Overruling Erie: nationwide class actions and national common law.

• Author: Sherry, Suzanna

The Class Action Fairness Act of 2005 (1) (CAFA) reflects a sharp change of direction in contemporary thinking about federalism. It expands federal jurisdiction substantially, placing many more state law claims into federal court. In so doing, it highlights and attempts to resolve the tension that has always existed between state and national interests.

In this Commentary, I argue that the enactment of CAFA amounts to swimming halfway across a river. Professor Linda Silberman's thoughtful and well-argued proposal (2) is a valiant attempt to keep from drowning while treading water in the middle of the river. I suggest that instead of treading water, we should swim the rest of the way.

The river I am talking about is federal mistrust of state judges. And the far shore we are heading toward is Swift v. Tyson. (3) In other words, I want to make the proposal that Judith Resnik describes as "a more energetic claim" than those made so far, (4) that Richard Marcus and Stephen Burbank say is foreclosed by CAFA's legislative history, (5) and that even Samuel Issacharoff and Catherine Sharkey offer only as a descriptive prediction of possible future developments: (6) CAFA should be read as overruling Erie Railroad Co. v. Tompkins, (7) at least for the national-market cases (8) that it places within federal court jurisdiction.

The skeleton of my argument is as follows: First, mistrust of state interference with national markets should be viewed as the norm, and Erie as the aberration. Second, despite its explicit intent to leave Erie in place, Congress squarely based CAFA on this norm. Finally, we cannot have it both ways: either we trust state judges or we do not. To the extent that Congress tried to adopt both positions, courts must find a single consistent interpretation of the statute--they must head for one shore or the other. I know which way I would swim.

Let me start with mistrust of state interference with the national-market economy. Justice Story's opinion in Swift eloquently presented the conventional nineteenth-century view. My favorite quotation, however, comes from a case two decades before Swift. In 1821, Chief Justice Marshall rejected one of Virginia's many attempts to place its own policy above that of Congress: "That the United States form, for many, and for most important purposes, a single nation, has not yet been denied. In war, we are one people. In making peace, we are one people. In all commercial regulations, we are one and the same people." (9) Marshall thus put national economic policies on the same plane as foreign relations--a federalized subject that must be protected from individual state obstruction.

Diversity jurisdiction, properly understood, also fosters a national market and a federalized commercial liability regime. The origins of diversity are famously unclear. (10) Under any reading, however, protecting national-market interests from state interference was a significant purpose behind diversity jurisdiction. (11) But that purpose is in direct tension with Erie: it does not make much sense, in terms of protection from aberrant or biased states, to place cases in federal courts but let state judges make the substantive rules. It is, in Professor Silberman's memorable characterization from an analogous context, paying more attention to where the defendant will be hanged than to whether he will be. (12) Indeed, it is probably not logical to let state legislatures determine the law in cases involving national commerce, and in that sense Swift did not go far enough (but that argument is beyond the scope of this Commentary).

So why did the same generation--and even many of the same individuals--that adopted Article III also enact the Rules of Decision Act? (13) Why, in fact, did the same Congress enact both the predecessor of § 1332 (14) and the Rules of Decision Act? This question is difficult only if we assume Erie's reading of the Rules of Decision Act--but Erie was probably wrong.

As other scholars have shown, the more historically accurate interpretation of the Act is that it was either limited to state legislative enactments (as Swift held), or that it was meant as a directive to apply American rather than English law, not as a directive to apply state rather than federal law. (15) The best textual evidence for the latter interpretation is the use of the term "the laws of the several states" rather than "the laws of the respective states." (16) In eighteenth-century parlance, that distinction...