Daniel N Hauser's Website
I am a post doctoral researcher at Aalto University.
Dynamic Games, Reputation, Social Learning, Model Misspecification
I consider a model in which a firm invests in both product quality and a costly signaling technology, and the firm's reputation is the market's belief that its quality is high. Signaling influences the rate at which consumers’ receive information about quality: the firm can either promote, which increases the arrival rate of signals when quality is high, or censor, which decreases the arrival rate of signals when quality is low. I study how the firm's incentives to build quality and signal depend on its reputation and current quality. The firm's ability to promote or censor plays a key role in the structure of equilibria. Promotion and investment in quality are complements: the firm has stronger incentives to build quality when the promotion level is high. Costly promotion can, however, reduce the firm's incentive to build quality; this effect persists even as the cost of building quality approaches zero. Censorship and investment in quality are substitutes. The ability to censor can destroy a firm's incentives to invest in quality, because it can reduce information about poor quality products.
"Social Learning with Model Misspecification: A Framework and A Robustness Result," joint with Aislinn Bohren (R&R at Econometrica)
This paper explores model misspecification in the canonical observational learning framework. An agent's type specifies how he interprets private and public signals and how he believes others draw inference. This framework captures behavioral biases such as confirmation bias, underweighting or overweighting information, optimism bias and correlation neglect, as well as models of inference such as level-k and cognitive hierarchy. We develop a simple criterion to identify how behavioral biases in information processing and inference impact asymptotic learning. Depending on the nature of the bias, beliefs may converge to the incorrect state, the correct state, or not converge at all. Agents with different biases may hold different asymptotic beliefs, even though they are all observing the same sequence of prior actions. Next, we explore the optimal strategy of an information designer who can release costly public information and has a preference over the asymptotic beliefs of agents. The optimal timing, frequency and strength of public information depends crucially on the structure of model misspecification. This contrasts with the fully rational model, in which the timing, frequency and strength of information releases are irrelevant for asymptotic learning.
Work in Progress
"Misinterpreting Social Outcomes and Information Campaigns," joint with Aislinn Bohren (Extended Abstract)
We study how to design information to help misspecified learners learn the true state of the world. Agents learn by observing exogenous signals and the choices of others in addition to signals generated by the designer. With the correctly specified model, these agents would eventually learn the truth without any intervention, in the presence of misspecification this is no longer the case. We characterize how the degree and type of misspecification affect the optimal information policy. Depending on the misspecification, it may be optimal for the designer to release very one very precise signal or to continually release small amounts of information.