Essays in Political Economy and Strategic Experimentation

Abstract

This dissertation consists of four chapters on topics in political economy and strategic experimentation.

The first chapter studies the consequences of systems of ``joint'' versus ``separate'' political appointments for accountability and political selection. These systems govern the electoral accountability of political agents that I refer to as the `proposer' and `veto player'. A system which ties the electoral fate of the veto player to the proposer forces the veto player to internalize the consequences of pandering to the voter by speciously rejecting the proposer's policies. Applications are given to parliamentary versus presidential government, as well as other sub-national constitutions.

The second chapter examines how non-majoritarian run-off rules can facilitate entry deterrence, or the electoral domination, of a potential challenger party by established parties. I also bound the challenger's performance across all equilibria. My results corroborate historical accounts of run-off rule adoption in Latin America.

The third chapter, co-authored with Dan Bernhardt, studies the incentives of political agents to refrain from fully exploiting their contemporary political power. We consider a setting in which there is uncertainty about who will hold future power to propose policies and to reject them. Agents may not exploit their immediate power to change policy both if they anticipate either an unfavorable or favorable shift in both future proposal and veto power. We uncover circumstances in which an agent who is opposed to the initial status quo may propose less initial reform than an agent who is favorable to the initial status quo.

The fourth chapter studies the problem of a Principal who experiments with a technology of unknown quality. The Principal invests funds and hires an Agent to work on the project. The Agent has incentives to shirk in order to manipulate the Principal's beliefs about project quality. I show that the Principal may over-invest relative to the first-best. She does in order to induce sufficiently pessimistic beliefs about the project, conditional on early failure, that she quits the project. This denies the Agent rents from the repeated interaction.