Improving models of wood products plant locations with restricted access data

Abstract

The United States is the world's largest producer and consumer of forest products, but due to industry aggregation and data suppression for privacy reasons, we have limited knowledge of determinants of wood products manufacturing establishments' locations. We use restricted-access microdata to explore location determinants of several disaggregated employer wood products industry groups and compare results to the aggregated employer and non-employer categories in the wood products subsector. Results suggest important but differing relationships between restricted access variables and establishment counts at the local level. While trucking employment is consistently important, the significance and sign of clustering variables, electricity costs, and production wages vary depending on the specific activity. More specifically, Sawmills and Other Wood Products both exhibit within- and across-industry group externalities of agglomeration. However, not all wood product industry groups interact with one another in the same way. Veneer, Plywood, and Engineered Wood establishments show evidence of only benefiting from locating near downstream manufacturers while Other Mill Work establishments benefit from clustering near other similar establishments. The results should be useful to policy makers as they formulate ways to best position natural resources to contribute sustainable jobs to the local region and prevent future industry bottlenecks.