ORDER ON MOTION FOR AWARD OF ATTORNEY FEES AND LITIGATION EXPENSES

HORNBY, District Judge.

I. INTRODUCTION

This is an award of attorney fees and expenses, out of a $3.3 million class action settlement for arrestee strip searches at the York County Jail. Although I approved final settlement in my Orders of August 18 and September 8, 2005, I reserved ruling on the motion for attorney fees. I conclude that the preferred method for determining a reasonable attorney fee is a market-mimicking analysis. Although the market data here is scant, it leads me to award 25% of the settlement, $825,000. Final expenses remain to be determined. But it appears that approximately $2,400,000 (including interest of about $70,000) will go to the class members.

II. PROCEDURAL BACKGROUND

The settlement agreement requires York County to establish a common fund of $3.3 million in satisfaction of all its liabilities (including attorney fees). The fund is to be distributed to class members after deduction of costs and fees. The agreement also provides that the plaintiffs' lawyers will ask the Court to award attorney fees out of the fund, in the amount of 30%, along with reimbursement for costs and expenses. Obviously the agreement does not and cannot dictate what amount I will actually award.

I held a preliminary hearing on the proposed settlement, and approved class-wide notice. Prior to the final fairness hearing, the plaintiffs submitted a Motion for Attorney's Fees and Litigation Expenses. After the final fairness hearing, I approved the settlement in all respects but one, and gave the parties the opportunity to amend. At that time I reserved ruling on the request for attorney fees and expenses. The parties amended the settlement to cure the offending term, and on September 8, 2005, I granted final approval of the settlement. The motion for attorney fees, therefore, is now ripe for decision.

The plaintiffs' lawyers have asked me to award attorney fees, using the "percentage-of-funds" method, in the amount of 30% of the total settlement fund (irrespective of interest.) As the settlement fund is worth $3.3 million, they are asking for $990,000. (On the basis solely of hourly rates and hours spent, the lawyers estimated in June, 2005, that they would accrue somewhat less than $520,000 by the time everything is complete, although that estimate may now be higher, see Pls.' Rep. on Admin. of Settlement 5 (Docket Item 181).) They also request reimbursement of litigation expenses and claims administration costs, in the amount of 3% of the total settlement, i.e., $99,000. If awarded 30% in fees, they have agreed to limit cost and expense reimbursement to this 3%, and propose to pay any additional expenses out of attorney fees.

III. ANALYSIS

Under the Federal Rules of Civil Procedure, in a class action I may award "reasonable attorney fees and nontaxable costs authorized by law." Fed.R.Civ.P. 23(h). Had this lawsuit proceeded to a successful judgment for the plaintiffs, they could have recovered from the defendant York County reasonable attorney fees and costs under 42 U.S.C § 1988, on top of any damages they received. Instead, York County settled the lawsuit before trial for a lump sum of $3.3 million, covering all its liabilities, including attorney fees. Second Am. Settlement Agreement ("Settlement Agreement") ¶¶ 6-7 (Docket Item 168). That settlement amount reflects both the class members' compensable injuries and their statutory claim to attorney fees. Unless the lawyers who have produced the successful outcome are paid, the class members will be unjustly enriched. See Florin v. Nationsbank of Ga. ( Florin I), 34 F.3d 560, 564 (7th Cir. 1994) ("[T]he settlement agreement seems to anticipate that the amount paid by the defendants into the fund includes an unspecified sum for class counsel's fees. . . ."); Boeing Co. v. Van Gemert, 444 U.S. 472, 478 (1980) ("[P]ersons who obtain the benefit of a lawsuit without contributing to its cost are unjustly enriched. . . ."); Restatement (Third) of Restitution and Unjust Enrichment § 30 cmt. b, at 75 (Tentative Draft No. 3, 2004) (when a fee-shifting case settles through creation of a common fund, "[q]uestions of unjust enrichment may guide the court's inquiry"). By the terminology of some of the caselaw, this settlement is a "common fund," and the common fund doctrine allows me to award attorney fees from it. In re Compact Disc Minimum Advertised Price Antitrust Litig. ( In re Compact Disc I), 216 F.R.D. 197, 216 n. 43 (D. Me. 2003) (citing In re Thirteen Appeals Arising out of the San Juan Dupont Plaza Hotel Fire Litig., 56 F.3d 295, 307 (1st Cir. 1995)); see also 4 Alba Conte Herbert B. Newberg, Newberg on Class Actions § 14:6 (4th ed. 2002).

Making a fair fee award from a common fund in a class action settlement is a difficult determination for a judge. There are no adversarial presentations to test the fee claim, and our legal system does not ordinarily expect judges to behave as inquisitors, gathering testimony and collecting information on their own. Presented with an unopposed request, therefore, I depend upon my own analysis and secondary research — against a backdrop of popular dissatisfaction with large and highly publicized fees. Third Circuit Task Force Report, Selection of Class Counsel, 208 F.R.D. 340, 343-44 (2002) (" 2002 Task Force Report") ("[T]here is a perception among a significant part of the non-lawyer population . . . that class action plaintiffs' lawyers are overcompensated for the work that they do."). But the lawyers here are highly skilled and experienced civil rights attorneys. Their professional performance was exemplary; they represented the class members' interests zealously, achieving an excellent result for the class under the circumstances. For these reasons, they deserve a reasonable fee that duly recognizes their professional excellence and performance and provides an appropriate incentive for lawyers to take on future meritorious cases on behalf of a client class. At the same time, they do not deserve a windfall at the expense of the class and I do not want the size of the award to encourage frivolous litigation that benefits primarily lawyers.

In candor, I simply do not know the precise fee award that meets those concerns. I can only detail the process I have used.

(1) Method: Lodestar or Percentage-of-Funds. In the First Circuit, courts have discretion to award fees from a common fund "either on a percentage of the fund basis or by fashioning a lodestar." In re Thirteen Appeals, 56 F.3d at 307 (approving allocation of fee award using percentage-of-funds method); United States v. 8.0 Acres of Land, 197 F.3d 24, 33 (1999). As between the two methods, the First Circuit has noted that the percentage-of-funds method is the prevailing practice, and that it may have distinct advantages over the lodestar approach. In re Thirteen Appeals, 56 F.3d at 307; see also 2002 Task Force Report, 208 F.R.D. at 422 (suggested procedures for reviewing fee requests in non-auction cases); In re Compact Disc I, 216 F.R.D. at 215. According to the Third Circuit, the percentage-of-funds method is preferred in common fund cases "because it allows courts to award fees from the fund in a manner that rewards counsel for success and penalizes it for failure." In re Rite Aid Corp. Sec. Litig., 396 F.3d 294, 300 (3d Cir. 2005) (quoting In re Prudential Ins. Co. of America Sales Practices Litig., 148 F.3d 283, 333 (3d Cir. 1998) (internal quotation marks omitted). I shall use the percentage-of-funds method here — i.e., is the total fee reasonable when examined as a percentage of recovery?

(2) The Limited Role of the Lodestar. As I have previously recognized, the "lodestar approach (reasonable hours spent times reasonable hourly rates, subject to a multiplier or discount for special circumstances, plus reasonable disbursements) can be a check or validation of the appropriateness of the percentage-of-funds fee, but is not required." In re Compact Disc I, 216 F.R.D. at 215-16; see also Vaughn R. Walker Ben Horwich, The Ethical Imperatives of a Lodestar Cross-Check: Judicial Misgivings About "Reasonable Percentage" Fees in Common Fund Cases, 18 Geo. J. Legal Ethics 1453, 1468, 1470 (2005) (arguing that courts reviewing percentage fees have ethical obligation to perform cross-check); but see 2002 Task Force Report, 208 F.R.D. at 422 (expressing skepticism of even this limited use of the lodestar; "the lodestar is at most a relevant factor . . . and it should not receive exaggerated importance."). Here, the lodestar figure is lower than the requested percentage fee, approximately $520,000 vs. $990,000.

When the lodestar cross-check shows that the percentage fee is lower than the fees the lawyers have accrued on a time-and-services basis, it is relatively easy to support a percentage-based fee. See, e.g., In re Compact Disc I, 216 F.R.D. at 217; In re Compact Disc Minimum Advertised Price Antitrust Litig. ( In re Compact Disc II), 370 F.Supp.2d 320, 323 (D. Me. 2005). But what if the percentage fee is higher than the lodestar (as it is here), maybe substantially higher?

Under binding Supreme Court precedent, lodestar enhancements or multipliers in fee-shifting cases are almost completely unavailable, because many considerations that could lead to enhancement are contained within the lodestar itself. Penn. v. Del. Valley Citizens' Council for Clean Air, 478 U.S. 546, 566, 106 S.Ct. 3088, 92 L.Ed.2d 439 (1986) ("[T]he lodestar figure includes most, if not all, of the relevant factors constituting a `reasonable' attorneys' fee."); see also, e.g., Blum v. Stenson, 465 U.S. 886, 898-99, 104 S.Ct. 1541, 79 L.Ed.2d 891 (1984) (case's "novelty and complexity" normally reflected in number of hours expended, and lawyers' "experience and special skill" normally reflected in hourly rate). Outright prohibitions also exist. Id. at 901, n. 16, 104 S.Ct. 1541 (no enhancement for "great number of persons benefited."); City of Burlington v. Dague, 505 U.S. 557, 567, 112 S.Ct. 2638, 120 L.Ed.2d 449 (1992) (no enhancement for contingency or risk of litigation); but see Blum, 465 U.S. at 901, 104 S.Ct. 1541 (enhancement may still be available in cases of "exceptional success"). The First Circuit is emphatic in rejecting multipliers to enhance the lodestar. See Lipsett v. Blanco, 975 F.2d 934, 942 (1st Cir. 1992) (enhancement for quality of representation is a tiny exception that may not "eclipse the rule," and only available "where a combination of sterling performance and exceptional results could conceivably justify a premium fee"); Adams v. Zimmerman, 73 F.3d 1164, 1178 (1st Cir. 1996) (holding for second time post- Dague that contingency enhancement unavailable). In fact, since Dague, the First Circuit has never awarded a lodestar enhancement or multiplier. The unenhanced lodestar figure is strongly presumed to be the reasonable attorney fee that the losing defendant should pay in a fee-shifting case.

Do these constraints on multipliers apply when I am not making an award against a losing defendant, but instead determining the reasonable fee that the lawyers should recover from the plaintiff class's common fund settlement? I conclude that they do not.

First, it is well established that a defendant may settle, for a single lump sum, all outstanding claims in a fee-shifting case, including claims for attorney fees. Evans v. Jeff D., 475 U.S. 717, 731-32 (1986) (Congress intended fee awards to be part of civil rights plaintiff's "arsenal of remedies," and settlement including waiver of award is consistent with this goal).

Second, it is equally well established for nonclass actions that the statutory amount that a defendant is required to pay under fee-shifting principles does not limit the amount that the plaintiff must pay the plaintiff's lawyer. Venegas, 495 U.S. at 90 (fee-shifting statute "controls what the losing defendant must pay, not what the prevailing plaintiff must pay his lawyer"). The Supreme Court recently reaffirmed this principle in Gisbrecht v. Barnhart, 535 U.S. 789 (2002). There, it held that the plaintiff was bound to pay his lawyer the agreed-upon contingency fee of 25%, rather than the lower lodestar amount the district court assessed against the defendant under a fee-shifting statute. Id. at 806 ("[W]e again emphasize . . . [that] the lodestar method was designed to govern imposition of fees on the losing party. . . . [N]othing prevents the attorney from gaining additional fees, pursuant to contract, from his own client."); accord Gobert v. Williams, 323 F.3d 1099, 1100 (5th Cir. 2003) (plaintiff's obligation to pay lawyer determined by retainer agreement, independent of fee award made by court).

These cases, read together, uncouple the fee analysis in determining an award against the losing defendant (the unenhanced lodestar) from the fee analysis for determining an award to the lawyer from the amount that he or she has recovered. Although, unlike Gisbrecht, there may be no fee contract for the plaintiffs' lawyers in a typical class action, I conclude that the reasoning of these cases permits me to award a percentage attorney fee (so long as it is reasonable) from a common fund in a class action settlement even if the fee effectively represents a multiplier of the lodestar amount. Accord Florin I, 34 F.3d at 564-65 (the policy considerations forbidding "multiples in statutory fee-shifting cases have little force in common fund cases"). (3) The Reasonableness of the Requested Percentage. In a class action, any fee I award must be reasonable. Fed.R.Civ.P. 23(h); see also Boeing Co., 444 U.S. at 478 (lawyer creating common fund benefiting non-clients is entitled to "reasonable attorney's fee" from fund). Thus, I must determine whether the 30% requested is reasonable under the circumstances, bearing in mind that my role takes on special importance because there are no adversaries here to dispute the fee and because I must act as a fiduciary of the class. See Fed.R.Civ.P. 23(h) advisory committee note ("Active judicial involvement in measuring fee awards is singularly important to the proper operation of the class-action process . . . [e]ven in the absence of objections."); In re Rite Aid Corp. Sec. Litig., 396 F.3d at 307 (when determining fees, judges "must protect the class's interest by acting as a fiduciary"). First Circuit cases give no guidance on how to determine whether a given percentage is reasonable. First, therefore, I canvass the approaches of other Circuits.

(a) Other Circuit Approaches. The majority of Circuits review percentage-of-funds fee awards by using multifactor tests in which the district court must examine and set forth findings on each factor. The Second Circuit requires analysis of six factors. Goldberger v. Integrated Res., Inc., 209 F.3d 43, 50 (2d Cir. 2000); see also Wal-Mart Stores, Inc., 396 F.3d at 121. The Sixth Circuit also uses six factors. Ramey v. Cincinnati Enquirer, Inc., 508 F.2d 1188, 1196 (6th Cir. 1974); see also Smillie v. Park Chem. Co., 710 F.2d 271, 275 (6th Cir. 1983). The Third Circuit, both in caselaw and through its influential Task Force, requires seven. Gunter v. Ridgewood Energy Corp., 223 F.3d 190, 195 n. 1 (3d Cir. 2000); 2002 Task Force Report, 208 F.R.D. at 423-24; see also In re Rite Aid Corp. Sec. Litig., 396 F.3d at 301.

Other circuits explicitly adopt the twelve factors of a seminal lodestar case, Johnson v. Georgia Highway Express, Inc., 488 F.2d 714, 717-19 (5th Cir. 1974) (adopting factors as "consistent with" the Model Code of Prof'l Responsibility). For example, the Fourth and Tenth Circuits mandate "pure" Johnson factors for percentage-of-funds fee awards. In re MRRM, P.A., 404 F.3d 863, 867-68 (4th Cir. 2005) (applying factors of Barber v. Kimbrell's, Inc., 577 F.2d 216, 226 (4th Cir. 1978) (adopting Johnson), to percentage award); Brown v. Phillips Petroleum Co., 838 F.2d 451, 454-55 (10th Cir. 1988). The Eleventh Circuit also mandates the Johnson factors, but adds still more factors, similar to those of the non- Johnson circuits. Camden I Condo. Ass'n v. Dunkle, 946 F.2d 768, 775 (11th Cir. 1991).

It is easy to see that the multifactor tests largely overlap. Almost all include factors related to the lodestar amount (time and labor), the complexity and difficulty of the case, the quality of the lawyers or the representation (including skill, standing, and efficiency), the size of the fund (including the value of the benefit to the class), and the risk of nonpayment or contingency of the case. Additionally, both the Third Circuit and the three Johnson circuits add comparison to other awards in similar cases, and the Third and Eleventh Circuits also look to whether there are objections. These, then, comprise the bulk of the factors used in the multifactor jurisdictions.

Three other circuits do not mandate specific multifactor tests. Instead, the Ninth Circuit uses a benchmark of 25%, from which deviation is permitted upon consideration of various case-specific factors. Vizcaino v. Microsoft Corp., 290 F.3d at 1047-48. The Eighth Circuit and the District of Columbia Circuit have not ruled on the appropriate way to review a percentage-of-funds fee award, and have not engaged in any extended discussion of reasonableness. But both those circuits have pointed to "benchmark" percentage ranges to justify the reasonableness of particular fees. Petrovic v. Amoco Oil Co., 200 F.3d 1140, 1157 (8th Cir. 1999) (24% fee found reasonable by citing 1985 Task Force Report's proposition that fees in range of 20-25% are reasonable); Democratic Cent. Comm. of D.C. v. Wash. Metro. Area Transit Comm'n, 3 F.3d 1568, 1575 (D.C. Cir. 1993) (fee found reasonable in part because it "falls well within the range usually awarded in common fund cases," 20% to 30%); see also Camden I Condo. Assoc., 946 F.2d at 775 (finding that district courts increasingly view 25% as "bench mark").

The Seventh Circuit determines reasonableness through a market-mimicking approach with the goal of awarding a fee that is the "market price for legal services, in light of the risk of nonpayment and the normal rate of compensation in the market" at the outset of the case. In re Synthroid Mktg. Litig., 264 F.3d at 718. According to the Seventh Circuit, reasonableness is not an ethical or philosophical question; "it is not the function of judges in fee litigation to determine the equivalent of the medieval just price. It is to determine what the lawyer would receive if he were selling his services in the market rather than being paid by court order." In re Cont'l Ill. Sec. Litig., 962 F.2d 566, 568 (7th Cir. 1992). This approach emulates the incentives of a private client-attorney relationship; the market price should take into account "the risk of nonpayment," "quality of performance," "the amount of work," and "the stakes of the case." Id. at 721. For the Seventh Circuit, the key inquiry is what a private plaintiff would have negotiated with the lawyers, if they had bargained at arm's length at the outset of the case. Id. at 718.

The Seventh Circuit approach presents special concerns in the context of class actions, of course, for generally no contract exists between lawyers and the class. After all, attorney fees in a class action require court, not client, approval. Thus, there is no readily apparent source of information about the market. See id. at 573. That is the major criticism leveled at the Seventh Circuit approach: that for some fee-shifting class action claims, any "market" is simply illusory and speculative, Vizcaino, 290 F.3d at 1049-50, and that instead of any sort of privately negotiated fee, for many noncommercial cases the fee is set entirely by judicial reference to what is reasonable. Id. at 1049 (the market is "simply counsel's expectation of court-awarded fees"); see also Goldberger, 209 F.3d at 51-52 ("[M]arket rates, where available, are the ideal proxy for [lawyers'] compensation. The problem is that we cannot know precisely what fees common fund plaintiffs in an efficient market for legal services would agree to. . . ."). While the Seventh Circuit has recognized that this is a legitimate concern, it retorts that a "consider-everything," factor-based method of setting fees "assures random and potentially perverse results," and that a "list of factors without a rule of decision is just a chopped salad." In re Synthroid Mktg. Litig., 264 F.3d at 719.

(b) The Method I Shall Adopt in this Case. My decision on attorney fees is of great importance to the lawyers and class members in this case. Every dollar I award the lawyers is a dollar that the class members will not recover and vice-versa. It also has great importance for future lawyers and plaintiff classes in this District, for in the absence of First Circuit law, they deserve an intelligible and predictable standard. I hope that an ongoing exposure to attorney fee requests in these class actions has begun to inform my thinking. I shall try to explain fully the rationale for my award for guidance in future attorney fee cases.

In past awards from common funds, I have recognized the difficulty and arbitrariness of assessing the reasonableness of a given fee. In re Compact Disc I, 216 F.R.D. at 216 ("Although some of the cases and commentaries call upon district judges to give a precise analysis of how they choose the percentage, I do not know how to do so in any meaningful way that would defend 10% over 11% or 9%, etc."). In none of those cases did I specify a methodology for determining reasonableness; rather, I referred more generally to finding the particular percentages reasonable under the circumstances. In re Compact Disc II, 370 F.Supp.2d at 323; In re Compact Disc I, 216 F.R.D. at 216. In each of them, the fee I awarded was lower than the lodestar figure accrued by the lawyers at the time of settlement. See also Tr. Final Fairness Hr'g at 37, Ramirez v. DeCoster, No. 2:98cv186 (D. Me. final approval of settlement Dec. 3, 2002) (Docket Item 264) (awarding attorney fee of one-third of recovery, lower than lodestar). Today, in the first case in which I have been asked to award a percentage that is higher than the lodestar, I shall identify a specific methodology.

The path of least resistance is to employ the multifactor approach to reasonableness. The majority of circuits use it and the fee-seeking lawyers have used it here. It will support virtually any percentage fee I award in a range from 16% to 33-1/3%.

But the preceding statement is the first and primary reason I reject the multifactor approach. It offers little predictability to either the awarding court, or the lawyers who seek fee awards. Report of the Third Circuit Task Force, Court Awarded Fees, 108 F.R.D. 237, 244-45 (1985) (" 1985 Task Force Report") (noting criticism of factor-based standard that it does not guarantee "rational" awards and that it encourages district courts to make conclusory awards subject to reversal and remand). In this case, for example, it would support equally a fee award of 16%, 20%, 25%, 30%, or 33-1/3%. That is not a rule of law or even a principle. Instead, it allows uncabined discretion to the fee-awarding judge. A judge who likes lawyers and remembers the hazards of practice can be generous; a judge who cares more about public reaction or who never used contingent fees in practice can be stingy. It is difficult to contradict the judge's statement about the case's complexity or lack thereof, the difficulties of discovery, the quality of the lawyering, etc. These are all highly subjective judgments.

The second reason I reject the multifactor approach is that some of the factors seem inconsistent with the reason for using a percentage-of-funds method in the first place, which is designed to create incentives for the lawyer to get the most recovery for the class by the most efficient manner (and penalize the lawyer who fails to do so). See In re Thirteen Appeals, 56 F.3d at 307 (percentage-of-funds method eliminates incentive to be inefficient, as inefficiency just reduces the lawyer's own recovery); Wal-Mart Stores, Inc., 396 F.3d at 121 (the percentage method "directly aligns the interests of the class and its counsel" and provides a powerful incentive for efficiency and early resolution) (citations omitted). Why, then, should we adjust the percentage depending upon whether the lawyer was efficient, how much time the lawyer put into the case, etc.?

The third reason I reject the factor-based approaches is that they consume significant lawyer and judicial resources, a consideration that has led to the criticism (and in some quarters, the abandonment) of the lodestar method in favor of the percentage-of-funds method. See In re Thirteen Appeals, 56 F.3d at 307 (stating lodestar method more "burdensome to administer" than percentage-of-funds method).

I believe that in setting a fee, a judge, consciously or unconsciously, necessarily compares what lawyers typically get paid for equivalent services — i.e., the market price. The judge may do that based on recollections from when that judge was a lawyer, from information generated in other cases, or from general lawyer/judge gossip, but inevitably the judge takes such information into account. The Seventh Circuit appropriately makes this measure explicit, rather than a Gestalt lurking behind the multifactor review. Making the standard explicit, in turn, allows the lawyers or objectors to provide evidence to correct judicial misimpressions.

There is good reason for using a market-oriented approach. If a consumer wanted to determine a reasonable plumber's, mechanic's or dentist's fee, the consumer would have to look to the market. Why should lawyers be different? Perhaps more important, the market is the implicit if not explicit standard when a jury awards damages that include reasonable medical expenses in a personal injury case. We do not use a multifactor approach then. We even look at the market to a degree in lodestar cases, because we purport there to look at market rates for what a lawyer can charge as an hourly rate.

I therefore adopt the methodology of the Seventh Circuit as most reflective of what a judge does instinctively in setting a fee as well as most amenable to predictability and an objective external constraint on a judge's otherwise uncabined power: "courts must do their best to award counsel the market price for legal services, in light of the risk of nonpayment and the normal rate of compensation in the market at the time." In re Synthroid Mktg. Litig., 264 F.3d at 718. The market-mimicking approach has its own shortcomings but it is better than the fuzzier alternatives.

Unfortunately, in this case I do not have direct data on the market price for civil rights class action lawyers, or for strip search class action cases inside or outside of Maine. To some degree, that is unavoidable for the reasons already described: the class action "market" is controlled by judges. That is why this is a market "mimicking" approach. But there should be contextual market information available, such as what lawyers charge as percentage fees for various types of litigation. As I suggest later, I shall probably ask for such information in future cases. Moreover, because I have not used this method of testing reasonableness previously and because First Circuit caselaw is silent, in fairness I will re-open the question if the lawyers wish to present additional evidence and arguments addressing it. If they do so request, however, I shall also consider appointing a special master under Fed.R.Civ.P. 53, or an independent expert under Fed.R.Evid. 706, and also entertain the defendant's offer to reveal information about its attorney fee arrangements.

In the meantime, I look to such evidence as is available to suggest what hypothetical negotiation would have produced at the outset as an agreed-to fee for a case such as this.

(1) Standard Contingency Fees. First, I consider standard contingency fees and the limitations imposed upon them by statutes or regulations. Although there is considerable variation by type of case and by stage of litigation, one-third of the amount recovered is considered by most to be the general standard in personal injury litigation. 1 Robert L. Rossi, Attorneys' Fees, § 2:9 (3d ed. 2001). At least two of the named plaintiffs in this case did negotiate contracts for a straight contingency fee of one-third of the recovery. Michele Nilsen signed the contract in September, 2002, a month before the complaint was filed. At that time it appears that there was a general discussion regarding the possibility of filing the case as a class action. Nilsen Aff. ¶ 5. (The complaint was filed as a putative class action on October 15, 2002.) Later, in May, 2003, Michael Goodrich signed a contract for the same one-third contingency fee. Goodrich Aff. ¶ 2. Mr. Goodrich had learned about the case and believed that he was a class member. Id. at ¶ 1. At the time that he agreed to the fee, Mr. Goodrich also explicitly agreed to serve as a class representative, and learned of his obligations in this role. Id. at ¶ 2. Obviously such agreements bind neither the class nor the court, but they are of modest limited evidentiary value on what the market might produce. Nevertheless, the paramount question remains what the lawyer would hypothetically charge in a class action specifically. The class action provides for the possibility of a significantly larger recovery that, despite the inherent increased complexity, may induce the lawyer to take the case for a lower percentage charge. Thus, while this information is relevant, it does not compel a particular result.

Looking to other facets of contingent fee practice, I observe that Maine statutes limit the contingency fee in medical malpractice claims to what is ultimately a lower percentage. 24 M.R.S.A. § 2961(1) (2000) (in medical malpractice actions, total contingent fee that may be contracted is 33.33% of first $100,000 of recovery, 25% of next $100,000, and 20% of any amount over $200,000). For the amount recovered here, the blended fee would be 20.6%. Maine also places declining limits on attorney fees in proceedings before the Workers Compensation Board. 39-A M.R.S.A. § 325(4)(B) (2001) (if case tried to completion, fees may not exceed 30% of benefits accrued; for lump-sum settlements post-expenses, fees restricted to 10% of first $50,000, 9% through 6% of the next increments of $10,000, and 5% of any amount over $90,000). For the award requested here, the blended fee would be 5.1%. I am not aware of any other Maine caps on what a lawyer can charge a client.

For certain claims, Federal statutes also limit the amount that a lawyer may collect from a client. See generally Gisbrecht v. Barnhart, 535 U.S. at 802-03 (2002) (collecting representative statutes). For example, courts may approve contingency fees to a maximum of only 25% of past due benefits awarded to a successful plaintiff in Social Security proceedings. 42 U.S.C. § 406(b) (2003). The Federal Tort Claims Act allows courts to approve contingency fees from the successful plaintiff's recovery to a maximum of 25% if the fund is created through judgment, or 20% if through settlement. 28 U.S.C. § 2678; see also Veterans' Benefit Act, 38 U.S.C. § 5904(d)(1) (1994) (20% of past due benefit). Several other federal statutes are less generous and allow a maximum of only 10% of the recovery to go toward attorney fees. See, e.g., War Claims Act, 50 U.S.C. App. § 2017m (1990) (10% of award).

Both the state and federal percentage caps are instructive, for they represent constraints on the fee a lawyer may recover in certain legal arenas. Sometimes, the maximum allowable percentage then becomes the market norm in that arena, rather than merely the upper ceiling. See Gisbrecht, 535 U.S. at 800, 122 S.Ct. 1817 (contingent fee contracts for fees of 25%, maximum allowable percentage, are "most common fee arrangements" for Social Security claimants). At the same time, I recognize that there is no statutory cap for attorney fees in civil rights cases like this one.

(2) Awards in Other Cases. For other strip search class actions resulting in a common fund settlement, the lowest percentage awarded as a fee that I have discovered was 16% ($1 million of a $6.25 million common fund). Haney v. Miami-Dade County, No. 04-20516-CIV (S.D. Fla. judgment of dismissal Oct. 7, 2005); see also Kahler v. County of Rensselaer, No. 03-cv-01324 (N.D.N.Y. final approval of settlement Sep. 23, 2004) (fee of approximately $443,000 from total potential fund of $2.7 million, representing about 16%, but with no explanation as to method, whether lodestar or percentage). The highest percentage explicitly awarded by a court was 33.33%, or $3,833,333 of an $11.5 million fund. Eddleman v. Jefferson County, No. 3:91CV-144 (W.D. Ky. final approval of settlement May. 5, 1999). Other cases seem to be distributed fairly evenly within that range.

Median attorney fee awards in other class actions generally ( i.e., not limited to strip searches) range within a few percentage points on either side of 30%. Thus, a survey of 1,120 common fund class actions found that the median award of attorney fees and expenses was 31.6% of the fund for cases in which the class recovery ranged from $3 million to $5 million. Stuart J. Logan et al., Attorney Fee Awards in Common Fund Class Actions, 24 Class Action Rep. 167, 167 (2003). According to a recent Federal Judicial Center study of 621 class actions, an award of 29% for attorney fees and expenses was "typical." Thomas E. Willging Shannon R. Wheatman, An Empirical Examination of Attorneys' Choice of Forum in Class Action Litigation 52 (2005). An earlier Federal Judicial Center study found that median fee rates "ranged from 27% to 30%." Thomas E. Willging et al., Empirical Study of Class Actions in Four Federal District Courts 69 (1996). (3) The Award Here. Maine practice is certainly relevant. This is a Maine lawsuit, with a Maine defendant, and mostly Maine plaintiffs. Therefore, to the extent I know them, I consider both the general range of contingent fee agreements and the limits that Maine has placed statutorily on contingent fee recovery for certain kinds of cases.

I can glean here that the Maine legislature concluded that 20.6% is a reasonable fee for medical malpractice cases of this magnitude. It appears that lawyers continue to take such cases with that fee limitation. On the other hand, there is no limit for civil rights lawsuits like this one, and therefore the lawyers are able to bargain for a higher contingent fee in individual civil rights cases. It is hard to generalize which kind of lawsuit has the higher risk and the greater complexity. Since the 20.6% fee limit stays the same regardless of how much the recovery exceeds $200,000, the medical malpractice lawyer has the opportunity to generate a huge fee with a huge award that may not be available in civil rights class actions like this one. Fees, of course, are ordinarily negotiated at the outset of the litigation before the size of recovery is known. Thus, those factors suggest that the market fee hypothetically negotiated for a civil rights class action might be somewhat higher than the 20.6% effective rate for medical malpractice cases of this size.

The workers compensation limits are unduly low for a civil rights class action. Workers compensation is largely an administrative proceeding, generally without complex factual issues, extensive discovery, or difficult arguments such as class certification, liability is strict, and the legislature has tried to reduce lawyer involvement. The medical malpractice numbers are more instructive and persuasive; like class actions, that cause of action includes complex factual and legal issues, as well as substantial barriers to recovery. Therefore, I put little weight on the workers compensation fee limits.

The Federal numbers are somewhat relevant. This was primarily a federal cause of action with a statutory fee award available at a successful conclusion. However, the federal limits from other statutes may be too restrictive, for many of the claims at stake under those statutes are administrative and straightforward. Those limits may also reflect a legislative desire that attorney fees not diminish certain statutory entitlements. Therefore, I conclude that the higher end of the federal 10%-25% range is more relevant than the lower end. I also recognize as I did before that without a federal statutory cap for this type of case, the lawyers hypothetically are able to negotiate for a higher percentage. A primary difficulty with all these numbers is that lawyers customarily practice in one of these fields or another, not across fields. Thus, the numbers have limited relevance.

I also consider that two of the named plaintiffs agreed to one-third contingency fees, and that a one-third contingency fee agreement is widely recognized in personal injury litigation. I draw no firm conclusions from that data, because the two individual contracts were signed before class status existed, and the individuals, contemplating their own individual recovery, had little incentive to bargain for a lower fee that might be available in a class action. I have already explained why the general contingency fee agreement is of limited help for the peculiarities of class actions, supra, note 34.

Other judicial fee awards are relevant to some extent. Other courts' awards necessarily affect the expectations of lawyers and, therefore, what they might agree to in voluntary negotiation. See Vizcaino, 290 F.3d at 1050 (lawyers' reasonable expectations may be based on "range of fee awards out of common funds of comparable size"). What they show here is that lawyers continue to take such cases in the face of fee awards ranging from 16% to 33-1/3%. Human nature suggests, therefore, that lawyers taking such a case might hope for 33-1/3% but recognize that it might be as low as 16%.

The above analysis leads me to an attorney fee award here of 25% ($825,000). This percentage is at the higher end of the federal statutory limitation range, somewhat above the effective Maine malpractice limit, much higher than the Maine workers compensation percentage, lower than the standard one-third contingent fee and lower than the two fee agreements entered into before this became a class action. The percentage is in the mid-range of court awards in strip search cases, and it is more than 6% lower than the average award in general class actions with recoveries of this size, as reported by Logan, et al., supra, at 167.

I am not confident that 25% truly is a good approximation of the market-mimicking rate. I could be wrong in either direction. On the information I have, I just don't know. Ironically, therefore, on this record my use of the Seventh Circuit's market-mimicking analysis may not be much better than the multifactor approach. But I hope that it will generate better evidence of the attorney fee market in future cases and more rational and predictable awards.

My reduction from the lawyers' requested 30% is no reflection whatsoever upon the quality of representation, or the nature of this litigation; these lawyers were excellent. Further, it is not a criticism of the percentage that they requested, given the precedents that they cite from other jurisdictions.

In future class actions that involve court-awarded attorney fees from a common fund, I will expect to receive evidentiary materials bearing on what lawyers negotiate for contingent fees or otherwise in comparable cases. I may well request detail about the percentages that would apply at different settlement stages of the case. See In re Continental Illinois Sec. Litig, 962 F.2d at 569 ("[R]isk of loss varies over the life of a case."). I may also consider using my authority under Fed.R.Evid. 706 to appoint an expert to advise me on the attorney fee market. Manual for Complex Litigation, supra, § 14.231, n. 590 (recognizing courts' use of court-appointed experts to assist with attorney fee determinations). Finally, I may at least consider ordering lawyers to propose fee arrangements at the outset.

B. LITIGATION COSTS AND CLAIMS ADMINISTRATION EXPENSES

In class action litigation, a district court may award reasonable costs. Fed.R.Civ.P. 23(h); see also In re Fidelity/Micron Sec. Litig., 167 F.3d 735, 737 (1st Cir. 1999) (prior to adoption of Rule 23(h), holding that class counsel creating a common fund is entitled to recover "expenses, reasonable in amount, that were necessary to bring the action to a climax"). Here, the lawyers requested 3% of the $3.3 million fund for litigation costs and claims administration expenses, and agreed to cap this request and pay the remainder from their own fee award if I awarded them the full 30% in fees. Mem. 16. Because I have not awarded the requested 30%, I direct the lawyers to promptly file with the court documentation of accrued and projected litigation costs and claims administration expenses, which I will then review.

C. LATE CLAIMS

The period for submitting a claim form to the Administrator or an extension request to the Court expired on September 9, 2005. Settlement Agreement ¶ 31. Since that time, several individuals have written letters to the Court requesting an extension of the time in which to submit a claim. At the request of the plaintiffs' lawyers, I have extended that deadline to November 10, 2005. Order on Pending Requests by Class Members for Extension to File Settlement Claims 1 (Docket Item 182). I will await the claims administrator's decision on these claims and entertain all appeals at one time.

IV. CONCLUSION

I award 25% of the settlement in attorney fees, specifically $825,000. I also direct the lawyers to file with the Court prompt documentation of costs and expenses.

SO ORDERED.

APPENDIX I

1. The size of the fund created and the number of persons benefited. The common fund created by the settlement agreement totals $3.3 million, plus interest now approaching $70,000. It will be distributed in its entirety after payment of fees, costs, claims administration expenses, and incentive payments. Although not part of the common fund created, there is additional value in the requirement that York County maintain a written policy prohibiting challenged strip searches, value akin to that of injunctive relief, which will benefit future arrestees. Order on Mot. for Final Approval of Class Settlement ("Order") 11 (Docket Item 159).

There are approximately 7,500 class members. From claims filed it appears that with a participation rate of approximately 17.78%, over 1,300 persons will directly benefit from this common fund. Decl. of Myong J. Joun 1 (Docket Item 143). The participation rate is particularly noteworthy as it was revised upward from an earlier estimate of 12-13% during the class notice period, indicating that even more people will benefit than had previously been thought. Order 7-8.

2. The presence or absence of substantial objections. I have previously addressed the issue of objectors to the settlement. Due to the nature of this lawsuit and its potential for embarrassment in publicly stepping forward, I do not consider the relative absence of objections to settlement a significant factor. Order 11. 3. The lawyers' skill and efficiency and the quality of representation. I have previously noted the skill of class counsel in this case, and there is no need to recite this aspect again. See Order 12-13 (lawyers were "highly qualified and experienced," "effective and thorough," and "advocated zealously and capably for the plaintiffs"). I do take additional note of class counsel's "considerable and highly successful effort" to reach as many class members as possible during the notice period. Id. at 7. This effort resulted in an upwardly revised participation rate estimate, which in turn contributed directly to the number of people benefited by this common fund.

4. The complexity and duration of the litigation. This litigation lasted two years and occasioned extensive discovery. Order 12. There were significant factual disputes as to whether the corrections officers actually viewed the arrestees' naked bodies while they changed into Jail clothing. Id. at 14. There was additional complexity due to the individualized nature of the alleged dignitary harm, which could have made the damages stage of the case procedurally complicated if the defendant moved to decertify the class on the this issue. Id. Further, class certification was not guaranteed, and was in fact vigorously disputed by York County all the way through appeal. See Tardiff v. Knox County, 365 F.3d at 3. Finally, the parties had to sort through "complex issues regarding York County's insurance coverage" before reaching an agreement. Order 13.

5. Contingency and risk of nonpayment. Class counsel in this case faced a substantial risk of nonpayment. The Jail contested liability factually from the outset, and there was a "serious risk" that a jury would credit the corrections officers' testimony over that of the arrestees. Order 14. Even if liability were established, it would still have been difficult for the plaintiffs to prove and recover damages for harms that were often dignitary rather than pecuniary. Id. at 9; see also Mot. For Final Approval, Ex. B, Decl. of Charles Harvey, Esq., ¶ 10 (Docket Item 133) (noting his belief that class members may receive nothing or only very small or nominal damage awards from a jury; recovered fees would therefore be correspondingly less). York County was reluctant to engage in settlement discussions, and even after two lengthy mediation sessions, the two parties remained far apart. Order 13. Finally, the risk of nonpayment was increased substantially by the fact that the settlement was funded largely by insurance, and the amount available to fund it was finite and declined as York County spent more on defense costs. Id. at 10.

6. The amount of time devoted to the case by plaintiffs' counsel. There is no doubt that class counsel devoted a long period of time to this litigation. The lodestar figure (the product of reasonable hours worked times the reasonable hourly rate) shows that in June, 2005, class counsel estimated that it would spend in total an amount somewhat less than $520,000.00 worth of time on this case, at least 4,084 hours, over 2 years. Mem. 11-13. 7. The awards in similar cases. In the main body of this Order, I have compared at length the percentage awards in other cases. Suffice it to say that the cases that are most similar (strip search common fund settlements) range from 16% to 33%. See Appendix II.

I can assess the lodestar cross-check similarly. The 25% awarded here represents a lodestar multiplier of approximately 1.6. Of the analogous strip search class action cases I reviewed for which lodestar data was available, I found a multiplier range from 1.19 to 2.68. See Appendix II. For class actions generally, in 142 cases with a class recovery ranging from $3 million to $5 million the median multiplier was 1.89. Stuart J. Logan et al., supra, at 167. But these averages and medians are of limited utility because the multiplier appears to differ greatly from case to case, based on the size of the fund and the type of class action. See id. at 167, 196; see also 4 Conte Newberg, supra, § 14:6 (numbers "ranging from one to four are frequently awarded"); Walker Horwich, supra, at 1468 (initial hypothesis that courts would award reasonable fees with multipliers ranging from 1.5 to 2.5, but actual empirical research showed ranges between 1.0 to 5.0).

The 1.6 multiplier in this case is lower than all but one of the strip search cases, and lower than the average multiplier of 1.89 found in class actions generally, and thus it is not unreasonable.

I conclude that under a multifactor approach to determining the reasonableness of a percentage fee, the preceding factors fully support an award of 25% from a common fund of $3.3 million, as well as many other percentages.

APPENDIX II

1. Haney v. Miami-Dade County, No. 04-20516-CIV (S.D. Fla. judgment of dismissal Oct. 7, 2005).

The fee in this case was awarded after a little more than a year and a half of litigation including discovery, a preliminary fairness hearing, class notice, and the court's final approval of the settlement. The district judge approved the lawyers' requested 16% award ($1 million) from the common fund settlement of $6.25 million. The lodestar was reported as $535,000, resulting in a multiplier of 1.74.

2. Bynum v. District Columbia, 384 F.Supp.2d 342, 352 (D.D.C. preliminary approval of settlement Aug. 31, 2005).

I draw no conclusions from this case because the proposed settlement merely went out for review by the class on August 31, 2005, after more than two years of litigation including discovery and class certification. In the settlement, class counsel seeks 33% ($4.0 million) from the proposed common fund of $12 million. In the fee petition, the lawyers give the total projected lodestar as $2,099,420, resulting in a lodestar multiplier of 1.91.

3. Kahler v. County of Rensselaer, No. 03-cv-01324 (N.D.N.Y. final approval of settlement Sep. 23, 2004).

This case consolidated two actions, and settled after less than a year including class certification and preliminary approval. The settlement agreement created a common fund in an amount contingent upon the number of claims submitted, not to exceed a total obligation upon the defendant of $2.7 million. The defendant was to deposit $1,000 for every timely claim, as well as fees and costs determined by the court. The claims administrator declared that 835 timely claims were submitted.

The lawyers requested 20% ($580,000) of the defendant's maximum obligation of $2.7 million. The court reduced the amount to $442,702 (without a written reason), which represented 16.4% of the total potential liability that the defendant could have incurred.

But because of a reversion of unclaimed amounts, the total fund was ultimately much lower, approximately $1,282,702 ($835,000 for claimants plus $442,702 for the fees). The lawyers therefore received approximately 34.5% of the fund. The court may not have considered this outcome when it made the fee award.

The lodestar was reported as approximately $371,396, resulting in a multiplier of 1.19.

4. Connor v. Plymouth County, No. 00-10835-RBC (D. Mass. final approval of settlement Mar. 11, 2004).

After almost four years of litigation including discovery, class certification, denial of a decertification motion, preliminary approval, and class notice, this case was settled by the creation of a $1.35 million fund. The court awarded the lawyers' requested 31% ($418,500) in fees. The lodestar in this case was approximately $155,813, resulting in a multiplier of 2.68.

5. Williams v. Block, No. 97-03826-CW (C.D. Ca. final approval of settlement Nov. 27, 2002).

This case consolidated more than a dozen actions, and settled creating a common fund of $21.5 million, after more than five years of litigation including discovery, numerous interlocutory appeals by both parties, denial of class certification several times, and several motions to dismiss. The court awarded 20% ($4,184,000) of the fund, although the lawyers had requested a fee of 25%.

The lodestar was $4,260,643, resulting in a multiplier of 0.98.

6. Mack v. Suffolk County, No. 98-12511-NG (D. Mass. final approval of settlement Oct. 1, 2002).

This case settled after more than five years of litigation including discovery, class notice, preliminary approval, and summary judgment including a judicial determination of full liability. Settlement (at which the only remaining issue was damages) created a common fund of $10 million. The court awarded fees in the amount of the requested 30%, or $3.0 million. The lodestar was estimated to be approximately $645,635, resulting in a multiplier of 4.65.

7. Doan v. Watson, 2002 WL 31730917 (S.D. Ind. 2002) (No. 99-4-C-B/S).

The settlement agreement in this case created a common fund in a contingent amount equal to $1,000 for each class member. The fund was exclusive of attorney fees. It appears that there were approximately 2,600 class members; thus the total amount contributed by the defendant to the fund was apparently $2.6 million. However, only about 630 actual claims appeared to be filed at the cutoff date. (It is unclear what happened with the fund's remaining $1.97 million.)

The court approved fees and expenses to be paid by defendant in addition to the settlement fund, in the amount of $700,000, at a time when the judge expected the fund to be $2.7 million. That would represent a 21% fee of the defendant's total potential liability.

If actual payout by the defendant was 1.33 million ($630,000 for the claimants plus $700,000 for the attorneys), then attorney fees represented over 50% of that total.

However, because the defendants paid the attorneys separately and in addition to the claimants, and the court did not evaluate the fees in relation to the settlement fund, it is difficult to determine where this case fits.

8. Gary v. Sheahan, No. 96-C-7294 (N.D. Ill. 1998).

At first glance this case appeared to be governed by common fund principles. The court awarded fees after five years of protracted litigation and summary judgment for the class on the merits. The parties subsequently settled the damages, creating a fund of $6,882,500, exclusive of attorney fees.

The court then held a separate proceeding to determine the fees that the defendant would have to pay, above and beyond the fund amount. The court awarded $3.0 million to class counsel (without a written explanation as to the basis for this amount).

Thus, this was not a common fund case (although there is confusing contrary language in the order).

9. Moser v. Anderson, No. C-93-634-B (D.N.H. final settlement approval Jan. 13, 2000).

The lawyers here assert that in this case the judge awarded 33.3% of a $3.0 million settlement in attorney fees. Joun Aff. ¶ 39. However, the Joint Stipulation of Settlement attached as Exhibit 15 to the Memorandum does not contain this information.

10. Maneely v. City of Newbergh, No. 01-CV-02600, 256 F.Supp.2d 204 (S.D.N.Y. stipulation of discontinuance and dismissal entered Sep. 6, 2005).

The lawyers here assert that this case is relevant to my fee award, and filed the Order for Preliminary Approval of Settlement in the case. It appears from this document that the case settled after more than four years of discovery, class certification, and preliminary approval, with the funding by the defendant of a settlement fund of $1.2 million, exclusive of attorney fees. Each eligible member (not a named plaintiff) is to receive $1,000 from this fund, and the rest will revert to the defendant. There is no information available for the number of claims processed.

The court awarded $550,000 in attorney fees, to be paid directly by the defendant (without a written explanation as to the basis for the award). Information about the amount requested by the lawyers is unavailable. The fee award represents 31.4% of the defendant's total potential liability, but the reversion of unclaimed amounts may make the ultimate percentage much higher.

The fee award does not appear to be governed by common fund principles, and thus I find it of little assistance.

12. Eddleman v. Jefferson County, No. 3:91CV-144-J (W.D. Ky. final approval of settlement May 24, 1999).

This case was a consolidated action of numerous cases. After nearly eight years including class certification, discovery, interlocutory appeals, summary judgment motions, it settled with a common fund of $11.5 million. This was to be the defendant's total potential liability, and any unclaimed money would revert to the defendant. At the time of final approval, the court awarded attorney fees in the amount of 33.33% (equaling approximately $3,833,333) of the total fund. Because of the reversion of unclaimed funds to the defendant, the actual percentage awarded as a fee may have been higher.