Seokbeom Kwon, a systems management engineering professor at Sungkyunkwan University in South Korea, published a journal article in Science claiming that over a quarter of U.S. patents on inventions derived from federally funded research do not disclose that funding as required by law. Kwon also claimed that research is less likely to acknowledge government research the larger the “private stake” in its application. Kwon’s assertions fueled anti-IP narratives and seemed designed to provide lawmakers with a pretext to enact new regulations that would burden collaborations between the public and private sectors.
However, the study’s methodology and reasoning are deeply flawed. A closer examination of Kwon’s research methods — and the assumptions implicit in them — reveals why policymakers should not take his sweeping conclusions at face value.
| Claim: Disparities between patent-paper pairs must be evidence of wrongdoing. |
| In reality: The Science article examines pairs of academic papers and patent filings — known as patent-paper pairs (PPPs) — under the assumption that if the patent and paper are associated with the same research, they must make the same disclosures. However, papers and patents operate under different standards. Disclosure requirements for patents are governed by law, while requirements for papers are largely decided by the specific conventions of each journal. Consequently, disparities between patents and papers are to be expected, and patents and papers that do not include the same information as one another can both simultaneously be truthful. The Science article’s interpretation that these disparities are necessarily dishonest falsely inflates its claimed rate of non-reporting. |
| Claim: Inventorship on a patent is equivalent to authorship of an academic article. |
| In reality: The Science article’s research methodology makes the implicit assumption that inventorship and authorship must be equivalent when it raises the alarm about grant recipients who are listed as an author on a paper but not as an inventor on the affiliated patent. Yet this is an incorrect assumption to make as, once again, patents are held to different standards than papers. U.S. patent laws are strict about who may be listed on a patent, requiring that only “true and only” inventors be named, and applicants who fail to comply with these rules may have their patents invalidated. On the other hand, who is listed as an author or contributor to a paper is affected by many factors — including traditions of academia — and has been rising steadily over time, meaning that less significant contributors may be included in a paper even when they are not eligible to be listed on a patent. Therefore, if a government-funded researcher is listed as an author on a paper but the corresponding patent does not disclose that person’s funding, it could simply be due to the differing rules regarding authorship and inventorship, rather than being evidence of deceit. |
| Claim: The publication of academic papers always precedes the patenting of an invention. |
| In reality: While the Science article operates under the assumption that all patents must report whatever government grants are referenced in the corresponding academic article, this is not necessarily the case because government funds may sometimes enter the equation after a patent has already been filed. For instance, if an invention is developed and patented using only private-sector funding, government funding may be awarded to follow-on research, which results in a published journal article. In that scenario, there would be no reason to expect a disclosure of federal funding on the patent, and there would be nothing untoward about the discrepancy between the patent and the article. |