Behavioral Economics Workshop

Jeevant Rampal (IIM Ahmedabad)

with Shantanu Khanna, Vinay Jha & Eduardo Fabres

To increase “national integration” (NI), in 2017 the Indian government changed its mechanism for assigning Indian Administrative Services officers to cadres. NI is low if officers are assigned cadres close to their home states. In contrast to the previous mechanism, the 2017 mechanism is associated with higher NI but it is not strategy proof while reporting cadre preferences. We propose a strategy-proof mechanism that can achieve arbitrary NI. Using surveys and experiments with aspirants, students, and IAS officers we study the causal impact of the previous, 2017, and proposed mechanisms on misreporting, regret, NI, perceived fairness, and difficulty. We find that the 2017 mechanism has the highest degree of misreporting, and that reporting behavior is similar under the 2017 and proposed mechanisms. Evidence shows that with uncertainty, NI is higher in the 2017 mechanism when compared with the previous mechanism. Surveys of aspirants reveal that the 2017 mechanism is the most difficult in terms of reporting, and is considered the least fair of the three mechanisms.

Anubhav Jha (Ashoka University)

with A. de Albuquerque, F. Finan, L. Karpuska & F. Trebbi

We present a methodology for decoupling taste-based versus statistical discrimination in political behavior. We combine a flexible empirical model of voting, featuring vertical and horizontal candidate differentiation in gender, ability, and policy positions, with a large-scale micro-targeted electoral experiment aimed at increasing female candidate vote shares. Our structural econometric approach allows us to separately identify preference parameters that drive taste-based discrimination and belief parameters that drive statistical discrimination via expectations about the ability and policy positions of female politicians. Our application to Brazilian municipal elections uncovers substantial levels of both taste-based and statistical discrimination. Counterfactual political campaigns show promise in reducing both.

Kalyani Chaudhuri (Ashoka University)

with Anujit Chakraborty

Recent evidence shows women are more likely than men to volunteer for non-promotable tasks in organizations, which perpetuates gender gaps in career advancement. Using a series of laboratory and survey experiments, we show that these gender differences in strategic volunteering are driven by women’s beliefs that others are unlikely to volunteer for these tasks, as well as social norms about who ought to complete these tasks. Specifically, our lab experiment indicates that men and women have similar prior beliefs about others’ likelihood of volunteering, but women update their beliefs more pessimistically than men in response to information that others are free-riding. In our surveys, women report volunteering more than men only in tasks that are stereotypically associated with females, and this gap disappears when beliefs about others’ volunteering are exogenously fixed for both men and women.

Priyoma Mustafi (Ahmedabad University)

with K. Pun Winichakul, G. Lezama, M. Lepper, A. Wilson, D. Danz & L. Vesterlund

To assess the threat of experimenter demand, we ask whether a hypothetical ‘ill intentioned’ researcher can manipulate inference. Four classic behavioral comparative statics are evaluated, and the potential for false inference is gauged by differentially applying strong positive and negative experimenter demand across the relevant decision pair. Evaluating three different subject pools (laboratory, Prolific, and MTurk) we find no evidence of experimenter demand eliminating or reversing directional effects. The response to experimenter demand is very limited for all three subject pools and is not large enough to generate false negatives, though we do find evidence of false positives when testing precise nulls in larger online-subject pools.

Yan Chen (University of Michigan, School of Information)

To evaluate the impact of group identity on belief formation, we conducted online experiments before and after the 2020 US presidential election. We elicit participants’ beliefs about future unemployment statistics and health system rankings and provide relevant news summaries. We find that people pay money to avoid information from political outgroups and attribute lower weight to this information when updating beliefs. An intervention which unlabels information sources decreases outgroup information avoidance by 50%, with a significant impact on only groupish participants. A debiasing intervention equalizing instrumental values of information sources reduces only universalists’ WTP for a private signal. The interventions have little effect on information processing. The evidence is consistent with a source-utility channel contributing to polarization.