dissenting:
The majority finds error in two aspects of the District Court’s award of attorney fees pursuant to 42 U.S.C. § 2000e-5(k) *38(1982) and 29 U.S.C. § 216(b) (1982). First, drawing a sharp line between private and nonprofit law offices, the court concludes that, for the former, historical billing rates are virtually determinative1 of the “reasonable hourly rate” component of the lodestar calculation required by Copeland v. Marshall, 641 F.2d 880 (D.C.Cir.1980) (en banc). Second, the majority finds, despite the carefully reasoned and meticulously documented opinion of Chief Judge Aubrey E. Robinson, Jr., see Laffey v. Northwest Airlines, Inc., 572 F.Supp. 354 (D.D.C.1983), that the District Court abused its discretion in doubling the lodestar figure to account for the significant risk incurred by plaintiffs’ attorneys of not prevailing and thus not recovering a fee.
In my judgment, neither conclusion is warranted under existing law. Given the unambiguous intent of Congress, recent pronouncements by the Supreme Court, and the controlling precedent in this circuit, there is simply no room for the majority’s view, that the amount of the fee award should turn on the nature of the practice the lawyer has chosen to follow rather than the customary fee prevailing in the community for lawyers of comparable skill, reputation, and experience. Nor can I agree that the District Court erred in its calculation of a contingency multiplier. The legislative history of the relevant attorney fees acts and the en banc decision in Copeland, supra, preclude the view that increasing the award to account for the risk of non-success constitutes an error of law. See also Blum v. Stenson, — U.S.-,-, 104 S.Ct. 1541, 1550, 79 L.Ed.2d 891 (1984) (Brennan, J., concurring). (Supreme Court decisions not yet appearing in U.S. Reports will hereinafter be cited only to Supreme Court Reporter.) Thus reversal is appropriate only if we find that the District Court abused its discretion. Hensley v. Eckerhart, 461 U.S. 424, 103 S.Ct. 1933, 1941, 76 L.Ed.2d 40 (1983) (opinion by Powell, J.); Copeland, supra, 641 F.2d at 901. Yet, presumably because the thoroughness of the District Court’s opinion makes the task all but impossible, the majority makes no serious effort to justify its finding of abuse of discretion. Instead, under the guise of applying the appropriately deferential standard of review, the majority substitutes its judgment for that of the District Court. This approach not only disregards the clear mandate of the en banc decision in Copeland, but also, contrary to the Supreme Court’s express admonition to avoid engendering a “second major litigation” over attorney’s fees, Hensley, supra, 103 S.Ct. at 1941, serves to encourage disappointed fee litigants to persist in pursuing their claims in this court. Accordingly, I dissent.
I
A
Buoyed by its confidence in the inevitable beneficence and efficiency of the free market’s “invisible hand,” the majority holds that the District Court erred in awarding fees at variance with those charged other clients in private litigation. I agree that the appropriate inquiry is to determine the prevailing market rate for complex federal *39litigation. Blum v. Stenson, supra, 104 S.Ct. at 1547 n. 11. I need not, however, plumb the mysteries of economic analysis to evaluate the validity of the majority’s view that the historical billing rate of a private law firm necessarily corresponds with the market value of its services. In my judgment, the analysis adopted by the majority is expressly foreclosed by the legislative history of the attorney fees acts, Supreme Court precedent construing those acts, and an unbroken line of cases in this circuit.
The task of determining the appropriate methodology for evaluating the reasonableness of fee awards “begins and ends with an interpretation of the attorney’s fees statute.” Blum, supra, 104 S.Ct. at 1546. Congress has made clear its intention that the same standards govern under all of the myriad statutes permitting the award of attorney fees to prevailing parties.2 See, e.g., S.Rep. No. 94-1011, 94th Cong., 2d Sess. 1, 6 (1976) (hereinafter cited as Senate Report) (42 U.S.C. § 1988 adopted to achieve consistency in attorney fees provisions of civil rights laws); Hensley, supra, 103 S.Ct. at 1939 n. 7; Jordan v. U.S. Dep’t of Justice, 691 F.2d 514, 523 n. 89 (D.C.Cir.1982). Yet the legislative history of none of the attorney fees statutes offers support for the majority’s conclusion that a private attorney’s historical billing rate is presumptively congruent with the market value of his services. Indeed the legislative history clearly negates precisely the analysis the majority today adopts. In enacting the Civil Rights Attorneys’ Fees Awards Act of 1976, 42 U.S.C. § 1988 (1982), Congress indicated that “[t]he appropriate standards [for measuring the reasonableness of attorney fees] * * * are correctly applied in such cases as Stanford Daily v. Zurcher, 64 F.R.D. 680 (N.D.Cal.1974) * * Senate Report at 6. In Stanford Daily the court, in determining the appropriate fee award to a private attorney, explicitly stated, “This court does not accept the attorneys’ usual billing rates as definitively fixing their billing rates for this litigation." 64 F.R.D. at 684 (emphasis added).3
*40Nor will the legislative history support the majority’s view that Congress endorsed a different calculus for measuring the reasonableness of attorney fees depending on whether counsel is a public interest lawyer or is in private practice.4 As the Supreme Court unambiguously found in Blum, supra, “Congress did not intend the calculation of fee awards to vary depending on whether plaintiff was represented by private counsel or by a nonprofit legal services organization.” 104 S.Ct. at 1547. For all attorneys, whatever the nature of their practice, the District Court’s task is to determine whether the “requested rates are in line with those prevailing in the community for similar services by lawyers of reasonably comparable skill, experience and reputation.” Id. at 1547 n. 11. See also Davis v. County of Los Angeles, 8 E.P.D. ¶ 9444 (C.D.Cal.1974) (cited in Senate Report at 6 as applying the appropriate standard). While the rates charged in private representations “may afford relevant comparisons,” Blum, supra, 104 S.Ct. at 1547 n. 11, the billing practice of the law office is merely one component in an otherwise essentially objective calculation. In many instances rates charged private clients will serve an evidentiary function of substantiating or rebutting other evidence adduced in the fee request, and thus will be a valuable indicator of prevailing rates in the community. Blum and the legislative history, however, quite simply leave no room for the conclusion that the District Court commits an error of law by awarding attorney fees that are at variance with fees charged by the private attorney in other litigation.
I readily acknowledge, as has the Supreme Court, that the determination of the market rate is “inherently difficult.” Id. Perhaps also, though I have my doubts,5 the methodology proposed by the majority is simpler and more administrable than that envisioned by Congress. But that, of course, is irrelevant given the clarity of the congressional design. In my judgment, “[t]he Court has simply fixed upon what it believes to be good policy and then patched together a rationale as best it could.” Washington Metropolitan Area Transit Authority v. Johnson, 104 S.Ct. 2827, 2837, 81 L.Ed.2d 768 (1984) (Rehnquist, J., dissenting). Indeed, had Congress intended so straightforward an approach as equating market rates with historical billing rates, it would have said so. That it did not is itself a strong indication that a law office’s billing practice is to be but one consideration among many in the required calculation.
In an effort to align its policy judgment with the legislative history, the majority notes that Congress explicitly admonished courts to guard against fee awards that “produce windfalls to attorneys.” Senate Report at 6. To award appellees a fee in excess of that charged other clients in private practice, the majority concludes, would constitute just such a windfall. Maj. op. at 16. Whatever its appeal in the abstract, this reasoning was expressly rejected both by the Supreme Court and by this court sitting en banc. In Blum the Court brushed aside the suggestion that *41the correct fee for a Legal Aid office is the cost of rendering the service plus a reasonable profit. 104 S.Ct. at 1546 n. 6. The appropriate rate for Legal Aid attorneys, no different from those in private practice, is that “prevailing in the community for similar services by lawyers of reasonably comparable skill, experience and reputation.” Id. at 1547 n. 11. Even if application of the standard yields income in excess of that ordinarily received by the law office, this result cannot be “viewed as the kind of ‘windfall profits’ [Congress] expressly intended to prohibit.” Id. at 1547.
Copeland v. Marshall, supra, 641 F.2d 880 (en banc), a case involving a private law firm,6 makes this point with even greater clarity. The court observed that “paying low-salaried attorneys the prevailing market rate normally will yield a larger fee than that to which they are accustomed,” but found “no flaw” in this result. Id. at 899. In fact, the court concluded, computing fees “differently depending on the identity of the successful plaintiff’s attorney” would produce a windfall for the defendant. Id. In many instances the incentive of the employers not to discriminate would be diminished if fee awards are made to turn on the nature of plaintiff’s counsel’s practice. Id. Moreover, the defendant is “subject to a lesser incentive to settle a suit without litigation than would be the case if a high-priced private firm undertook plaintiff’s representation.” Id.7
Though the juxtaposition in the above quoted passage is between a “private” and a “public interest” firm, the reasoning is wholly applicable to the case of a private firm whose lower rates reflect “personal professional interests and the ability of [its] clients to pay.” Nat’l Ass’n of Concerned Veterans v. Secretary of Defense, 675 F.2d 1319, 1325 (D.C.Cir.1982) (per curiam ).8 No less than when the plaintiff is represented by a public interest firm, the defendant’s incentive to cease discriminating or settle will be proportionally less if he knows he will benefit from the happenstance that plaintiff’s counsel historically charged lower fees than those prevailing in the community.
Indeed, in case after case in this circuit the court has given short shrift to the connected arguments that a bright line separates public interest and private law firms, and that, for the latter, the market rate presumptively is tied to the historical billing rate. The majority’s adoption of *42these twin propositions in the teeth of an unbroken line of contrary precedent is, to say the least, startling. In Nat’l Treasury Employees Union v. Nixon, 521 F.2d 317, 322-323 (D.C.Cir.1975) (Wilkey, J., writing for a unanimous court), the court noted that where private counsel “serve organizations for compensation below that obtainable in the market because they believe the organizations further the public interest,” upward adjustments in the attorney’s billing rate are necessary to align the compensation with its reasonable market value.9 See also Nat’l Ass’n of Concerned Veterans v. Secretary of Defense, supra, 675 F.2d at 1326 (where counsel customarily exercises billing judgment by not billing at the market rate, this factor must be considered in calculating appropriate rate). Similarly, in Sierra Club v. Gorsuch, 684 F.2d 972, 975 (D.C.Cir.1982) (per curiam), rev’d on other grounds sub nom. Ruckelshaus v. Sierra Club, 462 U.S. 680, 103 S.Ct. 3274, 77 L.Ed.2d 938 (1983),10 this court observed that no evidence of historical billing rates to private clients had been submitted to support the fee request, but found such submissions unnecessary where the private attorney represents primarily nonprofit public interest organizations.11 It is difficult to imagine more explicit refutation of the proposition that rates charged by private attorneys presumptively coincide *43with the “reasonable hourly rate” for their services.
B
In sum, the clear intent of Congress and controlling precedent in the Supreme Court and this circuit leave little room for a free-ranging policy analysis of the “best” method for determining reasonable hourly rates under the attorney fees acts. Even if the issue were an open one — and, most emphatically, it is not — I am not persuaded by the policy arguments advanced by the majority to support tying the reasonable hourly rate to the firm’s own billing rate.
The majority’s reasoning reduces to the view that setting hourly rates according to the established rate is easier and less arbitrary to administer. The court effectively discards the methodology applied in this circuit without exception since Copeland because it considers that approach to be a “charade,” an “arbitrary divination,” and a “complex and expensive overlay of delusivk mathematical form.” Maj. op. at 19-20. We are told that a series of benefits will flow from the adoption of a more “predictable” approach, including the avoidance of a second major litigation over attorney fees and a generally more accurate measure of the market value of the legal services rendered.
In my judgment, the promise of greater predictability and ease of administration is largely illusory.12 At first glance, using a private firm’s historical billing rate as a touchstone for determining market value seems elegantly simple. On closer inspection, however, the proposed approach proves no less difficult and no more precise than the established inquiry of ascertaining rates “prevailing in the community for similar services by lawyers of comparable skill, experience and reputation.” Blum, supra, 104 S.Ct. at 1547 n. 11. Many private firms have no uniform billing rate clearly etched in stone. Concerned Veterans, supra, 675 F.2d at 1325. Moreover, as a practical matter there is often substantial play in the joints of even the most established billing practice. The rate may well depend on the nature of the service rendered, the relationship with the client, its ability to pay, or myriad other considerations. In many instances the rate charged by a firm may have changed substantially during the course of the litigation that is the basis of the fee request. The changes may reflect the rising or sinking fortunes of the firm or the effects of inflation. See, e.g., Concerned Veterans, supra, 675 F.2d at 1325 (“inflation perforce induces rapid changes in billing practices”); E. Larson, Federal Court Awards of Attorney’s Fees 191 (1981). Indeed, in this very case litigation on the merits wended its way through the courts for over 14 years, largely coinciding with a period of rampant inflation. In short, the seemingly simple variable of the private firm’s historical billing rate often will prove to be a rapidly shifting, multi-determined complex of figures. It cannot be said with certainty that focusing on historical billing rates will generate fewer discovery requests, less litigation, and greater prospects for negotiated settlement than the methodology the majority today supplants. Indeed, in Copeland, 641 F.2d at 896, the court rejected a “cost-plus” approach to setting attorney fees precisely because it would necessitate considerable litigation over the inner workings of a law firm.
Thus I remain unconvinced that the approach endorsed by the majority is more administrable or will prevent the fee application from devolving into a “second major *44litigation.” Maj. op. at 21, quoting Hensley v. Eckerhart, supra, 103 S.Ct. at 1941. Indeed, there is a certain irony in the court’s reasoning, for a virtually inevitable consequence of the majority’s effort to make new law in this case is to give disappointed fee litigants the clear message that even if they don’t prevail in District Court under the controlling legal standard, they should continue to press their claims vigorously at the appellate level.13
In my view, the better approach is also the required approach. Consistent with the standard set out in Copeland and the method of proof enunciated in Concerned Veterans, the fee applicant must provide specific evidence of the prevailing rate for comparably complex federal litigation. Relevant documentation may include “affidavits reciting the precise fees that attorneys with similar qualifications have received,” and “[r]ecent fees awarded by the courts or through settlements to attorneys of comparable reputation.” Concerned Veterans, supra, 675 F.2d at 1325.14 No doubt, rates charged in the firm’s prior representations “may afford relevant comparisons.” Blum, supra, 104 S.Ct. at 1547 n. 11 (emphasis added). Nor do I question that in some, but by no means all, instances the applicant’s customary billing practice will provide the “best evidence” of the market rate. Concerned Veterans, supra, 675 F.2d at 1325. But neither the policy underlying the attorney fees acts nor the precedent construing them support the suggestion that a private firm's billing rate is presumptively dispositive and that any in-congruence between that rate and the fee award is a reversible error of law. Because, in my judgment, the District Court meticulously applied the correct standard, I would affirm its determination of reasonable hourly rates.
II
I would also affirm the District Court’s decision to adjust the lodestar to account for the risk of not prevailing. Were it my task to canvass the evidence de novo and fix an appropriate contingency multiplier, conceivably I would disagree with the determination below. But such is not my task, nor should it be. We may reverse an adjustment in the lodestar only if it constitutes an abuse of discretion. Copeland, supra, 641 F.2d at 893. This deferential standard is “appropriate” not only “in view of the district court’s superior understanding of the litigation,” but as well because of “the desirability of avoiding frequent appellate review of what essentially are factual matters.” Hensley, supra, 103 S.Ct. at 1941. Review under a more exacting standard encourages protracted appellate litigation, increases the uncertainty and expense of bringing a civil rights suit, and, in so doing, may well discourage other victims of civil rights violations from pressing their claims. Id. at 1951 (Brennan, J., concurring in part and dissenting in part). Purporting to apply the appropriately deferential standard, the majority finds a contingency adjustment inappropriate in this case. Because, however, the District Court indisputably applied the correct legal standard, carefully evaluated the substantial *45submissions on the issue, and fully articulated the basis for its determination, quite frankly I am at a loss to understand how the majority can conclude that the court below abused its discretion.
The majority’s analysis is succinct." It reads Blum, supra, to suggest that contingency adjustments are proper only in “exceptional” cases and then finds the odds of success at the outset of this 14-year-old litigation not “unusual” enough to meet this standard. Maj. op. at 29. Distilling from Blum an “exceptional cases” test is no mean feat. Blum, as the majority recognizes elsewhere, explicitly declined to analyze the appropriateness of an upward fee adjustment to account for the risk of not prevailing. 104 S.Ct. at 1550 n. 17.15 The Court did suggest that adjustments to reflect the quality of representation would be “exceptional” because that factor ordinarily would be subsumed by the reasonable hourly rate variable. Id. at 1549. But this reasoning does not apply to the quite different question of the contingency multiplier. Indeed, consistent with Blum’s concern to avoid “double counting,” in Concerned Veterans, supra, 675 F.2d at 1328, we admonished the District Court not to include the risk premium in calculating the lodestar figure. The “exceptional cases” test fashioned by the majority thus finds no support in the prior decisions of the Supreme Court.16
After careful consideration the District Court found that the odds of success at the outset of the litigation were approximately even. 572 F.Supp. at 379. Provided only that the lodestar figure does not already subsume the risk of non-success, whether these odds may be deemed “exceptional” is irrelevant in calculating the contingency adjustment.17 Congress made a deliberate judgment18 to encourage litigants to press *46claims for violation of civil rights even where the chances of success are uncertain. The legislative decision to permit courts to make adjustments to compensate for the risk of not prevailing reflects the “experience of the marketplace that lawyers generally will not provide legal representation unless they receive a premium for taking that risk.” Berger, Court Awarded Attorney’s Fees: What Is Reasonable?, 126 U.Pa.L.Rev. 281, 325 (1977). No doubt, Congress did not intend to encourage suits even where the prospects of success are extremely remote.19 But where, as here, the probabilities approached 50 percent, adjusting the lodestar to account for the prospect of large, uncompensated outlays of time and money is entirely consistent with — indeed, mandated by — the congressional design. Blum, supra, 104 S.Ct. at 1550-1551 (Brennan, J., concurring); E. Larson, supra, at 215.
Conclusion
In Copeland this court, sitting en banc, recognized that the process of setting the lodestar figure and calculating adjustments is “inherently imprecise and that certain estimations must be made.” 641 F.2d at 893. But for the policy reasons discussed above we concluded that the District Court was in the best position to determine the appropriate figure and that we would not second guess that judgment absent an abuse of discretion. In my view, the majority has done just that. Substituting its judgment for that of the District Court, the court today disregards this circuit’s carefully crafted approach to civil rights attorney fees awards and, in so doing, frustrates the intent of Congress and injects unnecessary confusion into an area of law I had thought settled long ago.
Accordingly, I respectfully dissent.
. The majority describes the methodology it adopts as "tying the 'reasonable hourly rate’ to the firm’s own billing rates.” Majority opinion (maj. op.) at 18. At the last minute, however, the court steps back from holding that a private attorney’s billing rate conclusively determines the maximum rate allowable for a fee award. Having determined the rates charged in private representation, the District Court "may then ‘bracket’ this rate by establishing that it falls within rates charged by other firms for similar work in the same community. * * * So long as the firm's own rate falls within the rate brackets, it is the market rate.” Maj. op. at 25 (emphasis in original). As a practical matter, however, it makes little difference whether the private billing rate is conclusive dr merely is presumptively determinative of the market rate. As this case demonstrates, attorneys in a given community charge a wide range of rates. Almost invariably, the defendant will be able to find at least one private firm in the community that charges a lower rate than the fee applicant for similar work. Thus in all but the rarest cases "bracketing” the fee applicant’s billing rate is an empty gesture that will yield results no different from conclusively setting the reasonable hourly rate as equivalent to the firm’s historical billing rate.
. See E. Larson, Federal Court Awards of Attorney’s Fees 301-302 (1981) (listing 54 federal statutes authorizing award of attorney’s fees).
. The Senate Report also cited Swann v. Charlotte-Mecklenburg Board of Education, 66 F.R.D. 483 (W.D.N.C.1975), as applying the correct methodology. In Swann the court, in determining the amount of the fee award, looked to the rate charged by opposing counsel. Id. at 485. Like Stanford Daily, Swann makes clear that Congress did not intend to use historical billing rates as a cap on fee awards. E. Larson, supra note 2, at 196.
The majority’s reading of Stanford Daily to support its holding, maj. op. at 22-23, requires the use of mirrors. The Stanford Daily court — consistent with the approach adopted in Copeland v. Marshall, 641 F.2d 880 (D.C.Cir.1980) (en banc), applied by the District Court in this case and endorsed in this dissent — looked at the historical billing rates of the private attorney and concluded that they conformed to the customary rate prevailing in the community. It did not suggest that the billing rate was dispositive provided only that some attorney somewhere could be found who charged more than that fee and some attorney somewhere could be found who charged less. No such mechanical and artificial "bracketing” was applied. Indeed, the court took pains to explain that the billing rates of the private attorney were the appropriate yardstick only because they "compare favorably with the rates charged by other attorneys in this area for work involving complex questions of fact and law.” 64 F.R.D. at 685. To defend its strained reading of the case, the majority must seize on the word "definitively” in the statement in Stanford Daily that ”[t]his court does not accept the attorneys’ usual billing rates as definitively fixing their billing rates for this litigation.” Id. at 684. As argued earlier, see note 1 supra, there is no meaningful distinction in practice between “presumptively” conclusive and "definitively” conclusive.
The majority misreads this dissent to suggest that "because the legislative history does not endorse the approach we adopt today that it' therefore precludes it.” Maj. op. at 23. By incorporating by reference the methodology of the cases cited in Senate Report, Congress revealed its intent that historical billing rates not unilaterally determine reasonable hourly rates in civil rights attorney fee awards. Relying on a transparently empty distinction between "presumptively” and "definitively” conclusive, the majority today adopts a methodology directly at odds with that intent. I cannot say it more plainly.
. Applying a different methodology depending on whether the fee applicant can be characterized as a private lawyer with regular billing rates, maj. op. at 24-25, is troubling for other reasons in addition to the plain incompatibility of this approach with the congressional design. First, the District Court will often face a difficult definitional problem. The status of the private attorney acting pro bono publico, for example, is not clear. Second, the majority's methodology will often generate anomalous results. It is entirely conceivable that the rate awarded a first year lawyer at a Legal Aid office would exceed that awarded to a vastly more experienced attorney whose practice, though private, was deliberately geared towards low paying clients. Indeed, the majority's disposition seems to assure precisely this result. In Blum v. Stenson, 104 S.Ct. 1541, 1544 n. 4, 79 L.Ed.2d 891 (1984), the Supreme Court upheld a resonable hourly rate of $105 per hour to a Legal Aid Lawyer with one and a half years of litigating experience. This amount appears to exceed the average fee for Bredhoff & Kaiser partners with vastly more extensive experience. Brief for appellees at Appendix B. Congress did not intend such an incongruous result.
. See text at 36-37 infra.
. Copeland involved the fee request of a private firm acting pro bono publico. 641 F.2d at 882 n. 1.
. The majority argues that under the methodology it adopts the parties will have a greater incentive to settle because the reasonable hourlv rate will be more predictable. Maj. op. at 21-22. As discussed in text at 36-37, I have little confidence that setting fees according to historical billing rates will be any more predictable than the current approach. In any event, even more important than providing an incentive to settle on the amount of attorney fees after liability for a civil rights violation has been established is the statutory objective of encouraging potential defendants to cease violating civil rights in the first place. Copeland recognizes that the prospects of accomplishing the latter goal decrease significantly if the potential defendant assumes that the costs of continuing its course of action and defending the civil rights suit will be less simply by virtue of the nature of the practice plaintiff's private counsel has chosen to pursue.
. Repeatedly, this court has recognized that some attorneys bill at rates below those obtainable in the market because their clients serve the public interest. Nat'l Treasury Employees Union v. Nixon, 521 F.2d 317, 322 (D.C.Cir.1975) (Wilkey J., writing for a unanimous court); Sierra Club v. Gorsuch, 684 F.2d 972, 975 (D.C.Cir.1982), rev’d on other grounds sub nom. Ruckelshaus v. Sierra Club, 462 U.S. 680, 103 S.Ct. 3274, 77 L.Ed.2d 938 (1983). Nonetheless, the majority concludes that weighing this consideration in the fee calculation is inappropriate because it "invites courts to reward those plaintiffs who represent causes the judge involved finds estimable.” Maj. op. at 14 n. 69 (emphasis deleted). But this argument assumes its own conclusion. The trial court’s task is not to weigh which causes it finds deserving but objectively to determine the rate prevailing in the community for similar services. If the court finds a differential between the firm’s billing rate and the prevailing rate, it may reflect the firm’s choice to bill its public interest clients below market rates. But at no point in the calculation is the court asked to pass on whether it approves or disapproves of the causes the firm has chosen to represent.
. The organization referred to in National Treasury Employees, as in the instant case, was a union.
. The majority labors mightily to distinguish away the line of precedent in this circuit that unambiguously conflicts with its approach. Maj. op. at 14-15 n. 69. The decisions, however, simply will not support the majority’s strained reading.
In Sierra Club v. Gorsuch, supra note 8, 684 F.2d at 975, the court did not even require a private lawyer to submit records of his historical billing practice. The majority dismisses this clear refutation of its methodology on the ground that the attorney had no "established billing practice.” Maj. op. at 14 n. 69. One searches the opinion in vain for documentation of this point. In fact, the attorney had an extensive private practice "representing primarily environmental .orgnizations and Indian tribes before state and federal courts.” 684 F.2d at 975. Nothing in the opinion suggests the absence of an established billing practice. Given the holding in Gorsuch that submissions of billing rates of private attorneys are not required, I simply do not understand how the majority can hold as a matter of law that historical billing rates are presumptively conclusive of the reasonable hourly rate.
The majority’s treatment of Nat’l Treasury Employees Union v. Nixon, supra note 8, is equally unconvincing. That the award in National Treasury Employees was not pursuant to a statute is entirely irrelevant. It is an undisputed truism to suggest that the legislative history of the statutes relevant to this case reveals Congress' intent to avoid "windfalls” for attorneys. In its numerous invocations of that phrase, however, the majority entirely fails to mention the Supreme Court’s crystal-clear holding that paying attorneys more than their usual rate is not the kind of "windfall" Congress intended to preclude. Blum v. Stenson, supra note 4, 104 S.Ct. at 1547.
The majority’s reading of Nat’l Ass’n of Concerned Veterans v. Secretary of Defense, 675 F.2d 1319 (D.C.Cir.1982) (per curiam), to support its holding requires a remarkable feat of interpretive sleight-of-hand. In Concerned Veterans the court listed numerous evidentiary sources the District Court may look to in establishing the prevailing community rate, including rates charged by other lawyers in comparable cases and recent fee awards by courts. Historical billing rates "may provide important substantiating evidence." Id. at 1326 (emphasis added). But the case simply does not suggest that the market rate is presumptively congruent with the attorney’s billing practice. Maj. op. at 17 n. 86.
Nor can the majority’s holding coexist with the en banc decision in Copeland v. Marshall. As in the recent decision in Murray v. Weinberger, 741 F.2d 1423 (D.C.Cir.1984), the fee in Copeland did correspond with the fee generally charged by the firm. But as Copeland makes explicit, 641 F.2d at 902, the District Court must look at evidence of fees charged by other lawyers engaged in comparable litigation in addition to the firm's historical rates. The court in Copeland did not undertake a mechanical "bracketing" of the firm’s rate. It merely used that rate as one factor among many in determining the hourly rate prevailing in the community. Indeed the entire thrust of the analysis is that the rates charged by attorneys are not presumptively dispositive.
. See also Donnell v. United States, 682 F.2d 240, 251-252 (D.C.Cir.1982) (Wilkey, J., writing for a unanimous court) ("some attorneys may receive fees based on rates higher than they normally command if those higher rates are the norm for the jurisdiction in which the suit was litigated”), cert. denied, 459 U.S. 1204, 103 S.Ct. 1190, 75 L.Ed.2d 436 (1983).
. The majority suggests that rather than " 'nickel and diming’ the guidelines" it offers, I should “flesh out" the approach I endorse. Maj. op. at 20. To do so would be entirely superfluous. The approach I defend is that which was announced by the court en banc in Copeland and, since that decision, has been successfully applied in dozens of cases, including this one. It is the majority that effectively abandons the settled law of this circuit, and the burden is thus on it to advance a policy argument so weighty that it justifies creating an internal conflict in the law of this circuit. In my judgment, that burden has not been carried. Rather than feeling “incompetent,” maj. op. at 23, to engage in a free-ranging policy analysis, I simply do not feel I have license to do so in this case.
. The Supreme Court has made clear that it disfavors any approach to fee calculations that encourages litigants to pursue appellate review of the District Court’s award. Hensley v. Eckerhart, 461 U.S. 424, 103 S.Ct. 1933, 1941, 76 L.Ed.2d 40 (1983).
In systemic terms, attorney’s fee appeals take up lawyers’ and judges’ time that could more profitably be devoted to other cases, including the substantive civil rights claims that § 1988 was meant to facilitate. Regular appellate scrutiny of issues like those in this case also generates a steady stream of opinions, each requiring yet another to harmonize it with the one before or the one after. Ultimately § 1988’s straightforward command is replaced by a vast body of artificial, judge-made doctrine, with its own arcane procedures, which like a Frankenstein’s monster meanders its well-intentioned way through the legal landscape leaving waste and confusion * * * in its wake. * * *
Id. at 1951 (Brennan, J., concurring in part and dissenting in part).
. The court may consider the fee charged by the losing party's counsel. See Swann v. Charlotte-Mecklenburg Board of Education, supra note 3, cited in Senate Report and discussed in note 3 supra.
. The Court observed that the fee applicants had not asserted and had made no effort to document a claim that the riskiness of the suit warranted an upward adjustment in the fee. Accordingly, it found that the District Court had erred in relying on the prospects for success in the litigation in making its calculation. Blum v. Stenson, supra note 4, 104 S.Ct. at 1550.
. While this case was pending another panel of this court did suggest that the contingency multiplier should be available only in "exceptional circumstances within the meaning of Blum v. Stenson." Murray v. Weinberger, supra note 10, 741 F.2d at 1431 (opinion by Wilkey, J.). Murray's reliance on Blum mirrors that of the majority and, in my view, is equally incorrect. In any event, the holding of Murray is irrelevant to this case. Determining whether a case is “exceptional within the meaning of Blum v. Stenson" is, of course, difficult when the Blum court expressly eschewed discussion of the contingency factor. The only actual use of the word "exceptional” in the opinion refers to the extent of the plaintiff’s success and provides no guidance at all in measuring the appropriateness of a contingency adjustment. Blum v. Stenson, supra note 4, 104 S.Ct. at 1548-1549. "[Ejxceptional within the meaning of Blum” can only refer to the Court’s repeated admonition that in weighing adjustments to the lodestar the District Court not engage in “double counting” by compensating the attorney for factors already included in the lodestar figure. Three times the Blum Court returned to the theme: “only in the rare case” will the "quality of service rendered”
not already be built into the reasonable hourly rate; "the novelty and complexity of the issues presumably were fully reflected in the number of billable hours recorded by counsel * * * [and the] reasonable hourly rate”; " ‘results obtained’ generally will be subsumed within other factors used to calculate a reasonable fee.” Id. Although this interpretation effectively collapses the first and third factors listed in Murray, it is the only coherent way to extrapolate from Blum’s discussion of other lodestar adjustments an "exceptional circumstances” test applicable to the contingency factor.
In this case the District Court scrupulously and explicitly followed our suggestion in Nat'l Ass’n of Concerned Veterans v. Secretary of Defense, supra note 10, 675 F.2d at 1328, that the risk of non-success not be incorporated in the lodestar figure. Laffey v. Northwest Airlines, Inc., 572 F.Supp. 354, 379 n. 50 (D.D.C.1983). The majority does not suggest otherwise. Therefore, the concern of the Blum Court and, derivatively, the Murray court embodied in the “exceptional circumstances” test has no bearing on this case.
. The majority reasons that a 50% chance of success is the "norm for litigated cases — in the average case, one side will win, one will lose." Maj. op. at 29 (emphasis deleted). This is analogous to suggesting that on any given day the chances of rain are even — either it will rain or it won’t.
. At least two of the cases cited in Senate Report as applying the appropriate standard ad*46justed the award to account for the risks incurred by the attorney. See Stanford Daily v. Zurcher, 64 F.R.D. 680, 682 (N.D.Cal.1974); Davis v. County of Los Angeles, 8 E.P.D. ¶ 9444 (C.D.Cal.1974). See generally E. Larson, supra note 2, at 215 (proper use of the contingency factor fulfills the congressional design).
. The majority draws a distinction between a "contingency multiplier” and a “contingency enhancement.” I agree that the District Court’s task goes beyond merely ascertaining the numerical probability of success at the outset of the litigation and multiplying the lodestar by the reciprocal of that figure. Congress did not make clear how much encouragement it wished to give suits of "varying degrees of promise," Leubsdorf, The Contingency Factor in Attorney Fee Awards, 90 Yale L.J. 473, 499 (1981), and I am reluctant to attribute to Congress the intent to encourage suits irrespective of the remoteness of the chance of success. Id. at 481. But the fact remains that Congress believed that effectuation of the civil rights laws required encouraging suits even where the prospects of success were uncertain. See note 15 supra. Thus it falls to the District Court to determine whether, measured from the outset of the litigation, the probabilities of success were of sufficient magnitude that providing the attorney with some incentive to take the case is consistent with the congressional design. Once having made this determination, the District Court must exercise its judgment as to how much monetary incentive the rational attorney would require to compensate for the risk of large, uncompensated outlays of time and money.