Pennsylvania v. Delaware Valley Citizens' Council for Clean Air

483 U.S. 711 (1987)

PENNSYLVANIA ET AL.
v.
DELAWARE VALLEY CITIZENS' COUNCIL FOR CLEAN AIR ET AL.

No. 85-5.

Supreme Court of United States.

Argued March 3, 1986 Reargued October 15, 1986 Decided June 26, 1987 CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT

Jay C. Waldman reargued the cause for petitioners. With him on the briefs on reargument were Henry G. Barr, John P. Krill, and John M. Hrubovcak. With him on the briefs on the original argument were Spencer A. Manthorpe and Messrs. Barr, Hrubovcak, and Krill.

Donald B. Ayer reargued the cause for the United States as respondent under this Court's Rule 19.6 in support of petitioners. Kathryn A. Oberly argued the cause for the United States on the original argument. With her on the brief were Solicitor General Fried, F. Henry Habicht II, and Deputy Solicitor General Geller.

James D. Crawford reargued the cause for respondents. With him on the brief was Joyce S. Meyers.^[*]

Briefs of amici curiae were filed for the State of Arizona et al. by Francis X. Bellotti, Attorney General of Massachusetts, and Suzanne E. Durrell, Assistant Attorney General, Robert K. Corbin, Attorney General of Arizona, Joseph I. Lieberman, Attorney General of Connecticut, Michael J. Bowers, Attorney General of Georgia, Richard G. Opper, Attorney General of Guam, Corinne K. A. Watanabe, Attorney General of Hawaii, Jim Jones, Attorney General of Idaho, Neil F. Hartigan, Attorney General of Illinois, Linley E. Pearson, Attorney General of Indiana, David L. Armstrong, Attorney General of Kentucky, William J. Guste, Jr., Attorney General of Louisiana, James E. Tierney, Attorney General of Maine, Stephen H. Sachs, Attorney General of Maryland, Frank J. Kelley, Attorney General of Michigan, and Louis J. Caruso, Solicitor General, Edward Lloyd Pittman, Attorney General of Mississippi, William L. Webster, Attorney General of Missouri, Stephen E. Merrill, Attorney General of New Hampshire, Lacy H. Thornburg, Attorney General of North Carolina, Nicholas J. Spaeth, Attorney General of North Dakota, Anthony J. Celebrezze, Jr., Attorney General of Ohio, Michael C. Turpen, Attorney General of Oklahoma, T. Travis Medlock, Attorney General of South Carolina, Jeffrey L. Amestoy, Attorney General of Vermont, William Broaddus, Attorney General of Virginia, Kenneth O. Eikenberry, Attorney General of Washington, Charles G. Brown, Attorney General of West Virginia, and A. G. McClintock, Attorney General of Wyoming; and for Twelve Small Private Civil Rights Law Firms by John Leubsdorf.

JUSTICE WHITE announced the judgment of the Court and delivered an opinion, Parts I, II, and III-A of which represent the views of the Court, and Parts III-B, IV, and V of which are joined by THE CHIEF JUSTICE, JUSTICE POWELL, and JUSTICE SCALIA.

This case involves the award of an attorney's fee to the prevailing party pursuant to § 304(d) of the Clean Air Act, 42 U.S. C. § 7604(d).^[1]

I

We set forth a detailed statement of the facts underlying this litigation in Pennsylvania v. Delaware Valley Citizens' Council for Clean Air, 478 U.S. 546 (1986), and recite only an abbreviated version of those facts here. In 1977, the Delaware Valley Citizens' Council for Clean Air (hereinafter respondent) and the United States each filed suit to compel the Commonwealth of Pennsylvania to comply with certain provisions of the Clean Air Act. The parties entered into a consent decree, approved by the District Court in 1978,^[2] which obligated the Commonwealth to establish a program for the inspection and maintenance of vehicle emissions systems in 10 counties in the Philadelphia and Pittsburgh areas by August 1, 1980. The Commonwealth failed to implement the program by this date, and protracted litigation ensued. Ultimately, in May 1983, the parties agreed to set June 1, 1984, as the date on which the Commonwealth would commence the inspection and maintenance program. Shortly after this agreement, respondent petitioned the District Court for attorney's fees and costs for the work performed after the issuance of the consent decree. In determining the amount of fees to be awarded, the District Court divided the work performed by respondent's counsel into nine phases. See 478 U.S., at 549-553. After computing the lodestar for each phase, the District Court adjusted this figure upward in phases four, five, and seven by doubling the lodestar to reflect the risk presumably faced by respondent that it would not prevail on these phases of the litigation. The District Court observed:

"The contingent nature of plaintiff's success has been apparent throughout this litigation. Plaintiffs entered the litigation against the U. S. Government and the Commonwealth of Pennsylvania. The case involved new and novel issues, the resolution of which had little or no precedent. . . . [P]laintiffs have had to defend their rights under the consent decree due to numerous attempts by defendants and others to overturn or circumvent this court's Orders." 581 F. Supp. 1412, 1431 (1984).

The Court of Appeals for the Third Circuit affirmed the District Court's enhancement of the fee award for contingency of success, 762 F.2d 272, 282 (1985), a judgment that we now reverse.^[3]

II

We first focus on the nature of the issue before us. Under the typical fee-shifting statute, attorney's fees are awarded to a prevailing party and only to the extent that party prevails. See, e. g., Maher v. Gagne, 448 U.S. 122, 129-130 (1980); Hensley v. Eckerhart, 461 U.S. 424, 435 (1983). Hence, if the case is lost, the loser is awarded no fee; and unless its attorney has an agreement with the client that the attorney will be paid, win or lose, the attorney will not be paid at all. In such cases, the attorney assumes a risk of nonpayment when he takes the case. The issue before us is whether, when a plaintiff prevails, its attorney should or may be awarded separate compensation for assuming the risk of not being paid. That risk is measured by the risk of losing rather than winning and depends on how unsettled the applicable law is with respect to the issues posed by the case and by how likely it is that the facts could be decided against the complainant. Looked at in this way, there are various factors that have little or no bearing on the question before us.

First is the matter of delay. When plaintiffs' entitlement to attorney's fees depends on success, their lawyers are not paid until a favorable decision finally eventuates, which may be years later, as in this case. Meanwhile, their expenses of doing business continue and must be met. In setting fees for prevailing counsel, the courts have regularly recognized the delay factor, either by basing the award on current rates or by adjusting the fee based on historical rates to reflect its present value. See, e. g., Sierra Club v. EPA, 248 U. S. App. D. C. 107, 120-121, 769 F.2d 796, 809-810 (1985); Louisville Black Police Officers Organization, Inc. v. Louisville, 700 F.2d 268, 276, 281 (CA6 1983). Although delay and the risk of nonpayment are often mentioned in the same breath, adjusting for the former is a distinct issue that is not involved in this case. We do not suggest, however, that adjustments for delay are inconsistent with the typical fee-shifting statute.

Second, that a case involves an issue of public importance, that the plaintiff's position is unpopular in the community, or that defendant is difficult or obstreperous does not enter into assessing the risk of loss or determining whether that risk should be compensated. Neither does the chance that the court will find unnecessary and not compensate some of the time and effort spent on prosecuting the case.

Third, when the plaintiff has agreed to pay its attorney, win or lose, the attorney has not assumed the risk of nonpayment and there is no occasion to adjust the lodestar fee because the case was a risky one. See, e. g., Jones v. Central Soya Co., 748 F.2d 586, 593 (CA11 1984), where the court said that "[a] lawyer may not preserve a right of recourse against his client for fees and still expect to be compensated as if he had sacrificed completely his right to payment in the event of an unsuccessful outcome."

III

A

Although the issue of compensating for assuming the risk of nonpayment was left open in Blum v. Stenson, 465 U.S. 886 (1984), JUSTICE BRENNAN wrote that "the risk of not prevailing, and therefore the risk of not recovering any attorney's fees is a proper basis on which a district court may award an upward adjustment to an otherwise compensatory fee." Id., at 902 (concurring). Most Courts of Appeals are of a similar view and have allowed upward adjustment of fee awards because of the risk of loss factor.^[4] The First Circuit, in Wildman v. Lerner Stores Corp., 771 F.2d 605 (1985), for example, takes this approach and allows an upward adjustment to the lodestar to account for the contingency factor. In that case, the District Court entered judgment on a jury verdict finding an employer liable for violating the Age Discrimination in Employment Act, 29 U.S. C. § 621 et seq., and two Puerto Rican statutes. The court awarded the prevailing party a lodestar fee amount of $56,500 and then increased that figure by 50% to account for the fact that because of the difficulties of the action and the novelty of the issue, "the plaintiffs' attorneys . . . faced a contingency of losing all their time and effort." 771 F.2d, at 610. In sustaining the enhancement of fee awards based on contingency, the Court of Appeals relied on the legislative history of 42 U.S. C. § 1988, detailed several additional reasons as to why it is necessary to increase the lodestar figure for contingent-fee cases, and concluded that rather than compensating lawyers for unsuccessful claims, an adjustment of the lodestar figure may be necessary in particular cases to provide for the reasonable attorney's fee envisioned by Congress.^[5]

This construction of the fee-shifting statutes has not been universal. The District of Columbia Circuit is particularly skeptical of the purpose served by enhancing the lodestar amount to account for the risk of not prevailing. In Laffey v. Northwest Airlines, Inc., 241 U. S. App. D. C. 11, 746 F.2d 4 (1984), cert. denied, 472 U.S. 1021 (1985), the court reversed the trial court's decision to double the lodestar based on the risk factor, citing a wide variety of problems with such an approach. The court found that, in theory, there should be no limit on the size of the fee if risk enhancement is permitted, for the less likely the chances of success in a particular case, the more "entitled" the prevailing party should be to have the fee award reflect acceptance of this risk. In a similar vein, the contingency factor penalizes the losing parties with the strongest and most reasonable defenses, thus "creating a perverse penalty for those least culpable." 241 U. S. App. D. C., at 33, 746 F.2d, at 26. Moreover, even if the risk of loss should be taken into account, "the chances of winning could not be set with anything approaching mathematical precision, and so vast increases in attorneys [fees] would derive from a spurious mathematical base." Id., at 33-34, 746 F.2d, at 26-27 (footnote omitted).

On a more fundamental level, the court found that using the risk of loss to increase the lodestar figure compensates attorneys not only for their successful efforts in one case, but for their unsuccessful claims asserted in related cases. This not only "encourag[es] marginal litigation," but raises "the reasonable question of `why the subsidy [for unsuccessful litigation] should come from the defendant in another case.' " Id., at 34, n. 138, 746 F.2d, at 27, n. 138 (citations omitted).

Such a scheme was deemed to be manifestly inconsistent with Congress' intent to award attorney's fees only to prevailing parties. Relying on this Court's holding in Hensley that attorney's fees could not be awarded for claims unrelated to those on which the party ultimately prevailed, the court reasoned:

"The same logic which restricts compensation to those portions of a lawsuit directly related to the relief procured also forbids multiplying attorneys fees so as effectively to compensate counsel for other, losing claims which may be brought. The prevailing party may expect full compensation for prevailing claims; there is no provision for compensating losing, unrelated claims in the same case, or other losing cases which might or might not involve the same parties. Any crude multiplier derived simply from the plaintiff's chance of success must be rejected as contrary to the congressional scheme." Id., at 34-35, 746 F.2d, at 27-28.

Finally, the court held that even if a contingency enhancement, as opposed to a contingency multiplier, could be used to reflect the party's initial chance of success, Blum made clear that such enhancements were proper only in the most exceptional of cases, and because "this case did not present an exceptional level of risk, no risk enhancement should be awarded." Id., at 36, 746 F.2d, at 29.^[6]

The bar and legal commentators have been much interested in the issue.^[7] Some writers unqualifiedly have endorsed the concept of increasing the fee award to insure that lawyers will be adequately compensated for taking the risk of not prevailing. "The experience of the marketplace indicates that lawyers generally will not provide legal representation on a contingent basis unless they receive a premium for taking that risk." Berger, Court Awarded Attorneys' Fees: What is "Reasonable"?, 126 U. Pa. L. Rev. 281, 324-325 (1977).^[8] See also, Developments in the Law — Class Actions, 89 Harv. L. Rev. 1318, 1615 (1976); Comment, 122 U. Pa. L. Rev. 636, 708-711 (1974).

Others have been considerably more reserved in their endorsement of a contingency bonus, focusing on four major problems with the use of this factor. First, evaluation of the risk of loss creates a potential conflict of interest between an attorney and his client, for in order to increase a fee award, a plaintiff's lawyer must expose all of the weaknesses and inconsistencies in his client's case, and a defendant's attorneys must either concede the strength of the plaintiff's case in order to keep down the fee award, or "allo[w] the fee to be boosted by the contingency bonus [by] insisting that the plaintiff's victory was freakish." Leubsdorf, The Contingency Factor in Attorney Fee Awards, 90 Yale L. J. 473, 483 (1981) (Leubsdorf). Second, in order to determine the proper size of the contingency bonus, a court must retroactively estimate the prevailing party's chances for success from the perspective of the attorney when he first considered filing the suit. Not only is this mathematically difficult to compute, but "once the result is known, it is hard for judges and lawyers to regain a perspective of ignorance and to treat the result as only one of several that were initially possible." Id., at 486.

The third problem with increasing the fee award to account for the risk of not prevailing is the same one identified by the courts which have questioned this practice: it penalizes the defendant with the strongest defense, and forces him to subsidize the plaintiff's attorney for bringing other unsuccessful actions against other defendants. Id., at 488-491. See Note, 80 Colum. L. Rev. 346, 375 (1980). Finally, because the contingency bonus cannot be determined with either certainty or accuracy, it "cannot be justified on the ground that it provides an appropriate incentive for litigation." Leubsdorf 496. Cf. Note, 96 Harv. L. Rev. 677, 686, n. 51 (1983); Comment, 53 U. Chi. L. Rev. 1074 (1986).

There are other considerations. Fee-shifting removes the interest a paying client would have in ensuring that the lawyer is serving the client economically; the task of monitoring the attorney is shifted to the judge in separate litigation over fees if the plaintiff wins. Fee litigation occurs on a case-to-case basis and is often protracted, complicated, and exhausting. There is little doubt that it should be simplified to the maximum extent possible. If the decided cases are any measure, assessing the initial risk of loss when the case is over is a particularly uncertain matter, especially for a judge who is confident that he has correctly decided for the plaintiff, but then must inquire how weak the plaintiff's case was and how likely it was that he, the judge, would have been mistaken. It may be absurd to ask the judge to "determine the probability that he would have decided the case incorrectly." Id., at 1094.

B

The disagreement among the Circuits and commentators indicates that Congress has not clearly directed or authorized multipliers or enhancements for assuming the risk of loss. Neither the Clean Air Act nor § 1988 expressly provides for using the risk of loss as an independent basis for increasing an otherwise reasonable fee, and it is doubtful that the legislative history supports the use of this factor. In concluding that risk-enhancement is authorized, JUSTICE BRENNAN in Blum, 465 U. S., at 902, relied on the fact that one of the items to be relied on in setting a fee and enumerated in Johnson v. Georgia Highway Express, Inc., 488 F.2d 714 (CA5 1974), is whether the fee is fixed or contingent, and that Congress endorsed consideration of this factor. See S. Rep. No. 94-1011, p. 6 (1976) (S. Rep). But a careful reading of Johnson shows that the contingency factor was meant to focus judicial scrutiny solely on the existence of any contract for attorney's fees which may have been executed between the party and his attorney. "The fee quoted to the client or the percentage of the recovery agreed to is helpful in demonstrating the attorney's fee expectations when he accepted the case." 488 F.2d, at 718. See Leubsdorf 479, n. 38. At most, therefore, Johnson suggests that the nature of the fee contract between the client and his attorney should be taken into account when determining the reasonableness of a fee award, but there is nothing in Johnson to show that this factor was meant to reflect the contingent nature of prevailing in the lawsuit as a whole.

JUSTICE BRENNAN also noted that Congress cited Stanford Daily v. Zurcher, 64 F. R. D. 680 (ND Cal. 1974) (subsequently aff'd, 550 F.2d 464 (CA9 1977), rev'd on other grounds, 436 U.S. 547 (1978)), as one of several cases which "correctly applied" the Johnson factors. Blum, supra, at 903. The court there increased the lodestar based, in part, on contingency-of-success considerations. But Congress also cited two other cases which it found also "correctly applied" the Johnson criteria. In Davis v. County of Los Angeles, 8 EPD ¶ 9444, p. 5047 (CD Cal. 1974), the District Court added a "Result Charge" to the basic fee award. This award was not intended to compensate the lawyers for assuming the risk of not prevailing on the merits; instead, as the label suggests, the court increased the award because "counsel [had] achieved excellent results," and "[t]he nature of the case made it difficult to litigate. . . ." Id., at 5048. The court in Swann v. Charlotte-Mecklenburg Bd. of Ed., 66 F. R. D. 483 (WDNC 1975), the third illustrative case cited with approval by Congress, did not increase the basic fee award at all. Instead, after reviewing nine factors similar to those listed in Johnson, the court reduced the prevailing party's fee request by nearly 15%, choosing to "err on the conservative side in dealing with any fee question" rather than "contribute unnecessarily to the overpricing of litigation in this or any other court." 66 F. R. D., at 486. Given the divergence in both analysis and result between these three cases, the legislative history is, at best, inconclusive in determining whether Congress endorsed the concept of increasing the lodestar amount to reflect the risk of not prevailing on the merits.

We must nevertheless come to a decision and have concluded that the judgment must be reversed.

IV

We are impressed with the view of the Court of Appeals for the District of Columbia Circuit that enhancing fees for risk of loss forces losing defendants to compensate plaintiff's lawyers for not prevailing against defendants in other cases. This result is not consistent with Congress' decision to adopt the rule that only prevailing parties are entitled to fees. If risk multipliers or enhancement are viewed as no more than compensating attorneys for their willingness to take the risk of loss and of nonpayment, we are nevertheless not at all sure that Congress intended that fees be denied when a plaintiff loses, but authorized payment for assuming the risk of an uncompensated loss. Such enhancement also penalizes the defendants who have the strongest case; and in theory, at least, would authorize the highest fees in cases least likely to be won and hence encourage the bringing of more risky cases, especially by lawyers whose time is not fully occupied with other work. Because it is difficult ever to be completely sure that a case will be won, enhancing fees for the assumption of the risk of nonpayment would justify some degree of enhancement in almost every case.

Weighing all of these considerations, we are unconvinced that Congress intended the risk of losing a lawsuit to be an independent basis for increasing the amount of any otherwise reasonable fee for the time and effort expended in prevailing. As the Senate Report observed: "In computing the fee, counsel for prevailing parties should be paid, as is traditional with attorneys compensated by a fee-paying client, `for all time reasonably expended on a matter.' Davis, supra; Stanford Daily, supra, at 684." S. Rep. 6.

The contrary argument is that without the promise of multipliers or enhancement for risk-taking, attorneys will not take cases for clients who cannot pay, and the fee-shifting statutes will therefore not serve their purpose. We agree that a fundamental aim of such statutes is to make it possible for those who cannot pay a lawyer for his time and effort to obtain competent counsel, this by providing lawyers with reasonable fees to be paid by the losing defendants. But it does not follow that fee enhancement for risk is necessary or allowable. Surely that is not the case where plaintiffs can afford to pay and have agreed to pay, win or lose. The same is true where any plaintiff, impecunious or otherwise, has a damages case that competent lawyers would take in the absence of fee-shifting statutes. Nor is it true in those cases where plaintiffs secure help from organizations whose very purpose is to provide legal help through salaried counsel to those who themselves cannot afford to pay a lawyer. It is also unlikely to be true in any market where there are competent lawyers whose time is not fully occupied by other matters.

The issue thus involves damages cases that lawyers would not take, not because they are too risky (the fee-shifting statutes should not encourage such suits to be brought), but because the damages likely to be recovered are not sufficient to provide adequate compensation to counsel, as well as those frequent cases in which the goal is to secure injunctive relief to the exclusion of any claim for damages. In both situations, the fee-shifting statutes guarantee reasonable payment for the time and effort expended if the case is won. Respondent's position is that without the prospect of being awarded fees exceeding such reasonable payment, plaintiffs with such cases will be unable to secure the help that the statutes aimed to provide.

We are not persuaded that this will be the case. Indeed, it may well be that using a contingency enhancement is superfluous and unnecessary under the lodestar approach to setting a fee. The reasons a particular lawsuit are considered to be "risky" for an attorney are because of the novelty and difficulty of the issues presented, and because of the potential for protracted litigation. Moreover, when an attorney ultimately prevails in such a lawsuit, this success will be primarily attributable to his legal skills and experience, and to the hours of hard work he devoted to the case. These factors, however, are considered by the court in determining the reasonable number of hours expended and the reasonable hourly rate for the lodestar, and any further increase in this sum based on the risk of not prevailing would result not in a "reasonable" attorney's fee, but in a windfall for an attorney who prevailed in a difficult case.^[9]

It may be that without the promise of risk enhancement some lawyers will decline to take cases; but we doubt that the bar in general will so often be unable to respond that the goal of the fee-shifting statutes will not be achieved. In any event, risk enhancement involves difficulties in administration and possible inequities to those who must pay attorney's fees; and in the absence of further legislative guidance, we conclude that multipliers or other enhancement of a reasonable lodestar fee to compensate for assuming the risk of loss is impermissible under the usual fee-shifting statutes.

Even if § 304(d) and other typical fee-shifting statutes are construed to permit supplementing the lodestar in appropriate cases by paying counsel for assuming the risk of nonpayment, for the reasons set out below, it was error to do so in this case.

V

Section 304(d), like § 1988, does not indicate that adjustment for risk should be the rule rather than the exception; neither does it require such an adjustment in any case. At most, it leaves the matter of risk enhancement to the informed discretion of the courts. There are, however, severe difficulties and possible inequities involved in making upward adjustments for assuming the risk of nonpayment, and we deem it appropriate, in order to guide the exercise of the trial courts' discretion in awarding fees, to adopt here the approach followed in Blum in dealing with other multipliers. As in that case, payment for the time and effort involved — the lodestar — is presumed to be the reasonable fee authorized by the statute, and enhancement for the risk of nonpayment should be reserved for exceptional cases where the need and justification for such enhancement are readily apparent and are supported by evidence in the record and specific findings by the courts.^[10]Blum, 465 U. S., at 898-901. For several reasons, the circumstances of this case do not justify the risk multiplier employed by the District Court.

First, the District Court doubled the lodestar in three phases of the case in recognition of the risk of loss, saying that the "contingent nature of plaintiffs' success has been apparent" from the outset, that plaintiffs entered the litigation against the United States and the Commonwealth of Pennsylvania, and that the case involved new and novel issues, the resolution of which had little or no precedent. Furthermore, they had to "defend their rights under the consent decree due to numerous attempts by defendants and others to overturn or circumvent this court's orders." 581 F. Supp., at 1431. This case, however, concerns only the reasonable fee for work done after the consent decree was entered, and fees have already been awarded for work done before that time. The risk of nonpayment should be determined at the beginning of the litigation. Lewis v. Coughlin, 801 F.2d 570, 576 (CA2 1986); Ramos v. Lamm, 713 F.2d 546, 558 (CA10 1983).^[11] Whatever counsel thought the risk of losing was at the outset, it is doubtful that counsel anticipated a similar risk in enforcing a decree if plaintiff was successful in having one entered. In any event, the District Court did not specifically identify any new and novel issues, and we fail to discern any, that emerged in the long process of enforcing the court decree in accordance with its terms. And whether the Commonwealth of Pennsylvania was a substantial opponent or whether it tried to circumvent the decree has little or nothing to do with whether the there was a real risk of not persuading the District Court to enforce its own decree. The matter may have been difficult, wearing, and time consuming, but that kind of effort has been recognized in the lodestar award.

Second, if it be assumed that this is one of the exceptional cases in which enhancement for assuming the risk of nonpayment is justified, we conclude that doubling the lodestar for certain phases of the work was excessive. We have alluded to the uncertainties involved in determining the risk of not prevailing and the burdensome nature of fee litigation. We deem it desirable and an appropriate application of the statute to hold that if the trial court specifically finds that there was a real risk-of-not-prevailing issue in the case, an upward adjustment of the lodestar may be made, but, as a general rule, in an amount no more than one-third of the lodestar. Any additional adjustment would require the most exacting justification. This limitation will at once protect against windfalls for attorneys and act as some deterrence against bringing suits in which the the attorney believes there is less than a 50-50 chance of prevailing. Riskier suits may be brought, and if won, a reasonable lodestar may be awarded, but risk enhancement will be limited to one-third of the lodestar, if awarded at all. Here, even assuming an adjustment for risk was justified, the multiplier employed was excessive.

Third, whatever the risk of winning or losing in a specific case might be, a fee award should be informed by the statutory purpose of making it possible for poor clients with good *731A claims to secure competent help. Before adjusting for risk assumption, there should be evidence in the record, and the trial court should so find, that without risk enhancement plaintiff would have faced substantial difficulties in finding counsel in the local or other relevant market.^[12] Here, there were no such findings.

Accordingly, the judgment of the Court of Appeals is Reversed.

*731B JUSTICE O'CONNOR, concurring in part and concurring in the judgment.

For the reasons explained by the dissent I conclude that Congress did not intend to foreclose consideration of contingency in setting a reasonable fee under fee-shifting provisions such as that of the Clean Air Act, 42 U.S. C. § 7604(d), and the Civil Rights Attorney's Fees Awards Act, 42 U.S. C. § 1988. I also agree that compensation for contingency must be based on the difference in market treatment of contingent fee cases as a class, rather than on an assessment of the "riskiness" of any particular case. But in my view the plurality is also correct in holding that the "novelty and difficulty of the issues presented, and . . . the potential for protracted litigation," ante, at 726, are factors adequately reflected in the lodestar, and that the District Court erred in employing a risk multiplier in the circumstances of this case.

The private market commonly compensates for contingency through arrangements in which the attorney receives a percentage of the damages awarded to the plaintiff. In most fee-shifting cases, however, the private market model of contingency compensation will provide very little guidance. See Riverside v. Rivera, 477 U.S. 561, 573-576 (1986). Thus it is unsurprising that when courts have enhanced fee awards to compensate for risk, "[p]inpointing the degree of risk [has been] one of the most subjective and difficult components of the fee computation process, and one which [has been] apt to lead to imprecision in the final award." 2 M. Derfner & A. Wolf, Court Awarded Attorney Fees, ¶ 16.04 [c] [i], p. 16-88 (1986). Although the dissent suggests a method of calculating compensation for contingency that is theoretically more satisfying than the practice of speculating on the riskiness of each case, the dissent does not explain how the theory should be put into practice. For example, how should a court translate the extra economic risk endured by smaller firms, see post, at 750-751, or by firms that take unpopular cases, see post, at 751, n. 15, into a percentage enhancement?

Moreover, although the dissent offers no defense of this method of compensating for risk, it leaves the door open for "extra enhancement" for "exceptional cases" that pose great " `legal' risk." Post, at 751-752. The "extra enhancement" presumably would be calculated based on the likelihood at the time the litigation was commenced that the particular legal claims raised by the prevailing party would have been rejected by the court. This type of enhancement clearly is subject to the many difficulties described by the plurality. Ante, at 721-723. The dissent suggests that the plurality's objections "lose much of their force" because the cases in which "extra enhancement" is granted will be rare. Post, at 752, n. 16. But, an arbitrary or unjust result is no less so for its rarity. Furthermore, the difficulties created by this type of enhancement will arise not only when the enhancement is granted, but also whenever it is sought.

To be "reasonable," the method for calculating a fee award must be not merely justifiable in theory but also objective and nonarbitrary in practice. Moreover, if the concept of treating contingency cases as a class is to be more than symbolic, a court's determination of how the market in a community compensates for contingency should not vary significantly from one case to the next. I agree with the plurality that without guidance as to the trial court's exercise of discretion, adjustment for risk could result in "severe difficulties and possible inequities." Ante, at 728. In my view, certain constraints on a court's discretion in setting attorney's fees are appropriate.

First, district courts and courts of appeals should treat a determination of how a particular market compensates for contingency as controlling future cases involving the same market. Haphazard and widely divergent compensation for risk can be avoided only if contingency cases are treated as a class; and contingency cases can be treated as a class only if courts strive for consistency from one fee determination to the next. Determinations involving different markets should also comport with each other. Thus, if a fee applicant attempts to prove that the relevant market provides greater compensation for contingency than the markets involved in previous cases, the applicant should be able to point to differences in the markets that would justify the different rates of compensation.

Second, at all times the fee applicant bears the burden of proving the degree to which the relevant market compensates for contingency. See Blum v. Stenson, 465 U.S. 886, 898 (1984) ("The burden of proving that such an adjustment is necessary to the determination of a reasonable fee is on the fee applicant"); Hensley v. Eckerhart, 461 U.S. 424, 437 (1983) ("Where settlement is not possible, the fee applicant bears the burden of establishing entitlement to an award and documenting the appropriate hours expended and hourly rates"). I would also hold that a court may not enhance a fee award any more than necessary to bring the fee within the range that would attract competent counsel. I agree with the plurality that no enhancement for risk is appropriate unless the applicant can establish that without an adjustment for risk the prevailing party "would have faced substantial difficulties in finding counsel in the local or other relevant market." Ante, at 731.

Finally, a court should not award any enhancement based on "legal" risks or risks peculiar to the case. The lodestar — "the product of reasonable hours times a reasonable rate," Hensley v. Eckerhart, supra, at 434 — is flexible enough to account for great variation in the nature of the work performed in, and the challenges presented by, different cases. "The novelty and complexity of the issues" raised in a case "presumably [would be] fully reflected in the number of billable hours recorded by counsel." Blum, 465 U. S., at 898. The same can be said for most other problems posed by the litigation, such as the tenacity of the defendant. The "special skill and experience of counsel should be reflected in the reasonableness of the hourly rates." Ibid. Thus it is presumed that when counsel demonstrates considerable ability in overcoming unusual difficulties that have arisen in a case, counsel will be compensated for those accomplishments by means of an appropriate hourly rate multiplied by the hours expended.

Based on the above guidelines, the enhancement for risk awarded by the District Court in this case must be reversed. The enhancement is not supported by any findings of fact concerning the degree to which contingency is compensated in the relevant market. Neither the findings nor the evidence indicate that the large enhancements in this case were necessary to attract competent counsel in the relevant community. Moreover, it is clear that the District Court based the enhancement on "legal" risks and risks unique to the case. The considerations used by the District Court to justify the enhancement — the "new and novel issues" raised by the case, and the stubbornness of the defendants, 581 F. Supp. 1412, 1431 (1984) — should already be reflected in the number of hours expended and the hourly rate, and cannot be used again to increase the fee award.

Accordingly, I concur in Parts I, II, and III-A of the plurality and concur in the judgment reversing the judgment of the Court of Appeals.

JUSTICE BLACKMUN, with whom JUSTICE BRENNAN, JUSTICE MARSHALL, and JUSTICE STEVENS join, dissenting.

In enacting fee-shifting statutes, Congress stressed that the fee awarded must be "adequate to attract competent counsel, but . . . not produce windfalls to attorneys." S. Rep. No. 94-1011, p. 6 (1976). Today, a plurality of the Court ignores the fact that a fee that may be appropriate in amount when paid promptly and regardless of the outcome of the case, may be inadequate and inappropriate when its payment is contingent upon winning the case. By not allowing an upward adjustment for a case taken on a contingent basis, the plurality undermines the basic purpose of statutory attorney fees — ensuring that "private citizens . . . have a meaningful opportunity to vindicate the important Congressional policies which these laws contain." Id., at 2.^[1]

I

A

In the private market, lawyers charge a premium when their entire fee is contingent on winning. The Canons of Professional Ethics of the American Bar Association, as first promulgated in 1908, recognized that "[i]n determining the amount of the fee, it is proper to consider . . . (5) the contingency or the certainty of the compensation." Canons of Ethics § 12, 33 A. B. A. Rep. 575, 578 (1908). The ABA Model Code of Professional Responsibility, originally promulgated in 1969 and subsequently adopted by nearly every State, see Nix v. Whiteside, 475 U.S. 157, 167, n. 4 (1986), likewise provides that one of the "[f]actors to be considered" in determining a reasonable fee is "[w]hether the fee is fixed or contingent." Model Code of Professional Responsibility, DR 2-106(B)(8) (1980). The ABA's most recently formulated ethical standards, the ABA Model Rules of Professional Conduct, adopted in 1983, continue to reflect a consensus among lawyers that "whether the fee is fixed or contingent" is one of "[t]he factors to be considered in determining the reasonableness of a fee." Model Rule 1.5(a) and 1.5(a)(8).

The premium added for contingency compensates for the risk of nonpayment if the suit does not succeed and for the delay in payment until the end of the litigation — factors not faced by a lawyer paid promptly as litigation progresses. See Clermont & Currivan, Improving on the Contingent Fee, 63 Cornell L. Rev. 529, 556-557, 561-566 (1978); Schwartz & Mitchell, An Economic Analysis of the Contingent Fee in Personal-Injury Litigation, 22 Stan. L. Rev. 1125, 1150-1154 (1970); F. MacKinnon, Contingent Fees for Legal Services 28, 62 (1964). All else being equal, attorneys naturally will prefer cases where they will be paid regardless of the outcome, rather than cases where they will be paid only if they win. Cases of the latter type are inherently riskier and an attorney properly may expect greater compensation for their successful prosecution. See Lindy Bros. Builders, Inc. v. American Radiator & Standard Sanitary Corp., 487 F.2d 161, 168 (CA3 1973) (" `No one expects a lawyer whose compensation is contingent upon his success to charge, when successful, as little as he would charge a client who in advance had agreed to pay for his services, regardless of success' "), quoting Cherner v. Transitron Electronic Corp., 221 F. Supp. 55, 61 (Mass. 1963); Wildman v. Lerner Stores Corp., 771 F.2d 605, 612 (CA1 1985) (significant difference between a "case taken on a full retainer and a case in which an attorney spends many hours over a period of months or years with no assurance of any pay if the suit is unsuccessful"). See also E. Larson, Federal Court Awards of Attorney's Fees 224-225 (1981).

In the private market, the premium for contingency usually is recouped by basing the fee on a percentage of the damages recovered. The premium also could be computed as part of an hourly rate that the lawyer bills after the litigation succeeds. See Clermont & Currivan, 63 Cornell L. Rev., at 546-547, 567; See, An Alternative to the Contingent Fee, 1984 Utah L. Rev. 485, 499-503. Under either approach, the market-based fee or hourly rate that is contingent on success is necessarily higher than the hourly rate charged when payment is current and certain. This fee enhancement ensures that accepting cases on a contingent basis remains an economically attractive and feasible enterprise for lawyers. See generally MacKinnon, at 3-6.

B

In directing courts to award a "reasonable" attorney's fee to a litigant who vindicates various statutory rights, e. g., 42 U.S. C. § 7604(d) (Clean Air Act), Congress made clear that the winning lawyer should be paid at a rate that is basically competitive with what the lawyer is able to earn in other cases. Congress' purpose — extensively described in the legislative history of the Civil Rights Attorney's Fees Awards Act, 42 U.S. C. § 1988, but fully applicable to statutes that protect the environment, see Ruckelshaus v. Sierra Club, 463 U.S. 680, 691-692 (1983) — was to encourage the enforcement of federal law through lawsuits filed by private persons. Congress found that the market itself would not provide an adequate supply of interested lawyers because many potential plaintiffs lacked sufficient funds to hire such lawyers. See H. R. Rep. No. 94-1558, p. 1 (1976); S. Rep. No. 94-1011, at 2. Thus, fee awards were considered to be "an essential remedy" in order to encourage enforcement of the law. Ibid. And unless the fee reimbursement was "full and complete," the statutory rights would be meaningless because they would remain largely unenforced. H. R. Rep. No. 94-1558, at 1. See also Note, Promoting The Vindication of Civil Rights Through the Attorney's Fees Awards Act, 80 Colum. L. Rev. 346, 350-351, 372 (1980); Berger, Court Awarded Attorney's Fees: What Is "Reasonable"?, 126 U. Pa. L. Rev. 281, 306-310 (1977).

Congress determined that the public would be best served by the award of fees similar to what "is traditional with attorneys compensated by a fee-paying client." S. Rep. No. 94-1011, at 6. "It is intended that the amount of fees awarded. . . be governed by the same standards which prevail in other types of equally complex Federal litigation, such as antitrust cases and not be reduced because the rights involved may be nonpecuniary in nature." Ibid. See also H. R. Rep. No. 94-1558, at 9. Thus, in Blum v. Stenson, 465 U.S. 886 (1984), the Court emphasized: "The statute and legislative history establish that `reasonable fees' under § 1988 are to be calculated according to the prevailing market rates in the relevant community." Id., at 895.

Congress found that a broad variety of factors go into the computation of a "reasonable" attorney's fee. One such consideration is the contingency that the attorney will be paid only if he wins the case. Three of the four major cases cited as examples of "the appropriate standards . . . correctly applied," S. Rep. No. 94-1011, at 6 — and noted by this Court in Blum v. Stenson, 465 U. S., at 895, 897, n. 13 — mentioned this risk as a factor for a court to weigh. See Johnson v. Georgia Highway Express, Inc., 488 F.2d 714, 718 (CA5 1974);^[2]Stanford Daily v. Zurcher, 64 F. R. D. 680, 685-686 (ND Cal. 1974), aff'd, 550 F.2d 464 (CA9 1977), rev'd on other grounds, 436 U.S. 547 (1978); Swann v. Charlotte-Mecklenburg Board of Education, 66 F. R. D. 483, 486 (WDNC 1975). In the Stanford case, the court expressly increased the fee award for the contingent nature of the attorneys' work, noting that such an increase "parallel[s] the American Bar Association's determination that attorneys deserve higher compensation for contingent than for fixed-fee work." 64 F. R. D., at 685. The court explained: "From the public's standpoint, the contingent fee helps equalize the access of rich, middle-class, and poor individuals to the courts by making attorney decisions concerning representation turn on an action's merits rather than on the size of a client's income." Ibid. Thus, contrary to the plurality's assertion, see ante, at 723-724, Congress envisioned that district courts would take the fact of contingency into account when calculating a reasonable attorney's fee so that the resulting fee would be equivalent to prevailing market rates.^[3]

As courts have gained more experience with fee calculations, many have begun to utilize as a "lodestar" the reasonable hours worked multiplied by a reasonable hourly rate. See Lindy Bros. Builders, Inc. v. American Radiator & Standard Sanitary Corp., 487 F. 2d, at 167-168; Hensley v. Eckerhart, 461 U.S. 424, 433 (1983) (approving use of the lodestar approach). The lodestar, however, was designed to simplify, not to circumvent, application of the Johnson factors where appropriate. See Copeland v. Marshall, 205 U. S. App. D. C. 390, 400, 641 F.2d 880, 890 (1980) (en banc); Berger, 126 U. Pa. L. Rev., at 286-287. Thus, a statutory fee cannot be computed solely by reference to rates charged by corporate firms, which obtain many payments from their clients through monthly billings.^[4] Rather, in order to arrive at a "reasonable" attorney's fee, a court must incorporate a premium for the risk of nonrecovery, for the delay in payment, and for any economic risks aggravated by the contingency of payment, at a level similar to the premium incorporated in market rates. The risk premium can be reflected in the hourly rate that goes into the lodestar calculation, or, if the hourly rate does not include consideration of risk, in an enhancement of the lodestar. See Blum v. Stenson, 465 U. S., at 903 (concurring opinion). Under either approach, adding a premium simply brings the fee up to the "reasonable" level contemplated by Congress. See 2 M. Derfner & A. Wolf, Court Awarded Attorney Fees, ¶ 16.04, p. 16-84 (1986) ("The increase in the fee to account for contingency. . . is part of the fine tuning which courts must do to make a fee reflect the true market value of an attorney's efforts").^[5]

An adjustment for contingency is necessary if statutory fees are to be competitive with the private market and if competent lawyers are to be attracted in their private practice to prosecute statutory violations. See The Commitee on Legal Assistance Committee Report on Counsel Fees in Public Interest Litigation, 39 Record of N. Y. C. B. A. 300, 317 (1984). This is simply the law of supply and demand. If lawyers can earn substantially higher pay from other cases in the private sector, they will tend either to reject statutory-enforcement cases or they effectively will be penalized for taking such cases. See Brief for Twelve Small Private Civil Rights Law Firms as Amici Curiae 6-29 (describing financial difficulties in depending on contingency cases with statutory attorney's fees); see also Darden v. Illinois Bell Tel. Co., 797 F.2d 497, 505 (CA7 1986) (concurring opinion) ("Unless fees under § 1988 are enhanced to take account of the risk of losing — whether through a multiplier or by the contingent fee device when the stakes are high — fees will be systematically too small"). Allowing adjustments for contingency similar to that given to attorneys in the private market thus appropriately " `enable[s] counsel to accept apparently just causes without awaiting sure winners.' " Yates v. Mobile County Personnel Bd., 719 F.2d 1530, 1533 (CA11 1983), quoting Jones v. Diamond, 636 F.2d 1364, 1382 (CA5 1981).

Not surprisingly, the Courts of Appeals are in agreement that adjustments for the risk of nonrecovery are appropriate in most circumstances.^[6] And while this Court, in Blum v. Stenson, 465 U. S., at 901, n. 17, left open the question of contingency adjustments, the Court also made clear that Congress intended statutory fees to be competitive with the private market for lawyers' services. Id., at 895. Thus, competitive awards, rather than being "the kind of `windfall profits' [Congress] expressly intended to prohibit," ibid., are, instead, the way to implement congressional intent.

C

If it were the law of the land, the plurality's decision, in Part IV of its opinion, to foreclose any compensation for the risk of nonrecovery would reduce statutory fees below the market rate and inevitably would obstruct the vindication of federal rights.^[7] Because fewer lawyers would be attracted to the work, some persons who now are able to bring valid claims would be unable to find a lawyer. They likely would be persons of modest means who could not afford to augment their lawyer's fee to what the market would charge — precisely the persons Congress sought to assist. Even plaintiffs who somehow could manage to attract lawyers at below the market rate usually would get what they pay for: lawyers who are less than fully employed or who are less capable. Without compensation for contingency, "[b]usy and successful attorneys simply could not afford to accept contingent employment . . . . Such an arrangement would ill serve policies of enormous national importance." Yates v. Mobile County Personnel Bd., 719 F. 2d, at 1534. As Congress recognized, effective enforcement of complex cases requires the services of experienced attorneys. Such lawyers are less likely to be underemployed. They therefore will tend to demand rates approaching what can be obtained in the private sector. Berger, at 314-315.

The plurality offers assurances that enforcement would not end completely because public-interest groups would still take these cases. See ante, at 726. In effect, the plurality would place the entire burden of injunctive actions and modest damages claims on the shoulders of the public-interest bar. But it is unrealistic to think that 600 public-interest lawyers in 90 public-interest law centers around the country would be able to pick up the slack from the rest of the bar, with its approximately 400,000 lawyers. "The services of the pro bono bar, which is concentrated in the eastern urban centers, simply [are] not available to most people." Berger, at 313 (footnote omitted).

Significantly, the plurality's opinion would validate payment of public-interest lawyers at substantially less than what would be competitive with the private market. In Blum, however, this Court made clear that nonprofit legal-aid organizations should receive no less in fee awards than the hourly rate set by the private market for an attorney's services. 465 U.S., at 895.^[8] See also New York Gaslight Club, Inc. v. Carey, 447 U.S. 54, 70, n. 9 (1980). The plurality today attempts to accomplish indirectly what the Court refused to do directly in Blum.

The plurality further defends its approach by asserting that plaintiffs bringing large damages claims could continue to attract private lawyers. See ante, at 726. But those plaintiffs might be able to hire counsel in any event through private contingency arrangements. Congress provided for fee shifting precisely because it concluded that too many plaintiffs would be unable to obtain representation in this manner. The plurality's solution would slight actions that seek injunctive relief or relatively small damages awards, on which the vindication of many federal rights depends. See Riverside v. Rivera, 477 U.S. 561 (1986).

II

In view of Congress' desire that statutory fees be competitive with the private market, the plurality needs a compelling reason in order to reject the market approach for determining what constitutes a reasonable fee. Although the plurality suggests some reasons, its objections are all based on a fundamental mischaracterization of the enhancement for contingency in awarding attorney's fees. The Court states that the issue before it is whether an attorney "may be awarded separate compensation for assuming the risk of not being paid," and explains that "[t]hat risk is measured by the risk of losing rather than winning and depends on how unsettled the applicable law is with respect to the issues posed by the case and by how likely it is that the facts could be decided against the complainant." Ante, at 715-716. Having framed the issue in this fashion, the Court discovers significant problems in allowing enhancements based on the likelihood of success in particular cases. The Court, for example, says it is disturbed that evaluation of the risk of loss may create a conflict of interest between attorneys and their clients, as both plaintiff and defense attorneys try to characterize their cases in a manner that will increase or decrease attorney's fees, ante, at 721-722, and that it is difficult for a court to estimate retroactively a prevailing party's chance of success once a court knows the outcome. Ante, at 722. The Court further notes that such enhancements have the perverse result of penalizing defendants with the strongest defenses and causing those defendants to subsidize plaintiffs' attorneys for other unsuccessful actions that they may bring. Ibid. This last concern is of great significance to the plurality because it appears to be "[in]consistent with Congress' decision to adopt the rule that only prevailing parties are entitled to fees." Ante, at 725. The Court also expresses alarm, echoing the opinion of the Court of Appeals for the District of Columbia Circuit in Laffey v. Northwest Airlines, Inc., 241 U. S. App. D. C. 11, 33, 746 F.2d 4, 26 (1984), cert. denied, 472 U.S. 1021 (1985), that "in theory, there should be no limit on the size of the fee if risk enhancement is permitted, for the less likely the chances of success in a particular case, the more `entitled' the prevailing party should be to have the fee award reflect acceptance of this risk." Ante, at 719; see also ante, at 725.

The underlying flaw in all of these objections is that the appropriate enhancement for risk does not depend, in the first instance, on the degree of risk presented by a particular case. Enhancement for risk is not designed to equalize the prospective returns among contingent cases with different degrees of merit.^[9] Rather, it is designed simply to place contingent employment as a whole on roughly the same economic footing as noncontingent practice, in order that such cases receive the equal representation intended by Congress. Enhancement compensates attorneys for the risk of nonpayment associated with contingent employment, a risk that does not exist in noncontingent cases. As discussed above, without the possibility of enhancement for contingency, an attorney, from a simple economic point of view, would prefer noncontingent employment to contingent employment. This is because even contingent cases with the best merit may sometimes fail, because delay in payment is inherent in any contingent arrangement, and because other economic risks may be aggravated by the contingency in payment.^[10] Thus, contrary to the plurality's vision of an enhancement that radically increases as a case's chance of success decreases, an enhancement for contingency in ordinary cases will not be based on the relative likelihood of success of a particular case.^[11]

Once it is recognized that it is the fact of contingency, not the likelihood of success in any particular case, that mandates an increase in an attorney's fee, the frightening difficulties envisioned by the plurality disappear. There is no reason to assume that, in most cases, a court will have to delve into the strengths and weaknesses of a particular case, that potential conflict of interests will arise between attorneys and clients, or that large enhancements disproportionate to the success of a case will be granted. Rather, a court's job simply will be to determine whether a case was taken on a contingent basis, whether the attorney was able to mitigate the risk of nonpayment in any way, and whether other economic risks were aggravated by the contingency of payment. The Court of Appeals for the First Circuit correctly points out that "it is the actual risks or burdens that are borne by the lawyer or lawyers that determine whether an upward adjustment is called for." Wildman v. Lerner Stores Corp., 771 F. 2d, at 613.^[12]

The American Bar Association sets forth a reasonable approach for determining when, and to what degree, enhancement is appropriate in calculating a statutory attorney's fee. Brief for American Bar Association as Amicus Curiae 18-22. A court first must determine whether an attorney has taken a case on a contingent basis. If a client has contracted to pay the "lodestar" fee (i. e., reasonable hours times a reasonable hourly rate), regardless of the outcome of the case, and has paid the attorney on a continuing basis, then the attorney has clearly avoided the risk of nonpayment and enhancement is not appropriate. See, e. g., Ohio-Sealy Mattress Mfg. Co. v. Sealy Inc., 776 F.2d 646, 660 (CA7 1985) (plaintiff paid lawyers on a regular hourly basis and therefore "lawyers neither risked noncompensation nor endured a delay before payment"); Jones v. Central Soya Co., 748 F.2d 586, 592 (CA11 1984) (no indication that case was taken on a contingent basis and therefore upward adjustment not appropriate). The court also must determine if an attorney has been able to mitigate the risk of nonpayment in any way. For example, if a client has agreed to pay some portion of the lodestar amount, regardless of the outcome of the case, the attorney has mitigated the risk of loss to some extent, although the percentage of total expenses paid by the client will indicate how much of a mitigating factor this contribution should be considered to be. See, e. g., Stanford Daily v. Zurcher, 64 F. R. D., at 686 (client agreed to pay $5,000 retainer and to undertake fundraising effort, but attorneys clearly not guaranteed payment for most of the hours expended). Or, for example, if the attorney has entered into a contingent-fee contract in a suit seeking substantial damages, the attorney again has mitigated the risk to some extent by exchanging the risk of nonpayment for the prospect of compensation greater than the prospective lodestar amount. Even in such cases, of course, a court must still calculate a reasonable attorney's fee to be assessed against the defendant. There is no reason to grant a defendant a "windfall" by excusing payment of attorney's fees simply because a plaintiff has entered into a contingent-fee contract. See, e. g., Hamner v. Rios, 769 F.2d 1404, 1408-1410 (CA9 1985); Sargeant v. Sharp, 579 F.2d 645, 649 (CA1 1978).

If an attorney and client have been unable to mitigate the risk of nonpayment, then an enhancement for contingency is appropriate. In many cases, a client will be unable to pay for counsel or will be unwilling to assume the risk of liability for attorney's fees, even if the public interest may be significantly aided by the private litigation. Other cases simply will not generate significant funds, even if they are successful. Many actions seek only declaratory or injunctive relief, many are hampered by immunity doctrines and special defenses available to the defendants, and many will generate only small awards. See H. R. Rep. No. 94-1558, at 9; Riverside v. Rivera, 477 U.S. 561 (1986); see also Rowe, The Legal Theory of Attorney Fee Shifting: A Critical Overview, 1982 Duke L. J. 651, 676.

As the American Bar Association points out, a court must not only determine whether an attorney has been able to mitigate the economic risk of nonpayment, it must also determine whether specific aspects of the case have aggravated that economic risk. Brief for American Bar Association as Amicus Curiae 21-22. The enhancement for contingency compensates the attorney primarily for the risk of spending numerous hours, indeed often years, on a case, with the knowledge that no payment may ever be recovered. Other aspects of a case, however, can aggravate the economic risk inherent in contingency payments.

First, although the Court treats delay in payment as independent of the risk of nonpayment, see ante, at 716, such delay is an integral aspect of contingency payments for which compensation is appropriate.^[13] Delay in payment causes cash-flow problems and deprives an attorney of the use of money, thus magnifying the economic risk associated with the uncertainty of payment. See, e. g., Copeland v. Marshall, 205 U. S. App. D. C., at 403, 641 F.2d, at 893; Brief for Twelve Small Private Civil Rights Law Firms as Amici Curiae 6-31 (describing difficulties related to delayed payments). Indeed, some types of litigation, such as cases seeking institutional reform or involving complex environmental issues, have a potential for such significant delay that attorneys must be assured of an appropriate enhancement in order to offset the financial disincentive to taking such cases.

Second, a case may present greater economic risks because of a particular attorney's circumstances. For example, contingent litigation may pose greater risks to a small firm or a solo practitioner because the risk of nonpayment may not be offset so easily by the presence of paying work, and because such paying work may have to be turned away once a contingent case is accepted.^[14] Conversely, where responsibility for a case is shared among several firms, the relative economic risk borne by each firm may be diminished. See, e. g., In re Fine Paper Antitrust Litigation, 98 F. R. D. 48, 84 (ED Pa. 1983), aff'd in part and rev'd in part, 751 F.2d 562 (CA3 1984).^[15]

In most cases in which an enhancement for contingency is sought, therefore, a court will not need to inquire into the relative likelihood of success of the particular case before it. It is possible, however, that in a few, unusual cases the likelihood of success may appropriately be taken into account. Sometimes, the "legal" risks facing a case may be so apparent and significant that they will constitute an economic disincentive independent of that created by the basic contingency in payment. When the result achieved in such a case is significant and of broad public interest, an additional enhancement is justified in order to attract attorneys to take such cases, which otherwise might suffer from lack of representation. Extra enhancement for such cases, however, should be awarded in exceptional cases only. In most cases where the "legal" risks are high, and the case therefore novel and difficult, attorneys may be expected to spend a greater number of hours preparing and litigating the case. Courts should consider this seriously in determining the number of "reasonable" hours to be incorporated in the lodestar and should be careful not to reduce unduly the number of hours in a novel and difficult case. If a court finds, however, that an attorney has taken a significant legal risk in a case of important public interest, and that this risk has not been compensated adequately by the court in the number of hours represented in the lodestar, the court may then grant an enhancement above that awarded for the basic contingency risk. In such a case, the court must make detailed findings regarding the particular legal risks that were apparent at the outset of the litigation and the importance of the result obtained — findings that would justify the additional enhancement.

Almost all of the plurality's objections to enhancements for contingency become irrelevant once such enhancement is seen, as a general matter, to be completely independent from the likelihood of success in particular cases. Under the approach outlined above, there is no reason for a court to assess the success of a case retroactively, no cause for a conflict of interest to arise between attorney and client, and no possibility of a grant of huge multipliers simply because the odds against a case were significant.^[16] The only remaining objection is that awarding higher fees to lawyers who accept contingent cases gives such lawyers the economic stability with which to bring other, possibly unsuccessful, lawsuits. In the plurality's view, this contravenes Congress' intent to award attorney's fees only to prevailing parties. Ante, at 725. But this objection must ultimately fail. The fact is that an attorney still recovers fees only when that attorney's client prevails in a lawsuit. If the attorney represents a client in an unsuccessful contingent litigation, no fees are recovered. That the attorney may use the fees obtained in the successful contingency lawsuit to bring other lawsuits — some of which will not be successful — does not contravene in any way Congress' mandate that fees be awarded solely to prevailing parties.^[17] There is certainly no indication in the legislative history of § 1988 that Congress' restriction of attorney's fees to prevailing parties was intended to deny reasonable fees to attorneys who would use their income to subsidize unsuccessful litigation. Indeed, that would be contrary to Congress' intent that attorneys bringing statutory violation cases be compensated in a manner similar to that by which attorneys in the private market are compensated. As in the private market, what a successful attorney does with earned fees is the attorney's own business.

The basic objective for courts to keep in mind in awarding enhancements for risk is that a "reasonable attorney's fee" should aim to be competitive with the private market, even if it is not possible to reflect that market perfectly. Thus, an enhancement for contingency, whether calculated as an increase in the reasonable hourly rate used to arrive at the lodestar or added to the lodestar as a bonus or a multiplier, is not designed to be a "windfall" for the attorney of the prevailing party. Rather, it is designed to ensure that lawyers who take cases on contingent bases are properly compensated for the risks inherent in such cases. Vindication of the statutory rights passed by Congress depends on the continued availability and willingness of highly skilled lawyers to take cases for which they will receive a statutory attorney's fee. In setting such fees, courts must ensure that the fees are "reasonable" — i. e., that the fees properly compensate an attorney for the risks assumed.

III

Respondent Delaware Valley "clearly prevailed in attaining what [it] sought and [won results that] would not have occurred without [its] efforts." 581 F. Supp. 1412, 1420 (ED Pa. 1984). As a prevailing party, Delaware Valley is entitled to a statutory fee award. Newman v. Piggie Park Enterprises, Inc., 390 U.S. 400, 402 (1968).

This case also appears to be a candidate for a contingency adjustment because the plaintiffs' lawyers apparently accepted the case on the expectation that they would be paid only if their clients prevailed. The District Court, however, explained its award of a multiplier in three phases of the litigation in the following brief statement:

"The contingent nature of plaintiff's success has been apparent throughout this litigation. Plaintiffs entered the litigation against the U. S. Government and the Commonwealth of Pennsylvania. The case involved new and novel issues, the resolution of which had little or no precedent. Commencing in Phase IV and continuing up until the present, plaintiffs have had to defend their rights under the consent decree due to numerous attempts by defendants and others to overturn or circumvent this court's Orders." 581 F. Supp., at 1431.

I conclude that we should vacate the award and remand the case to the District Court for further findings. First, as I have explained, the District Court should determine whether respondent's attorneys took this case on a contingent basis, whether they were able to mitigate the risks of nonpayment in any way, and whether other economic risks were aggravated by the contingency of payment. The court then should arrive at an enhancement for risk that parallels, as closely as possible, the premium for contingency that exists in prevailing market rates. The court should thereby arrive at an enhancement that appropriately compensates the attorneys for the risks assumed.

Second, the court might also determine whether this case deserves an extra enhancement because of the significant legal risks apparent at the outset of the litigation and because of the importance of the case. I would note, however, that respondent's attorneys began this litigation in order to enforce a consent decree — a situation that does not usually entail difficult legal risks. If the District Court were to believe that the case nonetheless did involve significant legal risks at the outset of the litigation, it would make specific findings to that effect and would not simply state that the "case involved new and novel issues." Ibid.^[18]

The plurality's per se ruling, in Part IV of its opinion, against enhancements for contingency contravenes Congress' express intent that statutory attorney's fees should be equivalent to prevailing market rates so that highly skilled lawyers will be available to vindicate the statutory rights conferred by Congress. At the least, however, the majority of this Court leaves open the opportunity for district courts to award enhancements for contingency in selected cases.

I respectfully dissent.

NOTES

[*] Briefs of amici curiae urging affirmance were filed for the American Bar Association by Eugene C. Thomas, John R. Hupper, Thomas D. Barr, and John H. Pickering; and for Joseph A. Bonjorno et al. by Henry T. Reath and Michael M. Baylson.

[1] Section 304(d) provides, in relevant part:

"The Court, in issuing any final order in any action brought pursuant to subsection (a) of this section may award costs of litigation (including reasonable attorney and expert witness fees) to any party, whenever the Court determines such award is appropriate."

Last Term in Pennsylvania v. Delaware Valley Citizens' Council for Clean Air, 478 U.S. 546 (1986), we agreed with the Court of Appeals that in awarding attorney's fees under § 304(d) the courts should follow the principles and case law governing the award of such fees under 42 U.S. C. § 1988, which provides that in the actions specified in that section "the court, in its discretion, may allow the prevailing party, other than the United States, a reasonable attorney's fee as part of the costs."

[2] At this time, respondent was awarded an attorney's fee for work done by its counsel, the Public Interest Law Center of Philadelphia (PILCOP), prior to the date of the consent decree.

[3] We granted certiorari last Term, 474 U.S. 815 (1984), heard argument, and issued an opinion holding that respondent was entitled to attorney's fees under § 304(d) for its counsel's work done in certain administrative proceedings because the work "was crucial to the vindication of Delaware Valley's rights under the consent decree . . . ." 478 U.S., at 561. We also concluded that the District Court erred by enhancing the fee award based on the "superior quality" of counsel's performance, reasoning that respondent did not show "why the lodestar did not provide a reasonable fee award reflecting the quality of representation . . . ." Id., at 567. We did not, however, address the merits of the question now before of us, an issue that was left open in Blum v. Stenson, 465 U.S. 886 (1984). We thought that reargument on this issue would be beneficial. We therefore restored this aspect of the case to the docket for decision this Term. 478 U.S., at 568.

[4] Numerous Courts of Appeals, acting under fee-shifting statutes, have approved an upward adjustment of the lodestar to compensate for the risk of not prevailing. See, e. g., Crumbaker v. Merit Systems Protection Board, 781 F.2d 191, 196-197 (CA Fed. 1986); Vaughns v. Board of Ed. of Prince Georges County, 770 F.2d 1244 (CA4 1985), aff'g 598 F. Supp. 1262, 1285-1286 (Md. 1984); Riddell v. National Democratic Party, 712 F.2d 165, 169-170 (CA5 1983); Kelley v. Metropolitan County Bd. of Ed., 773 F.2d 677, 683, 686 (CA6 1985) (en banc), cert. denied, 474 U.S. 1083 (1986); Craik v. Minnesota State University Bd., 738 F.2d 348, 350-351 (CA8 1984); White v. Richmond, 713 F.2d 458, 462 (CA9 1983); Ramos v. Lamm, 713 F.2d 546, 557-558 (CA10 1983); Jones v. Central Soya Co., 748 F.2d 586, 591 (CA11 1984).

In addition to the Courts of Appeals for the District of Columbia Circuit and the Seventh Circuit, other courts have refused risk enhancement for a variety of reasons. See, e. g., Lewis v. Coughlin, 801 F.2d 570, 576 (CA2 1986) (upward adjustment vacated for failure to evaluate risk of loss); Lanasa v. New Orleans, 619 F. Supp. 39, 50-51 (ED La. 1985) (settlement for low money damages figure could have been agreed to much earlier in litigation); Littlejohn v. Null Mfg. Co., 619 F. Supp. 149, 152 (WDNC 1985) (attorney received fully compensatory fee without adjustment); Bennett v. Central Telephone Co. of Illinois, 619 F. Supp. 640, 653 (ND Ill. 1985) (lack of supporting evidence and high hourly rates); EEOC v. Burlington Northern Inc., 618 F. Supp. 1046, 1061-1062 (ND Ill. 1985) (high hourly rates and risk of nonsuccess not unusually high); Litton Systems, Inc. v. American Telephone & Telegraph Co., 613 F. Supp. 824, 835 (SDNY 1985) (no great incentive needed to encourage appellee to defend its $276 million antitrust judgment on appeal); Cook v. Block, 609 F. Supp. 1036, 1043-1044 (DC 1985) (counsel guaranteed payment by client even if suit was unsuccessful); Cherry v. Rockdale County, 601 F. Supp. 78, 80-81 (ND Ga. 1984) (insufficient evidence supporting adjustment); Inmates of Maine State Prison v. Zitnay, 590 F. Supp. 979, 987 (Me. 1984) (contingency already reflected in lodestar); Rank v. Balshy, 590 F. Supp. 787, 799-800 (MD Pa. 1984) (contingency already reflected in lodestar).

[5] What the court viewed as the simple economics of the practice of law played a major part in the Court of Appeals' analysis:

"[T]he lodestar figure alone does not differentiate between the case taken on a full retainer and a case in which an attorney spends many hours over a period of months or years with no assurance of any pay if the suit is unsuccessful. Even if the client ultimately prevails, the burden of supporting salaried employees and fixed costs during the course of the contingent litigation can be substantial.

"Moreover, the attorney may face a second risk once his clients has prevailed — that the court will find some of his time duplicative, unnecessary, or inefficiently expended.

.....

"We think it clear that Congress did not intend that the enforcement of civil rights be limited primarily to those able to pay an attorney a full retainer or attract one of the few pro bono legal service organizations to their cause . . . [to] deny all considerations of the added burden and additional risks an attorney under a contingent fee agreement may have to bear does not strike us as `reasonable.' " 771 F.2d, at 612-613 (citations omitted). We note that some of the factors mentioned by the Court of Appeals are, in our mind, irrelevant to whether there should be separate compensation for assuming the risk of nonpayment.

[6] The Seventh Circuit has ruled that "the risk of losing `alone does not justify the use of a multiplier.' " McKinnon v. Berwyn, 750 F.2d 1383, 1392 (1984) (citations omitted). That court followed the reasoning of Laffey v. Northwest Airlines, Inc., 241 U. S. App. D. C. 11, 746 F.2d 4 (1984), finding that "[t]he fundamental problem of a risk bonus is that it compensates attorneys, indirectly but effectively, for bringing unsuccessful. . . suits, even though the attorney's fee statute is expressly limited to cases where the party seeking the fee prevails." 750 F.2d, at 1392. The court also reasoned that, in cases where the attorney has entered into a contingent-fee contract with his client, the attorney is already being compensated for the risk of loss, and "is not entitled to more insurance in the form of a risk multiplier." Id., at 1393. Ohio-Sealy Mattress Mfg. Co. v. Sealy Inc., 776 F.2d 646 (CA7 1985), and Kirchoff v. Flynn, 786 F.2d 320 (CA7 1986), may evidence some withdrawal from that position, but in a still later case the Court of Appeals, citing McKinnon, said that "this circuit has not favored the use of risk multipliers." In re Burlington Northern, Inc., 810 F.2d 601, 608 (1986).

[7] Hearings before a congressional Subcommittee also illuminate the differing views about the desirability and necessity of enhancement for the risk of loss. Hearings on S. 2802 before the Subcommittee on the Constitution of the Committee on the Judiciary, 98th Cong., 2d Sess. (1984); Hearings on S. 1580 et al. before the Subcommittee on the Constitution of the Committee on the Judiciary, 99th Cong., 1st Sess. (1985).

[8] In a similar vein, "[i]f we want to encourage private attorney general suits, risky plaintiffs' test litigation, or claims for nonmonetary relief, forbidding the shifting of compensation for risk could deter the bringing of such cases." Rowe, The Legal Theory of Attorney Fee Shifting: A Critical Overview, 1982 Duke L. J. 651, 676.

[9] The District Court employed an interesting approach in denying a risk multiplier in Cherry v. Rockdale County, 601 F. Supp., at 80-81. The court noted that risk enhancement is justified only if it is needed to provide compensation at a sufficient level to attract capable advocates. Because no evidence had been proffered showing what level of compensation was necessary to so attract lawyers, there was an insufficient evidentiary base upon which to award a fee enhancement. The court then argued "by way of illustration only," why in the case before it the evidence would not support a risk multiplier. According to a national survey to which the court had access, if the two attorneys requesting fees were paid at the rate that the top 25% of law firm partners admitted at the same time were paid, they would earn $77,800 and $89,800, respectively, per year. If they were associates, their annual salaries would be $56,800 and $61,300. This same survey showed that practitioners in small firms had an average overhead expense of $47,000 per lawyer. The court assumed that a reasonably diligent lawyer should bill 2,000 hours per year (40 hours per week multiplied by 50 weeks). The court relied on the parties' affidavits that stated a reasonable hourly wage for these attorneys was $100. The court then concluded that if the attorneys lost one-third of their cases a year, their compensation would still be within the upper 25% of compensation for all lawyers — i. e., $85,000 (($200,000 minus $68,000 (uncollectible)) minus $47,000 (overhead) equals $85,000). Any enhancement, the court observed, was simply unnecessary. If the lawyers lost one-half of all their cases in a year, some enhancement might be necessary, as compensation based on this loss rate would be only $53,000.

[10] We note the argument advanced by amici Arizona et al., but not dealt with by the parties or the courts below, that the attorneys for respondent, PILCOP, could not have properly accorded any weight whatsoever to the perceived risk of not prevailing when deciding to undertake representation in this case, due to PILCOP's tax-exempt status under the Internal Revenue Code. Brief for Arizona et al. as Amici Curiae 56-57. In its fee petition to the District Court, PILCOP asserted that it is "a non-profit, tax exempt law corporation," and that it was prohibited by certain Internal Revenue Service (IRS) "regulations" from accepting fees from its clients, including respondent. App. 161a-162a. PILCOP was undoubtedly referring to Rev. Proc. 71-39, 1971-2 Cum. Bull. 575, which provides that a public interest law firm desiring tax-exempt status may not accept fees for its services except in accordance with procedures approved by the IRS. Subsequently, the IRS issued Rev. Proc. 75-13, 1975-1 Cum. Bull. 662, which amplified Rev. Proc. 71-39 by setting forth procedures under which a public interest law firm could accept fees for its services and maintain its charitable organization, tax-exempt status. These procedures included the requirement, among others, that the public interest law firm "not use the likelihood or probability of a fee award as a consideration in its selection of cases." The argument advanced is that the tax-exempt law firm really risks nothing in cases like this and that because PILCOP undertook to represent respondent's cause without regard to the likelihood of eventually recovering fees under § 304(d) of the Clean Air Act, it follows that PILCOP could not validly have entertained the notion that if respondent did ultimately succeed in the litigation, its fee award would possibly have been enhanced due to the risk of not prevailing. Amici note that the Court of Appeals for the Second Circuit does not allow a contingency multiplier in awarding fees to nonprofit law firms. New York Assn. for Retarded Children, Inc. v. Carey, 711 F.2d 1136 (1983). We do not pass on the submission of the amici.

[11] "The test . . . should be an objective one based on the likely response of the bar to the case's pre-trial merits, rather than on the judge's subjective opinion of the merits." Lewis v. Coughlin, 801 F. 2d, at 575.

[12] "[A]n attorney's fee award should be only as large as necessary to attract competent counsel" and "one relevant factor bearing on high-risk is whether other counsel had declined to take the case because there was little or no prospect of earning a fee." Lewis v. Coughlin, supra, at 576.

[1] The concurrence recognizes that Congress did not intend to foreclose enhancements for contingency in the setting of reasonable attorney's fees. See ante, at 731. The plurality also recognizes, after a fashion, that fee-shifting statutes might be "construed to permit supplementing the lodestar in appropriate cases by paying counsel for assuming the risk of nonpayment." See ante, at 728. Neither opinion, however, follows through on its analysis by remanding the case to the District Court for application of the proper standards.

[2] In Johnson v. Georgia Highway Express, Inc., the court essentially adopted the factors that the ABA Model Code of Professional Responsibility DR 2-106(B) sets forth as guidelines for determining the size of an appropriate attorney's fee, including "[w]hether the fee is fixed or contingent." 488 F.2d, at 718 (emphasis omitted). The other factors cited by the court included: the time and labor required for the case, the novelty and difficulty of the questions, the skill required to perform the legal service properly, the preclusion of other employment, the customary fee, time limitations imposed by the client, the experience of the attorneys, the undesirability of the case, the nature of the relationship with the client, and awards in similar cases. Id., at 717-719. Many of these factors can be included within the "lodestar," in which the reasonable hours worked are multiplied by a reasonable hourly rate. See Blum v. Stenson, 465 U. S., at 898-899.

[3] Bills have been introduced in Congress to prohibit "bonuses or multipliers," including adjustments for the risk of nonrecovery, where a suit is brought against the United States, a State, or a local government. See H. R. 5757, 98th Cong., 2d Sess., §§ 6(a)(1) and (2) (1984); S. 2802, 98th Cong., 2d Sess., §§ 6(a)(1) and (2) (1984); H. R. 3181, 99th Cong., 1st Sess., §§ 6(a)(1) and (2) (1985); S. 1580, 99th Cong., 1st Sess., §§ 6(a)(1) and (2) (1985). So far these efforts to limit recovery of attorney's fees have been unavailing.

[4] Lawyers who are paid on an hourly basis, of course, face some risk of nonpayment, because not all hours worked can be billed. The hourly rate charged may reflect this. See Murray v. Weinberger, 239 U. S. App. D. C. 264, 272, 741 F.2d 1423, 1431 (1984).

[5] I therefore have little quarrel with the plurality's statement that "[n]either the Clean Air Act nor § 1988 expressly provides for using the risk of loss as an independent basis for increasing an otherwise reasonable fee." Ante, at 723 (emphasis added). The difficulty is that, on purely economic terms, a statutory attorney's fee is not "otherwise reasonable" if it fails to include some premium for the contingency in payment.

[6] After the Court left the issue open in Blum v. Stenson, 465 U. S., at 901, n. 17, almost all Courts of Appeals have upheld enhancements for contingency or have ruled that such enhancements are allowable in appropriate circumstances. See, e. g., Wildman v. Lerner Stores Corp., 771 F.2d 605, 613 (CA1 1985); Lewis v. Coughlin, 801 F.2d 570, 573-574 (CA2 1986); Durett v. Cohen, 790 F.2d 360, 364, and n. 3 (CA3 1986); Vaughns v. Board of Ed. of Prince George's County, 770 F.2d 1244, 1246 (CA4 1985); Sims v. Jefferson Downs Racing Assn., Inc., 778 F.2d 1068, 1084 (CA5 1985); Kelley v. Metropolitan County Bd. of Ed., 773 F.2d 677, 683, 686 (CA6 1985) (en banc), cert. denied, 474 U.S. 1083 (1986); Moore v. Des Moines, 766 F.2d 343, 346 (CA8 1985), cert. denied, 474 U.S. 1060 (1986); Planned Parenthood v. Arizona, 789 F.2d 1348, 1353-1354 (CA9 1986); Ramos v. Lamm, 713 F.2d 546, 558 (CA10 1983), cited with approval in Jordan v. Heckler, 744 F.2d 1397, 1401-1402 (CA10 1984); Jones v. Central Soya Co., 748 F.2d 586, 591 (CA11 1984); Crumbaker v. Merit Systems Protection Board, 781 F.2d 191, 196 (CA Fed. 1986).

The Courts of Appeals for the District of Columbia and the Seventh Circuits have questioned the propriety of risk enhancement. See Laffey v. Northwest Airlines, Inc., 241 U. S. App. D. C. 11, 36, 746 F.2d 4, 29 (1984), cert. denied, 472 U.S. 1021 (1985); McKinnon v. Berwyn, 750 F.2d 1383, 1392-1393 (CA7 1984). The Court repeats the analyses of these cases in some detail. See ante, at 719-720, and n. 6. Panels in each of those two Circuits, however, have indicated that enhancements would be appropriate in certain situations. See Murray v. Weinberger, 239 U. S. App. D. C. 264, 272-273, 741 F.2d 1423, 1431-1432 (1984); Ohio-Sealy Mattress Mfg. Co. v. Sealy, Inc., 776 F.2d 646, 661 (CA7 1985). Moreover, as the plurality does in this case, the courts in both Laffey and McKinnon misconstrued the nature and purpose of enhancements for contingency.

[7] The plurality's conclusion in Part IV of its opinion is, of course, not the governing law because five Members of this Court (those who are parties to this opinion and JUSTICE O'CONNOR) believe that an enhancement for contingency is appropriate in specified cases. See ante, at 731 (concurrence) ("I conclude that Congress did not intend to foreclose consideration of contingency in setting a reasonable fee under fee-shifting provisions such as that of the Clean Air Act, 42 U.S. C. § 7604(d), and the Civil Rights Attorney's Fees Awards Act, 42 U.S. C. § 1988").

[8] A nonprofit, public-interest law firm cannot obtain additional revenue by charging its clients for services (other than expenses) and still retain its tax-exempt status as a charitable organization. See 26 CFR § 1.501(c)(3)-1 (1987); Rev. Proc. 71-39, § 3.02, 1971-2 Cum. Bull. 575; Rev. Proc. 75-13, 1975-1 Cum. Bull. 662; Rev. Rul. 75-74, 1975-1 Cum. Bull. 152; Rev. Rul. 75-75, 1975-1 Cum. Bull. 154. Acceptance of statutory fee awards within limits, however, does not jeopardize that status. See Rev. Rul. 75-76, 1975-1 Cum. Bull. 154. And public-interest firms generally may not use money they receive from the Federal Government, such as the limited public funding received through the Legal Services Corporation, in cases where fees are available. See 42 U.S. C. § 2996f(b)(1); Hensley v. Eckerhart, 461 U.S. 424, 446, n. 6 (1983) (opinion concurring in part and dissenting in part).

[9] Such equalization is the result under the plurality's view of risk enhancement. Under that view, an attorney who takes relatively weak contingent cases would prevail infrequently, but could expect to receive large enhancements. This attorney would thus receive approximately the same compensation as an attorney who takes stronger contingent cases, prevails more often, and accordingly receives smaller enhancements.

[10] See, e. g., Hensley v. Eckerhart, 461 U. S., at 449 (opinion concurring in part and dissenting in part) ("Courts applying § 1988 must also take account of the time-value of money and the fact that attorneys can never be 100% certain they will win even the best case"); Crumbaker v. Merit Systems Protection Board, 781 F. 2d, at 197 (argument that no risk enhancement should be allowed, because plaintiff's attorney must have known, on the basis of law and facts in the case, that the defendant could not prevail, is "inane"; case required the best ability of counsel, defendant vigorously contested the case, and no case is certain to prevail).

[11] See Leubsdorf, The Contingency Factor in Attorney Fee Awards, 90 Yale L. J. 473, 501 (1981) (Leubsdorf) ("A contingency award does not require an inquiry into the likelihood of success in each case"); Note, Attorney Fees and the Contingency Factor Under 42 U.S. C. § 1988: Blum v. Stenson, 465 U.S. 886 (1984), 64 Ore. L. Rev. 571, 588 (1986) ("[S]tandard contingency adjustment, unrelated to the risks of any particular case, [should be] used in calculating section 1988 attorney's fees").

Indeed, it is ironic that the Court draws as heavily as it does on the article by Professor Leubsdorf, see ante, at 721-722 — an article on which the Court of Appeals for the District of Columbia Circuit in Laffey and the Court of Appeals for the Seventh Circuit in McKinnon likewise rely. In his article, Professor Leubsdorf clearly sets forth the difficulties that arise when contingency enhancements are based on the relative likelihood of success in particular cases, Leubsdorf, at 482-497 — an approach that he terms "the Lindy-Grinnell approach" because of two leading cases in the area, id., at 478-482. But Professor Leubsdorf begins his analysis with a recognition that some enhancement for contingency is necessary if attorneys are to receive the fair market value of their work. Id., at 480 ("A lawyer who both bears the risk of not being paid and provides legal services is not receiving the fair market value of his work if he is paid only for the second of these functions. If he is paid no more, competent counsel will be reluctant to accept fee award cases"). Thus, as the professor explains: "Although economic reasoning justifies a contingency bonus, it does not by itself explain the Lindy-Grinnell approach to calculating the size of the bonus" (first emphasis added). Ibid. Professor Leubsdorf subsequently offers alternative approaches for calculating and awarding contingency enhancements. Id., at 501-512.

[12] Thus, contrary to the implication in the Court's opinion, ante, at 717-718, the Court of Appeals in Wildman did not approve of contingency enhancements based on the likelihood of success in particular cases. Rather, the court vacated the multiplier granted by the District Court — which had held, without elaboration, that the contingent nature of the case and the quality of the attorney's work warranted an enhancement — and remanded the case for the District Court to determine whether an enhancement for contingency was warranted. 771 F.2d, at 613-614. Among the factors to be considered on remand were: what, if any, payment the attorneys would have received had the suit not been successful; what, if any, costs or expenses the attorneys would have incurred if the case had been lost; the extent to which the attorneys were required to compensate associates and to carry overhead expenses without assurance of compensation; and whether other attorneys refused to take the case because of the risk of nonpayment. Id., at 614. All these factors focus solely on the extent of contingency in payment and not on the likelihood of success based on the law or facts in the particular case.

[13] Although the Court errs in not viewing delay as an integral component of contingency, it is gratifying to note that the Court does "not suggest . . . that adjustments for delay are inconsistent with the typical fee-shifting statute." Ante, at 716.

[14] See, e. g., Brewer v. Southern Union Co., 607 F. Supp. 1511, 1532 (Colo. 1984) (small firm); Uzzell v. Friday, 618 F. Supp. 1222, 1227 (MDNC 1985) (solo practitioner). A case to which a substantial percentage of an attorney's law practice is devoted, see, e. g., Craik v. Minnesota State University Bd., 738 F.2d 348, 350-351 (CA8 1984), or in which additional personnel must be hired without assurance of compensation, may also be riskier than an ordinary contingent case.

[15] Economic risks might also be increased if a case foreclosed participation in otherwise available business because of potential conflicts of interest, or if the case was so unpopular as to risk loss of other business. See, e. g., Johnson v. Georgia Highway Express, Inc., 488 F. 2d, at 719 (one of the factors to consider is the " `undesirability' of the case" because of the effect it may have on obtaining other business) (emphasis omitted); York v. Alabama State Board of Education, 631 F. Supp. 78, 85 (MD Ala. 1986) (successful plaintiffs' civil rights lawyers often not hired for other types of sophisticated federal litigation). Although these circumstances probably will exist in comparatively few cases, attorneys should nonetheless be compensated if they have undertaken representation despite such risks.

[16] The cases in which an extra enhancement is granted for the low likelihood of success and for the importance of the result achieved will be sufficiently rare that the plurality's concerns lose much of their force. Moreover, those cases will tend to be the important test cases that should be encouraged through an award of attorney's fees.

[17] The justification for a contingency premium can be explained either as an inducement to persuade an attorney to invest his time in a matter that may be unproductive — even a lawyer who has all the hourly fee work that he can handle may be willing to accept a contingency if he concludes that the value of the potential premium justifies the risk of nonpayment — or as a means of providing a level of compensation that will offset the losses on other cases that fail. Either explanation is simply a reflection of how the private market compensates for contingency. See, e. g., Berger, Court Awarded Attorney's Fees: What is "Reasonable"?, 126 U. Pa. L. Rev. 281, 325 (1977).

[18] Both the District Court and the Court of Appeals relied heavily on the fact that respondent faced serious and persistent opposition in this case. See 581 F. Supp., at 1431; 762 F.2d 272, 281 (1985). The fact that an attorney faces strong opposition by a well-funded opponent should certainly be given significant weight by a court. This factor, however, ordinarily becomes relevant in the assessment of the reasonableness of the hours expended by the attorney in preparing and litigating the case. Like the novelty or difficulty of a case, a strong and persistent opponent usually demands a significant increase in the hours devoted to a case. In the few cases in which an extra enhancement is justified for significant legal risks faced at the outset of litigation, a court may take into account as well the perceived strength of the opposition at the outset. A court, however, should grant such extra enhancement only if it determines that the attorneys have not already been adequately compensated through the number of reasonable hours that constitute the lodestar.