This chapter examines the US government’s intramural research and development over a 40-year period, drawing together multiple human capital, spending, and patent datasets. The US Federal Government innovates along four dimensions: technological, organizational, regulatory, and policy. After discussing these dimensions, we focus on the inputs to and outputs of government intramural technological innovation. We measure innovative effort and results by accounting for government scientists and dollars committed to R&D and patents created with government involvement. Overall, we show that intramural innovations, measured by government-assigned patents, are slightly more original and general, but less cited, than patents awarded to companies and non-government organizations patenting in the same technology classes. The majority of the 200,000 federal government scientists work at the Department of Defense, Department of Energy, and NASA, and are largely in physical science and engineering occupations; other agencies’ scientific expertise is weighted toward mathematics, social sciences, and data analytics. As these latter disciplines’ innovative outputs are less readily patented, measuring federal government innovative output with government-assigned patents is likely to over-emphasize innovations in engineering and physical sciences while underreporting innovations in other disciplines. We discuss implications of our findings for public- and private-sector innovation and identify questions for future research.

We examine how the U.S. Federal Government selects governance structures for R&D contracts with private-sector firms. The government chooses between two contractual forms – grants and cooperative agreements – where the latter provides the government with substantially greater discretion over, and monitoring of, project progress. Using novel data both on R&D contracts and on the technical expertise available in specific government bureau locations, we test implications from the organizational economics and capabilities literature. We find that cooperative agreements are more likely to be used for early-stage projects and when the local government bureau personnel have relevant technical expertise; in turn, cooperative agreements yield greater innovative output as measured by patents and citations, controlling for the endogeneity of contract form. The results are consistent with multi-task agency and transaction cost approaches that emphasize decision rights and monitoring.

Is it better to be a specialist or a generalist over one’s career? Specialization has historically been identified as a successful strategy in human capital enhancement and job signaling. However, recent evidence suggests generalists may be reaping greater economic rewards than expected, due to both skill complementarities with specialists and through increased social capital. This empirical paper compares and tests theories of specialist advantage from labor economic, management, and sociological perspectives using a comprehensive administrative data set on the careers of U.S. federal civil servants over a 22-year period. While specialization is indeed a successful strategy for some employees, career generalism confers benefits on other employees, particularly in settings where specialists outnumber generalists. Furthermore, coherent generalist careers, comprised of diverse experiences confined to a single broad occupational field, are more advantageous than just specializing in a single niche within an occupational area. Atypical careers are also shown benefit employees, though the return to atypicality is substantially reduced if it renders employees incomparable with their local peers. The paper concludes with general observations about the changing nature of employment and proposes future directions for research.