Kremer and Levy on peer influence

From The Journal of Economic Perspectives, Summer 2008. No wonder my roommates had such poor grades!

We exploit a natural experiment in which students at a large state university were randomly assigned roommates through a lottery system. We find that on average, males assigned to roommates who reported drinking in the year prior to entering college had a Grade Point Average (GPA) one quarter-point lower than those assigned to nondrinking roommates. The effect of initial assignment to a drinking roommate persists into the second year of college and possibly grows. The effect is especially large for students who drank alcohol themselves in the year prior to college.

While these conclusions are fairly obvious, the secondary findings are quite interesting:

In contrast to the males, females' GPAs do not appear affected by roommates' drinking prior to college. Furthermore, students' college GPA is not significantly affected by roommates' high school grades, admission test scores, or family background...Surprisingly, the policy of segregating drinkers by having substance-free housing could potentially lower average GPA in the university.

Link to abstract: JEP, Vol. 22, No. 3
Link to (non-gated, older version of) paper: June 2005