Evaluating and extending innovation indicators for innovation policy

Innovation policymakers have struggled with the problem of how to develop better indicators to capture innovative activity. This study reviews and evaluates a variety of commonly used innovation indicators such as R&D, patents, and innovation. Then, using data from multiple data sources, such as US Patent and Trademark Office patent data, National Science Foundation surveys, private US innovation surveys, and Community Innovation Survey data, this study examines the overlap and differences across these indicators. Furthermore, this study suggests the need for a broader understanding of innovation and introduces new measures that complement existing measures to better capture the full population of innovations. These new measures focus on the distinction between innovations originating from R&D and those from outside of R&D (non-R&D innovation). The results show that 12% of triadically patented inventions come from non-R&D, and among US manufacturing firms, 11% of internally generated new-to-market innovations are non-R&D. This study contributes to better conceptualizing innovation by cross-validating innovation indicators, developing innovation indicators beyond the R&D-based perspective on innovation, and emphasizing the importance of understanding the broad universe of innovative activity.