An essay on the economics of education

This thesis examines the effect of class size and of a school voucher program on student achievement using data from Minnesota and Wisconsin, respectively. Simple comparisons of test scores of students in classes of different sizes produce biased estimates because children in small and large classes differ in a variety of observed and unobserved ways. This thesis uses an IV method introduced by Hoxby (2000) to assess the impact of class size on the test scores of children in grades 3 and 5 in Minnesota public schools. The method exploits random variation in class size due to variation in births from year to year in school catchment areas. While Hoxby found no effect of class size on the test scores of 4th and 6th graders in Connecticut, this thesis finds positive effects of small classes on math and reading test scores. But the positive class size effects found in this thesis are not very large, and given that high cost of reducing class size the results cast doubt on the usefulness of class size reduction policies.; This thesis also investigates the effect of a school voucher program using data from Wisconsin, especially from Milwaukee. The Milwaukee school voucher program provides school vouchers to low income students in grades K-12 that can be used to pay for selected private schools. To examine whether the program benefited students in the program, this thesis examines the test scores of the four cohorts of Black and Hispanic program applicants. To examine how the program affected the quality of Milwaukee public schools (MPS), this thesis compares the test score growth of MPS students not eligible for the program with that of students in the non-MPS in Wisconsin. This thesis finds that the program had no effects on math and reading test scores of students in the program. In addition, the program had negative effects on the math, reading, and language test scores of MPS students, and the negative effects were larger for schools that have more economically disadvantaged students. However, no effects were found on science and social studies test scores of MPS students.