The Process Innovation Patent Dataset classifies claims of all utility patents awarded by the U.S. Patent and Trademark Office (USPTO) since 1976 into process and non-process. Standardized legalistic language used in the drafting of patent claims allows to accurately distinguish process from non-process claims using textual analysis techniques. A process innovation typically describes a new way to produce an existing good, while a non-process innovation typically describes a new good that did not exist before.

Based on this data, in Bena and Simintzi (2019) and Bena, Ortiz-Molina, and Simintzi (2020), we create measures of process and non-process innovation at firm-year level to explore the link between labor markets and process innovation. A detailed description of the dataset, including its key features and how it was constructed, can be found in these papers.

Bena, Jan and Simintzi, Elena, Machines Could Not Compete with Chinese Labor: Evidence from U.S. Firms’ Innovation (January 20, 2019). Available at SSRN: or

Bena, Jan and Ortiz-Molina, Hernan and Simintzi, Elena, Shielding Firm Value: Employment Protection and Process Innovation (September 3, 2020). Available at SSRN: or

We'll make the dataset available after considering the intended use. The dataset is for academic or other non-commercial purposes only. Please contact us at

Share of process innovation

Share of process innovation by sections of the Cooperative Patent Classification