Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We show that the provision of social information influences a high-stakes decision and this influence persists over time. In a field experiment involving thousands of admits to Teach For America, those told about the previous year's matriculation rate are more likely to accept a teaching job, complete training, start, and return a second year. To show robustness, we develop a simple theory that identifies subgroups where we expect larger treatment effects and find our effect is larger in those subgroups. That social information can have a powerful, persistent effect on high-stakes behavior broadens its relevance for policy and theory.