Nemanja Antic
"Communication among Shareholders" joint with with Nicola Persico.
Abstract: This paper provides a model of information transmission among shareholders in an investment venture. An expert shareholder chooses how much information to communicate about the return on investment to a controlling shareholder who controls the investment strategy. Relative to the cheap-talk literature there are two key innovations in our model: first, that the shareholdings drive the incentive to communicate; second, that these shareholdings are determined endogenously. We embed this new model into two settings. In the first setting, equity ownership is determined by investors buying shares on a competitive equity market. We provide conditions under which share-trading delivers perfect communication and full risk-sharing. The second setting is a principal-agent relationship where the equity is granted as compensation by a principal (board) to an agent (CEO). Within this setting, the model highlights a novel trade-off between incentivizing effort provision and promoting information transmission. Furthermore, the principal-agent setting delivers insights into the effects of prudential regulation.
Miguel Ballester
"Separating Predicted Randomness from Noise".
Abstract: Given observed stochastic choice data and a model of stochastic choice, we offer a methodology that allows to separate the data representing the randomness that is inherent to the model from what is noisy behavior. In so doing, we quantify the maximal fraction of the data consistent with the model. We show how to implement our approach to any model of stochastic choice. We then study the case of four well-known models, that capture different notions of randomness. Finally, we illustrate our results with an experimental dataset.
James Best
''Persuasion for the Long Run''.
Abstract: We examine a persuasion game where concerns about future credibility are the sole source of current credibility. A long-run sender plays a cheap talk game with a sequence of short-run receivers. Even with perfect monitoring, long-run incentives do not perfectly substitute for ex-ante commitment to reporting strategies at the stage game. We then show that different methods of augmenting or garbling history can better harness long-run incentives and expand the Pareto frontier. Of particular interest is that a `review aggregator' can be used to implement average payoffs and information structures arbitrarily close to those available under ex-ante commitment. We examine applications to e-commerce, social media and `fake-news’, team management, central banking, and finance.
Pierre Boyer
"Politically feasible reforms of non-linear tax systems" joint with Felix Bierbrauer.
Abstract: This paper presents a conceptual framework for the analysis of politically feasible reforms of non-linear tax systems. First, we provide a generalization of median voter theorems for monotonic reforms, i.e. reforms where the change in tax payments is a monotonic function of income. The characterization of optimal tax systems using the perturbation method is based on such reform. Actual tax reforms are also often monotonic in practice. Second, we develop a method for diagnosing whether a given tax system admits welfare-improving and politically feasible reforms. This reveals the political feasibility of negative marginal tax rates (e.g. EITC) and of sharp increases of marginal tax rates around median income observed in several tax systems.
Paweł Dziewulski
"Eliciting the just-noticeable difference".
Abstract: The evidence from psychophysics suggest that people are unable to discriminate between alternatives unless the options are noticeably different. This can not be reconciled with weak order and utility maximisation. In this paper we provide the testable restrictions for a model of consumer choice that addresses the issue of noticeable differences by characterising consumer preference with a semiorder.
We introduce a necessary and sufficient condition under which a finite dataset of consumption bundles and corresponding budget sets can be rationalised with such a relation. Our restriction weakens the well-known generalised axiom of revealed preference. Furthermore, the method allows to elicit the just-noticeable difference required for the agent to discern between bundles. In fact, we argue that this value constitutes an informative measure of how well the preference revealed in the data could be approximated by a weak order. We conclude by applying our test to a household-level scanner panel dataset of food expenditures.
Jeff Ely
"Moving the goalposts" joint with Martin Szydlowski.
Georgy Egorov
"Strategic Communication with Minimal Verification" joint with Gabriel Carroll.
Abstract: A receiver wants to learn multidimensional information from a sender, but she has capacity to verify only one dimension. The sender's payoff depends on the belief he induces, via an exogenously given monotone function. We show that by using a randomized verification strategy, the receiver can learn the sender's information fully if the exogenous payoff function is submodular. If it is (strictly) supermodular, then full learning is not possible. In a variant of the model that allows for severe punishments when the sender is found to have lied, we can give a complete characterization of when full learning is possible. Our full learning result does not critically rely on perfect verifiability of one dimension: in an example with noisy verification, the receiver's ex-post perceived distribution of information converges in distribution to the true value as the noise vanishes.
Peter Eso
"Incomplete language as an incentive device" joint with Dezso Szalay.
Yingni Guo
"The Interval Structure of Optimal Disclosure".
Peter Klibanoff
"Incomplete Information Games with Ambiguity Averse Players" joint with Eran Hanany and Sujoy Mukerji.
Abstract: We study incomplete information games of perfect recall involving players who perceive ambiguity about parameters or types and may be ambiguity averse as modeled through smooth ambiguity preferences (Klibanoff, Marinacci and Mukerji, 2005). Our focus is on equilibrium concepts satisfying sequential optimality – each player’s strategy must be optimal at each information set given the strategies of the other players and the player’s conditional beliefs. We show that for the purpose of identifying strategy profiles that are part of a sequential optimum, it is without loss of generality to restrict attention to beliefs generated using a particular generalization of Bayesian updating. We also propose and analyze a strengthening of sequential optimality analogous to sequential equilibrium (Kreps and Wilson, 1982). Comparative statics in ambiguity aversion are provided, and we show that robustness to increases in ambiguity aversion implies a type of belief robustness. Examples illustrate new strategic behavior, including strategic use of ambiguity, that can arise under ambiguity aversion.
Maël Le Treust
"Persuasion with limited communication resources" joint with Tristan Tomala.
Abstract: We consider a Bayesian persuasion problem where the persuader communicates with the decision maker through a fixed communication channel. The channel has a given number of messages and is subject to exogenous noise. Imperfect communication entails a loss of payoff for the persuader. We show that if the persuasion problem consists of a large number of independent copies of the same base problem, then the persuader achieves a much better payoff by linking the problems together. We measure the payoff gain in terms of the capacity of the communication channel.
Raphaël Lévy
"Horizontal reputation and strategic audience management", joint with Matthieu Bouvard.
Abstract: We study how a decision maker uses his reputation to simultaneously influence the actions of multiple receivers with heterogenous biases. The reputational payoff is single-peaked around a bliss reputation at which the incentives of the average receiver are perfectly aligned. We evidence two equilibria characterized by repositioning towards this bliss reputation that only differ through a multiplier capturing the efficiency of reputational incentives. Repositioning is moderate in the more efficient equilibrium, but the less efficient equilibrium features overreactions, and welfare may then get lower than in the no-reputation case. Finally, we highlight how strategic audience management (e.g., delegation to third parties with dissenting objectives, centralization) alleviates inefficient reputational incentives, and how multiple organizational or institutional structures may arise in equilibrium as a result.
Daniel Martin
"Price Uncertainty Model".
Abstract: In this paper we propose a general model of consumption for consumers who are uncertain about prices. The behavioral implications of this model can be summarized by a set of non-parametric and testable restrictions on stochastic demand functions that are simple and bear a resemblance to classical restrictions on deterministic demand functions, but are novel in their construction. The empirical performance of these restrictions is assessed using grocery store consumption data.
Eric Mengus
"Credibility and Monetary Policy".
Abstract: This paper revisits the ability of central banks to manage private sector's expectations depending on its credibility and how this affects the use of interest rate rules and pegs to achieve monetary policy objectives. When private agents can only provide limited incentives for the central bank to follow a policy, we show that resulting limited credibility allows a central bank to prevents the inflation from diverging by defaulting on past promises if necessary. As a result, the Taylor rule, when expected, anchors inflation expectations on a unique equilibrium path as long as the Taylor principle is satisfied. Finally, we also show that limited credibility restricts the impact of long-term interest rate pegs, so as to make current conditions less dependent on future policy changes.
Margaret Meyer
"Information Design: Insights from Orderings of Dependence and Heterogeneity".
Abstract: I analyze models of private Bayesian persuasion with multiple receivers, focusing on settings where the sender’s strategic problem can be decomposed as follows: First, familiar single-agent arguments determine the optimal marginal distribution of signals for each receiver, while second, the optimal joint distribution of signals is determined by the interaction between the heterogeneity of receivers and the way their actions affect the sender’s payoff. I show how statistical orderings of heterogeneity and dependence can be used to reformulate the second step of the sender’s information design problem. For heterogeneity of receivers, I show how the majorization ordering of dispersion in vectors can be generalized to orderings of heterogeneity in univariate distributions (of signals), and I derive new results showing how greater heterogeneity in marginal distributions (of signals) affects the maximum achievable expected value for a symmetric supermodular or a symmetric payoff function for the sender. This approach allows me to reformulate some existing models and to derive new characterizations and new comparative statics results.
Stefania Minardi
"Cases and Scenarios in Decisions under Uncertainty" joint with Itzhak Gilboa and Larry Samuelson.
Abstract: We offer a model that combines and generalizes case-based decision theory and expected utility maximization. It is based on the premise that an agent looks ahead and assesses possible future scenarios, but may not know how to evaluate their likelihood and may not be sure that the set of scenarios is exhaustive. Consequently, she also looks back at her memory for past cases, and makes decisions so as to maximize a combined function, taking into account both scenarios and cases. We allow for non-additive set functions, both over future scenarios and over past cases, to capture (i) incompletely specified or unforeseen scenarios, (ii) ambiguity, (iii) the absence of information about counterfactuals, and (iv) some forms of case-to-rule induction ("abduction") and statistical inference. We axiomatize this model. Learning in this model takes several forms, and, in particular, changes the relative weights of the two forms of reasoning.
Joshua Mollner
"High-Frequency Trading and Market Performance" joint with Markus Baldauf.
Abstract: We study the consequences of high-frequency trading using a model with costly information acquisition. Faster speeds increase the efficacy of certain high-frequency trading strategies, with two principal effects. First, fundamental research is crowded out, since there is less time to profit by trading on information before detection by high-frequency traders. Second, the bid-ask spread declines, since there are fewer informational asymmetries. We characterize the frontier of the tradeoff between small spreads and intensive research. Although ubiquitous, the limit order book does not generally implement outcomes on this frontier. Two alternatives do: delaying all orders except cancellations and frequent batch auctions.
Eduardo Perez
"Information Design Under Falsification" joint with Vasiliki Skreta.
Abstract: We study test design when cheating is possible. An agent has a continuum of ex-ante identical, IID, items that can be either good or bad. A test generates a probability distribution of signals as a function of the state. Given realized test results, a decision-maker approves an item when her belief that it is good is above a threshold. The test designer wants approvals of only good items, whereas the agent always seeks approval and has a cheating technology that allows bad (good) items to generate the distribution of signals of good (bad) ones. In a Bayesian framework, cheating--even if costless--has a trade-off: while it may lead to better signals, it lowers the beliefs associated with them. We show that optimal tests can be derived among the class of cheating-proof ones, have a single ``failing" signal but a continuum of signals associated with approvals, and make the agent indifferent between any level of cheating. Good items are always approved. A fully revealing test, while optimal in the absence of cheating is pessimal here: The agent can cheat in this test (and in any two-signal test) in way that results to his best possible payoff--his Bayesian Persuasion payoff-- yielding zero payoff to the principal and the DM. Optimally designed three-signal tests can be ``close" to being fully optimal when cheating is costless and fully optimal when cheating is costly.
Daniel Quigley
“Inside and Outside Information” joint with Ansgar Walther.
Abstract: We analyze a model of communication where uninformed parties receive verifiable inside information, which is disclosed strategically by self-interested parties, as well as outside information such as reviews, ratings or policy reports. For a range of parameters, the classic “unraveling” spiral works in reverse, and information becomes fragile: Second-order changes in the distribution of outside information can trigger first-order reductions in inside disclosures. This yields new insights into policy questions such as the optimal transparency of financial systems. We also show that the importance of outside information hinges on the shape of insiders’ payoff functions, which leads to testable predictions for corporate disclosures, bilateral trade and political contests.
Bruno Strulovici
"Sequential investigations, fabrication, and institutional unraveling".
Abstract: Whether they concern crimes, political corruption, or scientific knowledge, many facts must be learned through the intermediation of individuals with special access to information, such as law enforcement officers, public officials, and experts. This paper considers whether societies and organizations can learn such facts when evidence is cheap to manipulate and is produced through sequential investigations. The answer depends on an asymptotic scarcity condition pertaining to the amount of evidence available about the fact. The condition distinguishes, for example, between reproducible scientific evidence and the evidence generated by a crime. Extensions to other monitoring structures and signals will be discussed.
Rafael Veiel
"A network structure for hierarchies of beliefs".
Abstract: We provide a representation of the universal type space of rationalizable beliefs in a given game with incomplete information. We exhibit a network structure between rationalizable strategies, and define a "strategic type” as a stochastic path along this network. We show that strategic types encompass all strategically relevant information from the universal type space, with the advantage of their space being much coarser. We then use this construction to provide necessary conditions for belief spaces to have strategically relevant m (or higher) order beliefs in terms of all beliefs of order lower than m. We deduce that a type space can only sustain strategically relevant higher order beliefs if the number of distinct lower order beliefs remains above a certain linear function, and there are enough distinct finite order strategic types. In particular, if lower order beliefs are specified too sparsely, then higher order beliefs become strategically irrelevant