Semi-presidentialism: The logic of institutional conflict and implications for public administration design

The dissertation examines the effects that the constitutional choice of semipresidentialism has on the political process and bureaucratic design in post-Communist democracies. The first part of the dissertation analyzes how the variation in semipresidential constitutional norms and party organization in parliament affects the functioning of major government institutions: the presidency, the legislature, and the cabinet. I introduce a multiple principal-agent analytical framework to explain the patterns of interactions among the political actors who comprise these institutions. I utilize a number of tools developed in spatial and game-theoretic modeling to generate a set of testable propositions about the effects of semipresidential constitutional norms on the politicians' behavior. Cabinet identity, cabinet stability and the likelihood of institutional conflict under semipresidentialism are found to be influenced by the specific provisions of constitutional design.; The second part of the dissertation explores the link between the design of "grand" institutions and the organization of public bureaucracy. To understand how the institutional interactions under semipresidentialism affect the organization and functioning of public bureaucracy, I rely on two research strategies. One is a comparative case study. The other is large-N quantitative analysis. Both research strategies lead to the similar conclusions: a semipresidential constitutional framework produces powerful disincentives for the presidents and prime ministers to engage in efficiency-enhancing reform of central government. The research findings show that semipresidential regimes have more cumbersome structures of central bureaucracy and larger cabinet size than parliamentary regimes. Semipresidential institutions are demonstrated to have adverse effects on a country's ability to restructure its executive government.