In its seminal ruling in Metropolitan Life Insurance Co. v. Glenn, 128 S. Ct. 2343 (2008), the U. S. Supreme Court held unanimously that a “structural” conflict of interest exists in situations where the same entity evaluates claims for benefits and pays benefit claims. Id. at 2348. The Court went on to state that the existence of such a conflict would be one factor among many in determining whether there has been an abuse of discretion, and the alleged conflict is of greater importance where circumstances suggested a “higher likelihood” that the conflict affected the benefits decision, or where there was a history of biased claims administration. Id. at 2351.

The Supreme Court’s Glenn ruling has left the courts to address two issues in benefit claim litigation: (1) whether a conflict of interest exists; and (2) if a conflict of interest does exist, what, if any, discovery related to that conflict of interest should be permitted.

Although it was hoped that the ruling in Glenn would lead to greater uniformity and predictability in the adjudication of claims for benefits, Glenn’s instructions have left courts struggling in determining how to apply them. Not only have the lower courts divided on the scope of discovery to permit where an inherent conflict is identified, but they have also disagreed on the criteria for finding an inherent conflict in the first place. The disagreement among the lower courts has now reached the circuit court level, causing one to wonder how much guidance Glenn – the first Supreme Court case to address the standard of review in benefits cases in over 20 years – has really provided.

Identifying Structural Conflicts in Taft-Hartley Plans

With respect to the first issue (whether a structural, or inherent, conflict of interest exists), Glenn provided specific guidance only with respect to insured single employer plans. With respect to multi-employer Taft-Hartley plans, which typically are administered by an evenly divided committee of employer and union designated trustees, courts have divided sharply on how best to apply Glenn’s rationale.

In Anderson v. Suburban Teamsters of Northern Illinois Pension Fund Board of Trustees, 588 F.3d 641 (9th Cir. 2009), the Ninth Circuit held that trustees of a multi-employer plan did not have a structural conflict of interest. In so holding, the court reasoned that, because the board of trustees consists of both employer and employee representatives, there are no pecuniary interests at stake. The Ninth Circuit held that, given the lack of economic motivation, the administrative decision-making process did not fit Glenn’s definition of a structural conflict of interest.

By contrast, in Durakovic v. Building Service 32 BJ Pension Fund, 609 F.3d 133 (2d Cir. 2010), the Second Circuit held that Taft-Hartley funds are inherently conflicted when making benefit determinations because, like insurance companies and employers administering self-insured plans, they both determine and pay claims. The court reasoned that, while the employer representatives on the board of trustees have fiduciary interests that weigh in favor of the trusts’ beneficiaries, they also have representational and other interests that weigh to the contrary. The fact that union representatives have an equal say in benefit determinations, the court stated, does not negate the conflict, but rather may impact the weight the conflict is afforded.

In Griffin v. N.Y. State Nurses Association Pension Plan & Benefits Fund, No. 10-CV-824, -- F. Supp. 2d --, 2010 WL 5342069 (E.D.N.Y. Dec. 22, 2010), (E.D.N.Y. Dec. 22, 2010), a district court cited to Durakovic and found a conflict existed because “the Plan is a multi-employer benefit trust fund and the board of trustees consists of fourteen members, comprised equally of representatives of the New York State Nurses Association and the management of participating facilities, as required by the Taft-Hartley Act,” who are responsible for paying and determining benefit claims.

The Fourth Circuit viewed the issue differently, albeit under somewhat unusual circumstances. In Parsons v. Power Mountain Coal Co., 604 F.3d 177 (4th Cir. 2010), the Fourth Circuit considered an employer’s allegation of an alleged conflict in the administration of benefits in a multi-employer plan where the underlying factual and eligibility determinations were made exclusively by union trustees. The employer was apparently concerned that the union trustees would be motivated to find eligibility, at the employer’s potential expense, because contributing employers would ultimately bear the consequences of a plan shortfall resulting from overly generous benefit payments. The Fourth Circuit upheld the trustees’ determination of eligibility and rejected the employer’s conflict of interest claim. The finding that there was no conflict was based on the observations that employer contributions to the plan are prescribed by the collective bargaining agreements and are not impacted by benefit determinations. To the extent that the benefit claim experience results in a shortfall in funding, the plan’s remedy is to reduce benefit levels. Thus, there is no improper motivation for the trustees to grant or deny benefits.

These varied approaches to the application of Glenn to Taft-Hartley plans puts into focus some of the issues of interpretation left open by Glenn, including with respect to when a structural conflict of interest exists. The Second Circuit’s ruling appears to be based on a literal reading of the opinion insofar as Glenn suggested that a structural conflict exists whenever the plan administrator and the entity paying benefits are one and the same. The Ninth and Fourth Circuit Courts, by contrast, favored a more practical approach that inquires into whether the benefit determination structure is one that could logically be impacted by financial considerations. The approach taken has a direct impact on whether conflict discovery will be required. And, as demonstrated below, once the gates to discovery are opened, there is no telling when they will close.

What Is the Scope of Conflict of Interest Discovery?

For any plan found to have a structural conflict – including presumably Taft Hartley plans – the court must consider what, if any, conflict of interest discovery should be conducted. With respect to this issue, three potential areas of discovery have been evaluated: (1) claims administration policies and manuals; (2) treatment of similar past claims; and (3) relationships among the entities providing and deciding benefit claims.

Claims Administration Policies and Manuals. Over the past year, courts have consistently permitted conflict of interest discovery of documents and information relating to claims policies, procedures, and/or manuals. For example, in Kruk v. Metropolitan Life Insurance Co., 267 F.R.D. 435 (D. Conn. May 27, 2010), the district court permitted limited discovery as to any statement of policy or guidance with respect to the plan and concerning the denied treatment option or benefits for the claimant’s diagnosis, without regard to whether such advice or statement was relied upon in making the benefit determination. Similarly, in Rauthe v. Metropolitan Life Insurance Co., No. CV 10-47-M-DWM-JCL, 2010 WL 3522487 (D. Mont. Sept. 1, 2010), the court permitted discovery as to the procedures and policies MetLife followed to investigate claims. In Emery v. American Airlines, Inc., No. 08-CV-22590, 2010 WL 457151 (S.D. Fla. Feb. 4, 2010), the district court permitted discovery relating to claim manuals, procedures, guidelines, and handbooks used for assessing the claim or relating to safeguards for following plan procedures and reducing bias; and documents showing management checks that penalize inaccurate decision-making or show active steps to reduce bias that relate specifically to Emery’s claim. In Prado v. Allied Domecq Spirits and Wine Group Disability Income Plan, No. 09-CV-4419, 2010 WL 3119934 (N.D. Cal. Aug. 2, 2010), the district court permitted limited discovery of all the plan documents and any policies guiding the insurer’s (Liberty’s) decision-making to determine if Liberty abused its discretion in denying plaintiff’s claims, as well as limited discovery into the nature, extent, and effect of Liberty’s conflict of interest on its decision-making process.

Similar Types of Claims. A second scope of conflict of interest discovery relates to past treatment by the decision-maker of similar types of claims. In Hall v. Life Insurance Co. of North America, 265 F.R.D. 356 (N.D. Ind. Feb. 25, 2010), the district court permitted limited discovery relating to how many similar types of claims were reviewed over the last five years where claimants were initially found not disabled, and then whether the claim was denied on appeal. In Zewdu v. Citigroup Long Term Disability Plan, 264 F.R.D. 622 (N.D. Cal. Feb. 12, 2010), the district court permitted discovery on the number of disability claims reviewed, granted, and denied by the particular physician who reviewed plaintiff’s claim. In addition, the court permitted discovery of documents pertaining to MetLife’s training of its medical staff regarding handling of disability claims generally and claims involving plaintiff’s particular medical issue. With respect to both discovery requests, the district court permitted the discovery because the plaintiff’s request was found to be narrowly tailored.

However, courts have denied discovery of past treatment of similar claims where the request is overly broad and/or not sufficiently tailored to the claim at issue. For example, in Price v. Hartford Life & Accident Insurance Co., -- F. Supp. 2d --, 2010 WL 3998039 (E.D. Mich. Oct. 12, 2010), although the district court permitted limited discovery as to the conflict of interest, the court denied plaintiff’s discovery request seeking broad information concerning all claims made under the policy at issue for the last 10 years. Similarly, in Heim v. Life Insurance Co. of North America, No. 10-CV-1567, 2010 WL 5300537 (E.D. Pa. Dec. 22, 2010), the district court denied plaintiff’s request for all of LINA’s Reading Hospital and Medical Center underwriting files as overly broad and not tailored to assess potential procedural and structural conflicts and the influence of such conflicts on the denial of plaintiff’s claim for benefits.

Relationships Among the Decision-Making Entities. The third area of conflict of interest discovery that has been permitted relates to the relationships among the various entities that contract with respect to claims administration. For example, in Price, the district court permitted identification of each doctor, file reviewer, or surveillance company who provided information considered in the claim, and information on the contractual relationship, fees paid, and frequency of their employment by Hartford. In Hackett v. Standard Insurance Co., No. CIV-06-5040-JLV, 2010 WL 1494772 (D. S.D. Apr. 14, 2010), the district court permitted discovery as to the financial incentives for denying claims and the relationship between the administrator and the vendors it hired to review claims. In Zewdu, the district court permitted discovery on MetLife’s compensation arrangement with the retained physician. Similarly, in Benson v. Hartford Life & Accident Insurance Co., 724 F. Supp. 2d 1187 (D. Utah 2010), the court afforded discovery to investigate whether conflicts of interest between Hartford and University Disability Consortium (the entity charged with oversight of employer provided benefits), or Hartford and the employer, unfairly biased the review of claimant’s disability status. In Sullivan v. Deutsche Bank Americas Holding Corp., No. 08-CV-2370, 2010 WL 391821 (S.D. Cal. Feb. 2, 2010), the district court permitted limited discovery of performance evaluations for the 11 individuals involved in the evaluation of plaintiff’s claim for long-term disability benefits over a three-year period.

Presence of Conflict of Interest Does Not Necessarily Mean Discovery

To further complicate the development of the law as to what, if any, conflict of interest discovery is permitted, the Eighth Circuit has departed from the majority of circuit courts, based on language in Glenn that suggests that the mere presence of a conflict of interest is an insufficient reason alone to expand the administrative record. In Atkins v. Prudential Insurance Co., No. 09-CV-3561, 2010 WL 5060613 (8th Cir. Dec. 13, 2010), the Eighth Circuit affirmed the district court’s rejection of plaintiff’s claim that discovery was warranted as to Prudential’s alleged structural conflict. Specifically, plaintiff “sought materials regarding how disability claims are adjusted, whether there were additional oral communications, how claims like Atkins’ have been handled in other cases, and the procedures Prudential has followed to ensure consistent decision making in similar cases.” The Eighth Circuit agreed with the district court that the administrative record was sufficient to permit a fair evaluation of whether Prudential’s decision to deny plaintiff’s disability claim was arbitrary and capricious.

Similarly, in Jones v. ReliaStar Life Insurance Co., 615 F.3d 941 (8th Cir. 2010), the Eighth Circuit affirmed the district court’s order denying plaintiff’s request for discovery to explore the insurer’s conflict of interest. In so doing, the court emphasized that, in ERISA cases, the general rule is that review is limited to evidence that was before the administrator, and plaintiff presented no convincing reason to create an exception in this case, especially where the insurer conceded that it was also the administrator of the plan, so discovery was not needed to establish a conflict.

Proskauer’s Perspective

Examination of the post-Glenn case law on structural conflicts and what, if any, conflict discovery should be granted leads to the realization that the seemingly simple questions posed by the Supreme Court do not lend themselves to simple, rigid answers. The rulings in Anderson and Parsons indicate that determining whether a Taft-Hartley fund suffers from a structural conflict may require more than the application of a simple economic test of who is paying the benefits. Similarly, the rulings in the discovery cases indicate that, while there may be a consensus on what types of discovery may be permitted, the question of how extensive the discovery should be does not lend itself to a uniform answer. Rather, the answer needs to be decided on a case-by-case basis, based on the facts and circumstances presented.

The hope that Glenn would generate uniformity and predictability in benefits claim litigation may thus be short-lived. For plan counsel, this is an unfortunate development because it means that the defense of these types of cases will continue to be risky and expensive.