''Efficiency and Surplus Distribution in Majoritarian Reputational Bargaining'', Journal of Economic Theory, 2023
(An older version is here. )
``Slippery Protests and Information Aggregation'' (with Zhengqing Gui)
A government may have concerns about "slippery slope" when she evaluates a reform: For some unknown distribution of citizens' ideologies, this reform paves the way to future undesirable outcomes. We propose a simple model to investigate whether a protest can aggregate information about the correctness of a reform in the presence of such concerns. We find that the government must use a non-monotonic interval rule in response to the protest: She reforms only if the number of protesters is neither too small nor too large. This non-monotonicity exposes all citizens to an endogenous risk of over-protesting. Due to this risk, information aggregation is not achievable when citizens place sufficiently high values on a correct reform.
"All-pay Campaign and Electoral Accountability " (New version coming soon)
In an electoral campaign, all candidates can spend resource on valence issues in order to win voters' support. When an incumbent faces an imminent election in the form of such an all-pay campaign, does she have any incentive to serve the interest of a representative voter? We study this question in a two-term model in which candidates have hidden ideologies and no policy commitment power. Depending on office benefit and spending effectiveness, the all-pay campaign may provide stronger accountability than the retention (Banks and Sundaram, 1998) or incentivize outright extremism.
''Competing Auctions with Informed Sellers'' (with Nicolas Riquelme)
We study a game of competing second-price auctions in which each seller privately observes the quality of her object and chooses a reserve price for her auction. Buyers observe the reserve prices and then decide on participation and bidding in at most one auction. We show: (1) there exists a symmetric perfect Bayesian equilibrium (PBE) when there are finitely many sellers and buyers; (2) symmetric PBE converge to a class of competitive ``limit equilibria'' as the number of players increases to infinity; (3) there is no separating limit equilibrium in which the lowest-quality seller posts a reserve price equal to her opportunity cost -- full separation entails exclusion of buyers who can generate social surplus from trade.
''Extreme Agenda Setting Power in Dynamic Bargaining Games'' (with John Duggan)
We investigate a canonical model of bargaining with a fixed agenda setter, and we show that when players are impatient or the set of alternatives is one-dimesional, the equilibrium outcome from the static model obtains; but when players are patient and the alternatives are multidimensional, the equilibrium outcome typically converges to the ideal point of the agenda setter.
Work in Progress
"A Note on Asymptotic Stationary Equilibrium Payoffs of Divide-the-Dollar"
In a Baron-Ferejohn divide-the-dollar model with n agents and q-quota voting rule, consider the mapping from interior recognition protocols to asymptotic (as agents become patient) stationary equilibrium payoffs. The image of this mapping is of dimension n-q, instead of the whole interior of the n-simplex, which is the case with proper discounting.