In Re Synthroid Marketing Litigation , 201 F. Supp. 2d 861 ( 2002 )


Menu:
  • 201 F.Supp.2d 861 (2002)

    In re SYNTHROID MARKETING LITIGATION.

    No. 97 C 6017.

    United States District Court, N.D. Illinois, Eastern Division.

    May 9, 2002.

    *862 Marvin Alan Miller, Miller Faucher and Cafferty, LLP, Chicago, IL, Michael D. Hausfeld, Lisa M Messetti, Cohen, Milstein, Hausfield & Toll, Washington, DC, Lawrence W. Schonbrun, Berkeley, CA, for Plaintiffs.

    Howard Brian Prossnitz, Law Offices of Howard Prossnitz, Chicago, IL, for Julianne Rausch.

    Lowell E. Sachnoff, Sachnoff & Weaver, Ltd., Chicago, IL, for Sandora Davila, Richard Saenz.

    Scott Rhead Shepherd, Shepherd, Finkelman & Gaffigan, LLC, Media, PA, for Mae Fischer.

    Gary D. McCallister, Gary D. McCallister & Associates, Ltd., Chicago, IL, for Blue Cross and Blue Shield of Alabama, Louisiana Health Service & Indem. Co., Alabama Medicaid Agency, Dept. of Health and Hospitals.

    Lynde Selden, II, Selden Law Firm, San Diego, CA, Theodore J Holt, Hackard & Holt, Gold River, CA, Richard H. Rosenthal, Law Offices of Richard H. Rosenthal, A Professional Corp., Carmel Valley, CA, for Rosalie Lynch.

    Marc David Sherman, Rena M. Honorow, Marc D. Sherman & Associates, Lincolnwood, IL, Lynde Selden, II, Selden Law Firm, San Diego, CA, Theodore J Holt, Hackard & Holt, Gold River, CA, for Jeanette C. Miller.

    Theodore J Holt, Hackard & Holt, Gold River, CA, Richard H. Rosenthal, Law Offices of Richard H. Rosenthal, A Professional Corp., Carmel Valley, CA, for Gordon Fjeld.

    James L. Johnson, Johnson Law Firm, Dallas, TX, for Cigna Healthcare.

    Elizabeth Leigh Thompson, Lindley J. Brenza, Bartlit, Beck, Herman, Palenchar & Scott, Chicago, IL, Joseph C. Kohn, Joanne Zack, Douglas A. Abrahams, Kohn, Swift & Graf, P.C., Philadelphia, PA, Gregory P. Hansel, Preti, Flaherty, Beliveau, Pachios & Haley, L.L.C., Portland, ME, for Aetna U.S. Healthcare, Inc., Aetna Life Ins. Co., Corporate Health Ins. Co.

    James F. Flug, Duncan, Weinberg, Miller & Penbroke, P.C., Washington, DC, for State of Michigan.

    Michael D. Monico, Monico, Pavich & Spevack, Chicago, IL, James W Bergenn, Theodore M Space, Ross Garber, Shipman & Goodwin LLP, Hartford, CT, Mark D. Fischer, Rawlings & Associates, Louisville, KY, for Blue Cross and Blue Shield of New Jersey, Inc.

    Delilah Brummet, Chicago, IL, Tyrone C Fahner, Diane Julia Romza-Kutz, Mayer, Brown, Rowe & Maw, Chicago, IL, *863 Duane C. Quaini, Christopher Qualley King, Sonnenschein, Nath & Rosenthal, Chicago, IL, Terrence A. Callan, Mary B. Cranston, Roxane A Polidora, Saul D Bercovitch, Pillsbury Winthrop LLP, San Francisco, CA, Kenneth J. Merlino, Power, Rogers & Smith, P.C., Chicago, IL, Diane Green-Kelly, Mayer Brown & Platt, Chicago, IL, for BASF Corp.

    Delilah Brummet, Chicago, IL, Tyrone C Fahner, Diane Julia Romza-Kutz, Mayer, Brown, Rowe & Maw, Chicago, IL, Duane C. Quaini, Christopher Qualley King, Sonnenschein, Nath & Rosenthal, Chicago, IL, Terrence A. Callan, Mary B. Cranston, Roxane A Polidora, Saul D Bercovitch, Pillsbury Winthrop LLP, San Francisco, CA, Kenneth J. Merlino, Power, Rogers & Smith, P.C., Chicago, IL, Richard L. Josephson, Travis James Sales, Baker & Botts, L.L.P., Houston, TX, Diane Green-Kelly, Mayer Brown & Platt, Chicago, IL, for Knoll Pharmaceutical Co.

    Terrence A. Callan, Mary B. Cranston, Roxane A Polidora, Saul D Bercovitch, Pillsbury Winthrop LLP, San Francisco, CA, for Neil Kurtz, Carter H Eckert, Gilbert H. Mayor, Boots-Flint Inc., B A Buhler.

    Michael D. Monico, Barry A. Spevack, Monico, Pavich & Spevack, Chicago, IL, Terrence A. Callan, Mary B. Cranston, Roxane A Polidora, Saul D Bercovitch, Pillsbury Winthrop LLP, San Francisco, CA, for Boots Pharmaceuticals.

    Charles Gilbert Fergus, Illinois Atty. General's Office, Chicago, IL, Rebecca Desiree Ward, Foley & Lardner, Chicago, IL, Terrence A. Callan, Mary B. Cranston, Roxane A Polidora, Saul D Bercovitch, Pillsbury Winthrop LLP, San Francisco, CA, Sandra S. Paguaga, Ill. Atty. General's Office, Chicago, IL, Randall L. Mitchell, James Dominick Adducci, Marshall Lee Blankenship, Adducci, Dorf, Lehner, Mitchell & Blankenship, P.C., Chicago, IL, for Flint Laboratories.

    R. Stephen Griffis, R. Stephen Griffis, P.C., Birmingham, Al, for Ethylyn Pugh.

    John Robert Sears, Missouri Atty. Gen., Jefferson City, MO, for State of Missouri.

    John B. Howard, Jr., Office of the Atty. Gen. of Maryland, Baltimore, MD, for State of Maryland.

    Lawrence W. Schonbrun, Berkeley, CA, for Ruby H. Davis, Karen Caronna.

    Jane Elizabeth Reames, Brenner, Ford & Monroe Ltd., Chicago, IL, George W. Cochran, George W. Cochran, Streetsboro, OH, for Cynthia L. Artim, Lisa Delgado, Sherrie A. Sitton, Lorrie Mead.

    Kearney Dee Hutsler, Thompson, Hutsler Law Firm, Birmingham, AL, Robert Patrick Cummins, Thomas Cusack Cronin, Cummins & Cronin, LLC, Chicago, IL, for Chromium Industries.

    Robert Patrick Cummins, Thomas Cusack Cronin, Cummins & Cronin, LLC, Chicago, IL, for Margaret Waid.

    R. Steven Baker, The Southern Law Group, P.C., Birmingham, AL, for Dana Williamson, Debra Jones, Betty Mullinax.

    Michael D. Monico, Monico, Pavich & Spevack, Chicago, IL, for Health Benefits Payers.

    Adelheid Mueller, West Palm Beach, CA, pro se.

    MEMORANDUM OPINION AND ORDER

    BUCKLO, District Judge.

    This opinion discusses attorneys' fees and costs in a large class action lawsuit. Synthroid is the brand name of a synthetic version of levothyroxine sodium, a drug used to treat thyroid disorders. In 1990, Betty J. Dong, a pharmacy professor at the University of California, San Francisco, *864 wrote a report for Knoll Pharmaceutical Company concluding that there were cheaper drugs that were bioequivalents. Knoll refused to allow publication of the report, but when Dong's conclusions were publicized anyway, Knoll, embarrassed, relented, and the Dong report appeared in the Journal of the American Medical Association in 1997. Meanwhile, after the initial release of Dong's conclusions, consumers and third party payers filed class action lawsuits against various defendants, including Knoll, BASF Corp., and Boots Pharmaceuticals, alleging that they had lied about or suppressed the existence of bioequivalents and overcharged consumers and their insurers for their brand name product, Synthroid. The various claims were transferred and consolidated before me under the multi-district litigation statute in August 1997. The parties came to me with a settlement proposal that I rejected as inadequate in an order of August 17, 1998. I certified two classes, one of consumers and one of third party payers.

    On August 5, 2000, I approved a new settlement proposal creating a common fund of over $150 million to be divided between the two classes. See 110 F.Supp.2d 676 (N.D.Ill.2000). The part of my opinion approving the settlement was affirmed by the Seventh Circuit, 264 F.3d 712 (7th Cir.2001). That court, however, remanded those portions of my order that dealt with attorneys' fees and costs. Consumer class counsel renew their request for 29% of the consumer settlement amount, as of October 2001, about $96,200,000 (including interest), and around $1,500,000 in costs. Again, I give them less than they ask for, but a lot of money nonetheless. Counsel for the third party payer ("TPP") class has again asked for attorneys' fees in the amount of about 22% of the class settlement funds (principal amount plus interest), which fund, as of October 2001, stood at about $48,500,000 as well as for around $620,100 in expenses. I approve this fee petition. Objectors' counsel also asks for attorneys' fees, and I deny these.

    I. Procedural Background

    In my previous opinion, I noted that attorneys' fees in a class action require court approval, in part because at this stage of the case, the role of class counsel now changes from that of fiduciaries to claimants against the fund created for the clients' benefit. 110 F.Supp.2d at 683 (citing Cook v. Niedert, 142 F.3d 1004, 1011 (7th Cir.1998)). In awarding fees, a court must protect the clients' interests without undermining the incentive for attorneys to bring meritorious class actions on an inescapably contingent basis. Id. (citing Florin v. Nationsbank of Georgia, N.A., 60 F.3d 1245, 1247 (7th Cir.1995)). Referring to the factors that bear on attorneys' fees that are discussed in the Manual for Complex Litigation (Third) § 24.121, at 190 (1995), I noted that this settlement created two large funds and benefitted many people; the objections were insubstantial; class counsel were able and efficient; the litigation was complex but fairly brief; and class counsel devoted much time to the case, but not a lot in view of the size of the settlement. Id. at 684.

    As the Manual suggests, I also considered the awards in similar cases. I noted that there are few so-called "megafund" cases with settlements of over $100 million, but that in such cases, courts have been reluctant to award class counsel attorneys' fees in the range of 22% or 29%, see id. at 684-85 (citing cases), first, because, other things being equal, such awards would be "an indefensible windfall," id. (citing In the Matter of Superior Beverage, 133 F.R.D. 119, 124 (N.D.Ill.1990)), or "manifestly unjust," id. (citing In re Unisys Corp. Retiree Med. Benefits ERISA Litig., 886 *865 F.Supp. 445, 462 (E.D.Pa.1995)).[1] Second, courts have thought that economies of scale kick in with additional recoveries on work expended, id. at 684 (citing In re Domestic Air Transp. Antitrust Litig., 148 F.R.D. 297, 351 (N.D.Ga.1993)), meaning that in a case of this size, lesser amounts of legal work are required for each additional dollar in the recovery. So I agreed with the Third Circuit Task Force on Court Awarded Attorney Fees that "absent unusual circumstances, the percentage will decrease as the size of the fund increases." Id. (citing In re Chambers Dev. Sec. Litig., 912 F.Supp. 852, 861 (W.D.Pa.1995)). Accordingly, I observed, most awards in cases involving very large settlements had been in the 4-10% range. Id. (citing Unisys., 886 F.Supp. at 462).

    I did not, however, state that the existence of this range implied a percentage cap as a matter of law on an attorneys' fee award in a large settlement. Rather the rationale was an economic analysis. My reasoning was similar to that of the District of Wyoming, which said that when "a common fund is extraordinarily large, the application of a benchmark or standard percentage may result in a fee that is unreasonably large for the benefits conferred." In re Copley Pharm., Inc., 1 F.Supp.2d 1407, 1413 (D.Wyo.1998) (citing Herbert P. Newberg, Attorney Fee Awards § 2.09 (1986)). As that court found, "empirical research ... reveals that courts are sensitive to this problem, reducing percentage awards as the size of the recovery increases." Id. (citing William J. Lynk, The Courts and the Plaintiff's Bar: Awarding the Attorney's Fee in Class-Action Litigation, 23 J. Legal Stud. 185, 201 (1994)). The issue is mainly one of economies of scale and diminishing marginal returns:

    It is not one hundred fifty times more difficult to prepare, try, and settle a $150 million case than it is to try a $1 million case, but application of a percentage comparable to that in a smaller case [may] yield an award 150 times greater. Thus where fund recoveries range from $51-$75 million, fee awards usually fall in the 13-20% range. In megafund cases like this one, wherein a class recovers $75-$200 million (or more), courts most stringently weigh the economies of scale inherent in class actions in fixing a percentage yielding a recovery of reasonable fees. Accordingly, fees in the range of 6-10% and even lower are common in mega-common fund cases.

    Id. (citing Newberg, at § 2.09; Lynk, 23 J. Legal Stud. at 201). Naturally, in a competitive market, economies of scale should be reflected in the market rate. Because my research had indicated that courts had concluded that this was the market rate for legal work leading to recoveries of that size, I awarded counsel for each class 10% recovery for their respective classes, coming to about $9 million for the consumer class counsel and about $4.5 million for the third party payer class counsel. I said that this was "at the high end of the megafund range." Id. at 685.

    *866 Class counsel moved for reconsideration, arguing, as I wrote in a minute order of September 7, 2000 (here quoted virtually in full) that "setting attorneys' fees in a megafund case at a flat 10% risks creating perverse incentives for attorneys to settle for a lower amount at a higher percentage. If there is a set cutoff at which a lower `megafund' attorney fee rate kicks in, say $40 million, an irresponsible and greedy class counsel might be tempted to settle for $39 million to get 25% rather than 10%." I agreed with that argument, but said, "[o]f course, no responsible judge would approve such a settlement, nor would any honest attorney propose it, and in any case this is an argument against a fixed cutoff for what counts as a megafund settlement, which I did not use, rather than against a flat 10% fee award." Id. Therefore I denied the motion to reconsider. I did not state that the flat 10% fee award here was derived from a 10% cap imposed as a matter of law on any "megafund" recovery, but only that it represented the high end of the range of what courts had in fact awarded in recoveries of that size.

    Class counsel (among others) appealed, and won a remand on the issue of attorneys' fees. The Seventh Circuit stated that "when deciding on appropriate fee levels in common-fund cases, 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." 264 F.3d at 718. Furthermore, the Seventh Circuit interpreted my statements in the opinion approving the settlement about "the high end of the megafund range" as imposing, as a matter of law, a cap of 10% on any "megafund" recoveries. Although I rejected, in the order of September 7, 2000, the idea that there is a cutoff for what counts as a megafund, see supra, the Seventh Circuit said that I had defined "megafunds" as settlements of at least $75 million. Id. at 718. The Seventh Circuit stated that a fixed cap plus a cutoff for what counts as a megafund recovery, creates a perverse incentive for attorneys to settle for less to get more for themselves; that is true, indeed, that is why, in the order of September 7, I rejected any such caps and cutoffs.

    The appeals court stated that there is no "`megafund rule' [that] trumps the[ ] market rates" and that it is not the case that "as a matter of law no recovery can exceed 10% of a `megafund.'" Id. Very large recoveries, then, are not to be treated differently merely because they are very large, although if the market treats them differently, the courts may also. I do not take this as a directive that an award of attorneys' fees in a large settlement cannot be 10% or less, just that the award is to reflect market rates insofar as these can be determined. Although the Seventh Circuit has made it clear that bigger awards are permitted even with very large settlements, I do not take the court of appeals to hold that I am to ignore economic factors such as economies of scale and diminishing marginal returns in ascertaining the market value of services; on the contrary, see 264 F.3d at 721; accord In the Matter of Continental Illinois Sec. Litig., 962 F.2d 566, 572 (7th Cir.1992). Anything that goes into ascertaining the market rate is to be taken into consideration. Considerations of fairness are not precluded, however, as long as the award mimics market outcomes.

    In fashioning an appropriate attorneys' fee award, I am to "estimate the terms of the contract that private plaintiffs would have negotiated with their lawyers, had bargaining occurred at the outset of the case (that is, when the risk of loss still existed)," 264 F.3d at 718, which is axiomatic, but necessary to state expressly. *867 The Seventh Circuit suggests that I should have followed "the private model by setting fee schedules at the outset of class litigation — [perhaps] by auction, ... [or] negotiation, sometimes for a percentage of recovery, [or] for a lodestar hourly rate and a multiplier for riskbearing." 264 F.3d at 719. "But in this case the district judge let the opportunity slip away, turning to fees only ex post." Id. However, ex ante determination was impossible because the case came to me in a posture ripe for settlement, indeed, with a proposed settlement for my consideration, as the result of a multi-district consolidation, rather than in the form of a filing in my court from the ground up. So I could not have designed, "[b]efore litigation occur[red], ... a fee structure that emulate[d] the incentives a private client would put in place." Id. I had no choice but to do what the appeals court designates as a second best choice, to "set a fee by approximating the terms that would have been agreed to ex ante, had negotiations occurred." Id.

    The Seventh Circuit pointed to three kinds of guideposts or benchmarks that can be used to mimic ex post what the market might have produced ex ante: (1) "the fee contracts some [third party payers] signed with their attorneys; (2) data from large common-pool cases where fees were privately negotiated; and (3) information on class-counsel auctions, [instances] where judges have entertained bids from different attorneys seeking the right to represent a class." 264 F.3d at 719. I proceed by going through each benchmark indicated by the Seventh Circuit in the order indicated. I am to bring to bear the factors that the Seventh Circuit did list for each separate group of class counsel (consumer and third party payer), based solely on its own settlement and not the total amount recovered by the two classes, and I am to make this determination in view of each group's own risk and productivity. Id. at 722. The Court of Appeals did not expressly direct any particular assignment of weight to the two factors of risk versus productivity, but its statement that, ideally, fees are to be determined ex ante, and its emphasis on the need to maintain incentives to bring risky lawsuits suggests that risk is more important. Id. at 720. However, productivity also matters. See, e.g., In re Folding Carton Antitrust Litig., 84 F.R.D. 245, 257-58 (N.D.Ill.1979) (appropriate to reward efficiency). The Seventh Circuit did not say which of the benchmarks I am to weight most heavily or indicate how to reconcile any conflicts that might arise among them. This will occupy me at some length in the following. Finally, I may select a fee structure that awards a "percentage[ ] of the recovery to cover for litigation expenses as well," and if I do, I should not separately reimburse expenses. 264 F.3d at 720. Accord In re Fidelity/Micron Securities Litigation, 167 F.3d 735, 737 (1st Cir.1999).

    II. Award to Consumer Class Counsel

    A. Evidence from Fee Contracts in the Case

    I begin, therefore, with the attorneys' fees for consumer class counsel. As of October 31, 2001, the consumer class fund was, for purposes of this motion, about $96.2 million; the 29% award that consumer class counsel request would come to about $27.9 million. Consumer class counsel request about $1.5 million in litigation expenses as well. Consumer class counsel's first argument for this award is that the actual agreements that exist in the case support a "floor" of more than 25%, and thus support the 29% class counsel requests. I am to take into account the risk of nonrecovery faced by the counsel at the outset of the case. The contingent fee agreements with two law firms of some of the third party payers *868 who chose to be individually represented (the Health Benefit Payers, a group of more than 100 insurers) reached before joining the class provided for contingent fees of 25%. The Porter Wright Group (a group of 18 third party payers) negotiated an agreement that provided, among other options, for the client to pay monthly costs plus 15% of the final settlement. Because these insurers were "sophisticated purchasers of legal services," their agreements "define[d] the market." 264 F.3d at 720 (emphasis in original). However, these parties became involved with the litigation at a later stage when the risk of loss was slight, see id., and therefore needed less incentive to be involved than consumer class counsel did at the start when the risk of loss was high.

    Consumer class counsel argue that the Seventh Circuit established as "the law of the case" that I am to disregard the lower fee arrangements (15% rather than 25%) made with the Porter Wright Group, see id. (the Porter Wright contracts "provide little guidance" because consumer class counsel were not paid on an ongoing basis), but the Seventh Circuit said of both sets of agreements, the lower Porter Wright and the higher Health Benefit Payers contracts, that they are of "limited utility." Id. Nonetheless, the appeals court discussed both sets of agreements (and no others) as examples of the "actual agreements" I should consider, see id., and most of its opinion involved explaining the limitations of various ex post methods for guessing at what an ex ante result might be. I cannot treat the Porter Wright figures as conclusive, but if I give much weight to actual agreements at all, I cannot disregard them either, nor can I treat the Health Benefit Payers' figures as the last word, rather than as a benchmark of some real but limited utility.

    Consumer class counsel also point to their fee agreements with the representative parties, the named individual plaintiffs, which provided (as usual) for one third to 40% of the net settlement. They cite no authority that this is a relevant consideration, and contracts with the representative parties are not listed by the Seventh Circuit as examples of actual agreements that I should consider. It does not follow from the fact that some people agreed to such a percentage rate that their agreements reflect the market rate for all — lawyers have an incentive to list as the named parties those who agree to the highest contingent attorneys fees. Long ago the Seventh Circuit rejected such agreements "as wholly immaterial to the issue before the court," the determination of "a fair and reasonable fee." Milwaukee Towne Corp. v. Loew's, Inc., 190 F.2d 561, 570 (7th Cir.1951). More recently, several courts in this district have rejected the use of such contracts to set class fees. See In re Amino Acid Lysine Antitrust Litig., 918 F.Supp. 1190, 1194-95 (N.D.Ill.1996) (Shadur, J.) ("An individual named plaintiff — who may be ... perfectly competent to make judgments for itself and to have the necessary economic interest and leverage to do that — cannot fairly be permitted to impose its own determination, made on its own behalf, as a decision that automatically will bind [others]."); State of Ill. v. Harper & Row Publishers, Inc., 55 F.R.D. 221, 223 (N.D.Ill.1972) (not "fair to the class members who were unrepresented when the fee contracts were made"). Therefore I set aside the higher percentage agreements in the contracts with the named parties.

    I now turn to the issue of the level of risk. Dramatically illustrating how, at the fee petition stage, the role of class counsel shifts from fiduciary to competitor, consumer class counsel argue that their risk of loss ex ante was high because their case *869 was extremely weak. They cite several reasons provided by the Seventh Circuit why "the plaintiffs would have had a headache" with the case, 264 F.3d at 716-17,[2] some of which could have been known ex ante and therefore factored into a market-based fee agreement: the fraud claim was doubtful because "it would be hard to say" that a disclosure in the Dong report that there were interchangeable drugs was "compelled" in view of the existence of some contrary studies; winning on the merits would be dicey on the antitrust and RICO claims because of the indirect nature of the consumers' injuries. Id.; see also 100 F.Supp.2d at 680, and damages would be hard to prove when many consumers are insured, 264 F.3d at 716.

    Consumer class counsel's denigration of their own case is to be taken with a quantity of salt at this stage of the litigation. That defendants would offer to settle the case fairly quickly for a lot of money was not known beforehand, but I agree with the defendants' professional judgment that there was ex ante a serious chance of a recovery had the case gone to trial. The antitrust claim was a long shot for the consumers in view of the fact that many of them were probably indirect purchasers, and the RICO claim was colorable but difficult. See 110 F.Supp. at 681. The fraud claim was much less problematic. Although there were studies contradicting the Dong report about whether there existed bioequivalents to Synthroid, see 264 F.3d at 716-17, a jury might well have concluded that it was fraud to bury a reputable report that suggested that there were such bioequivalents. Showing damages would have involved a long but not hopeless causal chain. See 110 F.Supp. at 681 (noting that Synthroid sales fell significantly after the publication of the study). The case was not that weak. The risk of loss (whatever it was) was indeed greater at the beginning of the litigation than the risk faced of third party payers who entered the case in the middle of litigation, when settlement negotiations were already far advanced. Consumer class counsel, however, have not established that at the onset of litigation, they faced a "substantial" risk requiring an incentive at the level requested. But they faced a more serious risk than the third party payers whose agreements were at issue. Their risk could not be described as "slight."

    Risk, however, is not the only factor I am to consider. The Seventh Circuit also directs me to consider productivity. The Third Circuit explains that:

    Quality [of legal services] must be evaluated in light of the results obtained for the class, considering the complexity of the case, and the professional methods utilized in processing the case, — rewarding the use of efficient methods to expedite the case....

    In re Fine Paper Antitrust Litig., 751 F.2d 562, 584 (3d Cir.1984) (lodestar context). Because the parties came to me essentially with a settlement proposal in hand (which I rejected) if I certified the classes (which I did), I do not have the sort of information about this factor that I would have had if it been litigated before me from the start. The results obtained for the consumer class were ultimately quite good, see 264 F.3d at 717. The case was moderately but not overwhelmingly complex, and much of the complexity would have arisen only had it gone to trial, e.g., in showing causation and calculating *870 damages. The efficiency of the methods used to obtain the good results leaves something to be desired. In particular, the initial settlement proposal offered to me when the case was consolidated before me was quite unsatisfactory.[3] As far as I can tell from the record, the defendants were not especially refractory. The difficulty with attaining a satisfactory settlement was mainly on the plaintiffs' side. Had the parties come to me with a reasonable settlement proposal to begin with, the case could have been over and settled much earlier, perhaps in 1998. Consumer class counsel, therefore, do not receive high marks for productivity, and they are not entitled to recover full attorneys' fees for the work that merely prolonged the litigation. I have indeed said that for purposes of approving the settlement the litigation was indeed fairly brief compared to the size of the recovery; but there I was appraising the fairness of the settlement and not an award of attorneys' fees. I do not say that having a settlement proposal rejected by the court as not fair or reasonable is ipso facto proof of inefficiency, but it is something that can be taken into consideration in the totality of circumstances. "Time spent by counsel should be considered" in determining the percentage award, 2 Mary Francis Derfner & Arthur R. Wolf, Court Awarded Attorneys Fees § 15.02[2](i), at 15-32, 15-33, (Lexis 2001), but "[t]he courts have not as a rule spelled out how the time factor affected the percentage chosen." Id. However, it is common sense and good economics that attorneys are not entitled to payment for time wasted. See Halderman v. Pennhurst State School & Hosp., 49 F.3d 939, 942 (3rd Cir.1995) (denying full fees for "the wasteful use of highly skilled and highly priced talent"). Other factors, as noted, include the behavior of the various parties and intrinsic difficulties in reaching an acceptable agreement.

    I cannot, therefore, accept consumer class counsel's argument that they are entitled to a 29% contingency fee on the grounds that their risk was greater than those of some third party payers merely because one group of those insurers had agreed, before joining the class, to a 25% fee arrangement at the low level of risk then presented. Another group of insurers in similar circumstances agreed, as an alternative, to only a 15% fee arrangement, plus monthly costs. In determining the appropriate rate of compensation ex post, I must look at productivity as well as risk, and consumer class counsel's productivity was mediocre at best even if their overall results were good.

    B. Evidence from Large Securities and Antitrust Cases Where Contracts were Privately Negotiated

    I turn, therefore, to the second benchmark indicated by the Seventh Circuit, data from securities and antitrust lawsuits where sophisticated clients have chosen to hire counsel up front. See 264 F.3d at 720, 721. Consumer class counsel offer several contingency contracts in individual antitrust cases where sophisticated commercial clients agreed to fees of 25%-33.33% of the recovery, plus expenses. Class Counsel's Appendix Exs. M-P. Moreover, they also offer some fee structures in securities class actions established by courts near the outset of litigation through some sort of negotiation *871 process, with the orders providing for contingent fees of at least 25-28%. Exs. Q & R.[4] Consumer class counsel suggest that these should be used as "floors" rather than as "benchmarks" because class actions face additional hurdles, notably those of winning class certification. However, this does not follow. While some classes face serious difficulties in obtaining certification because of problems with the Rule 23 requirements, for others certification is less onerous. For the reasons explained in my order certifying the class, 188 F.R.D. 295, certification in this case was not such a hurdle that I should depart from the Seventh Circuit's directions to treat as benchmarks rather than floors any securities and antitrust agreements where sophisticated parties have hired counsel up front. Contrary to consumer class counsel's suggestion, the fact that the Court of Appeals reviewed the decision to certify does not show that class certification was difficult.

    Moreover, the bare fact that in the cases listed the parties assented to the fee structures indicated is uninformative without further information, not provided, about the nature of these cases, from which I might glean an idea about how analogous they might be to the case before me, or how representative they are of the market. Class counsel have not tried to disabuse me of the worry that the examples given might be neither analogous in difficulty and risk or representative, but rather constitute a biased sample chosen to make a point. I cannot, therefore, give much weight to the "data" provided by these agreements.

    C. Evidence from Auction Cases

    1. Lead Counsel Auctions and The Evidence

    The third benchmark that the Seventh Circuit indicates is the outcome of "auctions for the right to be lead counsel." 264 F.3d at 720. In this procedure, pioneered by Judge Vaughn R. Walker of the Northern District of California, and Judge Milton Shadur of this district,[5] the court, acting as sort of a virtual agent of the absent class members, solicits bids at the beginning of litigation from several firms that compete for the right to represent the class as lead counsel. These bids are appraised in terms of both the cost of the representation and the quality of counsel. Advocates of the method claim that "[i]n theory, an auction will mimic a market transaction and result in reasonable quality, low-cost representation for the class." In re Cendant Corp. Litig., 264 F.3d 201, 258 (3d Cir.2001). What is most relevant from the ex post position that faces me here are two considerations identified by the Third Circuit: (1) "auctions may lead to lower-priced representation. Under the traditional method, lead counsel (who has already been appointed) tries to get as much as it can from the court in terms of fees. Under the auction method, in contrast, prospective lead counsel compete to submit the lowest reasonable bid." Id. at 259. In addition, (2) "assuming a sufficiently large number of bidders, an auction will likely better approximate a market transaction than having a judge set attorneys fees after the fact." Id. It is an especial advantage of auctions that they offer information about what lawyers are *872 actually willing to work for, as opposed to what they would prefer to work for, that ordinary and even sophisticated consumers of legal services cannot obtain.

    The auction method has been criticized on a number of grounds, e.g., that the low bidder may not be the most adequate representative, that the prospect of lower recoveries may deter counsel from prosecuting meritorious class actions, or in some cases, may provide incentives for a sell-out if a low bidder got stuck with a case that turned out to be no good. See Cendant, 264 F.3d at 258-60. The Seventh Circuit also acknowledges reservations about auctions, see 264 F.3d at 720 ("At first thought, auctions appear to be a poor mechanism for replicating the market price of legal services."), and explains the limited utility of the auction mechanism in the ex post context, stating that it is a way of "provid[ing] detailed analysis of the market rate for attorneys facing different levels of risk." Id. Even if the levels of risk are hard to determine based on the information in the opinions or to compare with the present case, "a court can examine the bids and the results to see what levels of compensation attorneys are willing to accept." Id. at 721. Since I am not deciding ex ante whether to choose lead counsel by auction, but am appraising the data from cases where courts have chosen that method for insight into the appropriate level of compensation to award ex post, I need not dwell on the objections to adopting the auction method.

    The key thing is the level of compensation that attorneys are willing to accept at various levels of risk. And here, the most striking aspect of the results of the auctions is that, in auctions, attorneys will bid and accept figures far below what they contract for in the sort of agreements considered in the first benchmark category. Consumer class counsel argue from the data gleaned from the first two benchmarks (actual agreements in the case and actual agreements by sophisticated counsel in other cases) that actual contracts hover in the neighborhood of a 25% contingent fee recovery, plus costs.[6] Even supposing this is so, those five of the baker's dozen of auction cases[7] that actually provide bid *873 figures come in at around 15% or less for a recovery of the size we are considering. See In re Oracle Securities Litig., 132 F.R.D. 538 (N.D.Cal.1990) (providing 30% of the first million, 25% of the next $4 million, then 20% of the next $10 million, and 15% of everything above $15 million, less if the litigation was concluded in under a year), and Bank One Shareholders, 96 F.Supp.2d 780 (N.D.Ill.2000) (providing for 17% of first $5 million, 12% of next $10 million, 7% of the next ten million, with no fee for higher recoveries). In Amino Acid, most of the bidders were in that range. Judge Shadur rejected as "clearly inferior" a bid where one firm offered to take 25% of the total recovery, or 30% if the case went to trial, but to absorb its own expenses. 918 F.Supp. at 1199. In Oracle, the rejected attorneys offered higher bids in the range of the request here, but even the one rejected bid that the court found in line with the demands of the market only started at 25% and declined in steps thereafter. 132 F.R.D. at 541. In Quintus, the winner proposed to take 8.5% of the first four million, declining to 5.5% of $20 million or above, if the case got as far as a motion to dismiss, 148 F.Supp.2d 967, 978 (N.D.Cal.2001); providing for 7% of a hypothetical $36 million recovery, id. at 983; the other bids led to similar results, with only one reaching as high as 20% of such a hypothetical recovery. Id. Finally, only one of the 12 prospective lead class counsels in In re Cendant, 191 F.R.D. 387 (D.N.J.1998), came in with a bid as high as 20% for any portion of the recovery; the rest started at 0% for the first $100 million, then ranged upwards to between 1.5% and 10% of any recovery over $500 million. Id. at 388. The winner proposed a sliding scale of 2% for the first $100 million to 9% for upwards of $500 million. Id. at 391.

    These results suggest, then, that even if many attorneys can induce their clients to sign contingent fee contracts providing for recoveries of 25% or more plus costs, many attorneys whom judges have appraised as highly able are willing, in the competitive bidding situation of a lead counsel auction, to work for a lot less — as much as 10 percentage points less, which in this case, given the consumer class fund, would mean up to $10 million or more for the class — and that much less for class counsel.

    It should be noted that in arriving at the 15% figure referred to above, I consider the spread of the bids. If I were to consider the winning bids only, the figure would be about 9%. See Laural L. Hooper & Marie Learn, Auctioning the Role of Class Counsel in Class Action Cases: A Descriptive Study, at 8 & Table 1 ((Federal Judicial Center 2001) ("FJC Study")). *874 Porter Wright Groups' Response to Third Party Pay[e]r Class Counsel's Motion on Remand, Ex. D. If there were more cases in the sample, I would feel more confident about any convergence on a lower figure, but in any case all the data are relevant to the question that the Seventh Circuit has directed me to investigate, viz., what class counsel will accept by way of compensation at various levels of risk.

    2. Assessing the Evidence from the Auction Cases

    The question is what I am to make of these results, and whether, having taken into consideration the data derived from the auction cases, I should use these results to determine consumer class counsel's fees, and if, so, how. There are three alternatives: (1) I might, in view of the theoretical problems with auctions, the data available, and the facts of this case, set aside the results of the auction cases as inapplicable here after having given them due consideration. (2) Or I might give those results some but not conclusive weight, also giving weight to data derived from the other two benchmarks indicated by the Seventh Circuit, actual agreements and agreements in securities and antitrust cases. (3) Finally I might follow the lead of the auction cases and order a fee structure based on the level of compensation indicated therein. The Seventh Circuit has not provided any direction in choosing among these alternatives. For reasons that will become clear from discussion of the theoretical basis of the auction approach, the second alternative is hard to maintain in a principled way; the arguments suggest, rather, that I should accept an award based on either the auction data or the actual contract data, but not both.

    a. Legal and Economic Considerations on the Market for Consumer Class Action and Other Legal Services

    Consumer class counsel ask me not to apply the auction benchmark at all because (they say) lead counsel auctions do not replicate the private market for legal services. They cite the Report of the Third Circuit Task Force on Selection of Class Counsel ("Report"), released in October 2001, which expresses criticisms of the method for, among other things, failing to account for other factors than mere cost that go into the selection of counsel, such as choice of the most qualified lawyers. See Report at 58-60.[8] However, the Third Circuit itself provisionally approved the method for some uses, though not in securities cases, and not the district court's rationale for it, in Cendant. See 264 F.3d at 279. More to the point in this district, the Seventh Circuit, after carefully considering the limitations of auctions, expressly directs me to consider data derived from such auctions at least as relevant to determining what class counsel's compensation should be. 264 F.3d at 719.

    But consumer class counsel correctly identify a tension in the Seventh Circuit's opinion. The appeals court states, on the one hand, that certain agreements made by some insurers, who are sophisticated purchasers of legal services, "define the *875 market" (though unfortunately too far into the case to accurately reflect the risk at the outset). 264 F.3d 712 at 720. Consumer class counsel suggest that I read this to say that such agreements provide a floor on attorney compensation. But on the other hand, in directing me to consider the auction cases, which point to much lower figures, as source of data on attorney willingness to accept certain levels of compensation, the Seventh Circuit implies that the outcomes of actual markets are not necessarily conclusive of the result in a market-based inquiry. The following theoretical and empirical considerations may dispel the appearance of paradox.

    First, it is axiomatic that the justification for a market-based approach is that it maximizes consumer welfare. According to standard economic theory, competitive markets do this by driving down price to marginal cost, while giving producers of services an incentive to be more productive. See, e.g., Business Electronics Corp. v. Sharp Electronics Corp. 485 U.S. 717, 757, 108 S.Ct. 1515, 99 L.Ed.2d 808 (1988) (Stevens and White, JJ., dissenting on other grounds) (antitrust context). As is well known, that conclusion depends on certain simplifying assumptions — ideally competitive markets, perfect consumer information, zero transactions costs, and the like. See, e.g., Elizabeth Hoffman & Mathew L. Spitzer, The Coase Theorem: Some Experimental Tests, 25 J.L. & Econ. 73 (1982). Those conditions do not strictly obtain in the real world, yet if they were not approximately true in many cases, then it would be a miracle if economic theory had any degree of predictive or explanatory power. But if these conditions fail to obtain to a sufficient degree, even the rough efficiency of the market outcome can no longer be presumed. This is a simple matter of logic: the conclusion (that outcomes are efficient) no longer follows from the premises (that the conditions of efficient markets obtain). See, e.g., Eyal Zamir, The Inverted Hierarchy of Contract Interpretation and Supplementation, 97 Colum.L.Rev. 1710, 1780 1801 (1997). "To the extent that market failures prevail, due to the existence of monopolies and cartels, information problems, problems of public goods, prisoner's dilemmas, and high transaction costs, ... enforcing contracts that appear to reflect the unfettered will of the parties will not maximize social utility." Id. at 1790.

    Second, the actual market for legal services departs rather sharply from the ideal conditions under which markets are efficient, and therefore in theory conducive to consumer welfare. Professor Hadfield of the University of Toronto Law School, an economist as well as a legal scholar, sums up a thoughtful analysis by saying:

    The market for lawyers is fundamentally noncompetitive. As a consequence of the complexity of legal reasoning and procedure, the profession's derived monopoly on the legitimate use of coercion, and the unification of the profession to serve the diverse needs for access to law, the price of law that emerges from the free market for lawyers is too high.

    Gillian K. Hadfield, The Price of Law: How the Market for Lawyers Distorts the Justice System, 98 Mich.L.Rev. 953, 999 (2000). Especially relevant to assessing attorneys' fees for consumer class counsel in particular is Professor Hadfield's conclusion that higher prices are due to "the noncompetitive nature of the market, which places the emphasis on the wealth of clients rather than on the cost of services." Id. at 963. The legal services market is segmented between individual clients, like the consumers here, and large, sophisticated, institutional actors with market power, like the insurers. The structure of the market means that, even taking into account the necessity for higher fees to provide *876 an incentive for attorneys to take risky cases involving contingencies, the bias of the commercial segment inflates the prices in the individual segment. "The result is that of a classic monopoly: price is determined by wealth, not cost." Id. at 997.

    Discussing consumer contingency fee arrangements, Professor Brickman of Cardozo Law School points to an additional source of noncompetitiveness in this segment of the legal services market. He notes that there, unlike the situation in a fairly competitive market, "the average consumer of legal services has little idea of what alternative prices are charged or available for various legal products.... [T]he attendant ability to comparison shop remains minimal."[9] Lester Brickman, Contingent Fees Without Contingencies: Hamlet Without the Prince of Denmark?, 37 UCLA L.Rev. 29, 104 (1989). He concludes that "[c]ontingent fee lawyers are charging monopoly rents — the earnings yielded by restrictions on competition." Id.

    In support of the proposition that "contingent fees yield more than competitive rates of return," Brickman offers the following considerations: (1) the fact that the standard contingent fee percentage remains one-third "despite the substantial decrease in the risk of nonrecovery during the past twenty-five years"[10] and "this [is] undoubtedly reflected, as well, in settlements." Id. (citation omitted). Moreover, (2) despite a substantial increase in the average award in real dollars in tort cases, the one-third fee remains the standard "without any concomitant increase in risk or effort." Id. at 106 (same). Brickman points out, further, (3) that "the existence of a standard one-third contingent fee itself indicates the absence of competitive pricing," because "[i]f competitive market forces were effectively operational in the contingent fee setting process, the percentages charged by lawyers would come to reflect, albeit roughly, the likelihood of success in each case. With a few exceptions, price appears unrelated to risk." Id. (same).

    In the context of the consumer class counsel contingency fees that I address here, these sorts of considerations are even more pertinent, for reasons related to the ones that Judge Shadur adumbrates in rejecting the individual fee contracts of the representative party as a benchmark in setting class counsel compensation levels. See Amino Acid, 918 F.Supp. at 1194-95 (improper to let individuals make binding decisions for others). In a related context, Judge Walker explains:

    An individual client can search for and strike a bargain for representation with any lawyer or law firm that does the type of work needed. But class members do not go shopping for their lawyers, or indeed do anything else in the case other than agree to lend their names to the litigation. Plaintiff[s'] lawyers in class actions generate the market for their own legal services.

    In re Wells Fargo Sec. Litig., 156 F.R.D. 223, 226 (N.D.Cal.1994) (rejecting a "steering *877 committee" of plaintiffs' lawyers). That is not a bad thing, but it strongly suggests that the market for the legal services of consumer class counsel deviates sharply from the conditions under which markets benefit consumers.

    If legal markets in general are moderately uncompetitive, the market for contingent legal services, especially among consumers, is highly uncompetitive, and the market for consumer class representation, especially of absent parties, most of all. Imperfect competition is the rule in the real world, see Thomas C. Arthur, The Costly Quest for Perfect Competition: Kodak and Nonstructural Market Power, 69 N.Y.U.L.Rev. 1, 13 (1994) ("[P]erfect competition is an artificial, abstract model which can exist only in theory."). Actual markets do not normally utterly fail to realize the benefits that ideal markets promise in principle, but the more that actual markets deviate from the ideal conditions, the less likely they are to maximize consumer welfare. In the market I consider here, it is highly questionable whether the conditions under which legal services are provided allow actual markets to fully realize the virtues of competition, or whether the actual market prices are efficient.

    Several conclusions seem to follow from this analysis. First, the results of a procedure such as an auction may, in the proper circumstances, be better at mimicking the outcome of a genuinely competitive market than the results of the rather distorted actual markets for legal services for consumer class counsel. Those circumstances include: a reasonable number of law firms actually engaging in competitive bidding, adequate information before the "purchaser," that is, the judge, as to price and quality of the law firms, and reasonably low costs for making a decision that would further the interests of the class members. If these circumstances obtain, then I should adopt the auction approach as my main benchmark, and give less if any weight to the information about actual agreements derived from real world markets that are distorted in the ways described. Second, if the circumstances of the auctions were not such that they better approximated the efficient results of competitive markets, I should accept consumer class counsel's criticisms, and set aside the auction benchmark altogether, as class counsel argues that I should do. Third, if auctions are either better or worse than actual markets at approximating efficient outcomes,[11] an attempt to compromise between the results of actual markets and auctions would be unwarranted: it would amount to intentionally deviating from better towards worse results, whichever method produced the worse results.

    b. Applications

    Applying these results, I conclude that in the circumstances, the auction benchmark is the better market-mimicking guide to an efficient outcome in a truly competitive market. The auction cases for which I have the bid figures present sufficiently large numbers of firms bidding to create real competition: six in Quintus, 148 F.Supp.2d at 982, twelve in Cendant, 191 F.R.D. at 388-89; eight in Amino Acid, 918 F.Supp. at 1197-1200; four in Oracle, 132 F.R.D. at 539-541; and nine in Bank One S'holders, 96 F.Supp.2d at 785-87. The bidding in all cases was competitive, and the judges who ran the auctions were informed in detail about the amount of compensation the bidders would accept and their qualifications as well as about *878 the qualifications of counsel, and almost certainly more informed than even sophisticated clients about what compensation the attorney would accept. There is no reason to think that the costs of gaining this information were particularly high. The dimensions on which these auctions diverge from actual markets, in competitiveness and information, here favor the auction benchmark over the actual agreement benchmark.

    Consumer class counsel do not effectively argue the contrary. Aside from referring me to the reservations in the Third Circuit Report about whether auctions simulate actual markets or ought to be used to choose lead counsel, neither of which are pertinent here for reasons explained, consumer class counsel really offer only one argument that, in the circumstances presented here, the results of the auction cases should be disregarded. They attempt to distinguish the case at bar from several auction cases on the basis that this was a riskier case. They argue, implausibly and inconsistently, that I should ignore all securities cases (except the ones that came out with figures in their ballpark) on the grounds that these are intrinsically less risky because more likely to be certified. This aside, they note that in Cendant, which provides the lowest figures, the Third Circuit stated that the basic liability issue was simple. 264 F.3d at 285. However, that court also said: "[t]he issue of damages against Cendant is different in character, for it involves technical accounting issues and hence can be quite complex." 264 F.3d at 234. But the very low range of all but one of the bids in Cendant suggest that competent counsel appraised it as a fairly low-risk case, so I will treat it as an outlier and not give its outcome much weight. Consumer class counsel also attempt to distinguish Amino Acid on the grounds that prior to the auction, reports about the defendants' participation in a price-fixing conspiracy, including incriminating tape recordings, had been widely circulated in the press. Actually, however, that suggests an analogy to the present case, which was triggered by the discovery and subsequent publication of a report that led counsel to conclude that the defendants had committed fraud, antitrust and RICO violations.[12]

    Consumer class counsel have not suggested persuasive reasons why the variance between the attorneys' bids in the auction cases and the attorneys' fees in the actual agreements should be explained by the variances in the level of risk in the cases involved. Moreover, given the record before me and the information in the auction cases, it is not evident that the level of risk in the cases where bid figures are available, except for Cendant, varies as much from the moderate but not substantial risk faced by consumer class counsel here as would be required to account for the difference between what attorneys were willing to take there and what consumer class counsel ask here. Because the auction cases that I consider involve a method that best tracks the efficient result of an ideal market, I use as my main benchmark the auction cases, and I determine the attorneys' fees for consumer class counsel as follows.

    D. The Award

    I am directed to attempt to estimate ex post what the market for attorneys' fees would have produced ex ante, in view of my estimates of the risk faced at the time of filing and productivity shown subsequently. The data I have include (a) two *879 mid-case, low-risk insurers' attorneys' fee contracts providing for fee structures ranging from a contingent fee of 25% to an alternative of 15% (with monthly costs paid), and (b) some contracts providing for fees of about 25% in securities and antitrust cases where sophisticated clients retained counsel on a contingency basis, but where consumer class counsel have not provided enough information about the agreements for me to know whether the cases are analogous or representative. Finally, I have (c) the auction results in four cases where bids are provided, where the accepted bids range in the neighborhood of 15% of the overall recovery, with the exception of Cendant, where the winning bid was much lower. The majority of the winning bids came in at around 9%; I take the rough average of all the bid figures, or 15% because that enlarges the sample size somewhat, but if consumer class counsel appeal, it is possible that the Seventh Circuit will conclude that only the winning bids matter, and reduce the award by six percentage points, from 15% to 9%. For the reasons explained, I give comparatively less weight to the actual agreements, and I treat the very low figures in Cendant as anomalous or irrelevant to my determination here.

    A sliding scale is indicated for the usual economic reasons — the declining marginal productivity of attorney effort and the existence of returns to scale,[13]see Matter of Continental Illinois Sec. Litig., 962 F.2d 566, 572 (7th Cir.1992) ("In large commercial litigation with prospects of multimillion dollar recoveries the percentage frequently is tapered — it might be 33 percent of the first million, 25 percent of the next million, and so on down."), but it should not involve a recovery cap or other disincentive. The Seventh Circuit disparaged Judge Shadur's choice of a bid in Amino Acid that set a cap on a fee award, "eliminating any incentive to push for a larger recovery." 264 F.3d at 720. But the appeals court looked with more favor on sliding scales without caps chosen in cases like Oracle. Because marginal returns decline continuously, a smooth curve is preferable, as in Oracle.

    I award consumer class counsel attorneys' fees, including litigation costs and expenses, on a scale calculated as follows: 30% on the first $10 million recovered, 25% on the second $10 million, 20% on the third $10 million, 15% on the fourth ten million, and 10% on all amounts above that. This comes to about $14.62 million of the consumer class's share of the common fund based on the October 2001 figures, and corresponds to an award of a little more than 15% of the baseline consumer class recovery of $96.23 million. This is in the range of the auction cases, which shows that it is what attorneys will in fact accept to prosecute a case like this. Insofar as this provides relevant information, it is also in accord with the low range of data from the actual insurers' agreements in this case (the agreement for (in the alternative) 15% contingency plus monthly costs). Here class counsel's ex ante risk was moderately high, and their results were good, but their productivity was moderately low because of the unsatisfactory first settlement proposal in the circumstances. Risk must be discounted against less than impressive productivity even in view of the good results. As the Seventh Circuit has authorized, see 264 F.3d at 722, I have made a substantial allowance for costs and expenses in my *880 percentage award rather than awarding costs separately.

    III. The Fee Award to the Third Party Payers

    Counsel for the class of third party payers (the "TPP class") seek 22% of that class's share of the settlement. That fund was about $45.5 million in October 2001, so they seek about $10 million. They also seek more than $600,000 in expenses. In support of their attorneys' fee petition, and following the Seventh Circuit's directions as set out above, they offer the following considerations. First, there are fee agreements that about ten of the third party payers made in this case before they joined the class. TPP class counsel say that none of these provided for a sliding scale, most were contingent fee agreements, and they ranged from 15% to 33.33% of the recovery (plus costs), with the majority of the agreements clustering around 25-30%. Only two of the almost 3,000 third party payers objected to the 22% fee request, and the 22% rate, the TPP class counsel say, represents an approximation of what the whole of the class would have agreed to at the outset of the case. Second, with regard to other data, TPP class counsel offer a number of contracts and cases indicating that the 22% request is actually somewhat below market rates. Thus James Dugan, lead TPP class counsel, offers an affidavit stating that in other class actions involving prescription drugs, he represented several other insurers for, on average, a 25% contingency fee. Ex. 3 at 3-4. TPP class counsel also point out that Judge Kocoras, construing the Seventh Circuit's standard, awarded 25% of a $697 million fund in another prescription drugs class action. See In re Brand Name Prescript. Drugs Antitrust Litig., No. 94 C 897, 2000 WL 204112, at *2 (N.D.Ill. Feb.10, 2000) (noting, however, that this was a huge, lengthy, and difficult case, with superb results). They offer no data based on auction cases.

    The first issue is whether to use the figures based on the auction cases as the main benchmark for the attorney's fees for TPP class counsel, or to here follow the benchmarks indicated by the actual agreements reached in the real world market. Many of the theoretical considerations that suggest that the data from the auction cases provide the best benchmark for determining appropriate compensation for consumer class counsel are not applicable here. If the overall market for legal services is distorted by an inflationary bias towards commercial customers, as Professor Hadfield suggests, see above, in a way that at least raises doubts about using those figures in the instance of consumer class counsel, the insurers here are the commercial customers in question. If ordinary markets for contingent legal services are uncompetitive because individual clients lack the time, ability, or knowledge to comparison shop, the insurers are "sophisticated purchasers of legal services," 264 F.3d at 720, whose choices define the market.

    Concerns remain about the competitiveness of even this market in view of comparatively flat percentage agreements in the neighborhood of 25%. Moreover, the auction case that seems most relevantly analogous because the class members were sophisticated commercial clients is Amino Acid, which is in the usual 15% range for such a case. 918 F.Supp. at 1196-1201. The other auction cases that give the counsel's bids, Cendant aside, involve shareholder classes that are more like the individual consumers in the consumer class here than like the insurers in the TPP class.

    However, these are comparatively frail considerations, particularly because here, if I set aside the shareholder cases, there *881 is really only one data point from an auction case. Some of the theoretical arguments discussed above that explain the variation of the auction figures from the market figures in the consumer cases do not apply at all in large commercial ones (lack of sophistication, bias towards commercial customers), or apply only weakly and generally (complexity of legal argument, professional monopoly on provision of legal services). Therefore, it is hard to be confident, based on the evidence available, that a percentage figure in the 15% range of Amino Acid represents the market rates for this segment of the market. If there were three or four such cases involving commercial clients with figures in a similar range, the result might be different. I should still want a theoretical explanation of the variance, but at least I would have a tendency that required explanation. A single case does not a tendency make. In the circumstances, I would not be comfortable using the auction data as my benchmark based on a single analogous case. Therefore, with the TPP class counsel, I give little weight to the auction cases, and adopt as the main benchmarks the data from the actual agreements both here and in other cases where sophisticated customers purchased legal services.

    TPP class counsel argue that, using these benchmarks, their 22% attorneys' fee is reasonable, even below market. I am concerned that this award is high, especially in view of the fact, noted by the Seventh Circuit, that the risk of loss to the TPPs, entering the case after class certification and with a settlement proposal on the table, was "slight." 264 F.3d at 719. Nonetheless, the Seventh Circuit so commented in the context of insisting that the market rate is reasonable. The data indicate that it is quite likely that had the TPP class members struck agreements with counsel before litigation commenced, they would have been in that range. The individual contracts of the TPPs, mainly drawn up after the likelihood of settlement was clear, show that the TPPs were willing to pay at least 22% for the most part even in those circumstances.

    As noted in discussing the consumer class counsel award, the Seventh Circuit has also stated that in making an ex post determination about the appropriate attorneys' fee for class counsel, I am to take into consideration not only the risk viewed, to the extent possible, from the perspective it would have been seen ex ante, but also productivity. 264 F.3d at 722. The results for the TPP class were excellent, and there is no similar ground for discounting TPP counsel's productivity as the ground that applied in the case of consumer class counsel. TPP class counsel did not participate in the litigation that produced the initial unsatisfactory settlement proposal, and indeed, intervened when it became clear that the initial settlement proposal would bar any recovery for their clients. See 1998 WL 526566, at *2. Therefore I cannot reduce the award of TPP class counsel on this basis. Nor is it clear that I should. They did well by their clients who would have received nothing themselves had they failed to intervene effectively; and who, moreover, once in, might possibly have taken much of the benefit that ultimately went to the consumer class had they "held back and sued in subrogation," 264 F.3d at 717, but did not do so. Therefore I find the award appropriate. TPP class counsel requests reimbursement for expenses in the amount of $620,998.40. I have reviewed the supporting materials for this request, and given what sophisticated clients appear willing to pay for like outlays, I grant the petition, as per Seventh Circuit directions, 264 F.3d at 722.

    *882 The Health Benefit Payers Group objects to such an award, arguing that the Seventh Circuit's opinion requires that: (1) any award to TPP class counsel be no greater than that contemplated by a fee agreement of July 1998 with the individually represented companies and their retained counsel, limiting the TPP class counsel fee to $5 million, or (2) I award TPP class counsel fees on the portion of the fund attributable to class members other than the individually represented companies, either as a matter of equity or because this was the fee agreement actually reached. However, as I have explained in my previous opinion, "[t]he Health Benefits Payers and the insurers that opted out lack standing to object. Nonparties have no standing [to object to a settlement proposal]...." 110 F.Supp.2d 676, 682 (citing Agretti v. ANR Freight Sys., Inc., 982 F.2d 242, 246 (7th Cir.1992)). No more do the Health Benefit Payers have standing to kibitz about the attorneys' fee awards. They are not parties. Moreover, they admit that the TPP class withdrew from the purported agreement before it was finalized. If the Health Benefit Payers Group has some enforceable agreement with the TPP class counsel that has not been honored, that is another case, and they can sue in the usual way if the suit is not time barred or otherwise will not lie.

    The Porter Wright Group of insurers also responds to TPP class counsel's request, and, as I have held, the Porter Wright Group does have standing to object for the class members among its clients who have not opted out, 110 F.Supp.2d at 682, and, thus, to object to the TPP class counsel fee petition on behalf of the class members among its clients. The Porter Wright Group valiantly upholds the correctness of my original 10% award for TPP counsel using the Seventh Circuit's benchmarks. Although I agree that there is an inverse proportion between the size of the claim to a share of the recovery and the fee counsel will accept, I do not think that the Porter Wright Group's arguments carry the day. The central argument is that Aetna, which the Porter Wright Group states is the largest class member with the most covered lives, and an entity that "dwarfs" the other class members, reached an actual agreement for 15% of the class recovery. But I do not see why I should take the market to be defined by the biggest player rather than an average, which supports a higher figure. The only thing that would support such a reduction would be if I took as my benchmark the accepted bids (and not just the offered ones) in the auction cases, but I have explained above that this would be inappropriate here, because it limits the small size of the already small sample, and because I am directed to use the data from the auction cases as evidence of what attorneys will accept in cases like this, and all the bids, not just the winning ones, are presumably evidence of what the bidders would accept.

    IV. Objectors' Counsel's Fee Petition

    Objectors' counsel petitions for attorney's fees of $134,550 and expenses of $3,828.21. As objectors' counsel acknowledges, "an objector to a class action settlement is not generally entitled to an award of counsel fees." In re Domestic Air Transp. Antitrust Litig., 148 F.R.D. 297, 358 (N.D.Ga.1993). Where the objections filed, however, "produced a beneficial effect upon the progress of the litigation, an award of fees in appropriate." Frankenstein v. McCrory Corp., 425 F.Supp. 762, 767 (S.D.N.Y.1977). As the Northern District of Georgia stated, "[i]f objectors' appearance sharpens the issues and debate on the fairness of the settlement, their performance of the role of devil's advocate warrants a fee award." Domestic Air Transp., 148 F.R.D. at 358. That is not *883 the case here. The objectors' argument that they sharpened the issues turns on the fact that the Seventh Circuit agreed with them about their right to intervene. This is the "profound benefit Objectors have conferred [on the class]." "But for the Objectors' participation, the standards granting both intervention and filing of contingent notices of appeal would not have been so clearly articulated," they say. I disagree, and so did the Seventh Circuit: The standards for intervention and the contingent notice of appeal procedure were stated in Blair v. Equifax Check Services, Inc., 181 F.3d 832, 834-35 (7th Cir.1999), and were "obvious to class members in other cases." 264 F.3d at 716. See, e.g., Cusack v. Bank United of Texas FSB, 159 F.3d 1040 (7th Cir.1998). The objectors, however, "failed to file a contingent appeal and called the approach of Blair and Cusack nonsensical. Why they spent time and money arguing the point rather than filing a precautionary springing notice of appeal from the final judgment is puzzling. After oral argument, however, the objectors took our advice...." Id. Whatever benefit the objectors may have conferred does not merit an award of attorneys' fees.

    V. Conclusion

    Consumer class counsel are awarded attorneys' fees, including litigation costs and expenses, on a scale calculated as follows: 30% on the first $10 million recovered for the class, 25% on the second $10 million, 20% on the third $10 million, 15% on the fourth ten million, and 10% on all amounts above that. I award third party payer class counsel attorneys' fees of 22% of the third party payer class fund, and $620,998.40 in litigation costs and expenses. I deny the objectors' counsel's petition for attorneys' fees and costs.

    NOTES

    [1] Evidence that the TPP have put in the record, in the form of a research memorandum by Professor John Coffee of Columbia Law School, indicates that more recently, courts have been more willing to allow attorneys' fees in the 29% range in "megafund" cases. Ex. B1 at 9-10. Different studies, however, give different figures: In a recent case, Judge Hart of this district cites an analysis of Beverly Moore, editor of Class Action Reports, based on fees and costs awarded in 757 common fund cases that used both the percentage method and the lodestar/multiplier method: the fees and costs represented 19.6% of the recoveries on average. See Montgomery v. Aetna Plywood, Inc., 1999 WL 172313, at *4 (N.D.Ill.1999).

    [2] The Seventh Circuit also cites several reasons that concern developments that occurred after the filing of the case-FDA and physician behavior since that filing, and Synthroid's current sales. Id. at 716-17. These factors could not have been known ex ante and are therefore properly disregarded.

    [3] As I have explained in previous opinions, first, under that proposal, the consumer class would get too little of the settlement. See 1998 WL 526566, at *2. Second, the number of claims, in theory as many as 5 million (it turned out to be about .75 million), was unknown, and consumer class members would receive as little as $13 a person. Id. Third, the notice provisions of the proposed settlement — putting terse and uninformative ads in mass audience major publications — would not do. See 110 F.Supp.2d at 680.

    [4] Elsewhere in their brief they argue that I should not use auction results at all because these are not properly market-mimicking, and I should not use securities cases at all because these are intrinsically less risky.

    [5] The approach was pioneered by Judge Walker in In re Oracle Securities Litig., 131 F.R.D. 688 (N.D.Cal.1990). Judge Shadur's auction cases include, e.g., In re Amino Acid Lysine Antitrust Litig., 918 F.Supp. 1190 (N.D.Ill.1996). The others are listed below.

    [6] This is considerably lower — 7 percentage points — than the average attorneys' fees awarded by courts in class actions, according to Professor Coffee's Declaration, Third Party Payer Ex. B1, see note 1 above; Professor Coffee also cites research indicating that the average attorneys' fee award in a class action is around 32%. Declaration at 6-7. That suggests that courts routinely overcompensate attorneys in class actions of all sorts. With large recoveries, the difference matters a great deal. In the case of the consumer class fund here ($96.2 million), seven percentage points is $6.7 million dollars.

    [7] In Cendant, the Third Circuit located only thirteen such cases by August 2001, 264 F.3d at 258 & n. 35 (citing Oracle, 132 F.R.D. 538 (N.D.Cal.1990) (Walker, J.); In re Cendant Corp. Sec. Litig., 109 F.Supp.2d 235 (D.N.J. 2000) (Walls, J.), In re Amino Acid Lysine Antitrust Litig., 918 F.Supp. 1190 (N.D.Ill. 1996) (Shadur, J.); In re Auction Houses Antitrust Litig., 197 F.R.D. 71 (S.D.N.Y.2000) (Kaplan, J.); In re Wells Fargo Sec. Litig., 156 F.R.D. 223 (N.D.Cal.1994) (Walker, J.); Raftery v. Mercury Fin. Co., No. 97 C 624, 1997 WL 529553 (N.D.Ill.1997) (Lefkow, J.); Sherleigh Assoc., LLC v. Windmere-Durable Holdings, Inc., 184 F.R.D. 688 (S.D.Fla.1999) (Lenard, J.); Wenderhold v. Cylink Corp., 188 F.R.D. 577 (N.D.Cal.1999) (Walker, J.); In re Lucent Tech., Inc. Sec. Litig., 194 F.R.D. 137 (D.N.J.2000) (Lechner, J.); In re Bank One S'holders Class Actions, 96 F.Supp.2d 780 (N.D.Ill.2000) (Shadur, J.); In re Comdisco Sec. Litig., 141 F.Supp.2d 951 (N.D.Ill.2001) (Shadur, J.); In re Quintus Sec. Litig., 201 F.R.D. 475 (N.D.Cal.2001) (Walker, J.)).

    Most of these cases do not provide or, in at least one instance, disclose, the bid figures; some only order that bidding be done. I have discussed all the ones that do provide bids in the text. In addition, subsequently to the Third Circuit's opinion in Cendant, Judge Walker disclosed the bids in Quintus, see 148 F.Supp.2d 967 (N.D.Cal.2001), and I have taken these into account as well. Judge Walker had another auction case where the bid spread was unclear, In re Calif. Micro Devices Sec. Litig., No. C-94-2817-VRW, 1995 WL 476625 (N.D.Cal. Aug.4, 1995), because the court asked for new bids. The final accepted bid was 15.7%. See Laural L. Hooper & Marie Learn, Auctioning the Role of Class Counsel in Class Action Cases: A Descriptive Study, at 8 & Table 1 ((Federal Judicial Center 2001) ("FJC Study")). Porter Wright Groups' Response to Third Party Pay[e]r Class Counsel's Motion on Remand, Ex. D.

    I note that pursuant to the Seventh Circuit's directions to look to the auction cases for information on what compensation attorneys will accept, 264 F.3d at 721, rather than for persuasive legal authority, I take judicial notice of the facts about the bids made and accepted. These are facts in the contents of court records about which judicial notice is appropriate. See General Elec. Capital Corp. v. Lease Resolution Corp., 128 F.3d 1074, 1081 (7th Cir.1997); Fed.R.Evid. 201(b). In view of this, I omit the subsequent histories of the cases in most instances, because only these findings matter, whatever the ultimate disposition of the case.

    [8] I do not think the Report does justice to the weight the "auctioneer" judges gave to non-price considerations, including the caliber of class counsel. See, e.g., Oracle, 132 F.R.D. at 542 ("All of the bidders are prominent, well established firms specializing in the type of litigation at bar. All appear adequately qualified to serve as class counsel, and it is impossible objectively to distinguish among these firms in terms of their background, experience and legal abilities."); Amino Acid, 918 F.Supp. at 1193-96 (agreeing that qualitative factors such as experience, knowledge, reputation, and resources are highly relevant); Auction Houses, 197 F.R.D. at 82 (noting that auctions receiving many bids involve large numbers of able plaintiffs' attorneys).

    [9] Judge Walker usefully discusses the experience of the proposed lead plaintiffs in Quintus, which, if typical, may be illustrative. One contacted only one firm, with which he signed a fee agreement; another did "some preliminary research," but contacted only one firm; a third interviewed lawyers from at least eight firms before hiring counsel, a fourth had not signed. 201 F.R.D. at 479-80. These were all sophisticated investors, college-educated individuals who alleged substantial personal losses from $59,000 to $633,000. And only one engaged in serious comparison shopping.

    [10] Professor Brickman states that the success rate at trial for contingent fee lawyers has doubled in that period. Id.

    [11] In principle they might be equally good methods, but then it would be hard to explain why their results diverged so sharply.

    [12] Consumer class counsel also argue that in Auction Houses, criminal liability was established and the class had been certified. However, I do not have the bids in Auction Houses, and so do not use figures from that case.

    [13] Consumer class counsel's argument that a sliding scale is inconsistent with the Seventh Circuit's statement that actual agreements for flat percentages "define the market," 264 F.3d 712 at 720, is itself inconsistent with the Seventh Circuit's direction to consider the auction cases, and its approval of such awards in those cases.