110 Yale L.J. 1187 (2001)

Narrowly construed, Pennhurst is a sensible (even if not necessary) process-based limitation on Congress's power to bind states to costly burdens. If read to mean that a state can never be bound by a grant condition when the statute itself does not unmistakably speak to a particular set of circumstances, however, Pennhurst becomes a substantive limitation on federal authority that significantly impairs Congress's ability to accomplish national goals. And if one reads Pennhurst to support the accountability model, then Pennhurst not only unduly impairs federal authority, but also undermines the interests served by Chevron; indeed, so read, Pennhurst would jeopardize the functioning of the administrative state itself. As I have explained, the question whether a reasonable agency interpretation that postdates a state's receipt of funds should bind the state is a close one. A conclusion that it should, although plausible, would be easier to swallow if some exceptions to the rule were created. Perhaps, for example, a state should not, notwithstanding an express waiver of Eleventh Amendment immunity in accepting federal funds, be held liable for damages in a private suit challenging conduct that is unlawful only under the agency's belated interpretation. And perhaps, if one were to adopt an approach of categorically deferring, there could be an exception when the agency reverses the interpretation that was in force when the state accepted funds. At bottom, however, the question requires a difficult balancing between values that are not easily quantified: state decisionmaking autonomy, on the one hand, and the separation of powers and the application of specialized governmental expertise to concrete problems, on the other. Whichever approach is correct, the question has assumed increased importance as the Supreme Court has closed (or at least erected barriers to) other avenues for the exercise of federal authority over the states. As the federal government increasingly turns to conditioned spending as a means of accomplishing national objectives, this question becomes correspondingly less academic. Its resolution will provide a degree of welcome clarity both for regulators in the federal government and for state recipients of federal funds.

110 Yale L.J. 1287 (2001)