increases. The variable is a measure of the pool of potential candidates for principal and
was expected to increase labor match efficiency. The hypothesis that matching efficiency
could be influenced by Tiebout-type competition between school districts is not
supported by the data. Results across specifications in table 2 indicate the number of
school districts in the metropolitan area does not significantly determine the variation in
teaching experience.
Model III controls for state level fixed effects.10 Public grade schools are
financed and regulated primarily at the state level where minimum qualifications for
principals are set. The point estimate of the effect population size has on skills variation
decreases dramatically in the model although the standard error is largely unchanged; the
covariate turns statistically insignificant. The results imply the level of urbanization
determines skills variation across school districts but cannot be used to explain, at
standard significance levels, strictly within-state variation. The two school district
characteristics, student size and mean principal quality, remain significant after
controlling for the fixed effects.
VI. Conclusion
This study uses an intuitive implication of the standard labor matching hypothesis
to examine the effect urbanization has on matching efficiency. Urbanization is found to
foster more efficient labor matches in the market for public school principal. The
heterogeneity in teaching experience among school principals is found to decrease with
urban size. The effect appears to be large. Although the variation in other skills
measures, primarily previous principal experience, could not be explained by urban
10 The sample for Model III is smaller than in models I or II because those states that had only one school
district in the data were taken out.
11