The facts, as recited by the Court, were as follows (citations omitted), and concern the rights of the surivors of William Kennedy, whose employer ERISA plan had an SIP part and a Pension part:
In 1971, William married Liv Kennedy, and, in 1974, he signed a form designating her to take benefits under the SIP, but naming no contingent beneficiary to take if she disclaimed her interest. William and Liv divorced in 1994, subject to a decree that Liv “is . . .divested of all right, title, interest, and claim in and to . . . [a]ny and all sums . . . the proceeds [from], and any other rights related to any . . . retirement plan, pension plan, or like benefit program existing by reason of [William’s] past or present or future employment.” William did not, however, execute any documents removing Liv as the SIP beneficiary, even though he did execute a new beneficiary-designation form naming his daughter, Kari Kennedy, as the beneficiary under DuPont’s Pension and Retirement Plan, also governed by ERISA.
On William’s death in 2001, petitioner Kari Kennedy was named executrix and asked DuPont to distribute the SIP funds to William’s Estate. DuPont, instead, relied on William’s designation form and paid the balance of some $400,000 to Liv. The Estate then sued respondents DuPont and the SIP plan administrator (together, DuPont), claiming that the divorce decree amounted to a waiver of the SIP benefits on Liv’s part, and that DuPont had violated ERISA by paying the benefits to William’s designee.2
2 The Estate now says that William’s beneficiary-designation form for the Pension and Retirement Plan applied to the SIP as well, but theform on its face applies only to DuPont’s “Pension and Retirement Plan.” In the District Court, in fact, the Estate stipulatedthat William “never executed any forms or documents to remove orreplace Liv Kennedy as his sole beneficiary under either the SIP or [a plan that merged into the SIP].”
Thus, despite the fact that the right to receive SIP benefits from the plan was expressly waived in a divorce decree (in a portion apparently not submitted as a QDRO), and despite the fact that William Kennedy tried to designate a new beneficiary for the plan, although he screwed up by not realizing that his plan had two components, the ex-spouse gets the plan benefits because William Kennedy failed to fill out a new beneficiary form specifically for this part of the plan.
It also bears mention that under the state law of every state, divorce decrees (and indeed any documents authorized by an ex-spouse) can be considered in litigation over beneficiary designations, and that all non-probate beneficiary designations in favor of an ex-spouse are revoked by operation of law upon divorce in many states, when ERISA pre-emption does not apply.
The notion that ERISA covered benefits should be administered without regard to generally applicable state law is bad policy and has produced a host of bad results like this one. (The Court notes without addressing the fact that its isn't even clear that homicide would diqualify a beneficary, despite the fact that it does so under every state's law.
Justice Souter, writing for the unanimous Court isn't departing from precedent in his ruling, although the law probably had enough wiggle room in it for a credible decision to be written the other way. But, his policy arguments set forth below, are unimpressive (citations omitted):
The point is that by giving a plan participant a clear set of instructions formaking his own instructions clear, ERISA forecloses any justification for enquiries into nice expressions of intent, in favor of the virtues of adhering to an uncomplicated rule: “simple administration, avoid[ing] double liability, and ensur[ing] that beneficiaries get what’s coming quickly, without the folderol essential under less-certain rules."
And the cost of less certain rules would be too plain. Plan administrators would be forced “to examine a multitude of external documents that might purport to affectthe dispensation of benefits,” and be drawn into litigation like this over the meaning and enforceability of purported waivers. The Estate’s suggestion that a plan administrator could resolve these sorts of disputes through interpleader actions merely restates the problem with the Estate’s position: it would destroy a plan administrator’s ability to look at the plan documents and records conforming to them to get clear distribution instructions, without going into court.
The Estate of course is right that this guarantee of simplicity is not absolute. The very enforceability of QDROs means that sometimes a plan administrator must look for the beneficiaries outside plan documents notwithstanding §1104(a)(1)(D); §1056(d)(3)(J) provides that a “person who is an alternate payee under a [QDRO] shall be considered for purposes of any provision of [ERISA] a beneficiary under the plan.” But this in effect means that a plan administrator who enforces a QDRO must be said to enforce plan documents, not ignore them. In any case, a QDRO enquiry is relatively discrete, given the specific and objective criteria for a domestic relations order that qualifies as a QDRO,12 requirements that amount to a statutory checklist working to “spare [an administrator] from litigation-fomenting ambiguities[.]" This is a far cry from asking a plan administrator to figure out whether a claimed federal common law waiver was knowing and voluntary, whether its language addressed the particular benefits at issue, and so forth, on into factually complex and subjective determinations.
12 To qualify as a QDRO, a divorce decree must “clearly specif[y]” the name and last known mailing address of the participant and the nameand mailing address of each alternate payee covered by the order; theamount or percentage of the participant’s benefits to be paid by theplan to each such alternate payee or the manner in which such amount or percentage is to be determined; the number of payments or period towhich the order applies; and each plan to which such order applies. A domestic relations order cannot qualify as a QDRO if it requires a plan to provide any type or form of benefit, or any option,not otherwise provided under the plan; requires the plan to provideincreased benefits; or requires the payment of benefits to an alternate payee that are required to be paid to another alternate payee under another order previously determined to be a QDRO. A plan is required to establish written procedures for determiningwhether a domestic relations order is a QDRO.
These are good and sufficient reasons for holding the line, just as we have done in cases of state laws that might blur the bright-line requirement to follow plan documentsin distributing benefits. Two recent preemption cases are instructive here. Boggs v. Boggs, held that ERISA preempted a state law permitting the testamentary transfer of a nonparticipant spouse’s community property interest in undistributed pension plan benefits. We rejected the entreaty to create “through case law . . . a new class of persons for whom plan assets are to be held and administered,” explaining that “[t]he statute is not amenable to this sweeping extratextual extension.” And in Egelhoff we held that ERISA preempted a state law providing that the designation of a spouse as the beneficiary of a nonprobate asset is revoked automatically upon divorce. We said the law was at fault for standing in the way of making payments “simply by identifying the beneficiary specified by the plan documents,” and thus for purporting to “underminethe congressional goal of ‘minimiz[ing] the administrative and financial burden[s]’ on plan administrators[.]"
What goes for inconsistent state law goes for a federal common law of waiver that might obscure a plan administrator’s duty to act “in accordance with the documents and instruments.” And this case does as well as any other in pointing out the wisdom of protecting the plan documents rule. Under the terms of the SIP Liv was William’s designated beneficiary. The plan provided an easy way for William to change the designation, but for whatever reason he did not. The plan provided a way todisclaim an interest in the SIP account, but Liv did not purport to follow it. The plan administrator therefore did exactly what §1104(a)(1)(D) required: “the documents control, and those name [the ex-wife].”
It is no answer, as the Estate argues, that William’s beneficiary-designation form should not control because it is not one of the “documents and instruments governing the plan” under §1104(a)(1)(D) and was not treated as a plan document by the plan administrator. That is beside the point. It is uncontested that the SIP and the summary plan description are “documents and instruments governing the plan.” Those documents provide that the plan administrator will pay benefits to a participant’s designated beneficiary, with designations and changes to be made in a particular way. William’s designation of Liv as his beneficiary was made in the way required; Liv’s waiver was not.
The litigation allergy expressed in this line of cases has no reasonable basis. A plan administrator can avoid a thicket of conflicting laws simply by having a choice of law provision in the plan. The wealth of litigation of ERISA benefits, because the results are frequently wrong in any reasonable set of laws, requires great resort to the expensive federal courts and to intricate, not widely understood, federal law. The case before the U.S. Supreme Court, where a court entered divorce decree memorializes an express waiver of the benefits in question, puts the lie to the notion that cases like this one have to be evidence intense and costly to litigate. In real life, Will and Trust contests are incredibly rare, there are a couple dozen a year in the State of Colorado which has 4.6 million residents, and most are quite inexpensive to litigate. The only thing that makes ERISA cases complicated is the Alice in Wonderland like law free zone into which it plunged Plan Administrators and everyone who must deal with them.