Argument preview: When can an ERISA limitations period start to run?
![](http://sblog.s3.amazonaws.com/headshots/tejinder-singh.jpg)
on Oct 14, 2013 at 5:08 pm
On Tuesday, October 15, the Justices will hear argument in Heimeshoff v. Hartford Life & Accident Insurance Co., in which they will consider whether the statute of limitations for a federal lawsuit alleging the wrongful denial of benefits under the Employee Retirement Income Security Act of 1974 (ERISA) can begin to run before the beneficiary has completed an administrative procedure that is a mandatory prerequisite to filing the lawsuit, when the ERISA plan documents themselves specify this earlier accrual date.
Petitioner Julie Heimeshoff says “no”: such an “approach is not only incompatible with the text, structure, and purposes of ERISA,” which requires beneficiaries to exhaust their administrative remedies before suing, “but it violates long-settled accrual rules established by this Court.” Moreover, by injecting uncertainty into when the statute of limitations would accrue and how it would be applied, it “deals a wild card into a regime that is designed to provide a uniform and predictable path to the fair resolution of benefit claims.” Respondent Hartford Life & Accident Insurance Company (“Hartford”) responds that ERISA plans are ordinarily enforced according to their terms, and the parties can agree, if they wish, to start the statute of limitations before the administrative process runs its course. Hartford argues that by pressing for a contrary result, “[p]etitioner asks this Court to rewrite or ignore a plan term based on no more than a sense that the term seems ‘odd’ or potentially inconsistent (in theory, though not in practice) with notions of fair remedial policy. Categorical invalidation of plan terms on such a basis would upend parties’ reasonable reliance interests and defeat the purposes of ERISA’s written plan requirement.”
There is nothing even approaching controlling law on this question, so all you fans of ERISA and insurance-related policy arguments should get ready for a very exciting morning. Are both of you ready? Let’s begin.
Heimeshoff worked as a senior public relations manager for Wal-Mart, which offered a group disability policy administered by Hartford and governed by ERISA. The plan documents state that any lawsuit alleging a wrongful denial of benefits must be brought within three years of the time that a beneficiary is required to submit “proof of loss” – an explanation and substantiation of the claim for benefits, submitted near the start of the claims process – to the plan administrator.
Heimeshoff alleges that in June 2005, she became disabled and could no longer work. She filed a claim for disability benefits on August 22, 2005. Her proof of loss was due on December 8, 2005. Heimeshoff went through the administrative claims process, including significant back and forth with the plan administrators, obtaining a medical report, and an internal appeal, which involved multiple extensions. Her claim was ultimately denied on November 26, 2007.
On November 18, 2010, just shy of three years after the final denial, Heimeshoff sued in federal district court to challenge Hartford’s adverse decision. The district court dismissed her action as time-barred, and the Second Circuit – applying circuit precedent in a brief, per curiam opinion – agreed. The court of appeals held that for her claim to be timely, she would have had to file her claim within three years of the date that proof of loss was due. The Supreme Court granted certiorari.
Heimeshoff argues that a statute of limitations cannot accrue until the plaintiff has a valid cause of action – in this case, after the final denial. She contends that this is the ordinary rule under practically every federal statute, and that while Congress is free to alter it, congressional silence means that it intends for this background presumption to govern.
Heimeshoff further urges that ERISA’s remedial scheme contemplates a nonadversarial administrative process before litigation, and that it would unduly burden this process if the statute of limitations ran during its progression. In the extreme case, it’s possible to imagine that the administrative process could last three years, which means that the beneficiary would have no chance to sue at all. But even in a less striking case, time spent in the administrative process would trade off one to one with time to sue. Thus, beneficiaries would have an incentive to cut the administrative process short to get to court before time runs out, and it would give plan administrators an incentive to drag their feet during the claims process to make it more difficult for a beneficiary to sue in the event of a denial. Deferring the accrual date until after the final denial solves this problem because no matter what happens in the administrative proceeding, the time to sue does not begin to run until it is over.
Hartford begins its response by asking the Court to imagine a hypothetical plan specifying that the statute of limitations shall be one year from the date of the final denial. The parties agree that this would be totally fine. But here’s the curious thing: if Heimeshoff’s claim was under that hypothetical, unobjectionable plan, then she would have had less time to sue than she had in this case. The hypothetical plan would have given Heimeshoff until November 26, 2008 to sue. But under the plan that actually governs her claim, she had until December 8, 2008 to sue. How can Heimeshoff complain about the statute of limitations, Hartford argues, when she had more time than she would have under a plan that everybody agrees is valid?
Hartford also notes that the norm in ERISA cases is to respect determinations made in the plan documents. Here, the plan document sets an accrual date and a length for the statute of limitations. Even if a limitations period would ordinarily run after a final denial, the parties should be able to contract around that if they choose to. That would be the case in any private contract action, and there is no reason to treat ERISA plans differently.
Moreover, Hartford points out – and Heimeshoff agrees – that measuring the statute of limitations from the date that proof of loss is due is not only commonplace, but often required by state insurance laws. The accrual provision in this ERISA plan is practically copied and pasted from accrual provisions that are indisputably valid (if not required) under state law.
Hartford responds to Heimeshoff’s arguments about inequity by arguing that if the administrative process runs for too long, the statute of limitations could be tolled, or courts could determine that it was not reasonable to bar an action based on the facts of the case. These equitable exceptions to the statute of limitations, Hartford argues, address any concern that comes from setting the accrual date too early.
In both her opening and reply brief, Heimeshoff responds to these arguments by saying that the key issue is not how long the beneficiary has to sue, but instead whether she can determine the length of that time period with any reasonable accuracy at the start of the claims process, and plan accordingly. Uncertainty and unpredictability, Heimeshoff argues, are incompatible with ERISA’s remedial scheme. They impose unnecessary costs, and will ultimately lead to unnecessary litigation as beneficiaries at the margins try to prove that the statute of limitations should be tolled in their particular case, or why it would be unreasonable to apply it as written. Heimeshoff makes a similar argument in response to Hartford’s claim about state law: none of the state insurance regimes that Hartford points to involves mandatory administrative exhaustion like ERISA does. Therefore, there is no administrative regime that would be compromised by allowing the statute of limitations to accrue from the date that proof of loss is due.
Heimeshoff also maintains that Hartford’s hypothetical plan proves why its approach is pointless, and serves only to inject complexity into a regime that would operate far more smoothly and efficiently if all statutes of limitations begin to run from the date of the final denial. If Hartford really wants shorter statutes of limitations, Heimeshoff argues that the proper mechanism is not to start running the statute of limitations early and open the door to unpredictable inquiries about tolling or reasonableness; instead, the proper mechanism is to set a shorter limitations period with a predictable start date. According to Heimeshoff, this is a win-win because it gives all sides clarity, and eliminates ripple effects on the administrative process. In the alternative, Heimeshoff argues that any time she spent in the administrative process should be tolled.
The United States filed a brief supporting Heimeshoff. The government’s arguments largely mirror hers: it argues that statutes of limitations cannot begin to run until a claim accrues, and also that ERISA’s particular remedial scheme is incompatible with an accrual date that begins before the administrative process ends. It dismisses Hartford’s attempt to invoke state law norms as incompatible with ERISA.
In this case, the Justices granted certiorari from a per curiam order of the Second Circuit applying settled circuit precedent. Often, such grants suggest that something about settled circuit precedent rubs the Justices the wrong way. I would expect the Justices to ask Catherine Carroll, appearing for Hartford, to reconcile her client’s proposed approach with ERISA’s remedial scheme, and to elaborate on why it wouldn’t make more sense to adopt Heimeshoff’s rule and shorten the statute of limitations to achieve the same outcome. I would expect the Justices to ask Matthew Wessler, appearing for Heimeshoff, and Ginger Anders, arguing for the United States, some tricky hypotheticals – for example, if a plan document states that the statute of limitations accrues from the date when proof of loss is due, but gives the beneficiary twenty years to sue, then how could that possibly compromise the administrative process or lead to an unfair result?
Another topic that may arise is the dearth of data on this issue. Put simply, nobody really knows how long ERISA administrative claims typically take to resolve (indeed, it likely varies significantly between plan administrators), and so nobody knows if running the statute of limitations from the start of that process is likely to bar many claims. The absence of hard numbers will harm whichever side the Justices think ought to carry the burden of proof, and so I expect the skilled advocates for each side to try and force each other into the position of supporting their claims with evidence.
For those looking to learn something about advocacy, I predict that this argument will provide a very good illustration of how to present policy arguments to a legal audience. As I mentioned, neither party can stake a claim to controlling precedent. In all likelihood, the side that wins will be the one that convinces the Justices that its rule and result make the most sense in this marketplace. Such presentations are difficult to execute because without black-letter law to ground them, they invite the Justices to inject their own intuitions into the debate, and an argument can quickly get away from an advocate. It should be a good show – for an ERISA case.