The Industry R&D Survey: Patent Database Link Project
Executive Summary — The development and diffusion of new innovations are central to economic growth, and understanding the firm-level underpinnings of technology progress is important to academics, policymakers, and business managers. While many researchers have examined (either separately or together) corporate research and development and technology diffusion, they run into two significant data constraints. William R. Kerr and Shihe Fu describe how they developed a new dataset for studying corporate innovation that encompasses three important existing datasets. This paper summarizes the Industry R&D Survey for researchers who want to study innovation through the Census Bureau's data. Key concepts include:
- The developed platform offers an unprecedented view of the R&D-to-patenting innovation process and a close analysis of the strengths and limitations of the Industry R&D Survey.
- This R&S platform can, through the Census Bureau's file structure, facilitate other studies of how innovation leads to productivity.
This paper details the construction of a firm-year panel dataset combining the NBER Patent Dataset with the Industry R&D Survey conducted by the Census Bureau and National Science Foundation. The developed platform offers an unprecedented view of the R&D-to-patenting innovation process and a close analysis of the strengths and limitations of the Industry R&D Survey. The files are linked through a name-matching algorithm customized for uniting the firm names to which patents are assigned with the firm names in Census Bureau's SSEL business registry. Through the Census Bureau's file structure, this R&D platform can be linked to the operating performances of each firm's establishments, further facilitating innovation-to-productivity studies.