Estimation of Industry-level Productivity with Cross-sectional Dependence by Using Spatial Analysis

We examine aggregate productivity in the presence of inter-sectoral linkages. Cross-sectional dependence is inevitable among industries, in which each sector serves as a supplier to the other sectors. However, the chains of such interconnections cause indirect relationship among industries. Spatial analysis is one of the approaches to address cross-sectional dependence by using a priori a specified spatial weights matrix. We exploit the linkage patterns from the input-output tables and use them to assign spatial weights to describe the economic interdependencies. By using the spatial weights matrix, we can estimate the industry-level production functions and productivity of the U.S. from 1947 to 2010. Cross-sectional dependencies are the consequences of indirect effects, and they reflect the interactions among industries linked via their supply chain networks result in larger output elasticities as well as scale effects for the networked production processes. However, productivity growth estimates are reportedly comparable across various spatial and non-spatial model specications.