Patents are often the most important asset and the main revenue source for pharmaceutical firms. Typically, patents are built upon the knowledge of other patents. The common perception is that firms use knowledge that spills over from publicly funded research for the development of marketable products. Our study quantifies the knowledge spillovers between non-profit and the private sectors in the pharmaceutical industry.
Methods
The National Bureau for Economic Research (NBER) provides multiple comprehensive datasets compiled from the United States Patent and Trademark Office (USPTO). The dataset we used includes all utility patents from January 1, 1963 through December 30, 1999. Each US patent is assigned a current United States Class Code, which helps identify to which sector the patent applies. To screen for patents that apply to the pharmaceutical industry, we relied upon a variable constructed by the NBER, called Drugs and Medical, using sub-category code 31, “Drugs.” We randomly selected patents within sub-category 31 from 1981, 1984, 1987, 1990, 1993, 1996, and 1999. For reliability, we confirmed that the patent sampled was assigned one of the USPTO class codes related to pharmaceuticals. Our sample consists of 2,274 U.S. patents. We distinguish between patents of for-profit firms and non-profit organizations, as well as between the “type” of knowledge, for-profit or non-profit, that these patents are built upon. Moreover, we collect information about the number of future patents that cites a particular patent (that is, forward citations). For a robustness check, I used a negative binomial regression and Ordinary Least Squares (OLS) regression.
Results
Our results suggest that the impact of non-profit research and the development of knowledge spillovers may not be fully realized immediately after the patent granting process. As such, for-profit firms seem to rely on the research activity of non-profit organizations further downstream. Moreover, when a for-profit firm relies upon at least one non-profit patent, it will lead to 11.58 and 1.24 additional patents further downstream at the industry and the firm levels, respectively.
Conclusions
Through sampling patent data, our research identifies and quantifies knowledge spillovers between non-profit and for-profit patents. We show that when publicly funded research is used as prior knowledge, the number of future patents at the industry and firm levels increases significantly. Since patents allow firms to charge a monopoly price and that consumers, through taxes, pay for the publicly funded research, the results may have policy implications with respect to the price of drugs.