|dc.description.abstract||All over the world, governments are increasingly introducing reforms intending to promote school integration and equal access within the educational systems. Many aim to provide more school opportunities to families and include education vouchers, open enrollment, inter-district choice, and centralized assignment policies, among others. This dissertation consists of three essays that seek to provide evidence for the design of effective and inclusive education policies.
Although many of these reforms implicitly or explicitly concern peer effects, -as they affect the composition of classrooms and schools-, they are often not fully internalized in their design [Hoxby, 2020, Heckman,1999]. The presence of peer effects may change the direction of a policy, having consequences on both the equity and the students' academic outcomes.
Hence, in Chapter 2, I develop a dynamic model of social interactions at school to study the impact of peers on student outcomes. I take advantage of the dynamics of the model to deal with the main identification issues that have been documented in the literature. Then, I use this model to estimate the impact of classroom peers on students' academic achievement by using student-level administrative data from Chile.
I find that social effects are positive but only significant for math. When exploring heterogeneous effects, I find that exposure to opposite-gender peers leads to higher academic performances in math for boys but not for girls. This result could be related to gender stereotypes that affect girls' attitudes and self-perceptions toward math. I also find that higher achievers experience the largest effects from high-ability peers. Conversely, students with low initial achievement levels appear to benefit less and may even experience adverse effects from top ability students. This suggests that some degree of tracking should be preferred to thoroughly mixed or random classrooms.
In the following two chapters, I analyze the effects of two large-scale school choice reforms aimed at reducing social segregation across schools by influencing enrollment patterns. In Chapter 3, I study the impact of a targeted voucher program on a series of cognitive, non-cognitive, and behavioral outcomes, ranging from student test scores to aggressive behavior (bullying). I exploit a national targeted voucher policy implemented in Chile, which increased the funding for disadvantaged students by 50%. Using a difference-in-differences strategy, I confirm previous findings that this policy did not significantly improve students' test scores. However, I show that the program led to meaningful improvements in students' non-cognitive and behavioral outcomes, for both low- and high-SES students. Finally, I provide evidence of the primary channel behind my results: schools participating in the policy have used the additional resources to hire more learning support staff, especially from the psycho-social area.
My last chapter, written in collaboration with Sergio Urzúa and Shanjukta Nath, investigates whether adopting a centralized school admission system can alter within-school socio-economic diversity. We examine a centralized school admission system that the Chilean government introduced in 2016 as the central component of a major education reform aimed at promoting social inclusion and reducing the high levels of school segregation. We use a difference-in-differences strategy, exploiting the sequential introduction of the reform across regions to quantify its heterogeneous impact on segregation. We find no impact of the new policy on school segregation. However, we document that the reform increased within-school segregation in areas with high levels of pre-existing residential segregation and with higher provision of private education.||en_US