A Conundrum on Federal Court Determinations of State Law Issues

In my legal research class, I recently went over the general principle for how a federal court sitting in diversity should determine state law issues. The federal court looks first to see if the issue has been decided by the applicable state court of last resort. In the absence of an opinion from the state court of last resort, it must predict how that court would decide the issue based on input that would be considered by that court.

I came up with a “hypo” for my students:

You are a federal district judge presiding over a diversity case. You are presented with a question of state law on which the high court of the state has not spoken. However, there is a published decision from your circuit court of appeals that predicts how the state’s high court would decide the issue. Must you follow that decision of the court of appeals, or should you decide independently whether the state’s highest court would still reach that decision?

Think the answer is obvious? I don’t think so at all, which is why I thought it was a good hypo. It sets up a collision between the mandatory authority of the court of appeals and the rule for deciding issues of state law.

Then, before I could present the hypo in the next class, I discovered as I scanned the internet for blog subjects last night that my scenario isn’t so hypothetical anymore. . . .

Howard Bashman was having similar thoughts at How Appealing. My attention was caught by a post titled When should a federal appellate court change its prediction of how a state court of last resort would rule on a question of state law?

The case that prompted the post at How Appealing is the Third Circuit case of Jaworowski v. Ciasulli, case no. 05-1423 (June 18, 2007), in which the district court had granted a motion to dismiss on statute of limitations grounds, feeling itself bound by a nearly 20-years-old Third Circuit Court of Appeals determination of the state law question. The New Jersey Supreme Court still hadn’t weighed in on the issue, so the Third Circuit took another look at the issue on appeal and, based on intervening state developments, decided the New Jersey Supreme Court would decide the issue differently than the Third Circuit had predicted it would nearly 20 years earlier.

Here is Bashman’s invitation to consider the hypo:

It is interesting to consider whether it would have been appropriate, in the first instance, for the federal district court to ignore the earlier Third Circuit ruling if the federal district court were confident that the earlier Third Circuit ruling had incorrectly predicted how New Jersey’s highest court would rule on the issue presented, even though New Jersey’s highest court hadn’t yet ruled on the issue.

Of course, a district court is bound to follow the decisions of its circuit court of appeals. But should that extend to questions of how a state’s highest court would decide a state law question? After all, case after case says that a federal court must determine how a state’s highest court would decide the issue, and case after case looks to state intemediate appellate court decisions, state statutes, legislative histories, etc.

It seems to me that a federal court decision on state law has a limited shelf life. Such a determination in the absence of a decision from the state’s high court is, by definition, a prediction subject to being proven wrong. And the older the federal decision, the greater the chance that state law has changed in the interim.

A logical compromise between the two principles would be for the district court to start with the court of appeals decision and then evaluate whether the analysis still holds up in light of intervening events at the state level. If there were no changes in legislation, no new decisions, or other changes that would influence the question, then the district court would be bound to follow the circuit court determination. But if there were intervening changes at the state court level, the district court logically should have to evaluate them in the context of the circuit court’s prior analysis to see if they change the result.

I kow it sounds blasphemous to suggest that a district court should not be bound by its circuit court in any situation, but it seems the practical result in this scenario.