Pragmatic Administrative Law

The history of administrative law, according to James Freedman, consists of an extended sense of crisis over the legitimacy of the regulatory state. The specific nature of the crisis has differed in each historical stage depending on the dominant concern of each era, but it has always been related to the difficulty of reconciling the administrative state with traditional American constitutional and political values. In the Reformation of Administrative Law, Richard Stewart discusses the crisis of legitimacy that characterized the 1960s and 1970s and the reforms that were adopted as a result. Stewart was uncertain what might follow the reformation, but now we know. The reformation has been followed by a “counterreformation” that is based on a set of premises that run directly counter for the premises of the reformation. Recently, some scholars, including Professor Stewart, have sought to move beyond the counterreformation, which they find insufficient to produce sound and legitimate government. This literature, like the earlier literature on the reformation and the counterreformation, adopts interest group pluralism as the basis of the administrative process. This essay evaluates the reformation, the counterreformation and the most recent scholarship through a different lens. As I have in other recent work, I propose that the American tradition of philosophical pragmatism offers the best methodology to evaluate and justify the administrative process. This approach leads me to three general conclusions. First, the reformation has been a greater success than Professor Stewart recognizes in the Reformation or his subsequent work. Second, the counterreformation has produced changes in the administrative process that cannot be justified as either improving the rationality of regulation or the legitimacy of the process. Finally, we should be quite cautious about implementing recent proposals by Professor Stewart and others because the available evidence indicates the methods that they favor only work in some specific contexts.