School Spending and Student Achievement: New Evidence from Longitudinal Data

Public school spending has increased dramatically in the United States during the 20th century; concerns about growing wage inequality and slow growth in labor productivity have led many observers to call for even greater increases in such expenditures in the future. Yet despite the ongoing political and legal attention devoted to public school funding, the current research literature is quite unclear as to whether increases in spending will improve the academic achievement of students. Using student-level data from the National Educational Longitudinal Study (NELS), this paper provides some evidence to suggest that the ³state of the art² estimators used in previous studies may be misspecified, thus offering one explanation for the puzzling pattern of results found in the literature. The longitudinal structure of the NELS data is then exploited to produce new estimates. The findings imply that increases in resources, particularly for teacher salaries, lead to modest gains in student achievement. We also consider the economic implications of these estimates. Our rough benefit-cost calculations suggest that even interventions that produce modest achievement gains may produce net benefits to society.