This document analyzes the effect of bullying on student performance in the short and in the long term. The level of bullying is estimated from the answers of 9 th grade students in Saber test (2005)(2006), while student performance is measured both in Saber 9º test (short term) and Saber 11º, a standardized test answered by students two years later (long term). Because of the existence of reverse causality, the relationship is estimated through instrumental variables. The results suggest that

more »

... one standard deviation increase in the within-school level of bullying decreases individual test score by 0.55 standard deviations in Saber 9º and by 0.4 standard deviation in Saber 11º. Furthermore, the results indicate that bullying´s effect is stronger for younger students in the short term but that there are no differences by gender. However, in the long run the existing differences disappear. Additionally, this document finds that there are differences in the relationship between the level of bullying and academic performance depending on student´s academic achievement.