Examining the maldistribution in teacher quality: A spatial analysis of the distribution of credentialed educators in California schools

Teacher quality is a primary factor influencing student achievement, which subsequently affects future earnings. Studies show that quality teachers are not distributed equally across the U.S., resulting in a maldistribution of quality teachers that disfavors minority groups. However, despite analyzing an outcome that involves distribution across geographical space, these studies do not employ the spatial econometric techniques needed to ensure accurate results. Exploratory Spatial Data Analysis (ESDA) and spatial regression analysis are used to test for spatial autocorrelation in the distribution of credentialed teachers throughout California's unified school districts. Those results are compared with a non-spatial regression analysis to uncover the implications of eschewing spatial modeling on these types of data. Spatial econometrics reveal that credentialed teachers are not distributed equally - non-random clustering of teachers exists to the disadvantage of areas with higher populations of traditionally disadvantaged minorities. However, non-spatial techniques overestimate the significance of race and fail to uncover the significance of other important variables affecting the distribution of teacher quality: the distribution of neighboring districts and pre-established student achievement. This reaffirms claims that utilizing non-spatial techniques on spatial data can lead to bias and incorrect estimates.