Social learning under acute stress.

Abstract

Individual decisions are often made simultaneously under social influence and acute stress, yet despite its importance, it has been largely unknown how stress influences the weight which people place on others' decisions. To answer this I ran a laboratory experiment where 140 subjects were exposed to an acute stressor or a control procedure, immediately before and after which we tested their behavior in a simple Bayesian-updating task. Using three measures (cortisol, heart-rate and mood questionnaire) I show that subjects in the treatment group were under considerable levels of stress. Although stress was expected to increase the weight they put on information coming from the observation of others, I see no effect of stress on subjects' behavior, either after private or public signals, or on the precision of the updating behavior. This holds across different specifications and after the addition of various personal controls, including the Big-Five personality traits and the psychological measure of conformity.