The job market paper presented by Santiago Pereda Social Spillovers in the Classroom: Identication, Estimation and Policy Analysis displayed a new method of identifying and estimating the strength of social spillovers in the classroom and the distribution of teacher and student effects. Identification depended on the assumptions of double randomization of teacher and students to classrooms and the linear in means equation of test scores. The linear independent factor representation of test scores allowed more efficient estimates of the social multiplier and the combination of all joint moments of different orders to be obtained. “This is for those interested not in causal or mean effects but for those whom the distributional effects are important as a way of assessing how policy is addressing inequality!,” summarized Santiago.

He also presented a theoretical model of social interactions in the classroom that yielded the linear in means equation of test scores. In the model, the teacher and students play a game in which they choose how much effort to exert. The method allowed the estimation of more features of the distribution of teacher and student effects than mean and variance. For estimation, he used a minimum distance procedure that combines information coming from different moments. He used data from the Tennessee Project STAR dataset.

Notably, distributions of teacher and student's abilities seemed to depart from the usual normality assumption, and student's distribution exhibited a high degree of heteroskedasticity in class size. Based on these estimates, he performed several counterfactual social planning experiments, comparing losers and winners under different assignment rules. Assignment of good teachers to large classrooms increased the average test scores, with students in the left tail benefiting more than the rest.

“Concluding,…”said Santiago, “Assignment of good students to small classrooms increases student test scores in the right tail of the distribution, while students in the left tail get lower test scores with an overall increase in mean test scores. Mixing good and bad students together results in a small effect on mean test scores but reduces inequality”.