The impact of uncertainty on school quality capitalization using the border method

Abstract

Hedonic valuation methods are commonly used to measure school quality capitalization to assess the value of school reforms and gain insight into how households choose schools and neighborhoods. These methods rely on observed school quality as a proxy for future school quality, but often ignore the potential for changes in people's expectations of future school quality. In particular, the border method is commonly used to measure school quality capitalization, as it differences out unobserved neighborhood characteristics that may be spatially correlated with school quality. However, it may produce estimates that are biased downwards since it relies on home sales near school borders, where residents' expectations of future school assignment may be different from contemporary expected school quality in the interior. I use 22,604 homes sales in DeKalb County, Georgia that occurred between 2003 and 2012 to investigate if information about potential school reassignment biases border method estimates downward. During this time, the school system sent differential signals to residents regarding their future school quality by announcing potential areas for reassignment long before the approval of final plans. Exploiting the variation between certain and uncertain borders, I find evidence that school assignment uncertainty reduces capitalization of school quality. The results provide evidence that residents' expectations of future school assignment are important factors contributing to the degree of school quality capitalization and suggest that future research should consider factors that influence people's expectations of school quality.