Can Experience Extinguish Decision Biases In Conflict Settings? Evidence From A Five-Year Large-Scale Natural Field Experiment

Whether experience can reduce decision biases in conflict settings is controversial. Some evidence suggests experience may quickly extinguish biases after just a few interactions; however, other evidence suggests biases can persist even among very experienced decision-makers. We reconcile these competing findings using a natural field experiment where volunteer crisis counselors were repeatedly and randomly assigned to resolve harder (versus easier) conflicts for five years, starting from career onset (i.e., zero hours of experience). We examine whether the “end effect” bias in counselors’ turnover decisions (disproportionately overweighting one’s most recent task assignment when evaluating one’s overall job) is moderated by experience. Counselors do learn to make less biased turnover decisions over time, but this process is extremely slow: each task performed shrinks the “end effect” bias by just 0.07%. It takes 1,356 tasks (~1,265 hours) to completely extinguish the “end effect” bias–but only 0.7% of counselors attain that much experience.