In the March 1988 issue of this
Review, B. Dan Wood invoked a principal-agent perspective to establish the importance of democratic hierarchies in shaping the outputs of public bureaucracies. He tested the model with air pollution enforcements of the Environmental Protection Agency (EPA) between 1977 and 1985. His results suggested some limitations on bureaucracies' responsiveness to elected political institutions, consistent with a principal-agent perspective with emphasis on the agent. Brian J. Cook challenges features of the design and empirical testing of Wood's hierarchical model. He questions the accuracy of Wood's characterization of the extent of EPA autonomy and power. Cook argues that a proper model requires recognition of the multiple principal nature of the U.S. system, and he advocates consideration of the normative foundations of principal-agent theory. Wood responds to both the substantive and methodological issues raised and suggests an eclectic approach in future research.