Essays on the Redistributive Consequences of Democratic Constitutions

Abstract

Why do some democracies redistribute more than others? In this dissertation, I examine how executive-legislative and electoral institutions, two fundamental dimensions in the design of democratic constitutions, influence redistributive policy. To do so, I develop and empirically evaluate three analytical models that emphasize that these political institutions matter for taxes and transfers because they shape the outcome of partisan conflict over redistribution. Taken together, the theoretical arguments and empirical findings suggest that candidate selection, the partisan composition of government and the bargaining power of the chief executive constitute an important set of mechanisms through which democratic constitutions influence policy. The emphasis on institutional explanations based on political parties that represent divergent societal interests stands in contrast to the common view in the political economy literature that institutions mainly matter for policy because they shape the incentives of identical rent-seeking politicians.

In the first part, I argue that public spending is lower in presidential than in parliamentary democracies because the separation of executive and legislative power under presidentialism reduces the frequency and bargaining power of left governments. Using cross-sectional and panel data covering a global set of democracies, I find evidence consistent with key implications of this argument. In the second part, I suggest that electoral institutions shape partisan conflict over redistribution differently than previously thought. In particular, parties competing for the power to set redistributive policy in majoritarian electoral systems can use legislative recruitment to make credible appeals to pivotal middle-class voters despite the absence of external commitment. In line with this logic of endogenous commitment, I find cross-national evidence that left parties are more centrist but equally likely to control the chief executive in majoritarian compared to proportional electoral systems. In addition, I exploit a natural experiment in Britain to compare candidate selection under alternative electoral systems. In the final part, I analyze why governments sometimes pursue economic reforms that deviate significantly from their electoral promises and why presidents are more likely than prime ministers to do so. I develop and test the argument that both executive-legislative institutions and economic conditions are crucial determinants of the program-policy link.