Precision vs. Scope: A Methodological Framework for Linking Procurement and Patent Data in Defense Innovation Research

Abstract

Understanding innovation within the U.S. Defense Industrial Base (DIB) is a critical objective of researchers seeking to inform defense procurement policy, yet empirically linking government funding to innovation indicators remains challenging. This paper addresses the core methodological problem of connecting government procurement data with patent data, two systems that are not naturally aligned. We propose a matching framework centered on two distinct data-matching strategies: award linkage (L_A), which matches patents to specific government awards, and organization linkage (L_O), which matches patent assignees to contractor entities. We argue that these strategies present researchers with a structural trade-off. Award linkage offers high precision and defense specificity but suffers from narrow scope and under-classification of innovative activity, while organization linkage provides broad scope to analyze firm-level capacity and knowledge spillovers but risks over-classification by including non–defense-related patents. We develop and validate our matching framework by reviewing how recent literature has navigated data matching challenges, formalizing the most prevalent linkage strategies, and then presenting a practical demonstration of the proposed L_A – L_O framework by analyzing the patenting activity of the government’s SBIR programs over the past decade.