Files in this item



application/pdfNing_Jiang.pdf (2MB)
(no description provided)PDF


Title:Essays in political economy
Author(s):Jiang, Ning
Director of Research:Polborn, Mattias K.
Doctoral Committee Chair(s):Polborn, Mattias
Doctoral Committee Member(s):Deltas, George; Gahvari, Firouz; Krasa, Stefan
Department / Program:Economics
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Differentiated Candidates
Political Competition
Sequential Games
Ideal Points
Abstract:This dissertation studies game theoretical models for political competition and a class of coordination games, it also includes a chapter focusing on estimation of the U.S. legislators' ideal points in multiple meaningful dimensions. The first chapter presents a formal political competition model with differentiated candidates. In this model, the government expenditure must be financed through a proportional income tax, candidates in an electoral competition are exogenously differentiated in how to allocate tax revenue on two public goods when elected but still are able to propose different tax policy to compete for voters. Citizen voters also have different trade-offs between the two public goods. In equilibrium, both candidates compete for support from a common cut-off voter type, who is indifferent between the two candidates and in general is not the expected median voter. All other voter types have strict preference over exact one of the two candidates. Conditions are characterized under which policy convergence and divergence may emerge respectively in equilibrium. From a welfare perspective, the equilibrium policies are not ex-ante majority efficient. The model also predicts that more extremeness in spending government revenue on one public good is detrimental for candidates. The second chapter examines sequential coordination games where players can decide whether to join in an adventure and when to join in. A player opting to delay is able to observe his opponent's action before him. The first mover has a privilege to claim more share of benefit from the adventure but do so at his own peril, in the sense that if the other player does not follow him to join in, an extra cost will be imposed. It studies how the introduction of endogenous timing and reward for the first mover affects coordination games with asymmetric information. An equilibrium in the form of monotone strategy with symmetric cutoffs exists. It also shows that harsher punishment can suppress a player's incentive to move first, but give rise to more follow-up from the other player. Similarly, more reward for the first mover leads to higher probability of moving first, but less follow-up. As private signals become infinitely informative, endogenous timing has positive welfare improvement on coordination with asymmetric information. Numerical analysis demonstrates that endogenous timing and the increasing of reward for the first mover can effectively help coordination. The third chapter proposes a novel topic-factorized ideal point estimation model for a legislative voting network extracted from the U.S. Congress roll call voting data. Traditional ideal point models, either one-dimensional or multi-dimensional, do not give explicit meaning to the latent policy space. Utilizing the text information contained in bills, a unified model that combines voting behavior modeling and topic modeling is presented, which directly estimates the ideal points of legislators and bills for multiple meaningful dimensions. Besides, the generation of topics are guided by the voting records in addition to the bill texts. An iterative estimation strategy is proposed to learn the topics of bills as well as the topic-factorized ideal points of legislators and bills. By comparing with the state-of-the-art ideal point estimation models, the new method has a much better explanation power in terms of held-out log-likelihood and other measures.
Issue Date:2015-01-21
Rights Information:Copyright 2014 Ning Jiang
Date Available in IDEALS:2015-01-21
Date Deposited:2014-12

This item appears in the following Collection(s)

Item Statistics