Three Essays on Resources in Education.

Abstract

The first chapter of this dissertation examines how a large increase to state aid to schools in New Jersey affects the expenditures of the affected school districts. The 30 school districts affected by the Abbott policy received a 10% boost to school budgets following a court ruling. Using the funding induced by the court ruling as an instrument for state aid to school districts, I measure the uses of increased state aid. In specifications that control for unobserved differences across districts, I find that 60-70% of the increased funding passes through to increased school expenditures and that the rest is used to reduce local contributions to school revenues in Abbott districts relative to comparison districts. This funding went largely to instructional purposes, paying for increased hiring of teachers and support services personnel.

The second chapter looks at the achievement effects of the expenditure changes examined in chapter 1. Using restricted student-level test scores for New Jersey's 11th grade assessment, I find that the policy improved math and reading scores of minority students by 0.2-0.25 standard deviations. Trends in scores at earlier grades also show shrinking achievement gaps.

The third chapter, written with James Sallee and Paul Courant, develops a simple model for allocating students and resources to schools in a system of higher education. Starting with a distribution of student ability and a fixed quantity of resources we show that only a few assumptions are required to lead to a tiered higher education system that sorts students by ability. Our assumptions, that student ability and school resources are complementary and that there is a fixed cost to establishing a school, lead to a tiered system even without peer effects.