Does being grouped in a class with higher-achieving students lead to more learning?
Economists and other social scientists have long been interested in measuring "peer effects" – how individual behavior is affected by one's peer group. In this paper, we take a well-known model from labor economics that makes predictions about how workers sort into different jobs. The key principle is that job assignments are based on comparative advantage. If there's a shortage of teachers in a given city, say, then other types of workers might decide to enter the teaching profession to fill that need. In a city with an ample number of teachers, those same workers would likely stay in their own sector.
Applied to classroom interactions, we suppose that students choose whether to be "nerds" or "troublemakers," depending on the composition of their peers. Using the same logic as above, we suppose that, in every classroom, there is "demand" for both studiousness and mischief. So, depending on the natural abilities of the classroom, one might be more inclined to spend time making trouble or studying. The precise change in behavior depends jointly on one's natural abilities and the composition of the classroom, the latter of which determines the social benefit payoff to making mischief rather than hitting the books.
Read the full paper.
Two key findings emerged: First, when comparative advantage is the guiding principle of peer group interaction, the effects of grouping a student into a class with higher-achieving peers would depend on the new student’s natural position within the ability distribution in that classroom and the effective “wages” associated with academic achievement in that particular “social market.” Second, since a student’s comparative advantage in a given peer group is typically unobserved, individual choice may still influence student behavior and achievement even under random assignment. In other words, a student’s relative position within the range of achievement levels in a given classroom is the critical factor predicting his or her performance and behavior in that classroom.