Policy Outcomes and Efficiency in Political Systems

University Of Pennsylvania, Philadelphia PA

Investigators

Abstract

This project models political systems and studies the efficiency properties of their out-comes. The first part investigates how, in a Downsian model of electoral competition, the number of candidates affects the efficiency properties of the equilibrium. Initial results suggest that, the larger the number of candidates, the larger is the incentive for candidates to engage in inefficient tactical redistribution. This suggests that, in political systems (or countries) with a large number of parties, political platforms should feature a greater fraction of pork-barrel projects, tax cuts, and other redistributive policies, at the expense of more efficient (but less targetable) public goods. The second part builds a model that contrasts two electoral systems, a proportional system where seats in an assembly are assigned proportionally to the vote shares, with a "winner--take-all" system where the party with the highest number of votes receives all the spoils of office (for instance, the presidential election in the US has this feature). This project shows that politicians competing in a proportional system tend to favor policies that are less risky, in terms of the way that they are perceived by voters, than in a winner-take-all system. Because politicians in a winner-take-all system choose to compete on riskier platforms, vote shares should display greater variability than in a proportional system. Preliminary empirical evidence seems consistent with this prediction. The third part investigates a way to obtain uniqueness of equilibrium in a plurality voting game with N candidates (and strategic voting).