Greek membership and academic performance: evidence from student-level data

Compared to other inputs of the ‘education production function’, less is known about the effect of social and non-classroom choices. This study examines whether joining fraternities and sororities improves academic performance. In order to account for the self-selection bias of Greek membership, I exploit plausibly exogenous cross-class variation in Greek student composition at the course-instructor level. My estimate shows that a 10 percentage point increase in Greek student composition translates to a 0.02 standard deviation increase in course grade for Greek students relative to non-Greek students. I further find that this effect is driven largely by low-achieving and middle-achieving white Greek members.