Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
I examine the impact of student incentives in Benin, using three different designs that can be implemented relatively cheaply and with administrative data. The first design is a standard incentive structure where students receive monetary rewards for reaching a performance target. In the other two designs, teams of four students receive incentives based on either their performance level as a group or in a team tournament scheme. I find a large and similar average treatment effect across designs, ranging from 0.27 to 0.34 standard deviations (Standard errors do not allow to rule out that the three designs are equally effective).