This appeal raises important questions about the reasonable calculation of contingent counsel fees in class actions resulting in the creation of a common fund payable to plaintiffs. We hold that the proper measure of such fees in a common fund case is a percentage of the fund. In addition, on the facts before us, we conclude that the District Court did not abuse its discretion in calculating both the percentage to be used and the amount of the fund that resulted from the efforts of counsel. Therefore, we affirm.

This case is the endgame in a series of lawsuits attacking a Department of Health and Human Services ("HHS," the "Department," or the "Secretary") policy concerning reimbursement by HHS for photocopying costs incurred by hospitals in meeting requirements of the Medicare program. In the underlying case, plaintiff hospitals and HHS entered into a settlement agreement, approved by the District Court, in which HHS agreed to pay $27.8 million to the hospitals. Today's dispute is over what portion of that pie goes to plaintiffs' lawyers.

The District Court applied a "percentage-of-the-fund" method in determining a reasonable fee award for class counsel, and decided that the attorneys should receive twenty percent of the common fund produced by their efforts. However, the court awarded only $2 million in fees, reasoning that because the efforts of plaintiffs' attorneys had contributed only $10 million to the value of the settlement fund, the attorneys were entitled to twenty percent of only that amount.

Plaintiffs appeal, arguing that their attorneys were entitled to twenty percent of the entire $27.8 million fund, or about $5.6 million. The Secretary cross-appeals, arguing that the fees awarded were too high; that under governing circuit and Supreme Court precedent, class counsel's fee should be limited to the product of the hours reasonably spent by the attorneys and their reasonable hourly rates (the "lodestar"), resulting in a fee of no more than $619,000.

The Social Security Amendments of 1983 require all hospitals participating in the Medicare program to enter into agreements with Medicare "peer review organizations" ("PROs"), which review the quality and medical necessity of hospital services rendered to Medicare beneficiaries. 42 U.S.C. Sec. 1395cc(a)(1)(F) (1988). The statute directs HHS to reimburse hospitals for the costs of maintaining such agreements. Id.

From the outset, however, HHS has refused to reimburse some costs incurred by hospitals in maintaining PRO agreements. One regulation, known as the "photocopying rule," specifically prohibited reimbursing costs incurred by hospitals in furnishing photocopies of medical records to PROs for mandatory review. 42 C.F.R. Sec. 466.78(b)(2) (1985), as amended by 57 Fed.Reg. 47,779 (1992) (codified at 42 C.F.R. Sec. 466.78(b)(2) (1993)).

Several hospitals filed a series of lawsuits challenging the legality of the photocopying rule. In Burlington Memorial Hosp. v. Bowen, 644 F. Supp. 1020 (W.D.Wis.1986), a district court concluded that the photocopying rule was arbitrary and capricious because the Department's basis for denying reimbursement for photocopying costs mandated by the statute--that these costs were already reimbursed in other payments--was without any reasonable basis in the rulemaking record. The court accordingly enjoined the Secretary's enforcement of the rule. HHS appealed the Burlington decision, but later dropped the appeal and settled with the seventeen plaintiffs in that case, agreeing to pay ten cents per page for PRO photocopies.

Because HHS did not change its photocopying rule, hospitals filed at least six other cases seeking reimbursement. In Beverly Hosp. v. Bowen, Medicare & Medicaid Guide (CCH) p 36,738, 1987 WL 192217 (D.D.C.1987), a consolidation of four of those cases involving several hundred hospitals, the district [303 U.S.App.D.C. 97] court held the photocopying rule illegal. That court, however, denied plaintiffs' requested relief of prospective and retrospective reimbursement, holding that the rulemaking process would provide an adequate remedy. The hospitals appealed.

During the pendency of the Beverly appeal, HHS issued a notice of proposed rulemaking, proposing to reimburse hospitals for photocopies furnished to PROs at the rate of $.0498 per page. 53 Fed.Reg. 8,654 (1988). The proposed rule would have limited retrospective relief to the previous three years, and would have imposed a number of procedural and substantive restrictions on hospitals seeking relief.

We rejected HHS's proposed approach, and remanded the case to the District Court "with instructions to assure that the agency affords the hospitals a fair opportunity to recover photocopying costs they were made to pay due to the Secretary's unlawful regulation." Beverly Hosp. v. Bowen, 872 F.2d 483, 487 (D.C.Cir.1989) (per curiam). We defined the Department's "task ... [as] conscientiously to remold the situation to approximate fairly what it should have been initially." Id.

Beverly, by its terms, applied only to the few hundred hospitals which were plaintiffs in that case. Perceiving that HHS was likely to pursue a policy of "non-acquiescence," several hospitals brought this class action, seeking declaratory and injunctive relief and damages for all other hospitals participating in the Medicare program. Two principal issues separated the parties: how much HHS should pay per page and how to calculate retroactively the number of pages copied going back to 1984. Most hospitals kept records of the relevant copying expenses starting in 1987, but the parties could not agree how to calculate the number of copies made between 1984 and 1987. HHS initially insisted on documentation for all copies reimbursed, whereas the hospitals urged, and eventually the settlement embraced, an extrapolation formula for determining the number of copies in the earlier years.

Shortly before the trial was scheduled to begin, after eighteen months of litigation involving considerable discovery, the parties entered a settlement agreement requiring HHS to pay prospectively for relevant copying costs, and to pay $27.8 million into a settlement fund for distribution to the class for copies made between 1984 and 1991. Before approving the settlement agreement, the District Court ordered notice of the proposed settlement sent to class members. In a separate mailing, class members received the details of the petition for attorneys' fees, requesting twenty percent of the common fund. The fee petition notice advised class members of their right to file objections to any element of the proposed settlement or fee petition. None of the plaintiff hospitals objected to class counsel receiving twenty percent of the common fund; indeed, a number of hospitals, including several of the class members with the largest economic stake in the fund, submitted declarations affirmatively favoring that disposition.

The District Court, applying a percentage-of-the-fund methodology, granted plaintiffs' requested twenty percent share as a rate, but concluded that counsel could "claim credit only for enhancing the fund" by payment at roughly $.07 per page instead of approximately $.0498 per page "which was apparently on the table when negotiations opened." Swedish Hosp. Corp. v. Sullivan, Medicare & Medicaid Guide (CCH) p 39,730, at 28,743, 1991 WL 319154 (D.D.C.1991). As that difference generated a "fund for which ... counsel is responsible" of approximately $10 million, rather than the $28 million, the twenty percent rate produced a fee of $2 million.

The district judge based his assessment of the attorneys' contribution to the fund on a number of factors. He found that plaintiffs in this case were at least in part piggybacking on the success of the Beverly plaintiffs; that plaintiffs' attorneys never faced the risk of zero recovery because the central issue of the illegality of HHS's photocopying rule was resolved in Beverly; that because the government was the defendant, plaintiffs and their attorneys never faced a risk of nonpayment; and that "the government acquiesced early in the treatment of the dispute as a class action." Id. The court also observed that the class members "were not paupers, unable to pay counsel for their time in the [303 U.S.App.D.C. 98] remote event of no recovery beyond" the Secretary's initial offer. Finally, the court was of the opinion that plaintiffs' attorneys had exhibited no extraordinary legal skills in their representation.

Both parties appeal the fee award, plaintiffs arguing it is too small and HHS that it is too large.

This case raises two main questions. The first is whether the District Court erred in not utilizing the lodestar to determine the appropriate fee. Because the answer is "no," we must address a second question, whether the District Court abused its discretion in setting a twenty percent fee in this case, and in choosing to apply that percentage only to that part of the fund for which counsel was responsible. The answer to this question is "no" as well. We address both questions in turn below.

In general, each party to litigation in the United States bears its own attorneys' fees absent a specific fee-shifting statute. Over time, courts have fashioned several equitable exceptions to this "American rule." One of the earliest, and still most common, exceptions is the "common fund" doctrine typically applied in class actions like the present one. See Sprague v. Ticonic Nat'l Bank, 307 U.S. 161, 164, 59 S. Ct. 777, 779, 83 L. Ed. 1184 (1939) (fee award from fund generated is within "the historic equity jurisdiction of the federal courts"). That doctrine allows a party who creates, preserves, or increases the value of a fund in which others have an ownership interest to be reimbursed from that fund for litigation expenses incurred, including counsel fees. It is by now well established that "a litigant or lawyer who recovers a common fund for the benefit of persons other than himself or his client is entitled to a reasonable attorney's fee from the fund as a whole." Boeing Co. v. Van Gemert, 444 U.S. 472, 478, 100 S. Ct. 745, 749, 62 L. Ed. 2d 676 (1980). The underlying justification for attorney reimbursement from a common fund, as explained by the Supreme Court in three early cases, is that unless the costs of litigation are spread to the beneficiaries of the fund they will be unjustly enriched by the attorney's efforts. See Sprague, 307 U.S. at 166-67, 59 S.Ct. at 779-80; Central R.R. & Banking Co. of Georgia v. Pettus, 113 U.S. 116, 126-27, 5 S. Ct. 387, 392-93, 28 L. Ed. 915 (1885); Trustees v. Greenough, 105 U.S. 527, 532, 26 L. Ed. 1157 (1882).

When awarding attorneys' fees, federal courts have a duty to ensure that claims for attorneys' fees are reasonable. See Hensley v. Eckerhart, 461 U.S. 424, 433, 103 S. Ct. 1933, 1939, 76 L. Ed. 2d 40 (1983); Pettus, 113 U.S. at 127, 5 S.Ct. at 392-93. Special problems exist in assessing the reasonableness of fees in a class action suit since class members with low individual stakes in the outcome often do not file objections, and the defendant who contributed the fund will usually have no interest in how the fund is divided between the plaintiffs and class counsel.1

2. The Percentage-of-the-Fund Method for Calculating Fees

Historically, courts exercised considerable discretion and applied a reasonableness standard, focusing upon the particular circumstances of a case, in determining the amount of a common fund fee award. The percentage[303 U.S.App.D.C. 99] -of-the-fund method of calculating attorneys' fees in common fund cases was most common. See Pettus, 113 U.S. at 128, 5 S.Ct. at 393 (paying successful attorneys a percentage of the fund recovered for the class). See also Court Awarded Attorney Fees, Report of the Third Circuit Task Force, 108 F.R.D. 237, 242 (1986) ("Third Circuit Task Force Report ").

The application of a percentage-of-the-fund approach sometimes resulted in large fee awards, and in the 1970s several courts began a movement to alternative methods of calculating attorneys' fees. In 1973, the Third Circuit led the way in Lindy Bros. Builders, Inc. of Philadelphia v. American Radiator & Standard Sanitary Corp., 487 F.2d 161 (3d Cir.1973), instructing judges in that circuit to first compute the product of the reasonable hours expended and the reasonable hourly rate to arrive at the "lodestar." That amount could then be adjusted upward or downward, based upon additional factors such as the contingent nature of the case and the quality of the attorneys' work. This so-called "lodestar/multiplier" approach thus shifted the emphasis from a fair percentage of recovery to the value of the time expended by counsel.

A year later in a case involving a fee-shifting statute rather than a common fund, the Fifth Circuit identified twelve factors as relevant to determining a reasonable fee award: 1) the time and labor required; 2) the novelty and difficulty of the questions involved; 3) the skill requisite to perform the legal services properly; 4) the preclusion of other employment by the attorney due to acceptance of the case; 5) the customary fee; 6) whether the fee is fixed or contingent; 7) time limitations imposed by the client or other circumstances; 8) the amount involved and the results obtained; 9) the experience, reputation, and ability of the attorneys; 10) the "undesirability" of the case; 11) the nature and length of the professional relationship with the client; and 12) awards in similar cases. Johnson v. Georgia Highway Express, Inc., 488 F.2d 714 (5th Cir.1974), abrogated in part on other grounds, Blanchard v. Bergeron, 489 U.S. 87, 109 S. Ct. 939, 103 L. Ed. 2d 67 (1989).2

After the emergence of Lindy and Johnson, the federal courts experimented with combinations of the Lindy lodestar and Johnson twelve-factor approaches in both common fund and fee-shifting contexts. In the context of the fee-shifting statutes, the lodestar approach but without enhancement for the Johnson factors has emerged as the prevailing method of fee calculation. See, e.g., City of Burlington v. Dague, --- U.S. ----, 112 S. Ct. 2638, 120 L. Ed. 2d 449 (1992); King v. Palmer, 950 F.2d 771 (D.C.Cir.1991) (en banc), cert. denied, --- U.S. ----, 112 S. Ct. 3054, 120 L. Ed. 2d 920 (1992). However, at the same time, many courts and commentators have noted differences between the fee-shifting and common fund cases, raising questions as to whether the same methodology should apply. A Third Circuit task force appointed to compare the respective merits of the lodestar and percentage-of-the-fund approaches concluded that the lodestar technique, at least as encumbered with the Johnson factors, is a "cumbersome, enervating, and often surrealistic process of preparing and evaluating fee petitions that now plagues the Bench and Bar." Third Circuit Task Force Report, 108 F.R.D. at 255.

The task force enumerated nine deficiencies in the lodestar process: 1) it "increases the workload of an already overtaxed judicial system"; 2) the elements of the process "are insufficiently objective and produce results that are far from homogeneous"; 3) the process "creates a sense of mathematical precision that is unwarranted in terms of the realities of the practice of law"; 4) the process "is subject to manipulation by judges who prefer to calibrate fees in terms of percentages of the settlement fund or the amounts recovered by the plaintiffs or of an overall dollar amount"; 5) the process, although designed to curb abuses, has led to other abuses, such as "encouraging lawyers [303 U.S.App.D.C. 100] to expend excessive hours engag[ing] in duplicative and unjustified work, inflat[ing] their 'normal' billing rate[s], and includ[ing] fictitious hours"; 6) it "creates a disincentive for the early settlement of cases"; 7) it "does not provide the district court with enough flexibility to reward or deter lawyers so that desirable objectives, such as early settlement, will be fostered"; 8) the process "works to the particular disadvantage of the public interest bar" because, for example, the "lodestar" is set lower in civil rights cases than in securities and antitrust cases; and 9) despite the apparent simplicity of the lodestar approach, "considerable confusion and lack of predictability remain in its administration." Id. at 246-49.

The task force, though recommending the retention of a lodestar approach in statutory fee cases, concluded that in common fund cases the best determinant of the reasonable value of services rendered to the class by counsel is a percentage of the fund. Id. at 255.

The Eleventh Circuit has, after reviewing criticisms of the lodestar method and the findings of the Third Circuit task force, specifically established the percentage-of-the-fund, not the lodestar, approach as applicable in all common fund cases in that circuit. Camden I Condominium Ass'n, Inc. v. Dunkle, 946 F.2d 768, 774 (11th Cir.1991). Other cases also offer evidence of a trend toward percentage-of-the-fund calculations in common fund cases. See, e.g., Paul, Johnson, Alston & Hunt v. Graulty, 886 F.2d 268, 271 (9th Cir.1989) (setting attorney fee benchmark of 25% of common fund); Brown v. Phillips Petroleum Co., 838 F.2d 451 (10th Cir.) (holding fee award representing 16.5% of fund not a per se abuse of discretion), cert. denied, 488 U.S. 822, 109 S. Ct. 66, 102 L. Ed. 2d 43 (1988); In re Avon Prods., Inc. Sec. Litig., Fed.Sec.L.Rep. (CCH) p 97,061, 1992 WL 349768 (S.D.N.Y.1992) (awarding 30% fee); In re Workers' Compensation Ins. Antitrust Litig., 771 F. Supp. 284 (D.Minn.1991); In re SmithKline Beckman Corp. Sec. Litig., 751 F. Supp. 525 (E.D.Pa.1990); In re Activision Sec. Litig., 723 F. Supp. 1373, 1377-78 (N.D.Cal.1989) (adopting policy of awarding 30% of the fund "absent extraordinary circumstances that suggest reasons to lower or increase the percentage"); Mashburn v. National Healthcare, Inc., 684 F. Supp. 679 (M.D.Ala.1988); Pray v. Lockheed Aircraft Corp., 644 F. Supp. 1289, 1311 (D.D.C.1986); In re Warner Communications Sec. Litig., 618 F. Supp. 735, 749-50 (S.D.N.Y.1985), aff'd, 798 F.2d 35 (2d Cir.1986); In re Superior Beverage/Glass Container Consol. Pretrial, 133 F.R.D. 119 (N.D.Ill.1990); see generally Third Circuit Task Force Report, 108 F.R.D. at 247 n. 32 (listing cases).

Neither this Court nor the Supreme Court has recent authority clearly controlling the question of the appropriate fee award methodology in the common fund context. We have at times applied a lodestar approach in common fund cases. See, e.g., National Treasury Employees Union v. Nixon, 521 F.2d 317, 320-22 (D.C.Cir.1975). More recently, we engaged in both a lodestar/multiplier analysis and a percentage-of-the-fund calculation in the same case using the results of each to bolster the other. In Bebchick v. Washington Metro. Area Transit Comm'n, 805 F.2d 396, 406-07 (D.C. Cir.1986), we awarded a fee representing twenty-five percent of the common fund which, we noted, approximated the lodestar plus a sixty percent enhancement for the contingency of success and compensation, the high quality of representation, and the public benefit involved.3

The Supreme Court, while frequently dealing with the lodestar approach in the fee-shifting context, see, e.g., cases collected in Burlington v. Dade, --- U.S. ----, 112 S. Ct. 2638, 120 L. Ed. 2d 449 (1992), has not lately spoken to the common fund fee issue. However, [303 U.S.App.D.C. 101] the latest guidance from the High Court counsels the use of a percentage-of-the-fund methodology. InBlum v. Stenson, 465 U.S. 886, 104 S. Ct. 1541, 79 L. Ed. 2d 891 (1984), where the Court approved the use of the lodestar method in a statutory fee-shifting context, the Court distinguished the common fund cases stating: "[u]nlike the calculation of attorney's fees under the 'common fund doctrine,' where a reasonable fee is based on a percentage of the fund bestowed on the class, a reasonable fee under [the fee-shifting statute before the Court] reflects the amount of attorney time reasonably expended on the litigation." Id. at 900 n. 16, 104 S. Ct. at 1550 n. 16. Although this language is dicta, as Blum involved no common fund, it is entirely consistent with the Court's decision four years earlier in Boeing Co. v. Van Gemert, 444 U.S. 472, 100 S. Ct. 745, 62 L. Ed. 2d 676 (1980), in which the Court approved a fee award based on the percentage-of-the-fund method in a common fund case. More importantly, the Blum footnote makes it plain that that decision's approval of the lodestar method in the fee-shifting context was not intended to overrule prior common fund cases, such as Boeing.

HHS argues that Burlington and King, though involving fee-shifting statutes, supply reasoning applicable to common fund fee awards because both situations require the court, rather than the market, to determine the value of the lawyer's services. According to HHS, in a common fund case such as this, the only possible reason to pay counsel more than the lodestar is to compensate them for the risk that they would recover nothing. But, HHS argues, Burlington and King "prohibit paying an attorney's fee higher than the lodestar when such an enhancement is designed to compensate counsel for the risk that they would lose the case and thus receive no fees." HHS Br. at 22-23. The Secretary argues that because Burlington and King abandon prior methods of calculating the lodestar and then enhancing it for risk of loss in a common fund fee award, and because Burlington deals in detail with the theoretical and practical concerns of court-ordered fees higher than the lodestar in contingent situations, it, rather than the brief footnote in Blum, supplies the controlling law. HHS concludes that the hospitals' attorneys cannot receive more than the $619,000 they have already been paid, which constitutes the District Court's already generous calculation of the lodestar.

We disagree with the proposition that Burlington and King mandate an unenhanced lodestar approach in common fund cases. An important assumption underlying the Secretary's assertion that Burlington and King forbid the award of attorneys' fees beyond the lodestar--that fee-shifting cases are in all relevant respects similar to common fund cases--does not withstand scrutiny. In our judgment, Burlington and King do not govern common fund awards because of several important, and ultimately decisive, differences between the two types of cases.

First and most obviously, there is often no resulting fund in fee-shifting cases, so the alternative of using a percentage-of-the-fund method to calculate attorneys' fees is not necessarily available. As a leading treatise on attorney fee awards explains:

Fee awards authorized by statute are payable by a losing defendant whether or not there has been any monetary recovery for the named plaintiffs or for a class, in contrast to common fund fee awards that are payable of the fund recovered. Because statutory fees are payable to prevailing parties to encourage private enforcement of statutes and deter violations, and because the results obtained are often nonmonetary or modest recoveries, a formula for a reasonable statutory fee award based on a percentage of the recovery is not usually available to the courts.

Newberg, ATTORNEY FEE AWARDS Sec. 1.10, at 17 (1986) (emphasis added) (footnotes omitted).

Second, and perhaps more importantly, using the lodestar approach in common fund cases encourages significant elements of inefficiency. First, attorneys are given incentive to spend as many hours as possible, billable to a firm's most expensive attorneys. Second, there is a strong incentive against early settlement since attorneys will earn more the longer a litigation lasts. See In re Union Carbide Corp. Consumer Prods. Business [303 U.S.App.D.C. 102] Sec. Litig., 724 F. Supp. 160, 167 (S.D.N.Y.1989); In re Activision Sec. Litig., 723 F. Supp. 1373 (N.D.Cal.1989). Arguably, the lodestar method may encourage some inefficiency even in the fee-shifting context, but there the risk is not as great nor the alternative as readily available as in the common fund cases. So far as the degree of risk of inefficiency, attorneys in fee-shifting cases not involving a common fund know that the two most determinative factors affecting their fees after Burlington are winning, without which there is no award, and reasonableness of time expended and rate as determined in adversarial context. Working additional hours beyond those efficiently allocated to the case does not appreciably enhance the first yet risks unrewarded effort determined in the second.

In the common fund case, by contrast, victory is still the key factor, but, as in the present case, the monetary amount of the victory is often the true measure of success, and therefore it is most efficient that it influence the fee award. That is, in the common fund case, if a percentage-of-the-fund calculation controls, inefficiently expended hours only serve to reduce the per hour compensation of the attorney expending them. On the other hand, if we apply the lodestar method to the common fund case, then the attorney inefficiently expending an excess amount of time does stand to gain by that inefficiency if the awarding court does not ultimately recognize the inefficiency in the far-from-exact testing of the fee award hearing. The danger that the court will not recognize unreasonably expended hours is magnified by the fact that in the common fund case the only party having an adverse interest at the time of the award will be the attorney's own clients, often a diverse and scattered group with small individual stakes. The opposing party in the common fund case, unlike the loser in a fee-shifting case, stands to lose no more if the attorneys' fee award is greater and therefore cannot be relied upon to provide an adversarial approach to deleting unreasonable time entries.

Furthermore, a percentage-of-the-fund approach more accurately reflects the economics of litigation practice. The district court in Howes v. Atkins, 668 F. Supp. 1021 (E.D.Ky.1987), noted that "[p]laintiffs' litigation practice, given the uncertainties and hazards of litigation, must necessarily be result-oriented. It matters little to the class how much the attorney spends in time or money to reach a successful result." Id. at 1025 (internal quotation marks omitted). Making a similar point, Judge Posner has recently argued that a percentage-of-the-fund approach most closely approximates the manner in which attorneys are compensated in the marketplace for these types of cases. Writing for the court in In re Continental Illinois Sec. Litig., 962 F.2d 566, 572 (7th Cir.1992), he noted:

The judicial task might be simplified if the judge and the lawyers spent their efforts on finding out what the market in fact pays not for the individual hours but for the ensemble of services rendered in a case of this character.... The class counsel are entitled to the fee they would have received had they handled a similar suit on a contingent fee basis, with a similar outcome.

Obviously, a court setting a fee does not perfectly replicate the marketplace by assigning counsel a percentage of the common fund, but the device can approximate the market with reasonable accuracy. Mechanisms which may facilitate a judge in more closely approximating the market include: encouraging class counsel to enter into preliminary non-binding fee agreements with class members; surveying class members for their views about appropriate levels of compensation; and, as the Third Circuit task force suggests, considering the fee issue early in the litigation. Although the district judge in this case did not employ exactly those mechanisms, his approximation of the market appears well supported, especially given the notice to and response of the class.

Additionally, a percentage-of-the-fund approach is less demanding of scarce judicial resources than the lodestar method. The lodestar method makes considerable demands upon judicial resources since it can be exceptionally difficult for a court to review attorney billing information over the life of a complex litigation and make a determination [303 U.S.App.D.C. 103] about whether the time devoted to the litigation was necessary or reasonable. This Court has reiterated the Supreme Court's warning that " '[a] request for attorney's fees should not result in a second major litigation.' " Bebchick, supra, at 401 (quoting Hensley v. Eckerhart, 461 U.S. 424, 437, 103 S. Ct. 1933, 1941, 76 L. Ed. 2d 40 (1983)). It is much easier to calculate a percentage-of-the-fund fee than to review hourly billing practices over a long, complex litigation.

A related weakness in the lodestar approach is that it often results in a substantial delay in distribution of the common fund to the class. The lodestar procedure requires detailed involvement by the District Court, evaluating the reasonableness of expenditure of attorney time and effort, and making comparative inquiries on reasonable rates for those services. Given the complexity of many class action lawsuits, combined with the degree of detailed review required and considering the heavy workload of most district court judges, lodestar calculation is likely to cause significant delay between the creation of a common fund and remuneration of class counsel. In contrast, the application of a percentage-of-the-fund methodology is relatively straightforward and much less time consuming.

For similar reasons, a percentage-of-the-fund approach is less subjective than the lodestar approach; under the former, the court need not second-guess the judgment of counsel as to whether a task was reasonably undertaken or hours devoted to it reasonably expended. See Third Circuit Task Force Report, 108 F.R.D. at 246 ("The elements of the Lindy process are insufficiently objective and produce results that are far from homogeneous.").

HHS advances two further arguments in support of its conclusion that the attorneys' fees in this case should be limited to the lodestar, which we briefly address. First, HHS reminds this Court that it has "a high obligation to do equity," and urges that we should not grant a fee award that excessively diverts money from the common fund to pay for the lawyers. HHS asserts that the portion of the fund awarded to attorneys "could otherwise have been used by [the hospitals] for many far better purposes, such as to improve their service or their facilities, to provide medical care to the indigent, or to help them gain control over their costs." HHS also suggests that we have a special obligation to "do equity" because the government, and by extension taxpayers, are the source of the common fund.

We find this argument to be without merit for multiple reasons. First, there is no evidence on how the money reimbursed to the hospitals will be spent. To the extent such a consideration is even appropriate or relevant, there is no reason on the record before us to conclude the hospitals would use the money they recover from the common fund for purposes any "better" than class counsel would. Second, HHS has failed to cite any case law or other authority for the novel proposition that a common fund case such as this one should be treated differently than other common fund cases simply because it successfully challenged illegal government action. Third, we are not convinced that a percentage-of-the-fund methodology will necessarily, or even routinely, result in larger fees. Admittedly it does in this case, but the Secretary provides, and we can discern, no reason to believe that this will be the general rule. We adopt a percentage-of-the-fund methodology not because it will pay lawyers more, but primarily because it is more efficient, easier to administer, and more closely reflects the marketplace.

Fourth, HHS's argument can be made with equal validity (or invalidity) as to any amount of fee and in no readily discernable way distinguishes between this method and any other. We further note that the Secretary's stated concern for money getting to the hospitals rings particularly hollow here given the history of this litigation. HHS would have this Court ignore the fact that the government, left to its own devices, would have paid nothing to reimburse hospitals for their photocopying expenses. HHS would also have us ignore the lengths to which the Department has gone, through repeated litigation, to avoid its statutory obligations in this matter.

[303 U.S.App.D.C. 104] HHS's second argument is that this Court should assign no relevance to the fact that none of the plaintiff hospitals objected to their attorneys receiving twenty percent of the entire $27.8 million dollar fund. HHS asserts that the average economic stake in the fee issue for any hospital is only about $900, which is too small an amount to justify careful attention to the issue or the trouble of objecting to the proposed fee arrangement. Furthermore, HHS argues, there is no reason to believe that the hospitals who declared their support for the attorneys receiving twenty percent of the entire fund are a fair, representative sample of the opinions of the roughly 6,000 hospitals in the class.

We find this argument unconvincing as well. First of all, even if no significance is to be found in the hospitals' failure to object to the proposed fee arrangement, this does not at all explain why significance should not be attached to the fact that a large number of hospitals, in contravention of their economic interests, took the trouble to affirmatively support the proposed fee settlement. Second, many of the hospitals, or health care systems, with the largest economic stake in this litigation were among the approximately 450 class members expressly endorsing the requested twenty percent of the common fund as a reasonable fee award.

In sum, we join the Third Circuit Task Force and the Eleventh Circuit, among others, in concluding that a percentage-of-the-fund method is the appropriate mechanism for determining the attorney fees award in common fund cases. We now turn to the District Court's application of that methodology to this case.4

B. Application of Percentage-of-the-Fund Methodology in this Case

The hospitals argue the District Court appropriately employed a percentage-of-the-fund approach in determining the fee, but abused its discretion in applying that methodology because it based the fee award on a number of erroneous factual and legal conclusions. HHS argues that even if a percentage-of-the-fund methodology is appropriate, the trial court abused its discretion in granting twenty percent rather than a smaller percentage of the fund.

We consider three distinct issues: first, whether twenty percent of the fund represents a reasonable fee in this case; second, whether the District Court has the discretion to apply a percentage of the fund only to that part of a fund for which counsel is responsible; and third, whether the District Court reasonably concluded that in this case counsel was only responsible for $10 million of the value of the fund. We conclude the answer to each question is "yes."

In general, a trial court enjoys substantial discretion in making reasonable fee determinations. See Hensley v. Eckerhart, 461 U.S. 424, 437, 103 S. Ct. 1933, 1941, 76 L. Ed. 2d 40 (1983). That discretion is premised upon the district court's "superior understanding of the litigation and the desirability of avoiding frequent appellate review of what essentially are factual matters." Id.; see also Pierce v. Underwood, 487 U.S. 552, 563, 108 S. Ct. 2541, 2549, 101 L. Ed. 2d 490 (1988) (holding that abuse-of-discretion standard governs appellate review of the trial court's attorney fee determination).

[303 U.S.App.D.C. 105] The hospitals are of the view that the normal degree of deference is not due here because "this case was unusual in that it required no meaningful substantive participation by the District Court until the stipulation of settlement was offered for approval...." Appellants' Br. at 20. According to appellants, "the District Court simply did not have the opportunity to develop the familiarity with the issues or the positions of the parties that would have resulted from disputatious discovery or a full-blown trial." Id.

We find this argument to be without merit. We can hardly imagine a more futile and foolhardy endeavor than struggling to review each district court's degree of familiarity with a case to decide how much deference to grant its findings and conclusions. Furthermore, it is far from unusual for a case to settle before trial--in fact, it may be the usual expectation. And, even if we accepted this argument as a general proposition, it would not apply here as Judge Oberdorfer had a high degree of familiarity with this case. Not only was he responsible for this case throughout the eighteen months of pretrial preparation, he was also the trial judge in the Beverly case, in which the primary legal issue underlying this litigation was decided.

We are of the opinion that the District Court acted within its discretion in setting the percentage of the fund at twenty percent. The twenty percent figure is well within the range of reasonable fees in common fund cases. As suggested in our discussion above, a review of similar cases reveals that a majority of common fund class action fee awards fall between twenty and thirty percent. Also, the already noted supportive response of the class members to the attorneys' fee notice, though not decisive, is relevant to the District Court's reasonableness determination.

We also hold that the court was within its discretion in basing its fee calculation only on that part of the fund for which counsel was responsible. The District Court's conclusion that, to a considerable extent, this case rode "piggyback" on the Beverly case is entitled to deference, especially as Judge Oberdorfer was the district court judge in both cases. The hospitals argue that Beverly, by its terms, applied only to the plaintiffs in that case, so the issue of whether HHS was obliged to pay the copying expenses of the hospital plaintiffs in this case was still an open question. This argument is unpersuasive as Beverly represented binding precedent in this Circuit when the plaintiffs in this case filed their complaint.

Finally, the hospitals' attorneys argue that the conclusion that they were responsible for only $10 million in added value to the fund was based on misleading information supplied by HHS and misunderstandings of the record. We disagree. The District Court concluded that the method for calculating the number of pages to be reimbursed was not a matter of much dispute. The court also concluded that at the time this lawsuit was brought, HHS had proposed a regulation that would have paid $.0498 per page, whereas the eventual settlement agreement provided payment of $.07 per page. Based on the difference between these two figures, the District Court concluded that class counsel contributed about $10 million to the value of the common fund. We find substantial evidence in the record to support these conclusions and do not think the court abused its discretion in reaching them.

For the reasons stated herein, we conclude that percentage-of-the-fund is the proper method for calculating fees in a common fund case. We also hold that in this case the District Court did not abuse its discretion in calculating either the percentage to be used or the amount of the fund that resulted from the efforts of counsel. Accordingly, we affirm the District Court's judgment in all respects.

So ordered.

Circuit Judge D.H. GINSBURG, concurring in part and dissenting in part:

I join in the opinion of the court insofar as it upholds the district court's decision to reduce the fee award from $5.6 million to $2 [303 U.S.App.D.C. 106] million. I dissent, however, insofar as the court relies upon the percentage-of-the-fund approach as the only permissible measure of a reasonable fee in a common fund case.

The authority controlling a common fund case in this circuit is Bebchick v. W.M.A.T.A., 805 F.2d 396, 406-07 (D.C.Cir.1986). Under that precedent, the lodestar is the starting point for calculating a fee award and the percentage-of-the-fund it represents is merely a secondary check upon the reasonableness of the resulting award.

Neither this court nor the Supreme Court has ever in a common case cast doubt upon the continuing vitality of this approach. Cf. Blum v. Stenson, 465 U.S. 886, 104 S. Ct. 1541, 79 L. Ed. 2d 891 (1984) (dictum in a statutory fee shifting case). Both City of Burlington v. Dague, --- U.S. ----, 112 S. Ct. 2638, 120 L. Ed. 2d 449 (1992), and King v. Palmer, 950 F.2d 771 (D.C.Cir.1991), were brought under fee-shifting statutes. According to the Supreme Court, the plaintiffs were not entitled to a risk enhancement principally because the statute authorized the award of a fee only to a "prevailing part[y]." Because that constraint is absent in a non-statutory common fund case, we find the question of enhancement in such a case precisely where Bebchick left it.

Far from being indicated by the Supreme Court's decision in City of Burlington, as the court implies, the percentage-of-the-fund approach seems actually to be at odds with that decision. The Supreme Court there clearly assumed that the lodestar was the appropriate starting point in calculating a reasonable fee. Contrary to my colleague's implication, however, that assumption was not compelled; many cases brought under fee-shifting statutes produce a common fund that could provide the measure of a reasonable fee. See e.g. Kientzy v. McDonnell Douglas Co., 990 F.2d 1051 (8th Cir.1993) ($600,000 award in sex discrimination case); King v. Palmer, 950 F.2d 771 (award of back pay). Were a percentage-of-the-fund approach so clearly preferable, it could be applied to such cases, with the lodestar reserved for instances in which there is no fund from which to calculate a reasonable award.

I realize that the percentage-of-the-fund approach is appealingly simple to administer. Whatever the problems associated with applying the lodestar, however, see supra Ct. Op. at 1267, 1268-69, they are no greater in a common fund case than in a fee-shifting case. Moreover, reliance upon the percentage-of-the-fund approach without any regard for the lodestar may produce excessively high awards and thus encourage even relatively non-meritorious cases to be brought. This is a case in point: since the fee award here comes to about 3.3 times what it would be using the lodestar, the case would have been worth bringing (i.e., would have given counsel an ex ante probability of earning the lodestar rate) even if the plaintiff had only a 30% chance of success.

In order to ensure that we do not encourage the litigation of less meritorious claims, any enhancement above the lodestar should be limited to what is reasonable in the particular case--again as illustrated in Bebchick. The district court ought to be required to give some special reason for authorizing any amount more than twice the lodestar--which is to say, any amount that makes it remunerative for lawyers to bring cases with less than a 50% chance of success. Perversely, however, the approach adopted by the court today will encourage counsel to bring just such relatively non-meritorious claims. The potential for recovering a fee that is all out of proportion to the resources expended on the litigation means that counsel need win only occasionally in order to make a profit. The result will be to burden with crap-shoot cases a system already overburdened with close cases.

In this case the district court offered no justification whatsoever for an award so disproportionate to (3.3x) the lodestar. I would remand the matter for the district court either to explain or to revise its award of $2 million for legal work that, if billed at counsels' hourly rate, would have fetched only $619,000.