Job Market Candidate
Available for interviews at:
Graduate School of Business
655 Knight Way
Stanford, CA 94305
- European Job Market in Rotterdam (Dec 18-19)
- ASSA Annual Meeting in San Diego (Jan 3-5)
Job Market Paper |
Learning from Unknown Information Sources [Online Appendix]
Using both experimental and observational data, I show that when the accuracy of information sources is uncertain, people's reactions to news display two patterns: under-reaction and pessimism. Relative to the case of known information accuracy, subjects in a lab experiment under-react to information with uncertain accuracy. In addition, the under-reaction is more pronounced for good news than for bad news. Evidence from the US stock market also corroborates these findings. In response to upward revisions in analyst earnings forecasts, stock prices under-react more if the revisions are issued by analysts with no forecast records. In contrast, for downward forecast revisions, the degree of under-reaction does not depend on whether the issuing analysts have records or not. These patterns are inconsistent with standard expected utility theory, but are consistent with a theory of belief updating where agents are insensitive and averse to compound uncertainty and ambiguity. I also find that individuals' reactions to information with uncertain accuracy are uncorrelated with their evaluations of bets with uncertain odds. This suggests that people have distinct attitudes toward uncertainty in fundamentals and uncertainty in information accuracy.
Information-Dependent Expected Utility
In decision problems under uncertainty, the subjective evaluation of an outcome can depend on the information content of its realization. To accommodate this dependence, we introduce and axiomatize a model of information-dependent expected utility by allowing the utility of an outcome to flexibly depend on its information content in an (Anscombe-Aumann) act. Subjective beliefs are identified in a special class of our model where the utility of an outcome can be decomposed as the sum of consumption utility and information utility. Our model allows for both information seeking and information averse preferences, as well as a comparative theory of information preferences. For information seeking preferences, we introduce a Hidden Acts representation where the value of information is as if induced from the expected utility of the optimal choice in a fictitious future decision problem given that information.
Social Comparison, Employee Attrition and Productivity: A Workplace Experiment
(with Shannon X. Liu and Hugh Xiaolong Wu) [Status: fieldwork in progress]
How should firms shape the social comparison process in the workplace to improve employee productivity and to reduce turnover? We conduct a 7-month field experiment in a multi-national spa chain with 160 stores and 5000 spa workers in China. We provide spa workers with bi-weekly messages on either the current performance of a co-worker with similar work experience (T1) or the performance trajectory of a senior co-worker (T2). First, we study how workers' outlook about their future performances and reference points for performance comparison depend on their knowledge about their co-workers' current and past performance. Second, we measure the effects of information treatments on the workers' beliefs about their own future productivity, reference points, stress levels, productivity, attrition, and pro-social behaviors.
[AEA RCT Registration]
Truth-Telling and the Design of Flexible Commitment Contracts: A Field Experiment
(with Shengmao Cao and Tony Fan) [Status: pilot completed]
We propose a novel commitment contract (CC) with exemption clauses to provide incentives for physical exercise while retaining flexibility. Like a traditional, rigid CC, some money is deposited into an account (by the participant or by a third party) and the participant is allowed to withdraw the money if she attends the gym. Unlike a rigid CC, if a participant reports that she did not go to the gym because of illnesses, injuries, unanticipated obligations, or other pre-specified conditions, she is allowed to withdraw the deposit as well. The participants' reports are not verified, so the effectiveness of such a CC depends on the participants' aversion to lying. We conduct a field experiment at the Stanford athletic facilities to evaluate the demand for and the effectiveness of a CC with exemptions, in comparison to a rigid CC and a control contract without incentives.
Over-Detection of Causal Patterns
(with Collin Raymond)