|
Rewards based on class rank are ubiquitous in schools throughout the world. In this paper we examine the consequences for student effort and performance when students are rewarded based on their relative class standing using game-theoretic competition models to guide our analysis Based on a competition model in which students within similar-ability peer groups recognize that their efforts will be matched, and in which stronger students assist weaker students, we derive comparative-static predictions for how changes in class student composition differentially affect student outcomes and effort across ability groups. We find empirically using a variety of identification methods applied to variations in school student composition due to the resettlement of large numbers of Southeast Asian refugees in the late 1970's and 1980's throughout the United States that, consistent with the competition cum peer-effect model but inconsistent with a Nash equilibrium competition model, increases in the number of (refugee) strong students reduce the effort levels of strong and weak incumbent students, decrease the test scores of incumbent strong students, but increase the test scores of weak incumbent students. Conversely, increases in the number of weak students decrease the effort and test scores of weak incumbent students but increase the effort and test scores of strong incumbent students. We also show that the findings we obtain on test score and effort level effects cannot be explained by the endogeneity of initial refugee location assignments, teacher adjustments in curriculum, selective classroom disruptions, parent shifts in homework assistance, labor-market effects, or by strong-student school exit.
|