A Value Chain Approach to Measuring the U.S. Production Structure

This paper develops a framework to examine the factor composition embedded in U.S. value added across final demand components by integrating value-added data with Integrated Industry-Level Production Account (KLEMS) statistics. This approach quantifies factor-specific value added within U.S. global value chains (GVCs) using the U.S. Bureau of Economic Analysis’ (BEA’s) domestic requirements tables and the BEA-BLS Integrated Industry-Level Production Account. The decomposition tracks both direct and indirect factor shares across upstream networks, revealing several patterns across final demand categories: exports are most intensive in research and development (R&D) capital, investment relies heavily on non-college labor, government spending has the highest share of college-educated labor, and consumption is least intensive in R&D and information technology (IT) capital. Moreover, investment, consumption, and government spending all contain notable embedded indirectly imported content, exceeding 5%. Sectoral comparisons highlight structural asymmetries: chemicals have the highest direct R&D intensity, financial services intensively use college-educated labor, and transportation and textiles primarily rely on non-college labor. These divergences highlight how upstream industries differ significantly from final-stage producers in factor intensity.