Mark Petri v. Stericycle, Incorporated

In the United States Court of Appeals For the Seventh Circuit ____________________ No. 20-2055 IN RE: STERICYCLE SECURITIES LITIGATION ST. LUCIE COUNTY FIRE DISTRICT FIREFIGHTERS’ PENSION TRUST FUND, et al., Plaintiffs-Appellees, v. STERICYCLE, INC., et al., Defendants. APPEAL OF: MARK PETRI, Objector-Appellant. ____________________ Appeal from the United States District Court for the Northern District of Illinois, Eastern Division. No. 1:16-cv-07145 — Andrea R. Wood, Judge. ____________________ ARGUED DECEMBER 8, 2020 — DECIDED MAY 18, 2022 ____________________ Before EASTERBROOK, KANNE, and HAMILTON, Circuit Judges. 2 No. 20-2055 HAMILTON, Circuit Judge. This appeal challenges an attorney fee award of 25 percent of a class-action settlement in this securities fraud case. Class member Mark Petri objected to the award and asked the district court to permit discovery into potential “pay-to-play” arrangements between class counsel and one of the public pension funds serving as a lead plaintiff. The court denied both requests, concluding that the fee award was reasonable and that the pay-to-play allegations lacked merit. Petri has appealed. We conclude that the district court did not give sufficient weight to evidence of ex ante fee agreements, all the work that class counsel inherited from earlier litigation against Stericycle, and the early stage at which the settlement was reached. We vacate the fee award and remand for a fresh determination more in line with what an ex ante agreement would have produced. With respect to the objector’s request for discovery into possible pay-to-play arrangements, we find no abuse of discretion, though we also would not have found an abuse of discretion if the discovery had been granted. I. Facts and Procedural History Stericycle is a waste management company with both government and private customers. Several years before this securities fraud case was filed, a former Stericycle employee brought a qui tam action under the federal False Claims Act and analogous state laws. United States ex rel. Perez v. Stericycle, Inc., No. 08 C 2390, 2016 WL 369192 , at *1–2 (N.D. Ill. Feb. 1, 2016). The whistleblower alleged that Stericycle was imposing illegal price increases on government customers with fixed-price contracts. Id. at *2. After investigation, New York settled with Stericycle for $2.4 million in 2013, and the other governments later settled for a total payment of $28.5 million. No. 20-2055 3 Id. Private customers also filed suit based on similar allegations and eventually settled for $295 million. In re Stericycle, Inc., Steri-Safe Contract Litigation, No. 13 C 5795, 2017 WL 4864874 , at *2 (N.D. Ill. Oct. 26, 2017). In October 2015, as these claims mounted and customers were leaving the company, the price of Stericycle’s common stock dropped from $149.04 per share to $120.31. The price of Stericycle’s depositary shares also fell, from $106.34 to $92.56. On behalf of the company’s investors, two Florida pension funds filed this securities fraud class action against Stericycle, its executives, members of its board, and the underwriters of its public offering. The complaint alleged that the defendants had inflated the stock price by making materially misleading statements about Stericycle’s fraudulent billing practices. Using the procedures of the Private Securities Litigation Reform Act, the district court appointed two other pension funds— the Public Employees’ Retirement System of Mississippi and the Arkansas Teacher Retirement System—as lead plaintiffs and Bernstein Litowitz Berger & Grossmann LLP as lead counsel for the class. Pleadings and motion practice followed for almost two years. The plaintiffs filed multiple amended complaints, and Stericycle countered with corresponding motions to dismiss. No merits discovery was conducted, which is also consistent with the Private Securities Litigation Reform Act. See 15 U.S.C. § 78u-4(b)(3)(B). With motions to dismiss still pending, the parties agreed to settle for $45 million. Lead counsel moved for a fee award of 25 percent of the settlement fund, as well as reimbursement of costs. Petri, a member of the class, objected only to the fee award, arguing that the amount was unreasonably high given 4 No. 20-2055 the low risk of the litigation and the early stage at which the case settled. Petri also moved to lift the stay the court had entered while the settlement agreement was pending so that he could seek discovery regarding class counsel’s billing methods, the fee allocation among firms, and counsel’s political and financial relationship with the Mississippi fund. The district court approved the $45 million settlement. The court also approved the proposed 25 percent attorney fee, finding the fee reasonable based on the contingent nature of the litigation and the positive outcome for the class. The court denied Petri’s discovery motion, reasoning that the fee award was based on a percentage of the fund rather than on billable hours or a lodestar calculation, the funds had already explained how they planned to distribute the award, and Petri had not provided any evidence of wrongdoing in the relationship between lead counsel and the Mississippi fund. Petri appealed both the attorney fee award and the discovery ruling. 1 II. The Fee Award We review class action fee awards deferentially, for abuse of discretion, recognizing that the district court is closer to the case than we are, and that a reasonable fee will often fall within a broad range. Birchmeier v. Caribbean Cruise Line, Inc., 896 F.3d 792, 797 (7th Cir. 2018). A district court abuses its discretion, however, if it “reaches an erroneous conclusion of 1 The district court rejected lead counsel’s argument that the fee should be calculated based on the gross settlement amount, without first netting notice and administration costs from that amount. The court also addressed lead counsel’s reimbursement request, concluding that all expenses were reasonable except for charges for online research. Those decisions are not at issue on appeal. No. 20-2055 5 law, fails to explain a reduction or reaches a conclusion that no evidence in the record supports as rational.” In re Southwest Airlines Voucher Litigation, 898 F.3d 740, 743 (7th Cir. 2018), quoting Harman v. Lyphomed, Inc., 945 F.2d 969, 973 (7th Cir. 1991). We also review de novo whether the district court’s legal analysis and method conformed to circuit law. Id.; see also Williams v. Rohm & Haas Pension Plan, 658 F.3d 629 , 635–36 (7th Cir. 2011) (approving district court’s method where “it weighed the available market evidence and it assessed the amount of work involved, the risks of nonpayment, and the quality of representation” (internal citation omitted)). In assessing the reasonableness of a fee request, a district court must attempt to approximate the fee that the parties would have agreed to at the outset of the litigation without the benefit of hindsight. Birchmeier, 896 F.3d at 796–97. The court should do its best “to award counsel the market price for legal services, in light of the risk of nonpayment and the normal rate of compensation in the market at the time.” Camp Drug Store, Inc. v. Cochran Wholesale Pharmaceutical, Inc., 897 F.3d 825 , 832–33 (7th Cir. 2018), quoting Sutton v. Bernard, 504 F.3d 688, 692 (7th Cir. 2007). The market rate for legal work “depends in part on the risk of nonpayment a firm agrees to bear, in part on the quality of its performance, in part on the amount of work necessary to resolve the litigation, and in part on the stakes of the case.” In re Synthroid Marketing Litigation (Synthroid I), 264 F.3d 712, 721 (7th Cir. 2001). This estimation ex post is “inherently conjectural,” In re Trans Union Corp. Privacy Litigation, 629 F.3d 741, 744 (7th Cir. 2011), yet district courts can look to actual fee agreements, data from similar cases, and class-counsel auctions to guide their analysis. Synthroid I, 264 F.3d at 719 . 6 No. 20-2055 The district court here acknowledged that it needed to make this effort to try to replicate what a pre-lawsuit market arrangement would or should have been. The court decided to award class counsel a percentage of the settlement—rather than calculating the award based on the lodestar of hours times hourly rates—and then said that it needed to “attempt to give counsel an amount that the parties themselves might have bargained for.” 2 From there, however, the court’s analysis was incomplete for three reasons. First, the court failed to consider an actual ex ante fee agreement between one of the funds and its counsel. Second, the court’s assessment of the risk of nonpayment did not give sufficient weight to the prior litigation involving Stericycle, which substantially reduced the risk of nonpayment. Third, in evaluating lead counsel’s efforts, the court did not give sufficient weight to the early stage at which the case settled. The cumulative effect of these issues leads us to conclude that the district court’s analysis did not sufficiently “re- flect the market-based approach for determining fee awards that is required by our precedent.” Sutton, 504 F.3d at 692 . We therefore vacate the 25 percent attorney fee award and remand for recalculation. A. Ex Ante Fee Agreement District courts deciding on attorney fee awards for class actions “must do their best to recreate the market by considering factors such as actual fee contracts that were privately 2 A district court may choose either the percentage method or the lodestar method, Americana Art China Co. v. Foxfire Printing & Packaging, Inc., 743 F.3d 243, 247 (7th Cir. 2014), and here the parties do not challenge the district court’s choice. No. 20-2055 7 negotiated for similar litigation, information from other cases, and data from class-counsel auctions.” Taubenfeld v. Aon Corp., 415 F.3d 597, 599 (7th Cir. 2005), citing Synthroid I, 264 F.3d at 719 . An ex ante agreement between the parties is a particularly useful guidepost for determining the market rate. See Synthroid I, 264 F.3d at 719 (“Only ex ante can bargaining occur in the shadow of the litigation’s uncertainty; only ex ante can the costs and benefits of particular systems and risk multipliers be assessed intelligently.”). The district court here did not address a September 2016 retention agreement between lead counsel and the Mississippi Attorney General. The agreement authorized lead counsel to represent the Mississippi fund in the Stericycle litigation and to seek a percentage of the recovery achieved for the class as compensation. That percentage, however, would be limited to “the percentage corresponding to [the Mississippi fund’s] estimated individual recovery set forth in Exhibit B … plus reasonable and necessary costs.” Exhibit B outlined the following tiered recovery structure: Twenty-five percent (25%) of any recovery up to Ten Million Dollars ($10,000,000.00); plus Twenty percent (20%) of any portion of such recovery between Ten Million Dollars ($10,000,000.00) and Fifteen Million Dollars ($15,000,000.00); plus Fifteen percent (15%) of any portion of such recovery between Fifteen Million Dollars ($15,000,000.00) and Twenty Million Dollars ($20,000,000.00); plus 8 No. 20-2055 Ten percent (10%) of any portion of such recovery between Twenty Million Dollars ($20,000,000.00) and Twenty-Five Million Dollars ($25,000,000.00); plus Five percent (5%) of any portion of such recovery exceeding Twenty-five Million Dollars ($25,000,000.00). In this schedule, “recovery” refers to the estimated recovery that [the Mississippi fund] is awarded as its share of the recovery achieved for the class. Exhibit B thus provides for increasing attorney fees, but declining percentages, as the settlement fund increases, which is generally consistent with widespread practices in cases generating funds to be distributed. See In re Synthroid Marketing Litigation (Synthroid II), 325 F.3d 974, 975 (7th Cir. 2003) (“[T]he market rate, as a percentage of recovery, likely falls as the stakes increase….”). At oral argument here, however, lead counsel asserted that this sliding scale structure applies only to the amount recovered by the Mississippi fund itself—not to the total amount recovered by the class. In this case, for instance, the fund’s share of the $45 million settlement presumably was far less than $10 million, so the Exhibit B schedule would be consistent with a 25 percent fee award. 3 3 In their motion for appointment as lead plaintiffs, the Mississippi and Arkansas funds asserted that they had purchased 462,826 shares of common stock during the class period (as it was defined at the time) and sustained losses of around $13 million. The Arkansas fund had also purchased 67,700 depositary shares. The parties’ settlement agreement estimated that if all eligible class members participated in the settlement, the average recovery—before deducting any fees, expenses, or costs—would No. 20-2055 9 That interpretation is consistent with the last sentence of Exhibit B of the agreement, but the limitation is improbable, arbitrary, unreasonable, and not consistent with a class representative’s fiduciary duty to class members. To see why, imagine a hypothetical settlement for $1 billion where the Mississippi fund suffered 1 percent of the class’s total losses. In that case, the fund would recover $10 million, and lead counsel would be entitled to a fee award of 25 percent of the entire settlement fund—$250 million. But a 25 percent fee in a $1 billion settlement would be well above fees awarded for such large funds, especially where counsel launched the case after others had done most of the heavy lifting and then settled early. See Stefan Boettrich & Svetlana Starykh, NERA, Recent Trends in Securities Class Action Litigation: 2018 Full-Year Review, at 33 tbl.2 (Jan. 29, 2019) (average fee award plus expenses for ten largest settlements since 2000, all over $1 billion, was 10.45 percent); id. at 41 fig.32 (median fee award for settlements over $1 billion was 7.6 percent in 1996–2013 and 15.4 percent in 2014–2018); Brian T. Fitzpatrick, An Empirical Study of Class Action Settlements and Their Fee Awards, 7 J. Empirical Legal Stud. 811, 839 tbl.11 (2010) (median fee award for 2006–2007 federal class action settlements over $1 billion was 9.5 percent); Theodore Eisenberg & Geoffrey P. Miller, Attorney Fees and Expenses in Class Action Settlements: 1993–2008, 7 J. Empirical Legal Stud. 248, 265 tbl.7 (2010) (median fee award for settlements over $175.5 million was 10.2 percent). be around $0.27 per share of common stock and $0.22 per depositary share. Based on those numbers, the funds’ estimated recovery would have been around $140,000, or around 0.31 percent of the $45 million settlement. 10 No. 20-2055 Lead counsel’s interpretation ties the fee award to only the Mississippi fund’s portion of the losses, but the award affects the fortunes of the entire class. As class counsel’s compensation increases, each class member’s recovery decreases. Class representatives owe fiduciary duties to class members. See Cohen v. Beneficial Industrial Loan Corp., 337 U.S. 541 , 549–50 (1949). It is hard to see how those class members would be well served by an agreement where they recover less if the Mississippi fund’s share of the losses is, for example, 20 percent rather than 50 percent. 4 The tiered structure in Exhibit B of the retention agreement makes more sense, and fits with a lead plaintiff’s fiduciary duty to class members, if it is understood to apply to the entire settlement fund—not just the Mississippi fund’s portion of the recovery. In our hypothetical $1 billion settlement, for instance, class counsel’s fee would be calculated as follows: 4The math works as follows: If the Mississippi fund’s share of the losses in this case had been 50 percent, then it would have received $22.5 million. Based on counsel’s interpretation of Exhibit B, class counsel would receive 25 percent of the first $10 million, 20 percent of the next $5 million, 15 percent of the next $5 million, and 10 percent of the remaining $2.5 million. That would work out to a total fee of $4.5 million, 10 percent of the fund. Now suppose the Mississippi fund’s share of the losses had been 20 percent, meaning it would have received $9 million. Looking again to counsel’s interpretation of Exhibit B, class counsel apparently would be entitled to a fee award of 25 percent of the entire settlement, or $11.25 million. In that scenario, class counsel would take almost $7 million more out of the total fund—significantly reducing the class’s recovery—even though the only thing that would have changed would be a smaller share of total losses for the Mississippi fund. No. 20-2055 11 $ 10,000,000 * (0.25) $ 5,000,000 * (0.20) $ 5,000,000 * (0.15) $ 5,000,000 * (0.10) $ 975,000,000 * (0.05) $ 53,500,000 That schedule would result in a fee of $53.5 million, or 5.35 percent of the total settlement, a number more consistent with the empirical evidence cited above. And in this case, applying that schedule to the $45 million settlement fund would result in an award of $5.75 million, or 12.78 percent of the settlement. The 25 percent fee that the district court approved is almost twice that amount. 5 We have recognized that sliding scale fee arrangements, such as the one set out in Exhibit B, are often the product of arms-length negotiations ex ante. See Silverman v. Motorola Solutions, Inc., 739 F.3d 956, 959 (7th Cir. 2013) (“[N]egotiated fee agreements regularly provide for a recovery that increases at a decreasing rate.”). In Silverman, which involved a $200 million settlement fund, we explained that such arrangements are logical because many litigation costs do not depend on the amount of damages, so it is “hard to justify awarding counsel as much of the second hundred million as of the first.” Id.; accord, Synthroid I, 264 F.3d at 721 (“Both negotiations and auctions often produce diminishing marginal fees when the 5These calculations do not account for any interest the settlement fund has accrued, see Synthroid II, 325 F.3d at 980 , or for removing notice and administration costs. 12 No. 20-2055 recovery will not necessarily increase in proportion to the number of hours devoted to the case.”). We also noted that our logic would apply with equal force to a $50 million settlement. Silverman, 739 F.3d at 959 . In an ex ante negotiation, therefore, it would make sense that a sophisticated, repeat-player plaintiff like the Mississippi fund would prefer a sliding scale arrangement to a flat rate. Cf. Aranda v. Caribbean Cruise Line, Inc., No. 12 C 4069, 2017 WL 1369741 , at *5 (N.D. Ill. Apr. 10, 2017) (“[I]n a hypothetical bargaining situation, well-informed class members … likely would shop around to see if any other firm would be willing to take their case and pursue a large recovery for a sliding-scale fee.”), aff’d sub nom. Birchmeier v. Caribbean Cruise Line, Inc., 896 F.3d 792 . That is not to say, of course, that a district court may never award a flat-rate fee. See Synthroid I, 264 F.3d at 721 (recognizing that in some circumstances a sliding scale arrangement will not be ideal). But where, as here, there is evidence of an ex ante agreement for a sliding scale fee structure, we expect a district court to give that evidence substantial weight in assessing the reasonableness of the proposed award. 6 6Petri also refers to a prior retention agreement between lead counsel and the Arkansas fund that authorized a 25 percent fee only in “smaller cases with special circumstances.” There is no copy of that agreement in the record. Petri cites testimony from an Arkansas fund executive in Arkansas Teacher Retirement System v. Bankrate, Inc., 18 F. Supp. 3d 482 (S.D.N.Y. 2014), that discusses the agreement. We are unable to determine conclusively from the transcript whether the executive was describing an agreement unique to the Bankrate litigation or a general retention agreement for all cases involving lead counsel and the Arkansas fund. On remand, lead counsel should provide an actual copy of that agreement as well as any retention agreements between lead counsel and the Arkansas No. 20-2055 13 B. Risk of Nonpayment Another important factor in a district court’s evaluation of a fee award is the risk counsel take that they will be paid nothing at all. Synthroid I, 264 F.3d at 721 . In a high-risk case, counsel is more likely to come away with nothing and thus would negotiate a higher contingent fee ex ante. See Birchmeier, 896 F.3d at 797 . Accordingly, a court trying to approximate a market rate must account for the relative risk of the litigation. See Trans Union, 629 F.3d at 746–48 (modifying fee award in part because special master who recommended award gave only “perfunctory” consideration to relative risk of loss). The district court here said only that the risk of nonpayment was “substantial” because the plaintiffs needed to establish intent to defraud and because there was a motion to dismiss pending when the case settled. That analysis was incomplete. The court failed to consider the prior litigation involving Stericycle’s billing practices and the subsequent, very substantial settlements, which signaled that class counsel were taking on a significantly reduced risk of nonpayment. Other relevant risk factors also weigh in favor of reconsideration. 1. Prior Litigation First, the district court did not discuss the qui tam action against Stericycle, the later investigation, or the eventual settlements the company reached with its government and private customers. While the existence of a prior criminal or civil proceeding is not dispositive, it can be a useful “proxy for fund that were prepared specifically for the Stericycle litigation. The district court should then take the terms of any such agreements into consideration when reevaluating the fee award. 14 No. 20-2055 assessing risk.” In re Dairy Farmers of America, Inc., 80 F. Supp. 3d 838, 848 (N.D. Ill. 2015). After the whistleblower filed her qui tam action, the New York Attorney General investigated Stericycle’s billing practices and obtained “documents consisting of contracts, invoices, payments, and correspondence that demonstrated that government customers were being charged price increases on a regular basis.” Stericycle, 2016 WL 369192 , at *2. Stericycle also produced “data consisting of its transactions with government customers from 2002 through June 2014.” Id. As noted above, Stericycle eventually agreed to settlements totaling over $325 million with its government and private customers. The prior litigation strengthened the securities plaintiffs’ case and substantially reduced lead counsel’s risk of nonpayment. Even lead counsel seemed to acknowledge as much when, before filing the fourth amended complaint, they asked the court to take judicial notice of the preliminary settlement of the private customers’ case. Stericycle had disclosed the settlement in a Form 8-K and also said that it was amending its debt agreements to establish a settlement fund and was developing guidelines for future price increases. At the time, lead counsel said: “These disclosures further corroborate Plaintiffs’ allegations here that Stericycle knowingly or recklessly hid from investors that it was engaged in a widespread scheme to automatically increase its prices without informing its customers….” In other words, lead counsel thought—quite No. 20-2055 15 sensibly—that the settlement made the plaintiffs’ case even stronger, lowering the risk of nonpayment. 7 The fourth amended complaint itself also relied heavily on information uncovered by the prior litigation. The complaint alleged, for instance: “The contours of the fraud have been confirmed in the settlement agreements in the Government Case and the Customer Case, through information provided to Lead Plaintiffs from former employees of the Company, and in sworn deposition testimony provided in the Customer Case—including testimony from Stericycle’s most senior executives.” The complaint then cited testimony from the private customers’ case to show that the billing increases “were 7 In the district court, lead counsel noted that the parties did not reach a settlement in the private customers’ case until August 2017, around one year after lead counsel had filed the original complaint in this case and “incurred the risk of non-recovery.” Even at the time that complaint was filed, however, litigation in the private customers’ case was well under way. See Stericycle, 2017 WL 4864874 , at *1 (observing that MDL panel had consolidated actions in August 2013 and that operative complaint had been filed in March 2016). And Stericycle’s settlement with the state of New York had already been announced in January 2013, on the same day that the whistleblower’s qui tam complaint was unsealed. See A.G. Schneiderman Announces $2.4 Million Settlement with Stericycle for Overcharging NY State and Local Entities, N.Y. State Off. of the Att’y Gen. (Jan. 8, 2013), https://ag.ny.gov/press-release/2013/ag-schneiderman-announces-24-million-settlement-stericycle-overcharging-ny-state. The settlement with the other governments, likewise, had been announced in October 2015. See Stericycle Announces Settlement to End 7-Year Qui Tam Suit, Stericycle (Oct. 8, 2015), https://investors.stericycle.com/press-releases/news-details/2015/Stericycle-Announces-Settlement-to-End-7-Year-Qui-TamSuit/default.aspx. So in an ex ante negotiation, there would have been ample evidence telling a sophisticated plaintiff such as the Mississippi fund that the prior litigation was likely to reduce the risk of nonpayment and to negotiate accordingly. 16 No. 20-2055 made with the specific purpose of inflating the Company’s publicly-reported revenue numbers in order to impress Wall Street.” In summarizing the allegedly fraudulent billing practices, likewise, the complaint repeatedly cited deposition testimony given by Stericycle executives and employees in the private customers’ case. And the complaint went on to allege that the qui tam case “further confirms the existence of Stericycle’s fraud with respect to the Company’s governmental customers.” Additionally, in attempting to establish a violation of the Securities Exchange Act, the complaint alleged that the defendants had made statements that were materially false and misleading in part because Stericycle was fraudulently increasing its rates. That allegation was based on the qui tam action and subsequent investigation. Lead counsel here also used the prior litigation to help demonstrate scienter, alleging that testimony from Stericycle executives in the private customers’ case showed that “the Officer Defendants were directly involved in developing and implementing the fraudulent automatic price increases.” The complaint alleged that the officer defendants “attempted to hide during the Customer Case that customers did not authorize the price increases.” None of this is to say that class counsel were wrong to rely on the prior litigation. Quite the contrary. We also recognize that class counsel still faced meaningful challenges. The prior settlements were with Stericycle’s customers, not investors, and the company still denied any wrongdoing, albeit after paying over $325 million in settlements. But even if class counsel carried the securities fraud ball across the goal line, the prior litigation gave them excellent starting field position, No. 20-2055 17 strengthening the plaintiffs’ case and substantially reducing class counsel’s risk of recovering nothing. That reduced risk would have been taken into account in any ex ante auction or market transaction for representation of the securities fraud class. Accordingly, the district court should have given more substantial weight to the effect of the prior litigation in considering class counsel’s risk of nonpayment. 2. Other Risk Factors Several other risk factors also weigh in favor of reconsidering the award. For one, class counsel points to the fact that no other firm filed a securities fraud case against Stericycle as evidence of the litigation’s high risk. It is true that a lack of interest from other firms “suggests that most members of the securities bar saw [the] litigation as too risky for their practices.” Silverman, 739 F.3d at 958 . In this case, however, a Minnesota pension fund initially moved for appointment as lead plaintiff. The fund asked the court to designate another experienced securities firm, Berman DeValerio, as lead counsel. So this was not a case where “no other law firm was willing to serve as lead counsel.” Id. In addition, class counsel rely on Stericycle’s reporting only $52 million in available cash at the time of the settlement as further evidence that there was a significant risk of nonpayment. But the analysis should be based on the risk that existed when the litigation began—not at the time of settlement. See Birchmeier, 896 F.3d at 796–97 (“When awarding fees to class counsel, district courts must approximate the fees that the lawyers and their clients would have agreed to at the outset of the litigation given the suit’s risks….”). 18 No. 20-2055 Finally, class counsel did not present any expert testimony to the district court about the magnitude of the risk they faced. While such evidence is not required, it can be useful to a district court that is attempting to estimate the ex ante risk of the litigation. See, e.g., Silverman, 739 F.3d at 958 (observing that expert report characterized case as “unusually risky” and that defendant “might well have prevailed on summary judgment but for some unanticipated facts plaintiffs’ lawyers turned up in discovery”); Hale v. State Farm Mutual Automobile Insurance Co., No. 12-0660, 2018 WL 6606079 , at *9 (S.D. Ill. Dec. 16, 2018) (noting that class counsel “were not assisted by any governmental investigations or prosecution” and that three class action fee experts “all opined that this case was as risky as they come”); Aranda, 2017 WL 1369741 , at *2–4 (relying in part on plaintiffs’ expert testimony about value generated by counsel in concluding that unique circumstances of case merited higher-than-average fee award). C. Amount of Work A third factor that affects the market rate for legal fees is “the amount of work necessary to resolve the litigation.” Synthroid I, 264 F.3d at 721 . A reduction may be warranted if the requested fee award is “disproportionate to the amount of work expended by class counsel.” Camp Drug Store, 897 F.3d at 833 . We have recognized that some bids in class-counsel auctions compensate lawyers based on how far the litigation progresses. Synthroid I, 264 F.3d at 721–22; see also Hale, 2018 WL 6606079 , at *10 (noting that “sophisticated market players typically set higher fee percentages when a case resolves during or after trial”). All other things being equal, a case that settled before the motion-to-dismiss stage, for instance, would be No. 20-2055 19 expected to result in a lower fee than a case that proceeded all the way to trial or beyond. Such terms are common in private fee agreements and “tie the incentives of lawyers to those of the class by linking increased compensation to extra work.” Synthroid I, 264 F.3d at 722 . 8 Here, the district court did not give sufficient weight to the early stage at which the case settled. The court made a passing reference to the early settlement in concluding that lead counsel had secured a good outcome for the class. (We are not as convinced the settlement was a good outcome, see note 3, above, but neither Petri nor anyone else is challenging here the $45 million settlement total.) But the court did not address whether the preliminary stage of the litigation warranted a reduction in the requested fee. See Camp Drug Store, 897 F.3d at 833 (upholding district court’s reduction of fee award where there was no paper discovery, no depositions taken, and no substantive motions filed). To be sure, this lawsuit involved more than “merely filing a complaint and negotiating a settlement.” Id. Because of the early settlement and the information lead counsel already had, however, it was not a case where the firm had to engage in extensive discovery or defend against a summary judgment motion. Cf. Silverman, 739 F.3d at 958–59 (approving 27.5 percent fee award where case settled after defendant’s motion for summary judgment was 8 Lead counsel themselves appear to have previously signed a retention agreement that tied attorney fees to how far the litigation progressed. In In re RH, Inc. Securities Litigation, No. 17-00554, 2019 WL 5538215 (N.D. Cal. Oct. 25, 2019), a securities class action brought by a Chicago pension fund, lead counsel agreed ex ante to a 15 percent fee “if a settlement was reached after a ruling on a motion to dismiss and before a ruling on summary judgment.” Petri App. A202. 20 No. 20-2055 denied and counsel spent more than $5 million on discovery and experts). As noted above, moreover, the district court did not discuss the impact of the prior litigation. That groundwork reduced not only the risk of nonpayment but also the amount of work required of class counsel. Without it, class counsel would have had to spend much more time and resources gathering evidence and taking depositions of Stericycle executives and employees. Again, we do not doubt that class counsel still needed to shoulder a substantial burden to achieve the result they did. They have earned a multimilliondollar fee here. But the prior litigation reduced that burden substantially. The district court should have given much greater weight to that factor in evaluating the fee request. Because the district court did not adequately consider the ex ante fee agreement, the risk of nonpayment, and the amount of work involved, we remand for reconsideration of the 25 percent fee award consistent with this opinion. 9 III. Discovery Issues Petri also appeals the district court’s denial of his motion to lift the stay on discovery. According to Petri, discovery was warranted to investigate potential pay-to-play arrangements between lead counsel and the Mississippi fund. We review discovery rulings for abuse of discretion. Allen-Noll v. Madison Area Technical College, 969 F.3d 343, 350 (7th Cir. 2020). Under 9Petri also challenges the district court’s failure to conduct a lodestar crosscheck. We have said, however, that “consideration of a lodestar check is not an issue of required methodology.” Williams, 658 F.3d at 636 . Here, the district court did not abuse its discretion when it concluded that a lodestar crosscheck was unnecessary. No. 20-2055 21 that standard, we will reverse only if “the judge’s ruling lacks a basis in law or fact or clearly appears to be arbitrary.” Kuttner v. Zaruba, 819 F.3d 970, 974 (7th Cir. 2016). The district court identified three categories of discovery sought by Petri: information about (1) counsel’s billing methods; (2) how the fee award would be divided among the participating firms; and (3) lead counsel’s financial and political relationships with the Mississippi fund and public officials who controlled it. The court concluded that discovery of the first two categories was unnecessary because the fee award was based on a percentage of the fund and because the plaintiffs had already explained that they planned to divide the total award based on each firm’s lodestar (i.e., hours times hourly rates). As for the relationship between lead counsel and the Mississippi fund, the court was not convinced by Petri’s pay-to-play allegations and denied the motion. The standard of review is decisive here. We find no abuse of discretion, though we also would find no abuse of discretion if the court had decided these issues the other way or somewhere in-between. A. Billing Methods First, we agree with the district court that discovery regarding counsel’s billing methods was not required. In requesting that information, Petri relies heavily on allegations by a former attorney at lead counsel’s firm that surfaced in Bernstein v. Bernstein Litowitz Berger & Grossmann LLP, 814 F.3d 132 (2d Cir. 2016). Bruce Bernstein, who was of-counsel at the firm, filed a complaint under seal against lead counsel and five individual partners in 2014. His allegations centered on another securities fraud class action in which lead counsel represented the Mississippi fund. Bernstein claimed that the 22 No. 20-2055 firm paid a local Mississippi attorney $112,500 in fees for a useless memo produced weeks after the case settled. He later learned that the attorney had little experience and was married to a lawyer in the Mississippi Attorney General’s Office. Id. at 136–38. According to Petri, scrutinizing lead counsel’s billing methods in this case would ensure that there is no such “pointless legal work undertaken for the excuse of generating a bill.” The relevance of the Bernstein allegations to this case is minimal. As a preliminary matter, the Second Circuit made clear that it was not assuming the truth of the allegations, noting that complaints frequently “contain allegations that range from exaggerated to wholly fabricated.” 814 F.3d at 143 (citation omitted). In fact, after interviewing witnesses and reviewing relevant documents, Bernstein himself said that he had “received information that seriously challenges my claims.” And Bernstein Litowitz has continued to serve as lead counsel—with the Mississippi fund as lead plaintiff—in other cases after the complaint was unsealed. See, e.g., In re Signet Jewelers Ltd. Securities Litigation, No. 16 Civ. 6728, 2019 WL 3001084 , at *9 (S.D.N.Y. July 10, 2019); cf. In re Merck & Co. Securities, Derivative & “ERISA” Litigation, MDL No. 1658, 2016 WL 8674608 , at *1–2 (D.N.J. Dec. 23, 2016) (relying in part on Bernstein’s later statement to conclude that Bernstein allegations did not meet Rule 60(b)(2) standard for relief from fi- nal judgment based on newly discovered evidence). More fundamental, there is little reason to think the firms were engaged in similar practices in this case. In Bernstein, lead counsel did not disclose—and was not required to disclose—the payment to the local Mississippi attorney when it submitted its fee petition to the court. 814 F.3d at 137 & n.2. No. 20-2055 23 Here, by contrast, lead counsel informed the district court that both Gadow Tyler PLLC and Klausner, Kaufman, Jensen & Levinson would be receiving attorney fees for their work on the case. Lead counsel added that no other firms would receive any fees. Attorneys from those two firms also submitted declarations describing their work. A partner from Gadow Tyler asserted that his firm’s participation included legal research in preparation of the third amended complaint, legal research prepared in opposition to Defendants’ motion to dismiss, meeting with Bernstein Litowitz attorneys to discuss case staffing and strategy, attending and participating in the mediation session held in Chicago, and participating in ongoing discussions about litigation strategy, settlement negotiations, and the settlement approval process. Furthermore, Gadow Tyler reviewed and edited certain lead plaintiff submissions, engaged in regular communications with the Office of the Mississippi Attorney General about case developments, and prepared and submitted regular reports to [the Mississippi fund]. A Klausner partner submitted a similar statement, affirming that his firm’s 27.8 hours on the case involved assisting lead counsel with the initial complaint filed on behalf of the two Florida pension funds. Both declarations also indicated that the firms had relevant experience: Gadow Tyler with securities class actions and Klausner with public employee retirement issues. And the overwhelming majority of their work 24 No. 20-2055 was completed before the case settled, further distinguishing this litigation from Bernstein. With all this information before it, and in granting a percentage-of-fund award, the district court did not abuse its discretion in denying Petri’s request for more detailed billing information. B. Fee Allocation Nor is discovery required regarding the allocation of attorney fees among the firms. As the district court observed, “Lead Plaintiffs have already explained how they intend to distribute the fee award.” Specifically, they had indicated that the firms will divide the fees according to their respective lodestars. The parties disagree about whether that disclosed allocation is sufficient. In his opening brief, Petri insisted that any other fee-sharing agreements would be “probative to the concern about political kick-backs.” Lead counsel responded: “The undisputed record is that there are no other fee sharing agreements—the sharing of fees between counsel has been fully disclosed.” Petri then suggested that this statement leaves open the possibility that there were earlier fee-sharing agreements with different terms. Even assuming Petri is right about the ambiguity of lead counsel’s statement, we do not see how earlier fee-sharing agreements would be relevant to our analysis. Suppose, for example, that lead counsel had initially agreed to give 20 percent of the fee award to Gadow Tyler and Klausner. Petri says that such an agreement would explain the 25 percent fee request: lead counsel had to ask for a huge fee to make up for the 20 percent portion going to the other firms. We take the No. 20-2055 25 point, but lead counsel’s request was never going to be the final word on the subject—the fee still had to be approved by the district court. Even if prior fee-allocation agreements existed, the district court did not abuse its discretion in denying this discovery request. 10 C. Lead Counsel’s Relationship with the Mississippi Fund Finally, Petri sought discovery regarding lead counsel’s relationship with the Mississippi fund and the elected officials who oversee its operations. He argued that the fund “has a pattern and practice of awarding lucrative legal work to firms that support [former Mississippi Attorney General Jim Hood].” The district court rejected that request as well, concluding that Petri’s allegations about lead counsel’s political contributions were insufficient to justify discovery. Reasonable judges could differ, but we find no abuse of discretion. 1. Pay-to-Play in Securities Litigation As the Third Circuit has explained, securities litigation involves unique pay-to-play concerns. See In re AT&T Corp., 455 F.3d 160, 168 (3d Cir. 2006). In securities class actions, especially under the terms of the Private Securities Litigation Reform Act, massive publicly managed pension funds often serve as lead plaintiffs. Those circumstances present the risk of “so-called ‘pay-to-play’ arrangements, such as where a law firm makes campaign contributions to elected officials who control governmental pension funds and is selected as the fund’s lead counsel.” Id. Such arrangements can distort fair 10Our decision on this point does not affect our earlier statement in note 6, above, that the earlier fee agreement with the Arkansas fund should be disclosed to help the court approximate the results of an ex ante negotiation here. 26 No. 20-2055 fee arrangements for the benefit of the class, for “the adversarial process is often ‘diluted.’” Id. One empirical study found that when pension funds whose managers have received campaign contributions serve as lead plaintiffs, they “appear to be less vigorous in negotiating attorney fees.” Stephen J. Choi, Drew T. Johnson-Skinner & A.C. Pritchard, The Price of Pay to Play in Securities Class Actions, 8 J. Empirical Legal Stud. 650, 651 (2011) (analyzing securities class actions filed between 2002 and mid-2007). Accordingly, district courts handling these cases “should be particularly attuned to the risk of pay-to-play.” In re Cendant Corp. Litigation, 264 F.3d 201 , 270 n.49 (3d Cir. 2001). At the same time, district courts must also “take care to prevent the use of discovery to harass presumptive lead plaintiffs.” Id. We see no reason that logic should not extend to discovery requests at the fee-award stage. Requiring some preliminary evidentiary showing before allowing such discovery is standard practice in both securities litigation, see 15 U.S.C. § 78u-4(a)(3)(B)(iv) (permitting discovery as to whether a class member is the most adequate plaintiff “only if the plaintiff first demonstrates a reasonable basis for a finding that the presumptively most adequate plaintiff is incapable of adequately representing the class”), and class actions more generally. See Fed. R. Civ. P. 23 advisory committee’s notes to 2003 amendment (“If the [fee] motion provides thorough information, the burden should be on the objector to justify discovery to obtain further information.”). 2. Lead Counsel’s Campaign Contributions Petri’s allegations are based on lead counsel’s campaign contributions to former Mississippi Attorney General Jim Hood, who held that position from 2004 to 2020. The No. 20-2055 27 Mississippi Attorney General’s Office has full authority “to bring, decide and settle cases on behalf of [the Mississippi fund].” According to Petri, four Bernstein Litowitz partners contributed a total of $20,000 to Hood’s campaign in October 2016, one month after the Mississippi fund moved to have the firm appointed as lead counsel in this case. Hood’s gubernatorial campaign also received $21,800 from various partners in April 2019, not long after the district court issued its preliminary approval of the settlement. And Bernstein Litowitz previously contributed $100,000 to the Democratic Attorneys General Association (DAGA), which provided a significant portion of Hood’s 2015 campaign budget. Based on these publicly reported facts, we cannot say the district court abused its discretion in denying discovery of possible further contributions. The allegations resemble those in Cendant, which addressed pay-to-play concerns in selecting the lead plaintiff and class counsel in another securities class action. The Cendant district court recognized a consortium of three pension funds as the presumptive lead plaintiffs because of the funds’ financial stakes in the litigation. Two other plaintiffs objected. They argued that the consortium could not protect the interests of the class because its chosen counsel had made campaign contributions to an elected official overseeing one of the funds, which “created an appearance of impropriety.” 264 F.3d at 269 (citation omitted). The district court rejected that argument because the plaintiffs provided no evidence that the contributions had influenced the consortium’s selection process. The Third Circuit affirmed, concluding that “[a]llegations of impropriety are not proof of wrongdoing.” Id. at 270. 28 No. 20-2055 Cendant also discussed steps that courts can take to mitigate pay-to-play concerns. In cases involving publicly managed funds, for example, courts might require lead plaintiffs to disclose any contributions by counsel to elected officials who oversee the fund. 264 F.3d at 270 n.49. If there is evidence of such contributions, the fund might be required to submit “a sworn declaration describing the process by which it selected counsel and attesting to the degree to which the selection process was or was not influenced by any elected offi- cials.” Id. The Third Circuit’s suggestions for guarding against payto-play activity may be useful at the fee-award stage as well. In this case, however, much of the suggested information is already in the record. Cf. Wal-Mart Stores, Inc. v. Visa U.S.A., Inc., 396 F.3d 96, 120 (2d Cir. 2005) (holding that decision to grant or reject objector’s motion for discovery regarding fairness of settlement depended on “whether or not the District Court had before it sufficient facts intelligently to approve the settlement offer” (citation omitted)). First, Petri submitted publicly available information about lead counsel’s contributions to Attorney General Hood’s campaigns and to DAGA. The district court reasonably concluded that the campaign contributions themselves did not justify discovery. See Cendant, 264 F.3d at 270 n.49 (concluding in context of lead plaintiff appointment that “evidence of campaign contributions, standing alone, does not create ‘a reasonable basis’ sufficient to justify party-conducted discovery”); see also In re Diamond Foods, Inc., Securities Litigation, 295 F.R.D. 240, 256 (N.D. Cal. 2013) (upholding choice of class counsel after requiring pension funds and counsel to describe No. 20-2055 29 selection process and to disclose certain contributions to Mississippi campaigns and DAGA). Petri also wants information about in-kind contributions and contributions by attorneys’ family members. If the district court had found that information about such contributions was needed to assess the reasonableness of the fee, it could have followed up on the issue. See Cendant, 264 F.3d at 270 n.49 (observing that evidence of campaign contributions would be sufficient “for the court, on its own initiative, to seek further information from the presumptive lead plaintiff”). But given the intrusive nature of the discovery and the limited value it seemed likely to provide, it was not an abuse of discretion for the court to deny Petri’s motion. Cf. Hemphill v. San Diego Ass’n of Realtors, Inc., 225 F.R.D. 616, 619 (S.D. Cal. 2005) (noting in settlement context that objectors “should be allowed ‘meaningful participation in the fairness hearing without unduly burdening the parties or causing an unnecessary delay’”), quoting In re Domestic Air Transportation Antitrust Litigation, 144 F.R.D. 421, 424 (N.D. Ga. 1992). Lawyers are “free to exercise their right to donate to politicians who support their views,” In re Countrywide Financial Corp. Securities Litigation, 273 F.R.D. 586, 604 (C.D. Cal. 2009), and the same is certainly true of lawyers’ family members. Second, an assistant attorney general in the Mississippi of- fice submitted an affidavit explaining the process for selecting counsel in securities cases. The office relies on a panel of eleven law firms to monitor the Mississippi fund’s investment portfolio. According to the assistant attorney general, those firms were selected based on their track records, resources, and reputations; campaign contributions “have no consideration in the selection process.” The office also has a “first-to30 No. 20-2055 approach” policy for selecting lead counsel, meaning that whichever firm initially flagged the case is selected. Here, Bernstein Litowitz was the only panel member that alerted the office to the Mississippi fund’s potential claims against Stericycle. We can imagine a district court finding such explanations not sufficiently persuasive, but in this case the court did not abuse its discretion in thinking that the selection process did not appear to have been tainted by political contributions. See Cendant, 264 F.3d at 269 (noting that objecting plaintiffs “had no evidence that the contributions, themselves legal, had influenced the [consortium’s] selection process”); see also In re Bank of New York Mellon Corp. Forex Transactions Litigation, 148 F. Supp. 3d 303 , 308–09 (S.D.N.Y. 2015) (acknowledging pay-to-play concerns but also recognizing that “no evidence” cast doubt on deputy attorney general’s assertion that campaign contributions did not affect selection of lead counsel). On this record, it was not an abuse of discretion for the district court to deny the requested discovery. Nor would we be inclined to reverse if the court had come out the other way or somewhere in-between. These issues are case- and fact-specific, and the district judge “is in the best position to decide the proper scope of discovery.” Scott v. Chuhak & Tecson, P.C., 725 F.3d 772, 785 (7th Cir. 2013) (citation omitted); see also Fields v. City of Chicago, 981 F.3d 534 , 550–51 (7th Cir. 2020) (“District court judges are accorded broad discretion in discovery matters, and therefore our review is deferential….”). Based on Petri’s evidence and allegations, we are not persuaded that the district court was required to order the requested discovery. No. 20-2055 31 IV. Motion for Sanctions We face one final issue: Petri has moved for sanctions against lead counsel based on remarks in its response brief about Petri’s attorney. Federal courts have inherent power “to fashion an appropriate sanction for conduct which abuses the judicial process.” Chambers v. NASCO, Inc., 501 U.S. 32 , 44–45 (1991). Lead counsel’s brief referred to Petri’s attorney, Theodore Frank, as a “notorious professional objector” and characterized his firm as an “objection-factory.” We have previously disapproved such rhetoric. See Pearson v. Target Corp., 968 F.3d 827 , 831 n.1 (7th Cir. 2020) (noting our avoidance of the phrase “professional objector” because “the merits of an objection are relevant, not amateurism or experience”). These attempts to use Frank’s past work to undermine his substantive arguments are improper and not at all persuasive. At this point, Frank’s track record—which now includes his success in this case—speaks for itself. 11 Lead counsel’s ad hominem attack on Frank was not professional and served only to emphasize the weakness of lead counsel’s own arguments. Still, the use of this language falls short of the type of conduct we have deemed sanctionable. See, e.g., McCurry v. Kenco Logistics Services, LLC, 942 F.3d 783 , 790–92 (7th Cir. 2019) (imposing sanctions where “patently 11 In our circuit alone, see, for example, In re Subway Footlong Sandwich Marketing & Sales Practices Litigation, 869 F.3d 551 (7th Cir. 2017); In re Walgreen Co. Stockholder Litigation, 832 F.3d 718 (7th Cir. 2016); Pearson v. NBTY, Inc., 772 F.3d 778 (7th Cir. 2014); Redman v. RadioShack Corp., 768 F.3d 622 (7th Cir. 2014); Robert F. Booth Trust v. Crowley, 687 F.3d 314 (7th Cir. 2012). 32 No. 20-2055 frivolous” appeal represented “a shameful waste of judicial resources” and counsel submitted “an overly long, borderline-unintelligible brief”). We exercise our discretion not to impose more formal sanctions in the still-optimistic hope that the rhetorical attacks might be de-escalated. But we reiterate what we said in Pearson: this kind of ad hominem criticism is unwarranted and counterproductive. * * * The attorney fee award is VACATED, and the case is REMANDED to the district court for recalculation. The denial of Petri’s motion for discovery is AFFIRMED. 12 12 Petri requests that thecase be reassigned under Circuit Rule 36, but we have no doubt that Judge Wood will handle the remand ably and fairly. The request is denied.