School quality beyond test scores: The role of schools in shaping educational outcomes

I study how schools affect student performance and attainment throughout secondary education. I develop a dynamic structural model of cognitive skill accumulation and educational decisions in which students have imperfect information about their ability and learn about it from academic evaluations. The estimation relies on administrative data covering the universe of students in public middle schools in Barcelona (Spain). I use the estimated model to simulate the educational outcomes that each student would achieve in every school in the sample. The results show that school quality cannot be adequately summarized by any single ranking: students with differing characteristics and ability levels benefit from different school inputs, and the school environment is particularly important for the educational attainment of students from disadvantaged backgrounds or with lower ability. In particular, the schools that would maximize these students’ final test scores are often not those that maximize their probability of graduating or pursuing further education. These findings suggest that evaluating schools using only standardized assessments may fail to identify the schools that best support disadvantaged students. I then evaluate a counterfactual policy requiring students to remain in school through the final grade and take the final assessment rather than only until they turn sixteen. The policy substantially increases graduation rates, especially among disadvantaged students, while having only modest effects on the skill composition of high school entrants.