Paid activity and time spent in graduate school.

Summary This paper investigates a causal mechanism through which a student's work can influence exam success. Students who have a paid job will a priori face increased difficulties in attending classes and in their personal work. The empirical analysis, based on five cross-sectional surveys conducted in France from 1997 to 2010, examines the joint decision of work and study time for nearly 80,000 students enrolled in the first to fifth years after the baccalauréat. We turn to estimating selection models on unobservables and observables to account for the endogeneity of paid activity. We find that students with paid employment spend less time on both attending classes and personal work. The negative impact on study time appears to be much higher for those with at least half-time paid employment.