This month we look at part three of our three part series on Class Actions. In part three, Robert Rachal and M. Todd Mobley address the role of experts in class certification post Wal-Mart and Comcast and how to use and attack experts during the class certification phase of class action litigation. Please look for parts one and two of this series in our previous Newsletters.
As always, be sure to review the Rulings, Filings, and Settlement of Interest where we discuss the Supreme Court's decision to resolve a circuit split in health care reform cases, the Supreme Court's decision on the enforceability of plan limitations provisions, the availability of remedies under ERISA, IRS guidance on in-plan Roth rollovers, benefits for same-sex spouses and non-discrimination testing relief, and valid assignments of rights.
Labor and Employment and ERISA Class Actions After Wal-Mart and Comcast—Practice Points for Defendants (Part III – Experts)*
By Robert Rachal and M. Todd Mobley
Introduction and Overview Part III: Wal-Mart and Comcast and the Central Role of Experts in Class Certification
This is the final installment of a three-part Bloomberg BNA Insight article addressing the impact of Wal-Mart and Comcast on labor and employment and ERISA class actions (197 PBD, 10/10/13; 207 PBD, 10/25/13; 40 BPR 2427, 10/15/13; 40 BPR 2537, 10/29/13). This is a hotly contested and developing area—cases such as the U.S. Court of Appeals for the Fourth Circuit's recent ruling suggesting a class may be possible regarding management discretion indicate that the full meanings of Wal-Mart and of Comcast are still being developed. This part focuses on experts, and how defendants may be able to use experts to defeat or limit class certification.
Wal-Mart and Comcast offer a good place to begin a discussion on class certification and experts, since the failures of expert proof in those cases led to class decertification. In Wal-Mart, plaintiffs sought to prove commonality for their pay and promotion claims using two forms of expert proof: (i) a "social framework" analysis purporting to show that Wal-Mart had a corporate culture that made it susceptible to gender bias; and (ii) a statistical analysis that showed disparities in pay and promotions. In answering the central question as to "why was I disfavored," the Court found the social framework analysis of plaintiffs' expert useless for class purposes—it provided no "glue" to show that discretion was exercised in a common and discriminatory manner—since it could not answer whether 0.5 percent or 95 percent of the employment decisions at Wal-Mart were determined by stereotyped thinking on gender.
The Court also found plaintiffs' statistical evidence deficient. The Court noted that the presence of disparities at the national or regional level does not establish the existence of disparities—or discrimination—at the store level where the challenged decisions were made. The statistics' more fundamental flaw was that they failed to address possible sex-neutral reasons, including the relative availability of qualified and interested women at the store level, that could rebut any bottom-line disparities. Proof of bottom-line disparities does not answer the common question. Rather, the plaintiff must identify the particular employment practice causing the disparity and show that it caused the disparity through a common mode of acting.
In Comcast Corp. v. Behrend, the Court addressed the proof of harm and damages—which is almost always done through experts—required to certify a class. In an antitrust claim regarding a proposed class of cable subscribers, plaintiffs proffered four theories of antitrust injury that they argued drove-up cable subscription rates. The district judge found only one of these, the "deterrence of overbuilding" theory, capable of class-wide proof, and that the others could not be determined in a manner common to the class. Plaintiffs' economics expert admitted that he had not isolated the damages resulting from the different theories of antitrust impact, instead including the non-class theories in his model.
The Court concluded that this expert evidence failed to carry plaintiffs' burden of proof on Rule 23's requirements. Specifically, the Court found that plaintiffs failed to satisfy the predominance requirements of Rule 23(b)(3) because they could not show damages capable of class-wide proof. The Court held that the damages model must be consistent with the liability model—i.e., that any model purporting to serve as evidence of damages in a class action must measure only damages attributable to the class-wide theory of harm. In the view of the dissent, the expert evidence tendered was sufficient for class purposes since it purported to show that Comcast's conduct resulted in higher prices, even though it failed to show causation tied to the class theory of harm. The majority imposed a far more rigorous standard: Plaintiffs must prove that the claimed class-wrong caused the injury class-wide, free of taint from individual factors. Absent such proof, plaintiffs cannot satisfy Rule 23's requirement that common issues predominate for class claims seeking damages.
As detailed below, Wal-Mart and Comcast provide significant grounds to challenge expert opinions supporting class certification.
Going on Offense: Using and Attacking Experts in Class Certification
The facts necessary to establish—or disestablish—whether Rule 23 has been met typically require expert analysis and opinion, e.g., analysis of whether there are common issues or of whether everyone in the class has a common interest or suffered a common injury. These are not facts typically found in the record, and rulings like Wal-Mart and Comcast illustrate how expert issues can affect class certification.
To prohibit abuse and enhance evidentiary reliability, there are a host of rules that control and limit expert evidence. In the class stage, an initial issue is whether class experts are subject to Daubert and the attendant reliability requirements imposed on expert evidence. To meet Daubert's reliability requirements, the expert must show that his testimony (i) is based on sufficient facts or data, (ii) is the product of reliable principles and methods, and (iii) that the expert has applied the principles and methods reliably to the facts at hand. In a not-too-distant era of "certify first, ask questions later," many courts declined to require that class experts meet the reliability standards imposed by Daubert. These days should be past; now class expert evidence must be not merely admissible but persuasive to pass class muster.
On defense strategy, class certification is typically procedurally advantageous to defendants. Unlike in summary judgment, at class certification plaintiffs bear the burden of proof. Further, class certification is often the first opportunity for a defendant to put on facts supporting its case and to show the defects in a plaintiff's class claims. Defendants can challenge plaintiffs' experts, and defendants can also put on their own experts to show the defects in plaintiffs' expert-analyses or to develop expert evidence showing that the Rule 23 requirements have not been met. Potential grounds to challenge experts and the Rule 23 requirements are discussed throughout this three-part Bloomberg BNA Insight article; some key points include:
On discrimination claims, are decisions made at the local store, office, or facility level? Does plaintiff's expert bundle up or "average out" the statistics? Conversely, can a defense expert show variability between the store, office, or facility on the challenged criteria?
On discrimination claims, are decisions made at multiple levels, with multiple actors and inputs? Does plaintiff's expert use a "bottom line" analysis that does not account for or break out the steps in the process? Conversely,
can a defense expert show the importance of the steps in the process, and any variability on the challenged criteria by the different actors and steps in the process?
On discrimination claims, has plaintiff's expert accounted for employee choice and interest? Conversely, can a defense expert show that there is not homogeneous interest in or qualifications for the job positions at issue?
On ERISA (or discrimination claims) have some in the proposed class benefitted from the challenged conduct? Can a defense expert statistically analyze and show these differences?
It sometimes may be worthwhile to file Daubert motions challenging the admissibility of class experts. Even if the motion is not granted, it can show flaws in the expert's analysis that undercut its persuasive value.
Consider using defense experts not just to show flaws in the analyses of plaintiffs' experts, but also to affirmatively show why class certification requirements have not been met.
Expert Issues Arising in Employer Discrimination Class Claims
Because they may rely on invalid or questionable assumptions, plaintiffs' class experts in discrimination claims are often ripe for challenge. Specifically, it is common for the plaintiff's class expert to assume homogeneity so as to infer causation and discrimination; for example, by assuming that everyone in the proposed class has the same job qualifications or interests or that the challenged decisions were made by the same decision-maker. But as explained in the Federal Judicial Center's Reference Manual on Scientific Evidence, if the data is not homogeneous, the statistical analysis combining that data is irrelevant and often misleading. Thus, as Wal-Mart noted, when discretionary decisions are made by different decision-makers, "demonstrating the invalidity of one manager's use of discretion will do nothing to demonstrate the invalidity of another's." Likewise, as illustrated in the Reference Manual on Scientific Evidence, if females are disproportionately applying to a position with a lower acceptance rate, those statistics cannot be reliably combined with other statistics to show gender discrimination. Similarly, association does not necessarily show causation, as there may be confounding variables that cause or substantially affect the challenged disparities. Statistics seeking to show disparities are also invalid if they fail to consider those who are similarly situated for the challenged action.
Expert issues worth noting at the class-certification stage of employment discrimination actions may include:
Failure to focus on the appropriate unit or level of analysis; and
Failure to account for confounding variables like individual choice or the interests of putative class members.
"Soft science experts," like survey and other social science experts, may also be challenged as unreliable or unpersuasive in showing that class requirements, such as commonality, have been met. Wal-Mart illustrated this, rejecting a sociological expert who could not specify with any degree of precision how often gender stereotyping affected the managers' decisions, and, thus, his testimony was irrelevant since it provided no "glue" to show there was a common mode of exercising discretion. Experts who purport to do surveys, or the like, to extrapolate to class damages also may be interdicted by Wal-Mart's bar on "Trial by Formula." Finally, it is important for defendants to consider using experts to undermine plaintiffs' statistical or other expert analyses and to develop alternative analyses showing, for example, that commonality, typicality, or adequacy are not met.
Failure to Focus on the Appropriate Unit or Level of Analysis
In Wal-Mart, commonality for plaintiffs' claims required showing that discretionary employment decisions were being made uniformly by individual store managers throughout all of Wal-Mart's 3,400 stores. In attempting to make such a showing, plaintiffs presented statistical evidence based on aggregated data collected from the regional and national corporate levels. As discussed above, the Court found this evidence insufficient, because disparities between men and women at the national or regional level could not establish the existence of the same disparities at the individual store level. The Court held that "[m]erely showing that Wal-Mart's policy of discretion has produced an overall sex-based disparity does not suffice." The Court explained that when claims are based on discretionary employment decisions, proof of discrimination and commonality must account for the actions of the actual decision-maker.
In line with Wal-Mart, lower courts have denied class certification when claims regarding discretionary employment decisions rely on statistical analyses that aggregate data across decision-makers. Indeed, in Bolden v. Walsh Construction Co. the Seventh Circuit summarized the post Wal-Mart world by explaining that "local discretion cannot support a company-wide class no matter how cleverly lawyers may try to repackage local variability as uniformity." In Bolden, plaintiffs (construction workers formerly employed by defendant) alleged that defendant's construction-site superintendents exercised their discretion in a racially discriminatory fashion over assigning overtime hours and managing on-site working conditions. Plaintiffs supported their allegations with statistical evidence based on data aggregated from all Chicago-area sites. The court, however, found that the analysis failed to focus on the "appropriate unit of analysis." The court explained that if defendant "had 25 superintendents, 5 of whom discriminated in awarding overtime, aggregate data would show that black workers did worse than white workers—but that result would not imply that all 25 superintendents behaved similarly, so it would not demonstrate commonality."
The Third Circuit encountered the issue of an expert's failure to focus on the appropriate unit of analysis in a slightly different context in Rodriguez v. National City Bank. In Rodriguez, the district court denied certification of the settlement class because, in Wal-Mart's wake, plaintiffs' regression analyses could not establish commonality and typicality. Plaintiffs' class action alleged that, because of defendant's discretionary pricing policy on certain mortgage fees, black and Latino borrowers received disproportionately higher non-risk-related charges than similarly-situated white borrowers. The district court found that statistical evidence of an overall race-based disparity was insufficient to establish commonality and typicality; rather, plaintiffs would need "to show the disparate impact and analysis for each loan officer or at a minimum each group of loan officers working for a specific supervisor[.]"  The Third Circuit agreed and explained that "the exercise of broad discretion by an untold number of unique decision-makers … undermines the attempt to claim, on the basis of statistics alone, that the decisions are bound together by a common discriminatory mode."
Failure to Account for Confounding Variables like Individual Choice of the Putative Class Members
Statistical analyses may also be vulnerable to attack for failure to account for the individual choice or interests of putative class members. When defendants are able to put forth evidence suggesting proposed class members may not have homogeneous interests, choices or qualifications, courts have rejected plaintiffs' attempts to gloss over such variables and make assumptions regarding the homogeneity (and, in turn, commonality and typicality among the members) of the proposed class.
For example, in York v. Starbucks Corp., a wage and hour case involving allegations that Starbuck's corporate policies incentivized managers to "shortchange" workers of their rights under California labor law, the court took issue with plaintiffs' statistical evidence for failure to take individual choice into account. The court explained that, with regard to lunch and rest breaks, "the statistical evidence cannot begin to show whether a break was skipped because a manager forbade the employee from taking it or whether it was not taken as a matter of individual choice." Accordingly, the court found that the statistical evidence failed to demonstrate what might have caused the alleged labor violations—corporate policy, or individual choices and desires.
Similarly, in Puffer v. Allstate Ins. Co., plaintiff commenced a Title VII sex-discrimination action against Allstate on behalf of all female Allstate employees in management positions. In an effort to establish commonality, plaintiff proffered a statistical analysis of Allstate's personnel data, which provided information regarding employment records and compensation. Plaintiff argued that the analysis identified disparities in job assignments, promotions, and levels of salary paid. However, the court disagreed, taking issue with the fact that plaintiff's expert failed to consider, among other things, "whether putative class members were interested in the management jobs that [the expert] found were underrepresented by women … ." The court noted that although it is not necessary for experts to include all possible measurable variables in their analyses, crucial variables may not be omitted or glossed over—and in Title VII cases particularly, such crucial variables include the identification of those who were qualified and interested in the position.
Failure to account for individual choices and interests—at least when the evidence suggests that this is a meaningful issue—also undermines statistical analyses by infecting the data with claims of putative class members who suffered no class harm. Such failures have led courts to reject employment-related statistical analyses as unreliable on proof of liability. This also may have added force on proof of class damages. In Comcast, the Supreme Court faced analogous issues in rejecting plaintiffs' expert's damages model; this was so because "at the class certification stage (as at trial), any model supporting a plaintiff's damages case must be consistent with its liability case … ." Although Comcast was not an employment case, its principle that class damages must be limited to the class harm uninfected by individual issues is a broad one. The same principle applies when plaintiff's statistical analyses or proofs of class harm fail to account for individual choices and interests in the employment context—the analyses become over-inclusive and incapable of reflecting the class harm.
Defendants' Use of Experts to Defeat Class Certification
As the cases above illustrate, challenging plaintiff's proffered statistical experts is often effective to defeat or limit the class. Notably, this strategy need not be strictly defensive, as defendants may also affirmatively employ expert testimony to defeat class certification.
For example, in Serrano v. Cintas Corp., the court found that defendants' statistical experts effectively undermined the statistical evidence proffered by plaintiffs, preventing plaintiffs from establishing a class-wide discriminatory impact. In Serrano, plaintiffs alleged that defendant engaged in race and gender discrimination when hiring for sales positions, and they proffered statistical experts to establish commonality. Defendant, in turn, presented statistical experts who demonstrated that although some of defendant's store locations under-hired women and racial minorities, other locations over-hired members of these groups. Further, defendants' expert showed that although some locations under-hired women, those same locations over-hired racial minorities. Indeed, some locations under-hired one or both groups one year and over-hired the same group or groups the following year. Defendant's expert thus provided evidentiary support for the lack-of-commonality issue flagged in Wal-Mart—that a court can neither assume that there was a "common mode of acting" nor establish commonality by aggregating data across decision-makers and facilities.
In Randall v. Rolls-Royce Corp., the district court denied class certification because, among other things, defendants' statistical expert demonstrated a lack of commonality among putative class members. In a gender discrimination claim over compensation, defendants' expert showed that plaintiffs' expert failed to account for important variables in compensation. The Rolls-Royce defendants also used statistical analyses to demonstrate that plaintiffs failed to satisfy Rule 23's typicality requirement. Here, defendants' expert showed that the named plaintiffs actually earned salaries equal to or greater than their male comparators during the class period. Thus, the court found "it is plainly true that the particular circumstances surrounding the named Plaintiffs' individual claims do not comport with the required element of typicality."
Closely examine statistical analyses to determine whether the expert utilized data based or focused on an inappropriate unit of analysis, e.g., regional or national data as opposed to local data.
Closely examine statistical analyses to determine whether statistical experts have considered factors such as individual choices or interests. Investigate whether there are other potential confounding variables lurking in the data and analysis.
Investigate whether there is reliability and "fit" to opinions offered by social science or survey experts.
Consider putting forth affirmative expert evidence showing lack of class commonality, typicality or class conflicts.
Expert Issues Arising in ERISA Class Claims
ERISA claims involving disclosures (ERISA imposes statutory and fiduciary duties on disclosing information to participants) or investments in 401(k) plans often raise class issues for experts. For disclosure claims, the assumption of homogenous understandings may be inaccurate. Many ERISA claims involve multiple disclosures over extended periods, each of which can raise issues about what the various class members knew, relied upon, or understood.
For example, in Groussman v. Motorola, Inc., the court denied class certification for former participants in a 401(k) plan offered by Motorola. Plaintiffs alleged that defendants imprudently managed the 401(k) plan, including by making misrepresentations related to their investments in Motorola's stock. The court found that plaintiffs could not satisfy Rule 23's typicality requirement because, among other reasons, they failed to identify the specific misrepresentations that were made and relied upon. In contrast, defendants showed that there was "a difference as to what each [p]laintiff understood at any given time, and that [p]laintiffs did not rely upon the same information or statements in making their investment decisions." Communications or statistical experts also may be coupled with this anecdotal evidence to defeat class; for example, a statistical analysis of actions taken may show that plaintiff's claim of uniform homogenous understandings and reliance is implausible.
Experts thus may prove useful when claims involve different communications made to different members of a proposed class—even when objective standards apply to whether a defendant is liable for those communications. In Luiken v. Domino's Pizza LLC, the issue was whether a delivery service charge imposed by defendant for pizza deliveries was a gratuity under a state statute. The court held that even though the statute used an objective standard to determine whether a service charge is a gratuity, the context of the communications as to each customer still mattered. Thus, the court found that there was no commonality under Wal-Mart, because there was no common answer as to what each customer was told and understood regarding their service charge payments. In this context communication experts could buttress this lack of commonality by showing how understandings vary based on the different communications made.
Experts also can be central to class claims challenging ERISA investments involving 401(k) or other participant-directed individual account plans. For example, experts may be able to show that the class's interests are not homogenous. This proved effective in Langbecker v. Electronic Data Systems Corp., where plaintiffs claimed that the defendants breached their fiduciary duties by offering the company's stock in the 401(k) plan. Defendants' expert showed that thousands of class members profited from this investment, and thousands more continued to invest in the company's stock after the plaintiffs claimed it should be eliminated. Moreover, the expert showed that putative class members had different interests on which theory of the case would maximize their claims and recovery. The court found that the expert's analysis defeated Rule 23's adequacy requirement; thus, the class was decertified.
Wal-Mart has reinvigorated this type of conflict analysis. In Groussman, defendants attacked the commonality of the proposed class by showing that the challenged investments had varying effects based on the putative class members' individual trading patterns. The court held this defeated commonality, finding it "likely that a portion of the proposed class members actually acquired a net gain during the class period as a result of the investments in Motorola stock." Similarly, in Spano v. Boeing Co., an action challenging defendant's 401(k) plan investment options and fees, the Seventh Circuit found that the named plaintiffs could not satisfy Rule 23's adequacy requirement, because the class was so broadly defined that many putative members suffered no harm or may have even profited from the challenged investment options. These class conflicts and lack of commonality can arise in multiple contexts, such as for claims based on fund withdrawal restrictions imposed in response to financial disruption caused by the Great Recession.
These intra-class conflict issues are also not limited to actions challenging investments in 401(k) or other participant-directed plans. For example, these issues may arise in actions challenging plan investments for closely held ESOPs. In Hans v. Tharaldson the court found current and former employees of an ESOP-owned company had divergent interests regarding the class litigation because the litigation risked harm to the company's value and sales prospects, thereby harming the ESOP investments of the current employees.
If the class claims challenge plan investments, consider using an expert to analyze whether there is a lack of commonality, typicality or even possible conflicting interests and actions regarding the challenged investments.
If the class claims challenge disclosures, consider using experts to analyze whether there were common, uniform understandings and actions in relation to the challenged disclosures.
Rulings, Filings and Settlements of Interest
Supreme Court to Resolve Circuit Split in Health Care Reform Cases
By Peter Marathas, Robert Rachal and Brian Neulander
The Supreme Court will review two of the numerous lawsuits challenging the Affordable Care Act's (ACA) requirement that group health plans and insurers cover, without cost-sharing, contraceptives and/or abortifacients (the "Contraceptive Mandate"). The plaintiffs in these suits are secular, for-profit corporations and their owners, and they assert that being forced to comply with the Contraceptive Mandate would violate their First Amendment religious rights and would also violate the Religious Freedoms Restoration Act ("RFRA").
All courts addressing the various Contraceptive Mandate suits have struggled with the issue of whether secular, for-profit corporations are covered under either the First Amendment or RFRA (which generally prohibits federal law from imposing a burden on any "person's" religious freedom). This year, a circuit split developed: the Seventh, Tenth, and D.C. Circuits adopted a "pass through" theory that allowed the corporations to assert the free exercise rights of their owners, and held that the Contraceptive Mandate places a substantial burden on the plaintiffs' religious freedoms. The Third and Sixth Circuits, in contrast, have rejected the argument that secular, for-profit corporations can exercise religion, and have held that the owners are not burdened since it is the corporation, not the owners, who would be funding this coverage.
The Supreme Court is likely to schedule oral argument for March, and a ruling is expected in June. Because the Court will address the religious rights of for-profit corporations, the ruling may have significance beyond the Affordable Care Act.
Valid Assignment Confers Beneficiary Status on Chiropractor
By Todd Mobley
A federal district court in Illinois recently issued a pair of rulings in cases where insurers sought to recoup payments from practitioners. Pennsylvania Chiropractic Association v. Blue Cross Blue Shield Association, 2013 U.S. Dist. LEXIS 159331 (N.D. Ill. Nov. 7, 2013); 2013 U.S. Dist. LEXIS 159491 (N.D. Ill. Nov. 7, 2013). Plaintiffs, three individual chiropractors and three chiropractic associations, alleged that defendants, Blue Cross Blue Shield of America and individual Blue Cross Blue Shield entities (collectively, "BCBS"), falsely or fraudulently determined that certain reimbursements they received for services rendered should never have been made. When plaintiffs refused BCBS's demands to repay the challenged reimbursements, BCBS withheld payments on subsequent, unrelated claims that plaintiffs submitted to BCBS for services provided to other BCBS insureds.
The court ruled that one of the individual chiropractors had standing to assert claims under ERISA because the chiropractor's patients (i.e., plan participants) had assigned to him their rights to payment under their health plans and was thus an ERISA beneficiary. The court found that the chiropractor was not afforded ERISA notice and appeal rights to which a beneficiary is entitled following an adverse benefit determination, and thus granted the chiropractor summary judgment as to liability. For the other two individual chiropractor plaintiffs, the court denied summary judgment because there remained triable issues regarding anti-assignment clauses in their provider agreements. In a related decision, the court determined that the chiropractic-association plaintiffs also had standing because the injunctive and equitable relief sought would inure to the benefit of plaintiffs' members.
Sixth Circuit Expands Availability of Remedies Under ERISA
By Brian Neulander
The Sixth Circuit recently concluded that a disability plan participant was entitled to relief consisting of benefits under the plan and disgorgement of defendant's profits for delaying payment. In so ruling, the Court found that this case presented a "a logical extension" of its precedent allowing a plaintiff to pursue in limited circumstances both a claim for benefits and a claim for breach of fiduciary duty where, as here, the claim for benefits would not prevent defendants' unjust enrichment. Rochow v. Life ins. Co. of N. Am., 2013 WL 6333440 (6th Cir. Dec. 6, 2013). Moreover, relying on copyright law that permits actual damages and disgorgement of profits, the Court determined that disgorgement did not result in double compensation. Nor did it represent punishment because it left defendant no worse off than it would have been had it paid benefits to plaintiff when they were due. Lastly, the Court noted that "[i]f no remedy beyond the award of benefits were allowed, insurance companies would have the perverse incentive to deny benefits for as long as possible, risking only litigation costs in the process." A sharply worded dissent criticized the ruling, not only as contrary to precedent, but also counter to ERISA's goal of expeditious claims resolution based on the administrative record.
USSC to Consider Presumption of Prudence in Employer Stock Litigation
By Russell Hirschhorn
On Friday, the US Supreme Court agreed to consider Fifth Third Bancorp v. Dudenhoeffer (U.S. No. 12-751, cert. granted 12/13/13). The Supreme Court stated that it will consider the following issue: "Whether, to state a claim that a fiduciary of an employee stock ownership plan violated the duty of prudence by continuing to invest plan assets in the employer's stock, a plaintiff must rebut a presumption that the fiduciary acted prudently by alleging that the employer faced imminent financial peril?" Click here for a summary of the Sixth Circuit's decision.
Supreme Court Affirms Enforceability of Plan Limitations Provision
By Amy Covert and Joseph Clark
Resolving a split among the Courts of Appeal, the United States Supreme Court affirmed the Second Circuit in finding enforceable a limitations provision in a long term disability ERISA plan that set forth the length of the limitations period as well as when the period commenced. The plan at issue required participants to file suit for benefit claims within three years after "proof of loss" is due. Heimeshoff v. Hartford Life & Accident Insurance Co., 2013 WL 6569594 (S. Ct. Dec. 16, 2013).
Defendant Hartford Life & Accident Co., the administrator of Wal-Mart's Long Term Disability Plan, issued a final denial of plaintiff Julie Heimeshoff's claim for long term disability benefits in 2007. Less than three years later, but more than three years after proof of loss was due, plaintiff filed suit, challenging the denial of her benefits pursuant to ERISA § 502(a)(1)(B). She did so despite the plan's limitation provision, which stated that "Legal action cannot be taken against The Hartford . . . [more than] 3 years after the time written proof of loss is required to be furnished according to the terms of the policy."
The District Court granted defendant's motion to dismiss, holding that plaintiff's action was barred by the plan's limitation provision, which the Second Circuit upheld. Affirming, the Supreme Court held that a plan limitation period is enforceable so long as it is not unreasonable or barred by statute. In so holding, the Court disagreed with plaintiff's contention that the limitations provision would "undermine ERISA's two-tiered remedial scheme," finding that the provision does not disrupt ERISA's internal review process or diminish the availability of judicial review.
IRS Provides New Guidance on In-Plan Roth Rollovers
By Lisa A. Berkowitz Herrnson
On December 11, 2013, the IRS issued IRS Notice 2013-74 which provides guidance on in-plan Roth rollovers. An in-plan Roth rollover is a rollover within a Section 401(k), Section 403(b) or governmental Section 457(b) plan to a designated Roth account in the same plan. Notice 2013-74 expands the types of contributions that may be rolled over by an in-plan Roth rollover and supplements previous guidance on this topic.
Designated Roth contributions, governed by Section 402A of the Internal Revenue Code (the "Code"), are elective deferrals to a retirement plan, otherwise excludable from gross income, that a participant elects to include in gross income. An individual's designated Roth contributions, along with associated earnings, must be maintained in a separate account under the plan. Pursuant to an in-plan Roth rollover, a distribution from an individual's account in a retirement plan, other than a designated Roth account, may be rolled over to the individual's designated Roth account in the same plan.
Until now, based on guidance under the Small Business Jobs Act of 2010 ("SBJA"), which first provided for in-plan Roth rollovers, in-plan Roth rollovers were limited to "otherwise distributable amounts" (where a plan participant had attained age 59 ½ or terminated employment). However, Code Section 402A(c)(4)(E) was added by the American Taxpayer Relief Act ("ATRS"), effective for in-plan Roth rollovers made after December 31, 2012. Under this new rule, a plan can permit an in-plan Roth rollover of "otherwise nondistributable amounts" in addition to the amounts already deemed rollover eligible under the SBJA. Accordingly, as explained further in Notice 2013-74, the following amounts are now eligible for in-plan Roth rollovers, regardless of whether the amounts otherwise satisfy conditions for distributions:
elective deferrals in 401(k) plans and 403(b) plans;
matching contributions and nonelective contributions, including qualified matching contributions and qualified nonelective contributions; and
annual deferrals made to governmental 457(b) plans.
Notice 2013-74 clarifies that an amount rolled over to an individual's designated Roth account pursuant to an in-plan Roth rollover remains subject to the plan's distribution restrictions otherwise applicable to that amount before the in-plan Roth rollover. As an example, the guidance describes a 401(k) plan participant who makes an in-plan Roth rollover of an amount not otherwise eligible for distribution because the participant has not yet terminated employment or attained age 59 ½. In that case, the amount rolled over and any applicable earnings may not be distributed from the designated Roth account prior to the occurrence of a distribution event.
The deadline for adopting an amendment providing for in-plan Roth rollovers is the later of the last day of the first plan year for which the amendment is effective or December 31, 2014, provided the amendment is effective as of the date the plan first operates in accordance with the amendment. This extended deadline also applies to plan amendments that allow for elective deferrals under the plan to be designated as Roth contributions, allow for the acceptance of rollover contributions by designated Roth accounts, or permit in-plan Roth rollovers of some or all otherwise distributable amounts. The guidance also notes that, subject to the normal nondiscrimination rules that apply to plan benefits, rights and features, plans may restrict the types of contributions eligible for in-plan Roth rollovers, as well as the frequency of such rollovers.
Those interested in amending plans to satisfy the new rules in Notice 2013-74 should consider their options carefully and consult with counsel to make sure the technical rules are all followed.
Federal District Court (In Its Capacity As An Employer) Must Reimburse Employee for the Cost of Health Benefits for her Same-Sex Domestic Partner
By Tulio Chirinos and Roberta Chevlowe
The Ninth Circuit Judicial Council, an administrative body that reviews decisions of the court's chief judge, recently weighed in on an issue involving same-sex domestic partner health benefits in the post-Windsor world. The decision is interesting insofar as it relies at least partially on the Windsor decision in awarding "spousal" benefits to an unmarried same-sex couple, even though Windsor only addressed the rights of same-sex married couples.
In In re Fonberg, 2013 U.S. App. LEXIS 23826 (9th Cir. Nov. 25, 2013), the Judicial Council ordered a federal district court in Oregon to reinstate a back pay award to its former employee (a law clerk at the court) to account for the cost of health benefits for her same-sex domestic partner. In 2009, the law clerk had requested (and was denied) enrollment of her partner in the health coverage provided by the district court. The coverage was denied because the district court only provided health benefits for spouses, not domestic partners. The law clerk and her partner were unable to marry in their state of residence (Oregon) due to the state's constitutional ban on same-sex marriage. However, under Oregon law, they were able to (and did) register as domestic partners and, under that law, domestic partnerships are conferred rights "on equivalent terms" to marriage.
After the clerk's request for domestic partner health coverage was denied by the Office of Personnel Management (the entity that administers benefit programs for employees of the federal government, including the district court), the clerk filed a discrimination complaint under Oregon law, which at the time prohibited discrimination based on sex and later was amended to include sexual orientation as a protected category. Although the clerk initially was awarded an allowance for the cost of providing her partner with health coverage, that directive was subsequently rescinded by the chief judge.
On appeal, the Judicial Council held that the denial of benefits violated Oregon's nondiscrimination law because the clerk and her partner were being treated differently from opposite sex partners who could marry and receive spousal health benefits from the federal government. The Council found that, while "Oregon's statutory scheme purports to confer upon same-sex domestic partners the same rights and legal status as those conferred on married partners," in actuality it does not, since those partners are denied benefits provided to married couples. Citing Windsor, the Council further concluded that the distinction drawn by the Office of Personnel Management based on the sex of the partners constitutes a deprivation of the clerk's due process and equal protection rights.
The Fonberg decision has far-reaching consequences inasmuch as it appears to require federal government employers to provide health benefits to unmarried same-sex domestic partners who reside in states that provide them with rights equivalent to marriage, even though Windsor only conferred rights on married partners (and has been interpreted as not applying to couples in domestic partnerships and civil unions.) As with other issues flowing from the Windsor decision, stay tuned for how this will play out.
IRS Issues Temporary Non-Discrimination Testing Relief for Closed Defined Benefit Plans and Request for Comments
By Paul Hamburger and Lisa A. Berkowitz Herrnson
On December 13, 2013, the IRS issued Notice 2014-5 which provides temporary relief for satisfying the nondiscrimination requirements under Section 401(a)(4) of the Internal Revenue Code (the "Code") for plan sponsors that maintain defined benefit plans which have been closed to
Many defined benefit plan sponsors have implemented "soft freezes" of their plans, closing them to new hires, but continuing defined benefit plan accruals for participants hired before the "soft-freeze." These employers might then implement a new or enhance an existing defined contribution plan for new hires. Over time, the defined benefit plan might no longer pass coverage testing under Code Section 410(b) on its own because of greater turnover among the non-highly compensated employee group covered by the plan relative to the highly compensated employee population. As a result, the frozen defined benefit plan must be aggregated with the defined contribution plan to satisfy the coverage requirements of the Code (the defined benefit plan, when aggregated with the defined contribution plan for testing is referred to as the "Aggregated Plan").
Then, if the defined benefit plan and the defined contribution plan are aggregated for coverage testing, they also must be aggregated for nondiscrimination testing. To satisfy the nondiscrimination requirements, that Aggregated Plan must be tested on the basis of "equivalent benefits" (i.e., by converting the defined contribution amounts to equivalent benefits and then aggregating those benefits with the benefits under the defined benefit plan) by satisfying one of three conditions. Specifically the Aggregated Plan must:
be primarily defined benefit in character;
consist of broadly available separate plans; or
meet the minimum aggregate allocation gateway.
Meeting one of the first two conditions above is most likely during the early years following a "soft freeze." Eventually, however, the demographics will change such that the Aggregated Plan will be required to satisfy the minimum aggregate allocation gateway, which requires that each nonhighly compensated employee in the Aggregated Plan have a minimum aggregate normal allocation rate that is a function of the highest aggregate normal allocation rate of the highly compensated employees. Because many defined contribution plans may not provide for allocations that would satisfy this gateway requirement, the Aggregated Plan will end up not satisfying the nondiscrimination requirements.
Notice 2014-5 offers temporary relief for plan sponsors that maintain closed defined benefit plans. This relief helps these plans meet the Code's technical nondiscrimination requirements for plan years beginning before January 1, 2016, even if the equivalent benefit conditions cannot be satisfied, as long as the defined benefit plan in the Aggregated Plan situation was amended to exclude new hires prior to December 13, 2013 and meets one of the following two conditions:
for plan year beginning in 2013, the defined benefit plan was part of an Aggregated Plan that was either primarily defined benefit in character; or
the defined benefit plan passed nondiscrimination testing on its own, without any aggregation, for the plan year beginning in 2013.
The IRS noted that all other provisions under Code Section 401(a)(4) concerning nondiscrimination testing apply during this period temporary relief.
Although the relief provided by Notice 2014-5 is certainly welcome, it is questionable how helpful it actually will be to plan sponsors. Perhaps more significantly, the Notice also includes a number of proposals for a permanent resolution to resolve these issues in the future. Plan sponsors facing the challenges addressed in the Notice should certainly consider filing comments on these proposals. Written or electronic comments are being accepted by the IRS until February 28, 2014.
More Post-Windsor Guidance from the IRS Relating to Benefits for Same-Sex Spouses: Cafeteria Plan Changes, Flexible Spending Accounts and Health Savings Accounts
By Bali Kumar and Roberta Chevlowe
Prior to the U.S. Supreme Court's Windsor decision that repealed Section 3 of the federal Defense of Marriage Act (DOMA), same-sex spouses were not recognized as spouses for federal tax and benefits purposes. In the immediate aftermath of Windsor, the Internal Revenue Service (IRS) issued Revenue Ruling 2013-17, which stated the IRS position that, for federal tax purposes, the term "spouse" now includes legally married same-sex couples regardless of whether their state of residence permits same-sex marriage. As a result, the value of employer-provided health coverage for a same-sex spouse would no longer be taxable under federal law, and employees could pay for the coverage on a pre-tax basis through an employer's cafeteria plan. Employees also could obtain reimbursement for same-sex spouses' expenses under health care and other reimbursement plans.
On December 16, 2013, the IRS supplemented that guidance with the release of Notice 2014-1, written in Q&A format with examples. Notice 2014-1 clarifies several issues for plan sponsors and administrators of cafeteria plans, flexible spending accounts (FSAs) and health savings accounts (HSAs). First, it allows an employer to permit an employee to make a mid-year election change under its cafeteria plan with regard to health coverage for a same-sex spouse. Second, it provides that an employee may be reimbursed from his or her health care FSA for expenses incurred by a same-sex spouse during the 2013 plan year, even before the date of the Windsor decision (but no earlier than the date of the marriage). Third, it confirms that a same-sex married couple is subject to the joint limits applicable to married couples under HSAs and dependent care plans.
Mid-Year Election Changes under Cafeteria Plans
In the wake of Windsor, employers were confronted with the question of whether they could allow an employee who was married to a same-sex spouse before the date of the Windsor decision to enroll the spouse in employer-provided health coverage or, if the spouse was already covered under the plan, to make an election change to pay for the coverage on a pre-tax basis. Since mid-year election changes under a cafeteria plan are permitted only in limited circumstances where there has been a "change in status" recognized under Section 125 of the Internal Revenue Code (the Code), the answer to this question was not clear. In the absence of specific guidance, some employers took the position that the Windsor decision itself was a change in legal marital status for such an employee (since the marriage was now, but was not previously, recognized under federal law) and, therefore, this was a permissible election change.
Notice 2014-1 confirms that it was (and is) permissible for a cafeteria plan to allow a participant who was already married to a same-sex spouse as of the date of the Windsor decision to make an election change, treating the employee as if he experienced a change in legal marital status due to the Windsor decision. Cafeteria plans may accept such election changes at any time during the plan year that includes June 26, 2013 (i.e., the date of the Windsor decision) or December 16, 2013 (the effective date of the Notice). In addition, a mid-year election change due to a marriage that occurs after the date of the Windsor decision is permissible.
On a related note, the IRS makes clear that the change in federal tax treatment of a same-sex spouse's health benefits (as a result of Windsor) does not constitute a "significant change in the cost of coverage" under the Code Section 125 change in status rules; but, in light of the legal uncertainty created by Windsor and the fact that the Notice in any event recognizes Windsor as a change in legal marital status, .employers that allowed such changes will not be treated as having failed to comply with Section 125.
Windsor-related election changes will generally be effective as of the date other changes become effective under the plan rules. In addition, with respect to election changes made between June 26 and December 16, 2013, the IRS explained that a plan will not be treated as failing to meet the Code Section 125 requirements to the extent that coverage becomes effective no later than the later of: (i) the date that coverage would be added under the plan's usual procedures for change-in-status elections, or (ii) a reasonable period of time after December 16, 2013.
Pre-Tax Premium Payments for Spousal Health Coverage
Notice 2014-1 states that if the employer is notified before the end of the plan year that includes December 16, 2013 (i.e., by December 31, 2013 for a calendar year plan) that a participant is married to his or her same-sex partner (who is covered under the plan), the employer must begin treating the employee contribution as a pre-tax salary reduction no later than the later of (i) the date that a change in legal marital status would be required to be reflected for income tax withholding purposes, or (ii) a reasonable period of time after December 16, 2013. Individuals may notify the employer of the marriage by making a mid-year election change to pay for coverage on a pre-tax basis (as permitted by the Notice) or by filing a revised Form W-4 reflecting the change in marital status.
To the extent that an employee paid for a same-sex spouse's health coverage on an after-tax basis, the guidance makes clear that, for the plan year including December 16 (and prior years for which the limitations period for filing a tax refund claim has not expired), the amounts paid by the employee for the spousal coverage are excluded from the employee's gross income and not subject to federal income and employment taxes, and an employee may seek a refund for taxes paid on such amounts.
Notice 2014-1 also provides that an employee may be reimbursed from his or her health care FSA for expenses incurred by a same-sex spouse during the 2013 plan year, even before the date of the Windsor decision (but no earlier than the date of the marriage). For example, an FSA with a calendar year plan year may reimburse expenses incurred by the spouse on or after January 1, 2013 (or the date of marriage, if later). The guidance further states that a same-sex spouse may be treated as covered by the FSA even if the participant initially elected coverage under a self-only FSA. These rules also apply to dependent care and adoption assistance FSAs.
Contribution Limits for HSAs and Dependent Care Assistance Programs
Notice 2014-1 confirms that a same-sex married couple (married as of the last day of the taxable year) is subject to the joint deduction limit for contributions made to an HSA where either spouse elects family coverage under a high deductible health plan. If the spouses' combined elections would exceed the $6,450 limit for the 2013 taxable year, the guidance provides that contributions for one or both of the spouses may be reduced for the remaining portion of the tax year to avoid exceeding the limit. If the contributions already exceed $6,450, the excess may be distributed from the HSAs of one or both spouses no later than the tax return filing deadline for the spouses. Any excess that remains undistributed as of such date will be subject to excise taxes under the Code.
The guidance further provides that a same-sex married couple (married as of the last day of the taxable year) is subject to the maximum annual exclusion from gross income for married couples under a dependent care FSA as of the 2013 taxable year. If the couple's combined contribution elections exceed $5,000 (or $2,500 per spouse, if married filing separately) for 2013, their contributions may be reduced for the remainder of the year to avoid exceeding the limit. Otherwise, the excess contributions will be includable in their gross income and will be taxed.
Written Plan Amendment
If a cafeteria plan explicitly allows mid-year election changes due to a change in legal marital status, a plan amendment generally is not required in order to permit changes relating to same-sex spouses as a result of the Windsor decision. However, if the plan sponsor chooses to permit Windsor-related election changes that were not previously included in the plan document, the plan must be amended on or before the last day of the first plan year beginning on or after December 16, 2013. For calendar year plans, this means that the plan must be amended by December 31, 2014. The guidance specifically states that the amendment may be effective retroactive to the first day of the plan year that includes December 16, 2013, provided that the plan operates in compliance with the guidance contained in the Notice.
Federal District Court Tosses $450 Million Retiree Health Care Suit Against GM
By Todd Mobley and Stacey Cerrone
A federal district court in Michigan dismissed a breach-of-contract suit against General Motors over a $450 million payment for retiree medical benefits. Int'l Union, UAW v. Gen. Motors, LLC, No. 10-11366, 2013 U.S. Dist. LEXIS 173793 (E.D. Mich. Dec. 10, 2013). As part of a settlement agreement with the UAW in 2007, General Motors promised to make a $450 million payment to a voluntary employee's beneficiary association (i.e., a trust established to mitigate medical costs for UAW-affiliated retirees). However, prior to making the $450 million payment, General Motors filed for chapter 11 protection. After emerging from bankruptcy, the newly restructured General Motors entered into a new agreement with the UAW (the "2009 Agreement"), which explicitly superseded all prior agreements between General Motors and the UAW regarding retiree medical benefits. The 2009 Agreement was silent on the $450 million payment obligation contained in the 2007 Agreement. When General Motors later refused to make the $450 million payment, the UAW sued for breach of the 2007 Agreement. Despite the complete absence of language in the 2009 Agreement as to the $450 million payment, the UAW argued that General Motors' $450 million obligation remained. According to the UAW, if the parties had wished to extinguish such a large obligation, they would have explicitly said so in the 2009 Agreement. The court disagreed, holding that the 2009 Agreement superseded the 2007 Agreement and explaining that General Motors assumed only the obligations contained in the 2009 Agreement. Essentially, the court found that the "UAW's efforts to turn the absence of language into language [was] reminiscent of the efforts to capture a 'will o' the wisp.'"
* Originally published by Bloomberg, BNA. Reprinted with permission.
 Scott v. Family Dollar Stores, Inc., No. 12-1610, 2013 BL 287115 (4th Cir. Oct. 16, 2013) (allowing complaint to be amended to assert class claims for decisions involving management discretion; distinguished Wal-Mart based on amended allegations involving higher-level managers than those at issue in Wal-Mart).
 Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2553-56 (2011).
 Id. at 2553-54
 Id. at 2555-56.
 133 S. Ct. 1426 (2013).
 Id. at 1430-31.
 Id. at 1431, n.3.
 Id. at 1433-35.
 Id. at 1433.
 Id. at 1441.
 Id. at 1433-35.
 Daubert v. Merrill Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), is the seminal Supreme Court ruling setting forth the reliability standards that apply before expert evidence is admissible. See also Fed. R. Evid. 702.
 Fed. R. Evid. 702; see also Daubert, 509 U.S. at 589-95.
 The ruling by the Wal-Mart district court was an apt example of the prior standards, admitting a social framework analysis that could not answer with any degree of confidence the class issue; the district court did so because it thought Daubert did not apply at the class stage. The Supreme Court pointedly noted "we doubt that is so," and proceeded to eviscerate and dismiss this evidence as it decertified the class. See Wal-Mart, 131 S. Ct. at 2553-54. After Wal-Mart, many courts apply Daubert to the admissibility of expert analysis of class issues; perhaps more important, like Comcast, they require this evidence to be not just admissible but persuasive in carrying plaintiff's burden to show that Rule 23's requirements have been met. See, e.g., Comcast, 133 S. Ct. at 1433-35; Ellis v. Costco Wholesale Corp., 657 F.3d 970, 982 (9th Cir. 2011) (expert's testimony must be admissible under Daubert and persuasive on the class issues under the "rigorous analysis" standard applied to class certification); Pedroza v. PetSmart, Inc., (C.D. Cal. Jan. 28, 2013) (explaining and applying same).
 Federal Judicial Center's Reference Manual on Scientific Evidence, 233-35 (3d Ed. 2011).
 See 131 S. Ct. at 2554.
 Reference Manual on Scientific Evidence, 233-35.
 Id. at 262-64; see also Randall v. Rolls-Royce Corp., No. 1:06-cv-860-SEB-JMS (S.D. Ind. Mar. 12, 2010) (in gender discrimination claim over compensation in which managers had substantial discretion to adjust pay, expert's report failed to account for components of the compensation system such as merit increases, annual bonuses, and critical skills adjustments).
 E.g., EEOC v. Bloomberg L.P., No. 07 Civ. 8383 (LAP), 2010 BL 31687 (S.D.N.Y. Aug. 31, 2010) (in pregnancy discrimination claim over leave, expert's report was irrelevant and unreliable because it failed to compare class members to other similarly situated Bloomberg employees who had also taken leave); Bolden v. Walsh Grp., No. 06 C 4104, 2012 BL 76095 (N.D. Ill. Mar. 30, 2012) (in race discrimination claim, the expert report was inadmissible because in coming up with labor market comparator to show disparities, it failed to take into account geography, commuting distance, ability and job interest).
 See 131 S. Ct. at 2553-55; see also, e.g., Jones v. YMCA, No. 09 C 6437 (N.D. Ill. Sept. 5, 2013) (experts' research on unconscious bias based on millisecond word associations in a laboratory cannot be reliably applied to managers' decisions in the workplace; such an application has no social science research support, and ignores numerous differences, including that managers' decisions have consequences and there are accountability mechanisms that counteract any unconscious bias).
 See Wal-Mart, 131 S. Ct. at 2561; cf., e.g., In re Taco Bell Wage & Hour Actions, No. 1:07-cv-01314-OWW-DLB., 2011 BL 244118 (E.D. Cal. Sept. 26, 2011) (to show class manageability under Rule 23(b)(3), plaintiff proposed an expert who would survey a sample to develop a statistical analysis of class liability and damages; court held expert's proposed method was too unreliable).
 See 131 S. Ct. at 2555.
 Id. at 2555-56.
 688 F.3d 893, 898 (7th Cir. 2012).
 Id. at 894.
 Id. at 896.
 Id. Similarly, in Bell v. Lockheed Martin Corp., No. 08-6292 (RBK/AMD), 2011 BL 316158 (D.N.J. Dec. 14, 2011), plaintiffs alleged that Lockheed engaged in gender discrimination by promoting men more quickly than women who were equally, if not more, qualified for the respective positions. As in Wal-Mart, decisions regarding promotions and wage increases were made pursuant to the discretion of individual managers; however, the statistical analyses proposed by plaintiffs were aggregated based on company-wide data. The court found that because the statistical proofs proffered by plaintiffs were essentially the same as the proofs rejected in Wal-Mart, plaintiffs failed to establish commonality. Id. at *23-24.
 726 F.3d 372 (3d Cir. Aug. 12, 2013).
 Id. But see Scott v. Family Dollar Stores, Inc., No. 12-1610, 2013 BL 287115 (4th Cir. Oct. 16, 2013) ("Wal-Mart is limited to the exercise of discretion by lower-level employees, as opposed to upper-level, top-management personnel…. Consequently, discretionary authority exercised by high-level corporate decision-makers, which is applicable to a broad segment of the corporation's employees, is more likely to satisfy the commonality requirement than the discretion exercised by low-level managers in Wal-Mart.").
 No. CV 08-07919 GAF (PJW) (C.D. Cal. Nov. 23, 2011).
 255 F.R.D. 450, 454-55 (N.D. Ill. 2009).
 Id. at 461-62.
 Id. at 462.
 Id. at 465; see also In re Taco Bell, 2011 U.S. Dist. LEXIS 109169, at *18 (finding that failure to consider all facts underlying terminations of putative class members eliminates ability to opine on class-wide terminations).
 Puffer, 255 F.R.D. at 461.
 Bolden v. Walsh Grp., No. 06 C 4104, 2012 BL 76095 (N.D. Ill. Mar. 30, 2012) (in race discrimination claim the expert report was inadmissible because it failed to take into account how geography and commuting distance would affect job interest); Aliotta v. Bair, 614 F.3d 556, 568 (D.C. Cir. 2010) ("If [expert's] statistics do not control for [employee choice], they tell us nothing about why older employees took the buyouts, and are therefore not relevant to determining whether FDIC discriminated against them."); EEOC v. Sears, Roebuck & Co., 839 F.2d 302, 320-21, 324-26 (7th Cir. 1988) (same—statistics were flawed and failed to show discrimination when they failed to account for different interest between genders for commissioned sales position).
 133 S. Ct. at 1433 (internal quotation omitted).
 No. 04-40132, 2009 BL 133763 (E.D. Mich. Mar. 31, 2009).
 Id. at 6.
 No. 1:06-cv-860-SEB-JMS (S.D. Ind. Mar. 12, 2010) aff'd 637 F.3d 818 (7th Cir. 2011).
 Id. (explaining that Plaintiffs' expert "examined components of the Rolls-Royce compensation system other than base salaries, such as the merit increases, annual bonuses, and critical skills adjustments to employee salaries, and found no evidence of disparities adverse to women. Conveniently, that conclusion is not disclosed in his report.").
 No. 1:10-cv-00911, 52 EBC 1965 (N.D. Ill. Nov. 15, 2011)(222 PBD, 11/17/11; 38 BPR 2142, 11/22/11).
 Id. at *5.
 705 F.3d 370, 372 (8th Cir. 2013).
 Id. at 374-76.
 476 F.3d 299, 39 EBC 2352 (5th Cir. 2007)(13 PBD, 1/22/07; 34 BPR 210, 1/23/07).
 Id. at 315.
 Id. at 315-16.
 No. 1:10-cv-00911, 52 EBC 1965 (N.D. Ill. Nov. 15, 2011).
 633 F.3d 574, 586-87, 50 EBC 1801 (7th Cir. 2011)(16 PBD, 1/25/11; 38 BPR 220, 2/1/11). Cf. Abbott v. Lockheed Martin Corp., 725 F.3d 803, 813-14, 56 EBC 2352 (7th Cir. 2013)(154 PBD, 8/9/13; 40 BPR 1959, 8/13/13)(in case challenging investment in a 401(k) plan, finding that the plaintiffs' use of the Hueler Index as a method of measuring damages was a reasonable way of excluding uninjured persons from the class and limiting intra-class conflicts).
 See In re Principal U.S. Property Account ERISA Litig., No. 4:10-cv-00198 at pp. 46-57, 2013 BL 282261 (S.D. Iowa Sept. 30, 2013)(198 PBD, 10/11/13; 40 BPR 2413, 10/15/13) (class members made individual investment choices in response to withdrawal restrictions that impacted who was in the class and who was harmed; plaintiff's claims also created conflicts between those who wanted higher liquidity versus those who wanted higher returns).
 No. 3:05-cv-115, 49 EBC 2194 (D.N.D. May 7, 2010), amended by No. 3:05-cv-115 (D.N.D. Aug. 27, 2010). For cases raising similar issues outside the ERISA context, see Gilpin v. American Federation of State, County, & Municipal Employees, AFL-CIO, 875 F.2d 1310, 1313 (7th Cir. 1989) (affirming decision to refuse to certify a class when different groups of employees had differing interests as to the remedies to be sought from a union), and United Independent Flight Officers v. United Air Lines, 756 F.2d 1274, 1284 (7th Cir. 1985) (affirming refusal to certify class when members had divergent and antagonistic interests regarding goals of the lawsuit and benefits sought).