New Insights Into the U.S. Production Structure
A recent U.S. Bureau of Economic Analysis (BEA) research paper, “A Value Chain Approach to Measuring the U.S. Production Structure,” introduces a new way to analyze how different types of labor and capital are used throughout the U.S. production system.
The approach traces the journey of goods and services through all upstream industries involved in their creation, providing valuable insights into the processes leading to the final use of products. By combining value-added data with detailed industry-level statistics, authors Jon D. Samuels and Gonca Senel are able to track not just the direct use of factors such as labor and capital in final products but also their indirect use in all the upstream industries that contribute to those products.
This research takes a threefold approach. First, the authors focus on the baskets of final demand (consumption, investment, government spending, and exports), which are central to how gross domestic product is measured and analyzed in the United States. Understanding the production structure behind these baskets helps clarify the origins of economic value and the demand for different factors of production.
Second, the paper emphasizes the importance of considering not just the final stage of production, but also the entire network of upstream industries that supply intermediate inputs. For example, while farm products may not directly use much research and development (R&D) capital, they rely heavily on chemical products, which are themselves R&D intensive. By tracing the factor content of these upstream inputs, the authors can estimate how much labor, R&D, and other forms of capital are truly embedded in each component of final demand.
Third, the paper seeks to distinguish between value added by domestic U.S. producers and value that is imported from abroad. This is important for understanding the true domestic content of U.S. exports and other final demand categories, as well as for analyzing the role of global value chains in the U.S. economy.
The findings reveal several clear patterns. Over time, the share of noncollege labor in value added has steadily declined across all components of final demand, and this share is more volatile than other inputs, suggesting it is more sensitive to economic cycles and sectoral changes. In contrast, the shares of software and R&D capital have gradually increased, indicating a slow shift toward more innovation-intensive production.
Looking at specific final demand categories, exports are found to be the most R&D intensive, while government spending has the highest share of college-educated labor. Investment relies most heavily on noncollege labor, even though this share is declining. Consumption, on the other hand, is least intensive in information technology and R&D capital, likely because it includes many basic consumer goods that do not require much innovation input.
The analysis also shows that investment, consumption, and government spending all contain significant amounts of indirectly imported content—more than 5 percent—that a notable portion of the value in these categories originates from abroad, even if it is not directly imported.
Sectoral comparisons highlight important structural differences. Chemicals have the highest direct R&D intensity, financial services make the most use of college-educated labor, and transportation and textiles rely primarily on noncollege labor. These differences underscore the importance of looking beyond the final stage of production to understand how resources are used throughout the value chain.
The paper concludes that focusing only on the final stage of production can give a misleading picture of the resources required to produce goods and services. By accounting for the entire upstream production structure, analysts and policymakers can gain a more accurate understanding of the U.S. economy and its evolving structure. The authors suggest that future research should expand to cover more historical periods, break down capital and labor into more detailed components, and further examine the contributions of goods and services within each upstream share.
