Abstract: We study interim strategy-proofness (ISP), a mechanism design criterion that extends the familiar concept of strategy-proofness to settings with interdependent values. ISP requires a mechanism to provide each agent with a strategy that guarantees the highest expected utility—where the expectation is taken with respect to her belief about the other agents’ types—regardless of the strategies played by the other agents. We show that ISP mechanisms are immune to the presence of agents subject to various cognitive limitations: those who do not have a correct conjecture about other agents’ strategies, those who are unaware of their higher-order beliefs, and those who entertain “cursed” beliefs. For a binary collective choice setting, we provide a characterization of ISP mechanisms. The main insight of this characterization is the identification of a constraint on how much strategic externality agents are allowed to impose on each other, with the constraint being less severe when preferences are less interdependent. This characterization enables us to find an optimal ISP mechanism for the jury model and prove that the Condorcet Jury Theorem does not hold when ISP is required.
Abstract: In this paper, we study ex post strategy-proofness (ESP), which extends the concept of strategy-proofness to settings with interdependent values which are more complex intrinsically. ESP requires that truthful reporting is always optimal for every agent, regardless of the strategies and types of other agents. We explore which social choice functions can be implemented by ESP mechanisms with transfers. For a broad class of environments, including single-good auctions, we find that implementability is precisely characterized by two conditions: 1. Monotonicity: a “higher” type leads to a “higher” alternative; and 2. Screenability: types that are “entangled” are treated identically, where entanglement is a condition that is more serious when types are densely distributed or when preference interdependence is significant. For more general settings, we show that implementability is characterized by a generalized cyclical monotonicity condition, or equivalently, by an optimality condition derived from an induced matching problem between types and alternatives. We also discuss how implementation becomes easier if agents can reason using iterated elimination of dominated strategies, and how it becomes more challenging when transfers are not available.
Theoretical Economics , 18(2): 463-479, May 2023
Abstract: We study ex post implementation in collective decision problems where monetary transfers cannot be used. We find that deterministic ex post implementation is impossible if the underlying environment is neither almost an environment with private values nor almost one with common values. Thus, desirable properties of ex post implementation such as informational robustness become difficult to achieve when preference interdependence and preference heterogeneity are both present in the environment.
Journal of Mathematical Economics ,98, January 2022
Abstract: We extend Harsanyi's (1955) utilitarianism theorem to an infinite-horizon multi-generation setting: Under some additional assumptions, the Pareto condition is equivalent to utilitarian aggregation and the utilitarian weights are unique. We analyze the properties of utilitarian weights, such as the limiting behavior of utilitarian weights for distant future generations, and the comparative statics of utilitarian weights as the social discount factor or the social risk attitude changes. Among other findings, we show that a higher social discount rate is associated with a more unequal assignment of utilitarian weights across generations.
Econometrica, 86(5): 1537-1567, September 2018 (Lead Article)
Abstract: The most critical issue in evaluating policies and projects that affect generations of individuals is the choice of social discount rate. This paper shows that there exist social discount rates such that the planner can simultaneously be (i) an exponential discounting expected utility maximizer; (ii) intergenerationally Pareto—that is, if all individuals from all generations prefer one policy/project to another, the planner agrees; and (iii) strongly non-dictatorial—that is, no individual from any generation is ignored. Moreover, to satisfy (i)–(iii), if the time horizon is long enough, it is generically sufficient and necessary for social discounting to be more patient than the most patient individual’s long-run discounting, independent of the social risk attitude.
Macroeconomics 18(1): 149-159, March 2016
Abstract: This paper studies the impact of monetary policy on income inequality through both theoretical analysis and empirical test. Theoretically, this paper probes into the four possible channels through which the monetary policy affects income inequality: income composition, portfolio, saving redistribution, and wage heterogeneity. The first two channels increase income inequality when central banks implement an expansionary monetary policy. The last two channels have opposite effects. This paper then conducted the empirical test on the relation between monetary policy and income inequality, based on the panel data of China, United States, Canada, and Germany from 1995 to 2012. The paper applied Gini coefficient as a measurement of income inequality, and the broad measure of money supply and loan interest rate, respectively, as measurements of monetary policy. After controlling certain variables such as financial policy, economy growth level, education, population growth, industrial structure, unemployment rate, and advancement in science and technology, we found that the expansionary monetary policy will significantly reduce income inequality. The empirical test exhibited that the saving redistribution and earning heterogeneity channels undertook more powerful effects.