Inference Based on Alternative Bootstrapping Methods in Spatial Models with an Application to County Income Growth in the United States

This study examines correlates with aggregate county income growth across the 48 contiguous states from 1990 to 2001. Since visual inspection of the variable to be explained shows a clear spatial relationship and to control for potentially endogenous variables, we estimate a two-stage spatial error model. Given the lack of theoretical and asymptotic results for such models, we propose and implement a number of spatial bootstrap algorithms, including one allowing for heteroskedasticity, to infer parameter significance. Among the results of a comparison of the marginal effects in rural versus non-rural counties, we find that outdoor recreation and natural amenities favor positive growth in rural counties, densely populated rural areas enjoy stronger growth, and property taxes correlate negatively with rural growth.