07-5580-cv Nichole McDaniel, et al. v. County of Schenectady, et al.
1 UNITED STATES COURT OF APPEALS 2 FOR THE SECOND CIRCUIT 3 4 August Term 2008 5 6 7 Argued: May 13, 2009 Decided: February 16, 2010 8 9 10 Docket No. 07-5580-cv 11 12 _____________________________________ 13 14 NICHOLE MARIE MCDANIEL, Individually and on behalf of a class of others similarly 15 situated, LESSIE LEE DAVIES, Individually and on behalf of a class of others similarly situated, 16 17 Plaintiffs-Appellants, 18 19 -v.- 20 21 COUNTY OF SCHENECTADY, HARRY BUFFARDI, Both individually and in his official 22 capacity as Sheriff of the County of Schenectady, GORDON POLLARD, Both individually and as 23 Undersheriff of the County of Schenectady, ROBERT ELWELL SR., Both Individually and as 24 Major in the Schenectady County Sheriff’s Department, 25 26 Defendants-Appellees. 27 _____________________________________ 28 29 30 Before: WALKER, LIVINGSTON, Circuit Judges, KAPLAN,* District Judge. 31 32 Plaintiffs-Appellants appeal from an order of the United States District Court for the Northern
33 District of New York (Sharpe, J.), approving the settlement of a class action arising from alleged
34 violations of their constitutional rights, but awarding less than the requested fee to their attorneys.
* The Honorable Lewis A. Kaplan, District Judge of the United States District Court for the Southern District of New York, sitting by designation. 1 We conclude that the district court did not abuse its discretion by declining to award attorneys’ fees
2 using a percentage-of-fund approach, in the extent of its reliance on our decision in Arbor Hill
3 Concerned Citizens Neighborhood Association v. County of Albany, 493 F.3d 110 (2d Cir. 2007),
4 superseded by 522 F.3d 182 (2d Cir. 2008), or in its application of the reasonableness factors set
5 forth in Goldberger v. Integrated Resources, Inc., 209 F.3d 43 (2d Cir. 2000).
6 Affirmed.
7 ELMER ROBERT KEACH, III, Law Office of Elmer Robert 8 Keach III, PC, Amsterdam, New York, JASON J. ROZGER, 9 BRUCE E. MENKEN, Beranbaum, Menken, Ben-Asher & 10 Bierman LLP, New York, NY, JONATHAN WATSON 11 CUNEO, Cuneo, Gilbert & LaDuca, Washington D.C., 12 GARY E. MASON, The Mason Law Firm, Washington D.C., 13 for Plaintiffs-Appellants. 14 15 WILLIAM G. GREAGAN, Goldberg Segalla, LLP, Albany, 16 NY, for Defendants-Appellees.
17 PHILIP K. HOWARD, JENNIFER O. FARINA, Covington 18 & Burling LLP, New York, NY, as Amicus Curiae in support 19 of Defendants-Appellees. 20
21 LIVINGSTON, Circuit Judge:
22 Plaintiffs-Appellants appeal from an order of the United States District Court for the Northern
23 District of New York (Sharpe, J.), approving the settlement of a class action arising from alleged
24 violations of the their constitutional rights, but awarding less than the requested fee to their attorneys
25 from the common fund established by the settlement. Rather than base its attorneys’ fees calculation
26 on a percentage of the fund, the district court elected to calculate fees using a variant of the lodestar
27 method described by this Court in Arbor Hill Concerned Citizens Neighborhood Association v.
2 1 County of Albany, 493 F.3d 110 (2d Cir. 2007), superseded by 522 F.3d 182 (2d Cir. 2008).
2 Considering the factors set forth in Goldberger v. Integrated Resources, Inc., 209 F.3d 43 (2d Cir.
3 2000), the district court concluded that counsel were adequately compensated by an award of fees
4 that reflected their normal hourly rates. Because we conclude that the district court did not abuse
5 its discretion in making these determinations, we affirm.
6 BACKGROUND
7 On October 19, 2001, this Court issued its decision in Shain v. Ellison, 273 F.3d 56 (2d Cir.
8 2001), in which it determined, relying on its previous decisions in Wachtler v. County of Herkimer,
9 35 F.3d 77 (2d Cir. 1994), Walsh v. Franco, 849 F.2d 66 (2d Cir. 1988), and Weber v. Dell, 804 F.2d
10 796 (2d Cir. 1986), that it was clearly established that a corrections officer may not perform a strip
11 search of a pre-trial detainee charged only with a misdemeanor absent an individualized, reasonable
12 suspicion that the detainee possesses contraband or weapons. See Shain, 273 F.3d at 59. In the
13 wake of this decision, a number of lawsuits challenging strip search policies were commenced
14 throughout New York State. See, e.g., Complaint, McBean v. City of New York, No. 02-cv-5426
15 (S.D.N.Y. July 15, 2002); Complaint, Kelsey v. Schoharie County, New York, No. 1:04-cv-00299
16 (N.D.N.Y. Mar. 19, 2004); Complaint, Pritchard v. County of Erie, No. 04 cv 0534 (W.D.N.Y. July
17 21, 2004).
18 On June 29, 2004, Appellants Nichole Marie McDaniel and Lessie Lee Davies filed a
19 complaint in the United States District Court for the Northern District of New York on behalf of
20 themselves and others similarly situated, asserting that the Schenectady County Sheriff’s Department
21 maintained a policy, implemented by senior officers including Appellees Harry Buffardi, Gordon
3 1 Pollard, and Robert Elwell, of strip-searching all individuals who were incarcerated at the
2 Schenectady County Jail and placed in jail clothing, regardless of the crime with which they were
3 charged. The complaint sought compensatory and punitive damages, as well as declaratory and
4 injunctive relief. The parties vigorously litigated this action for a period of more than three years,
5 with various attorneys for the plaintiff class spending more than 1000 hours working on the case.
6 Throughout the course of the litigation, Appellees maintained that the class members were
7 merely “required to change into jail uniforms in the presence of a corrections officer of the same sex”
8 and that the Schenectady County Jail had no formal policy of strip searching all detainees. Ultimately,
9 however, they agreed to the terms of a settlement, pursuant to which Appellees agreed to substantial
10 injunctive relief and the creation of a settlement fund totaling $2.5 million. The settlement agreement,
11 signed on July 31, 2006, indicated that Appellants’ counsel would petition the court for an award of
12 attorneys’ fees “in the amount not to exceed 26%” of the total settlement fund, and also provided for
13 the separate reimbursement of administrative expenses.
14 During the litigation of this case, counsel for Appellants also acted as counsel for the plaintiffs
15 in two other actions – each initiated prior to the filing of the complaint in this case – alleging that
16 other counties in New York maintained impermissible strip search policies. See Complaint, Kahler
17 v. Rensselaer County, No. 1:03-cv-1324 (N.D.N.Y. Oct. 31, 2003); Complaint, Mariott v. County
18 of Montgomery, No. 5:03-cv-00531 (N.D.N.Y. Apr. 29, 2003). In Kahler, the parties reached a
19 settlement in March 2004, about three months prior to the initiation of the instant suit, which the
20 district court ultimately approved in substantial part. See Motion to Certify Class and for Preliminary
21 Approval of Class Action Settlement, Kahler v. Rensselaer County, No. 1:03-cv-01324 (N.D.N.Y.
4 1 Mar. 18, 2004); Order, Kahler v. Rensselaer Country, No. 1:03-cv-01324 (N.D.N.Y. Sept. 23,
2 2004). Counsel in Kahler received attorneys’ fees of approximately $442,700 out of a total
3 settlement fund of $2.7 million, about 16% of the fund.1 In Mariott, the district court granted the
4 plaintiffs’ motion for partial summary judgment in April 2006, concluding that Montgomery County’s
5 jail policies included an unconstitutional strip search for which the county was liable (although
6 reaching no decision as to the scope of damages) and that plaintiffs’ counsel, as counsel to a
7 prevailing party in an action brought under 42 U.S.C. § 1983, were entitled to a statutory award of
8 interim attorneys' fees. Marriott v. County of Montgomery, 426 F. Supp. 2d 1, 6-12 (N.D.N.Y.
9 2006). The parties thereafter reached a settlement, which received final approval in May 2007,
10 resulting in an attorneys' fees award of $600,000 out of a total settlement fund of $2 million, or 30%
11 of the fund. See Motion to Approve Consent Judgment (Exhibit A, Settlement Agreement), Marriott
12 v. County of Montgomery, No. 5:03-cv-00531 (N.D.N.Y. Apr. 13, 2007); Order Granting Final
13 Approval of Class Action Settlement and Judgment, Marriott v. County of Montgomery, No.
14 5:03-cv-00531 (N.D.N.Y. May 15, 2007).
15 On September 5, 2007, the district court in this case conducted a final fairness hearing
16 regarding the proposed settlement agreement, during which it issued an oral decision regarding the
17 acceptability of the agreement. In determining the proper fees to be awarded to counsel from the
18 common fund, the court noted its duty to act as “a guardian of the rights of absent class members”
19 and the resulting need to “approach fee awards with an eye to moderation.” Oral Decision at 9.
1 Because unclaimed portions of the fund reverted to the defendant, attorneys’ fees as a percentage of the fund actually paid to plaintiffs were ultimately higher. See Nilsen v. York County, 400 F. Supp. 2d 266, 287 (D. Me. 2005) (discussing Kahler).
5 1 Further observing that the calculation of a reasonable fee was committed to its sound discretion, the
2 court indicated that “[i]n the past, both the lodestar and the percentage of fund methods have been
3 available to district judges in calculating attorneys’ fees in common fund cases,” and provided a brief
4 explanation of each approach. Id. at 9-10. As the court noted, under the lodestar method:
5 [T]he district court scrutinizes the fee petition to ascertain the number of hours 6 reasonably billed to the class and then multiplies that figure by an appropriate hourly 7 rate[.] [O]nce that initial computation has been made, the district court may, in its 8 discretion, increase the lodestar by applying a multiplier based on other less objective 9 factors such as the risk of litigation and the performance of the attorneys. 10 11 Id. at 9. The court stated that the percentage method, by contrast, involves setting “some percentage
12 of the recovery as the fee,” and recognized that, whether using the percentage method or lodestar
13 method of calculating attorneys’ fees, a court is guided by the factors articulated by this Circuit in
14 Goldberger in determining a “reasonable common fund fee.” Id. at 9-10; see Goldberger 209 F.3d
15 at 47, 50. Earlier in the fairness hearing, the court had noted that the requested amount of 26% of
16 the total settlement fund, or $650,000, was equivalent to a lodestar multiplier of 1.98 to 2.24.
17 The district court then embarked upon a discussion of this Court's decision in Arbor Hill
18 Concerned Citizens Neighborhood Association v. County of Albany, 493 F.3d 110 (2d Cir. 2007),
19 superseded by 522 F.3d 182 (2d Cir. 2008), quoting our observation that “the meaning of the term
20 ‘lodestar’ has shifted over time and its value as a metaphor has deteriorated to the point of
21 unhelpfulness.” Oral Decision at 10 (quoting Arbor Hill, 493 F.3d at 117). In place of the traditional
22 lodestar method, the district court characterized Arbor Hill as holding that:
23 The better course [and] the one most consistent with attorneys’ fees jurisprudence is 24 for the district court, in exercising its considerable discretion, to bear in mind all of 25 the case specific variables that this court and other courts have identified as relevant 26 to the reasonableness of attorneys’ fees in setting a reasonable hourly rate.
6 1 Accordingly, the reasonable hourly rate is the rate a paying client would be willing to 2 pay. 3 4 Id. at 10. The court indicated that the factors relevant to any such determination were those
5 described in Johnson v. Georgia Highway Express, Inc., 488 F.2d 714 (5th Cir. 1974), abrogated
6 on other grounds by Blanchard v. Bergeron, 489 U.S. 87, 92-93 (1989), as well as the underlying
7 intuitions that “a reasonable paying client wishes to spend the minimum necessary to litigate the case
8 effectively” and that a paying client's negotiating position may be strengthened by the attorneys’
9 “desire to obtain the reputational benefits that might accrue from being associated with the case.”
10 Oral Decision at 11. Finally, citing this Court’s decision in Porzig v. Dresdner, Kleinwort, Benson,
11 North America LLC, 497 F.3d 133 (2d Cir. 2007), the district court characterized Arbor Hill as
12 requiring a court to “determin[e] the reasonable hourly rate for each attorney and the reasonable
13 number of hours expended and multiply[] the two figures together to obtain the presumptively
14 reasonable award.” Oral Decision at 12.
15 Drawing upon “the factors . . . outlined,” the District Court then determined that application
16 of the percentage method would be inappropriate, choosing instead “to determine what amount [was]
17 reasonable using the factors articulated in Goldberger and the most recent Arbor Hill decision.” Id.
18 at 12. Regarding the first Goldberger factor, the time and labor expended by counsel in the course
19 of prosecuting the case, the district court recognized that “class counsel put substantial time and
20 effort into this case,” but also concluded that this time and effort was sufficiently compensated by
21 awarding fees determined by the attorneys’ ordinary hourly rates. Id. at 12-13. Moreover, the court
22 indicated that a multiplier was inappropriate in light of the fact that counsel had previously worked
23 on two similar cases in the same judicial district, in which “some of the ground work for this litigation
7 1 was already established.” Id. at 13. As to the second factor, the complexity of the case, the district
2 court determined that the case at bar was “not particularly complex” as it was “an ordinary civil rights
3 case in which liability appear[ed] reasonably certain.” Id. With regard to the third factor, the risks
4 involved in the particular litigation, it found that “liability was reasonably certain” and that although
5 “there was risk on the issue of damages,” that risk was of little importance to counsel, who would
6 have been statutorily entitled to attorneys' fees as the prevailing party in a § 1983 action, even if only
7 nominal damages were awarded. Id. at 13-14. The court further noted that, as to the fourth factor,
8 the quality of representation, counsel “did a fine job,” but were adequately compensated by the
9 ordinary hourly rate counsel charged. Id. at 14. As to the fifth factor, the portion of the overall
10 settlement allocated to attorneys’ fees, the percentage arising from its use of the Arbor Hill method
11 (roughly 13%) was “squarely within the range of reasonable percentages in such a case” as the instant
12 one. Id. With regard to the last Goldberger factor, public policy, the district court took note of “the
13 compelling public policy reasons for keeping an eye on attorneys' fees in class action cases.” Id. at
14 15. Finally, the court briefly noted that Arbor Hill also required a district court to consider whether
15 the burden of litigating the case had precluded an attorney from accepting other gainful employment.
16 After this discussion, the district court repeated its decision not to employ the percentage
17 method or apply any type of multiplier to the fee calculated by multiplying the attorneys’ hourly rates
18 by the number of hours worked. At the same time, the court observed that it was awarding attorneys’
19 fees based upon the rates ordinarily billed by class counsel, even to the extent that those rates were
20 substantially higher than the prevailing rate in the Northern District of New York. This method
21 yielded attorneys' fees in the amount of $344,795. The district court also allowed the reimbursement
8 1 of $9,001.50 in litigation expenses and $107,000 in administrative expenses. The total award thus
2 amounted to $460,796.50.
3 After counsel for Appellants objected to the reduction of their fees, the district court clarified
4 its reliance upon Arbor Hill in the following fashion:
5 What I'm saying about Arbor Hill is this, I'm simply dispensing with the classic 6 definition of the lodestar method. If that were to be applied as the way it was 7 consistently applied, it would be irrelevant where an attorney was from, it would be 8 what the prevailing rate is for attorneys in this area. That's the aspect of Arbor Hill 9 that I'm dispensing with when I say that's no longer the standard of what constitutes 10 a reasonable attorney's fee. That is as substantial an explanation why some of the fees 11 as calculated in the fee schedule submitted by class counsel, they would have 12 exceeded that traditional rate in the Northern District of New York. I've not 13 discounted them as a result of that. That's the impetus of the Arbor Hill decision.
14 Oral Decision at 26-27.
15 On November 5, 2007, the district court issued a final decision and order approving the
16 settlement and explicitly incorporating the content of its oral decision, including the award of
17 attorneys’ fees. The court slightly increased the total amount to be paid to counsel, granting
18 attorneys' fees in the amount of $343,744.50, reimbursement of litigation expenses in the amount of
19 $10,053.31, and administrative expenses in the amount of $107,233.40, resulting in a final award of
20 $461,031.21.
21 This appeal followed.
22 DISCUSSION
23 We review a district court’s award of attorney’s fees for abuse of discretion, Goldberger v.
24 Integrated Res., Inc., 209 F.3d 43, 47-48 (2d Cir. 2000), “which occurs ‘when (1) [the court’s]
25 decision rests on an error of law (such as application of the wrong legal principle) or clearly
9 1 erroneous factual finding, or (2) its decision – though not necessarily the product of a legal error or
2 a clearly erroneous factual finding – cannot be located within the range of permissible decisions.’”
3 Kickham Hanley P.C. v. Kodak Ret. Income Plan, 558 F.3d 204, 209 (2d Cir. 2009) (quoting
4 Vincenty v. Bloomberg, 476 F.3d 74, 83 (2d Cir. 2007)). “‘[A]buse of discretion’ – already one of
5 the most deferential standards of review – takes on special significance when reviewing fee decisions”
6 because “‘the district court, which is intimately familiar with the nuances of the case, is in a far better
7 position to make [such] decisions than is an appellate court, which must work from a cold record.’”
8 Goldberger, 209 F.3d at 47-48 (quoting In re Bolar Pharm. Co. Sec. Litig., 966 F.2d 731, 732 (2d
9 Cir. 1992) (per curiam)).
10 In challenging the district court’s attorneys’ fee determinations, Appellants make several
11 arguments. First, they contend that the district court committed legal error by failing to calculate
12 attorneys’ fees using the percentage method, suggesting “[t]he percentage method is the presumptive,
13 preferred method” in this Circuit. Appellants’ Br. at 9. Second, Appellants suggest that the district
14 court improperly relied on Arbor Hill to justify its decision not to use the percentage method, since
15 that case and others cited by the court involved statutory fee shifting rather than a common fund.
16 Third, they assert that, whether or not the district court was within its discretion in employing a
17 modified-lodestar method to calculate attorneys’ fees, it committed errors in its application of the
18 Goldberger factors, including the use of an unprecedented factor to penalize counsel for their
19 experience in litigating similar cases. We consider each of these arguments in turn.
20 I. The District Court’s Decision Not to Employ a Percentage-of-Fund Method
21 Appellants contend that the district court failed to consider basing its calculation of attorneys’
10 1 fees on a percentage of the fund secured by counsel, that it flouted a “clear trend in the law towards
2 [using] the percentage method,” Appellants’ Br. at 11, and even that it defied an instruction by the
3 Supreme Court in Blum v. Stenson requiring the use of the percentage method in common fund cases,
4 see 465 U.S. 886, 900 n.16 (1984) (observing that “[u]nlike the calculation of attorney’s fees under
5 the ‘common fund doctrine,’ where a reasonable fee is based on a percentage of the fund bestowed
6 on the class, a reasonable fee under [42 U.S.C.] § 1988 reflects the amount of attorney time
7 reasonably expended in the litigation”).
8 Appellants’ assertion that the district court “refused to consider using the percentage method
9 at all,” Appellants’ Br. at 15, is factually incorrect. The court described both the lodestar method and
10 the percentage method of calculating attorneys’ fees in its bench decision before indicating that it
11 would apply a variant of the former.2 In its written decision and order approving the settlement, the
12 court reiterated its decision to use neither a lodestar multiplier nor the percentage method in
13 calculating the fee award. Thus, the “assertion that the district court erroneously refused even to
14 consider using the percentage approach is squarely contradicted by the record. . . . [T]he court clearly
15 understood that it could have awarded a percentage fee in this case. It simply chose not to do so.”
16 Goldberger, 209 F.3d at 50-51.
17 Although we have acknowledged that “the trend in this Circuit is toward the percentage
18 method,” it remains the law in this Circuit that courts “may award attorneys’ fees in common fund
2 While the Arbor Hill panel indicated its preference for abandonment of the term “lodestar” altogether, the approach adopted in that case is nonetheless a derivative of the lodestar method. To emphasize the similar role of both the lodestar method and Arbor Hill’s method as alternatives to the percentage-of-fund approach, we refer to Arbor Hill’s approach as a “variant” or “modified” lodestar method throughout this opinion.
11 1 cases under either the ‘lodestar’ method or the ‘percentage of the fund’ method.” Wal-Mart Stores,
2 Inc. v. VISA U.S.A., Inc., 396 F.3d 96, 121 (2d Cir. 2005); see also Cent. States Se. v. Merck-Medco,
3 504 F.3d 229, 249 (2d Cir. 2007) (“In common fund cases, courts typically use either the lodestar
4 method or the percentage method to compute attorneys’ fees.”); Masters v. Wilhelmina Model
5 Agency, Inc., 473 F.3d 423, 436 (2d Cir. 2007) (“Of course, courts may continue to use the lodestar
6 approach alone in calculating attorneys fees in common fund cases.”). And this Court has never
7 treated the Supreme Court’s dicta in Blum as demanding a single method for calculating attorneys’
8 fees in common fund cases or otherwise foreclosing the use of a lodestar approach. See Goldberger,
9 209 F.3d at 49 (observing that “Blum indicates that ‘the percentage-of-the-fund method is a viable
10 alternative’” (quoting Savoie v. Merchants Bank, 166 F.3d 456, 460 (2d Cir. 1999)), and noting that
11 six other circuits, post-Blum, have given district courts the option of choosing either the percentage
12 or the lodestar method).
13 Appellants nonetheless urge this Court to adopt the percentage method as the presumptive
14 approach to fee awards or to abandon the lodestar approach altogether, noting that several other
15 circuits have done so. This court has considered the advantages and disadvantages of the lodestar
16 and percentage methods previously, see Goldberger, 209 F.3d 47-53, and we need only address them
17 briefly here.3
18 The lodestar method is not perfect. It creates an incentive for attorneys to bill as many hours
19 as possible, to do unnecessary work, and for these reasons also can create a disincentive to early
3 In addition to the cases and other sources cited infra, see In re Auction Houses Antitrust Litigation, 197 F.R.D. 71 (S.D.N.Y. 2000), for an expanded discussion of the disadvantages associated with the lodestar and percentage-of-fund approaches to common fund fee awards.
12 1 settlement. See id. at 48-49 (citing Savoie, 166 F.3d at 460-61). Under certain conditions, moreover,
2 lodestar awards can create the near opposite incentive, encouraging attorneys to settle before trial
3 even when it is not in their clients’ best interest. While under the lodestar method lawyers share the
4 “downside” risk of trial (i.e., the possibility of an adverse judgment, and hence no fee), they do not
5 share in the potential economic “upside” (i.e., fees as a percentage of a large common fund),
6 especially since trial requires comparatively fewer hours than the process of trial preparation. See
7 Janet Cooper Alexander, Do the Merits Matter? A Study of Settlements in Securities Class Actions,
8 43 STAN . L. REV . 497, 543 (1991); John C. Coffee, Jr., Understanding the Plaintiff’s Attorney: the
9 Implications of Economic Theory for Private Enforcement of Law through Class and Derivative
10 Actions, 86 COLUM . L. REV . 669, 717 (1986) (hereinafter Coffee, Understanding the Plaintiff’s
11 Attorney). Although the district court is charged with ensuring the fairness of a proposed settlement,
12 including any lodestar-based attorneys’ fee award, this task is often challenging in common fund
13 cases, especially because – since the attorneys’ fees are drawn from a common fund rather than being
14 paid separately by the defendants – there is little incentive for the defendants to contest the size of
15 the fee. To the contrary, plaintiffs’ and defendants’ lawyers share an interest in the approval of an
16 agreed upon settlement. Goldberger, 209 F.3d at 52-53. As a result, the district judge “los[es] the
17 benefit of an adversarial process, which may . . . inform[] and sharpen[] the judicial inquiry.” Doe
18 v. C.I.A., 576 F.3d 95, 107 (2d Cir. 2009); see also Baker v. Carr, 369 U.S. 186, 204 (1962);
19 Goldberger, 209 F.3d at 53 (“It is not without significance that when [lead counsel for plaintiffs on
20 appeal] stood up at oral argument to petition for a bigger slice of his clients’ recovery, no one sat
21 adjacent to him at opposing counsel’s table.”).
13 1 But the percentage method has its limitations as well. As we indicated in Goldberger, this
2 Circuit’s adoption of the lodestar method was precipitated by the perception that percentage fees
3 “tended to yield too little for the client-class, and an unjustified ‘golden harvest of fees’ for the
4 lawyer.” 209 F.3d at 48 (quoting City of Detroit v. Grinnell Corp., 495 F.2d 448, 468, 469 (2d Cir.
5 1974) (“Grinnell I”), abrogated on other grounds by Goldberger, 209 F.3d at 43). Particularly in
6 cases that result in a very large monetary award, the percentage method holds the potential to result
7 in attorneys’ fees many times greater than those that would have been earned under the lodestar of
8 hourly rate multiplied by hours worked. “The principal analytical flaw” in Appellants’ argument for
9 a presumptive percentage award as a “benchmark” in common fund cases lies in the “assumption that
10 there is substantial contingency risk in every common fund case” that would justify such a multiplier.
11 Id. at 52.
12 Moreover, although the percentage method has the advantage of aligning the interests of
13 plaintiffs and their attorneys more fully by allowing the latter to share in both the upside and downside
14 risk of litigation, it can create perverse incentives of its own, potentially encouraging counsel to settle
15 a case prematurely once their opportunity costs begin to rise. See Coffee, Understanding the
16 Plaintiff’s Attorney at 687-90. And as in the case of the lodestar method, neither defense counsel
17 nor the actual plaintiffs have much of an incentive under the percentage-of-fund approach to oppose
18 an award of attorneys’ fees, the latter since “[t]hey have no real incentive to mount a challenge that
19 would result in only a ‘minuscule’ pro rata gain from a fee reduction.” Goldberger, 209 F.3d at 53
20 (citing Cont’l Ill. Sec. Litig., 962 F.2d 566, 573 (7th Cir. 1992)).
21 In short, neither the lodestar nor the percentage-of-fund approach to awarding attorneys’ fees
14 1 in common fund cases is without problems. It is for reasons such as those just discussed that in
2 Goldberger we declined to “junk” the lodestar method in favor of the presumptive or exclusive use
3 of the percentage method, see id. at 47-53, and instead left the decision as to the appropriate method
4 to “the district court, which is intimately familiar with the nuances of the case.” Id. at 48. While
5 Appellants assert that there would be a benefit in allowing “district judges . . . [to] step away from
6 the business of analyzing and reviewing attorneys’ fee applications,” Appellants’ Reply Br. at 9, we
7 underscore the importance of the district court’s duty “to act as a fiduciary who must serve as a
8 guardian of the rights of absent class members,” City of Detroit v. Grinnell Corp., 560 F.2d 1093,
9 1099 (2d Cir. 1977) (“Grinnell II”) (internal quotation marks omitted) abrogated on other grounds,
10 Goldberger, 209 F.3d at 43, and reaffirm the requirement of a “searching assessment” regarding
11 attorneys’ fees “that should properly be performed in each case.” Goldberger, 209 F.3d at 52. Even
12 were we not bound by Circuit precedent on the matter, we would decline to hold otherwise.
13 II. The District Court’s Reliance on Arbor Hill
14 Appellants also contend that the district court impermissibly relied on this court’s decision in
15 Arbor Hill Concerned Citizens Neighborhood Association v. County of Albany, 493 F.3d 110 (2d
16 Cir. 2007), since that case addressed the use of the lodestar method in the context of statutorily
17 awarded attorneys’ fees, rather than a common fund.4 They suggest, inter alia, that reliance on Arbor
4 The district court in this case relied on the original version of the Arbor Hill opinion. See 493 F.3d 110 (2d Cir. 2007). Following a petition for rehearing, this Court amended its opinion sua sponte. See 522 F.3d 182 (2d Cir. 2008). While we cite to the amended version, the differences between the two opinions are immaterial for the purposes of this appeal.
As Appellants note, the district court also referred to other statutory fee-shifting cases, including Porzig v. Dresdner, Kleinwort, Benson, North America LLC, 497 F.3d 133 (2d Cir.
15 1 Hill may have improperly influenced the district court’s decision to select a modified lodestar
2 approach rather than a percentage-of-fund approach, though the court’s bench decision does not, in
3 fact, make it at all evident that the case influenced the court’s selection.
4 At the start, we note that it is unclear how any error in the district court’s decision to use
5 Arbor Hill’s modified lodestar approach, even assuming such error occurred, could compel a finding
6 that the court should have used the percentage method instead. Given that a district court has the
7 discretion to choose either the lodestar or percentage-of-fund approach in calculating attorneys’ fees,
8 see supra Section I, any error in a court’s application of the lodestar method would not necessitate
9 selection of the percentage method on remand.
10 To the extent that Appellants’ argument depends on the inference that, had the district court
11 not referenced Arbor Hill, Appellants’ suggested 26% fee would have seemed more reasonable, an
12 analysis of Arbor Hill and the district court’s reliance on that decision shows that Appellants’ concern
13 is unfounded. In Arbor Hill, this Court proposed the use of a modified version of the lodestar
14 approach and recommended abandonment of the term “lodestar” for the alternative term
15 “presumptively reasonable fee.” 522 F.3d at 183. In brief, the Court suggested that a district court’s
16 efforts to approximate the reasonable fee that a competitive market would bear – the implicit goal of
17 the lodestar approach – would be better served by considering case-specific “reasonableness” factors
18 earlier in the calculation. Instead of first determining the lodestar (multiplying an hourly rate by the
19 number of hours worked) and then adjusting the lodestar through a multiplicative factor to account
2007), and Johnson v. Georgia Highway Express, Inc., 488 F.2d 714 (5th Cir. 1974), abrogated on other grounds by Blanchard v. Bergeron, 489 U.S. 87, 92-93 (1989). Our discussion in this section accounts for the district court’s reliance on those cases as well.
16 1 for case-specific considerations, the Arbor Hill Court suggested that a district court should assess
2 case-specific considerations at the outset, factoring them into its determination of a reasonable hourly
3 rate for the attorneys’ work. See id. at 186, 190. This reasonable hourly rate could then be
4 multiplied by the number of hours worked to generate a “presumptively reasonable fee.”5 Id. at 190.
5 As Arbor Hill was a statutory fee shifting case, the Court pointed towards the twelve case-specific
6 factors outlined in Johnson, 488 F.2d at 717-19, as being among those relevant to a district court’s
7 determination of a reasonable market rate under this revised approach. See Arbor Hill, 522 F.3d at
8 190.
9 Arbor Hill also indicated that the district court in that case had been unduly restrictive in its
10 interpretation of the “forum rule,” though it ultimately upheld the district court’s decision to deny the
11 request of plaintiffs’ counsel to calculate their hourly fees based on their home district, rather than
12 the less expansive prevailing rates in the district where the litigation occurred. See id. at 190-94. In
13 clarifying the forum rule, the Court held that:
14 [A] district court may use an out-of-district hourly rate – or some rate in between the 15 out-of-district rate sought and the rates charged by local attorneys – in calculating the 16 presumptively reasonable fee if it is clear that a reasonable, paying client would have 17 paid those higher rates. We presume, however, that a reasonable, paying client would 18 in most cases hire counsel . . . whose rates are consistent with those charged locally. 19 This presumption may be rebutted – albeit only in the unusual case – if the party 20 wishing the district court to use a higher rate demonstrates that his or her retention 21 of an out-of-district attorney was reasonable under the circumstances as they would 22 be reckoned by a client paying the attorney’s bill. 23
5 The Court’s dissatisfaction with the term “lodestar” stemmed from its sense that the term had lost its meaning in the sense of a “star that leads,” since the number generated by the first step of the traditional lodestar approach, by definition, did not incorporate case-specific factors that would enable it to correctly approximate a market hourly rate for fees. See id. at 189-90.
17 1 Id. at 191. The Court further indicated that “the reasonableness of a prevailing party’s decision to
2 retain out-of-district counsel is best considered in setting the hourly rate – rather than deciding
3 whether to adjust a presumptively reasonable fee,” since, inter alia, this approach comported with
4 the focus on market rates underpinning Arbor Hill’s “presumptively reasonable fee” approach. Id.
5 at 191-92.
6 In the instant case, the district court observed early in its bench decision that the Plaintiffs’
7 suggested fee award of 26% represented, in its estimation, a multiplier of 1.98-2.24 beyond what
8 counsel would have earned based on their hourly rates. Appellants suggest that to the extent the
9 district court’s impression of the multiplier as being too large may have been responsible for its
10 disinclination to base attorneys’ fees on a percentage of the fund, the court’s reliance upon Arbor Hill
11 worked to their detriment. In clarifying its understanding of that case, however, the court indicated
12 that if “[the traditional lodestar method] were to be applied . . . it would be irrelevant where an
13 attorney was from, [and the appropriate rate] would be what the prevailing rate is for attorneys in this
14 area.” Oral Decision at 26. The court then stated that “some of the fees as calculated in the fee
15 schedule submitted by class counsel . . . would have exceeded [the] traditional rate in the Northern
16 District of New York. I've not discounted them as a result of that.” Id. at 26-27. In other words,
17 had the district court not referenced Arbor Hill, and instead relied on its understanding of the forum
18 rule as traditionally applied, Appellants’ suggested 26% fee would have represented an even higher
19 multiplier of the lodestar. If Appellants are correct that the court’s assessment of the multiplier
20 contributed to its decision not to use a percentage-based approach, the court’s reliance on Arbor Hill
18 1 cannot have hurt them.6
2 In any event, we find no error in the district court’s reliance on Arbor Hill in this case. At
3 bottom, Appellants’ arguments regarding Arbor Hill are founded upon the assumption that the
4 lodestar method applied in the common fund context is distinct from that employed in the statutory
5 fee-shifting context. While it is true some district courts in this Circuit have expressed uncertainty
6 as to the relevance of Arbor Hill in common fund cases, see In re Ramp Corp. Sec. Litig., No. 05
7 Civ. 6521, 2008 WL 58938, at *2 n.3 (S.D.N.Y. Jan. 3, 2008); In re AOL Time Warner ERISA
8 Litig., No. 02 Cv. 8853, 2007 WL 3145111, at *1 n.3 (S.D.N.Y. Oct. 26, 2007), this Court’s case
9 law does not suggest that there are two different lodestar methods, see, e.g., In re Agent Orange
10 Products Litig., 818 F.2d 226, 232 (2d Cir. 1987) (“[W]e have adopted a lodestar formula for
11 calculating fees in equitable fund and statutory fee contexts.”). To the contrary, in discussing the
12 lodestar method, our common fund and statutory fee shifting cases have employed the same definition
13 and referenced the same foundational cases.7 See, e.g. Arbor Hill, 522 F.3d at 186 (citing Lindy
6 We note in passing that this Court’s recent decision in Simmons v. New York City Transit Authority, 575 F.3d 170 (2d Cir. 2009), has further clarified the forum rule, emphasizing Arbor Hill’s language that use of an out-of-district rate is appropriate “only in the unusual case.” See Simmons, 575 F.3d at 174. Were we to reconsider the district court’s decision to permit some class counsel in this case to use an out-of-district rate, we would examine whether Appellants made not only “a particularized showing . . . that the selection of out-of-district counsel was predicated on experience-based, objective factors,” but also demonstrated “the likelihood that the use of in-district counsel would [have] produce[d] a substantially inferior result.” See id. at 176. Given that the forum rule was not raised as an issue in this appeal, and the fact that it would likely be within the district court’s discretion to award the same attorneys’ fee by reevaluating the Goldberger factors, we decline to remand for consideration of this issue alone. 7 We do not suggest that concerns present in the common fund and statutory fee-shifting contexts are identical in every respect, compare Goldberger, 209 F.3d at 53 (noting that in common fund context, plaintiffs have little incentive to negotiate size of attorneys’ fees because of
19 1 Bros. Builder, Inc. v. Am. Radiator & Standard Sanitary Corp., 487 F.2d 161 (3d Cir. 1973) as
2 foundational case for lodestar approach in statutory fee case); Goldberger, 209 F.3d at 48 (citing
3 Lindy Bros. in common fund case); Grinnell I, 495 F.2d at 470-74 (citing Lindy Bros. in justifying
4 adoption of lodestar approach in common fund case). It is therefore not surprising that several
5 district courts within this Circuit have referred to Arbor Hill’s “presumptively reasonable fee”
6 approach in common fund cases. See Silberblatt v. Morgan Stanley, 524 F. Supp. 2d 425, 434
7 (S.D.N.Y. 2007) (referencing Arbor Hill in use of lodestar as cross-check on reasonableness of
8 percentage award); In re Assicurazioni Generali S.p.A. Holocaust Ins. Litig., Nos. 97 Civ. 2262, 98
9 Civ. 9186, 00 Civ. 9413, 2009 WL 762438, at *2 (S.D.N.Y. Mar. 24, 2009) (relying upon Arbor Hill
10 to explain application of lodestar method); AOL Time Warner, 2007 WL 3145111, at *1 n.3
11 (concurring with conclusions of Arbor Hill-based cross-check of percentage award). From a
12 mathematical perspective, of course, it makes little difference whether a court, following Arbor Hill,
13 considers case-specific factors to estimate a reasonable rate for an attorney’s services, which is then
14 multiplied by the number of hours worked, or whether the court takes the traditional approach and
15 considers these same factors in calculating a multiplier to the lodestar. The benefit of Arbor Hill’s
16 methodology is that by considering case-specific factors at the outset, the district court’s focus on
17 mimicking a market is maintained. See Arbor Hill, 522 F.3d at 192.
18 To the extent that the specific factors considered by courts in the common fund and statutory
their minimal pro rata gain from any reduction in fees), with Arbor Hill, 522 F.3d at 184 (noting that in the fee-shifting context, plaintiffs have little incentive to negotiate a market-based rate structure with attorneys because fees are paid entirely by defendants), but simply that the lodestar approach is common to both.
20 1 fee-shifting contexts are somewhat different, compare Wal-Mart, 396 F.3d at 121, with Johnson, 488
2 F.2d at 717-19, we note that the district court’s reliance on Arbor Hill in this case was limited. While
3 the court referenced Johnson in its initial discussion of Arbor Hill, it later clarified that it was relying
4 on the latter for its “presumptive reasonable fee” approach and its more flexible application of the
5 forum rule. And in considering case-specific factors, the district court clearly followed Goldberger,
6 addressing each of the six factors outlined in that case. The one Johnson factor separately considered
7 by the court, “the preclusion of employment by the attorney due to acceptance of the case,” Oral
8 Decision at 15, weighed in favor of a higher fee award, and its consideration did not prejudice the
9 Appellants. We are therefore unconvinced the court “modified the Goldberger factors” in any
10 meaningful way by referencing Arbor Hill, and accordingly, we find no error in the district court’s
11 reliance on that case.
12 III. The District Court’s Application of the Goldberger Factors
13 Appellants’ final arguments concern the district court’s application of the case-specific factors
14 elucidated in Goldberger for determining the reasonableness of a common fund fee. Appellants
15 contend that, in addition to committing errors in relation to each of these factors, the district court
16 “invented” an additional factor, holding the attorneys’ experience in litigating strip search cases
17 against them, when in fact their experience should have weighed in favor of a higher fee.
18 The Goldberger factors are applicable to the court’s reasonableness determination whether
19 a percentage-of-fund or lodestar approach is used, see Wal-Mart, 396 F.3d at 121 (“Irrespective of
20 which method is used, the ‘Goldberger factors’ ultimately determine the reasonableness of a common
21 fund fee.”), and in the latter context, indicate whether a multiplier should be applied to the lodestar.
21 1 (Under the Arbor Hill approach, they would be factored into the appropriate hourly rate before
2 calculating the presumptively reasonable fee.) The Goldberger factors include: (1) counsel’s time
3 and labor; (2) the litigation’s magnitude and complexity; (3) the risk of the litigation; (4) the quality
4 of representation; (5) the requested fee in relation to the settlement; and (6) public policy
5 considerations. 209 F.3d at 50. Importantly, we defer to the district court’s determinations regarding
6 these factors. See id. at 47-48. Cf. Merck-Mecco, 504 F.3d at 249 (“The District Court applied the
7 Goldberger test and made specific and detailed findings from the record, as well as from its own
8 familiarity with the case . . . . Accordingly, we find no reason to disturb the District Court’s
9 [determination].”).
10 We address first Appellants’ contention that the district court improperly counted the
11 attorneys’ experience in litigating strip search cases against them when that factor should have
12 counted, if at all, in favor of the higher proposed fee award. Although attorney experience is not
13 explicitly enumerated among the Goldberger factors, it is clear that experience might be relevant to
14 several of them, including consideration of the time and effort expended by counsel, the complexity
15 of the litigation, and quality of the representation. In Goldberger itself, this Court upheld the district
16 court’s decision to award a lodestar fee with no multiplier in part based upon its observation that the
17 attorneys had been “helped enormously” by the “spadework” performed by federal authorities in
18 actions against the same defendants and because the legal issues presented were not novel. 209 F.3d
19 at 54, 56.
20 Here, the advance spadework that assisted counsel in their suit against the County of
21 Schenectady did not involve the previous unearthing of facts regarding the specific defendants, but
22 1 rather the prior mining of relevant case law and shoring up of legal arguments that the district court
2 reasonably concluded to have occurred during the course of two prior actions alleging
3 unconstitutional strip search policies. The district court’s decision to consider the benefit afforded
4 to counsel by this experience – relevant to several of the Goldberger factors – does not constitute an
5 error of law. Nor does Appellants’ observation that experience might in some cases enable counsel
6 to demand a higher fee make the district court’s conclusion that it cut the opposite way in this case
7 clearly erroneous. Indeed, our prior case law indicates that whether a given factor cuts in favor of
8 or against the use of a multiplier depends heavily on the facts of a case. The level of risk associated
9 with litigation, for example, which is “perhaps the foremost factor” to be considered in assessing the
10 propriety of a multiplier, and typically weighs in its favor, can weigh against the use of a multiplier
11 when the magnitude of risk stems only from a lawsuit’s dubious legal merit. See id. at 54 (internal
12 quotation marks omitted).
13 We similarly see no clear error in the district court’s remaining determinations as to the
14 Goldberger factors. Appellants find fault with the court’s determination regarding the first factor,
15 that counsel were adequately compensated for their time and effort using their ordinary rates, and
16 suggest that here attorney experience should have weighed in favor of some multiplier. They further
17 criticize the court’s finding regarding the fourth factor, quality of representation, and urge that the
18 fact that class counsel “did a fine job and obtained a good settlement,” Oral Decision at 14, also
19 should have counted in favor of a higher award. We have already addressed the experience argument,
20 and for similar reasons find Appellants’ contention regarding their performance to be insufficient to
21 show clear error. Even in Goldberger, where counsel were praised by the district court as “the cream
23 1 of the profession” whose “genius and dedication were vital in resolving the complexities of the
2 litigation,” this Court upheld a fee award with no multiplier. 209 F.3d at 55-56. Both here, and in
3 that case, “while the district court declined to award formal multipliers for . . . quality of
4 representation, the court did consider those factors by allowing counsel to recover generous hourly
5 fees.” Id. at 56. “[T]he question before us is not whether we would have awarded a different fee,
6 but rather whether the district court abused its discretion in awarding this fee.” In re Nortel Networks
7 Corp. Sec. Litig., 539 F.3d 129, 134 (2d Cir. 2008).
8 Nor are we convinced that Goldberger factors two and three, the complexity and risk
9 associated with the litigation, compel a higher fee award than that granted by the court below. In
10 response to the district court’s conclusion that the case was “an ordinary civil rights case in which
11 liability appear[ed] reasonably certain,” Oral Decision at 13, Appellants contend that the case was “far
12 from simple,” Appellants’ Br. at 24, that this Circuit’s holding in Shain regarding the
13 unconstitutionality of misdemeanor strip searches was “far from stable” at the time of their suit, Id.
14 at 25 (quoting McBean v. City of New York, 233 F.R.D. 377, 387 (S.D.N.Y. 2007)), and that they
15 faced substantial litigation risk due to the availability of an exhaustion of remedies defense under the
16 Prison Litigation Reform Act (“PLRA”) and because damages were difficult to prove in their case.
17 While appellants are correct that Shain was decided by a divided panel, we cannot say that the district
18 court’s sense of the strength of Appellants’ case was clearly erroneous. At the time the instant suit
19 was brought, see Goldberger, 209 F.3d at 55 (noting that “[i]t is well-established that litigation risk
20 must be measured as of when the case is filed”), a policy of conducting blanket strip searches of
21 misdemeanor detainees was clearly unconstitutional. See Shain, 273 F.3d at 62-66. Settlements had
24 1 already been reached in two other class actions involving blanket strip search policies, including
2 Kahler, in which Appellants’ counsel themselves were involved, and the issue of liability had been
3 resolved in favor of plaintiffs in a third. See Dodge v. County of Orange, 282 F. Supp. 2d 41
4 (S.D.N.Y. 2003) (finding defendants liable); Stipulation and Order of Settlement and Discontinuance,
5 Spinner v. City of New York, No. 1:01-cv-2715 (E.D.N.Y. June 9, 2004); Motion to Certify Class
6 and for Preliminary Approval of Class Action Settlement, Kahler v. Rensselaer County, No. 1:03-cv-
7 01324 (N.D.N.Y. Mar. 18, 2004). Moreover, Appellants’ complaint noted that the Schenectady
8 Police Department’s strip search policy, which they alleged was closely related to that of the
9 defendant county, had been found unconstitutional in Gonzalez v. City of Schenectady, 141 F. Supp.
10 2d 304 (N.D.N.Y. 2001). Compl. at ¶ 33. While Appellants note that the Appellees distinguished
11 the “change outs” pursuant to their policy from forbidden strip searches, and point to a purportedly
12 similar case in which a lawsuit was dismissed, that case was decided nearly a year after they brought
13 suit. See Judgment, Steinberg v. County of Rockland, No. 04 Civ. 4889 (S.D.N.Y. May 23, 2005).
15 Appellants’ assertions that the defendants in this case possessed potentially valid defenses and
16 that damages were difficult to prove are similarly insufficient to demonstrate clear error in the district
17 court’s findings with respect to litigation risk. Counsel must do more than “point to . . . general
18 hurdles – such as the defenses available to defendants –” and “argue that because their fee was
19 entirely contingent on their ability to overcome such hurdles, they must, as a matter of law, be
20 compensated by a fee enhancement.” Goldberger, 209 F.3d at 54. While the district court
21 acknowledged that an award of damages was uncertain in the case, it correctly noted that “[t]he
25 1 damages risk is of little consequence to an attorney for a prevailing plaintiff in a civil rights case
2 who’s entitled to obtain attorneys’ fees pursuant to 14 [U.S.C. §] 1988.” Oral Decision at 13-14.
3 Even if only nominal monetary damages had been recovered, Appellants’ attorneys stood to recover
4 a fee based on their success in obtaining injunctive relief. “[W]here both monetary and equitable relief
5 have been pursued, the size of the monetary recovery is not necessarily the proper measure of the
6 plaintiff’s success. . . . ‘[A] plaintiff who failed to recover damages but obtained injunctive relief . .
7 . may recover a fee award based on all hours reasonably expended if the relief obtained justified that
8 expenditure of attorney time.’” LeBlanc-Sternberg v. Fletcher, 143 F.3d 748, 758 (2d Cir. 1998)
9 (quoting Hensley v. Eckerhart, 461 U.S. 424, 435 n. 11 (1983)). Cf. McCardle v. Haddad, 131 F.3d
10 43, 54 (2d Cir. 1997) (“[W]e have held that the denial of attorneys’ fees was not an abuse of
11 discretion where the plaintiff recovered only nominal damages and received no other meaningful
12 relief.”).
13 Finally, we reject Appellants’ assertions that the district court erred in its findings as to the
14 fifth and sixth Goldberger factors, the reasonableness of the fee in relation to the size of the
15 settlement and public policy considerations relevant to fee awards. Appellants’ argument with regard
16 to the former essentially repeats their “benchmark” assertion that the fee award in this case was too
17 far from the percentage-of-fund awards in similar cases, adding that the district court erred in its
18 assessment that an award equivalent to 13% was “squarely within the range” of typical awards and
19 that Appellants’ suggested lodestar multiplier was at the low end of multipliers that have often been
20 awarded. While Appellants cite to Nortel as demonstrating this Court’s concern over a district
21 court’s failure to consider awards in similar cases, that case upheld a fee award despite its disparity
26 1 from other awards. See 539 F.3d at 134. Our case law makes clear both that no presumption applies
2 to the percentage of a common fund that should be awarded to plaintiffs’ counsel, and that an award
3 of no lodestar multiplier at all is within the district court’s discretion. See Goldberger, 209 F.3d at
4 51-57. A fee award does not “constitute an abuse of discretion simply because it deviates materially
5 from the [percentage] usually awarded in similar cases. Instead, we adhere to our prior practice that
6 a fee award should be based on scrutiny of the unique circumstances of each case, and a ‘jealous
7 regard to the rights of those who are interested in the fund.’” Id. at 53 (quoting Grinnel II, 495 F.2d
8 at 469). Even if a more thorough review by the district court of attorneys’ fee awards in strip-search
9 common fund cases might have indicated that the 13% fee awarded in this case was slightly below
10 the lower end of the spectrum, see Nilsen v. York County, 400 F. Supp. 2d 266, 281, 287-289 (D.
11 Me. 2005) (observing a percentage-of-fund range of 16-33% in strip search cases), we conclude that
12 the district court’s award was within its discretion given the considerations outlined in its discussion
13 of the Goldberger factors. See Nortel, 539 F.3d at 134.
14 We are mindful that public policy supports the pursuit of meritorious class action litigation
15 to vindicate constitutional rights. Appellants suggest that some reward beyond the ordinary
16 remuneration for legal services is needed to ensure that such suits are brought by competent counsel.
17 “On the other side of the ledger, however, is our longstanding concern for moderation.” Goldberger,
18 209 F.3d at 57. Civil rights cases may or may not raise the same danger of “routine
19 overcompensation” for risk that has troubled this court in the context of “mega-fund” class actions,
20 but it is because of the case-specific nature of the fee award inquiry that we have been loath to disturb
21 the determinations of district courts in this area. “When the exercise of . . . discretion is supported
27 1 by adequate findings and is consistent with our preference for moderation, as it was here, we will not
2 substitute our own predilections for the judgment of the district court.” Id.
3 CONCLUSION
4 For the foregoing reasons, the order of the District Court appealed from is AFFIRMED.