David Epstein and Sharyn O'Halloran produce the first unified theory of policy making between the legislative and executive branches. Examining major US policy initiatives from 1947 to 1992, the authors describe the conditions under which the legislature narrowly constrains executive discretion, and when it delegates authority to the bureaucracy. In doing so, the authors synthesize diverse and competitive literatures, from transaction cost and principal-agent theory in economics, to information models developed in both economics and political science, to substantive and theoretical work on legislative organization and on bureaucratic discretion. Professors Epstein and O'Halloran produce their own deductive specification of the conditions for making or delegating policy, gather a rich, original data set on delegation and discretion in the postwar era to test the propositions derived from their model, and devise appropriate statistical tests to assess the validity of their propositions. With implications for the study of constitutional design, political delegation, legislative organization, administrative law, and the role of the executive in policy making, this book redefines the study of legislative-executive relations under separate powers.

‘This is an important book, simultaneously an original argument and a synthesis of a quarter-century of work on legislative-executive relations. Using a variety of tools, ranging from juicy examples and toy models to systematic empirical analysis and formal theory, Epstein and O’Halloran have crafted a persuasive approach to the analysis of policy making in separation-of-powers regimes. It will surely influence the next generation of work.’

Kenneth A. Shepsle - Harvard University

‘The creativity and care with which the data in this book were collected and analyzed sets a standard for empirical political science research. Epstein and O’Halloran’s combination of theory and data yields an interesting and compelling case for their interpretation of delegation. This work is state of the art political science.’

David Brady - Stanford University