LAFFITTE v. ROBERT HALF INTERNAT., INC.

OPINION

WERDEGAR, J. —

A class action employment lawsuit settled before trial for $19 million, with the agreement that no more than a third of that recovery would go to class counsel as attorney fees. In seeking the trial court's approval of the settlement, class counsel sought the maximum fee amount, $6,333,333.33. After considering information from class counsel on the hours they had worked on the case, applicable hourly fees, the course of the pretrial litigation, and the potential recovery and litigation risks involved in the case, the trial court — over the objection of one class member — approved the settlement and awarded counsel the requested fee.

The objecting class member contends the trial court's award of an attorney fee calculated as a percentage of the settlement amount violates a holding of this court in Serrano v. Priest (1977) 20 Cal.3d 25 [141 Cal.Rptr. 315, 569 P.2d 1303] (Serrano III),1 to the effect that every fee award must be calculated on the basis of time spent by the attorney or attorneys on the case. (See Serrano III, at p. 48, fn. 23.) We disagree. Our discussion in Serrano III of how a reasonable attorney fee is calculated was made in connection with an award under the "private attorney general" doctrine. (See id. at pp. 43-47.) We clarify today that when an attorney fee is awarded out of a common fund preserved or recovered by means of litigation (see Serrano III, supra, at p. 35), the award is not per se unreasonable merely because it is calculated as a percentage of the common fund.

FACTUAL AND PROCEDURAL BACKGROUND

Three related wage-and-hour class action lawsuits were filed against Robert Half International Inc., a staffing firm, and related companies (hereafter collectively Robert Half) in Los Angeles County Superior Court. In September 2012, the parties jointly moved for an order conditionally certifying a settlement class and preliminarily approving a settlement. The trial court granted the motion and preliminarily approved the settlement. With the court's permission, the proposed settlement was amended in November 2012.

Under the settlement agreement as amended, Robert Half would pay a gross settlement amount of $19 million. It was agreed class counsel would request attorney fees of not more than $6,333,333.33 (one-third of the gross settlement amount), to be paid from the settlement amount. Robert Half would not oppose a fee request up to that amount, and if a smaller amount was approved by the court the remainder would be retained in the settlement amount for distribution to claimants, rather than reverting to Robert Half. The settlement agreement further provided that any unclaimed portion of the net settlement amount (resulting, for example, from class members choosing not to make claims or failing to qualify for compensation) would be reallocated to qualified claimants rather than returned to Robert Half or given to any third party.

Class member David Brennan objected to the proposed settlement on several grounds, including that the projected $6,333,333.33 attorney fee appeared to be excessive and class counsel had not provided enough information to evaluate it.

Class counsel subsequently made the anticipated request for $6,333,333.33 in attorney fees. A fee equal to one-third of the settlement fund recovered for the class, counsel asserted, is within a historical range of 20 to 50 percent of a common fund and is also within the range provided in contingent fee agreements signed by the named plaintiffs. Recovery of any fee was contingent on success in the litigation, "and the case presented far more risk than the usual contingent fee case." The requested fee, counsel also asserted, is also appropriate under the "lodestar" method, in which an attorney fee is based on the hours worked and an hourly billing rate, sometimes adjusted by a positive or negative multiplier. The firms acting as class counsel would collectively expend between 4,263 and 4,463 attorney hours, depending on whether the objector appealed approval of the settlement. Multiplying the individual attorneys' hours by rates assertedly tied to their skill and experience, counsel calculated a lodestar fee amount of between $2,968,620 and $3,118,620. The multiplier needed to reach the requested fee of $6,333,333.33 would thus be 2.03 to 2.13.

The totals of hours expended, the range of percentages in common fund cases and in the fee agreements, and the range of hourly rates applicable to class counsel were supported by data in the fee motion and supporting declarations. Class Counsel Kevin T. Barnes generally described the work performed in "one of the most heavily litigated cases I have ever been a part of and the extensive research and litigation for the past 8 ½ years. This litigation included extensive written discovery, extensive law and motion practice, 68 depositions, three Motions for Summary Judgment, a Class Certification Motion, subsequent Reconsideration Motion and then another Motion to Decertify, numerous experts, consultation with an economist regarding potential damage exposure and two full day mediations."

While tentatively approving the settlement and fee request, the trial court asked counsel for additional information and discussion on certain points. Barnes submitted a supplemental declaration that, in part, argued the calculated multiplier over the lodestar amount (2.03 to 2.13) was reasonable in light of counsel's "hard work and determination" in a difficult case and the "enormous" risks of nonpayment counsel undertook. Barnes's declaration detailed the risks that the actions would fail at the certification stage, would be deemed barred by arbitration agreements, or would fail on the merits because of findings the class members were exempt employees.

On April 10, 2013, the trial court overruled Brennan's objections and gave the settlement and attorney fee request its final approval. In its oral ruling the court stated: "On the amount of the attorneys fees, the court considers in this case that there is a contingency case, and so I do a double check on the attorneys fees by looking at the lodestar amount. I do believe I have sufficient information on the number of hours that were present and that the hourly rates charged therefore were within the norm and not overstated. [¶] Given the lodestar, I then also find I have information in the record which supports the multiplier that would be applied to lodestar if you're looking at a strict lodestar calculation, which we're not, we're looking at a contingency calculation, the amount of the contingency is not unreasonable. I'm considering the novelty and difficulty of the questions involved, the skill displayed in presenting them, the extent to which the litigation precluded other employment by the attorneys and the inherent risk whenever there is a fee award that is contingent. [¶] On that basis, I am granting final approval."

On objector Brennan's appeal from the judgment entered on the settlement, the Court of Appeal affirmed. The Court of Appeal held Serrano III did not preclude award of a percentage fee in a common fund case, that an award of one-third the common fund was in the range set by other class action lawsuits, and that the trial court did not abuse its discretion by cross-checking the reasonableness of the percentage award by calculating a lodestar fee and approving a multiplier over lodestar of 2.03 to 2.13.

We granted review on the objector's petition, which presented a single issue: whether Serrano III permits a trial court to calculate an attorney fee award from a class action common fund as a percentage of the fund, while using the lodestar-multiplier method as a cross-check of the selected percentage.2

DISCUSSION

We review attorney fee awards on an abuse of discretion standard. "The `experienced trial judge is the best judge of the value of professional services rendered in his court, and while his judgment is of course subject to review, it will not be disturbed unless the appellate court is convinced that it is clearly wrong.'" (Serrano III, supra, 20 Cal.3d at p. 49.) "Fees approved by the trial court are presumed to be reasonable, and the objectors must show error in the award." (Consumer Privacy Cases (2009) 175 Cal.App.4th 545, 556 [96 Cal.Rptr.3d 127].) We consider here whether a trial court abuses its discretion, when awarding a fee from a common fund created or preserved by the litigation, by calculating the fee as a percentage of the fund and checking the reasonableness of the fee with a lodestar calculation.

(1) California has long recognized, as an exception to the general American rule that parties bear the costs of their own attorneys, the propriety of awarding an attorney fee to a party who has recovered or preserved a monetary fund for the benefit of himself or herself and others. In awarding a fee from the fund or from the other benefited parties, the trial court acts within its equitable power to prevent the other parties' unjust enrichment. (Serrano IV, supra, 32 Cal.3d at p. 627; Serrano III, supra, 20 Cal.3d at p. 35; Farmers etc. Nat. Bank v. Peterson (1936) 5 Cal.2d 601, 607 [55 P.2d 867]; Fox v. Hale & Norcross S. M. Co. (1895) 108 Cal. 475, 476-477 [41 P. 328]; Lealao v. Beneficial California, Inc. (2000) 82 Cal.App.4th 19, 27 [97 Cal.Rptr.2d 797] (Lealao).)

Because it distributes the cost of hiring an attorney among all the parties benefited, a common fund fee award has sometimes been referred to as "fee spreading." In contrast, "fee shifting" refers to an award under which a party that did not prevail in the litigation is ordered to pay fees incurred by the prevailing party. (Lealao, supra, 82 Cal.App.4th at p. 26; Camden I Condominium Assn. v. Dunkle (11th Cir. 1991) 946 F.2d 768, 774.) California law permits fee shifting in favor of the prevailing party on certain statutory causes of action (e.g., Gov. Code, §§ 12965, subd. (b), 12974, 12989.2), when a plaintiff has acted as a private attorney general by enforcing an important right affecting the public interest (Code Civ. Proc., § 1021.5), and in contract cases where the contract provides for an award of fees to the prevailing party (Civ. Code, § 1717).

(2) Class action litigation can result in an attorney fee award pursuant to a statutory fee shifting provision or through the common fund doctrine when, as in this case, a class settlement agreement establishes a relief fund from which the attorney fee is to be drawn. Two primary methods of determining a reasonable attorney fee in class action litigation have emerged and been elaborated in recent decades. The percentage method calculates the fee as a percentage share of a recovered common fund or the monetary value of plaintiffs' recovery. The lodestar method, or more accurately the lodestar-multiplier method, calculates the fee "by multiplying the number of hours reasonably expended by counsel by a reasonable hourly rate. Once the court has fixed the lodestar, it may increase or decrease that amount by applying a positive or negative `multiplier' to take into account a variety of other factors, including the quality of the representation, the novelty and complexity of the issues, the results obtained, and the contingent risk presented." (Lealao, supra, 82 Cal.App.4th at p. 26.)

The two approaches to determining a fee contrast in their primary foci: "The lodestar method better accounts for the amount of work done, while the percentage of the fund method more accurately reflects the results achieved." (Rawlings v. Prudential-Bache Properties, Inc. (6th Cir. 1993) 9 F.3d 513, 516.) Each has been championed and criticized for its respective advantages and disadvantages. The lodestar method has been praised as providing better accountability and encouraging plaintiffs' attorneys to pursue marginal increases in recovery, but criticized for discouraging early settlement and consuming too large an amount of judicial resources in its application. (Id. at pp. 516-517; 5 Rubenstein, Newberg on Class Actions (5th ed. 2015) § 15:65, pp. 225-226 (hereafter Newberg on Class Actions).) The percentage method "is easy to calculate; it establishes reasonable expectations on the part of plaintiffs' attorneys as to their expected recovery; and it encourages early settlement, which avoids protracted litigation. However, a percentage award may also provide incentives to attorneys to settle for too low a recovery because an early settlement provides them with a larger fee in terms of the time invested." (Rawlings v. Prudential-Bache Properties, Inc., supra, at p. 516.) Where the class settlement is for a very large amount, a percentage fee may be criticized as providing counsel a windfall in relation to the amount of work performed. (Brundidge v. Glendale Federal Bank, F.S.B. (1995) 168 Ill.2d 235, 243 [213 Ill.Dec. 563, 659 N.E.2d 909]; 5 Newberg on Class Actions, supra, § 15:65, p. 224.)

Before discussing the percentage method's use in California, we review the history of the two fee calculation approaches in class action litigation nationally.

I. Lodestar-multiplier v. Percentage of the Recovery

The history of attorney fee awards in class actions has been one of reaction and counterreaction, divisible into three major eras. (See Walker & Horwich, The Ethical Imperative of a Lodestar Cross-check: Judicial Misgivings About "Reasonable Percentage" Fees in Common Fund Cases (2005) 18 Geo. J. Legal Ethics 1453, 1453-1454 (hereafter Walker & Horwich).)

In the first period, from the 1966 amendments to rule 23 of the Federal Rules of Civil Procedure (28 U.S.C.), which "heralded the advent of the modern class action" (Walker & Horwich, supra, 18 Geo. J. Legal Ethics at p. 1453), to the middle of the 1970s, awards based on a percentage of the recovery were common: "Judges relied on a variety of factors in setting reasonable amounts for fee awards, but most heavily emphasized was the size of the fund or the amount of benefit produced for the class. Awards often reflected what the court believed was a `reasonable percentage' of the amount recovered, with the percentages varying considerably from case to case. However, the percentage-of-recovery system sometimes resulted in strikingly large fee awards in a number of cases. Press reaction to these awards, and criticism from within the profession that the fees were disproportionate to the actual efforts expended by the attorneys, generated pressure to shift away from the percentage-of-recovery approach." (Court Awarded Attorney Fees: Report of the Third Circuit Task Force (1985) 108 F.R.D. 237, 242 (hereafter 1985 Task Force Report); see In re Thirteen Appeals Arising Out of San Juan Dupont Plaza Hotel Fire Litigation (1st Cir. 1995) 56 F.3d 295, 305 ["Traditionally, counsel fees in common fund cases were computed as a percentage of the fund, subject, of course, to considerations of reasonableness."].)

The second period ran from the Third Circuit's Lindy decisions in the mid' 1970s (Lindy Bros. Builders, Inc. v. American Radiator & Standard Sanitary Corp. (3d Cir. 1973) 487 F.2d 161 (Lindy I) and Lindy Bros. Builders, Inc. v. American Radiator & Standard Sanitary Corp. (3d Cir. 1976) 540 F.2d 102), which described and mandated the use of a lodestar-multiplier method in common fund class action cases in the Third Circuit, to the middle of the 1980s. In this period, the lodestar-multiplier method predominated in federal courts in fee spreading as well as fee shifting cases. The virtue of using a lodestar to determine fees, the court explained in Lindy I, is its seemingly direct relationship to the value of the services rendered: "[W]e stress ... the importance of deciding, in each case, the amount to which attorneys would be entitled on the basis of an hourly rate of compensation applied to the hours worked. This figure provides the only reasonably objective basis for valuing an attorney's services." (Lindy I, supra, 487 F.2d at p. 167.) Quoting from a district court decision, Lindy I expressed the fear "`that the bar and bench will be brought into disrepute, and that there will be prejudice to those whose substantive interests are at stake,'" if fee awards were not restrained by reference to the actual time spent and skill displayed by counsel. (Id. at p. 168.)

"The Lindy lodestar approach rather quickly gained acceptance in other federal courts throughout the country because it was viewed as a more reasonable approach than the percentage-of-benefit technique for making fee awards in modern complex litigation." (1985 Task Force Report, supra, 108 F.R.D. at p. 244.) Several federal appellate courts mandated use of the lodestar-multiplier method even in cases where class litigation had resulted in establishment of a common fund. (See, e.g., National Treasury Employees Union v. Nixon (D.C. Cir. 1975) 521 F.2d 317, 322; City of Detroit v. Grinnell Corp. (2d Cir. 1977) 560 F.2d 1093, 1098-1099; Grunin v. International House of Pancakes (8th Cir. 1975) 513 F.2d 114, 127.) In statutory fee shifting cases, where the prevailing party's fees are ordered paid by the nonprevailing party, the lodestar method was generally adopted, with United States Supreme Court approval. (Hensley v. Eckerhart (1983) 461 U.S. 424, 433 [76 L.Ed.2d 40, 103 S.Ct. 1933]; 5 Newberg on Class Actions, supra, § 15:38, pp. 124-129.)

The third period, which continues today, began in the mid-1980s. In 1984, in a statutory fee shifting case involving a lodestar-multiplier calculation, the United States Supreme Court distinguished common fund cases and indicated a different method would be used in such a case: "Unlike the calculation of attorney's fees under the `common fund doctrine,' where a reasonable fee is based on a percentage of the fund bestowed on the class, a reasonable fee under [42 United States Code section] 1988 reflects the amount of attorney time reasonably expended on the litigation." (Blum v. Stenson (1984) 465 U.S. 886, 900, fn. 16 [79 L.Ed.2d 891, 104 S.Ct. 1541].)

The next year, the Chief Judge of the Third Circuit convened a "task force" of judges, academics and attorneys from around the country to address "perceived deficiencies and abuses" that had arisen in the application of the Lindy lodestar method. (1985 Task Force Report, supra, 108 F.R.D. at p. 253.) The task force noted the main complaints that had been lodged against the lodestar method of determining an appropriate fee award. Prominent among these were that the emphasis on the number of hours worked creates a disincentive for the early settlement of cases and encourages lawyers to expend excessive hours; that the need for documentation and examination of detailed billing records had greatly increased the time and effort devoted to fee matters; and that the lodestar-multiplier method was neither as objective nor as precise as it appears facially because, for example, many plaintiffs' attorneys usually work on a contingency fee basis, making the assignment of a customary billing rate for lodestar purposes problematic. (1985 Task Force Report, supra, 108 F.R.D. at pp. 246-248.)

Distinguishing between fee spreading cases in which the fee award is to be taken from a common fund (including a class action settlement fund involving absent class members), and statutory fee shifting cases in which the award is a product of an adversary proceeding between the prevailing and nonprevailing parties (1985 Task Force Report, supra, 108 F.R.D. at pp. 250-251), the task force recommended courts generally use a percentage-of-the-fund method in common fund cases and a lodestar-multiplier method in fee shifting cases. "Accordingly, the Task Force recommends that in the traditional common-fund situation and in those statutory fee cases that are likely to result in a settlement fund from which adequate counsel fees can be paid, the district court, on motion or its own initiative and at the earliest practicable moment, should attempt to establish a percentage fee arrangement agreeable to the Bench and to plaintiff's counsel. In statutory fee cases the negotiated fee would be applied in the event of settlement; in all fully litigated statutory fee cases the award would continue to be determined in an adversary manner under the basic Lindy approach," with suggested modifications. (Id. at pp. 255-256, fn. omitted.)

By making a percentage fee award (which the task force envisioned being set early in the proceedings) in a common fund case, "any and all inducement or inclination to increase the number of Lindy hours will be reduced, since the amount of work performed will not be permitted to alter the contingent fee." (1985 Task Force Report, supra, 108 F.R.D. at p. 258.) Plaintiffs' counsel will have "a substantial inducement ... to settle the matter quickly, since the fee scale will have been established and counsel's compensation will not be enhanced by a delay." (Ibid.) Moreover, the percentage method "will eliminate the cumbersome, enervating, and often surrealistic process of preparing and evaluating fee petitions that now plagues the Bench and Bar under Lindy." (Ibid.) The lodestar method would, under the task force recommendations, continue to be used in statutory fee cases in which no common economic benefit, or only a fund insufficient to yield a reasonable fee, has been or is likely to be produced. (Id. at p. 259.)

In the years since the 1985 Task Force Report was released, the views expressed in it have gained general acceptance in federal and state courts. (See Walker & Horwich, supra, 18 Geo. J. Legal Ethics at pp. 1457-1458.) The Third Circuit itself holds that while both methods of calculating a fee may be used, "[t]he percentage-of-recovery method is generally favored 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. Securities Litigation (3d Cir. 2005) 396 F.3d 294, 300.) Currently, all the circuit courts either mandate or allow their district courts to use the percentage method in common fund cases; none require sole use of the lodestar method. (5 Newberg on Class Actions, supra, § 15.66, pp. 228-231.)3 Most state courts to consider the question in recent decades have also concluded the percentage method of calculating a fee award is either preferred or within the trial court's discretion in a common fund case.4 Thus, "[i]n the years since the Third Circuit's report ... federal and state courts alike have increasingly returned to the percent-of-fund approach [in common fund cases], either endorsing it as the only approach to use, or agreeing that a court should have flexibility to choose between it and a lodestar approach, depending on which method will result in the fairest determination in the circumstances of a particular case." (Strawn v. Farmers Ins. Co. of Oregon, supra, 297 P.3d at p. 446.)

The American Law Institute has also endorsed the percentage method's use in common fund cases, with the lodestar method reserved mainly for awards under fee shifting statutes and where the percentage method cannot be applied or would be unfair due to specific circumstances of the case. (ALI, Principles of the Law of Aggregate Litigation (2010) § 3.13.) "Although many courts in common-fund cases permit use of either a percentage-of-the-fund approach or a lodestar (number of hours multiplied by a reasonable hourly rate), most courts and commentators now believe that the percentage method is superior. Critics of the lodestar method note, for example, the difficulty in applying the method and cite the undesirable incentives created by that approach — i.e., a financial incentive to extend the litigation so that the attorneys can accrue additional hours (and thus, additional fees). Moreover, some courts and commentators have criticized the lodestar method because it gives counsel less of an incentive to maximize the recovery for the class." (Id., com. b.)

(3) While the percentage method has been generally approved in common fund cases, courts have sought to ensure the percentage fee is reasonable by refining the choice of a percentage or by checking the percentage result against a lodestar-multiplier calculation. (Walker & Horwich, supra, 18 Geo. J. Legal Ethics at pp. 1458-1461; 5 Newberg on Class Actions, supra, § 15:72, pp. 247-250.)

Some courts have employed a benchmark percentage, with upward or downward adjustments justified by a multifactor analysis. The Ninth Circuit has approved a 25 percent benchmark. (See Vizcaino v. Microsoft Corp. (9th Cir. 2002) 290 F.3d 1043, 1047 [approving 28 percent fee as justified by a benchmark of 25 percent adjusted according to specified case circumstances]; accord, In re Bluetooth Headset Products Liability Litigation (9th Cir. 2011) 654 F.3d 935, 942 [district courts in the circuit "typically calculate 25% of the fund as the `benchmark' for a reasonable fee award, providing adequate explanation in the record of any `special circumstances' justifying a departure"].) The Eleventh Circuit, similarly, stated in 1991 that "district courts are beginning to view the median of this 20% to 30% range, i.e., 25%, as a `bench mark' percentage fee award which may be adjusted in accordance with the individual circumstances of each case...." (Camden I Condominium Assn. v. Dunkle, supra, 946 F.2d at p. 775; see also Faught v. American Home Shield Corp. (11th Cir. 2011) 668 F.3d 1233, 1242 ["this court has often stated that the majority of fees in these cases are reasonable where they fall between 20-25% of the claims."].)

Other courts have mandated or suggested a sliding scale approach, an idea suggested by the Third Circuit's 1985 task force, in which the award in cases of larger recoveries is limited to a lower percentage to account for supposed economies of scale in litigating larger claims. (1985 Task Force Report, supra, 108 F.R.D. at p. 256; see, e.g., In re Cendant Corp. PRIDES Litigation (3d Cir. 2001) 243 F.3d 722, 736 ["[D]istrict courts setting attorneys' fees in cases involving large settlements must avoid basing their awards on percentages derived from cases where the settlement amounts were much smaller."].) As the court in Silverman v. Motorola Solutions, Inc. (7th Cir. 2013) 739 F.3d 956, 959, put the theory, "it is almost as expensive to conduct discovery in a $100 million case as in a $200 million case.... There may be some marginal costs of bumping the recovery from $100 million to $200 million, but as a percentage of the incremental recovery these costs are bound to be low. It is accordingly hard to justify awarding counsel as much of the second hundred million as of the first."

A further refinement of the sliding scale, championed in the Seventh Circuit, applies the lower percentages to the marginal amounts of the award over each step point. "Awarding counsel a decreasing percentage of the higher tiers of recovery enables them to recover the principal costs of litigation from the first bands of the award, while allowing the clients to reap more of the benefit at the margin (yet still preserving some incentive for lawyers to strive for these higher awards)." (Silverman v. Motorola Solutions, Inc., supra, 739 F.3d at p. 959.) Even without a well-developed sliding scale approach, some courts have approved fee awards representing a small percentage of the fund in cases involving very large settlements, the so' called "megafunds," in view of the "windfall" that would otherwise accrue to counsel. (See, e.g., Wal-Mart Stores, Inc. v. Visa U.S.A., Inc. (2d Cir. 2005) 396 F.3d 96, 103, 123 [where settlement fund was worth $3.05 billion, "the sheer size of the instant fund makes a smaller percentage appropriate"]; In re Bluetooth Headset Products Liability Litigation, supra, 654 F.3d at p. 942 ["[W]here awarding 25% of a `megafund' would yield windfall profits for class counsel in light of the hours spent on the case, courts should adjust the benchmark percentage or employ the lodestar method instead."].)5

(4) The most significant trend has been a blending of the two fee calculation methods, an approach in which one method is used to confirm or question the reasonableness of the other's result. Where the court uses the percentage method as its primary approach, the technique is referred to as a "lodestar cross-check," and has been described as follows: "First, the court computes a fee using the percentage method in the traditional manner, using a benchmark fee and adjustments as appropriate. Next, the court computes the fee using the lodestar method (absent any multiplier) in the traditional manner as described in Lindy I. At this point, the percentage-based fee will typically be larger than the lodestar-based fee. Assuming that one expects rough parity between the results of the percentage method and the lodestar method, the difference between the two computed fees will be attributable solely to a multiplier that has yet to be applied. Stated another way, the ratio of the percentage-based fee to the lodestar-based fee implies a multiplier, and that implied multiplier can be evaluated for reasonableness. If the implied multiplier is reasonable, then the cross-check confirms the reasonableness of the percentage-based fee; if the implied multiplier is unreasonable, the court should revisit its assumptions." (Walker & Horwich, supra, 18 Geo. J. Legal Ethics at p. 1463.)

Many federal circuits encourage or allow their district courts to conduct a lodestar cross-check on a percentage fee award (5 Newberg on Class Actions, supra, § 15:88, pp. 343-344),6 and empirical studies show the percentage method with a lodestar cross-check "is the most prevalent form of fee method" in practice. (Newberg, § 15:89, p. 348; see also Walker & Horwich, supra, 18 Geo. J. Legal Ethics at pp. 1461-1463.) We will return to the subject of lodestar cross-checks later, in reviewing the fee calculation in this case, which included such a cross-check. First, we address the use of the percentage method to calculate class action fee awards in California courts.

II. California Law After Serrano III

"Prior to 1977, when the California Supreme Court decided Serrano III, supra, 20 Cal.3d 25, California courts could award a percentage fee in a common fund case. (See, e.g., Melendres v. City of Los Angeles (1975) 45 Cal.App.3d 267, 284 [119 Cal.Rptr. 713].) After Serrano III, it is not clear whether this may still be done. (See Dunk v. Ford Motor Co. (1996) 48 Cal.App.4th 1794, 1809 [56 Cal.Rptr.2d 483] [`The award of attorney fees based on a percentage of a "common fund" recovery is of questionable validity in California.'].)" (Lealao, supra, 82 Cal.App.4th at p. 27.) Below, we clarify that Serrano III does not preclude award of a percentage fee in a common fund case.

In Serrano III, we reviewed an award of fees to attorneys who had obtained a judgment, affirmed in our Serrano II decision, that required reform of California's public school financing system to bring it into constitutional compliance. (Serrano III, supra, 20 Cal.3d at pp. 31-32.) The trial court had made the award on a private attorney general theory, rejecting reliance on the common fund and substantial benefit theories. (Id. at p. 33.)

(5) This court first addressed the common fund theory, under which "`when a number of persons are entitled in common to a specific fund, and an action brought by a plaintiff or plaintiffs for the benefit of all results in the creation or preservation of that fund, such plaintiff or plaintiffs may be awarded attorney's fees out of the fund.'" (Serrano III, supra, 20 Cal.3d at p. 34.) We agreed with the trial court that this equitable theory was inapplicable to the case because the plaintiffs had not, by their successful litigation efforts, "created or preserved any `fund' of money to which they should be allowed recourse for their fees." (Id. at p. 36.) To the extent the Legislature allocated additional moneys for public education in order to implement reforms, such expenditures were not required by the judgment itself and counsel did not propose their fee be paid out of any such increased expenditures. (Id. at pp. 36-37.) We went on to reject the theory of substantial benefit on similar grounds, explaining that "concrete `benefits' can accrue to the state or its citizens in the wake of Serrano [I & II] only insofar as the Legislature, in its implementation of the command of equality which that case represents, chooses to bestow them." (Id. at p. 41.)

We approved the fee award, however, under the private attorney general theory. We held that a fee award was within the trial court's equitable powers at least where the litigation had vindicated a public policy grounded in the California Constitution, the benefits flowed to a large number of Californians, and the nature of the litigation justified subsidizing the plaintiffs' efforts. (Serrano III, supra, 20 Cal.3d at pp. 46-47; see also Code Civ. Proc., § 1021.5 [codifying the private attorney general doctrine].)

Considering the amount of the fee, we rejected the contention by one of the firms representing the plaintiffs that it was inadequate in light of the circumstances. We explained that the trial court had considered the relevant circumstances in calculating a reasonable fee, using what would now be called a lodestar-multiplier method: "Fundamental to its determination — and properly so — was a careful compilation of the time spent and reasonable hourly compensation of each attorney and certified law student involved in the presentation of the case." (Serrano III, supra, 20 Cal.3d at p. 48, fn. omitted.)7 In the omitted footnote (originally numbered 23), we further addressed fee calculation: "We are of the view that the following sentiments of the United States Court of Appeals for the Second Circuit, although uttered in the context of an antitrust class action, are wholly apposite here: `The starting point of every fee award, once it is recognized that the court's role in equity is to provide just compensation for the attorney, must be a calculation of the attorney's services in terms of the time he has expended on the case. Anchoring the analysis to this concept is the only way of approaching the problem that can claim objectivity, a claim which is obviously vital to the prestige of the bar and the courts.' (City of Detroit v. Grinnell Corp. (2d Cir. 1974) 495 F.2d 448, 470; see also Lindy Bros. Bldrs., Inc. of Phila. v. American R. & S. San. Corp. (3d Cir. 1973) 487 F.2d 161, 167-169; see generally Dawson, Lawyers and Involuntary Clients in Public Interest Litigation [(1975)] 88 Harv. L.Rev. 849, especially pp. 925-929.)" (Serrano III, supra, 20 Cal.3d at p. 48, fn. 23.)

For his claim that Serrano III mandates primary use of the lodestar method in every case, the objector relies on these passages, in particular our allusions to "`the court's role in equity'" in awarding fees, a role that includes awards in common fund cases, and to the lodestar as the "`starting point of every fee award.'" (Serrano III, supra, 20 Cal.3d at p. 48, fn. 23, italics added.) The quoted text and footnote, however, concern calculation of a fee awarded under the private attorney general theory. In Serrano III, this court simply did not address the question of what methods of calculating a fee award may or should be used when the fee is to be drawn from a common fund created or preserved by the litigation. For this reason, the passages quoted cannot fairly be taken as prohibiting the percentage method's use in a common fund case.

To the contrary, in its earlier discussion of the common fund doctrine, Serrano III cited with approval several decisions in which a percentage fee was awarded. In Fox v. Hale & Norcross S. M. Co., supra, 108 Cal. at page 476, apparently our first case approving a common fund fee award, the award was for 25 percent of the moneys the plaintiff had collected. In Farmers etc. Nat. Bank v. Peterson, supra, 5 Cal.2d at page 607, we held the plaintiff in a suit for an accounting was properly awarded "5 per cent of the moneys received and recovered herein as an attorney's fee." And in Glendale City Employees' Assn., Inc. v. City of Glendale (1975) 15 Cal.3d 328 [124 Cal.Rptr. 513, 540 P.2d 609], a labor action by public employees, we upheld "the portion of the judgment awarding counsel for plaintiffs 25 percent of all retroactive salaries and wages received." (Id. at p. 341, fn. 19.) Having cited these decisions, together with a few others, as establishing and exemplifying the common fund attorney fee doctrine in California (Serrano III, supra, 20 Cal.3d at p. 35), the Serrano III court observed it could find no such fund in that case (id. at p. 36). Had we meant, in our later discussion of the lodestar calculation of a private attorney general fee, to disapprove the percentage method of calculation used in these common fund cases, we would have said so.

In emphasizing the objectivity provided by a lodestar calculation, Serrano III, supra, 20 Cal.3d 25, decided in 1977, was typical of its era. (As discussed in part I, ante, that period is considered to have begun with the Third Circuit's 1973 decision in Lindy I, supra, 487 F.2d 161, which we cited in the footnote passage quoted above.) Because the award in Serrano III was not made from a common fund and did not rest on the common fund theory, we had no occasion there to consider the comparative disadvantages of the lodestar-multiplier method that have since led the vast majority of courts nationwide to instead favor, or at least to allow, use of the percentage-of-the-fund method in common fund cases. As explained in part I, ante, both the Second and Third Circuits subsequently retreated from their endorsements, in City of Detroit v. Grinnell Corp., supra, 495 F.2d 448, and Lindy I, supra, 487 F.2d 161 — the two decisions cited in Serrano III's footnote 23 — of the lodestar method as the preferred or exclusive means of calculating a reasonable fee. (See Goldberger v. Integrated Resources, Inc., supra, 209 F.3d at pp. 48-50 [2d Cir.]; In re Rite Aid Corp. Securities Litigation, supra, 396 F.3d at p. 300 [3d Cir.].) Presenting as it did no common fund from which an award could be made, Serrano III was not a case for entertaining the policy grounds for allowing a common fund fee to be calculated as a percentage of the fund, considerations that have so heavily influenced later courts' decisions on this issue.

Since Serrano III, we have several times, in fee shifting cases, endorsed the lodestar or lodestar-multiplier method of calculating an attorney fee award; none of our decisions involved a case where the fee was to be awarded from a common fund created or preserved by the litigation. (See Graham v. DaimlerChrysler Corp. (2004) 34 Cal.4th 553, 579 [21 Cal.Rptr.3d 331, 101 P.3d 140] [award under Code Civ. Proc., § 1021.5]; Ketchum v. Moses (2001) 24 Cal.4th 1122, 1131-1132 [104 Cal.Rptr.2d 377, 17 P.3d 735] [award under Code Civ. Proc., § 425.16, subd. (c)]; PLCM Group, Inc. v. Drexler (2000) 22 Cal.4th 1084, 1094-1095 [95 Cal.Rptr.2d 198, 997 P.2d 511] [award under Civ. Code, § 1717]; Maria P. v. Riles (1987) 43 Cal.3d 1281, 1294-1295 [240 Cal.Rptr. 872, 743 P.2d 932] [award under Code Civ. Proc., § 1021.5]; Press v. Lucky Stores, Inc. (1983) 34 Cal.3d 311, 321-322 [193 Cal.Rptr. 900, 667 P.2d 704] [same].) And even with regard to such statutory fee shifting cases, we have noted the lodestar-multiplier method of determining a reasonable fee is not necessarily exclusive: "We emphasize, however, that although we are persuaded that the lodestar adjustment approach should be applied to fee awards under Code of Civil Procedure section 425.16, we are not mandating a blanket `lodestar only' approach; every fee-shifting statute must be construed on its own merits and nothing in Serrano jurisprudence suggests otherwise." (Ketchum v. Moses, supra, at p. 1136.)

The objector relies on several Court of Appeal decisions, the first being Jutkowitz v. Bourns, Inc. (1981) 118 Cal.App.3d 102 [173 Cal.Rptr. 248] (Jutkowitz). A minority shareholder plaintiff who had filed a putative class action over a proposed purchase of corporate stock, leading the buyer to increase the price offered, sought an augmented attorney fee based on the value he had created for shareholders who sold at the increased price, even though most of them were not members of the class. (Id. at pp. 106-109.) Although the plaintiff based his fee increase request on the common fund theory, he neither showed that any fund had been created from which the increased fee could be awarded nor, as far as the appellate opinion indicates, sought any particular percentage of the asserted fund as fee. (See id. at pp. 108-110.)

In rejecting the plaintiff's attempt to have the amount of his attorney fee enhanced, the Jutkowitz court observed: "While the size of the class may affect the complexity of counsel's task and the size of the fund created may reflect the quality of his work, the correct amount of compensation cannot be arrived at objectively by simply taking a percentage of that fund." (Jutkowitz, supra, 118 Cal.App.3d at p. 111, italics added.) Given that no fund had in fact been created from which an attorney fee could be taken, the italicized remark need not be read as barring the percentage method of calculating a fee award in a true common fund case. To the extent it could be read broadly as expressing such a general rule, however, we disapprove Jutkowitz v. Bourns, supra, 118 Cal.App.3d 102.

Salton Bay Marina, Inc. v. Imperial Irrigation Dist. (1985) 172 Cal.App.3d 914 [218 Cal.Rptr. 839] (Salton Bay), an inverse condemnation action, also involved no common fund. For its conclusion that the plaintiff was only entitled to reimbursement of a reasonable attorney fee measured by time expended by the attorney, without regard to the contingency fee agreement between attorney and client, the appellate court relied in part on the above passages from Serrano III and Jutkowitz. (Salton Bay, supra, at pp. 953-954.) The decision does not speak to how a fee award should be calculated in a class action settlement or other common fund case. Nor does People ex rel. Dept. of Transportation v. Yuki (1995) 31 Cal.App.4th 1754, 1767-1771 [37 Cal.Rptr.2d 616] (Yuki), an eminent domain case following Salton Bay in disapproving direct use of a contingency fee agreement to determine a fee award, address the issue before us today.

In Dunk v. Ford Motor Co., supra, 48 Cal.App.4th 1794, 1809 (Dunk), in the context of a class action settlement, the court disapproved an attorney fee award the plaintiff attempted to justify as a small percentage of the settlement's value. The court gave two reasons: "(1) The award of attorney fees based on a percentage of a `common fund' recovery is of questionable validity in California; and (2) even if it is valid, the true value of the fund must be easily calculated." (Ibid.) On the second point, the court explained that because the settlement at issue provided class members with coupons for discounts on purchases of new vehicles, its real value could not be ascertained until the end of the coupon redemption period. (Ibid.) On the first, the court cited Jutkowitz, supra, 118 Cal.App.3d 102; Salton Bay, supra, 172 Cal.App.3d 914,; and Yuki, supra, 31 Cal.App.4th 1754, as having "cast doubt on the use of the percentage method to determine attorney fees in California class actions." (Dunk, supra, at p. 1809.) The Dunk court, while finding the percentage method inapplicable to the settlement before it due to the lack of a readily valued common fund, did not purport to bar its usage generally in common fund cases.

Dunk was, in turn, cited as illustrating the doubt over use of the percentage method in California, in the passage from Lealao, supra, 82 Cal.App.4th at page 27, quoted at the beginning of this part. Lealao, a consumer class action over prepayment penalties charged by a lender, was settled by the lender's agreement to pay class members who filed claims 77 percent of the penalties they had paid, a settlement worth almost $15 million if every class member filed a claim. (Id. at pp. 22-23.) Though class counsel requested 24 percent of the recovery as a fee ($3.5 million, modified to $1.76 million after claims of only $7.35 million were filed), the trial court, believing itself precluded from awarding a percentage fee where no separate fund had been established from which the fee could be drawn, granted only a fee of $425,000, calculated as a lodestar without multiplier. (Id. at pp. 24-25.)

Relying on Serrano III, supra, 20 Cal.3d 25, the Lealao appellate court held a pure percentage fee is improper when, as in the case before it, the settlement does not establish a separate fund from which the fee is to be paid. (Lealao, supra, 82 Cal.App.4th at pp. 37-39.)8 But the trial court's lodestar fee could properly be enhanced through a multiplier based on the a percentage of the benefit obtained (Lealao, at pp. 39-50), employing "the common federal practice of `cross-checking' the lodestar against the value of the class recovery." (Id. at p. 45.) Such a cross-check is not prohibited by Serrano III, Jutkowitz or Dunk (Lealao, at pp. 44-45) and helps to determine a reasonable fee because a percentage-of-the-benefit analysis "provides a credible measure of the market value of the legal services provided" (id. at p. 49).

The Lealao court expressed doubt as to the wisdom of considering only the amount of the recovery in determining a fee award, but acknowledged that "[t]he federal judicial experience teaches that the `reasonableness' of a fee in a representative action will often require some consideration of the amount to be awarded as a percentage of the class recovery." (Lealao, supra, 82 Cal.App.4th at p. 53.) Since that decision, several other Court of Appeal panels have approved some form of percentage fee calculation. (See Consumer Privacy Cases, supra, 175 Cal.App.4th at p. 558 [use of percentage method under common fund doctrine "is not an abuse of discretion ... as long as the method chosen is applied consistently using percentage figures that accurately reflect the marketplace."]; Chavez v. Netflix, Inc. (2008) 162 Cal.App.4th 43, 63 [75 Cal.Rptr.3d 413] [under reasoning of Lealao, percentage calculation may be used to determine a lodestar multiplier; it was not an abuse of discretion "for the trial court to apply a percentage figure at the low end of the typical contingency contractual arrangement (21.8 percent) to calculate the multiplier in the context of this settlement"]; Apple Computer, Inc. v. Superior Court (2005) 126 Cal.App.4th 1253, 1270 [24 Cal.Rptr.3d 818] [observing that "attorney fees awarded under the common fund doctrine are based on a `percentage-of-the-benefit' analysis"]; Wershba v. Apple Computer, Inc. (2001) 91 Cal.App.4th 224, 254 [110 Cal.Rptr.2d 145] ["Courts recognize two methods for calculating attorney fees in civil class actions: the lodestar/multiplier method and the percentage of recovery method."].)

In summary, California decisions from Serrano III forward have shown some uncertainty as to the role a percentage-of-the-recovery calculation may play in determining court-ordered attorney fees, but have not established any rule prohibiting such a calculation when the fee is to be drawn from a common fund created by the litigation.

III. A Percentage Calculation with Lodestar Cross-check Is Permitted in a Common Fund Case.

(6) Whatever doubts may have been created by Serrano III, supra, 20 Cal.3d 25, or the Court of Appeal cases that followed, we clarify today that use of the percentage method to calculate a fee in a common fund case, where the award serves to spread the attorney fee among all the beneficiaries of the fund, does not in itself constitute an abuse of discretion. We join the overwhelming majority of federal and state courts in holding that when class action litigation establishes a monetary fund for the benefit of the class members, and the trial court in its equitable powers awards class counsel a fee out of that fund, the court may determine the amount of a reasonable fee by choosing an appropriate percentage of the fund created. The recognized advantages of the percentage method — including relative ease of calculation, alignment of incentives between counsel and the class, a better approximation of market conditions in a contingency case, and the encouragement it provides counsel to seek an early settlement and avoid unnecessarily prolonging the litigation (see pt. I, ante; Lealao, supra, 82 Cal.App.4th at pp. 48-49; Rawlings v. Prudential-Bache Properties, Inc., supra, 9 F.3d at p. 516) — convince us the percentage method is a valuable tool that should not be denied our trial courts.

We do not address here whether or how the use of a percentage method may be applied when there is no conventional common fund out of which the award is to be made but only a "`constructive common fund'" created by the defendant's agreement to pay claims made by class members and, separately, to pay class counsel a reasonable fee as determined by the court (see Lealao, supra, 82 Cal.App.4th at pp. 23-24, 28), or when a settlement agreement establishes a fund but provides that portions not distributed in claims revert to the defendant or be distributed to a third party or the state, making the fund's value to the class depend on how many claims are made and allowed. (See 5 Newberg on Class Actions, supra, § 15:70, pp. 236-242.) The settlement agreement in this case provided for a true common fund fixed at $19 million, without any reversion to defendant and with all settlement proceeds, net of specified fees and costs, going to pay claims by class members.

(7) The trial court in this case thus did not violate principles established in Serrano III, supra, 20 Cal.3d 25, or otherwise abuse its discretion, in using a percentage method for its primary calculation of the fee award. The choice of a fee calculation method is generally one within the discretion of the trial court, the goal under either the percentage or lodestar approach being the award of a reasonable fee to compensate counsel for their efforts. (Consumer Privacy Cases, supra, 175 Cal.App.4th at pp. 557-558.) Before approving the settlement agreement and percentage fee award in this case, the trial court supplemented its own familiarity with the case by obtaining additional information from class counsel on the risks and potential value of the litigation; the court carefully considered that information on contingency, novelty and difficulty together with the skill shown by counsel, the number of hours worked and the asserted hourly rates, which the court found were not overstated. On that basis, the trial court determined the fee request was for a reasonable percentage of the settlement fund.

Nor do we perceive an abuse of discretion in the court's decision to double check the reasonableness of the percentage fee through a lodestar calculation. As noted earlier, "[t]he lodestar method better accounts for the amount of work done, while the percentage of the fund method more accurately reflects the results achieved." (Rawlings v. Prudential-Bache Properties, Inc., supra, 9 F.3d at p. 516.) A lodestar cross-check thus provides a mechanism for bringing an objective measure of the work performed into the calculation of a reasonable attorney fee. If a comparison between the percentage and lodestar calculations produces an imputed multiplier far outside the normal range, indicating that the percentage fee will reward counsel for their services at an extraordinary rate even accounting for the factors customarily used to enhance a lodestar fee, the trial court will have reason to reexamine its choice of a percentage. (Walker & Horwich, supra, 18 Geo. J. Legal Ethics at p. 1463.)

The utility of a lodestar cross-check has been questioned on the ground it tends to reintroduce the drawbacks the 1985 Task Force Report identified in primary use of the lodestar method, especially the undue consumption of judicial resources and the creation of an incentive to prolong the litigation. (See 5 Newberg on Class Actions, supra, § 15:86, pp. 330-334 [describing, but largely rejecting, objections to cross-check]; Gilles & Friedman, Exploding the Class Action Agency Costs Myth: The Social Utility of Entrepreneurial Lawyers (2006) 155 U.Pa. L.Rev. 103, 140-142 [use of lodestar method, even as cross-check, undesirably limits deterrent potential of certain large-damages class actions by incentivizing pretrial settlement].) We tend to agree with the amicus curiae brief of Professor William B. Rubenstein that these concerns are likely overstated and the benefits of having the lodestar cross-check available as a tool outweigh the problems its use could cause in individual cases.

(8) With regard to expenditure of judicial resources, we note that trial courts conducting lodestar cross-checks have generally not been required to closely scrutinize each claimed attorney-hour, but have instead used information on attorney time spent to "focus on the general question of whether the fee award appropriately reflects the degree of time and effort expended by the attorneys." (5 Newberg on Class Actions, supra, § 15:86, p. 331; see, e.g., Goldberger v. Integrated Resources, Inc., supra, 209 F.3d at p. 50 [2d Cir.; "where used as a mere cross-check, the hours documented by counsel need not be exhaustively scrutinized by the district court"]; In re Prudential Ins. Co. America Sales Practice Litigation Agent Actions (3d Cir. 1998) 148 F.3d 283, 342 [agreeing with district court that "detailed time summaries were unnecessary where, as here, it was merely using the lodestar calculation to double check its fee award."]; Barbosa v. Cargill Meat Solutions Corp. (E.D.Cal. 2013) 297 F.R.D. 431, 451 ["Where the lodestar method is used as a cross-check to the percentage method, it can be performed with a less exhaustive cataloguing and review of counsel's hours."].) The trial court in the present case exercised its discretion in this manner, performing the cross-check using counsel declarations summarizing overall time spent, rather than demanding and scrutinizing daily time sheets in which the work performed was broken down by individual task. Of course, trial courts retain the discretion to consider detailed time sheets as part of a lodestar calculation, even when performed as a cross-check on a percentage calculation.

(9) As to the incentives a lodestar cross-check might create for class counsel, we emphasize the lodestar calculation, when used in this manner, does not override the trial court's primary determination of the fee as a percentage of the common fund and thus does not impose an absolute maximum or minimum on the potential fee award. If the multiplier calculated by means of a lodestar cross-check is extraordinarily high or low, the trial court should consider whether the percentage used should be adjusted so as to bring the imputed multiplier within a justifiable range, but the court is not necessarily required to make such an adjustment. Courts using the percentage method have generally weighed the time counsel spent on the case as an important factor in choosing a reasonable percentage to apply. (5 Newberg on Class Actions, supra, § 15:86, pp. 332-333; see, e.g., In re Thirteen Appeals Arising Out of San Juan Dupont Plaza Hotel Fire Litigation, supra, 56 F.3d at p. 307 ["even under the [percentage of fund] method, time records tend to illuminate the attorneys' role in the creation of the fund, and, thus, inform the court's inquiry into the reasonableness of a particular percentage."].) A lodestar cross-check is simply a quantitative method for bringing a measure of the time spent by counsel into the trial court's reasonableness determination; as such, it is not likely to radically alter the incentives created by a court's use of the percentage method.

(10) We therefore agree with the Court of Appeal below that "[t]he percentage of fund method survives in California class action cases, and the trial court did not abuse its discretion in using it, in part, to approve the fee request in this class action." We hold further that trial courts have discretion to conduct a lodestar cross-check on a percentage fee, as the court did here; they also retain the discretion to forgo a lodestar cross-check and use other means to evaluate the reasonableness of a requested percentage fee.

DISPOSITION

The judgment of the Court of Appeal is affirmed.

Cantil-Sakauye, C. J., Chin, J., Corrigan, J., Liu, J., Cuéllar, J., and Kruger, J., concurred.

LIU, J., Concurring. —

Appellant David Brennan devotes the lion's share of his briefing to issues beyond today's holding that trial courts may use the percentage method instead of the lodestar method to award attorneys' fees from a common fund. He argues that the lodestar method as applied does not comply with Serrano v. Priest (1977) 20 Cal.3d 25 [141 Cal.Rptr. 315, 569 P.2d 1303] (Serrano III); that courts demand too little documentation of attorney hours and do not subject such documentation to careful scrutiny; that named plaintiffs do not adequately monitor class counsel; and that courts using the percentage method, including the trial court in this case, have applied percentage numbers drawn from individual contingent fee cases without taking into account the economies of scale in class representation. To remedy these alleged abuses, Brennan urges us to explicitly state the requirements of the lodestar methodology, to require appointment of class guardians to protect absent class members through adversarial representation, and to appoint fee experts for absent class members where class counsel retains such an expert.

Although the court declines to address these arguments, I write separately to suggest practices that may help to promote accuracy, transparency, and public confidence in the awarding of attorneys' fees in class action litigation.

First and foremost, although disputes over attorneys' fees often arise in the context of a proposed settlement as in this case, courts and litigants need not and generally should not wait until the end of litigation to set the terms of attorney compensation. Whenever possible, the parties should negotiate, and the court should review and conditionally approve, the terms of attorney compensation at the start of litigation. The parties and the court may revisit the arrangement when the litigation concludes, and the court may make adjustments if unusual or unforeseen circumstances render the initial terms clearly unreasonable or unfair. But in general, the parties' initial bargain should be given substantial weight in determining the reasonableness of a fee award.

The Task Force on Selection of Class Counsel convened by the United States Court of Appeals for the Third Circuit has endorsed a version of this approach. While acknowledging that "a precise ex ante determination of fees is usually unworkable," the task force recommended that "the topic of attorney fees should be addressed at the early stages of the case as well as throughout the prosecution of the case. At the outset of the case, the court may be well-advised to direct counsel to propose the terms for a potential award of fees; the potential fees might be established within ranges, with the court making it clear to the parties that the fee remains open for further review for reasonableness. A preliminary fee arrangement may provide a helpful structure for the court when it conducts its reasonableness review at the end of the case." (Third Circuit Task Force, Selection of Class Counsel (2002), 208 F.R.D. 340, 420-421, fns. omitted (Task Force Report); see Baker et al., Is the Price Right? An Empirical Study of Fee-Setting in Securities Class Actions (2015) 115 Colum. L.Rev. 1371, 1432 (Baker et al.) [recommending ex ante fee arrangements for class actions governed by the federal Private Securities Litigation Reform Act of 1995 (PSLRA; Pub.L. No. 104-67 (Dec. 22, 1995) 109 Stat. 737) and urging that "the district court should apply the agreed terms unless unforeseen developments have rendered those terms clearly excessive or unfair"].)

This approach has doctrinal and practical virtues. Doctrinally, a court's authority to award attorneys' fees from a common fund stems from its equitable power to prevent unjust enrichment. (See Serrano v. Unruh (1982) 32 Cal.3d 621, 627 [186 Cal.Rptr. 754, 652 P.2d 985] [the "central theory underlying" fee awards from a common fund is "`prevention of an unfair advantage to the others who are entitled to share in the fund and who should bear their share of the burden of its recovery'"]; Serrano III, supra, 20 Cal.3d at p. 35 ["`one who expends attorneys' fees in winning a suit which creates a fund from which others derive benefits, may require those passive beneficiaries to bear a fair sharing of the litigation costs'"].) But a claim for unjust enrichment typically lies where it is impractical to bargain ex ante for a good or service in an arms-length negotiation. "[W]hen it is feasible for parties to bargain, restitution is typically denied to providers who confer benefits without negotiating for payment in advance." (Silver, A Restitutionary Theory of Attorneys' Fees in Class Actions (1991) 76 Cornell L.Rev. 656, 667.) "The effect of withholding compensation in contexts where parties can bargain is to demonstrate a preference for voluntary exchange." (Id. at p. 669.)

As a practical matter, "[t]he best time to determine [the rate of attorney compensation] is the beginning of the case, not the end (when hindsight alters the perception of the suit's riskiness, and sunk costs make it impossible for the lawyers to walk away if the fee is too low). This is what happens in actual markets. Individual clients and their lawyers never wait until after recovery is secured to contract for fees. They strike their bargains before work begins. Ethically lawyers must do this, but the same thing happens in markets for other professional services with different (or no) ethical codes.... Only ex ante can bargaining occur in the shadow of the litigation's uncertainty; only ex ante can the costs and benefits of particular systems and risk multipliers be assessed intelligently. Before the litigation occurs, a judge can design a fee structure that emulates the incentives a private client would put in place. At the same time, both counsel and class members can decide whether it is worthwhile to proceed with that compensation system in place." (In re Synthroid Marketing Litigation (7th Cir. 2001) 264 F.3d 712, 718-719.) Empirical evidence suggests that ex ante fee negotiation is a key mechanism for reducing agency costs between counsel and the class they represent. (See Baker et al., supra, 115 Colum. L.Rev. at p. 1394 [studying 431 securities class action settlements from 2007 through 2012, and finding that "fee agreements negotiated at the beginning of cases have a substantial moderating effect on fee requests" where public pension funds act as lead plaintiffs].)

Moreover, ex ante fee arrangements do not present the conflict of interest that inherently arises when attorneys seek fees from a common fund comprising their clients' recovery. "At the start of litigation, there is no money to divide. There is only the prospect of forming a joint venture between a client and a lawyer that seeks to maximize the parties' joint wealth by offering the lawyer compensation terms that will motivate the lawyer to work hard on behalf of the client. [¶] When fees are set at the end of litigation, by contrast, the amount to be recovered is already known. This heightens the conflict between the client and the attorney because every additional dollar for one means a dollar less for the other." (Baker et al., supra, 115 Colum. L.Rev. at p. 1440.)

Opponents of ex ante fee agreements in the class action context have argued that (1) there is no "functioning market" for plaintiffs' representation and thus no reliable benchmarks that can provide a "general solution to the problem of market failure in setting class counsel fees" (ABA Tort Trial and Insurance Practice Section, Report on Contingent Fees in Class Action Litigation (2006) 25 Rev.Litig. 459, 481, 482); (2) at the early stages of class action litigation, there are too many uncertainties for bargaining to occur (id. at p. 482); and (3) if fee arrangements are disclosed to defendants, this might disadvantage plaintiffs in settlement negotiations (Baker et al., supra, 115 Colum. L.Rev. at p. 1436).

As to the first point, courts evaluating ex ante fee arrangements may use "a simple benchmark: the percentage or range of percentages prevailing in the private market in similar contingent fee representations." (Silver, Dissent from Recommendation to Set Fees Ex Post (2006) 25 Rev.Litig. 497, 499.) "Plaintiffs have formed voluntary groups in mass accident cases, pollution cases, defective product cases, securities fraud cases, and cases of other kinds. Associations, including homeowners' associations, interest groups, unions, partnerships, and corporations have sued on behalf of their members or owners thousands of times.... All of these lawsuits are examples of aggregate litigation, and in all of them lawyers' fees have been set ex ante via negotiations." (Id. at pp. 499-500; see In re Synthroid Marketing Litigation, supra, 264 F.3d at p. 719 ["[A] court can learn about similar bargains. That is at least a starting point."].)

As to the second point, the principal virtue of an ex ante fee arrangement is its allocation of risk between attorney and client in the face of litigation uncertainty. At the end of litigation, when the amount of recovery and the outcomes of all other uncertainties are known, perceptions of risk are likely to be distorted by hindsight bias. (Baker et al., supra, 115 Colum. L.Rev. at pp. 1441-1444.) Uncertainty is the very reason why it is appropriate for negotiations over fees to occur at the start of litigation; the market price for legal services can be more accurately derived through bargaining behind the veil of ignorance. (In re Synthroid Marketing Litigation, supra, 264 F.3d at p. 719.) Moreover, the initial terms set by the parties and approved by the court are not etched in stone; as noted, the court may make adjustments if unusual or unforeseen circumstances render the initial arrangement clearly unreasonable or unfair.

As to the third point, concerns about disclosure can be alleviated by allowing plaintiffs and class counsel to submit their fee arrangements to the court under seal or "by discussing fees with class counsel in chambers on an ex parte basis." (Baker et al., supra, 115 Colum. L.Rev. at p. 1437.) Such approaches pose no unfairness to defendants, who "are indifferent to fee requests because the fees are paid out of the common fund." (Id. at p. 1419.)

Quite apart from the concerns above, a significant practical challenge to negotiating attorneys' fees in many class actions, whether at the start or end of litigation, is the lack of an active and interested class representative who can effectively bargain with and monitor plaintiffs' counsel. Some class actions, such as securities litigation, have managed to attract large institutional investors as lead plaintiffs. In that role, they closely evaluate and choose high-quality lawyers, and they actively bargain for favorable fee structures and secure ex ante fee arrangements more often than do other lead plaintiffs. (Baker et al., supra, 115 Colum. L.Rev. at pp. 1393-1394; see 15 U.S.C. § 78u-4(a)(3) [establishing process for appointment of lead plaintiff in class actions governed by the PSLRA].) By contrast, consumer class actions and wage-and-hour disputes often lack a class representative with sufficient incentive, resources, or expertise to negotiate with class counsel. Moreover, although Brennan came forward in this case as an objector, class objectors are too rare to be generally relied upon to monitor class counsel. (Eisenberg & Miller, The Role of Opt-Outs and Objectors in Class Action Litigation: Theoretical and Empirical Issues (2004) 57 Vand. L.Rev. 1529, 1549 ["Across all case types, ... the median objection rate is zero and the mean is 1.1 percent of class members."].) And the few who do object have had little to no demonstrable impact on attorneys' fees or settlement amounts. (Id. at p. 1563 ["We found no significant association between the number of dissenters and either the gross fee or the fee as a percentage of class recovery."].)

Although trial courts can exercise vigilance to ensure fairness in fee negotiations, doing so puts the judge in the position of "a fiduciary guarding the rights of absent class members" (In re Cendant Corp. Litigation (3d Cir. 2001) 264 F.3d 201, 231) while at the same time serving as a neutral arbiter of counsel's claims concerning the reasonableness of a proposed award. As Brennan puts it, the trial judge is asked "to simultaneously assume the conflicting roles of impartial judge and class advocate."

In many cases, trial courts may have no choice but to walk the fine line between protecting the interests of absent class members and impartially evaluating the reasonableness of a proposed fee award. In cases involving substantial sums, however, trial courts may take steps to insulate themselves from apparent conflicts by appointing a class guardian or "devil's advocate" so that arguments for and against the reasonableness of a fee arrangement may be presented in a genuinely adversarial process. (Cf. Rubenstein, The Fairness Hearing: Adversarial and Regulatory Approaches (2006) 53 UCLA L.Rev. 1435, 1454 [proposing court-designated attorney to serve as "devil's advocate" in evaluating class action settlements].) The class guardian would provide counterpoints to class counsel's arguments concerning the risks and difficulty of litigating the case. Perhaps most importantly, the class guardian or a fee expert retained by the guardian would provide information on prevailing market rates for similar litigation. The appointment of a guardian and a full-dress adversarial process would cost money (from the common fund) and time. But these costs, which would serve to enhance the accuracy and legitimacy of fee awards, would "pale[] in comparison to the significant amounts of money" to be divided between plaintiffs and counsel in high-value cases. (Id. at p. 1455.)

*

The suggestions above reflect the importance of fairness and reasonableness in attorney compensation. Ensuring "objectivity" in attorney compensation "`is obviously vital to the prestige of the bar and the courts.'" (Serrano III, supra, 20 Cal.3d at p. 48, fn. 23.) Moreover, "[p]robably to a unique degree, American law relies upon private litigants to enforce substantive provisions of law that in other legal systems are left largely to the discretion of public enforcement agencies.... The key legal rules that make the private attorney general a reality in American law today ... [are] those rules that establish the fee arrangements under which these plaintiff's attorneys are compensated. Inevitably, these rules create an incentive structure that either encourages or chills private enforcement of law." (Coffee, Understanding the Plaintiff's Attorney: The Implications of Economic Theory for Private Enforcement of Law Through Class and Derivative Actions (1986) 86 Colum. L.Rev. 669, 669-670, fns. omitted (Coffee).) "By setting fees too high or too low, judges would incentivize lawyers to bring too many class actions or too few. Excessive litigation would over-deter primary conduct that is desirable; insufficient litigation would under-deter primary conduct that is unwanted." (Baker et al., supra, 115 Colum. L.Rev. at p. 1375.)

It must be acknowledged that "there is a perception among a significant part of the non-lawyer population and even among lawyers and judges that the risk premium is too high in class action cases and that class action plaintiffs' lawyers are overcompensated for the work that they do." (Task Force Rep., supra, 208 F.R.D. at pp. 343-344.) I express no view on the degree to which this perception is anchored in reality. (Compare Task Force Rep., supra, 208 F.R.D. at p. 344 ["When there is a public reaction to an attorney fee award in a given case, the public is usually unaware of what the lawyers actually did, what risks they took, what investment they made, and how important their lawyering was to victory for the class."] with Coffee, supra, 86 Colum. L.Rev. at p. 714 ["At its worst, the settlement process may amount to a covert exchange of a cheap settlement for a high award of attorney's fees."].) But the perception itself may prompt some judges and policymakers to respond by narrowing substantive legal protections or by curtailing procedural mechanisms of enforcement.

Public confidence in the fairness of attorney compensation in class actions is vital to the proper enforcement of substantive law. Although there may be no single "right answer" to how much class counsel should earn in each case, ex ante fee arrangements with the possibility of ex post modification for unusual circumstances may provide a useful approach to estimating market rates, reducing the distortive effects of hindsight bias, and aligning the interests of counsel and the class they represent. Courts and litigants should be alert to this and other approaches that may help to promote greater public confidence in a form of litigation on which many people rely to obtain effective access to justice.