineteenth century had made appellate courts “ ‘impregnable citadels of technicality.’ ” United States v. Widgery, 778 F.2d 325, 329 (7th Cir.1985) (quoting Roger Trayner, The Riddle of Harmless Error 14 (1970)). What matters about an error is whether it caused substantial injustice to a party, not whether the appellate court feels strongly about the desirability of adherence to the rule. When we reverse because of a mere technicality, we cause injustice to the party that had finally obtained an end to the litigation. Where an error does not affect substantial rights, the party that obtained an end to the lawsuit is entitled to an appellate stake through the lawsuit’s heart. We can no more create “out of whole cloth” an exception to Federal Rule of Civil Procedure 61 than we could to the analogous criminal harmless error rule, Federal Rule of Criminal Procedure 52; the Supreme Court has more than once found it necessary to remind us of our lack of power regarding the latter rule. See, e.g., Johnson v. United States, 520 U.S. 461, 117 S.Ct. 1544, 1548, 137 L.Ed.2d 718 (1997).

B. Klingele error in this case was harmless.

The majority opinion says “an objective examination of the record may disclose that the pro se prisoner litigant has a complete understanding of Rule 56’s requirements gained from some other source,” but rejects the “subjective scrutiny of the prisoner’s pleadings in which the dissent engages.” The examples the majority gives of “objective” grounds for treating error as harmless are citation by the prisoner of Rand and finding a Klingele form in another of the prisoner’s lawsuit files.

The “subjective scrutiny” in which I engage is precisely the analysis that Federal Rule of Civil Procedure 61 requires. The rule and statute do not allow us to limit our consideration of harmless error to such mechanical issues as whether the prisoner received a Klingele form in another recent lawsuit. My “subjective scrutiny” consisted of reading the record.

The Rule 61 test is whether in light of the record in this case, affirmance “appears to the court inconsistent with substantial justice.” The record shows that Rand was aware of the requirements of Rule 56. He said in one of his own papers that he knew he “must respond” to the defense motion, and he did so. Rand had a fair chance to establish either a genuine issue of material fact, or that the people he sued were not entitled to judgment as a matter of law, and he did not do so.

The majority opinion reverses, despite the harmless error rule, because (1) the notice was not in plain enough English, and included citations that might confuse a layman; (2) it did not tell Rand that his case would be over if he lost on summary judgment. The majority’s first reason is stylistic and highly subjective, a suggestion' that our Appendix A is better in writing style than the California Attorney General’s statement. Rand used numerous citations himself, so we cannot reasonably infer that he would be intellectually overwhelmed by use of citations in the notice he was sent. Rand did respond, with extern sive, appropriate materials, so he plainly was not harmed by failure to tell him that if he did not respond, he would lose his case. Rand’s own motion for additional time to oppose the motion said he knew had to oppose it. We must do our duty to decide whether in the individual case we are adjudicating, the error affected substantial rights, and the answer, on this record, has to be “no.”

The Supreme Court held in United States v. Olano, 507 U.S. 725, 113 S.Ct. 1770, 1778, 123 L.Ed.2d 508 (1993), that for error to be prejudicial, “it must have affected the outcome of the District Court proceedings.” Under Rule 61 and Olano, we cannot treat Klingele error as prejudicial merely on the ground that the plaintiff was deprived of his procedural right to a Klingele warning.

The majority opinion seems to proceed on the unstated assumption that the burden is on the party seeking affirmance to demonstrate harmlessness. But that assumption is mistaken. This is a civil action by the prisoner, not a criminal case against him. The Supreme Court held in Palmer v. Hoffman, 318 U.S. 109, 116, 63 S.Ct. 477, 87 L.Ed. 645 (1943), that (in a civil case) the party seeking reversal “carries the burden of showing that prejudice resulted.” We need not decide whether this standard still applies to habeas corpus cases, see O’Neal v. McAninch, 513 U.S. 432, 115 S.Ct. 992, 130 L.Ed.2d 947 (1995), because Rand’s case is not a petition for a writ of habeas corpus. It is a civil action for damages and other relief. In a civil case, the standard for treating error as harmless is “less stringent” than in a criminal case. City of Long Beach v. Standard Oil, 46 F.3d 929, 933 (9th Cir.1995).

Of course in many kinds of errors, we cannot know for sure what would have happened, had the decision gone the other way. The clearest example is denial of a continuance. Yet we held, en banc, that a party cannot obtain reversal for failure to grant a continuance without “showing actual and substantial prejudice ... measured in terms of the outcome of the trial.” Martel v. County of Los Angeles, 56 F.3d 993, 995 (9th Cir.1995) (en banc). There is no justification for failure to apply the same rule here. Rand has not demonstrated that failure to give him a Klingele warning affected the outcome of the district court proceedings, so we have to affirm.

Conclusion.

Prisoners are not the only people who have rights that need to be protected at summary judgment. The people they sue have rights too. One of these rights is that judgment “shall be rendered forthwith” if the papers “show that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as' a matter of law.” Fed.R.Civ.P. 56(e). The.papers showed that, yet we have mistakenly held that the district judge lacked authority to render judgment “forthwith.” The cáse has been in litigation for nine years, and we have kept it alive. Reversing for lack of a boilerplate Klingele form, without regard to whether the omission mattered, construes Rule 56 in a way that undermines the “just, speedy, and inexpensive determination of every action.” Fed. R.Civ.P. 1.

Klingele unfortunately affects many people whose disputes have nothing to do with prisoners’ claims. Because prisoner petitions are about a quarter of our casé load, and are mostly frivolous and mostly pro se, if we cannot effectively filter out all the bad ones efficiently, then other people’s more substan tial cases wait too long to get to the front of the line for arguments. It is not sufficient to dispose of prisoners’ appeals rapidly by reversing on a mechanical and technical ground, because that just clogs the district court dockets. Litigants in district court are also entitled to get to the front of the line without waiting too long behind meritless prisoners’ eases.

In 1996, the last year unaffected by our decision to rehear Rand v. Rowland, 113 F.3d 1520 (9th Cir.1997), en banc, we reversed or vacated 39 summary judgments against pro se prisoners, on the sole ground that the district court had violated the Klin-gele rule. In some of these, the prisoners basically told us in their appeal papers, “I was entitled to a Klingele notice and didn’t get one, so I am entitled to reversal.” And they won. This is senseless. We should quit reversing judgments in prisoners’ lawsuits against other people for lack of a boilerplate form in the file.

Notes

1

. Although we took the entire case en banc, the only issue that concerns us is the fair notice requirement under Klingele. Rand, 113 F.3d at 1522-25. Part II.C. of the panel opinion, concerning appointment of counsel, is not affected by our en banc review and is not withdrawn.

2

. That notice stated:

NOTICE OF RULES RELATING TO SUMMARY JUDGMENT

Pursuant to Klingele v. Eikenberry, 849 F.2d 409, 411 (9th Cir.1988), the following information is provided to you:

A party is entitled to summary judgment where the documentary evidence produced by the parties permits only one conclusion. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 251-52, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986). A party seeking summary judgment bears the initial responsibility of informing the Court of the basis of its motion, and identifying those portions of "pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any,” which it believes demonstrates the "absence of a genuine issue of material fact." Celotex Corp. v. Catrett, 477 U.S. 317, 325, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986) (quoting Fed.R.Civ.P. 56(c)).

If the moving party meets its initial burden showing "the absence of a material and triable issue of fact,” "the burden then moves to the opposing party, who must present significant probative evidence tending to support its claim or defense.” Intel Corp. v. Hartford Acc. & Indent. Co., 952 F.2d 1551, 1558 (9th Cir.1991) (quoting Richards v. Neilsen Freight Lines, 810 F.2d 898, 902 (9th Cir.1987)). Plaintiff is advised that he must do more than simply show that there is some metaphysical doubt as to the material facts. Matsushita Elec. Industrial Co. v. Zenith Radio, 475 U.S. 574, 586, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986).

Rather, plaintiff must come forward with sufficient evidence demonstrating to the Court that there are genuine issues of material fact to be decided at trial. Fed.R.Civ.P. 56(e). Plaintiff may not simply rely upon the pleadings to designate specific facts establishing a genuine issue for trial. Fed.R.Civ.P. 56(e); see also Celotex Corp., 477 U.S. at 324, 106 S.Ct. 2548. The existence of a genuine issue of material fact may be demonstrated through the use of affidavits, depositions, answers to interrogatories, and admissions. Celotex Corp., 477 U.S. at 324; see also Fed.R.Civ.P. 56(c). If "the record taken as a whole could not lead a rational trier of fact to find for the nonmoving party, there is no ‘genuine issue for trial.' ” Matsushita, 475 U.S. at 587, 106 S.Ct. 1348.

Furthermore, pursuant to Local Rule 260, any party opposing a motion for summary judgment shall file a concise statement, and the source thereof in the record, of all material facts as to which títere is a genuine issue precluding summary judgment.

Additional information concerning summary judgment may be found in Local Rules 230 and 260, the cases cited above, and T.W. Electrical Serv. v. Pacific Electrical Contractors Assn., 809 F.2d 626 (9th Cir.1987).

3

. We note that the Fifth Circuit adopted a rule that "particularized additional notice of the potential consequences of a summary judgment motion and the right to submit opposing affidavits need not be afforded a pro se litigant.” Martin v. Harrison County Jail, 975 F.2d 192, 193 (5th Cir.1992) (citing Jacobsen v. Filler, 790 F.2d 1362, 1364-67 (9th Cir.1986)). However, Martin did not squarely address the issue before this court' — the propriety of the fair notice rule when restricted to pro se prisoners. Although the plaintiff in Martin was a pro se prisoner, the court specifically relied upon our decision in Jacobsen, 790 F.2d 1362, which explicitly excluded pro se prisoners from its holding. See id. at 1364 n. 4.

4

. Even trial and appellate judges have been known to misunderstand Rule 56.

A district court does not, of course, make "findings of fact” in ruling on a summary judgment motion. Findings of fact are made on the basis of evidentiary hearings and usually involve credibility determinations, which explains why they are reviewed deferentially under the clearly erroneous standard. See Fed. R. Civ. Proc. 52(a).

Epstein v. MCA, Inc., 126 F.3d 1235, 1253 n. 18 (9th Cir.1997). Yet, district courts, on occasion, have purported to make findings on summary judgment. See, e.g., Republic of Nicaragua v. Standard Fruit Co., 937 F.2d 469, 480 (9th Cir.1991) (reversing on ground that "district court relied improperly on credibility findings to support its grant of summary judgment on the parties’ intent"). On appeal, we have also mistakenly applied the "clearly erroneous” standard of review to uphold the grant of summary judgment. See, e.g., Bothke v. Fluor Eng’rs & Constructors, Inc., 834 F.2d 804, 810 (9th Cir.1987) (noting that prior panel in same case "erred in applying the ‘clearly erroneous’ test of Fed. R.Civ.P. 52(a) to the district court’s entry of summary judgment pursuant to Fed.R.Civ.P. 56”).

5

. "Should it appear from the affidavits of a party opposing the motion that the party cannot for reasons stated present by affidavit facts ’essential to justify the party's opposition, the court may refuse the application for judgment or may order a continuance to permit affidavits to be obtained or depositions to be taken or discovery to be had or may make such other order as is just.” Fed. R.Civ.P. 56(f).

6

. The dissent contends that we should reevaluate the Klingele rule in light of the recent passage of the Prison Litigation Reform Act of 1995 ("PLRA”). Pub.L. No. 104-134, Title VIII, 1996 U.S.C.C.A.N. (110 Stat.) 1321-66. However, both the PLRA's early screening procedures, 28 U.S.C. § 1915A, and its requirement that prisoner litigants must pay the filing fee even when proceeding in forma pauperis, 28 U.S.C. § 1915(c), insure that those actions which do reach the summary judgment stage are now more likely to he meritorious than was the case before the PLRA's enactment. See Crawford-El v. Britton, — U.S. -, -, 118 S.Ct. 1584, 1596, 140 L.Ed.2d 759 (1998) (summarizing provisions of PLRA and noting that the Act has had "its intended effect" of reducing "the number of prisoner civil rights suits filed in federal court”). Crawford-El itself rejected a supposedly-procedural rule of the District of Columbia Circuit which made a substantive change in the law in § 1983 cases, and which would have made it easier for prison officials to obtain summary judgment in prisoner civil rights litigation. See 118 S.Ct. at 1592, 1595 (rejecting requirement of proof of an unconstitutional motive by clear and convincing evidence).

7

. We also see no impediment to district courts’ adopting a local rule implementing the requirements of this opinion.

8

. A model notice is appended to this Opinion as Appendix "A”.

9

. Of course, although not strictly a summary judgment issue, we also recognize that the lack of notice cannot be harmful if the record discloses that the pro se prisoner cannot prove, any set of facts which would entitle him or her to relief under the familiar rule of Conley v. Gibson, 355 U.S. 41, 45-46, 78 S.Ct. 99, 2 L.Ed.2d 80 (1957). We have long held that this general rule applies equally to civil rights complaints brought by pro se prisoners. See Pena v. Gardner, 976 F.2d 469 (9th Cir.1992); Baker v. McNeil Island Corrections Ctr., 859 F.2d 124, 127 (9th Cir.1988); Maurer v. Los Angeles County Sheriff’s Dep’t, 691 F.2d 434, 435-37 (9th Cir.1982); Gillespie v. Civiletti, 629 F.2d 637, 640 (9th Cir.1980). We do not eschew that rule now.

10

. The record does not disclose that Rand was given notice or otherwise made aware of the requirements of Rule 56 of which defendants' notice did not inform him.

1

. A partial list of factors applicable to the overruling of circuit precedent by an en banc court maybe found at Aguon, 851 F.2d at 1175.

Case Details

Case Name: Lee A. Rand v. James Rowland Nadim Khoury, M.D., William Bunnell Roy Lee Johnson Leo R. Estes

Court Name: Court of Appeals for the Ninth Circuit

Date Published: Aug 27, 1998

Citations: 154 F.3d 952; 41 Fed. R. Serv. 3d 1381; 98 Cal. Daily Op. Serv. 6650; 98 Daily Journal DAR 9270; 1998 U.S. App. LEXIS 20944; 95-15428

Docket Number: 95-15428

Court Abbreviation: 9th Cir.