PER CURIAM:
Waiters and waitresses at Chili’s Restaurants sued them employer, claiming they were forced to share tips with ineligible employees. Fifty-five plaintiffs succeeded at trial. Their employer challenges the collective action, the district court’s rulings, the jury’s findings, and the award of attorney’s fees. We AFFIRM.
FACTUAL AND PROCEDURAL HISTORY
This is a collective action brought by waiters and waitresses (together, “servers”) of Chili’s Restaurants against their employer, Brinker International Payroll Company, L.P. The servers alleged that Brinker violated the Fair Labor Standards Act (“FLSA”) when its managers coerced servers to pay a portion of their tips to employees who were not entitled to tips.
One provision of the FLSA provides for collective actions:
An action to recover the liability prescribed in [this section] may be maintained against any employer (including a public agency) in any Federal or State court of competent jurisdiction by any one or more employees for and in behalf of himself or themselves and other employees similarly situated. No employee shall be a party plaintiff to any such action unless he gives his consent in writing to become such a party and such consent is filed in the court in which such action is brought.
29 U.S.C. § 216(b). As the last sentence just quoted provides, prospective claimants must opt-in under the FLSA, fundamentally distinguishing these suits from Rule 23 class actions in which a prospective plaintiff must opt-out.
Sandoz v. Cingular Wireless LLC,
553 F.3d 913, 916 (5th Cir.2008). Collective actions bind only the opt-in plaintiffs.
Id.
The FLSA permits employers to pay a sub-minimum wage to employees who “customarily and regularly receive tips.” 29 U.S.C. § 203(m). Employers may not avail themselves of the sub-minimum wage unless “all tips received by [a tipped] employee have been retained by the employee, except that this subsection shall not be construed to prohibit the pooling of tips among employees who customarily and regularly receive tips.”
Id.
The plaintiffs alleged that they did not retain all of their tips because they were forced to pool their tips with employees who did not customarily and regularly receive tips.
The allegedly tip-ineligible employees at the heart of this case are known as “expediters” or “Quality Assurance” workers (“QAs”). According to Brinker’s corporate documents and job descriptions, QAs at Chili’s inspect completed food orders from the kitchen, garnish plates, and delegate to servers and bussers the delivery of food to customers. A central question is whether QAs “customarily and regularly” receive tips or are otherwise entitled to share in a tip pool.
Brinker’s policy is that servers may voluntarily tip QAs. A company memo acknowledges that “QA positions (with little to no customer interaction), dishwashers, cooks, and janitors” were “occupations that would invalidate a tip pool.” The issue was instead whether Chili’s managers had a common practice of coercing tip-sharing.
Early in the litigation, the parties agreed to send an opt-in notice to all Chili’s servers nationwide. Approximately 3,556 servers opted-in as plaintiffs. During discovery, the parties deposed more than 50 plaintiffs and nearly 100 other persons, such as the deposed plaintiffs’ managers and coworkers. After discovery, and in a series of orders, the district court decertified the 3,556-person class, finding the question of coercion too individualized to be tried in one action. It retained 55 plaintiffs who had been deposed, though, concluding that their evidence showed they were similarly situated.
The plaintiffs proposed to call 14 plaintiffs to testify as representatives of the 55-person class. Before trial, Brinker conceded that these 14 plaintiffs had been coerced by their managers into tipping QAs. The jury trial was thus limited to Brinker’s affirmative defense, whether QAs were otherwise entitled to share in a mandatory tip pool. The parties stipulated to damages.
The jury answered two interrogatories. It found the 14 plaintiffs to be representative. It also found that Brinker had not proven “that QAs/Expos work in positions or an occupation that customarily and regularly receive tips.” The court accepted these findings, denied Brinker’s post-trial motions, and awarded attorney’s fees. Brinker timely appealed.
DISCUSSION
I. Decertification
Brinker argues that the district court should have decertified the 55-per-son collective action. It claims there was no common evidence — no “single, uniform, nationwide policy or practice of coerced tip-sharing” — because each coercive act was carried out by a “rogue manager” at a different restaurant. A similar argument is made as to Brinker’s affirmative defense, which turns on a QA’s individualized job duties and the amount of time spent performing those duties. Brinker asserts the dominance of individual situations means the collective trial was error.
The court also erred, Brinker argues, in going to trial without approving any of the plaintiffs’ proposed trial plans.
Like several other circuits, this court has never set a legal standard for collective-action certification.
Mooney v. Ararnco Servs. Co.,
54 F.3d 1207, 1216 (5th Cir.1995),
overruled on other grounds by Desert Palace, Inc. v. Costa,
539 U.S. 90, 123 S.Ct. 2148, 156 L.Ed.2d 84 (2003);
see Morgan v. Family Dollar Stores, Inc.,
551 F.3d 1233, 1259-60 & n. 38 (11th Cir.2008) (collecting cases). We have declined to choose between two standards, one involving a multi-factor “similarly situated” test, and the other akin to the standard for Rule 23 class actions.
Mooney,
54 F.3d at 1213-14
&
n. 7. The district court in this case applied the “similarly situated” analysis. It considered: “(1) the disparate factual and employment settings of the individual plaintiffs; (2) the various defenses available to defendant which appear to be individual to each plaintiff; and (3) fairness and procedural considerations.”
See id.
at 1213 n. 7. Brinker challenges only the court’s application of this standard. We review for abuse of discretion.
Id.
at 1212.
Brinker’s arguments for decertification were persuasive enough to remove approximately 3,500 plaintiffs from the class. They are unpersuasive as to the final 55, though, because evidence from each showed that they were similarly situated. All were subjected to some form of managerial coercion in tipping QAs. Their managers had directed them to tip QAs, collected tips for QAs, suggested tipping QAs, threatened to discipline them for not tipping QAs, or trained them to tip QAs, among other actions. Although there was no corporate policy mandating tip-sharing, the district court found the deposition testimony indicative of a pattern. This conclusion was not an abuse of discretion.
There also was evidence that the QAs were similarly situated. The parties agreed that QAs consistently performed certain functions, and that there was one company-wide job description. Several plaintiffs and some of Brinker’s witnesses agreed that “QAs never leave the pass-through area or interact with customers only rarely.” Corporate memoranda and documents also painted a picture of QAs as monolithic. Brinker “exaggerates the factual differences among employees on various shifts and in different departments. If one zooms in close enough on anything, differences will abound; ... [b]ut plaintiffs’ claims need to be considered at a higher level of abstraction.”
Frank v. Gold’n Plump Poultry, Inc.,
No. 04-CV-1018, 2007 WL 2780504, at *4 (D.Minn. 2007). There is need for care in evaluating distinctions among employees, but
those distinctions must make a difference relevant to the legal issues presented.
We would give Brinker’s arguments more credence if this case had proceeded on a truly representative basis. Instead, all 55 plaintiffs presented individualized evidence through testimony to the jury or deposition excerpts to the court. The plaintiffs claim Brinker was allowed to call witnesses to rebut the evidence as to all 55 plaintiffs, ensuring that the plaintiffs could not bury bad facts with the non-testifying plaintiffs. They further assert that after both sides presented their cases, the jury had heard evidence about 25 of the 55 plaintiffs, or 45 percent, and the court considered evidence on the remainder.
In any event, 55 separate trials would not have been justifiable under our FLSA caselaw. Section 216(b) collective actions are intended “to avoid multiple lawsuits where numerous employees have allegedly been harmed by a claimed violation or violations of the FLSA by a particular employer.”
Sandoz,
553 F.3d at 919
(quoting Prickett v. DeKalb Cnty.,
349 F.3d 1294, 1297 (11th Cir.2003)).
This suit validly proceeded as a collective action.
II. Representative Testimony
Brinker argues that the court erred by allowing the plaintiffs to handpick 14 “perfect plaintiffs” to testify, as opposed to a random sample. It contends that these 14 plaintiffs were unrepresentative and could not be imputed to the 41 non-testifying plaintiffs.
The district court had to develop a plan for this complex case. Our review of the plan itself is deferential because “the trial judge is in a much better position than an appellate court to formulate an appropriate methodology for a trial.”
In re Chevron U.S.A., Inc.,
109 F.3d 1016, 1018 (5th Cir.1997). Any arguable “error is presumed harmless until shown to be prejudicial.”
McClain v. Lufkin Indus., Inc.,
519 F.3d 264, 282 (5th Cir.2008) (citations omitted).
Before trial, Brinker requested representative testimony. The essence of its argument now is that the 14 testifying plaintiffs presented an unrepresentative slice of the evidence. But Brinker does not explain how the non-testifying plaintiffs would have been different, even though it deposed them and witnesses from all of their restaurants.
The plaintiffs assert that the jury heard evidence regarding a total of 25 plaintiffs. Evidence was presented to the court via deposition excerpts on the remaining 30. Thus, there was no
In re Chevron
problem of having a few handpicked “perfect plaintiffs” speculate as to several thousand unknown persons. A jury hearing evidence on 25 of 55 plaintiffs (45 percent) was sufficiently informed.
Brinker also does not state what the district court should have done. Within a few sentences in its appellate brief, Brink-er endorses “random selection using competent scientific or statistical analysis,” then argues that statistical sampling would not have been appropriate here because each server’s experience was different. We note that Brinker originally opposed statistical sampling.
There was no reversible error as to this issue.
III. Evidence on Coercion
Brinker argues that its due process rights were violated when it was prevented from presenting evidence to the jury that the non-testifying plaintiffs were not coerced to share tips. Although Brinker conceded that the 14 testifying plaintiffs had been coerced, it never conceded that
the 41 non-testifying plaintiffs had been coerced. Brinker cites its proffer of evidence on non-coercion, and specifically the deposition of Ted Davis, a non-testifying plaintiff who stated that his managers had never told him or encouraged him to tip QAs.
This court reviews evidentiary rulings for abuse of discretion.
Lyondell Chem. Co. v. Occidental Chem. Corp.,
608 F.3d 284, 295 (5th Cir.2010);
see Versai Mgmt. Corp. v. Clarendon Am. Ins. Co.,
597 F.3d 729, 740 (5th Cir.2010) (same for district court’s enforcement of a pretrial order).
Before trial, Brinker conceded that the 14 testifying plaintiffs were “coerced” under the district court’s definition. It “concede[d] for purposes of trial that the fourteen opt-in plaintiffs designated to testify at trial were not ‘free from any coercion whatsoever.’ ” Brinker acknowledged that the “principal issue remaining for trial” would be its “QA eligibility defense.”
Brinker did not concede that the other 41 plaintiffs had been coerced. The meaning of its reservation is unclear. By that point, the court had already ruled that the trial would proceed by representative proof. If the 14 representative plaintiffs did not need to present evidence of coercion at trial, it is not clear how the remaining 41 plaintiffs would get such evidence before the finder of fact.
A week later, Brinker filed a motion
in limine
seeking the following:
Defendant requests that the Court enter an Order to exclude and bar named-plaintiff Jennifer Roussell and all opt-in plaintiffs who are participating in this trial, as well as their counsel and all witnesses called on their behalf, from making any mention of, or conducting any interrogation regarding, or otherwise bringing before the jury, directly or indirectly, through documentary or testimonial evidence or otherwise, matters concerning Defendant coercing or forcing those Plaintiffs to share tips with QAs or involuntarily to “tip out” QAs, as that issue would no longer be part of this case.
The language is vague on whether the emotion applied to non-testifying plaintiffs. It sought to bar plaintiffs’ counsel from presenting coercion evidence as to the 14 testifying plaintiffs but not the remaining 41. That result would defeat the purpose of proceeding on a representative basis. The motion claimed there was “one issue remaining to be tried-whether Defendant’s QAs were eligible to participate in tip pools.” The motion was granted.
Immediately after seating the jury, the court and the parties agreed that Brinker could make an offer of proof outside the presence of the jury regarding the non-testifying plaintiffs’ non-coercion. The purpose was apparently to allow the court to decide whether the 14 plaintiffs were representative of all 55 plaintiffs. During trial, though, the court called for another discussion, finding misapprehension between the parties; both were objecting to each other’s attempted presentation of coercion evidence. Brinker once again confirmed that “the only issue for the jury was going to be the QA issue.”
Brinker’s argument on appeal is that the court erred by preventing evidence of non-coercion from going to the jury. But the record reveals that Brinker repeatedly sought to present this evidence outside the presence of the jury. Brinker cannot now claim this was error.
IV. Brinker’s Concession
Brinker argues that the court erred in extrapolating its concession to the 41 non-testifying plaintiffs. “There was no evidentiary basis to support a class-wide coercion finding, as the testifying opt-ins
admitted that they had no knowledge of tip-sharing outside of their own restaurants.” It asserts that the district court ruled on coercion as a matter of law, when it should have been a factual matter for the jury. Brinker contends that the court’s ruling relieved the plaintiffs from proving “both individual causation and class-wide liability.”
This court reviews “the district court’s findings of fact for clear error, and conclusions of law and mixed questions of law and fact de novo.”
French v. Allstate Indent,. Co.,
637 F.3d 571, 577 (5th Cir.2011).
The district court was in an unenviable position. Brinker did not want to try the case with all 55 plaintiffs, but objected to using representative testimony. Brinker conceded coercion of the 14 testifying plaintiffs, but that left no avenue for the 41 non-testifying plaintiffs to prove coercion through live testimony. The finding was reasonably left to the trial court.
The court’s resolution of this dilemma contains both factual and legal determinations. As a factual matter, it found Brink-er’s proffer of non-coercion “insufficiently persuasive” to warrant reversing its previous ruling that all 55 plaintiffs were similarly situated. As a legal matter, it found Brinker’s arguments for decertification unconvincing and kept the case as a collective action. It also upheld the jury’s factual findings that the 14 testifying plaintiffs were representative and that QAs were ineligible to join a mandatory tip pool.
The court did not err. As we have described, evidence from all 55 plaintiffs was presented through their depositions, and controverting evidence from other witnesses (coworkers or managers) at their respective restaurants was also presented. The district court sifted through the depositions and affirmed its previous decision that there was a pattern of managerial coercion at Chili’s.
Brinker’s argument that the court made a factual finding “as a matter of law” is illusory. The record reveals that Brinker asked the court to rule as a matter of law. Brinker cannot complain of an error, if any, that it invited.
V. QA Tip Eligibility
Brinker claims the district court erred when it denied Brinker’s motions for judgment as a matter of law and a new trial on the issue of QA tip eligibility. It argues that its witnesses provided uncontested evidence that some QAs “performed customer service duties sufficient to meet or exceed any reasonable legal requirement to allow them to participate in mandatory tip pools.” This is a sufficiency-of-the-evidence challenge to the jury’s finding that QAs were not eligible to participate in mandatory tip pools.
“A motion for judgment as a matter of law ... in an action tried by jury is a challenge to the legal sufficiency of the evidence supporting the jury’s verdict.”
Allstate Ins. Co. v. Receivable Fin. Co.,
501 F.3d 398, 405 (5th Cir.2007) (quotation marks and citation omitted). The motion should be granted “only when the facts and inferences point so strongly in favor of the movant that a rational jury could not reach a contrary verdict.”
Id.
(quotation marks and citation omitted). This court reviews the denial of the motion
de novo,
considering all the evidence and drawing reasonable inferences in the light most favorable to the verdict.
Id.
We are unconvinced that Brinker has met the high standard required of sufficiency challenges. The jury could have concluded that the witnesses Brinker cites in its brief were outliers or unbelievable. Accepting Brinker’s recitation of the testimony, even if some QAs interacted with customers, they still might not “customari
ly and regularly” receive tips. 29 U.S.C. § 203(m). We find evidence jurors could accept to support that QAs could not participate in a mandatory tip pool.
Brinker had the burden to prove it operated a legal tip pool, but failed to convince the jury of QAs’ eligibility. Viewing the evidence in the light most favorable to the verdict, Brinker’s insufficiency argument fails.
VI. The District Court’s Legal Interpretations
A. Defining “coercion” and “voluntary.”
Brinker argues that the district court’s definition of “coercion” was overly broad. It asserts that the court allowed servers to prove coercion by showing any managerial suggestion, recommendation, facilitation, or implied remark, which it says permitted them to avoid causation. Brinker seeks an objective definition, specifically, whether a reasonable server could show that a manager “dissuaded him, by force or threat, from believing that tipping out a QA was voluntary.” The plaintiffs reframe the dispute as one about the word “voluntary,” and claim the court correctly defined “voluntary” as “free from any coercion whatever and outside of any formalized arrangement or as a condition of employment.”
We review a district court’s legal rulings
de novo. French,
637 F.3d at 577. The FLSA does not define either “coercion” or “voluntary.” The district court relied upon the Department of Labor’s “Field Operations Handbook:”
[I]t does not appear that the Congress, even in requiring as a general principle that tipped employees retain all their tips, intended to prevent tipped employees from deciding, free from any coercion whatever and outside of any formalized arrangement or as a condition of employment, what to do with their tips, including sharing them with whichever co-workers they please.
Department of Labor Field Operations Handbook § 30d04(c) (Dec. 9, 1988),
available at
http://www.dol.gov/whd/FOH/ FOH_Ch30.pdf.
The court found the Handbook persuasive and held that servers may share tips as long as their choice is “free from any coercion whatever and outside any formalized arrangement or as a condition of employment.” It agreed with Brinker that this was an objective inquiry as in Title VII; only a manager’s actions that “might well” affect the server were actionable.
See Burlington N. & Santa Fe Ry. Co. v. White,
548 U.S. 53, 69, 126 S.Ct. 2405, 165 L.Ed.2d 345 (2006).
As for specific examples of coercive actions — where the server “might well” be influenced to pay tips involuntarily — the court looked to Brinker’s internal memos. Brinker’s “General Guidelines for Tip Pooling” state that “[a] practice is voluntary when the employee has full control over his/her tips without any recommendations, suggestions, requirements, or implications by management or trainers.” The document used similar terms to describe when tip-pooling would be deemed “mandatory.”
“[W]e construe the FLSA liberally in favor of employees....”
Allen v. McWane, Inc.,
593 F.3d 449, 452 (5th Cir.2010) (quotation marks and citation omitted). Applying that construction to the FLSA, the district court did not err in relying upon the Handbook language. In this case in particular, it is relevant that the court adopted Brinker’s own definitions for when a manager’s actions would make tipping “mandatory” and not “voluntary.”
B. Direct customer interaction.
Brinker next argues that the court erred by finding “the extent of [an employee’s] direct customer interaction” relevant to whether she can share in tips. It contends that “employees who perform important customer service functions are eligible to share tips regardless of whether they have direct customer interaction or not.” We review
de novo
this legal question.
French,
637 F.3d at 577.
The FLSA does not specify which employees may share in a tip pool; it merely authorizes tip-pooling “among employees who customarily and regularly receive tips.” 29 U.S.C. § 203(m). Customarily, front-of-the-house staff like servers and bartenders receive tips. Back-of-the-house staff like cooks and dishwashers do not, and thus cannot participate in a mandatory tip pool.
Direct customer interaction is relevant because it is one of the factors distinguishing these two categories of workers. “One can distinguish hosts from restaurant employees like dishwashers, cooks, or off-hour employees like an overnight janitor who do not directly relate with customers at all.”
Myers v. Copper Cellar Corp.,
192 F.3d 546, 550 (6th Cir.1999) (holding that salad preparers were not tip-eligible) (citation omitted).
We conclude that the district court reasonably found direct customer interaction “highly relevant” to tip-eligibility. It neither made direct interaction a prerequs-ite — as claimed by
amid
— nor did it reject Brinker’s suggestion that the performance of “important customer service functions” also is relevant to tip-eligibility. Brinker has not advanced a compelling reason why direct customer interaction should not be considered.
VII. Answering a Jury Question
Brinker contends that the district court erroneously responded to a jury question. We review the district court’s answering of a jury question for abuse of discretion, granting the court “wide latitude” in framing an answer.
United States v. Skelton,
514 F.3d 433, 446 (5th Cir.2008).
Before trial, the court found industry practice could “provide some evidence of the restaurant industry’s conclusion as to whether [QAs’] job duties are more like servers, busboys, or service bartenders or instead are more like cooks, dishwashers, and laundry room attendants.” The district court permitted testimony about similar positions at other restaurants. Plaintiff Roussell testified, over Brinker’s objection, that QAs at Applebee’s “did essentially the same things” as QAs at Chili’s. Other plaintiffs testified similarly.
After retiring to deliberate, the jury asked the court whether “the ‘occupation’ that regular[ly] and customarily receives tips [is] specific to Chili’s (Brinker) or this or similar positions in other restaurants?” After a conference with the attorneys, the court gave an answer that included this:
The occupation/positions of QA/Expo is not necessarily unique to Chili’s (Brink-er). However, as you consider whether there was competent evidence and/or testimony presented in this case that relates to QA/Expo-type occupation/positions at other restaurants, you should factor into your consideration whether the QA/Expo-type occupation/positions in those other restaurants is similar or dissimilar-to the extent the evidence or testimony allows you to make that determination-to the QA/Expo occupation/positions at Chili’s (Brinker)....
The court’s careful answer was not an abuse of discretion. Jurors were told to evaluate the evidence themselves. There
was evidence the position of QA was common in the restaurant industry, even if there were differences in the tasks performed by QAs at other restaurants. The jury could consider Brinker’s impeachment of each witness who testified about other restaurants, and determine how important those differences were. This is not reversible error.
VIII. The ‘Related Duties ’ Defense
Brinker asserts that the court erred by rejecting its “related duties” defense during summary judgment. Brinker claimed that when a server occasionally worked an entire shift as a QA, the employee is a “server” for tipping purposes and may share in a mandatory tip pool. There was evidence that servers performed the work of a QA during the course of their server duties, and therefore, Brinker argues, “the duties of the two positions were sufficiently related such that a Server who took a turn working a shift as a QA could be deemed to remain in the occupation of Server.” It asserts that this was a factual issue for the jury.
Summary judgment rulings are reviewed
“de novo,
applying the same standard as the district court.”
Holt v. State Farm Fire & Cas. Co.,
627 F.3d 188, 191 (5th Cir.2010). “Summary judgment is appropriate if there is no genuine issue as to any material fact and the moving party is entitled to judgment as a matter of law. The facts and evidence must be taken in the light most favorable to the non-mov-ant.”
Id.
(citation omitted).
Federal regulations shed light on this issue:
(e) Dual jobs. In some situations an employee is employed in a dual job, as for example, where a maintenance man in a hotel also serves as a waiter.... He is employed in two occupations, and no tip credit can be taken for his hours of employment in his occupation of maintenance man. Such a situation is distinguishable from that of a waitress who spends part of her time cleaning and setting tables, toasting bread, making coffee and occasionally washing dishes or glasses.
29 C.F.R. § 531.56(e). We give the regulation deference.
See Fast v. Applebee’s Int'l Inc.,
638 F.3d 872, 877, 879-81 (8th Cir.2011) (granting both the regulation and the Department of Labor’s interpretations in its Field Operations Handbook deference, and holding the interpretations reasonable).
A Department of Labor Opinion Letter states that a dual-job employee may continue to receive tips when he performs tasks incidental to his usual job, but may not receive tips “where there is a clear dividing line between the types of duties [performed in the two jobs]....” U.S. Dep’t of Labor, Wage & Hour Div., Op. Ltr. WH-502, 1980 WL 141336 (Mar. 28, 1980).
Undisputed facts show that the servers clocked in as QAs for an entire shift, under the QA job code, performed QA duties only, and were paid the full minimum wage. These shifts were not assigned to all servers. We agree that there was a clear dividing line between these jobs.
The question remains whether the court’s determination that there is a “clear dividing line” between the two jobs was a legal conclusion or a factual determination inappropriate for summary judgment. The question initially appears to be a factual one. On the other hand, answering the question requires applying facts to a standard expressed in an Opinion Letter.
The issue likely is better considered as a mixed question of law and fact. As we held in a Jones Act case, though, “summary judgment or a directed verdict is mandated where the facts and the law will reasonably support only one conclusion. Our review of such a mixed question is
plenary.”
Roberts v. Cardinal Servs., Inc.,
266 F.3d 368, 373 (5th Cir.2001) (quotation marks and citations omitted). We will review this issue
de novo. Williams v. Phillips Petrol. Co.,
23 F.3d 930, 934 (5th Cir.1994).
The undisputed facts fit cleanly into the Department of Labor guidance. A server working a QA shift, making QA pay, and performing QA duties, is not spending “part of her time” on QA work — she is a QA for that shift. We agree that “even when the nontip-producing duties are related to a tipped occupation, if they are performed for an entire shift, the employee is not engaged in a tipped occupation and is not subject to the tip credit for that shift.”
Fast,
638 F.3d at 880. It follows that servers working an occasional QA shift are, just like QAs, ineligible to participate in mandatory tip pools during those shifts. Had this issue gone to the jury, “the facts and the law will reasonably support only one conclusion.”
Roberts,
266 F.3d at 373. Brinker’s inability to present this to the jury was harmless.
See
Fed. R.CivJP. 61.
IX. Attorney’s Fees
Brinker argues that the award for attorney’s fees was excessive. It contends that the district court failed to compare what the plaintiffs sought in their complaint with what they obtained at trial. Brinker asserts that the plaintiffs’ fees should be substantially reduced because they lost on certification and recovered a fraction of the money they thought the case was worth. Brinker does not challenge the award for costs or expenses.
“A district court’s determination of attorneys’ fees is reviewed for abuse of discretion, and the findings of fact supporting the award are reviewed for clear error.”
McClain,
519 F.3d at 284. “A district court abuses its discretion when it bases its decision on an erroneous legal conclusion or on a clearly erroneous finding of fact.”
Jeter v. Astrue,
622 F.3d 371, 376 (5th Cir.2010) (quotation marks and citations omitted).
While rulings on fee awards should not “consume more paper than did the cases from which they arose[,] ... the district court’s findings and reasons must be complete enough to assume a review which can determine whether the court has used proper factual criteria in exercising its discretion to fix just compensation.”
In re High Sulfur Content Gasoline Prods. Liab. Litig.,
517 F.3d 220, 228-29 (5th Cir.2008) (quotation marks and citations omitted). “The essential goal in shifting fees (to either party) is to do rough justice, not to achieve auditing perfection,” and therefore “substantial deference” is owed the district court’s “overall sense of a suit.”
Fox v. Vice,
— U.S. -, 131 S.Ct. 2205, 2216, 180 L.Ed.2d 45 (2011).
Applying our usual method, the district court first calculates the “lodestar” by multiplying the reasonable hourly rate by the number of hours reasonably expended.
McClain,
519 F.3d at 284. It then evaluates whether the lodestar should be adjusted based on the
Johnson
factors, the most important of which is the degree of success obtained.
Abner v. Kansas City So. Ry. Co.,
541 F.3d 372, 376-77 (5th Cir.2008) (citing
Johnson v. Ga. Highway Express,
488 F.2d 714, 717-19 (5th Cir.1974)). The only
Johnson
factor relevant to this appeal is “[t]he amount involved and the results obtained.”
Johnson,
488 F.2d at 718.
Reasonable attorney’s fees and costs are authorized by the FLSA. 29 U.S.C. § 216(b). The parties argued the fee dispute in two stages. The first resolved the fees and costs incurred through a portion of the post-trial motions; the second resolved everything afterward. Brinker preserved its objections by filing notices of appeal after each award was entered.
During the first round, the plaintiffs sought $1.81 million in attorney’s fees. Brinker challenged whether the claimed hours were reasonable and necessary given the success obtained. A hearing was held. In a detailed, 28-page opinion, the court sifted through the competing evidence on each part of the request, found that “both parties have taken rather extreme positions on” the degree of success factor, and reduced the plaintiffs’ bill in two ways.
The court first reduced by 209.2 hours the time recoverable for document review, then broke out that reduction by occupation. On the degree of success factor, although the large class was ultimately decertified, the court said it “was a very close call,” “was not apparent from the onset of the litigation,” and happened “incrementally, over the course of three separate orders, and after much discussion and briefing by the parties.” Therefore substantial hours were justified, as well as entangled with the work done on the 55 plaintiffs that went to trial. Given that the plaintiffs’ success was “not that which was originally sought,” the court reduced the lodestar amount by another 20 percent beyond the plaintiffs’ reductions and the court’s specific reduction. The award during this round was approximately $1.43 million.
During the second round of adjudication, the court awarded $90,000 for fees incurred after trial — much of which was spent arguing over fees.
The court took into account the plaintiffs’ lack of success before trial. It considered that the plaintiffs lost on class certification and reduced their fees accordingly. Second, the amount sought in a plaintiffs complaint is relevant to the degree of success obtained, but is ultimately only one of several factors to consider, in the district court’s “measured exercise of discretion.”
Farrar v. Hobby,
506 U.S. 103, 114, 113 S.Ct. 566, 121 L.Ed.2d 494 (1992).
It is important for the district court to explain its fee award with clarity, “carefully considering the successful and unsuccessful portions of the case as well as the reasonable and unreasonable fees claimed by Plaintiffs.”
Abner,
541 F.3d at 384 (quotation marks and citation omitted). The district court’s thorough analysis in the present case was sufficient. Its order displayed a detailed understanding of the successful and unsuccessful portions of this case, drawn from years of managing it first-hand. The district court also noted that Brinker never offered to settle the case; thus, the plaintiffs could not have reduced their costs to a level more appropriate with the results achieved.
The award was not an abuse of discretion.
AFFIRMED.