IN RE DIET DRUGS

OPINION OF THE COURT

JORDAN, Circuit Judge.

Three law firms and some of their clients challenge the final award of attorneys' fees that the United States District Court for the Eastern District of Pennsylvania entered on behalf of class counsel in this landmark class action. For the following reasons, we will affirm the award.

I. Background

A. Diet Drugs Litigation and Settlement

This appeal arises from multidistrict mass tort litigation concerning the appetite suppressants fenfluramine, marketed as "Pondimin," and dexfenfluramine, marketed as "Redux." Over its decade-long course, the case1 has generated nearly 8000 separate orders from the District Court and numerous prior rulings and opinions from this Court. E.g., In re Diet Drugs Prods. Liab. Litig., 543 F.3d 179 (3d Cir.2008); In re Diet Drugs Prods. Liab. Litig., 431 F.3d 141 (3d Cir.2005); In re Diet Drugs Prods. Liab. Litig., 418 F.3d 372 (3d Cir.2005); In re Diet Drugs Prods. Liab. Litig., 401 F.3d 143 (3d Cir.2005); In re Diet Drugs Prods. Liab. Litig., 385 F.3d 386 (3d Cir.2004); In re Diet Drugs Prods. Liab. Litig., 369 F.3d 293 (3d Cir.2004); In re Diet Drugs Prods. Liab. Litig., 282 F.3d 220 (3d Cir.2002). Although we have already set forth the background of the case and the class action settlement agreement more than once, see, e.g., Diet Drugs, 385 F.3d at 389-93; Diet Drugs, 282 F.3d at 225-29, we do so once more in order to provide context for our discussion of the fee award entered by the District Court.

Beginning in 1997, a tide of products liability lawsuits arose after researchers discovered an association between some commonly prescribed appetite suppressants and a series of disorders generally known as valvular heart disease ("VHD"). The drugs involved were "fen-phen," a diet drug regimen that paired fenfluramine with phentermine, and dexfenfluramine, which was developed to produce the same anorectic effects as fen-phen without the need for a drug pairing. Evidence of serious coronary side effects from these drugs prompted the United States Food and Drug Administration ("FDA") to issue a public health advisory alert, and the pharmaceutical company Wyeth,2 which was responsible for the development and promotion of fenfluramine and dexfenfluramine, to withdraw the drugs from the market.3 Former fen-phen and dexfenfluramine users filed lawsuits and instituted class actions in numerous federal jurisdictions and state courts. Some of the earliest litigation took place in state courts in Texas, where attorneys, including Appellant Brian S. Riepen, brought numerous actions against Wyeth. Those suits generated extensive discovery, including the production by Wyeth of approximately three million pages of documents.

Many diet drug suits were also filed in federal courts. Pursuant to 28 U.S.C. § 1407, the Judicial Panel for Multidistrict Litigation transferred all of those cases, including more than 130 putative class actions, to the United States District Court for the Eastern District of Pennsylvania for coordinated and/or consolidated pretrial proceedings under MDL Docket No. 1203 (the "MDL"). Included among those actions were four cases filed by Riepen. Likewise, Appellants Freedland, Farmer, Russo, Behren & Sheller and Raymond Valori P.A. (collectively "Valori"4) represented approximately ten clients involved in the MDL.

The District Court appointed a plaintiffs' management committee (the "PMC") to oversee the litigation and to conduct discovery of general applicability to the MDL plaintiffs. The PMC began its discovery efforts in late 1998, and, by March 1999, it had taken 80 depositions and amassed approximately nine million pages of documents, from which it winnowed 5,000 documents that, the PMC claims, established Wyeth's liability. The PMC stored the discovery results in an electronic document depository and made it available to counsel for every plaintiff in the MDL. As part of the discovery process, representatives of the PMC attended regular status conferences held by a court-appointed special discovery master (the "Special Master") and prepared motions and responses regarding class-wide discovery, in addition to addressing a variety of other pre-trial issues.

Ultimately, the PMC filed a class action against Wyeth to pursue medical monitoring on behalf of former users of Wyeth's diet drugs. The PMC moved for class certification, and, on August 26, 1999, the Court granted the motion. By then, state courts in Illinois, New Jersey, New York, Pennsylvania, Texas, Washington, and West Virginia had also certified medical monitoring classes.5

In April 1999, Wyeth, the PMC, and a coalition of counsel involved in the state court class actions began to negotiate a nationwide settlement. On November 18, 1999, they executed an elaborate settlement agreement (the "Settlement Agreement") that contemplated a series of options for class members.6 At the outset, class members could obtain an echocardiogram at Wyeth's expense, to determine if they suffered from VHD, or they could exercise an initial opt-out from the settlement and pursue their claims in separate tort cases. Class members who chose not to take the initial opt-out and were diagnosed with VHD would have a second choice to make: they could receive a cash and medical services benefit or exercise an intermediate opt-out from the Settlement Agreement, which, again, would free them to turn to the tort system.7 Wyeth agreed to waive its statute-of-limitations defense against tort claims by those opting out at that point. In exchange, intermediate opt-out claimants were barred from seeking punitive damages against the company.

Class members who took the cash and medical services benefit and developed more serious VHD before 2015 could choose from yet a third pair of options. They could receive payment in accordance with a matrix of calculations that assigned compensation based on different levels of severity of medical conditions.8 The "matrix claims" would be processed based on the attestation of a physician, with Wyeth being able to test the foundations of the claims through an audit process permitting medical review of up to 15% of all such claims, unless the Court ordered an expanded audit for good cause shown. Alternatively, class members with worsening VHD could exercise a back-end opt-out so that they could pursue their claims in tort under the same conditions applicable to the intermediate opt-out claims.

To fund these various remedies, Wyeth agreed to create a $3.75 billion settlement fund to be administered by court-appointed trustees. The settlement fund was to be divided into two sub-funds: Fund A, into which $1 billion would be injected, would be designated for the payment of all non-matrix benefits. Fund B, which would receive $2.55 billion, would be designated for the payment of the matrix benefits. The remaining $200 million would go into an account denominated the "Fund A Legal Fees Escrow Account," from which attorneys who helped to create and implement the Settlement Agreement would be paid for services related to the non-matrix benefits.

On August 28, 2000, the District Court certified the settlement class and approved the Settlement Agreement. The settlement was hailed as an innovative departure from ordinary settlements requiring class members to make a "once-and-for-all choice" between a private remedy scheme and the tort system. Richard A. Nagareda, Autonomy, Peace, and Put Options in the Mass Tort Class Action, 115 Harv. L.Rev. 747, 796 (2002). Plaintiffs' counsel and Wyeth were praised in particular for creating the staged opt-out opportunities, which were viewed as a bold compromise between the competing concerns of individual autonomy for the class members and a comprehensive legal peace for the corporate defendant. See id. at 797-805.

By the summer of 2002, however, the number of matrix claims submitted to the trust and the number of class members who exercised their intermediate and back-end opt-out rights (collectively, "downstream opt-outs") had grown well beyond Wyeth's expectations. Doubting the veracity of many of these claims, Wyeth and the PMC sought, and were granted, an order directing the medical review of 100% of the claims submitted to the settlement trust. While the 100% audit helped to ensure the integrity of the claims review, it also increased the cost of trust administration. Between the influx of new claims and the heavy processing burden they created, Wyeth feared that it would not have sufficient funds to satisfy all of its diet drug liabilities. The District Court likewise concluded that the settlement was in jeopardy, commenting that "it is not unlikely, absent some curative amendment, that thousands of deserving class members may never receive any compensation for their medical conditions from ingesting Pondimin and Redux." In re Diet Drugs Prods. Litig., 226 F.R.D. 498, 509 (E.D.Pa. 2005).

Wyeth and the PMC thus worked to amend the Settlement Agreement. Most significantly for present purposes, one of the changes, denominated the "Seventh Amendment," altered the payment matricies so that pending matrix level I and II claimants became so-called "Category One Class Members." The claims of those Category One Class Members were administered separately from the settlement trust, subjected to independent medical review, and compensated from a separate fund called the "Supplemental Class Settlement Fund," into which Wyeth injected an additional $1.275 billion.9

B. Attorneys' Fees

The MDL and settlement process yielded four potential sources for fees to compensate the PMC and other attorneys who had a hand in creating common benefits for the enormous class of claimants (collectively, "Class Counsel"). First, through Pretrial Order ("PTO") 467, entered in 1998, the District Court ordered Wyeth to withhold 9% of the payments it made to plaintiffs whose cases were transferred to the MDL and place those funds in the "MDL Fee and Cost Account," from which Class Counsel would be compensated for providing case-wide services. Likewise, the Court provided for the sequestration of 6% of the value of claims in state court cases where the litigation was coordinated with the MDL.10 That money also went into the MDL Fee and Cost Account. The percentages were to be deducted from the fees due to the individual lawyers for the opt-out claimants who recovered against Wyeth.11

Second, as discussed above, Wyeth deposited $200 million into the Fund A Legal Fees Escrow Account pursuant to the Settlement Agreement. That account was the means of paying Class Counsel for services related to the non-matrix benefits.

Third, also pursuant to the Settlement Agreement, $229 million was transferred from Fund B into an account known as the "Fund B Legal Fees Escrow Account" to compensate Class Counsel who helped to create the matrix benefits.

Fourth and finally, the Seventh Amendment authorized the Court to award a "Common Benefit Percentage Amount," i.e., "common benefit fees to attorneys for professional services ... found by the Court to be of `common benefit' to Category One Class Members." (App. at 11882.) Those common benefit fees would be drawn directly from the Supplemental Class Settlement Fund, and not a separate legal fees account that correlated to the fund.12 Where a Category One Class Member was represented by an attorney, the common benefit award was to be deducted from the attorney's fees.13

To summarize, then, the Court was authorized to pay Class Counsel from four distinct "pots" of money: the MDL Fee and Cost Account, the Fund A Legal Fees Escrow Account, the Fund B Legal Fees Escrow Account, and the Supplemental Class Settlement Fund. The first pot was established to compensate Class Counsel for services, such as its efforts in obtaining and storing discovery, performed for the benefit of all class members, including those who were compensated outside the context of the Settlement Agreement because, for example, they suffered from PPH, opted out of the Settlement Agreement, or participated in coordinated state litigation. Each of the other pots corresponded to a particular fund established pursuant to the Settlement Agreement and was evidently intended to reward counsel for creating the particular benefits that claimants received from that fund. Collectively, we will refer to these latter three pots as the "Settlement Accounts."

1. Interim Fee Award

In the spring of 2001, the District Court established procedures that would govern its consideration of the fee award due to Class Counsel. As an initial step, it required all Class Counsel to submit time and expense records to a court-appointed auditor and to a lawyer designated as Plaintiffs' Liaison Counsel. The auditor, who was charged with determining which items of time and expense met previously established criteria for payment, reported that seventy-two law firms had performed 354,431.49 hours of compensable work and that a "lodestar value" of $101,076,658.54 was appropriate in view of their services.14

Each law firm claiming to be Class Counsel then had to submit to Plaintiffs' Liaison Counsel a fee presentation, which was to contain a litany of information relevant to the services rendered, including "[a] summary of the professional time for which compensation or reimbursement is claimed ..." and "[v]erified copies of all pertinent time records which were maintained contemporaneously ... throughout this litigation...." (App. at 7733.) The seventy-two firms that provided their records to the auditor filed fee presentations with Plaintiffs' Liaison Counsel, who, on February 15, 2002, submitted to the Court a thirty-volume compendium containing the fee presentations. On the same day, those same seventy-two firms filed a joint petition for attorneys' fees in which they requested a total of approximately $567 million from the four available funds.

There were nine objectors to the joint petition, including Riepen.15 Riepen argued that he should not have to pay an MDL assessment because he did not use PMC discovery, and he argued that the requested class fee was too high, given what he viewed as the low risk of non-compensation in the case. On March 4, 2002, the District Court entered an order permitting the objectors to request and, subject to court approval, to take limited discovery regarding the petition. Riepen participated in several discovery conferences, but did not seek any discovery. Other objectors deposed PMC lead counsel on subjects that included the details of the records submitted to Plaintiffs' Liaison Counsel and the contributions of contract attorneys to the PMC's efforts.16 Once discovery was complete, the District Court held a two-day hearing on the propriety of the fee award sought in the petition.

The District Court ruled on the fee petition on October 2, 2002, in an order designated as "PTO 2622." In re Diet Drugs Prods. Liab. Litig., Civ. Action No. 99-20593, 2002 WL 32154197 (E.D.Pa. Oct.3, 2002). Based on its findings that (1) "[t]he PMC faced significant risk at the beginning of the litigation that the work they did would be unsuccessful and uncompensated," (2) "[t]he discovery package created by the PMC ultimately paved the way for the class settlement and many individual settlements," and (3) "the PMC conferred great benefits on all litigants in the MDL and state-coordinated litigation [and]... performed their duties with admirable skill, diligence, and efficiency," the Court awarded Class Counsel 6% of the recoveries by claimants whose actions were part of the MDL and 4% of the recoveries by claimants in coordinated state actions (the "6% & 4% Assessment").17 Id. at *19. That entitled Class Counsel to a distribution of $76,861,455 from the MDL Fee and Cost Account.

As to the fees to be drawn from the Settlement Accounts, the Court found it "premature to perform a definitive ... analysis ... [because t]here is a significant amount of work still to be done ... in assisting the administration of the Settlement Agreement." Id. at *11. It concluded, however, that Class Counsel was entitled to a payment of almost $77 million— $38,430,728 from the Fund A legal fees escrow account and the same amount from the Fund B legal fees escrow account.18 Riepen and other objectors appealed, but we dismissed for lack of jurisdiction, holding that PTO 2622 was neither a final order nor a collateral order from which an appeal could be brought. Diet Drugs, 401 F.3d 143.

2. Final Fee Award

On January 5, 2007, the Court sought suggestions regarding the procedures and timetable it should use in determining a final fee award. It invited any interested party to submit a memorandum on the subject, and it scheduled a hearing for March 1, 2007.19 In response, the PMC filed a compendium of written agreements among the now ninety firms that claimed entitlement to common benefit fees and between those firms and a group of lawyers that came to be known as the "Major Filers." The Major Filers consist of more than fifty law firms that together represented approximately 97% of the downstream opt-out plaintiffs, 26,000 claimants who were compensated under the Seventh Amendment, and half of all class members who received matrix payments through May 31, 2007.20 Neither Riepen nor Valori was among the Major Filers.

Those agreements (the "Major Filer Agreements"21) reflected the shared understanding of Class Counsel and the Major Filers that "the amount to be awarded in common benefit fees ... is to be determined by the Court in the exercise of its sound discretion...." (App. at 12976-77.) Subject to that understanding, however, Class Counsel and the Major Filers agreed that Class Counsel would request the entire $200 million originally deposited in the Fund A Legal Fees Escrow Account, plus interest; $178,111,111 from the Fund B Legal Fees Escrow Account;22 and 7% of the benefits paid under the Seventh Amendment. Additionally, they agreed that it was appropriate for Class Counsel to levy a reduced assessment—2% in federal cases and 1.33% in state cases—on recoveries obtained by downstream opt-out claimants. The rationale behind the reduction, as stated in the Major Filer Agreements, was that recoveries in the downstream opt-out cases—as opposed to the initial opt-out and PPH cases—occurred in part because of the Settlement Agreement, which had its own mechanisms for compensating attorneys.23 In the PMC's words, the reduction "served as a prophylactic against `double dipping'" by Class Counsel. (Appellee's Br., No. 08-2387, at 25.) The Major Filers agreed that they would not object to the fees sought by Class Counsel, and the parties represented that there were no other "agreements, promises, or undertakings" among them. (App. at 13023.)

After the March 1 hearing, the District Court entered an order, in accordance with the procedures it had established to adjudicate the interim fee award, requiring that the auditor submit a report of the compensable time and expenses claimed by counsel, that Class Counsel submit supplemental fee presentations to Plaintiffs' Liaison Counsel, and that Plaintiffs' Liaison Counsel file a compendium of the fee presentations with the Court. The auditor reported that, from the inception of litigation, Class Counsel had expended 553,020.53 hours in common benefit time, thereby producing a lodestar value of $156,849,257.24. On July 16, 2007, Plaintiffs' Liaison Counsel filed the compendium of fee presentations and, on behalf of Class Counsel, a joint fee petition that complied with the terms to which Class Counsel and the Major Filers had agreed.

As it did in connection with the 2002 joint fee petition, the Court authorized those who objected to the petition to request limited discovery. Valori moved for discovery on August 7, 2007. The Special Master concluded that the motion was untimely and, because it did not adhere to the Court's instruction that any such motion set forth a concise statement of the need and legal basis for discovery, deficient. Nevertheless, the Court permitted Valori to depose the PMC's lead counsel regarding the terms and meaning of the compendium of agreements that lead counsel had submitted to the Court and whether any side agreements existed between Class Counsel and the Major Filers.

Thereafter, Valori filed an objection to the joint fee petition. The two primary arguments in the objection were that the requested award was too high because there was a low risk of non-payment, given the existence of the legal fees escrow accounts, and that the requested award did not allocate the burden of payment proportionally between the initial opt-out and PPH claimants, on one hand, and the downstream opt-out claimants, on the other, according to the benefits that each group received. Riepen also objected to the joint fee petition, renewing the arguments that he had made in response to the 2002 fee petition.

On April 8, 2008, the District Court ruled on the petition, in an order designated as "PTO 7763(A)." In re Diet Drugs Prods. Liab. Litig., 553 F.Supp.2d 442 (E.D.Pa.2008). It awarded Class Counsel a total amount of approximately $567 million, including the approximately $154 million that constituted the interim fee award, broken down in the following manner. First, the Court applied the percentage-of-recovery method24 to determine the appropriate fee award to be allocated from the Settlement Accounts. Id. at 467-85. In light of the reasonableness factors that we articulated in In re Prudential Ins. Co. Am. Sales Practice Litig. Agent Actions, 148 F.3d 283 (3d Cir.1998), and Gunter v. Ridgewood Energy Corp., 223 F.3d 190 (3d Cir.2000), it concluded that an award equaling 6.75% of the recoveries under the Settlement Agreement—which the Court valued at $6.44 billion—was fair and appropriate. Diet Drugs, 553 F.Supp.2d at 485. The Court then performed a lodestar cross-check "to gauge whether the ... award creat[ed] a windfall" for the petitioners. Id. By the Court's calculation, the award represented a lodestar multiplier of 2.6. Id. at 486. While it recognized that its multiplier was artificially low, given that the auditor's report did not separate the time spent on the Settlement Agreement from time spent on the MDL,25 it felt confident that the true multiplier—whatever it was—was not excessive for a "supermega fund case" such as this one.26 Id. at 487.

Second, the Court determined how to apportion the 6.75% award from among the Settlement Accounts. From the Fund A Legal Fees Escrow Account, it awarded Class Counsel $161,569,272, which, when added to its interim distribution of $38,430,728, equaled $200,000,000—the amount that was originally deposited into that account. Id. at 487. From the Fund B Legal Fees Escrow Account, it awarded $124,633,410.60, which, when added to its interim distribution of $38,430,728, equaled $163,064,138.60, which represented 6.39% of Fund B's original $2.55 billion value. Id. at 488. Finding no reason why the fee award pertaining to recoveries under the Seventh Amendment should be materially different from that related to matrix benefits,27 the Court also awarded Class Counsel 6.4% of the Supplemental Class Settlement Fund. Id. That distribution amounted to $71,447,638.10. Id.

Third, the Court determined that the justifications that supported the 6% & 4% Assessment in 2002 still applied with equal force, and it approved Class Counsel's proposed reduction in the assessment—to 2% in federal cases and 1.33% in state cases— for downstream opt-out claimants based on the logic that those claimants had incurred value from the Settlement Agreement that initial opt-out and PPH claimants did not incur, and Class Counsel was rewarded for creating that value from the Fund A Legal Fees Escrow Account. See id. at 491-96. It updated the award from the MDL Fee and Cost Account accordingly, adding $56,300,000 to the interim distribution of $76,861,455 for a total award of $133,161,455.28 Id. at 496.

The District Court then requested submissions regarding how to refund the money left over in the MDL Fee and Cost Account and the funds established pursuant to the Settlement Agreement, and it instructed Plaintiffs' Liaison Counsel to submit a plan regarding the allocation of the award among Class Counsel. On July 21, 2008, the Court completed adjudication of the fee award. PTO 7896 wrapped up the remaining refund and allocation issues, and PTO 7897 certified as final PTOs 2622, 7763(A), and 7896.

C. Appeal

Riepen filed a notice of appeal from PTO 2622, the interim fee award, and from PTO 7763(A), the final fee award, on May 6, 2008.29 Valori filed a notice of appeal from PTO 7763(A) on May 8, 2008. We have consolidated the appeals for disposition.

II. Discussion30

"[A] thorough judicial review of fee applications is required in all class action settlements." In re Gen'l Mots. Corp. Pick-Up Truck Fuel Tank Prods. Liab. Litig., 55 F.3d 768, 819 (3d Cir.1995). We review the District Court's fee determination for abuse of discretion, "which can occur if the judge fails to apply the proper procedures in making the determination, or bases an award upon findings of fact that are clearly erroneous." In re Cendant Corp. PRIDES Litig., 243 F.3d 722, 727 (3d Cir.2001) (internal quotations and citations omitted). To ensure that we will have a sufficient basis for review, we require district courts to set forth the reasoning in support of a fee award. In re Rite Aid Corp. Sec. Litig., 396 F.3d 294, 301 (3d Cir.2005).

Appellants object to three aspects of the fee award in this case: the level of transparency inherent in the process that led to the award, the size of the award derived from the Settlement Accounts, and the applicability of the MDL assessments to their individual cases. We address each challenge in turn.

A. Transparency of the Proceedings

Riepen claims that the fee award was the product of a flawed process in which the District Court accepted summaries from the auditor and Plaintiffs' Liaison Counsel instead of requiring the public filing of actual time and expense reports from Class Counsel. According to Riepen, the procedure adopted by the District Court violates the principles of transparency espoused by the United States Court of Appeals for the Fifth Circuit in In re High Sulfur Content Gasoline Prods. Liab. Litig., 517 F.3d 220 (5th Cir.2008). In High Sulfur, the lead plaintiffs' counsel in a class action persuaded the district court, during an ex parte hearing and without the benefit of supporting data, to divide a lump sum attorneys' fee award among more than six dozen plaintiffs' lawyers. Id. at 223. At that same hearing, the Court "accepted [l]ead [c]ounsel's proposed order sealing the individual awards; preventing all counsel from communicating with anyone about the awards; requiring releases from counsel who accepted payment; and limiting its own scope of review of objections to the allocation." Id. at 223-24. Because "the record [was] bereft of factual information essential to ... appellate review," and because the sealing of the record "protect[ed] no legitimate privacy interest that would overcome the public's right to be informed," the Court of Appeals vacated the award. Id. at 229-30.

There are two answers to Riepen's reliance on High Sulfur. First, this case is so factually distinct from that one that comparing the two is fruitless. Far from adjudicating the fee award in an ex parte hearing, the District Court solicited submissions from all interested parties concerning "what steps and procedures the court should implement, as well as a suggested timetable, in determining any final or other awards of attorneys' fees," and it held three public hearings on the matter. (App. at 12624.) Moreover, during adjudication of both the interim and final fee awards, the Court permitted objections and allowed objectors to take limited discovery, though it need not have granted any discovery at all. See Prudential, 148 F.3d at 338, 342 (recognizing that "discovery in connection with fee motions should rarely be permitted," and that "whether to grant discovery is committed to the sound discretion of the [district] court" (internal quotations and citation omitted)). Finally, the Court required the auditor and Plaintiffs' Liaison Counsel to submit volumes of data reflecting the time and money that Class Counsel spent on the diet drugs litigation—data that the Court put on the public record and used to support the fee award that it ultimately granted.31

Second, High Sulfur aside, the fee proceedings were amply transparent under our precedent. Indeed, it is difficult to discern what the District Court reasonably could have done to increase the level of transparency associated with the fee award. Riepen suggests that the Court should have considered and made public Class Counsel's individual billing records, but we have held that courts need not always engage in that time-consuming process. See Rite Aid, 396 F.3d at 306-07 ("[D]istrict courts may rely on summaries submitted by the attorneys and need not review actual billing records." (citing Prudential, 148 F.3d at 342)). In large cases, especially one of prodigious proportions like this, reliance on summaries is certainly within the discretion of the district court. Also, as the High Sulfur Court recognized, transparent fee proceedings are necessary, in part, so that we can engage in meaningful appellate review of the resulting award. The District Court's procedures in this case have been more than adequate to that end.32

B. Size of the Settlement Award

Appellants argue that the portion of the fee awarded from the Settlement Accounts, more than $434 million in all, was excessive. Riepen claims that the District Court improperly calculated the award as a percentage of the recovery, instead of using the lodestar method, and Riepen and Valori both contend that the award is based on an erroneous application of the reasonableness factors we have previously articulated.

1. Method of Calculation

"Attorneys' fees are typically assessed through the percentage-of-recovery method or through the lodestar method." In re AT&T Corp., 455 F.3d 160, 164 (3d Cir.2006) (citation omitted). The former "applies a certain percentage to the [settlement] fund." Id. (citation omitted). The latter "multiplies the number of hours class counsel worked on a case by a reasonable hourly billing rate for such services." 33 Id. (citation omitted).

Riepen argues that the District Court's employment of the percentage-of-recovery method was erroneous because, when a case involves fee shifting as does this one, the lodestar method should be used. Riepen's contention is misguided for two reasons. First, no "fee shifting," as that term is traditionally used, occurred in this case. Fee shifting—an exception to the so-called "American Rule," whereby parties pay their own attorneys' fees— occurs when one party is compelled by statute to bear the opposing party's fees. Alyeska Pipeline Svc. Co. v. Wilderness Soc'y, 421 U.S. 240, 269-70, 95 S.Ct. 1612, 44 L.Ed.2d 141 (1975). It is true that, under our precedent, the lodestar method is often applied in cases where fee-shifting statutes operate. Prudential, 148 F.3d at 333. But there is no such statute at work here. Wyeth voluntarily undertook the process of compensating opposing counsel, by establishing and funding various escrow accounts dedicated to the payment of claimants' legal costs. Second, even if this case could be said to involve fee shifting, Riepen does not complain about fee shifting at all. His problem is not that the burden of attorneys' fees was improperly "shifted" from the plaintiffs to Wyeth. Rather, Riepen is appealing the fee award to challenge the allocation of fees among the various attorneys who represented plaintiffs' interests.

Contrary to Riepen's characterization, this case falls under the common fund doctrine, a second exception to the American Rule. That "doctrine `provides that a private plaintiff, or plaintiff's attorney, whose efforts create, discover, increase, or preserve a fund to which others also have a claim, is entitled to recover from the fund the costs of his litigation, including attorneys' fees.'" In re Cendant Corp. Sec. Litig., 404 F.3d 173, 187 (3d Cir.2005) (quoting G.M. Truck, 55 F.3d at 820 n. 39). When calculating attorneys' fees in such cases, the percentage-of-recovery method is generally favored. Prudential, 148 F.3d at 333; see also The Manual for Complex Litigation § 14.121 (4th ed.2004) (reporting that "the vast majority of courts of appeals now permit or direct district courts to use the percentage method in common-fund cases").

It was entirely appropriate for the District Court to adhere to the general convention and apply the percentage-of-recovery method in this case. The lodestar method is "designed to reward counsel for undertaking socially beneficial litigation in cases where the expected relief has a small enough monetary value that a percentage-of-recovery method would provide inadequate compensation." Prudential, 148 F.3d at 333. Riepen contends, however, that the percentage-of-recovery method resulted in too much, as opposed to "inadequate," compensation for Class Counsel in this case. Moreover, the financial stakes in this case were enormous, and the lawyers involved were primarily concerned with obtaining relief for their clients and the members of the class, not with serving the public interest. Thus, the District Court correctly applied the method better designed to "reward[ ] counsel for success and penalize[ ] it for failure." G.M. Truck, 55 F.3d at 821.

2. Percentage-of-Recovery Analysis

In determining what constitutes a reasonable percentage fee award, a district court must consider the ten factors that we identified in Gunter, 223 F.3d 190, and Prudential, 148 F.3d 283. They are: (1) the size of the fund created and the number of beneficiaries, (2) the presence or absence of substantial objections by members of the class to the settlement terms and/or fees requested by counsel, (3) the skill and efficiency of the attorneys involved, (4) the complexity and duration of the litigation, (5) the risk of nonpayment, (6) the amount of time devoted to the case by plaintiffs' counsel, (7) the awards in similar cases, Gunter, 223 F.3d at 195 n. 1; Prudential, 148 F.3d at 336-40, (8) the value of benefits attributable to the efforts of class counsel relative to the efforts of other groups, such as government agencies conducting investigations, (9) the percentage fee that would have been negotiated had the case been subject to a private contingent fee arrangement at the time counsel was retained, and (10) any innovative terms of settlement, Prudential, 148 F.3d at 338-40; see also AT & T, 455 F.3d at 165.34

Trial courts must "engage in robust assessments of the [Gunter/Prudential] factors when evaluating a fee request," Rite Aid, 396 F.3d at 302 (citation omitted), and that occurred here. In an exhaustive opinion, the District Court gave thorough consideration to each of the factors. Appellants do not argue to the contrary. Instead, they challenge the Court's analysis of three particular factors: the presence or absence of substantial objections, the risk of nonpayment, and the value of benefits attributable to the efforts of other groups.35

i. Presence or Absence of Substantial Objections

The District Court found it "remarkable" that there were so few objections to the settlement terms and to the fees requested by counsel, given the approximately six million class members in the MDL. Diet Drugs, 553 F.Supp.2d at 473. By the Court's count, fewer than thirty objections to the Settlement Agreement, eleven objections to the interim joint fee petition, and only four objections to the final joint fee petition were filed. Id. Valori claims that it was clear error for the District Court to rely so heavily on the absence of objections to the final joint fee petition because the Major Filers, some of whom had vigorously contested the interim petition, agreed not to object. In essence, according to Valori, the Class Counsel improperly influenced this factor through their Agreement with the Major Filers, and the Court should not have considered it.36

Valori overstates the Court's reliance on the lack of objections. In fact, the Court explicitly declared that

Diet Drugs, 553 F.Supp.2d at 474 (quoting Cendant PRIDES, 243 F.3d at 730). Valori also fails to recognize the breadth of the Court's analysis. Whatever weight the Court gave to this factor it gave based on the dearth of objections throughout the settlement and fee adjudication process, instead of focusing only on the objections to the final joint fee petition. Finally, Valori distorts the effect of the agreement between Class Counsel and the Major Filers. The record indicates that only one Major Filer objected to the interim fee petition, and there is nothing but Valori's argument, unsupported by evidence, that suggests that more of the Major Filers would have objected to the final petition absent the agreement. In short, Valori's contention leaves us unpersuaded that the District Court erred—clearly or otherwise—in its consideration of this factor.

ii. Risk of Nonpayment

Appellants claim that the District Court applied the wrong legal standard to its risk-of-nonpayment analysis, made at least one erroneous factual finding, and neglected to consider an important risk mitigator. We disagree with them on each point.

Valori argues that the District Court erred as a matter of law by assessing the risk of nonpayment only at the beginning of litigation, instead of throughout the action. That risk, according to Valori, dissipated after the Settlement Agreement was reached and, pursuant to its terms, Wyeth agreed to fund escrow accounts from which Class Counsel would be compensated.

We have never addressed whether courts must reconsider the risk of nonpayment as the action evolves, and we need not do so here because, whether it was required or not, the District Court did reassess the risk in this case. Although the Court stated that it was evaluating the risk of nonpayment as of "the inception of the action and not through the rosy lens of hindsight," Diet Drugs, 553 F.Supp.2d at 478 (citing In re NASDAQ Market-Makers Antitrust Litig., 187 F.R.D. 465, 488 (S.D.N.Y.1998)), its analysis was more comprehensive than that. The Court specifically concluded:

Id. at 479. In practice, therefore, the Court evaluated the risk of nonpayment in the very manner that Valori advocates; Valori simply does not like the conclusion the Court reached.

In a related argument, Riepen takes issue with the Court's finding that Class Counsel "faced significant risk [of nonpayment] at the beginning of the litigation." Id. at 478. He claims that, to the contrary, the "deck [was] stacked against Wyeth from the very beginning." (Appellant's Br., No. 08-2363, at 41.) Wyeth, according to Riepen, faced potentially crippling liability, through state and federal litigation, and high-profile scholarship that established the link between the diet drugs and heart disease. As a result, it entered into settlement negotiations with the PMC very early in the litigation process, and as part of the settlement it agreed to pay more than $400 million into two funds from which class attorneys would be compensated.

Riepen confuses the risk of nonpayment at the inception of litigation with the risk immediately after the Settlement Agreement was executed. While the escrow funds undoubtedly reduced the risk of nonpayment, those funds were but one part of an intricate agreement that the PMC negotiated with Wyeth. If, as the District Court recognized, the Settlement Agreement "had not been structured to avoid a ruinous outcome for Wyeth, the efforts of [Class Counsel] would have been for naught." Diet Drugs, 553 F.Supp.2d at 479. Additionally, Riepen's view of the risk of nonpayment is more myopic than the Court's. As noted above, the Court assessed risk not at one fixed point in the action, but throughout its existence. Riepen does not challenge, for example, the Court's finding that the risk of nonpayment increased once the "second wave" of litigation threatened the Settlement Agreement, and, based on the record, he could not persuasively do so.

Lastly, Valori contends that the Court erred in evaluating the risk of nonpayment by neglecting to consider the "potentially billions of dollars in fees" that Class Counsel was receiving from their representation of diet drug claimants in concurrent state law cases. What individual counsel received in a particular state case, however, is irrelevant to the fee award here, which compensates Class Counsel for services that benefitted all class members, as well as the litigants in coordinated state actions. See, e.g., Rite Aid, 396 F.3d at 302 n. 11 (refusing to "conflate [ ] two distinct settlements" when considering the "reasonableness of the attorneys' fees based on one settlement fund").

iii. Value of Benefits Attributable to Others

In assessing whether Class Counsel had benefitted from "the efforts of other groups, such as government agencies conducting investigations," AT & T, 455 F.3d at 165 (citation omitted), the District Court noted that this case differed from the typical antitrust or securities litigation —in which the Gunter/Prudential factors are often considered—"where government prosecutions frequently lay the groundwork for private litigation," Diet Drugs, 553 F.Supp.2d at 481. The Court concluded that, while Class Counsel was in some sense beholden to the scholars who linked the diet drugs to VHD, and beholden as well to the FDA for its efforts to remove the drugs from the market, Class Counsel had not relied on "the government or other public agencies to do their work for them as has occurred in some cases." Id. at 481-82.

According to Riepen, the Court committed an error of law by limiting its analysis to the efforts of scholars and the FDA, and thereby ignoring the contributions of the lawyers who, while conducting contemporaneous diet drugs litigation in Texas state courts, obtained millions of pages of discovery from Wyeth and took 43 depositions before a single deposition took place in the MDL. Because the MDL trial docket lagged behind those in state court cases in Texas, Riepen believes that the Texas lawyers provided Class Counsel a "litigation road map ... [:] At the end of the day, the PMC only had to take depositions for a few months ... before [Wyeth] initiated settlement discussions with them [sic]." (Appellant's Br., No. 08-2363, at 39.)

Riepen is correct that the District Court did not mention the Texas lawyers' work in conjunction with this factor. That does not mean, however, that the Court ignored the contributory efforts of the Texas lawyers in determining an appropriate percentage of recovery. The issue was litigated during both the interim and final fee adjudication, and the Court determined that, whatever the Texas cases may have added, the recoveries arising from the MDL were due to the "herculean efforts" of the PMC37—in developing the case against Wyeth, in negotiating an agreement that allowed Wyeth to resolve the claims against it, and in amending the Settlement Agreement when it appeared to be in jeopardy.38 Diet Drugs, 553 F.Supp.2d at 474.

But even if we agreed that the District Court undervalued the Texas lawyers' contributions—or if we agreed with any one of Appellants' discrete challenges—we would not, on that basis alone, vacate the fee award. "The fee award reasonableness factors `need not be applied in a formulaic way' because each case is different, `and in certain cases, one factor may outweigh the rest.'" AT & T, 455 F.3d at 166 (quoting Rite Aid, 396 F.3d at 301). Our task is to discern whether the Court's percentage-of-recovery analysis, when examined in its totality, supports the fee that it finally determined was appropriate.

The Court made numerous findings pertaining to the Gunter/Prudential factors that Appellants do not dispute. For instance, it found that (1) the work of Class Counsel yielded a $6.44 billion settlement fund that benefitted more than 800,000 class members; (2) the Diet Drugs litigation was complex,39 and it endured significantly longer than did other super-mega-fund cases;40 (3) Class Counsel had devoted an extraordinary amount of time to the Settlement Agreement and the litigation surrounding it; (4) the requested award was, in percentage terms, slightly below the average award granted in the super-mega-fund cases; (5) the Major Filers' consent to the joint fee petition indicated that the petitioners were not seeking fees in excess of market value; and (6) many of the Settlement Agreement's features—including the multiple downstream opt-out rights—were innovative and "ha[d] already served as models for other cases."41 Diet Drugs, 553 F.Supp.2d at 472-85. On the basis of those extensive— and uncontested—findings alone, we would conclude that the District Court's fee award was not an abuse of discretion. As noted above, however, the aspects of the Court's analysis that Appellants challenge also support the percentage-of-recovery award that Class Counsel received.42

C. Applicability of the MDL Assessments to Appellants' Cases

Appellants argue that it was improper, under the common benefits doctrine, for the Court to levy assessments against their clients (1) who recovered against Wyeth on their PPH claims, which were not covered by the Settlement Agreement ("PPH clients"), and (2) who exercised initial opt-outs from the Settlement Agreement and recovered against Wyeth independently ("initial opt-out clients").43 In addition, Riepen contends that the MDL assessment should not have been levied against one of his client's recoveries because the District Court lacked subject matter jurisdiction over that client's claims.

"Under the common benefit doctrine,44 an award of attorney's fees is appropriate where `the plaintiff's successful litigation confers a substantial benefit on the members of an ascertainable class, and where the court's jurisdiction over the subject matter of the suit makes possible an award that will operate to spread the costs proportionately among them.'" Polonski, 137 F.3d at 145 (quoting Hall, 412 U.S. at 5, 93 S.Ct. 1943). Thus, in order to obtain common benefit fees, an attorney must confer a substantial benefit to members of an ascertainable class, and the court must ensure that the costs are proportionately spread among that class. Id. Both Riepen and Valori argue that Class Counsel conferred no benefit on their initial opt-out and PPH clients. Valori also contends that the initial opt-out and PPH plaintiffs bear a disproportionate burden of the assessment, when compared with the downstream opt-out claimants.

The PMC questions whether the common benefit doctrine even applies in multidistrict litigations such as this one, and suggests that the principal basis for the exercise of a district court's power to levy an assessment "derives from the docket management powers of the federal judiciary." (Appellees' Br., No. 08-2363, at 57.) The Judicial Panel for Multidistrict Litigation is empowered, under 28 U.S.C. § 1407, to transfer related cases to a single court, and that court has—and is expected to exercise—the ability to craft a plaintiffs' leadership organization to assist with case management. Inherent in that ability, according to the PMC, is the power to fashion some way of compensating the attorneys who provide class-wide services.

The PMC finds support for its position in In re Air Crash Disaster at Fla. Everglades on Dec. 29, 1972, 549 F.2d 1006 (5th Cir.1977). Like this case, Florida Everglades was a multidistrict litigation in which, at the fee award stage, the MDL court granted the plaintiffs' management committee a fee award drawn from the fees received by individual plaintiffs' attorneys. Id. at 1008-09. The United States Court of Appeals for the Fifth Circuit concluded, over the appellants' objection, that levying such an assessment was within the District Court's managerial power:

Id. at 1012.

The Fifth Circuit's observations are apt and apply with even greater force in this MDL, which dwarfed the size of Florida Everglades, with hundreds of thousands of class members spread all across the United States. That is not to say, however, that the District Court can ignore basic principles of fairness in applying an assessment. Florida Everglades is not to the contrary. Indeed, the Fifth Circuit vacated the fee award and remanded so that the district court could conduct a "hearing in the full sense of the word" and enter findings of fact and conclusions of law from which the court of appeals could determine whether the award constituted a fair and just enrichment of the plaintiffs' committee, should the district court's decision be appealed. Id. at 1021.

Likewise, we must ensure that the District Court granted Class Counsel a just award in this case. Whether we do so by applying the common benefits doctrine or independently assessing whether the District Court properly exercised its managerial power is of no real consequence. No matter how we label our analysis, we must determine whether the Court abused its discretion in concluding that Class Counsel conferred a substantial benefit on the initial opt-out and PPH plaintiffs or in how it spread the burden of the assessment among claimants who recovered outside of the Settlement Agreement.

1. Substantial Benefit

Appellants argue that their initial opt-out and PPH clients did not enjoy a substantial benefit from the PMC's services. By Appellants' lights, because those clients were not parties to the Settlement Agreement, they did not receive any of the benefits—such as medical testing or claims preservation—for which the PMC bargained. And, although the PMC obtained class-wide discovery to which all plaintiffs' attorneys had access, Riepen in particular contends that he did not use it in pursuing his initial opt-out clients' claims against Wyeth. Rather, he insists that the only pre-existing discovery that he used to develop his initial opt-out cases was procured by lawyers (including himself) in the Texas state court cases against Wyeth. Valori is less clear about whether he used the MDL discovery, but he nonetheless argues, generally, that the PMC's work product did not substantially benefit his initial opt-out and PPH clients.

According to the District Court, however, the PMC bestowed numerous benefits on initial opt-out and PPH claimants, even if their attorneys did not use the discovery that the PMC marshaled and retained. The mere availability of the discovery, in the Court's words, "substantially influenced [Wyeth's] evaluation of every plaintiff[']s case."45 (App. at 1194 (emphasis added).) More tangibly, the Court found that the PMC had, to the benefit of every claimant, helped to administer the MDL by tracking individual cases, distributing court orders, and serving as a repository of information concerning the litigation and settlement. Diet Drugs, 553 F.Supp.2d at 493. Furthermore, it obtained a number of favorable discovery and evidentiary rulings that applied on a litigation-wide basis, and it enforced a uniform procedure for the production of documents, deposition testimony, and expert disclosures that governed every MDL case against Wyeth.46

Appellants do not contest any of those findings, and each has substantial support in the record. We think it beyond reasonable denial that the initial opt-out and PPH claimants benefitted from Wyeth's loss of bargaining power due to the PMC's efforts. As the District Court noted, Wyeth had to defend itself against the initial opt-out and PPH claimants knowing that they had access to pertinent discovery and understanding that they, in turn, knew Wyeth was heavily invested in settling. It stands to reason, then, that those plaintiffs stood a better chance of recovery from Wyeth than they would have absent the PMC's efforts. Thus, the PMC conferred a substantial benefit on the initial opt-out and PPH claimants.47

2. Proportionality

Valori argues that the burden of the assessment fell disproportionately on the initial opt-out and PPH claimants and that the fee award must be vacated on that basis.48 As he notes, a 6% & 4% Assessment49 was levied against all opt-out recoveries pursuant to the interim fee award, but, as applied to downstream opt-out claimants, the Court, in its final analysis, accepted the PMC's recommendation that the assessment be reduced to 2% & 1.33%. The PMC explained the logic behind the disparate assessments thusly: Downstream opt-out claimants, as opposed to initial opt-out and PPH claimants, recovered in part because of the Settlement Agreement. The medical monitoring that Wyeth paid for allowed them to establish their injuries, and, without the claims preservation terms to which Wyeth agreed, they may have been frozen out of the tort system. Therefore, part of the fee award drawn from downstream opt-out recoveries came out of the Settlement Accounts and the assessments from the MDL Fee and Cost Account were reduced accordingly. The reduction prevented Class Counsel from being paid twice for the benefits conferred on the downstream opt-out claimants.

That the PMC created, and the District Court approved, a "prophylactic against `double dipping'" is laudable. (Appellee's Br., No. 08-2387, at 25.) However, it is also a non-sequitur as a response to Valori's contention that the burden of the fee award was not allocated proportionately to the benefits that each group of claimants received. Unlike the other aspects of the District Court's well-reasoned opinion that we have already addressed, its logic regarding the assessments, as allocated between the downstream opt-out claimants and the initial opt-out and PPH claimants, is assailable.50 Even if we credit the math employed by the PMC and the District Court and assume that, despite the disparate assessments, all opt-out and PPH claimants paid a roughly equal portion of the fee award, there is a sound argument that the downstream opt-out claimants received, for the same price, a greater tangible benefit from Class Counsel's services than the initial opt-out and PPH claimants received.

The inquiry we must make, however, is not whether a portion of the District Court's logic is subject to criticism, but whether the fee award itself constitutes an abuse of discretion. Certainly, limits on the reasoning behind an award may lead to the conclusion that the award itself cannot stand. But when, as in this case, the result can otherwise be justified, we are not compelled to find an abuse of discretion. Cf. In re Nortel Networks Corp. Sec. Litig., 539 F.3d 129, 134 (2d Cir.2008) (approving fee award even though the district court neglected to use awards granted in similar cases as benchmarks). We look, then, to the basis of Valori's proportionality attack to see whether the District Court's order can be justified in spite of the attack.

Valori contends that the District Court violated the principles espoused in Boeing, 444 U.S. 472, 100 S.Ct. 745, 62 L.Ed.2d 676.51 In Boeing, the Supreme Court asserted that it is proper to award fees to a common benefits attorney only where "the benefits of class recovery have `been traced with some accuracy' and the costs of recovery have been `shifted with some exactitude to those benefiting.'" Id. at 480-81, 100 S.Ct. 745 (quoting Alyeska Pipeline, 421 U.S. at 265 n. 39, 95 S.Ct. 1612). However, neither Boeing nor any other authority requires courts to ensure that the "costs of recovery" are shifted with exactitude among the various subclasses of claimants who benefitted from class counsel's efforts. Boeing and the other cases defining the contours of the common fund doctrine mandate only that the fee awarded to class counsel was reasonable given the benefit that these attorneys provided to the class members.52

Under Boeing, the pertinent question is not what the initial opt-out and PPH claimant paid in fees relative to what the downstream opt-out claimants paid.53 Rather, it is whether the initial opt-out and PPH claimants paid, in an absolute sense, a fair amount for the benefit of Class Counsel's services.54 The answer to that question, in our view, is a definite "yes." As noted above, the District Court found that Class Counsel substantially enhanced all claimants' chances for recovery by amassing meaningful discovery, drawing Wyeth to the bargaining table, and negotiating with Wyeth a comprehensive settlement that evinced the company's acute interest in resolving the claims against it. Appellants have not demonstrated that these findings were clearly erroneous. Indeed, they are amply supported by the record. Given all that the initial opt-out and PPH claimants received, the 6% & 4% Assessment levied on their recoveries was reasonable, and, thus, it satisfies Boeing's charge that costs and benefits be traced with some accuracy.

Any lingering concern that the fee award imposed a disproportionately heavy burden on the initial opt-out and PPH claimants shrinks when the proportionality issue is considered in the context of the fee award as a whole. Allocating the burden of the award among the claimants was but one part of the extraordinarily complicated endeavor of determining an appropriate award in this massive MDL. Even the discrete question of how to allocate the award was fraught with complex considerations, including how to treat the downstream opt-out claimants, who recovered outside the context of the settlement but still received valuable benefits under the Settlement Agreement, and what measures, if any, should be used to prevent Class Counsel from over-recovering—via the assessments and the Settlement Accounts—for their services that benefitted the downstream opt-out claimants. It would, in this case, be unwise to vacate the entire award based solely on how it was allocated, when the award is persuasively justified in all other respects and is justifiable in this one problematic respect.

Moreover, even if we were inclined to vacate the fee award based on the allocation, it is not clear to us that Valori's requested relief—a remand to the District Court with instructions to reallocate the award—would be feasible. Reducing the assessment on the downstream opt-out recoveries required the District Court to order refunds totaling more than $52 million to numerous law firms that, by prior court order, had paid the 6% & 4% Assessments into the MDL Fee and Cost Account. Months later, those refunds are not likely to be sitting in the bank accounts of the law firms that received them. It seems likely that taxes have been paid, referral counsel has been compensated, and, generally speaking, the refunds have, in all or in part, worked their way through the channels of commerce and, accordingly, would be difficult for the Court to reclaim.

We also find it significant— and surprising—that Valori, who has argued so vigorously that the allocation is unfair, never sought a stay of the refund distribution pending appeal. Had Valori moved for a stay, and had the Court granted his motion, the practical difficulties associated with administering the redistribution that he requests would be alleviated. When pressed on the matter during oral argument, Valori asserted that, in order to request a stay, he would have had to post a supersedeas bond—a bond that, given the enormous amount of money at issue in this case, he would not have been able to afford—so the Court probably would not have granted his request anyway. That defeatist stance is too convenient an excuse. Although Fed.R.Civ.P. 62(d) states that "[i]f an appeal is taken, the appellant may obtain a stay by supersedeas bond," courts may forego that requirement when there are other means to secure the judgment creditor's interests. See, e.g., Arban v. West Pub. Corp., 345 F.3d 390, 409 (6th Cir.2003) (expressing the view that Rule 62(d), which speaks to stays granted as a matter of right, does not constrain district courts from granting stays in accordance with their discretion); Fed. Prescription Serv., Inc. v. Am. Pharm. Ass'n, 636 F.2d 755, 759 (D.C.Cir.1980) (same); Munoz v. City of Phila., 537 F.Supp.2d 749, 751-52 (E.D.Pa.2008) (granting a stay without requiring a bond where the movant had sufficient funds to pay the judgment against it and there was "no basis to think that prompt payment [would] not take place should the judgment be sustained on appeal"). Here, the assessments and fees awarded pursuant to the Settlement Agreement were maintained in escrow accounts under the District Court's control. It is therefore quite possible, perhaps even likely, that the Court would have waived the bond requirement or required a substantially reduced bond in this case. All of that being said, we need not decide whether practical difficulties in administering a reallocation, or Valori's inaction in attempting to mitigate those difficulties, foreclose us from remanding the matter.55 As noted above, we do not believe that the District Court's allocation, viewed in context, constitutes an abuse of discretion.

3. Assessment on the Harris Case

According to Riepen, the MDL assessment, even if properly applied to the initial opt-out and PPH plaintiffs, should not have been applied to the recovery of his client Jana Harris because her case did not belong in federal court in the first place. In February 1999, Riepen field suit in a Kansas state court on the behalf of Harris, a citizen of Kansas. Among the defendants named in the suit was a pharmacy with its principal place of business in Kansas. Wyeth argued that the pharmacy was fraudulently joined to defeat removal, and it proceeded to file a notice of removal to shift the case to federal court. Riepen's co-counsel filed a motion to remand, but before the United States District Court for the District of Kansas could rule on the motion, the case was transferred to the Eastern District of Pennsylvania as part of the MDL. Harris settled her case with Wyeth, and the case was dismissed with prejudice pursuant to a motion that Riepen did not oppose, before the District Court denied the remand motion on December 7, 2000.56

"It is well-settled law that subject matter jurisdiction can be challenged at any point before final judgment," In re Kaiser Group Intern. Inc., 399 F.3d 558, 565 (3d Cir.2005) (citation omitted), but the necessary corollary is that subject matter cannot be challenged after such judgment is entered. See Hodge v. Hodge, 621 F.2d 590, 592 (3d Cir.1980) ("It was settled long ago ... that when a federal court proceeds to final judgment on the merits, the issue of subject matter jurisdiction is res judicata even though it was not litigated ...." (citation omitted)). Here, final judgment was entered in Harris's case when the District Court dismissed it with prejudice.

Riepen argues that the matter is still open because the District Court retained the ability to exempt him from the assessment until it issued its final, appealable attorneys' fee order in July 2008. But authority from the Supreme Court and our Court makes clear that a decision on the merits is separate from orders regarding attorneys' fees for the purposes of finality and appealability. See White v. N.H. Dept. of Employment Sec., 455 U.S. 445, 451-52, 102 S.Ct. 1162, 71 L.Ed.2d 325 (1982) ("[A] request for attorney's fees ... raises legal issues collateral to" and "separate from" the decision on the merits.); In re Colon, 941 F.2d 242, 245 (3d Cir.1991) ("treat[ing] attorneys' fees apart from the merits for purposes of appeal"). Thus, while Riepen may challenge the attorneys' fees and cost assessments that were imposed on him, he cannot do so by attacking subject matter jurisdiction on a case that was dismissed with prejudice almost ten years ago.57

III. Conclusion

The District Court set forth its reasoning in support of the fee award in a careful opinion that gives us a more than sufficient basis for review. It employed transparent procedures and undertook a thorough and proper analysis—based on the appropriate information—in determining the award. Given the duration of the litigation and the extraordinary efforts of Class Counsel, the amount of the award, though extraordinarily large, is not excessive in this extraordinary case, and while we have some reservations about the allocation of the assessments between the downstream opt-out claimants and the initial opt-out and PPH claimants, we do not believe the Court abused its discretion in apportioning the award as it did. We will therefore affirm the final fee award.

AMBRO, Circuit Judge, dissenting in part.

I join Judge Jordan's excellent opinion on all points save one—I believe the District Court abused its discretion in assessing the "downstream opt-out" plaintiffs at a lower rate for the case-wide services provided by the plaintiffs' management committee (the "PMC") than it assessed the "initial opt-out" and primary pulmonary hypertension ("PPH") plaintiffs. I would therefore grant the request of appellants—Freedland, Farmer, Russo, Behren & Sheller and Raymond Valori P.A. (collectively "Valori")—to vacate the District Court's order refunding fees exclusively to the downstream opt-out plaintiffs, and remand so that the refunds can be redistributed pro rata to all plaintiffs charged for the PMC's services.

To review, the District Court awarded Class Counsel attorneys' fees for assisting in the recoveries of two separate sets of plaintiffs whom its members did not represent: (1) plaintiffs who recovered within the class action (recoveries that essentially involved proving eligibility for the various funds created by the Settlement Agreement and its subsequent amendments); and (2) plaintiffs who recovered outside the class action (but whose recoveries were, according to the District Court, aided substantially by the PMC's case-wide services). The latter set of plaintiffs included three different groups: PPH plaintiffs (whose claims were not covered by the Settlement Agreement), initial opt-out plaintiffs (who opted out of the class action entirely and pursued individual tort actions against Wyeth), and downstream opt-out plaintiffs (who exercised their opt-out rights after receiving some benefits from the Settlement Agreement and with respect to whom Wyeth agreed, as part of the Settlement Agreement, to relinquish any statute-of-limitations defenses in exchange for those plaintiffs being barred from seeking punitive damages).

In 1998, the District Court ordered Wyeth to withhold 9% of all payments made to federal diet drug plaintiffs (whose cases had passed through that Court pursuant to the directions of the Judicial Panel for Multidistrict Litigation) and 6% of all payments made to plaintiffs in coordinated state cases.58 The money withheld was placed into the "MDL Fee and Cost Account," with the idea that it would later be used to compensate Class Counsel for the PMC's case-wide services.59 In 2002, the Court lowered those percentages to 6% and 4%, respectively. In 2008 (in one of the orders before us now), the Court ratified Class Counsel's proposal to refund a portion of the downstream opt-out plaintiffs' PMC fees while refunding nothing to the initial opt-out or PPH plaintiffs.60 The result of that refund was that the downstream opt-out plaintiffs ended up assessed at rates of (respectively) 2% and 1.33%, while the other plaintiffs subject to PMC fees were kept at 6% and 4%. Class Counsel made the refund proposal pursuant to an agreement it had reached prior to submitting its final fee petition with the so-called "Major Filers," a group that included lawyers for approximately 97% of the downstream opt-out plaintiffs (along with lawyers for a large number of plaintiffs who recovered within the class action). As part of that agreement, the Major Filers pledged to refrain from lodging any objections to Class Counsel's subsequent fee petition.

On its face, it seems suspicious that the one group that was charged less for the PMC's class-wide services also happened to be the one group that reached an outside deal with Class Counsel. Nonetheless, the District Court justified subjecting the downstream opt-out plaintiffs to the lower assessment through the following chain of reasoning. It reasoned, first, that, because the purpose of the PMC fees was to compensate Class Counsel for benefits provided to those who recovered outside the Settlement Agreement, it could not assess PMC fees on recoveries that were the product of the Settlement Agreement. The Court then inferred that, because in calculating the value of the Settlement Agreement it had credited that Agreement with producing half of the $2.3 billion recovered by the downstream opt-out plaintiffs,61 it would be improper for Class Counsel to receive PMC fees corresponding to the full amount recovered by the downstream opt-out plaintiffs. From that, the Court concluded that it should refund the excess back to the downstream opt-out plaintiffs. In taking this path, the District Court was (largely) adopting the reasoning that Class Counsel had laid out in its fee petition.62

As the majority recognizes, every step in the District Court's reasoning makes sense, except the last. It is true that Class Counsel's award for the Settlement Agreement included compensation for creating half the value recovered by the downstream opt-outs plaintiffs. But the money to fund that compensation did not come out of the downstream opt-out plaintiffs' recovery (even though it was compensation for enabling part of that recovery). Rather, as Class Counsel conceded during oral argument, that money came out of the $200 million Wyeth deposited into the Fund A Legal Fees Escrow Account. Accordingly, while it was appropriate for the Court to be concerned about authorizing double dipping (by allowing Class Counsel to recover twice, in two different capacities, for enabling the same recovery), there was no corresponding danger that the downstream opt-out plaintiffs would be double-charged. As such, the effect of the District Court's refund order was that the downstream opt-out plaintiffs ended up being charged less for the PMC's case-wide services than were the other groups subject to PMC fees,63 despite the fact that the downstream opt-out plaintiffs received no fewer benefits from those services, and, overall, certainly received more from the efforts of Class Counsel (since they benefitted both from the Settlement Agreement and the PMC's services).

The majority nonetheless holds that it was not an abuse of discretion for the District Court to order the excess PMC fees refunded solely to the downstream opt-out plaintiffs, rather than (as Valori urges us to do here) distributing the refunds among all those assessed such fees. The majority's reasoning essentially is that the applicable body of law merely requires that the fees assessed against a particular beneficiary be proportional to what that beneficiary received, not that the fees be proportional to those assessed against other beneficiaries. On that basis, the majority concludes that there was no abuse of discretion here because the PMC fees assessed against the initial opt-out and PPH plaintiffs were proportional to the benefits they received from the PMC's case-wide services, leaving those groups with basically no cause for complaint.

I agree that, viewed from the vantage point merely of the particular benefits the initial opt-out and PPH plaintiffs received, the PMC fees assessed against them were fair. I disagree, however, that that is the only legally relevant vantage point.64

For starters, I believe that the majority applied the wrong body of law. They derive their conclusion from Boeing Co. v. Van Gemert, 444 U.S. 472, 100 S.Ct. 745, 62 L.Ed.2d 676 (1980), a case that lays out the contours of the "common fund doctrine." Boeing, as the majority notes, does not articulate an explicit requirement that a district court, in awarding attorneys' fees, ensure that those fees are allocated proportionally across the entire class of beneficiaries.65 Yet the PMC fees were not assessed pursuant to the common fund doctrine.66 That doctrine applies only where (as in the case of the award for the Settlement Agreement) fees are being awarded as compensation for giving "each member of a certified class ... an undisputed and mathematically ascertainable claim to part of a lump-sum." Id. at 479, 100 S.Ct. 745. The PMC fees, on the other hand, were not awarded for creating a fund from which the different opt-out and PPH plaintiffs recovered. Rather, they were awarded for creating diffuse, non-monetary, benefits (e.g., discovery materials, drawing Wyeth to the bargaining table, etc.) that helped those plaintiffs recover outside the class action. As such, they were awarded pursuant to the "common benefit doctrine," which requires only that the party receiving the attorneys' fees have conferred a "substantial benefit" on "members of an ascertainable class," not that the benefit it conferred creates a recovery fund. Polonski v. Trump Taj Mahal Assocs., 137 F.3d 139, 145 (3d Cir. 1998).

The reason this distinction matters is that it is an explicit requirement of the common benefit doctrine that, in awarding fees, a "court ... ensure that the costs are proportionally spread among that class." Id. The District Court failed to do that here, and, accordingly, I believe it abused its discretion.67

In addition, I am more troubled than my colleagues that the District Court arrived at the lower rate of assessment for the downstream opt-out plaintiffs in response to an agreement reached between Class Counsel and a group, the Major Filers, that included almost all the downstream opt-out plaintiffs. I believe that, presented with a proposal that benefitted a group that was a party to the proposal (the downstream opt-out plaintiffs) at the expense of group that was not a party to it (the initial opt-out and PPH plaintiffs), the District Court was required to subject that proposal to extra scrutiny. That the District Court adopted Class Counsel's flawed reasoning more or less in full suggests to me that such scrutiny was not applied. That, too, was an abuse of discretion.

My suspicion is that what is driving the majority's reluctance to find an abuse of discretion here (despite agreeing that the District Court's reasoning was flawed) is its belief that it would be a shame "to vacate the entire award based solely on how it was allocated, when the award is persuasively justified in all other respects." Maj. Op. at 551. I share the view that the problem I am focusing on represents, at most, a minor blemish in the District Court's otherwise excellent, and persuasive, treatment of an extraordinarily difficult case. But I do not agree that rectifying the disproportionate allocation of the PMC fees requires anything so drastic as vacating the entire award. All that needs to be vacated is the separate order refunding the excess PMC fees exclusively to the downstream opt-out plaintiffs. That would leave the entire award to Class Counsel—both the $434,511,777.33 it received for the Settlement Agreement and the $133,161,455 it received in PMC fees—untouched.

I agree that there might be some administrative difficulties associated with reclaiming the $52 million in PMC fees that were already refunded to the downstream opt-out plaintiffs. I too share the majority's frustration with Valori's failure to request a stay of the distribution order he later challenged on appeal. I do not, however, consider such problems insoluble, since we deal here with purely fungible assets—money. For that reason, I consider wholly inappropriate Class Counsel's suggestion—wisely sidestepped by the majority, Maj. Op. at 552 n. 55—that we extend beyond the bankruptcy context the controversial doctrine of equitable mootness, which applies only to attempts to "unscrambl[e] complex bankruptcy reorganizations," and even then "`is limited in scope and should be cautiously applied.'" Nordhoff Invs., Inc. v. Zenith Elecs. Corp., 258 F.3d 180, 185 (3d Cir.2001) (quoting In re PWS Holding Corp., 228 F.3d 224, 236 (3d Cir.2000)).

In sum, I would vacate the order refunding the excess PMC fees exclusively to the downstream opt-out plaintiffs and remand with instructions that the excess be redistributed pro rata to all plaintiffs assessed such fees. Because the majority would affirm the District Court in all aspects, I respectfully dissent as to this issue only.