The impact of student's ordinal cognitive ability rank on school violence: Evidence from China

Abstract

School violence can exert immediate and enduring effects on the human capital development of students. Despite extensive discussions on the causes of school violence, the role of peer effects has been underexplored in the literature. To fill this gap, this paper examines the effect of cognitive ability rankings on students' propensity for committing school violence. We select schools that randomly assign students to classes to mitigate selection biases and identify the causal effects. Our findings show that improving a student's rank from the lowest to the highest reduces school violence by 0.5 standard deviations. A mechanism is the higher opportunity cost for students with a high cognitive ability rank. We also find that high-ranking students are more likely to stay with high-quality peers.