Journals
Nash equilibrium often does not seem to accurately predict behavior. In experimental game theory, it is usually assumed that the monetary payoffs in the game represent subjects’ utilities. However, subjects may actually play a very different game. In this case, mutual knowledge of preferences may not be satisfied. In our experiment, we first elicit subjects’ preferences over the monetary payoffs for all players. This allows us to identify equilibria in the games that subjects actually are playing (the preference games). We then examine whether revealing other subjects’ preferences leads to more equilibrium play and find that this information indeed has a significant effect. Furthermore, it turns out that subjects are more likely to play maxmin and maxmax strategies than Nash equilibrium strategies.
This paper introduces receiver ambiguity in a binary model of Bayesian persuasion. The sender has a well-defined prior, while the receiver considers an interval of priors and maximizes a convex combination of worst and best expected payoffs (α-maxmin preferences). We characterize the sender's optimal signal and find that the receiver's payoff differences across states given action (sensitivities), play a fundamental role in the characterization and the comparative statics: If the sender's preferred action is the least (most) sensitive one, then the sender's equilibrium payoff, as well as the sender's preferred degree of receiver ambiguity, is increasing (decreasing) in the receiver's pessimism. We document a tendency for ambiguous receivers to be more difficult to persuade.
Political representatives with criminal backgrounds are considered a great problem in many countries. In India, public disclosure of the large share of politicians currently facing criminal charges has sparked a heated public debate and emerging literature assessing the causes and effects. We develop two hypotheses based on our theoretical considerations. Based on the coding of published affidavits and a comprehensive set of three proxies to measure effort in the 14th Lok Sabha over the 2004-2009 legislative period, we put these hypotheses to an empirical test. Members of the parliament (MPs) facing criminal accusations exhibit on average about 5% lower attendance rates and lower utilization rates in a local area development fund, but only insignificantly lower parliamentary activity. In line with our hypotheses, these differences decline in the development level of the constituency - a proxy for higher rent-seeking possibilities and monitoring intensity. We argue and demonstrate why these negative relations should constitute an upper bound estimate of the causal effect, and show that even under conservative assumptions the effect is unlikely to be caused by unaccounted selection-bias.
[Online Appendix]Product differentiation decisions are frequently made under imperfect probabilistic information about consumer tastes (demand ambiguity). We investigate a Hotelling duopoly game of product-design-then-price choices that incorporates demand ambiguity. Our model allows for different levels of demand ambiguity. We find that the impact of ambiguity on product differentiation depends on firms' ambiguity attitudes. Furthermore, our model generalizes the probabilistic model of Meagher and Zauner (J Econ Theory 117:201 - 216, 2004) and the non-probabilistic model of Krol (Int J Ind Org 30:593 - 604, 2012).
Chapters in Books
Künstliche Intelligenz und Machine Learning in der Energiewirtschaft in Realisierung Utility 4.0. Digitales Business in der Energiewirtschaft erfolgreich gestalten - Transformation von Versorgungsunternehmen in der Praxis (Hrsg.: O. Doleski), 2019, S. 449-463. Wiesbaden: Springer.
Discussion Papers
Ambiguity aversion is commonly associated with a preference for mixing among acts. In this paper, we provide an axiomatic hierarchy that characterizes increasingly strong levels of ambiguity aversion. Each level is specified by k, the maximum number of outcomes of the mix that the decision maker is ensured to accept. For the class of invariant biseparable preferences (Ghirardato et al., 2004), we show that each level k of the hierarchy is equivalent to a specific property of the representation functional: Acts with at most k different outcomes are evaluated at the minimum expected utility w.r.t. the set of dominating measures. Furthermore, the extreme levels of the hierarchy correspond to the prominent definitions of ambiguity aversion introduced in Schmeidler (1989) and Ghirardato and Marinacci (2002). We show that in the Choquet expected utility model the hierarchy only has three levels. Furthermore, our level k = 2 axiomatizes exact capacities, providing the first such characterization.
[2018 Version]This paper investigates whether the strategic behavior of expected utility players differs from that of non-expected utility players in the context of incomplete information games where players can choose mixed strategies. Two conditions are identified where uncertainty-averse non-expected utility players behave differently from expected utility players. These conditions concern the use of mixed strategies and the response to it. It is shown that, if and only if these conditions fail, non-expected utility players behave as if they were expected utility players. The paper provides conditions, in terms of the payoff structure of a game, which are necessary and sufficient for behavioral differences between expected and non-expected utility players. In this context, games are analyzed that are especially relevant for the design of experiments.