Essays in Political Economy and Mechanism Design

Abstract

This thesis studies extended models of choice in political economy and mechanism design. In some situations, economic agents' decision problem does not fit within the traditional ``economic man'' framework of expected-utility maximization.

The first chapter looks at citizens' indirect political opposition to a dictator when the political institutions do not allow open elections and the dictator uses physical force to punish protesters. The first finding states that as the dictator's power is weakening over time, citizens' anticipation of his eventual downfall makes the uncertain time-frame of the revolution happen sooner. Secondly, the dictator is most oppressive when his power is moderate. Thus, the policing is non-monotone in the state: little in good times, progressively more during hardship right up to the tipping point when it is completely withdrawn. The government puts up an intense short-term fight to stay in power, even though the times are changing and its authoritarian grip is loosening.

The second chapter also looks at political choice -- this time it is a democracy with loss-averse voters and ``career-concerned'' politicians. This rational-expectations model confirms the empirical finding that the voters prefer incumbents during good times and take a chance on challengers when experiencing a bad shock. The model is also consistent with the second empirical finding that while incumbent's average disaster relief increases in the magnitude of an unrelated crisis, their average probability of winning decreases. The politician's decision involves a tradeoff between personal rent and increasing the probability of being elected by choosing a higher signal. Therefore, when the voters suffer a loss from a natural disaster, the incumbent cuts his rents and provides more public goods as the electoral race tightens. By combining ``career-concerned'' incumbents with behavioral voters, the same model can explain both facts, whereas individually these parts are not enough.

The third chapter looks at social choice from a mechanism designer's point of view, where some of the constituents make mistakes under the exclusive information setting. This theoretical chapter derives novel necessary and sufficient conditions for full implementation (matching desirable outcomes to equilibria), even when the faulty players lie about their private information.