Strategic uncertainty in market design

• Applicant: Bernhard Kasberger, Ph.D.
• Subject Area: Economic Theory
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 550419621

The project studies strategic uncertainty in market design. Strategic uncertainty describes a situation in which participants cannot fully anticipate others' behavior, perhaps because they face the strategic situation for the first time. Traditional game-theoretic methods like Bayes-Nash equilibrium may not adequately address such uncertainty. The aim is to develop new game-theoretic methods that feature strategic uncertainty, use them to enhance our understanding of the impact of strategic uncertainty on market outcomes and apply these methods to design market rules. To do so, we first define a flexible game-theoretic solution concept that accommodates varying degrees of uncertainty about others' behavior. Players are uncertain about the distribution of other players’ types and strategies and have non-probabilistic beliefs about the distribution of others' actions, which are characterized by simple statistics. The number of statistics is then a natural measure of how uncertain the player is about the behavior of the others. Such a modeling approach is inspired by the observation that the only available information about the other participants’ behavior is simple statistics in many markets. In equilibrium, the players’ beliefs are consistent with the optimal behavior and the latent type distribution. An objective of the project is to prove the existence of such an equilibrium. The solution concept will be applied to various market scenarios (e.g., auctions, school choice algorithms) to assess the impact of uncertainty on market outcomes such as welfare and revenue. Ideally, this leads to the design of more resilient markets. The new solution concept is also used to design optimal information policies in market settings, such as sponsored search auctions. Is a higher or lower transparency about the behavior of the other bidders optimal? The project includes the empirical assessment of theoretical predictions in existing and new laboratory experiments.

• DFG Programme: Research Grants