Prime ministers, presidents, and redistribution: a formal model of redistributive responsiveness and some evidence

This paper analyzes both theoretically and empirically how variation in the form of democratic government - presidential or parliamentary - shapes the responsiveness of elected policymakers to economic inequality. The paper provides an extensive form model of how executive-legislative institutions, market inequality, and party polarization jointly affect general-interest redistribution. In the basic set-up, there are two policy-motivated parties - a low-tax (conservative) party and a high-tax (socialist) party – that compete in multiple electoral districts under uncertainty about the position of the median voter. In the electoral stage of the game, parties announce binding policy platforms and the election determines the composition of the legislature. After the formation of a government, bargaining over policy takes place. The key institutional difference is that the chief executive has strong proposal rights under parliamentary government (in the form of a confidence vote procedure) and weak proposal rights under presidential government. A main theoretical result is that, holding the partisan composition of government and the level of market inequality fixed, there are no generic differences in redistribution across the form of government. Institutional effects emerge only under divided (or coalition) government. Altogether, the model differs fundamentally in its assumptions and implications from the canonical model of the economic consequences of the form of government (Persson, Roland, and Tabellini 2000). The empirical part of the paper tests some of the implications of the model using data on top-income shares and fiscal policy in a panel of 20 countries.