Cozen O’Connor v. Tobits: A Questionable Resolution of Some Complicated Choice of Law Problems

Thanks to the ever-alert Chris Geidner, I saw this post-DOMA decision by a federal district court in Pennsylvania, which is a good example of how complicated the post-DOMA choice-of-law problems can be (and how hard it is to get them right).

The case is an interpleader action dealing with ERISA problems (bear with me!) arising out of the death of Ms. Sarah Farley. Ms. Farley worked at a Pennsylvania law firm and married another woman, Ms. Tobits, in Canada. At the time of the marriage until Ms. Farley’s death, the couple were residents of Illinois. The firm’s ERISA plan, however, contains a clause saying that it is governed by Pennsylvania law (to the extent it isn’t preempted by federal law).

Under the rules of ERISA and the terms of the plan, Ms. Tobits gets the proceeds of the plan if she is Ms. Farley’s “surviving spouse.” If she isn’t, Ms. Farley’s parents get it. So the question is what law to look to — Pennsylvania law, Illinois law, Canadian law, something else? — to figure out if Ms. Tobits is a spouse.

The district court concluded that Ms. Tobits was a spouse. For reasons that are not clear, it seems to have decided that what matters is how the marriage would have been recognized in Illinois. (At times it notes that Illinois was “the couple’s place of domicile,” but earlier it asserted flatly that “Post-Windsor, where a state recognizes a party as a ‘Surviving Spouse,’ the federal government must do so with respect to ERISA benefits …” which implies that maybe another state would work too. It’s not clear why it thought Canada wasn’t enough.) It then concluded that Illinois law recognized the marriage, and so the Illinois marriage was incorporated into federal law.

Both of those conclusions are questionable. They might be right, but it’s not clear, and would require more explanation than the district court gave.

1. As for which state: The most natural thing to do in interpreting an ERISA plan is to do what the plan says, which is to follow Pennsylvania law. The district court’s explanation for why it didn’t do this is that “ERISA pre-empts Pennsylvania law entirely.” But that isn’t right. ERISA doesn’t contain a complete definition of spouse, or a general choice of law provision. The district court itself looked to a state law, which wouldn’t make sense if ERISA pre-empted state law entirely.

The district court could have said that “the federal common law of choice of law” controls instead of the plan’s choice of law provisions. That’s what the Sixth Circuit did in a similar ERISA case, though it didn’t involve same-sex marriage, DaimlerChrylser v. Durden, which Ms. Tobits urged the Court to follow. But the district court didn’t go that route, and a more recent Supreme Court decision says that federal courts shouldn’t create federal common law that conflicts with the plain terms of an ERISA plan. (These cases are discussed at pages 1396-1397 & 1420 of my DOMA article.) This would have been a better route than the one the district court took, but it still would have been dubious.

2, In any event, once the court decided to use Illinois law, it’s not clear to me that it did so properly. Illinois does not currently celebrate same-sex marriages, and an Illinois law enacted in 1996 says that “A marriage between 2 individuals of the same sex is contrary to the public policy of this State.” Normally that would mean that Illinois does not recognize the marriage.

But here things get slightly more complicated. A more recent Illinois law does allow same-sex couples to create a civil union, and it also says that a same-sex marriage “legally entered into in another jurisdiction, shall be recognized in Illinois as a civil union.” That means that the Canadian marriage is recognized in Illinois as a civil union.

Ordinarily, that would probably not be enough. Federal law and ERISA usually ask whether the parties are marriednot simply whether they have a lawful civil union. But then there’s another twist. A definitional provision of the Illinois civil union law says:

“Party to a civil union” means, and shall be included in, any definition or use of the terms “spouse”, “family”, “immediate family”, “dependent”, “next of kin”, and other terms that denote the spousal relationship, as those terms are used throughout the law.

That definitional provision means that Illinois actually defines “spouses” to include those in civil unions, even if it has not defined the word marriage to include them.

Even so, being a “spouse” under Illinois law is still probably not enough to qualify under the terms of the ERISA plan, which says that “‘Spouse’ means the person to whom the Participant has been married …” So even under the district court’s questionable choice-of-law assumption, the decision still seems probably wrong.

Now, I’ve done my best to follow the briefing and what I know about ERISA and choice of law, but I’m not sure I’m right about these things. But I am sure that these issues are quite complicated, and quite a bit harder than the district court made them seem. And I suspect we’ll see a lot more hard choice-of-law questions as the post-DOMA litigation keeps playing out.