This dissertation gathers evidence on three sources of education inequalities across different education levels (preschool, primary, secondary, and higher education) in the context of Spain. It revolves around the causal effects of large-scale educational policies on the efficiency and equity of educational systems.The first chapter focuses on the effects of financial aid for disadvantaged students in the context of higher education. National financial aid programs for disadvantaged students cover a large fraction of college students and represent a non-negligible component of the public budget. Using a reform in the Spanish need-based grant program, this paper tests the causal effect of receiving the same amount of grant under different intensities of academic requirements on student performance, degree completion and student dropout. I use administrative micro-data on the universe of applicants to the grant in a large university. Exploiting sharp discontinuities in the grant eligibility formula, I find strong positive effects of being eligible for a grant on student performance when combined with demanding academic requirements, while there are no effects on student dropout. Students improve their final exam attendance rate, their average GPA in final exams, and their probability of completing the degree. They also reduce the fraction of subjects that they have to retake. The second chapter centers on the gender differences in academic performance due to the testing-environment, in the context of primary and secondary education. However, little attention has been devoted to investigating how the organization of student testing may influence the relative performance of male and female students. This paper analyzes the gender gap in test scores that arises as a result of differential responses by boys and girls to the testing environment. To that end, we exploit a unique randomized intervention on the entire population of students in the 6th and 10th grades in the Region of Madrid (Spain). The intervention assigned schools to either internally or externally administered testing. We find that girls do worse than boys in exams that are externally administered, especially in male-dominated subjects. Additional survey evidence on stress, self-confidence, and effort suggests that lower relative female performance in externally administered tests results from a lower ability to cope with stressful situations as a result of less familiarity with the testing environment.The third chapter studies the relationship between school choice priorities and school segregation in the context of preschool education. This work aims at broadening the scope of market design questions to school choice by examining how government-determined school choice priorities affect families’ choices and pupil sorting across schools in the context of the Boston Mechanism. We use two large-scale school choice reforms in the school choice priority structure undertaken in the region of Madrid (Spain) as a source of variation. In particular, we exploit an inter-district school choice reform that widely expanded families’ choice set of schools. We combine an event study first difference across cohorts and a Difference-in-Difference design to identify the impact of the reforms. Using unique administrative data on parents’ applications to schools, this paper shows that families reacted to the reform exerting higher inter-district choice and applying to schools located further away from home than before the reform. We find distributional effects of the reform concluding that parents from the highest education levels and parents of non-immigrant students were those who reacted the most in absolute terms. We find a decrease in school segregation by parental education and an increase in school segregation by immigrant status.