Learning Management Through Matching: A Field Experiment Using Mechanism Design

05/04/2019 15:00


Abstract: We run a novel management placement for young professionals. Participants are placed in established firms for one month each, in order to shadow a middle manager in her or his daily work. We hypothesize that this experience provides a useful opportunity for young, highly educated and motivated individuals to learn the skills of management, and a stepping stone to new career opportunities. Our field experiment uses random assignment of young professionals to participate in the program, random assignment of potential host firms to participate in the program, and a deferred acceptance matching algorithm to assign young professionals to host firms. This empirical design allows us to recover both average treatment effects, through the randomized experiment, and marginal effects of being assigned to different types of host firms, through a propensity score. We find that treated individuals on average are more likely to have good wage jobs both on the extensive and intensive margins, and that there is a zero effect on self-employment, on average. However, among treated individuals, high ability individuals who are assigned to a firm with a high management practices score are significantly more likely to start a business. This effect is driven by the top end of the ability distribution as proxied by the propensity score, and seems to be accompanied by a substitution from wage work. Our results point to the importance of heterogeneity in treatment variety, and offer an empirical framework to analyse such heterogeneity.