Working papers
Temptation to Consume Information - R&R at American Economic Journal: Microeconomics
Presentations: EEA Congress, ESA Job Market Series, European Winter Meeting of the Econometric Society
Avoiding information can be desirable for individuals. We use an experiment to show that temptation can be an important barrier to this strategy. 75% of individuals who would like to avoid undesirable information in a day's time are also willing to pay for a commitment device that entirely eliminates it from their choice set. As in other studies of temptation, only a small share of individuals make dynamically inconsistent choices when offered the temptation against their preferences. However, we construct a novel analysis of self-control costs using response times and find evidence that commitment demand is motivated by costly self-control.
We explore whether beliefs are distorted to enable optimism about the consequences of past choices. In our experiment, involving over 1000 subjects, individuals are asked to recall a signal. In the treatment group alone, the signal is positively correlated with a favourable unobserved outcome of a prior choice. Our results show that subjects in the treatment group believe the signal was higher, even immediately after transmission. We also obtain suggestive evidence that belief distortions increase after a few days due to memory loss. Finally, subjects distort beliefs in precisely the same way even when they know they will find out the true value at the end of the experiment, implying that the ability to sustain belief distortions indefinitely does not influence the present motive to distort beliefs.
Presentations: European Winter Meeting of the Econometric Society, Barcelona School of Economics Summer Forum, Early Career Behavioural Economics Conference, CESifo Summer Institute, ASSA Annual Meeting (invited)
We use field data on teenagers' memories of mathematics grades to provide novel evidence on the dynamic relationship between biased recall and beliefs. Recall of grades is strongly positively biased, and becomes more biased when more time has passed since the relevant report card, supporting theories of motivated beliefs and memory loss. Students are also more likely to overconfidently predict their next grade in periods where they recall an incorrectly high grade. We estimate a structural model which finds support for the two-way relationship between beliefs and recall proposed by Kőszegi et al. (2022). Beliefs about ability are sticky: a single bad quarterly grade in mathematics negatively affects academic self-esteem for 2--3 years. Our model also reveals substantial heterogeneity across students: high achievers have disproportionately strong preferences for positively biased recall, making their self-esteem much more resilient to unfavourable grades. Simulations suggest informational interventions can harm welfare and possibly widen attainment gaps if not coupled with measures to protect self-esteem, especially for low achievers.
Pro-social Behaviour in Groups: Evidence From The Field (with Julia Shvets and Toke Aidt)
Contact:
Faculty of Economics
Austin Robinson Building, Sidgwick Ave
Cambridge CB3 9DD, United Kingdom
Email: vr277 [a t] cam.ac.uk
Phone: +447798760936