Job Market Paper
"When Biased Beliefs Lead to Optimal Action: An Experimental Study" [Paper]
(with Marco Castillo)
Abstract: Do biased beliefs always lead to sub-optimal actions in equilibrium? Heidhues et al. (2018) demonstrate that optimal action can be achieved with misspecified beliefs when output depends not on each of the inputs independently but solely on their aggregate. This study provides an experimental test of this proposition. Supporting the theory, Experiment A highlights the exacerbated inefficiency that arises when decision-makers allocate tasks to individuals separately, guided by their potentially incorrect beliefs about the relative productivity of each person. However, this harm can be mitigated when decision-makers allocate tasks to a group of individuals, focusing solely on the average productivity of the group. Experiment B further establishes a causal link by introducing exogenous belief biases. This study holds significant implications for how to address the negative impacts of belief biases, especially when belief biases are challenging to rectify.
Presented at: 2022 North-American Economic Science Association Conference (Santa Barbara, CA); Southern Economic Association (SEA) 92nd Annual Meeting (Fort Lauderdale, FL); Western Economic Association International 17th International Conference (Virtual); 2023 Asia-Pacific ESA Economic Science Association Conference (Seoul, Republic of Korea); SPUDM 2023 Warm up talks (virtual); 2024 ASSA Graduate Student Session (San Antonio, TX; scheduled)
Abstract: We examine giving to an in-group member relative to an out-group member in the third party (other-other) dictator game. Individuals are randomly assigned to real groups under a “weak” and a “strong” condition, and also assigned to artificial groups using the minimal group paradigm. We compare the effect of the type of group on subject’s allocation of an amount of money between an in-group member and a person who is not a member of the in-group, using a within-subject design with repeated measures. We find that the Weak and Strong real group conditions bracket the Minimal group condition.
Abstract: We ask whether social preferences measured in subjects who come to the laboratory when invited are systematically different from those of subjects who only respond when an online option is available. Subjects participated in two types of third-party (other-other) dictator games and a trust game, either in the lab or on-line. In the third-party dictator games, the dictator divides $20 between two other individuals, one of whom is a member of their in-group. (We also varied types of in-group between a real group and an artificial group.) In the trust game, the first-mover decides how much of the endowment to send to the second-mover. The second-mover receives the amount sent tripled by the experimenter and decides how much to send back to the trustee. Across all the games, we find no statistically significant differences in social preferences measured in-lab and on-line.
Abstract: How important are non-pecuniary costs incurred during college to the choice of college major? We investigate this question by implementing an information RCT that randomly provides truthful information on teaching quality, climate and inclusion and future earnings to college freshman and sophomores at a large U.S. university. Beliefs about college major characteristics and the probability of choosing among various majors are collected before and after information provision. We find significant effects of teaching and climate information, especially for choosing among business and economics majors. For these majors, average perception of teaching quality and climate improve with the intervention, with men most affected by climate information and women by teaching quality. Our study shows that non-pecuniary aspects of human capital accumulation experienced during college can have an impact on the allocation of talent across fields.
Abstract: Previous studies have shown that punishment opportunities can effectively reduce free riding in the public goods production. By varying the timing of punishment in a public good game, we design two distinct punishment rules, “pre” punishment and “post” punishment, which are involved with different levels of emotional arousal. We employ biometric measures (eye trackers and skin conductance response) in the experiment to capture the psychological process, which will shed light on the psychological process mediating punishment behavior, the response to punishment, and their impact on cooperative behavior. Our results show that consistent with previous findings, introducing punishment opportunities increases contributions and reduces free riding. And participants’ biological arousal is higher after introducing punishment opportunities. This study provides useful suggestions for policy makers and managers for designing proper penalty rules to increase cooperation, and will also contribute to the public good game literature by uncovering the psychological process underlying punishment institutions.
Work in Progress
Abstract: When individuals process new information, two motives may coexist. First, people want to protect their existing or preferred beliefs, which leads them to over- or under-evaluate the reliability of information that aligns or misaligns with those beliefs. Second, people have a preference for accurate information, regardless of its consistency with their beliefs. In this paper, we examine whether the order in which information and its informativeness are revealed influences the dominance of one motive over the other. To test our hypothesis, we conducted a series of online experiments. In one group, information content is revealed first, and then subjects learn of its informativeness. In the other group, the order is reversed. Our preliminary findings suggest that learning informativeness later intensifies motivated reasoning.
Presented at: 2023 North-American Economic Science Association Conference (Charlotte/UNC Charlotte, NC; scheduled); Southern Economic Association (SEA) 93nd Annual Meeting (New Orleans, LA; scheduled)
Abstract: (to be updated)
Contributions to Crowd-Sourced Research Projects
Abstract: With the help of more than 700 reviewers we assess the reproducibility of nearly 500 articles published in the journal Management Science before and after the introduction of a new Data and Code Disclosure policy in 2019. When considering only articles for which data accessibility and hard- and software requirements were not an obstacle for reviewers, the results of more than 95% of articles under the new disclosure policy could be fully or largely computationally reproduced. However, for almost 29% of articles at least part of the dataset was not accessible for the reviewer. Considering all articles in our sample reduces the share of reproduced articles to 68%. The introduction of the disclosure policy increased reproducibility significantly, since only 12% of articles accepted before the introduction of the disclosure policy voluntarily provided replication materials, out of which 55% could be (largely) reproduced. Substantial heterogeneity in reproducibility rates across different fields is mainly driven by differences in dataset accessibility. Other reasons for unsuccessful reproduction attempts include missing code, unresolvable code errors, weak or missing documentation, but also soft- and hardware requirements and code complexity. Our findings highlight the importance of journal code and data disclosure policies, and suggest potential avenues for enhancing their effectiveness.