The Effect of Local Property Taxes and Administrative Costs on Student Achievement Scores : Evidence from New Jersey Public Schools

In this study we theorize that student achievement varies cross-sectionally with the percentage of school district revenues raised from local taxes and with salary levels of school district administrators (Superintendents and Principals). This theory is tested using a sample of 218 Kindergarten – Grade 12 school districts in the state of New Jersey. Our regression model includes sixteen variables to explain student achievement scores on standardized exams on Language and Mathematics from years 2001-2006. We find that increases to the percent of school funds raised locally has a positive influence on test scores, consistent with agency models that suggest local property tax funding (opposed to centralized financing) provide a disciplinary mechanism to improve the quality of output from local public good providers (Hoxby 1999). Consistent with our predictions, we find that higher administrative salaries are associated with higher test scores. But we also find that other costs of administration (excluding salaries) are associated with lower test scores. Taken in combination, these results help explain the mixed findings in prior studies that explore the nexus between administrative costs and student achievement