Spatial econometric analysis of location drivers in a renewable resource-based industry: The U.S. South Lumber Industry

This research identified key factors driving the location of firms in a resource-based industry by expanding on Location Theory, New Economic Geography, and Cluster Theory. First, six latent location factors were identified based on a survey of lumber industry members using factor analysis. An exploratory analysis of sawmill location identified deviations from complete spatial randomness, suggesting the formal incorporation of a spatial dimension in an econometric model. A Bayesian spatial autoregressive approach was used to model the presence of the lumber industry at the county level in the U.S. South. Empirical results are congruent with theory. The strength of the spatial process is weak in the presence of descriptors with a direct geographic dimension. Counties that provide cost advantages, available labor, and appropriate road access are more likely to host the lumber industry. The model was used to identify counties where lumber industry developments could occur with a high probability of success. This research framework can be extended to other industries and regions.