Robin Feeko v. Pfizer Inc

AMBRO, Circuit Judge,

dissenting.

We must give a conflict of interest “greater importance, perhaps determinative importance, where the evidence suggests a greater likelihood that it affected the decision to deny benefits.” Howley v. Mellon Fin. Corp., 625 F.3d 788, 794 (3d Cir.2010) (emphasis added). In this case, three of the five members of Pfizer’s Administrative Committee came to the table with a conflict that prevented them from neutrally deciding the benefits claims. My colleagues nonetheless bless the Committee’s decision. Because I believe there was a debilitating conflict whose influence is apparent in a series of procedural and substantive missteps in the Committee’s deliberations, I respectfully dissent.

The question before the Administrative Committee was whether Benchmark was a successor company to Pfizer. If Benchmark was, then Appellants are not entitled to severance benefits; if it was not, then Appellants win. ERISA demands a neutral resolution to that question. But for three members of the Committee, which operates by majority vote, the result was a foregone conclusion. That is because, as the District Court noted, several months before Appellants filed their claims, these members were “personally involved” in the decision to take Appellants off Pfizer’s payroll and put them on Benchmark’s. Feeko v. Pfizer, Inc., Civ. Action No. 11-4296, 2014 WL 6473265 , at *5 (E.D.Pa. Nov. 18, 2014). During that process, the three members, acting in their business rather than fiduciary capacities, structured the deal with an intent to assure that Benchmark would be a successor company to Pfizer. Id. at *6. They did so under the belief that, as a result of this structure, severance payments would be “precluded,” Id. Indeed, during the meeting when the Committee considered Appellants’ claims, these members conceded that they had already been part of a decision that Appellants “would not be eligible for severance.” App. at 285 (Committee minutes). And though they disclosed this, the Committee made no effort to mitigate the conflict. Feeko, 2014 WL 6473265 , at *6. Pfizer now asks us to consider these members to be neutral arbiters on a question whose answer they had already decided. That asks too much.

The District Court acknowledged the severity of this conflict. The three members, it noted, had “significant incentive” to deny Appellants’ benefits claims. Id, In other words, they “would have been influenced to reach the same conclusion” in their fiduciary capacities as they did in their business capacities. Id. They might “seek to avoid contradicting their own interpretations so quickly; they would have significant reason to avoid subjecting Pfizer to liability for severance benefits when they had participated in drafting or approving language they believed precluded such liability.” Id.

Yet the District Court concluded, and the majority here agrees, that these conflicts, while substantial, can be overlooked. But this approach does little more than pay lip service to conflicts as a factor in evaluating a plan administrator’s decisions. The logical conclusion from the District Court’s opinion is that the conflict bore heavily on the Committee’s decision-making process. The conflict therefore undermined the neutrality that the Committee was supposed to have.

*107 Perhaps Pfizer could have overcome all of this if the other evidence clearly supported its position. Instead, the opposite is true. The administrative record is so littered with pitfalls that it only supports, rather than rebuts, the notion that a conflict of interest improperly intruded on the Administrative Committee’s deliberations.. As my colleagues acknowledge, the Committee based its interpretation on terminology found nowhere in the severance plan, mistakenly described the nature of the transaction between Pfizer and Benchmark (calling it a “sale” even though nothing was sold), and supported its determination using past practices that arose under entirely different factual situations. Maj. Op. at 104-05.

And most importantly — and here my view differs substantially from that of the majority — the Committee labeled Benchmark a “successor company” to Pfizer under circumstances that defy the common meaning of the term. The majority is correct that “successor company of the Company” can mean a number of things. See, e.g., Howard Johnson Co. v. Detroit Local Joint Exec. Bd., 417 U.S. 249, 257 , 94 S.Ct. 2236 , 41 L.Ed.2d 46 (1974) (noting that “mergers, consolidations, or purchases of assets” can, under the right circumstances, lead to successorship). But here the question is not what it can mean, but rather what it cannot mean. We might, for instance, reasonably debate whether a transfer of a small percentage of assets creates successorship. But under our facts, there are none of the minimal indicia of successorship — no consolidation, no 'merger, no purchase of assets. The only attempt that Pfizer makes to justify its interpretation as a matter of common usage is that “Benchmark qualified as a successor company of Pfizer with respect to the provision of credit union services.” Feeko, 2014 WL 6473265 , at *8. But that does not do the trick because Benchmark was already providing these services for Pfizer well before Appellants became Benchmark employees.

Benchmark concededly succeeded to Pfizer in the employment of Appellants, but that makes Benchmark the successor employer of Appellants. It does not make Benchmark a successor company to Pfizer, which is the relevant inquiry under the severance plan. Imagine that Pfizer, desiring healthy meal options for its employees, opens a McDonald’s on its campus and decides that it will staff the restaurant with Pfizer employees so that it can more effectively manage their performance. Growing tired of the arrangement, Pfizer transfers the employees to- McDonald’s. There is no logical way to view McDonald’s in this scenario as a successor company to Pfizer. For the same reasons, Benchmark did not, in any normal usage of the term, become a successor company to Pfizer.

The majority’s other principal argument — that the Committee’s construction is consistent with the purpose of the severance plan because Appellants are not what comes to mind when one thinks of severed employees — no doubt has surface appeal, but I am unconvinced. At the outset, Pfizer’s own actions — in sending Appellants a letter saying they had experienced a “termination of employment” and in providing them with a question-and-answer document saying that a period of unemployment was not a prerequisite to receiving salary continuation benefits — rebut this. See, e.g., App. at 203, 625. Additionally, given that Appellants have lost certain Pfizer benefits (such as pension plan contributions and free prescriptions) now that, they work for Benchmark, our own case law supports the conclusion that there is nothing incongruous about treating them as severed employees with all the severance rights accorded that status. See Ko-trosits v. GATX Corp. Non-Contributory Pension Plan for Salaried Employees, 970 *108 F.2d 1165 , 1171 (3d Cir.1992) (noting that severance payments can offset a reduction in benefits even when “employment continues without interruption with a new employer”).

So we are back to where we started. Under Howley , a conflict can be “determinative” of our outcome if it bears sufficiently on the decision of a plan administrator. 625 F.3d at 794 . If we are not willing to give determinative effect here— where we have an interpretation at odds with the key term’s common meaning, various procedural errors, and a conflict that provided, in the District Court’s- own words, a “significant incentive” to deny Appellants’ claims — I cannot see how we can ever factor conflicts meaningfully into our analysis. As a result, I would reverse the District Court and grant summary judgment to Appellants. 1

. Because the majority upholds the grant of judgment to Pfizer, it does not need to address the District Court's separate order denying class certification on the ground that the class was not sufficiently numerous, Given my view that Appellants are entitled to judgment, I would need to review the certification order. It rested on the belief that unnamed class members who have missed the deadline to exhaust their administrative remedies cannot, as a matter of law, count toward the numerosity requirement. I would hold that the District Court erred in this determination. The overwhelming consensus among other courts to consider the question in the ERISA context is that exhaustion by unnamed class members is not necessary where the named plaintiffs have exhausted their administrative remedies. See, e.g., Flinders v. Workforce Stabilization Plan of Phillips Petroleum Co., 491 F.3d 1180 , 1193 n. 5 (10th Cir.2007); In re Household Int’l Tax Reduction Plan, 441 F.3d 500, 501-02 (7th Cir.2006); Thomas v. Smith-Kline Beecham Corp., 201 F.R.D. 386, 395 (E.D.Pa.2001). Although exhaustion by unnamed class members is not required, the District Court still has discretion to determine whether to count unnamed members who did not exhaust toward the numerosity requirement. In exercising its discretion here, the District Court would have needed to consider whether counting the unnamed members would defeat the purposes that the exhaustion requirement typically serves: providing notice to the defendant of the nature of the claims prior to litigation and allowing an opportunity for internal resolution. Relevant factors include whether the unnamed members’ claims are "very similar to those” Pfizer already rejected administratively (such that Pfizer, even without exhaustion, had notice) and whether requiring exhaustion "would merely produce an avalanche of duplicative proceedings and accidental forfeitures.” In re Household Int’l Tax Reduction Plan, 441 F.3d at 501-02 . If the District Court, in exercising its discretion, had counted the unnamed members and found that the class was sufficiently numerous, it would then have needed to determine whether any other reasons precluded class certification. •