Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
There is substantial variation across occupations in outcomes, such as earnings, job satisfaction, and health, yet little is known about how occupational preferences are shaped and evolve. This paper assesses a channel that may influence occupational preferences—peer effects—by exploiting unique institutional features and a natural experiment at West Point to document the way students' (i.e., cadets') occupational preferences evolve, and to test for peer effects. After graduation, cadets must enter the U.S. Army as an officer in one of 16 occupational branches, and are required to rank-order their preferences for the branches repeatedly over four years. West Point randomly assigns cadets to social groups, both at broad company and narrow roommate levels. We leverage the timing of the randomization and preference elicitations to test for peer effects at the company, roommate, and older-peer levels. With the possible exception of the first few weeks of exposure, we find little evidence of peer effects. We also use a survey of cadets to explore mechanisms and find evidence that peer information may not be credible in this case, but rather cadets rely on mentors and job characteristics when making this decision.