Eleven law firms appeal from an order allowing fees out of a fund resulting from the settlement of class actions charging a conspiracy to fix prices in violation of Section 1 of the Sherman Act. 15 U.S.C. Sec. 1 (1982).1 The order was entered after a hearing on objections to fee applications made on behalf of some members of the plaintiff class.2 We remand for further proceedings consistent with this opinion.

On March 3, 1978, twelve separate fine paper pricefixing lawsuits were transferred by the Judicial Panel on Multi-district Litigation for coordinated and consolidated pretrial proceedings to the Eastern District of Pennsylvania, where four such actions were already pending.3 The first of these sixteen actions had been filed on July 18, 1977 in the Northern District of Illinois by the law firm of Specks & Goldberg, Ltd. Other actions were later filed in different districts around the country, and these, too, were eventually transferred to the Eastern District of Pennsylvania.

Prior to the March 3, 1978 transfer, some members of the law firms which had filed the initial cases met, first in Atlanta in January of 1978, and again in Chicago on February 3, 1978, to organize a common strategy for the handling of the litigation. At the February 3, 1978 meeting these plaintiffs' lawyers, acting as a committee of the whole, elected an Executive Committee, which designated Granvil I. Specks, Esq., of the Chicago firm of Specks & Goldberg, Ltd., and Harold Kohn, Esq., of the Philadelphia firm of Kohn, Savett, Marion & Graf, P.C., as co-chairmen.4 Both Specks and Kohn were experienced in the successful representation of plaintiffs in Fed.R.Civ.P. 23(b)(3) class actions. Those present at the February 13, 1978 meeting also agreed upon a lead counsel team, comprised of Specks, Kohn, Joseph Cotchett, John Noll and, later, Seymour Kurland.

Shortly after the March 3, 1978 transfer order, and ten months prior to any class certification, the trial court held a pretrial conference at which, over the objection of the Attorney General of California, who represented one plaintiff, it approved the organizational structure proposed by plaintiffs' counsel.5 App.A. 130-43. The court by order expressly approved the creation of the Executive Committee, and expressly contemplated the establishment by that committee of additional standing committees. At the time of this pretrial conference a number of subcommittees, including a Rule 37 Subcommittee, a Plaintiffs' Discovery Committee, a Compliance With Defendants' Discovery Committee, an Industrial Analysis Committee and a Finance Committee, had already been created by the Executive Committee. App. O 262-67.

The same Pretrial Order which approved the organizational structure of plaintiffs' counsel also fixed a 30 day time limit for the filing of a motion or motions for class certification, and a 90 day time limit for completion of discovery relating to class action issues. App. A 138-39. It was not until February of 1979, however, that the court certified a national class of direct purchasers of fine paper.6

Meanwhile, however, settlement discussions went forward with certain fine paper manufacturer defendants. On July 25, 1978, counsel for the plaintiffs received an offer from St. Regis Paper Co. to settle its total liability for $2 million. Five other defendants made settlement offers between that date and January 16, 1979, when settlement offers totaling $30 million were in hand. Shortly after the entry of the court's February 1979 order certifying a national class of direct purchasers of fine paper the plaintiffs moved, on March 5, 1979, for preliminary approval of the proposed settlements with the six defendants whose offers totaled $30 million. That proposed settlement was opposed by those paper manufacturer defendants who had not yet offered to settle, some of whom had previously filed crossclaims for contribution. Judicial approval of the partial settlements was not obtained until September 2, 1979. Pretrial Order No. 66.

By the time the $30 million settlement with six fine paper manufacturers had been approved, 23 additional separate lawsuits had been added for a total of 38. Fourteen of these were separated from the plaintiff class by virtue of the class certification order.7 Thus twenty-four individual actions were on file on behalf of the class at the time of court approval of the initial round of settlements. Several of these suits had been filed after the soon to be approved offers totaling $30 million were already in hand.

In December of 1979 the trial court entered an order which consolidated discovery in all the fine paper cases, directed that discovery be completed by July 3, 1980, and fixed September 22, 1980 as the trial date. The same order directed plaintiffs to file a pretrial memorandum on August 29, 1980, "in the form to be prescribed by the Court." Pretrial Order No. 76, App. A 227. Thereafter, on January 8, 1980, the court, in Pretrial Order No. 83, prescribed the form of the plaintiffs' pretrial memorandum.8 Other trial preparation went forward against the remaining defendants. That preparation ultimately involved deposing approximately 250 witnesses, inspecting and copying thousands of documents, briefing and arguing numerous motions, and preparing a 643 page pretrial memorandum.

A week before the September 22, 1980 trial date Champion International Corporation offered to settle its liability for $4 million, and on the first day of trial the remaining defendants offered an additional $16,650,000. Added to the already approved $30 million partial settlement, these offers brought the total settlement fund to $50,650,000. Under the terms of the settlements the defendants were to be relieved of liability both for damages and for attorneys' fees, and would immediately pay over the settlement fund so that it would earn interest on behalf of those interested in the recovery until it was disbursed. The settlement thus was structured to create a fund in court from which attorneys' fees could be paid; a practice approved by the court in Lindy Bros. Bldrs., Inc. of Phila. v. American Radiator & Standard Sanitary Corp., 487 F.2d 161 (3d Cir.1973) (Lindy I ).

The trial court fixed March 20, 1981 as the deadline for filing petitions for fees and expenses from the fund. Pretrial Order No. 143, as amended by Pretrial Order No. 147, App. J 472. Prior to that date negotiations took place among members of the plaintiffs' Executive Committee and others in an effort to reach agreement on a ceiling on fees which would be sought by the attorneys for the class. These negotiations were unsuccessful, and on March 20, 1981, 41 separate petitions were filed by private law firms and state attorneys general, seeking an aggregate of $21 million in fees and expenses for 70,000 attorney hours by 160 lawyers, and for 25,000 paralegal hours. Thus, in the aggregate, attorneys for the class sought over 40% of the $50,650,000 settlement fund.

It soon became apparent that the inability of the Executive Committee to work out an agreed upon maximum for fees and expenses was the result of a dispute among some members of that committee as to the value to the class of their respective services. In the fee application of Kohn, Savett, Marion & Graf, P.C., Harold Kohn took the position that his firm was chiefly responsible for the results achieved, and that the work claimed to have been performed by other applicants was exaggerated and of little value. On March 27, 1981, several counsel urged the court to appoint a Fee Review Committee, a procedure suggested in the Manual For Complex Litigation. Sec. 1.47(4) (1982). Kohn objected, and the trial court declined to appoint such a committee. 98 F.R.D. at 77. On April 24, 1981, 26 lawyers, representing the plaintiffs in 19 of the 24 settled lawsuits, filed objections to the Kohn firm's fee application. That objection was supported by affidavits of 17 plaintiffs' attorneys, to which were attached numerous exhibits. App. C 1-396. On May 15, 1981, the Kohn firm filed, on behalf of the two named plaintiffs it represented, a general objection, App. N 50-56, and sixteen specific objections to the fee applications filed by other plaintiffs' attorneys. In these objections Kohn contended that the applications of other attorneys had not been filed in good faith, and that most of the time which had been expended on the case had been solely for the purpose of generating fees. App. N 57-118.

On the same day that Kohn's objections were filed, class members listed in this appeal as appellees, none of which were named class representatives in the litigation, filed objections to the fees. App. L 18. This group of objectors, represented by Weil, Gotshal & Manges, moved for permission to review all fee applications and file supplemental written objections. App. L 28. That motion was granted. App. L 42. The court fixed June 9, 1981 for the commencement of hearings on the fee applications. At least in part because of disputes over discovery of documents in the files of some plaintiffs' attorneys, review of the fee applications by Weil, Gotshal & Manges took several months. The review was not completed by June 9, 1981, and the hearing on that date was continued so that a report on the Weil, Gotshal & Manges review could be completed.

At the July 23, 1981 hearing several applicants requested evidentiary hearings on disputed factual allegations, and the trial court observed:

Insofar as the request of the motion of Mr. Sloan and Mr. Specks for an evidentiary hearing, that of course will abide by the results of the report filed by Mr. Millstein [of Weil, Gotshal & Manges] and the response of the petitioning lawyers. If you want evidentiary hearings you will get it.

App. L 76.

Finally, on October 5, 1981, Weil, Gotshal & Manges filed on behalf of objecting class members a 522 page report, supported by a 164 page appendix of charts, and by three separately bound volumes of documentary exhibits.9 The trial court permitted the petitioning firms to file responses, and fixed times for hearings on the applications and objections.

The trial court scheduled a hearing on the fee applications and objections for each petitioning firm that requested one. These were scheduled between 4:00 P.M. and 6:00 P.M. on each court day. The hearings consumed 73.5 hours over 41 days.10 Eight law firms participated. During the course of those hearings petitioning attorneys objected that the Weil, Gotshal & Manges Report and supporting exhibits were not admissible evidence. Initially, at a hearing on December 17, 1981, the court ruled that they would not be accepted as evidence. App. L 200. Rather they would be treated only as argument. App. L 224. On the fifteenth day of hearings, however, when Specks & Goldberg, Ltd. had completed its case in support of its fee applications, Mr. Millstein of Weil, Gotshal & Manges offered in evidence, apparently against all fee applicants, the three volumes of exhibits referred to in his firm's report. App. F 395. Many of these exhibits were internal memoranda, or correspondence among some, but not all, of the petitioning law firms. The petitioning attorneys who were present objected to their admissibility on authenticity, hearsay and relevancy grounds. App. F 393-400. The court ruled that the objecting parties could take the volumes home over the weekend and then inform the court what items they objected to, at which time a ruling would be made. App. F 398. The court tentatively ruled that a firm's internal memorandum was not binding on any other firm "[u]nless I conclude there is a conspiracy." App. F 399.

The reference to a conspiracy is understandable in light of the nature of the fee objections made by Harold Kohn. In essence, Kohn charged that Granvil I. Specks and others entered into a conspiracy to overstaff the case, engaging attorneys and paralegals in "busy work" for the purpose of running up time in order to fatten the lodestar determination required by Lindy Bros. Bldrs., Inc. of Phila. v. American Radiator & Standard Sanitary Corp., 487 F.2d 161 (3d Cir.1973). If such a conspiracy was established by independent evidence, admissions in furtherance of it by each of its members could be used against the coconspirators. Fed.R.Evid. 801(d)(2)(E). On March 1, 1982, the sixteenth day of hearings, extended argument was held on admissibility of the tendered exhibits. App. M 317-45. The court observed "I am not going to rule on those particular items until we get everybody in here, everybody who wants to testify comes in." App. M 343. On May 20, 1982, after the evidentiary hearings were closed, the trial court, without making a finding as to the existence of a conspiracy, and without specifying the basis of admissibility, ordered that the tendered documents be admitted in evidence, with no limitation as to the parties against which or the purpose for which they might be used. App. I 370. These exhibits constitute 1,053 pages in App. N, O, and P.

The objectors offered no other evidence. The hearings consisted of testimony and exhibits offered by those petitioning law firms which requested evidentiary hearings, and of cross-examination of the witnesses whose testimony those firms presented. Both Weil, Gotshal & Manges, on behalf of the appellee objectors, and Kohn, Savett, Marion & Graf, P.C., on behalf of its client, waived the right to insist that each petitioning attorney appear for cross-examination. App. F 414-15. Thus evidentiary hearings were not held on all fee applications, but were held on all applications involved in these appeals.11 The last fee hearing was on May 12, 1982.

On March 3, 1983, the trial court filed a lengthy opinion dealing with the fee applications.12 Early in that opinion the judge indicates the general impression with which the proceedings left him, observing:

These fee petitions are grossly excessive on their face and, regrettably, lend substance to the widely-held and mostly unfavorable impressions of the plaintiffs' class action bar, sometimes referred to as the class action industry.

98 F.R.D. at 68. The opinion outlines the history of this litigation resulting in the $50,560,000 settlement. It then discusses formation of the organizational structure of plaintiffs' counsel, noting that after Specks & Goldberg, Ltd. filed the first action in Illinois on behalf of a national class of direct purchasers of fine paper, Granvil I. Specks circulated the complaint to several other lawyers with whom he had worked in prior class actions. These lawyers, the court observed, filed similar complaints on behalf of different named plaintiffs in other districts. Specks and the others then met and "began planning for distribution of patronage, that is, deciding to which firms the work assignments would be allocated." 98 F.R.D. at 71. The general tone of the court's discussion of the organizational meetings discloses disapproval of the steps leading to the filing of these separate lawsuits, and of the organization of plaintiffs' counsel for the purpose of allocating work.13 The opinion acknowledges that the committee structure agreed upon at the February 3, 1978 meeting was approved by court order, but observed:

In retrospect, I can recognize the force of Mr. Spiegel's argument [on behalf of California, opposing that structure] but at the time the court was persuaded by the array of distinguished counsel for the private plaintiffs and had no reason to anticipate that the case would not be prosecuted in the best interests of the class.

98 F.R.D. at 74. The opinion further notes that eight new firms came into the case after the first $30 million in settlement offers was in hand, and that the Executive Committee, over Kohn's objection, assigned work to some of these newcomers. 98 F.R.D. at 74. About the committee system the opinion notes:

It was inevitable that this type of structure would generate wasted hours on useless tasks, propagate duplication and mask outright padding.

98 F.R.D. at 75.

Turning to the objections, the court noted Kohn's assertion that the 97,000 hours claimed by plaintiffs' lawyers were excessive and that the same work could have been accomplished with the expenditure of 5,000 to 15,000 hours by one competent firm.14 98 F.R.D. at 78. The court also noted the Weil, Gotshal & Manges challenge to fee requests, and the recommendations in the Weil, Gotshal & Manges Report. 98 F.R.D. at 79. Regarding the objections filed to the Kohn firm's fee petition, the court observed:

Plaintiffs' counsel returned the volley by filing numerous objections to the Kohn firm's fee petition.14

Although these same objectors did not challenge any other fee petitions, the objections to the Kohn petition apply equally to many other petitions.15

98 F.R.D. at 79. About the responses to the Weil, Gotshal & Manges Report, the trial court observed:

Not unexpectedly, the petitioners (except Kohn) defended the organizational structure and the hours charged claiming that the division of labor had been the most effective way of tackling the massive task confronting them and that only their long hours, hard work and sheer persistence produced the fund which benefited the class.

Kohn, in his response, agreed with Weil, Gotshal that there had been tremendous waste, particularly in the deposition program.

98 F.R.D. at 80 (footnote omitted).

Following the foregoing preliminary observations, the court's opinion sets forth general guidelines to be applied in making the analyses required by Lindy Bros. Bldrs. Inc. of Phila. v. American Radiator & Standard Sanitary Corp., 487 F.2d 161 (3d Cir.1973) (Lindy I ), and 540 F.2d 102 (3d Cir.1976) (Lindy II ). The trial court proposed disallowing or reducing lodestar hours for the following tasks:

1. Pretrial memorandum time.

2. Read and review time.

3. Fee petition time.

4. Pretrial conference time.

5. Hours without adequate supporting data.

6. Industrial Analysis Committee time.

98 F.R.D. at 81-82. It also determined that in lieu of the specific rates historically charged by the petitioning attorneys there would be allowed:

1. Hourly rate for partner-level work performed by a partner--$100.00.

2. Hourly rate for associate-level work performed by an associate--$50.00.

3. Hourly rate for co-lead counsel (Specks and Kohn)--$150.00.

98 F.R.D. at 83.

The trial court also categorized the types of work appropriate for partners, associates, and paralegals. In instances in which experienced persons were found to have performed compensable tasks which in the court's view should have been assigned to persons with less experience or less professional training, the court included the time in the lodestar, but at uniform hourly rates for the less experienced or less professionally trained persons.

Addressing the Lindy I and II contingency factors, the court ruled that a 1.5 positive contingency multiplier would be allowed for lodestar time expended prior to the consummation of the $30 million settlements on January 16, 1979. 98 F.R.D. at 84. As to time expended after that date, the court ruled that "[o]nce the initial settlements were approved, there was no risk of non-payment for services or reimbursement of expenses." 98 F.R.D. at 84. And in justifying the selection of 1.5, rather than some higher rate, as the multiplier for time expended prior to January 16, 1979, the court observed:

I have not chosen a higher multiplier because prior to January 1979, the risk of an unsuccessful result in this case was spread among 25 law firms and 7 attorneys general. Moreover, it appears that this was not a case of high risk because there was a "large number of attorneys who were willing to prosecute this case on a purely contingent basis. This was quite obviously not a case where counsel were reluctant to invest considerable time because of the fear they would go unpaid." In re Penn Central Securities Litigation, 416 F. Supp. at 919 n. 30 [ (E.D.Pa.1976) ].

98 F.R.D. at 84 (footnote omitted).

Addressing the Lindy quality multiplier, the court ruled that none would be allowed, observing:

Under all the circumstances of the case, I am not persuaded that any counsel demonstrated such an unusually high degree of skill as would warrant compensation beyond that reflected in the hourly rates.

98 F.R.D. at 84.

One firm, Specks & Goldberg, Ltd., was singled out, however, for the application of a negative quality multiplier:

Specks and Goldberg ... because of its dominant leadership role, must be charged with primary responsibility for all the wasted hours, duplication and gross inefficiency which marked this case from its inception.

98 F.R.D. at 84. A 50% negative quality multiplier was applied to the Specks & Goldberg, Ltd. lodestar amount. 98 F.R.D. at 84.

The court also ruled that it would not allow reimbursement for expenses incurred by attorneys arising from activities for which their time would be disallowed, and that close scrutiny would be made of charges for meals consumed at hometown restaurants. 98 F.R.D. at 85.

Having laid down the foregoing general guidelines, the trial court proceeded to apply them to each separate fee application. The result, overall, was a series of awards aggregating approximately $4.3 million in fees, and expenses of approximately $1.1 million. Thus the aggregate fees and expenses awarded amounted to slightly more than 10% of the $50.6 million settlement fund.15 Every fee application was reduced to some extent. Listed in the margin are the requests and awards for those petitioners who have not appealed, or whose appeals have been withdrawn.16 The fees and expenses requested by appellants, and the court's rulings on them, are as follows:

Thus the appellants requests for fees were reduced by $6,340,761.50 and for expenses by $37,263.22, an overall disallowance rate of approximately 80%.

The listed fee requests include in each instance requested Lindy multipliers. To the eleven appellants, the court awarded a contingency multiplier for time expended prior to January 16, 1979 and included in the lodestar calculation, but not for time thereafter, as follows:

Freeman, Atkins & Coleman, Ltd.        9,892.50  App.  J 182
Philip C. Goldstick & Associates,
      Ltd.                              --  0--
Much Shelist Freed Denenberg
    Ament & Eiger, P.C.                4,972.50  App.  J 270
O'Brien & Hallisey                     5,405.25  App.  J 290
Walter E. Riordan                       --0--    App.  J 319
Rogers, Rude, Candlin, Faulkner &
    Sjostrom                            --0--    App.  J 346
Sachnoff, Weaver & Rubenstein, Ltd.    7,496.88
Saveri & Saveri                        7,462.50  App.  J 384
Sloan & Connelly, P.C.                 2,750.00  App.  J 400
Specks & Goldberg, Ltd.              115,900.00  App.  J 413
Lawrence Walner & Associates, Ltd.     6,001.25  App.  J 432
                                     ----------
Total                                159,880.88

The 50% negative quality multiplier applied to the court's calculation of a $789,330.50 lodestar amount for Specks & Goldberg, Ltd. reduced its lodestar amount to $394,665.75, and thus the pre-January-1979 contingency factor increased it to the amount awarded. $510.565.25. 98 F.R.D. at 215.

Applying its general guidelines the court also reduced the number of hours in each firm's lodestar calculation as follows:

(1) Pretrial Memorandum Time

The court, concluding that time spent in preparing the court-ordered pretrial memorandum was grossly excessive, reduced each firm's time expended on that task by one-half. 98 F.R.D. at 81. This resulted in elimination of 712.70 lodestar hours for the appellants' lodestar calculations as follows:

                                      Hours     Hours
Firm                                Requested  Allowed
Freeman, Atkins & Coleman, Ltd.
    Robert S. Atkins                    95.10    47.60  App.  J 175
Much Shelist Freed Denenberg
    Ament & Eiger P.C.
    Lawrence H. Eiger                   88.75    44.38  App.  J 251
    Michael B. Hyman                    88.45    44.23  App.  J 259
    Stephen A. Kanner                  105.95    59.98  App.  J 262
O'Brien & Hallisey
    Jeremiah F. Hallisey                62.25    31.13  App.  J 275
    Anne Treseder                       82.59    41.29  App.  J 278
Sachnoff, Weaver & Rubenstein,
  Ltd.
    Sarah R. Wolff                     333.25   166.63  App.  J 355
Sloan & Connelly, P.C.
    James Sloan                          8.40     4.20  App.  J 395
    Robert Davy                         14.80     7.40  App.  J 398
Specks & Goldberg, Ltd.
    Granvil I. Specks                  293.50   146.75  App.  J 404
    Perry Goldberg                      77.25    38.60  App.  J 408
Lawrence Walner & Associates, Ltd.
    Lawrence Walner                    108.10    54.05  App.  J 424
    William Sidlinger                   81.10    40.55  App.  J 427
                                    ---------  -------
Total                                 1439.49   726.79

The court also disallowed one half of an unspecified amount of all of the time of Saveri & Saveri included in time disallowance because records were too vague. See Part III.C. (5), infra. Since for the attorneys in question hourly rates ranging from $50 to $150 per hour were approved by the court, the 50% across the board elimination of time devoted to the pretrial memorandum produced a significant fee reduction.

(2) Read and Review Time

The court observed that "[h]undreds of hours were billed to the class by individual attorneys who spent time reading and reviewing a multitude of documents such as memoranda, correspondence, briefs, motions, and interrogatories, which crossed their desk during the course of this case, even though these attorneys were not closely involved with or responsible for the subject matter of these documents." 98 F.R.D. at 75-76. It therefore disallowed all hours billed for reading and reviewing documents except by the members of the fourteen-member Executive Committee. Id. at 81.17

(3) Fee Petition Time

In accordance with Lindy II, 540 F.2d at 111, the court eliminated from the lodestar all time spent in the preparation of fee petitions. Id.

(4) Pretrial Conference Time

The court observed that "counsel billed a total of 1446 hours for the 31 conferences. This large number of hours was primarily due to the attendance of attorneys who contributed nothing to the conferences but were merely there as spectators. The hours billed by these attorneys will not be compensated by the class." 98 F.R.D. at 81. The time thus disallowed totaled almost 1,400 hours. Among the appellants it was disallowed as follows:

Firm                                 Hours
Freeman, Atkins & Coleman, Ltd.
    Robert S. Atkins                  17     App.  J 175
Much Shelist Freed Denenberg
  Ament & Eiger, P.C.
    Michael J. Freed                  10     App.  J 254
O'Brien & Hallisey
    Jeremiah F. Hallisey              22     App.  J 274
Sachnoff, Weaver & Rubenstein, Ltd.
    Lowell Sachnoff                   41.75  App.  J 348
    Dean A. Dickie                     8     App.  J 351
Saveri & Saveri
    Guido Saveri                     144.5   App.  J 379
Lawrence Walner & Associates, Ltd.
    Lawrence Walner                   21.4   App.  J 424
                                     ------
Total                                264.65
 (5) Hours Billed Without Adequate Time Records.

The court observed that hundreds of hours were described in time records "by vague and meaningless terms or insufficiently documented." 98 F.R.D. at 81-82. Such hours the court refused to include in the lodestar. For the appellants the excluded time is as follows:

Firm                                    Hours
Freeman, Atkins & Coleman, Ltd.
    Robert S. Atkins                      32     App.  J 175
    Robert C. Goldberg                    14.3   App.  J 177
Philip C. Goldstick & Associates, Ltd.
    Philip C. Goldstick                    2.4   App.  J 185
    Alan L. Fulkerson                      5.5   App.  J 186
    Stew. M. Weltman                    11.1   App.  J 187
Much Shelist Freed Denenberg
  Ament & Eiger, P.C.
    Lawrence H. Eiger                     40.6   App.  J 251
    Robert A. Skirnick                     7.7   App.  J 256
    Anthony C. Valiulis                    5.6   App.  J 257
    Michael B. Hyman                      26     App.  J 259
    Stephen A. Kanner                     30.82  App.  J 261
    Paralegal time                        15.9   App.  J 265
O'Brien & Hallisey
    Jeremiah F. Hallisey                  68.25  App.  J 273
    Anne Treseder                         78.25  App.  J 277
    Gary Reiss                            81.45  App.  J 280
    Daniel B. Leraul                       6.38  App.  J 282
    George J. McNabb                        .3   App.  J 283
    Robert Murphy                          2.37  App.  J 285
    Don B. Kates                           2.7   App.  J 287
    Stanley S. Taylor                      3     App.  J 288
    Kenneth Goshorn                         .55  App.  J 288
Walter E. Riordan, P.A.
    Walter E. Riordan                    858.    App. J 314
Sachnoff.  Weaver & Rubenstein, Ltd.
    Lowell Sachnoff                        3.5   App.  J 349
    Bary Rosen                            28.25  App.  J 358
    Mitchell Goldsmith                      .5   App.  J 360
    Andrew Schatz                         24.25  App.  J 362
    Paralegals                           241.55  App.  J 360
    Outside Services                     223.75  App.  J 367
Saveri & Saveri
    Guido Saveri                         368.25  App.  J 379
                                         164.65  App.  J 379
Sloan & Connelly, P.C.
    James Sloan                           32.5   App.  J 395
    Michael Connelly                       4.1   App.  J 398
    Robert Davy                            3.6   App.  J 398
Specks & Goldberg, Ltd.
    Granvil I. Specks                    319.8   App.  J 403
    Perry Goldberg                        46.25  App.  J 407
    Howard L. Fink                         7     App.  J 410
Lawrence Walner & Associates, Ltd.
    Lawrence Walner                        7.6   App.  J 423
                                        -------
Total                                   2818.72
(6) Industrial Analysis Committee Time

The court excluded from the lodestar all of the approximately 1,800 hours of work performed for the Industrial Analysis Committee, one of the subcommittees formed by the Executive Committee, noting that "[t]he work of this committee was basically useless. It duplicated the efforts of others and resulted in no appreciable benefit to the class." 98 F.R.D. at 82. For the appellants the excluded time is as follows:

Freeman, Atkins & Coleman, Ltd.

    Robert C. Goldberg               58.1   App.  J 177
Saveri & Saveri
    Guido Saveri                     50.75  App.  J 379
Sloan & Connelly, P.C.
    James Sloan                       5.2   App.  J 395
Specks & Goldberg, Ltd.
    Granvil I. Specks                34     App.  J 404
Lawrence Walner & Associates, Ltd.
    Lawrence Walner                 216.5   App.  J 424
    William Sidlinger               100.5   App.  J 427
                                    ------
Total                               464.9

(1) Disallowance of Time of John A. Kithas and Charles Lamont

Saveri & Saveri, P.C. included in its petition 1,717.25 hours of work performed by John A. Kithas and Charles Lamont. The Weil, Gotshal & Manges Report objected to the inclusion of their time on the ground that Kithas & Lamont was, at the time the work was performed, a separate entity. Pretrial Order No. 143 provides:

(12) If the time or rate of any lawyer who is not a partner or employee of the petitioning firm is included in any petition, the petition shall contain a detailed statement prepared by the petitioning firm and such other lawyer, setting forth in detail the arrangements between them for compensation and defraying of cost disbursements during the litigation and the division or allocation of any fee or reimbursements awarded, together with a statement of the reasons why it was necessary for the petitioner to enter into such arrangements.

The court excluded from the lodestar for Saveri & Saveri, P.C. all time of Kithas and Lamont because it had not filed the statement required by Pretrial Order No. 143. 98 F.R.D. at 202-03.

(2) Disallowance of Time of Rogers, Rude, Candlin, Faulkner & Sjostrom

Rogers, Rude, Candlin, Faulkner & Sjostrom of Minneapolis filed a petition seeking compensation for 513.6 hours expended as associate counsel with the firm of Walter E. Riordan, P.A. The firm's petition was denied by the court because it did not comply with paragraph (14) of Pretrial Order No. 143, in failing to disclose the nature of its arrangement with Walter Riordan, P.A. The court also found that James W. Rude had agreed to work on the case for Walter Riordan for $15 an hour, and was in fact paid at that rate for 76.4 hours of work. 98 F.R.D. at 188-89.

(3) Sloan and Associates, P.C., Pricing Study Time

During the course of trial preparation a dispute arose among members of the Executive Committee over the best way to establish damages. Some 4,200 hours of time were devoted to pricing/damage studies. Granvil I. Specks turned over to Sloan & Connelly pricing materials obtained on discovery for a pricing study, although Harold Kohn opposed giving the assignment to that firm. Later the firm of Berger & Montague and the Kohn firm prepared new studies. The Weil, Gotshal & Manges Report urged that the trial court cut all time spent on pricing/damage studies by one-half, or, alternatively determine which attorneys' time benefited the class and which did not. Sloan & Connelly, P.C. devoted 747.9 attorney hours and 1,070.75 paralegal hours to the pricing study. On July 23, 1980, two months before the scheduled trial date, the Executive Committee was informed by one of its experts that the study had not yet produced support for a theory of parallel pricing. Further efforts were made to refine the study. The court ruled:

It is my considered judgment that this pricing project involved too many hours of too many attorneys. The end-product does not justify so substantial a charge against the class fund as Sloan & Connelly requests. Accordingly, to reflect my determination that an excessive number of hours are claimed for this work as compared to the results achieved, I will only permit half of the total 1818.65 hours to be compensated.

98 F.R.D. at 207. Thus the Sloan & Connelly, P.C. lodestar for compensation was reduced by 909.33 hours.

(4) Disallowance of Other Walter E. Riordan, P.A. Time

Walter E. Riordan discovered a witness, James Nelson, who had been President of General Paper Corporation, a now bankrupt paper firm, during the period when the conspiracy allegedly operated. According to Riordan, Granvil I. Specks of the Executive Committee anticipated that Nelson would be subject to rigorous cross-examination at deposition and for that reason directed Riordan to represent Nelson during those proceedings. The court excluded from the lodestar for Walter E. Riordan, P.A. 678 hours spent in these activities. The court also disallowed one-half of an additional 1,347.25 hours because of the nature of the tasks performed and the incompleteness of the supporting records. 98 F.R.D. at 177-78.

The appellants dispute and take considerable offense at the court's preliminary observations. Each appellant contests the legality of the court's general guidelines. Each, as separately affected, contests the factual basis for the court's application of those guidelines to its petition. The appellants impacted by the court's specific rulings dispute their factual and legal predicates. Several raise procedural objections to the manner in which the hearings were conducted, and evidentiary objections to the materials upon which the court relied. Thus each aspect of the court's decision referred to in Part III is challenged on some basis and must be addressed by this court.

As noted in Part I above, the underlying lawsuit was a class action brought pursuant to sections 4 and 16 of the Clayton Act, 15 U.S.C. Secs. 15, 26, to enforce section 1 of the Sherman Act, 15 U.S.C. Sec. 1 (1982), which was settled by the creation of a fund in exchange for the release of the defendants from liability both for damages and for statutorily authorized fees. In Lindy Bros. Bldrs., Inc. of Phila. v. American Radiator & Standard Sanitary Corp., 487 F.2d 161, 165 (3d Cir.1973) (Lindy I ), this court considered such a settlement for the first time. The court might have taken the position that in instances in which statutory fees are authorized the liability for damages must be settled separately, and only when the settlement fund for the class members was approved would the court either entertain litigation or consider a separate settlement of the defendants' liability for statutory fees.18 That course would have eliminated the possibility of a conflict of interest between attorneys for the class and class members, and would have established a framework in which the amount of fees to be awarded would be determined in a genuinely adversarial context. Had we taken that approach, the need for judicial scrutiny of fee requests would have been substantially reduced. In cases where settlements of fee requests are made with the defendants after prior approval of damage claim settlements, the court can, in most instances, assume that the defendants closely scrutinized the fee requests, and agreed to pay no more than was reasonable. Additionally, when a damage settlement is followed by litigation over statutory fee requests, the amount awarded would be determined by the normal workings of the adversary system.

Instead, mindful of the desirability of facilitating the commencement and disposition of Fed.R.Civ.P. 23(b)(3) class actions, we held that settlements releasing defendants from both damage and statutory fee liability were proper, and would result in a fund in court from which fees could be awarded under the equitable fund doctrine. Lindy I, 487 F.2d at 164-65. The consequence of that holding was to create an inevitable conflict of interest between the attorneys for Rule 23(b)(3) class members and the class members for whom they appeared. Whether the holding has in fact had the hoped for effect of facilitating settlements of class actions is an open question. Be that as it may, we recognized in Lindy I that by authorizing the conversion of statutory fee cases into fund in court cases we were creating the potential for such a conflict of interest. Thus we held that fee requests from the resulting equitable fund in court must be subjected to heightened judicial scrutiny.19 Those standards were refined in Lindy Bros. Builders, Inc. v. Am. Radiator, 540 F.2d 102 (3d Cir.1976) (Lindy II ).

This court recognizes that the first step in allocating a fund in court among the attorneys responsible for its creation and the class members for whose benefit it was created is a determination of how many hours were spent, by which attorneys, and in what manner. Next the court must determine the value of each attorney's services to the class:

The value of an attorney's time generally is reflected in his normal billing rate. A logical beginning in valuing an attorney's services is to fix a reasonable hourly rate for his time--taking into account the attorney's legal reputation and status (partner, associate).... Similarly, the court may find that the reasonable rate of compensation differs for different activities.

Lindy I, 487 F.2d at 167. Thus the hourly rate must be individually determined, separately for each attorney, and for separate categories of activities engaged in by each attorney.

Once this lodestar amount is determined the court must adjust it to reflect the contingent nature of the attorneys' undertaking. Lindy I, 487 F.2d at 168. This contingency adjustment involves two separate aspects. The first aspect to be determined is the likelihood of success in obtaining a judgment or settlement in the underlying lawsuit, to be measured at the point when the attorney's time was committed to the case. Id.; Lindy II, 540 F.2d at 113. The second aspect is the time value of compensation long delayed, when compared with normal billing and collection practice of law firms for fees and expenses. We explicitly recognized in Lindy II that the contingency factor must take into account delay in receipt of payment. 540 F.2d at 117. See also Copeland v. Marshall, 641 F.2d 880, 893 (D.C.Cir.1980). A second delay factor, unrelated to the contingency of the outcome, is discussed in Part V.C. (1), infra.

Finally we require that the court may adjust the amount determined in the lodestar calculation for the quality of an attorney's work. Lindy I, 487 F.2d at 168. Quality 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 and penalizing the use of methods the predominant purpose of which was to delay or obstruct the proceedings." Lindy II, 540 F.2d at 118.

Our equitable fund case law also makes clear that the attorneys' claim for fees from a fund in court is a cause of action, belonging to the attorneys, for the reasonable value of their services, for which a hearing is required. As we observed in Lindy I:

Much of the evidence on which the judge will base his award or denial of fees may be disputed; the evidence presented by an attorney petitioning for fees may be incomplete. The denial of fees obviously harms the petitioning attorney. Just as obviously, award of attorneys' fees harms the unrepresented claimant by reducing his net recovery. These opposing interests should be afforded a hearing to provide an evidentiary basis for resolution of disputed factual matters and to allow the parties to supplement possibly incomplete statements of opposing parties.

487 F.2d at 169. See also Merola v. Atlantic Richfield Co., 515 F.2d 165, 167 n. 5 (3d Cir.1975). Plainly, therefore, the hearing on a fee application in an equitable fund case requires compliance with those procedural rules which assure fair notice and an adequate opportunity to be heard. Equally plainly, the requirement of an evidentiary hearing demands the application in that hearing, of the Federal Rules of Evidence. Fed.R.Evid. 1101(b) (rules apply generally to civil actions).

Since we are reviewing a civil proceeding tried to the court, our review of the court's determination of fact questions is by the clearly erroneous standard of Fed.R.Civ.P. 52(a). On legal issues, including legal standards for the admissibility of evidence, our review is plenary. In re Japanese Electronic Products Antitrust Litigation, 723 F.2d 238, 258 (3d Cir.1983). Most of the issues presented by this appeal, however, fall into categories which, as noted specifically hereafter, we have held to be committed to the discretion of the trial court. In reviewing those issues we must first decide whether in determining the facts upon which the court predicated the exercise of discretion it relied upon properly admissible evidentiary materials and, if so, whether the findings are clearly erroneous. In reviewing the court's exercise of discretion in light of supportable factual premises, we must consider whether the court has taken into account properly relevant considerations. A trial court which gives weight to improper considerations, or ignores those properly relevant, abuses its discretion or commits legal error. Id. at 263, 265, 277, 302.

Finally, we do not approach the review process with any predisposition against what we know to be the fairly common practice in large class action litigation of encouraging the participation, on the plaintiffs' side of the case, of a number of law firms which may divide the work, and thus the risk. Such spreading of work, and hence of risk, seems to us consistent with the decision by Congress, in the antitrust laws and elsewhere, to encourage enforcement of federal statutory policies by private litigants.

With these general principles in mind, we turn to the specifics of the appellants' contentions.

B. Objections to the Manner of Conducting the Hearing

Some appellants complain about the manner in which the fee hearings were conducted, making particular reference to the failure of the objectors to offer testimony; to the court's excessive reliance upon the Weil, Gotshal & Manges Report; to the fragmentation of the hearings into two hour segments at the end of regular court business days; to the delay in ruling upon the admissibility of three volumes of documents until after the close of the hearings; and to the admission of those documents over objections on authenticity, hearsay and relevancy grounds.

The failure of the objectors to offer testimony is not dispositive of any issue in this appeal. The fee applicants had the burden of persuasion on the number of hours to be included in the lodestar, the rates to be applied, and the propriety of affirmative multipliers.20 The court was entitled to rule on the basis of their presentations, the cross-examination of their witnesses, and its own knowledge of the proceedings. See Lindy I, 487 F.2d at 169.

As noted in Part II above, the court indicated during the hearing that it would treat the Weil, Gotshal & Manges Report in the nature of a brief. App. L 224. Appellants point to striking similarities between that Report and the contents of the court's fee opinion. We give no weight to this criticism. There is no indication in the court's fee opinion that evidentiary weight was given to the Report. It is hardly surprising that there may be similarities between the text of a judicial opinion and that of a brief by counsel that the court found to be persuasive. Legal writing, including judicial opinion writing, is rarely entirely original.

The complaint about fragmentation of the hearing is more substantial. Requiring counsel from distant cities, applicants and objectors alike, to conduct hearings between four and six p.m. over a rather extended period was certainly an imposition on them. Moreover the fragmentation of the hearings into short segments produced a record which is, from the point of view of appellate review, less than ideally tidy. Nevertheless, the time for scheduling hearings is a matter which must of necessity remain in most instances a matter of virtually unreviewable district court discretion, for only the district court has the information required to balance the demands of any litigant against the competing time demands of other litigants. Moreover, no appellant has pointed to any way in which the fragmented manner in which the hearings were conducted affected the substantive results of which it complains. Thus we give no weight to criticism of the fragmented hearings.

The timing of the court's ruling on the admissibility of three volumes of documents, offered in evidence over objection on the sixteenth day of the hearings, and admitted in evidence without explanation of the basis for the ruling on May 20, 1982, is a far more troubling matter. Ordinarily in any hearing, and fee applications are no exception, litigants are entitled to a ruling on objections to the admission of evidence, if not simultaneously, at least sufficiently in advance of the close of the hearing so they know what record evidence must be met. Thus we have carefully considered whether, putting aside questions of admissibility, any prejudice resulted in this case from the delayed ruling. We conclude that there was none. No appellant has suggested that any additional evidence would have been offered had it been known that the three volumes of documents would be admitted. Since the court made it known on March 1, 1982 that admission of the documents was under consideration, the appellants did have the opportunity to make an offer of proof of any such additional evidence in anticipation of a ruling favoring admission. Following the entry of the May 20, 1982 order admitting the three volumes of documents, moreover, no appellant made any motion to supplement the record. Thus the delay in ruling on admission of the documents until after the close of the hearings, while hardly the preferred practice, cannot be found to have prejudiced the appellants.

That leaves for consideration the correctness of the court's disposition of the objection to admissibility.

As to authenticity, no more is required than evidence establishing prima facie that the matter in question (here documents from various attorneys' files) is what it purports to be. Fed.R.Evid. 901(a). Assuming that the court entirely disregarded the authenticity objections, such an error would on this record be completely harmless. No appellant has suggested that the documents are other than what the objectors claim them to be. There is ample evidence in the record suggesting, prima facie, the authenticity of each document. Once such evidence is in the record, the document is authenticated for all purposes against all parties. In re Japanese Electronic Products Antitrust Litigation, 723 F.2d 238, 285 (3d Cir.1983). Thus we attach no significance to the court's failure to deal specifically with the Rule 901(a) objection.

The hearsay objection cannot be disposed of so easily because the court failed to make the findings required by Fed.R.Evid. 801(d)(2)(E), namely that a conspiracy existed; that the parties against whom the documents were offered were coconspirators; and that the documents contained statements made during the course and in furtherance of the conspiracy. The court simply admitted the entire three volumes of documents against all appellants without limitation as to purpose.

The appellee fee objectors defend this ruling on several grounds. Some documents, they urge, are non-hearsay because they fall within Rule 801(d)(1) (prior statement by witnesses) or Rule 801(d)(2) (admission by party-opponent). That is undoubtedly the case for some of the documents, but even then such admissions may only be used against the party who made them. Some are duplicative of documents offered by the fee petitioners in support of their own applications, and are thus at most duplicative of other evidence properly before the court. Some are pleadings or memoranda filed with the court in the course of the proceedings, which may be presumed to have been given the consideration appropriate for such documents. Some are summaries of the contents of other documents properly before the court, such as the fee petitions. Finally, as to other documents, such as correspondence from the files of several counsel, the objectors rely on the catchall hearsay exception in Rule 803(24) as statements having circumstantial guarantees of trustworthiness. We place no reliance on Rule 803(24), however, because since the authors of the correspondence could have been called as witnesses, the statements cannot be found to be "more probative on the point for which [they are] offered than any other evidence which the proponent can procure through reasonable efforts." Fed.R.Evid. 803(24)(B). For the same reason we place no reliance on Rule 804.

Having examined the documents, we conclude that many of them, as introduced against many of the appellants, are inadmissible hearsay. That conclusion only commences our inquiry, however, for in this non-jury proceeding we have the benefit of a lengthy opinion by the trial court. A party may only raise a serious claim of prejudice if some evidentiary use of an inadmissible document was made against that party. The vast majority of the contested rulings on this appeal involve the application of rules of law to undisputed facts. In those instances where rulings as to specific appellants involve factual issues, we will determine whether any inadmissible hearsay significantly affected the court's determination of those factual issues. We will make the same determination with respect to the relevancy objections.

Two final objections to the manner in which the hearing was conducted must be addressed, both of which involve discovery.

Sloan & Connelly, P.C. complains that the trial court quashed subpoenas served upon objectors' counsel for the purpose of cross-examination to establish the objectors' motivation in pressing their objections to the requested fees. The court's ruling was, in essence, a relevancy ruling under Fed.R.Evid. 403. We find no abuse of discretion in the court's refusal to permit an already cluttered record to be further confused by an inquiry so completely collateral to the central issue of reasonableness of the fee requests. The position of the objectors, who have been referred to pejoratively by some appellants as "Fortune 500 companies" and potential class action defendants, is hardly unique. They are in no different position than was Humble Oil and Refining Company, one of the objecting class members in the seminal Lindy fee proceeding. We fail to see how motives of the objectors, other than the obvious financial one of maximizing their recovery, would make any fact of consequence to the determination of reasonable fees more or less probable. Fed.R.Evid. 401. Certainly we can find no abuse of discretion.

Lawrence Walner and Associates, Ltd. moved to have the court require the disclosure, in camera, of fees paid by the settling defendants in the underlying litigation, and by the corporate objectors in this and other antitrust litigation. The request was made for the purpose of enlightening the court as to reasonable hours and hourly rates for comparable lawyers in complex litigation. The information sought certainly was relevant, and arguably even helpful. See, e.g., Blowers v. Lawyers Co-op. Pub. Co., Inc., 526 F. Supp. 1324 (W.D.N.Y.1981); Naismith v. Professional Golfers Ass'n, 85 F.R.D. 552 (N.D.Ga.1979); Stastny v. Southern Bell Telephone & Telegraph Co., 77 F.R.D. 662 (W.D.N.C.1978). But see Samuel v. University of Pittsburgh, 80 F.R.D. 293 (W.D.Pa.1978). Discovery rulings are reviewed, however, for abuse of discretion. Marroquin Manriquez v. I.N.S., 699 F.2d 129, 134 (3d Cir.1983). Considering all the evidence offered on hours and rates, and the likelihood that such discovery would generate inquiries into collateral matters, such as privilege, we cannot hold that the court abused its discretion in denying the motion.

Summarizing, we conclude that with the possible exception of admission of some hearsay evidence, to which particular reference will be made hereafter, none of the objections to the manner in which the fee hearings were conducted affect the disposition of any issue presented in this appeal.

C. Objections to the Trial Court's Guidelines

We have outlined in Part III. B., supra, the general guidelines adopted by the trial court. We now turn to the appellants' objections to those guidelines.

(1) Contingency Factors

All appellants urge that the court committed legal error in several respects in its treatment of the Lindy contingency multipliers. As noted in Part V.A., supra, Lindy I and II require an adjustment to the lodestar amount for (1) the risk of lack of success inherent in the lawsuit, and (2) the time value of compensation long delayed. The appellants contend that the court erred (1) in limiting application of the risk of lack of success multiplier to time committed prior to January 16, 1979, and (2) in disregarding entirely the delay in payment factor.

We hold that the court committed legal error when it ruled that "[o]nce the initial settlements were approved, there was no risk of non-payment for services or reimbursement of expenses." 98 F.R.D. at 84. The major premise of the court's reasoning is that time expended after the initial $30 million settlement fund came into existence could, if no other amount was recovered, be paid for out of that fund. That premise is false as a matter of law, for under the equitable fund doctrine payment for those services can be charged to the fund only for services which benefit the class members interested in that fund. Sprague v. Ticonic National Bank, 307 U.S. 161, 59 S. Ct. 777, 83 L. Ed. 1184 (1939); Trustees v. Greenough, 105 U.S. 527, 15 U.S. 527, 26 L. Ed. 1157 (1881). Thus attorneys expending time, after January 16, 1979, in processing lawsuits against those paper manufacturers which had not settled did so at the risk of lack of success, and would have received nothing from the fund for such time had there been no further recovery.21

The objectors defend the ruling, not on the basis of the trial court's reasoning, but on the theory that once six defendants settled, it was inevitable that the remaining defendants would do so, and thus there was in fact no risk of lack of success. The trial court, however, made no such finding. In light of information in the record about the vigor with which the defendants remaining in the case after January 16, 1979 continued to contest it, and about concerns among plaintiffs' counsel about a viable damage theory, such a finding, had it been made, would be hard to support. Certainly it is not a finding which an appellate court can make as a basis for affirming on a theory which the trial court did not address.

We also hold that the court committed legal error in disregarding the delay in payment of compensation in calculating the contingency multiplier. In a fund in court case delay in payment occurs during two separate periods. First, the attorneys remain unpaid, and remain at risk of going without payment at all, during litigation until the time that a fund is created. Next, they are deprived of the use of their share of the settlement fund until the court finally determines what that share is. During the first period the delay and risk of lack of success elements are so interrelated that it is perhaps not feasible to measure them separately, although several appellants have suggested approaches for so doing. During the second period, on the other hand, the risk of lack of success is eliminated, and the delay in payment factor is precisely measurable. The attorneys are entitled to a share of the fund measured at the time it comes into existence, and the class members are entitled to the balance. When, as here, the fund is earning interest, the time value of the money is, to each party entitled to a share, equal to the interest earned on that share pending its distribution.

The court in its discussion of general guidelines made no mention of delay in payment as an element of the contingency multiplier. Nowhere else in the opinion is delay in payment prior to the creation of the fund discussed. At one point, in evaluating the application of a firm not before us as an appellant, the court observed:

Fine, Kaplan & Black argue for the higher hourly rate as reflected by the fees calculated on a current basis in order to compensate for the loss due to inflation and loss of interest over the time period during which they have had to wait for payment. Any delay in payment can be considered in determining whether to add a positive multiplier.... Since a substantial part of the delay in this case has been occasioned by the protracted litigation over fees among plaintiffs' counsel, I will not enhance any award for the delay factor.

98 F.R.D. at 124 n. 66 (citations omitted). The stated reason is legally insufficient as a justification for refusing to consider delay in payment as an enhancement factor. First, litigation over fees had nothing whatever to do with delay encountered prior to the creation of the settlement fund. Second, the litigation over fees pitted against each other competing claimants for a share in that fund. No rule of law or equity suggests that the entire time value of money withheld during the resolution of those claims should be borne by some but not other claimants.

Moreover, even if the court's reasoning were legally supportable, the statement that "a substantial part of the delay in this case has been occasioned by the protracted litigation over fees among plaintiffs' counsel" is simply not accurate. Nineteen months elapsed between the filing of the first complaint and the first round of settlements. Another nineteen months elapsed before approval of the final settlement on September 22, 1980. The court by order fixed a date six months thereafter for the filing of fee petitions. On May 15, 1981, the court granted Weil, Gotshal & Manges, representing objectors, additional time to review the fee applications and supplement their objections. On June 9, 1981, the court granted Weil, Gotshal & Manges additional time. The Weil, Gotshal & Manges Report was complete on October 5, 1981, almost seven months after the date fixed for filing fee petitions. Hearings extended into early May of 1982, another six months, but it was the court, not plaintiffs' counsel, which scheduled those hearings in two hour segments. After the hearings were closed, the court took another ten months to prepare an opinion. Thus while thirty-eight months elapsed from the time the first class action was filed until the September 22, 1980 settlement, and the fee opinion was not filed until 29 months thereafter, it is grossly unfair to suggest that plaintiffs' counsel, rather than the objectors and the trial court, are solely responsible for the delay. The objectors were entitled to object, and to be afforded the time necessary for the preparation of their objections. The court was entitled to fix a schedule for hearings consistent with the interests of other litigants, and to take such time as it deemed necessary to prepare an opinion reflecting numerous rulings on complex issues. But these entitlements are no reason for giving to the class members interest earned after the settlement fund was in hand not only on their share of the fund but on the attorneys' share as well. Neither the fee applicants nor the objecting class members should be given an incentive to protract litigation over the allocation of a fund in court. Evenhandedness demands that each party having an interest in that fund receive its share of fund earnings, pending distribution, in proportion to its interest.

Our conclusion that the court committed several legal errors in its consideration of the Lindy I and II contingency multipliers requires a remand for further proceedings in which an appropriate contingency multiplier shall be determined for all lodestar time up to September 22, 1980, and for calculation of the share of fund earnings which should be credited to the attorneys.

(2) The Quality Multiplier

Several appellants contend that the court abused its discretion in disallowing a quality multiplier for any attorney. They do not contend that the court misapprehended the factors relevant to the superior quality determination. Those factors to be considered are "[t]he complexity and novelty of the issues presented, the quality of the work that the judge has been able to observe, and the amount of the recovery obtained." Lindy I, 487 F.2d at 168. They contend, rather, that considering the recovery obtained, $50 million, in a price-fixing case in which no government agency filed any charges, and in an industry which successfully defended a somewhat similar action, a quality multiplier was required.

We disagree. It is not possible on the record before us to evaluate the significance of the recovery in the abstract. In the Lindy case, for example, the court was able to point out that the settlement was two to three times the actual price overcharge. Lindy II, 540 F.2d at 115. We have not been referred to any such concrete comparisons in this case. The petitioning fee applicants have not carried their burden of establishing that $50 million was an extraordinary settlement.

A quality multiplier might well be appropriate, even for a modest result, if the court were to conclude that it had been achieved with unusual efficiency, and with little expenditure of attorney time and expense. Obviously, however, the trial court is in the best position to determine whether the case was handled with unusual efficiency. The court determined otherwise, and the very magnitude of the requests for lodestar hours suggests that this determination cannot be deemed an abuse of discretion.

We hold, therefore, that the trial court did not abuse its discretion in denying an affirmative quality multiplier to any applicant.(3) Use of Uniform Hourly Rates

Most of the petitioning law firms submitted affidavits setting forth historical normal billing rates for each attorney whose time was included in the proposed lodestar amount. For the most part those affidavits are uncontested. In some instances, however, the Weil, Gotshal & Manges Report pointed out discrepancies between hourly rates sought for some lawyers and for other lawyers in the same or other firms having equivalent experience. The court made no use of historical billing rates. Instead, as noted in Part II. B., supra, it determined an hourly rate of $150 for Harold Kohn and Granvil I. Specks, $100 for all other partner level work, and $50 for associate level work, regardless of the relative levels of experience, skill, or prestige in the marketplace, of the attorneys in question. In justifying that approach the court observed that "[t]hese rates are in conformity with the regular hourly rates billed to noncontingent clients by private law firms in major metropolitan areas." 98 F.R.D. at 83. No record evidence is referred to in support of this observation, and the subject is not one on which judicial notice is appropriate.22 All appellants contend that in applying a hypothetical national billing rate for lead counsel work, partner level work, and associate level work, the court committed legal error, or at least an abuse of discretion.

In the seminal Lindy I opinion this court held:

The value of an attorney's time generally is reflected in his normal billing rate. A logical beginning in valuing an attorney's services is to fix a reasonable rate for his time--taking account of the attorney's legal reputation and status (partner, associate). Where several attorneys file a joint petition for fees, the court may find it necessary to use several different rates for different attorneys.

487 F.2d at 167. Ever since Lindy I this court, both in fund in court cases23 and in statutory fee cases,24 we have reiterated that individual determinations of reasonable billing rates are required for the lodestar determination. Our premise has been that the reasonable value of an attorney's time is the price that time normally commands in the marketplace for legal services in which those services are offered. Recently we have been given clear direction by the Supreme Court that our market standards rule is required as a matter of law. In his separate opinion in Hensley v. Eckerhart, 461 U.S. 424, 103 S. Ct. 1933, 76 L. Ed. 2d 40 (1983) (a statutory fee case), Justice Brennan wrote:

as nearly as possible, market standards should prevail, for that is the best way of ensuring that competent counsel will be available to all persons with bona fide civil rights claims. This means that judges awarding fees must make certain that attorneys are paid the full value that their efforts would receive on the open market in non-civil-rights cases, both by awarding them market-rate fees, and by awarding fees only for time reasonably expended ....

Id. at 447, 103 S.Ct. at 1947 (Brennan, J., concurring in part and dissenting in part) (citations omitted) (emphasis in original). Still more recently, in Blum v. Stenson, U.S. ----, 104 S. Ct. 1541, 79 L. Ed. 2d 891 (1984), Justice Powell for a unanimous Court, announced as the standard for lodestar determinations, even for cases litigated by not for profit law offices, the prevailing market rate in the relevant community. See also Laffey v. Northwest Airlines, Inc., 746 F.2d 4 (D.C.Cir.1984) (district court must base hourly rates on the firms' established billing rates, not on "true value" of attorneys' services).

Given this court's long prevailing standard, and the Supreme Court's endorsement of it in Hensley v. Eckerhart and Blum v. Stenson, we agree with the appellants that the approach taken by the trial court in this case, of applying hypothetical national rates to all attorneys, regardless of the market rate they would command in the community in which they practice, was legal error. This holding requires a remand for further proceedings in which appropriate individual reasonable rates shall be determined for the lodestar time of each attorney.

(4) Reduction of Partner Rates for Certain Activities

All appellants object that the court not only disregarded historical market billing rates, but also disregarded their professional judgment as to the level of experience required for certain tasks, when it reduced time charges for certain partner work to associate level charges and certain associate work to paralegal level charges. The appellants cannot successfully urge that as a matter of law their professional judgment controls, for as early as Lindy I we held that "the court may find that the reasonable rate of compensation differs for different activities." 487 F.2d at 167. Thus the court's ruling in this respect may be set aside only if predicated on clearly erroneous findings of fact or if it is so arbitrary and capricious as to amount to an abuse of discretion. The court observed:

In this litigation, it was not unusual for senior partners to devote a great deal of time to clerical and administrative and investigative tasks such as document review. Partners will not be compensated at partner-level rates for tasks which are customarily performed by junior associates or paralegals.

98 F.R.D. at 83. We have carefully examined the court's specific application of that standard to individual fee applications. With one exception (Walter C. Riordan, P.A.), the significant reductions and the court's reasons therefore are as follows:

Firm                                                         Hours
Freeman, Atkins & Coleman, Ltd.
  Robert S. Atkins partner                                     12.17     App.  J
                                                                           176
time for reviewing various discovery documents reduced to
associate level.
  Robert C. Goldberg partner                                  112.6      App.  J
                                                                           178
time for reviewing Champion discovery documents and
Federal Trade Commission documents, and attending discovery
of client Campbell Paper Co. reduced to associate
level.
Much Shelist Freed Denenberg
  Ament & Eiger, P.C.
  Lawrence H. Eiger partner                                    86.8      App.  J
                                                                           253
time for reviewing and collating documents, legal
research, summarizing depositions and organizing price
work sheets, reduced to associate level.
  Michael J. Freed partner                                     85.9      App.  J
                                                                           255
time for conferences or on phone with co-counsel reading
and reviewing various pleadings and documents filed
in the case, drafting a complaint for his client and
attending depositions taken by others, reduced to associate
  level.
  Robert A. Skirnick partner                                   74.4      App.  J
                                                                           256
time for document review, research, and drafting motions,
reduced to associate level.
  Anthony C. Valiulis partner                                   3.3      App.  J
                                                                           257
time for legal research and drafting motions reduced to
associate level.
O'Brien & Hallisey
  Jeremiah F. Hallisey partner                                219.55     App.  J
                                                                           276
time for general research, conference time, correspondence,
and review which could readily have been
performed by an associate, reduced to associate level.
Sachnoff, Weaver & Rubenstein, Ltd.
  Dean A. Dickie partner                                       53.75     App.  J
                                                                           352
time for scheduling, preparing indices, and inspecting
microfilm and documents
Sloan & Connelly, P.C.
  James Sloan partner                                           9.8      App.  J
                                                                           398
time for drafting document requests, notice and subpoenas,
reduced to associate level.  Scheduling, preparing
indices, and inspecting microfilm and documents reduced
to associate level.
  Michael Connelly partner                                     56.3      App.  J
                                                                           397
time for drafting stipulations and notices of depositions,
attending document production sessions, and initial
review of trade association documents reduced to associate
  level.
Specks & Goldberg, Ltd.
  Granvil I. Specks partner                                  1646.72     App.  J
                                                                           404
time reduced to associate level.  The court found:
      Included in this total is time spent on legal research and
      the drafting of interrogatories, notices of depositions, briefs,
      lists of defendants' subsidiaries and affiliates and discovery
      outlines.  Other tasks included are the payment of billB s,
  acquiring
      information regarding rates for certificates of deposit,
      making travel and conference room arrangements, persuading
      lawyers to pay their assessments, and preparing
      notices concerning meetings and drafting minutes of meetings.
      These 1646.72 hours also encompass time Mr. Specks
      spent reviewing Moodys, Standard & Poor's and the defendants'
      annual reports, preparing service lists, obtaining
      Lockwood's Directory and researching defendants' market
      shares.  Also included are nearly two hundred hours Specks
      spent on class certification discovery.
98 F.R.D. 211-12.
  Howard L. Fink partner                                     1923.9      App.  J
                                                                           411
time for reviewing documents, doing legal research,
scheduling depositions, arranging subpoenas, assisting
Mr. Specks at depositions, drafting notices of depositions,
interviewing potential third-party witnesses, collecting
materials relating to damages for use at trial,
and digesting depositions.
Lawrence Walner & Associates Ltd.
  Lawrence Walner partner                                     817.97     App.  J
                                                                           426
time for reviewing Wausau and Mead time documents,
preparing deposition notebooks, drafting documents and
deposition summaries, drafting replies to Rule 11
interrogatories, research, drafting motion for protective
order, reviewing defendants' first wave of interrogatories,
and assembling pricing survey for the Sloan & Connelly
firm.
Total                                                        5103.16

Thus for the listed appellants the court concluded that over 5,000 hours of partner time was spent on tasks which should have been assigned to associates.

The appellants do not seriously question the hours or the court's description of the nature of the activities. They vigorously dispute, however, the court's ruling that all those activities should have been assigned to associate level personnel. They urge that often it is far more efficient for an experienced antitrust lawyer to examine documents, for example, than to spend the time educating an associate as to what, in those documents, may have antitrust significance. They point out that fifteen minutes spent in the law library by an experienced antitrust lawyer may save hours of less productive research time by associates plowing unproductive fields. They contend, moreover, that the court, influenced excessively by the Weil, Gotshal & Manges Report, imposed on the aggregation of small firms conducting this litigation the hierarchical organization structure typical of large firms employing hundreds of lawyers.

Because of the magnitude of the reductions of partner time to associate level time, we have given particular scrutiny to the court's exercise of discretion. While we think that much of the time might reasonably be classified as appropriate for partners, we cannot hold that there was an abuse of discretion. In reviewing the court's allocations, we must look not only at the work which was classified as appropriate for associates, but also at the work which was classified as appropriate for partners. Since there is no serious factual dispute over the hours or nature of the services, we cannot hold that the court's findings are clearly erroneous. The court's overall approach in such allocations, with respect to the appellants listed above, appears to have been evenhanded. There was no abuse of discretion. Thus the listed reductions of partner time to associate levels of compensation will be affirmed.

The time of Walter Riordan, Esq. included in the lodestar was substantially reduced, and of the time allowed all 673.63 hours was reduced to associate level time. Mr. Riordan's appeal warrants separate consideration at Part V.C.6(d), infra.

(5) Elimination From the Lodestar of Certain Time Categories

We noted in Part III.C., that the court identified six categories of attorney activity for which time would be eliminated in whole or in part from the lodestar calculation. Each appellant affected by such elimination contends that the court erred. Thus we address each category.

(a) Pretrial Memorandum Time

The trial court labeled the pretrial memorandum as "one of the major examples of maladministration in this case." 98 F.R.D. at 81. The court found that a total of 4,500 hours were expended in preparation of this document; over 1,173 hours by appellants. Rather than isolating specific wasted hours, the court cut by one-half the hours charged by each firm to this task.

The appellants point out that the 643 page pretrial memorandum was in strict compliance with the trial court's preclusive Pretrial Order No. 83, quoted at n. 8, supra, as well as Pretrial Orders Nos. 70 and 76. They point out, quite correctly, that the required document was more than simply a memorandum. The court had required, with preclusive effect, an outline of the plaintiffs' entire case. That approach to pretrial proceedings in complex litigation has been expressly approved by this court. In re Japanese Electronic Products Antitrust Litigation, 723 F.2d 238, 259 (3d Cir.1983). Considering that over 250 depositions had been taken, and over one million documents had been amassed in pretrial preparation, in order to establish a conspiracy over ten years among fifteen defendants, it is hardly surprising that compliance with the court's preclusive pretrial order would require a massive pretrial memorandum.

The appellants challenge the trial court's finding that 4,500 hours were devoted to preparation of that memorandum. They point out that the figure appears to have been taken from Appendix XIII of the Weil, Gotshal & Manges Report, a document not in evidence, in which time charged by each affected firm to plaintiffs' trial memorandum and final contentions work is totaled. The court's opinion contains no explanation of the manner in which the 4,500 hour figure was compiled. Objectors have not contradicted appellants' contention that 902.65 hours were included in the calculation which were not, in the time records of the firms involved, attributable to work on the pretrial memorandum. As best we can tell in the absence of trial court findings, the discrepancy between plaintiffs' figure and the court's occurred because, between June 1, 1980 and August 15, 1980, the plaintiffs were working on two interrelated and overlapping projects: preparation of final answers to defendants' contention interrogatories and preparation of the court-ordered pretrial memorandum. After the work had already begun, on July 7, 1980, the court acceded to the request by Harold Kohn that he be permitted to answer the contention interrogatories in the pretrial memorandum. Thus it appears that the 4,500 hour figure includes not only the task of complying with the court's requirement for preparation of a pretrial memorandum, but also the task of answering defendants' contention interrogatories. The court does not address the latter task.25

But while we cannot on this record accept at face value either the court's finding that 4,500 hours were spent preparing a 643 page pretrial memorandum, or its observation about the worth of that document, there is nevertheless considerable evidence in the fee petitions themselves that, whether because of friction between Harold Kohn and other lead counsel, or because of the complexity of the organizational structure, there was great duplication of effort. The committee structure resulted in assignment of discovery of individual defendants to subcommittee members. The work of the subcommittee members had to be forwarded and reviewed by an overall discovery committee and by lead counsel. Thus it appears to us that the court's negative reaction to the inefficient manner in which the pretrial memorandum was prepared has some factual support.

That does not, however, justify an across the board cut which may have penalized both the efficient and inefficient alike. The court's award of fees for 50% of the pretrial memorandum time confirms that the preparation of that document pursuant to, and in the form specified by court orders, was to that extent beneficial to the class. Given the difficulty we have encountered in reconciling the 4,500 hour figure with the fee petitions, the absence of more detailed findings in support of that computation, the confusion over the pretrial memorandum and answers to defendants' contention interrogatories, and the Lindy I and II requirement of individualized treatment of attorneys fee requests, we are not able to affirm an across the board 50% reduction in pretrial memorandum time to be included in the lodestar. Some reduction probably was appropriate for duplication, since duplicated effort did not benefit the class. Without more specific findings, however, we are at a loss to determine what time was beneficial and what was not.

(b) Pretrial Conference Time

The affected attorneys contend that the court committed legal error, or abused its discretion, in disallowing time spent at pretrial conferences by all attorneys except lead counsel and their assistants. Typical of the arguments advanced in support of inclusion of this time is that on behalf of Lawrence Walner & Associates, Ltd., that it enabled the attending attorneys:

to obtain a first-hand impression of the attitude of the court toward the case. In the period before trial, pretrial conferences and hearings are one of the few opportunities afforded counsel to gauge the attitude of the court toward the case. A facial expression, a tone of voice, may communicate volumes.

Brief for appellant Lawrence Walner & Associates, Ltd., 38-39. What is missing in these arguments, however, is any indication of the manner in which observation of the court's demeanor by a multitude of attorneys in any way inured to the benefit of the class. The very organizational structure agreed upon by plaintiffs' counsel and approved by the court contemplated a division of responsibility, with lead counsel, from whom trial counsel ultimately would be selected, having primary responsibility for dealing with the court. Thus rather than suggesting an abuse of discretion in disallowing time for attending pretrial conferences, we enthusiastically endorse that ruling.

Saveri & Saveri, P.C. makes a distinct objection that Executive Committee meetings were scheduled to coincide with pretrial conferences. That firm does not argue, however, that the court erred in classifying time spent at Executive Committee meetings as pretrial conference time. Thus we attach no significance to this objection.

Disallowance for lodestar calculations of time of appellants spent at pretrial conferences was on this record entirely proper.

(c) Fee Petition Time

Lawrence Walner & Associates, Ltd. urges that this court reconsider the holding in Lindy II, 540 F.2d at 111, that in fund in court cases time spent in preparation of fee petitions does not benefit the class and cannot be included in the lodestar. This panel is bound by the law of the circuit announced in Lindy II. Id. Even if we were free to disregard the Lindy II holding, we do not find persuasive the Walner firm's argument that such time should in fund cases be treated as it is treated in statutory fee cases.26

(d) Hours Billed Without Adequate Time Records of Activity

The affected firms contend that the trial court applied far too strict recordkeeping requirements when it excluded hours for time described in vague or meaningless terms or with insufficient documentation. We have carefully reviewed the court's rulings in this respect. Our review has been rather severely handicapped. The opinion displays a great deal of meticulous attention, with determinations for each attorney, but with respect to such attorneys a typical entry is:

14.3 hours are supported by entries which are too vague to provide any basis for determining if the time spent was of value to the class.

98 F.R.D. at 129. In each instance but one the court merely mentioned a number of disallowed hours, with no cross-reference whatever to dates or documents. Thus we have no way of knowing to which entries the court refers. Complicating the review problem further, the fee petitions are included in a 19 volume appendix, but the underlying time records containing the allegedly inadequate descriptions are not. The parties to this appeal cannot be faulted for that omission, however, for because of the manner in which the opinion deals with hours disallowed because of vague entries, they had no more information about the identity of the entries in question than we have. Thus the appellees' brief is no more helpful on this issue than the opinion.

The trial court is unquestionably right that the burden is on the fee applicant in an equitable fund case to demonstrate to the court's satisfaction that the work for which compensation is sought benefited the class. Thus the need for adequate time record entries is plain. See Pawlak v. Greenawalt, 713 F.2d 972, 978 (3d Cir.1983); In re Armored Car Antitrust Litigation, 472 F. Supp. 1357, 1358 (N.D.Ga.1979), modified on other grounds, 645 F.2d 488 (5th Cir.1981); In re Equity Funding Corp. of America Securities Litigation, 438 F. Supp. 1303, 1327 (C.D.Cal.1977). Moreover we suspect from the meticulous care with which the court examined the time records that the conclusion about inadequacy of entries in this case may be sound. We are confronted, however, with a situation in which, because of the court's failure to provide any means of identifying the entries found to be inadequate, appellate review, even for the purpose of satisfying ourselves that the court's findings are not clearly erroneous, is not possible. We cannot affirm the disallowance of hours without adequate time record support unless the trial court identifies the entries in question. More specific findings are essential. See Prandini v. Nat'l Tea Co., 585 F.2d 47, 52 (3d Cir.1978) (Prandini II ).

(e) Industrial Analysis Committee Time

In reaching the conclusion that all of the work of the Industrial Analysis Committee was basically useless, and thus excludable from the lodestar, the court followed a strong recommendation in Part VIII. D. of the Weil, Gotshal & Manges Report. In support of the conclusion that the work of that committee was of no benefit to the class that Report relied on the contents of letters from Harold Kohn to other plaintiff attorneys critical of the progress of work of the committee, and of the level of staffing devoted to it. The correspondence relied upon is included in the documents objected to on hearsay grounds. In accepting the Weil, Gotshal & Manges recommendation that the time devoted to the work of the Industrial Analysis Committee be completely disallowed, the court relied explicitly on the Kohn letters, and so far as the opinion discloses, on no other evidence.27 Thus we are in this instance required to consider appellants' relevancy and hearsay objections.

The contents of Kohn's letters, disclosed in that part of the opinion quoted in the margin, establish their relevancy. Kohn's evaluation of the worth of the committee work was the judgment of counsel who was responsible for the trial of the case. Clearly, however, the letters are hearsay, and thus inadmissible against the objectors unless they qualify under one of the exceptions in Rule 803. We have in Part V.B., supra, held that they do not qualify under Rule 803(24)(B). The only other exception arguably relevant is Rule 803(6), records of regularly conducted business activity. Unfortunately the court made no Rule 104(a) finding as to the preconditions of admissibility of the Kohn letters as business records, and aside from the representation by counsel that the letters were obtained from files of counsel, no foundation evidence appears of record. Moreover, as appears on the face of the letters, they were written at a time when friction between Kohn and Specks over management of trial preparation had become substantial. Thus we can hardly assume the letters reflected a regular practice of recording information. Finally, the letters were used by the court not merely to establish what transpired in the work of the committee, but for the author's opinion as to the worth of what had transpired.

If the parties were in agreement that the Kohn letters accurately set out the facts, we could perhaps hold that their use by the court was harmless error, since the court's appraisal of the worth of the committee work and Kohn's appraisal may have coincided. But the appellants dispute Kohn's version of the facts. For example, the court relies on Kohn's representation that his firm had obtained its own industry information sufficient to conduct settlement negotiations, while Specks contends that Kohn received those materials from a member of the Industrial Analysis Committee who had previously obtained it in conjunction with his work in the committee. Reply Brief of Specks & Goldberg, Ltd. at 21. Moreover even the Kohn letters do not suggest that all work of the Industrial Analysis Committee was useless. Some of it appears to have been useful, for example, in establishing the commonality of questions of fact necessary for the class action determination.

The court's complete reliance on hearsay evidence, in excluding all of the Industrial Analysis Committee time from the lodestar determination prevents an affirmance of that exclusion. Our own examination of the fee petitions, however, convinces us that not all of that time should be included. The firms serving on the Industrial Analysis Committee logged 818.6 hours of lawyer time and 815.75 hours of paralegal time. Among the appellants a disproportionate amount of partner time may have been devoted to a task which, while perhaps necessary, did not require the attention that partners devoted to it. Moreover, some of the work product may well have been of such poor quality that it cannot be said to have benefited the class. Thus we are compelled to remand for findings, based on properly admitted evidence, with respect to the number of hours of Industrial Analysis Committee work which should be included in the lodestar.

(f) Read and Review Time

As noted in Part III. C. (2) none of the appellants challenge the exclusion of read and review time expended by counsel other than the member of the executive committee.

(6) Other Specific Rulings

(a) Saveri & Saveri, P.C. contends that the court erred in excluding from its lodestar calculation the time of John Kithas and Charles Lamont. As we note in Part III.D. (1) supra, the court excluded this time because Kithas & Lamont is a separate entity, and Saveri & Saveri, P.C. failed to comply with the disclosure requirements of paragraph 12 of Pretrial Order No. 143. This ruling adopts the recommendation of the Weil, Gotshal & Manges Report. In response to that Report, Saveri & Saveri, P.C. contends that although Kithas and Lamont, working out of Saveri's office, have their own clients, these two lawyers have in major lawsuits always been treated as associates. Saveri & Saveri, P.C. contends, further, that this arrangement is a not unusual one in San Francisco law firms. The failure to comply with paragraph 12 of Pretrial Order No. 143 was explained as an innocent misinterpretation on its part of the arrangements to which the paragraph applied.

We hold that the court did not abuse its discretion in declining to excuse noncompliance with the Pretrial Order requiring disclosure of the arrangement with Kithas and Lamont. Thus the exclusion of 1717.25 hours from the lodestar of Saveri & Saveri, P.C. was proper. 98 F.R.D. at 203.

(b) Rogers, Rude, Candlin, Faulkner & Sjostrom contends that the court erred in denying their fee request for violation of paragraph 12 of Pretrial Order No. 143. They contend that although they did have an arrangement with Walter E. Riordan, P.A., it was not the hourly rate which the court found, but some form of complicated loan arrangement. Thus, they urge the court erred in determining that the pretrial order applied to them. Considering the contents of the deposition taken by Mr. Faulkner, App. N 180, admissible against Rogers, Rude, Candlin, Faulkner & Sjostrom under Fed.R.Evid. 801(d)(2), and the late filing by that firm of its notice of association, we cannot hold that the trial court's findings of fact as to the nature of the relationship between the two firms is clearly erroneous. Given that finding, the court did not abuse its discretion in declining to award any fee to Rogers, Rude, Candlin, Faulkner & Sjostrom for noncompliance with the disclosure requirements of the Pretrial Order.

(c) Sloan and Connelly, P.C. contends that the court erred in disallowing from the lodestar 50% of its time devoted to a pricing study. The court concluded:

It is my considered judgment that this pricing project involved too many hours of too many attorneys. The end-product does not justify so substantial a charge against the class fund as Sloan & Connelly requests. Accordingly, to reflect my determination that an excessive number of hours are claimed for this work as compared to the results achieved, I will only permit half of the total of 1818.65 hours to be compensated.

98 F.R.D. at 206-07. In reaching this conclusion the trial court relied almost completely upon the submissions made by Sloan & Connelly, P.C. and on that firm's own work product. There is one reference, 98 F.R.D. at 206, to a hearsay statement in a letter by Harold Kohn to Granvil I. Specks, but in light of the court's overall treatment of the subject matter, we conclude that any error in considering that hearsay statement was harmless. The trial court's appraisal of the value to the class of the damage studies is entitled to deference since the court was thoroughly familiar with the liability and damage issues presented by the case. We hold, therefore, that the court did not abuse its discretion in eliminating 50% of the damage study hours from the lodestar as unbeneficial to the class.

(d) Walter E. Riordan, P.A. contends that the court erred in the treatment of his fee petition in several significant respects. Some relate to the manner in which the court handled the hearings, including its reliance on the Weil, Gotshal & Manges Report. We have addressed those contentions, common to other counsel, in Part V.B., supra.

The court disallowed 201 hours expended, after this class action was settled, in connection with the testimony of a potential witness in the majority states case which went to trial. That ruling is unquestionably sound.

Of the 1,347.25 hours which the court regarded as attributable to this case the court identified 100 hours as having been spent on tasks which could have been performed by non-lawyers. Walter Riordan, P.A. does not dispute the hours involved, although, like other appellants, it disputes the propriety of second guessing the professional judgment of the attorneys in allocating work. For the reasons set forth in Part V.C. (4) we hold that a reduction in the lodestar amount would be proper for these hours. The court, however, both reduced Riordan's requested rate to $50 and cut the 1,347.25 hours by 50%. In making the time reduction the court relied in part on the nature of the tasks performed. 98 F.R.D. at 178. Thus, in contrast with other counsel, Riordan was penalized twice for performing tasks which should have been assigned to associates. The court gives no explanation for this double reduction. We hold that it is inconsistent with Lindy standards and an abuse of discretion.

The court also lists 858 hours as reflected by vague time record entries. In this instance several specific entries are quoted, but in no case have we been able to correlate disallowed time and entries. Thus we have the same difficulty with these 858 hours as with the hours discussed in Part V.C. (5)(d), supra. Moreover, in Riordan's case the court did not simply eliminate specific hours from lodestar time. Relying both on the nature of the tasks performed and the vagueness of the records the court reduced the time by 50%. Our difficulty with that approach is the same as with the hours discussed in Part V.C. (5)(d), supra; we are simply unable to perform our review function absent more specific findings.

Walter E. Riordan, P.A. also contends that the court committed legal error in holding that he could not be compensated for any of the 678 hours expended by him when, at the direction of Granvil I. Specks, he appeared at depositions taken by the plaintiffs and the defendants, representing the witness James Nelson. The court ruled:

It is clear from a reading of the transcript that Riordan was not representing the class during these depositions. The class will not be required to pay for the 678 hours Riordan spent representing Mr. Nelson at his deposition.

98 F.R.D. at 177 (footnotes omitted). Riordan concedes that he entered an appearance at the deposition on behalf of Nelson. He urges that if the court has found that he was not representing the class as well, that finding is on this record clearly erroneous.

James Nelson was the President of General Paper Company, a paper distributor in Minneapolis, which went bankrupt. Riordan and another attorney, James Perry, interviewed Nelson, obtaining from him a detailed description of how the defendants fixed prices. Nelson also related that he had diaries and memoranda concerning price information. At the evidentiary hearing on his fee application Granvil I. Specks identified Nelson as the plaintiffs' most important witness. There does not appear to be any real dispute over the fact that after a document search of General Paper Company records was conducted Specks decided that Nelson should be deposed, and that Riordan should appear for him. We do not understand the court to have held otherwise. Rather, we understand the court's holding to be that Riordan could not, as a matter of law, act both on behalf of a witness whose testimony is favorable to the class action plaintiffs and on behalf of the class.

The record contains substantial evidence from which it could be inferred that the class had a significant interest in protecting Nelson as a witness from hostile examination. The performance of that task could well have benefited the class, and it is undisputed that lead counsel Specks thought so when he directed Riordan to appear for Nelson at the depositions. We hold, therefore, that the court erred as a matter of law in disallowing the 678 hours devoted to representing Nelson at depositions. The extent to which the class was benefited must be determined in the first instance by the trial court.

We will remand in the appeal of Walter E. Riordan, P.A. for redetermination of its lodestar in a manner consistent with this opinion. The court should also reconsider the rate at which Mr. Riordan should be compensated for such hours as are eventually considered in his lodestar.

(e) The Specks & Goldberg, Ltd. Negative Multiplier

Our final task is the most difficult one presented by this appeal: consideration of the Specks & Goldberg, Ltd. objection to the imposition on it of a 50% negative quality multiplier. An adjustment based upon quality may be made up or down to reflect "exceptional services only; it may be considered in the nature of a bonus or penalty," Lindy II, 540 F.2d at 118. Moreover, "[t]he heavy burden of proving entitlement to such an adjustment is on the moving party." Id. Thus we must determine whether the court's conclusion that the objectors carried the burden of justifying a negative quality adjustment is supported by the record.

Specks & Goldberg, Ltd.'s first objection is that the court failed to consider the great results obtained. In advancing this contention it seeks an affirmative quality multiplier, and in this respect Specks & Goldberg, Ltd. has no better case than the remaining appellants. For the reasons set forth in Part III.C. (2), we hold that the court did not abuse its discretion in denying to Specks & Goldberg, Ltd. an affirmative quality multiplier.

Next Specks & Goldberg, Ltd. urges that in imposing a negative quality penalty the court applied an incorrect legal standard. It urges that such a penalty can be imposed only if the court finds that the predominant purpose of the firm so penalized was to delay or obstruct the proceedings. In advancing this contention the appellant takes out of context some language in Lindy II, 540 F.2d at 118. The Lindy II court did not intend that the quality factor be so circumscribed. The test is whether "the lawyer discharged the professional burden undertaken with a degree of skill above or below that expected for lawyers of the caliber reflected in the hourly rates." Id. In this case the chief task undertaken by Specks & Goldberg, Ltd. was the management of trial preparation by the numerous law firms participating in such preparation. The court could properly ask, and did in fact ask, whether the objectors' complaints of poor management are justified.

Specks & Goldberg, Ltd. also urges that because the court eliminated from the lodestar of all firms so many hours for activities deemed not beneficial to the class, there was no justification for a quality penalty for its alleged bad management. We reject that contention. Lindy I and II plainly contemplate that the quality factor be taken into consideration after the determination of the lodestar amount.

Finally Specks & Goldberg, Ltd. contends that the court's findings in support of its ultimate conclusion that trial preparation was mismanaged is not supported by admissible evidence, and is clearly erroneous. The evidentiary objection is predicated upon the court's ruling admitting the documents offered on behalf of the objectors. We have carefully examined those parts of the court's opinion discussing the Specks & Goldberg, Ltd. management role (Part II, 98 F.R.D. at 70-76; Part XI, 98 F.R.D. at 210-16). The court refers to some of the documents to which Specks & Goldberg, Ltd. objects. Some, such as minutes of committee meetings, are admissible against committee members on several grounds. Some are hearsay, but in no instance do we find that the hearsay was given such significant weight by the trial court in making its mismanagement determination that its admission was prejudicial. Specks & Goldberg, Ltd. objects to certain pejorative references to Granvil I. Specks, such as the reference to his acting like a "potentate, recruiting supporters and insuring their loyalty by dispensing favors at the expense of the class," 98 F.R.D. at 215, his "recruiting" of law firms, id., and his plan from the beginning "for the distribution of patronage." 98 F.R.D. at 71. Although less pejorative descriptions might have been appropriate, none of these objections really address the central findings that the case was managed in a manner that involved too many attorneys spending too many hours, and that Granvil I. Specks, on the evidence presented in support of his own fee petition, was to a large degree responsible for that management.

The court's findings that there was mismanagement and that Specks was responsible for it are not clearly erroneous, and the application of a negative quality multiplier was not an abuse of discretion. Since there must be a remand for recalculation of the Specks & Goldberg, Ltd. lodestar the amount of that multiplier must, of course, be redetermined.

We have found more in the trial court's opinion of which we wholeheartedly approve than we have found necessary to reverse. The court's review of massive fee petitions was burdensome and that burden was discharged with considerable care and attention. The difficulties which the trial court encountered are those which will confront many other courts as a result of the emerging law on fee shifting, a law with which American courts have not until recent years had to cope. It is not surprising that on some issues we should disagree with the trial court in this dynamic area of the law.

The judgment will be affirmed insofar as it denies the fee petition of Rogers, Rude, Candlin, Faulkner & Sjostrom. With respect to all other appellants the judgment will be reversed and the case remanded for redetermination of the amounts of fees and expenses to be awarded in a manner consistent with this opinion.

Although I concur in virtually all of Judge Gibbons' opinion,1 I write separately to set forth a different perspective on three significant points.

1. I cannot join in the assertion in footnote 19 that "an examination of the case law in this circuit since Lindy I discloses that the principal beneficiaries of the heightened judicial scrutiny which that case required have not been class member beneficiaries of a settlement fund, but defendants resisting statutory liability for attorneys fees." That may or may not be the case, but I cannot endorse the unsupported assertion.

2. While I generally agree with Judge Gibbons' explication of the delay factor, I am troubled by the locution he uses to describe it. Specifically, I am concerned that his description of the multiplier for delay as being a function of the "time value of money" could be interpreted as requiring a mechanical calculation of the delay factor on the basis of interest rates.

Judge Gibbons properly divides the delay problem into two parts. The first part concerns the period of time before the settlement is paid into court. For this interval--which can, of course, last many years--Lindy II gives the district court discretion to award a multiplier for delay in payment. 540 F.2d at 117. Lindy does not, however, require that the district court award a delay factor equivalent to the prevailing or legal rate of interest for the relevant period of time; rather, the amount of the delay multiplier is also subject to the discretion of the court. This is in contrast to the calculation regarding the second delay factor, discussed at page 594 of Judge Gibbons' opinion; once the fund has been paid into court and is earning interest, counsel and litigants who are entitled to a share of the fund should also receive a pro-rata share of the interest earned by the fund pending its distribution.

I thus join in Part V.C(1) of the opinion on the understanding that the opinion means that the first part of the delay factor is awarded and calculated in the discretion of the district court.

3. I have several concerns relative to Judge Gibbons' discussion of the contingency multiplier. As a preliminary matter, I note my doubts that either Sprague v. Ticonic National Bank, 307 U.S. 161, 59 S. Ct. 777, 83 L. Ed. 1184 (1939), or Trustees v. Greenough, 105 U.S. 527, 15 U.S. 527, 26 L. Ed. 1157 (1881), is authority for the proposition that under the equitable fund doctrine, when a case brought against several defendants is settled with respect to some of those defendants, the costs of litigating the rest of the case may never be paid out of the fund established by the initial settlement.2

More generally, although I agree that, given the facts in this case, counsel were not entitled to rely on the proceeds of the first settlement to fund their pursuit of the second, I believe that the "equitable fund doctrine" does not universally preclude such a result. If there is a prior agreement between the litigants and counsel, or if, in the class action context, judicial approval of such a funding scheme is obtained after a partial settlement conference, an initial settlement fund might properly become a "war chest" in support of further litigation, so long as the parties involved in the suit against the remaining defendants are the same as those involved in the original settlement. I understand the holding in the majority opinion--that the district court's premise regarding the use of the earlier fund for the later litigation is "false as a matter of law"--to be limited to the facts in this case, where it is uncontroverted that no prior agreement or court approval of a funding plan is present. On this basis, I concur in that holding.3

An issue of more particular relevance to this case is whether a large partial settlement should have any impact on a district court's determination of the contingency multiplier that will be applied to calculate the amount of attorneys fees to be paid out of a second settlement. I believe that the first settlement should impact on the contingency determination regarding the second settlement in the following ways:

a) Certainly, if fees were partly or wholly reimbursable out of the original settlement already paid into court, see supra, any contingency multiplier as to the second settlement would have to be reduced to account for the fact that the counsel fees would otherwise have been reimbursed. This, however, is not a factor here.

b) I also believe that a district court might conclude that there is less risk attached to the pursuit of a second settlement because counsel are at least not at risk with respect to the time and funds that they have already expended in obtaining the first settlement. As a general rule, a person is willing to go only so far "in debt" in pursuit of a risky objective; the elimination of part of that debt may very well affect the person's perception of the risk involved in continuing the endeavor. This may have been one factor that the district court had in mind.

c) Although he intimates that the conditions are not met here, Judge Gibbons suggests that the contingency factor may be lowered if the initial settlement indicates an inevitability (or at least a strong likelihood) that a second settlement will follow. Supra at p. 594. I agree with this common sense proposition and offer the following commentary. In addition to examining the similarities and differences between the suits that settled initially and those that remained (with the assumption that, if at least one sophisticated defendant settled, it becomes very likely that other defendants faced with the identical suit will also settle), a district court might rely on various external factors pointing to the effect of the first settlement on the contingency of the second. For instance, it might be relevant that experienced practitioners who considered the suit initially too risky were willing to undertake a role in the case after the first settlement. This in fact happened in this case. Additionally, a court might take into account the fact that counsel reflected their subjective evaluation of the difference in risk by requesting a lower contingency multiplier for the second part of the litigation, as Harold Kohn in fact did here. Although Judge Gibbons might be correct that a finding that the first settlement made the second one "inevitable" would be "hard to support" on the record before us, supra at p. 594. I think the record shows that the first settlement made the second one much more likely.

I do not suggest that the factors I have adumbrated necessarily preclude a contingency multiplier, but only that they should be considered when determining the propriety and amount of a multiplier for the second stage of the litigation. Neither do I believe that Judge Gibbons' opinion rules out their application in this case. Rather, his opinion rejects only the overbroad rationale of the district court for denial of a contingency multiplier, i.e., that the risk in this case was eliminated because the attorneys were entitled to collect out of the first fund for subsequent litigation. I concur with the contingency multiplier part of the majority opinion on this understanding.