ECONOMIC THEORY WORKING GROUP
The Economic Theory Working Group meets on Fridays 12pm-1:30pm.
The objective of this working group is for students and faculty to present advanced or preliminary version of their work in an informal setting.
The focus is on theory and theoretical parts of the model.
Students are welcome to present papers in the fields including but not limited to
microeconomic theory, game theory, political economy, mathematical economics,
computation, industrial organization, and market design.
You can subscribe to the Econonmic Theory Working Group mailing list
here.
Please email Raluca Ursu (rursu) or Jose Tudon (jtudon) if you want to present.
Below is the list of current and past presentations.
SPRING 2015
05/08/2015: Peter Evangelakis
05/15/2015: Chen Yeh
05/29/2015: Gabriela Antonie
06/05/2015: Hanzhe Zhang
SPRING 2014
- 04/11/2014: Danxia Xie, "Innovation Financing, Systemic Risk and Crisis"
- 04/25/2014: Jose F. Tudon Maldonado, "Price Dispersion With ex ante Homogeneity"
- 05/02/2014: Hanzhe Zhang, "Pre-Matching Gambles"
- 05/09/2014: Jose F. Tudon Maldonado, "Price Dispersion With ex ante Homogeneity"
- 05/16/2014: Hanzhe Zhang, "Marriage Age and Schooling Patterns"
- 05/23/2014: John Hatfield,
- 05/30/2014: Ryan Yuhao Fang,
- 06/06/2014: Peter Evangelakis,
- 06/13/2014: Raluca Ursu,
WINTER 2014
- 01/17/2014: Ryan Fang, Media Bias, Political Polarization, and the Merits of Fairness.
[Show Abstract]
In this paper, we study the economic and social consequences of media bias and evaluate the merits of content regulations proposed to curtail such biases. We present a model in which Bayesian consumers might rationally prefer biased news to more balanced news and profit maximizing media outlets report news with significant biases to attract consumers. Our model fits a range of empirical regularities uncovered by recent studies. We find that, when the costs of consuming news are low, media outlets have the incentives to produce more balanced news in equilibrium, and, when competition intensifies, media outlets produce news with more bias. Thus, we are able to explain both the recent trend of increasing media bias in television news and the earlier trend of decreasing bias in newspapers. Our policy analysis shows that content regulations cannot be justified on the ground of protecting consumer welfare. Such regulations can never lead to Pareto superior outcomes but can lead to Pareto dominated ones. However, we also find that media bias can lead to political polarization, which content regulations might help to mitigate. The effectiveness of such regulations in mitigating polarization depends critically on market conditions. Under some circumstances they might in fact exacerbate polarization.
- 01/24/2014: Hanzhe Zhang, Marriage Age Patterns: A Unifying Theory and Global Evidence.
[Show Abstract]
Age at first marriage is systematically related to personal income earned in later life. Around the world, personal income-marriage age relationship for males has been persistently inverse-U shaped: those who marry earlier and and later earn less than those marry in around a median age. Personal income-marriage age relationship historically has been monotonically positive for females, but recently in the United States it has become inverse-U shaped similar to males'. I construct an equilibrium dynamic matching model to explain the gender-specific patterns and their evolutions over the years. The two key drivers are stochastic labor market returns from human capital investment and differential fecundity that reduces females' desirability in marriage market and interferes with females' investment decision but not males'. In addition to reconciles and extends previous theories on the relationship between marriage age and personal income, the model also sheds light on other patterns related to marriage, such as gender cap in median age at marriage, spousal matching patterns, and marriage market's influence on career choices.
- 03/07/2014: Hanzhe Zhang, Stability Induces Pre-Matching Gambles.
[Show Abstract]
I show that a risk-averse agent may prefer to take an unfair lottery that alters his or her match quality before entering a stable, transferable utilities matching market. The fundamental engine that drives the seemingly risky choice is a stable sorting effect: change in one's marginal payoff results from a changed partner from stable assignment in addition to change in one's own match quality. This effect induces risk-taking behavior regardless of the shape of the surplus function; specifically, complementarity, a key assumption that generates gambles in one-sided hedonic markets, is no longer needed. The risk-taking behaviors are actually socially efficient, and equilibrium wage inequality can be endogenously generated even when all the agents are born homogeneous, suggesting a connection between efficiency and inequality. Furthermore, the choices of lotteries resemble occupational choices, and a dynamic extension of the model offers explanations to the phenomenon that more men choose risky careers than women and sheds light on demographic patterns related to marriage age.
FALL 2013
- 10/11/2013: Aditya Bhave, Optimal Opacity in a General Equilibrium Model of Banking
[Show Abstract]
Opaque structured-finance securities were at the heart of the collapse of the shadow-banking system in 2008. This paper considers the efficiency of aggregate investment in opaque assets in a general equilibrium model of banking. The key mechanism is that crises lead to runs on banks with opaque assets, necessitating fire sales of these assets to banks with transparent assets. There are two resulting externalities. On the one hand, although there is cash-in-the-market pricing of assets during crises, banks do not internalize the fact that investing in transparent assets increases the price of opaque assets, reducing the inefficiency cost of runs on banks with opaque assets. On the other hand, fire sales give rise to a rent-seeking motive for investing in transparent assets. Depending on which externality dominates, there are either too many or too few opaque assets in equilibrium. If there are too few opaque assets, asset purchases by a planner can implement an efficient allocation. If there are too many opaque assets, under certain conditions, the planner would choose to purchase assets when a crisis occurs, even though a commitment to zero asset purchases would be optimal ex ante due to moral hazard considerations. In this case, pro-cyclical bailouts can implement an efficient allocation.
- 10/18/2013: Hanzhe Zhang, Courtship as an Investing Game
[Show Abstract]
We attempt to explain an inverse-U relationship between age at first marriage and lifetime income for men and a positive linear relationship for women: men who marry in the mid- to late-twenties earn the most throughout the lifetime, and women who earn more tend to marry later. Pre-marital investment with stochastic return, I argue, explains the inverse-U relationship and differential fecundity explains the gender difference. Because women cannot wait long to realize investment gains, they invest for immediate and safe returns. Men, on the other hand, make risky investments that take long time to realize and have volatile returns. Non-investors earn less and marry early; the investors who realize the gain quickly earn more and marry early, but those who fail earn less and marry late. Our simple equilibrium overlapping generations matching model also predicts other major patterns and trends pertaining to marriage age. We also prove that a stationary equilibrium exists uniquely and that the equilibrium private investments are socially efficient.
- 10/30/2013: Pietro Bonaldi, Existence of Equilibrium in Hart's Securities Model with Consumption in the First Period
[Show Abstract]
Hart (1974) established necessary and sufficient conditions for the existence of equilibrium in a securities model where investors' beliefs regarding security returns disagree. Under additional assumptions that prevent investors from risking unbounded losses, Hammond (1982) proposed the "overlapping expectations" property as an equivalent reformulation of Hart's conditions. In both models, investors live two periods and trade securities in the first one to maximize their future expected utility. No consumption takes place in the first period. As shown in the present work, however, when consumption in the first period is allowed, the "overlapping expectations" condition is no longer sufficient for the existence of equilibrium. If it is further assumed that investors' beliefs do not depend on security prices, the equilibrium is restored.
- 11/07/2013: Armin Rick, The Benefits of Miscommunication in Communication Games
[Show Abstract]
This paper investigates the effect of communication error on the performance of a communication technology with respect to equilibrium outcomes such as information transmission, communication costs, and welfare. In particular, I analyze a general class of communication games between an informed sender and an uninformed receiver that nests a variety of classic communication technologies such as cheap talk, verifiable messages, and costly signaling. In this class of communication games, communication error can often increase welfare by improving the tradeoff between information transmission and communication costs. Roughly speaking, communication error relaxes the relationship between the sender's communication strategy - which determines communication costs - and the induced distribution of posterior beliefs - which determines the informativeness of the communication mechanism and thus the quality of the receiver's decision making. As a result, communication error may improve the quality of information transmission without increasing communication costs and/or reduce communication costs without changing the quality of information transmission.
I illustrate such gains from communication error in a competitive version of the Spence signaling game. In particular, I characterize optimal communication error structures and derive necessary and sufficient conditions for which optimal communication error can make all types of agents strictly better off than in any equilibrium that is implementable without communication error. This requires perfect information transmission to be sufficiently costly (or not implementable) on the one hand and partial information transmission to be sufficiently valuable on the other. Thus, gains from communication error can be large in many settings in which information transmission is particularly valuable. In these cases, communication error can even transform a communication technology that makes all types of the sender worse off into one that makes everyone better off.
- 11/08/2013: Jin Yeub Kim, Endogenous Choice of a Mediator: Inefficiency of Bargaining
[Show Abstract]
In this paper, I build a theory of how privately informed parties choose mediators, which is a topic that is not well developed in the literature. I consider a bargaining problem in which two parties with private information about their types negotiate the choice of a mediator. A mediator, in my context, is equivalent to a communication mechanism that respects private information. I use a two-pronged approach, cooperative and noncooperative, to show that the selection of a mediator is endogenous and is driven by the "information leakage" problem. With the noncooperative approach, I consider a threat-secure mediator who survives a vote against any alternative mediators. For a benchmark class of examples, I show that the threat-secure mediator exists and is unique. In the cooperative approach, I find that there exists a unique neutral bargaining solution -- an incomplete information version of the Nash bargaining solution. I establish that the selected mediators in both approaches are the same. Moreover, the selected mediator is -- among all the interim incentive efficient mediators -- the worst for the parties ex ante and the least peaceful mediator. Therefore, I argue that the very process of selecting a mediator may exhibit an inherent inefficiency in bargaining if the parties already know their types. This paper represents a first attempt to build a bridge between cooperative and noncooperative game theories in the context of bargaining with incomplete information.
- 12/13/2013: Ryan Fang, Media Bias, Political Polarization, and the Merits of Fairness
[Show Abstract]
We present a model of media bias where Bayesian consumers might rationally prefer biased news to more balanced news. To accommodate such preferences, profit maximizing media outlets report news with significant biases in equilibrium. We apply the model to studying the industrial organization of the market for news and find that, consistent with the history of the US market, low costs of consuming news encourage balance in the news reported in equilibrium, while intense competition encourages bias. The biases in the news reported by media outlets are correlated with the average ideological positions of their readers / viewers. When news consumption costs are low, the media outlets' readership / viewership may overlap, and the overlap is larger if the outlets produce news with similar biases. We find that media bias can lead to political polarization, which, in the presence of externalities, may result in social welfare losses. Content regulations such as the FCC's controversial Fairness Doctrine can help mitigate polarization and correct distortions. However, the effectiveness of such regulations depends critically on market conditions. In some scenarios they might exacerbate rather than mitigate polarization.
SPRING 2013
- 04/26/2013: Seongkyu Park, Increasing Insiders to Improve Efficiency: Disclosure Inquiry-Response on Rumors
[Show Abstract]
Some financial markets has disclosure regulation that the market headquarters discloses an inquiry regarding the rumor information which the relevant firm must also publicly disclose within certain period of time. Since the inquiry is made publicly, some investors in the market may be able to acquire relevant information before the response disclosure by the firm through their private channels. Applying the Kyle (1985) and Holden and Subrahmanyam (1992) model to this idea, I show that the rumor disclosure regulation improves market efficiency by making the price converge to the true value faster. Comparison with the Huddart, Hughes, and Levine (2001) is made where it regulates the insiders to disclose their trading action.
- 05/10/2013: Stephane Wolton, Too Much Ado about Something - Communication Externality in Campaigns
[Show Abstract](with Carlo Prato - Princeton/Georgetown University)
We analyze a model of campaigns where successful communication of candidates' platforms requires effort (a la Dewatripont and Tirole, 2005) from both candidates (senders) and the voter (receiver). We uncover the existence of a communication externality in electoral campaigns. When the voter is more attentive to candidates' messages, her expected equilibrium payoff can decrease. This theoretical result is novel and has broad implications. First, in line with empirical findings, we find that crises are not a good time for reform: crises can lead to delayed or botched reforms. Second, we provide a new rationale for the existence of supermajority requirements, unrelated to the traditional tyranny of the majority argument. We also show that close elections intrinsically differ from lopsided ones, which challenges the external validity of regression discontinuity designs using close elections.
- 05/17/2013: Simcha Barkai
[Show Abstract]
I analyze a generalized moral hazard model with multiple agents and I show that feasible differentiable contracts cannot implement Pareto optimal allocations in Nash equilibrium.
- 06/07/2013: Jin Yeub Kim, Endogenous Choice of a Mediator
[Show Abstract]
In many economic and political situations, two parties in conflict
often get stuck with a "bad" mediator. Despite a significant amount of
the theoretical work on mediation, the theory of how mediators are
chosen is not sufficiently understood. My interest is in understanding
how the parties bargain over the choice of a mediator. My main insight
is that a mediator is endogenously selected badly for signaling
reasons. That is, since the parties do not want to reveal weakness,
they choose the one who mediates in a way that is better for the
strong type of both parties; and thus, a mediator with high
probability of failure making peaceful settlement is chosen, which, in
fact is (interim but) not ex ante incentive efficient. In this sense,
I argue that the very process of selecting a mediator itself create
inefficiencies and show a common feature of distortions in
equilibrium. Further, I relate my results to the solution concepts in
Holmstrom and Myerson (1983), Myerson (1984), and Cramton and Palfrey
(1995).
- 06/14/2013: Armin Rick, The Benefits of Information Garbling in a Spence Signaling Game
[Show Abstract]
Under what conditions can limiting the flow of information improve equilibrium outcomes in signaling games? What are the properties of Pareto-efficient garbling mechanisms if transmission of information is socially valuable? This paper investigates these questions in the context of a competitive labor market model with a costly but unproductive signaling action. Perhaps surprisingly, the equilibrium quality of information transmission is not necessarily monotone in the extent of information garbling in this model and there can be considerable welfare gains for all types of market participants from garbling action signals in an appropriate way-even though the sender's and the receiver's interest are not aligned and all types of the sender care about their individual reputation in the same way. I derive necessary and sufficient conditions for which any equilibrium without garbling is Pareto-dominated by an (informative) garbling mechanism and I describe properties of Pareto-efficient garbling mechanisms. In particular, garbling the action signal in an appropriate way can turn a signaling technology that reduces everyone's welfare into a technology that makes everyone better off precisely when the value of information transmission is sufficiently high.
WINTER 2013
- 01/11/2013: Seth Blumberg, Information and Inefficient Matching
[Show Abstract](with Xan A. Vongsathorn)
Stable matching with transferrable utility is known to be efficient when agents have complete information about potential partners. In recent work, Liu et al. (2012) show that this remains true even when agents are unsure about others' types, via an iterated belief-formation procedure. By relaxing their assumptions, I show that an inefficient match can be stable if the norms of surplus-sharing are such that the surplus enjoyed by the more mysterious side of the market has the opposite curvature from the total match surplus.
- 01/18/2013: Aditya Bhave, The Effect of Fire Sales on Liquidity Hoarding
[Show Abstract]
This paper shows how fire sales of assets cause private incentives for liquidity hoarding to deviate from first best. There are two externalities. First, fire sales might result in too little liquidity hoarding because price-taking agents do not account for the pecuniary externality of their actions on fire-sale prices. Second, fire sales might cause too much liquidity hoarding by encouraging rent-seeking behavior. I demonstrate that capital requirements can address the first externality, while a central counterparty can improve the allocation in the second case.
- 02/15/2013: Jin Yeub Kim, Why and How does Mediation Go Wrong?
[Show Abstract]
I examine the role of mediation in two-party conflicts that might lead to a dispute. I aim to build a game-theoretic model which answers questions of why and how mediators can be bad, and how parties get stuck with bad mediators. In my 'two principals - one agent' setting, an informed agent (a mediator) advises two principals (two parties) on selecting one of multiple proposals or an outside option (status quo). The mediator is privately informed about the proposals' benefits and shares the parties' preferences except that the mediator's preferences are to get a deal. I show that the mediator biases his proposal toward the one that would make both principals worse off compared to when some other proposal is adopted or when there is no mediator. This result extends the one in Che, Dessein, and Kartik (2013), but now with two principals, the degree of pandering worsens.
- 02/22/2013: Stephane Wolton, Election and Reforms in Democracy
[Show Abstract](with Carlo Prato - Princeton/Georgetown University)
We analyze a model of electoral competition with costly political communication. A representative voter elects one between two candidates, who compete by choosing (1) whether or not to commit to a reform (whose implementation is costly for them) and (2) how intensely to campaign on the issue they choose. Candidates have private information regarding their competence and the reform is beneficial to the voter only if implemented by a competent politician. Elections thus serve the dual purpose of screening competent candidates and providing incentives to carry out welfare-improving change. The key innovation is that a candidate successfully communicates her platform to the voter only if both exert effort as in Dewatripont and Tirole (2005). After characterizing the conditions for a separating equilibrium to arise, we show that reforms can only occur when the voter's benefit from policy changes is in an intermediate range: when this benefit is too small, the voter cannot provide enough electoral incentives; when the benefit is too large, he faces the risk of policy failure. Moreover, reforms are more likely to occur when the election is close.
- 03/01/2013: Armin Rick, Manipulation and Optimal Detection in a Principal-Agent Model With Career Concerns
[Show Abstract]
I consider a situation in which a principal delegates a decision to a career-minded expert who has private information about the state of the world. The previous literature has shown that full transparency about the agent's actions and their consequences may prevent the agent from using his private information even if this is socially desirable. I show that in this case, giving agents the opportunity to manipulate their action signal can always improve both the expected outcome of the agent's decision and the quality of equilibrium information transmission, provided that the principal can detect manipulation with sufficiently high probability and is able to commit to a detection policy ex ante. This result holds even if the gain from taking the correct action is state-specific and the principal cannot improve her welfare by limiting her attention to the consequences of the agent's actions, which has been suggested in the previous literature. Moreover, there exists ways in which the principal values information such that the manipulation mechanism with optimal detection is welfare maximizing among all mechanisms that scramble the action-state signal.
- 03/08/2013: Ryan Fang, Income, Inequality and Regime Stability
[Show Abstract](with Weifeng Zhong, Norwestern University)
We present a theory relating economic fundamentals to the stability of political regimes. We consider a government setting economic policies with the knowledge that it is going to face a rebellion. By applying a global game style perturbation to the citizens' coordination game in the event of a rebellion, we are able to select an essentially unique equilibrium, in which the citizens are less likely to support the rebellion if they are richer. Thus the regime has the incentive to enrich the citizens in order to improve its chances of staying in power. On the other hand, it would also like to steal money from the society whenever it is possible. After trading off the conflicting interests, the optimal amount of money stealing by the government is strictly increasing with the strength of the latter. The amount of redistribution implemented by the government is non-monotonic in the government's strength: the weakest governments implement no redistribution and, as the governments' strength increases past some treshold, the amount of redistribution implemented becomes positive and then decreases monotonically. We also examine the impact of economic shocks that affect the citizens' pre-policy income.
FALL 2012
- 11/02/2012: Martin Santamaria - On the Probability of Existence of Stable Networks
- 11/16/2012: Ryan Dorow - Optimal Term Limits for Executives
- 11/30/2012: Ryan Fang - The Demand and Supply of Biased News - A Bayesian Analysis
- 12/07/2012: Hanzhe Zhang - Optimal Mechanisms for Markets with Long-Lived Buyers
SPRING 2012
- 04/06/2012: Stephane Wolton - Signaling in the Shadow of Conflict
- 04/13/2012: Elnaz Bajoori - On the definition of perfect equilibrium in Bayesian games
- 04/20/2012: Martin Santamaria - On the Probability of Existence of Stable Networks
- 04/27/2012: Pablo Azar (MIT) - The Query Complexity of Scoring Rules
- 05/11/2012: Hanzhe Zhang - Optimal Auction Design in the Presence of Posted Price Competition
- 05/18/2012: Armin Rick - Hidden Signal Manipulation in Assignment Models With Noisy And Unbiased Default Signals
- 05/25/2012: Ryan Fang - Economic Development, Income Inequality and Regime Change
- 06/01/2012: Maryam Farboodi - Optimal Revenue Maximizing Mechanism in Common-Value Position Auctions
WINTER 2012
- 02/10/2012: Jin Yeub Kim - Candidate Integration and Voter's Choice
- 02/24/2012: Aditya Bhave - Information and Intervention in Financial Crises
- 03/09/2012: Adriaan Ten Kate - Undirected search with heterogeneous beliefs
FALL 2011
- 10/07/2011: Aditya Bhave - Primary-Market Auctions for Event Tickets
- 10/28/2011: Stephane Wolton - How Organizations' Investment in Lobbying Affects Policy-Making
SPRING 2011
- 04/08/2011: Enghin Atalay - Social Networks with Heterogeneous Agents
- 05/13/2011: Aditya Bhave - The Role of Liquidity in Bank Runs and the Contagion of Financial Crises
- 05/27/2011: Adriaan Ten Kate - Vertical Integration and Innovation