Changes
Jump to navigation
Jump to search
Murray Stern (2007) - Do Formal Intellectual Property Rights Hinder The Free Flow Of Scientific Knowledge (view source)
Revision as of 19:02, 27 March 2013
, 19:02, 27 March 2013no edit summary
*This page is referenced in the [[Patent Thicket Literature Review]]
*This page is listed on the [[PTLR Core Papers]] page
Although many scholars suggest that IPR has a positive effect on cumulative innovation, a growing anticommons perspective highlights the negative role of IPR over scientific knowledge. At its core, this debate is centered on how intellectual property rights over a given piece of knowledge affect the propensity of future researchers to build upon that knowledge in their own scientific research activities. This article frames this issue around the concept of dual knowledge, in which a single discovery may contribute to both scientific research and useful commercial applications, and finds evidence for a modest anti-commons effect. A key implication of dual knowledge is that it may be simultaneously instantiated as a scientific research article and as a patent. Such patent-paper pairs are at the heart of our empirical strategy.We exploit the fact that patents are granted with a substantial lag, often many years after the knowledge is initially disclosed through paper publication. The knowledge associated with a patent-paper pair therefore diffuses within two distinct intellectual property environments, one associated with the pre-grant period and another after formal IP rights are granted. Relative to the expected citation pattern for publications with a given quality level, the anti-commons perspective suggests that the citation rate for a scientific publication should fall after formal IP rights associated with that publication are granted. Employing a differences-in-differences estimator for 169 patent-paper pairs (and including a control group of other publications from the same journal for which no patent is granted), we find evidence for a modest anti-commons effect (the citation rate after the patent grant declines by approximately 10 to 20 percent). This decline becomes more pronounced with the number of years elapsed since the date of the patent grant and is particularly salient for articles authored by researchers with public sector affiliations.
==Review==
===Measures of thicket===
Patent thickets are expected to be associated with:
*Granting of patent potentially reflects thickets in the context of reducing flow to public knowledge (as measued by forward citations to the article covering the invention)
*Broader patents in complex areas with a high degree of prior art
*Patents over research tools
===Sample===
*340 peer-reviewed scientific articles between 1997 and 1999 appearing in "Nature Biotechnology", a high quality journal that focusses on real-world biotechnology applications.
**A subset of 169 of these articles have a corresponding patent grant as of October 2003;
**Patents from the USPTO are paired with articles by matching authors and insitutions in the article with those in patents, a reading of the article's abstract and patent's claims;
**Citation counts are drawn from Science Citation Index Expanded (SCI) and traced until the end of 2002.
*A total of 1,688 article year observations are analyzed.
===Results===
*Patenting has negative impact on flow of knowledge into public domain as measured by patent citations in the article associated with a patent grant.
:''"while patented articles enjoy approximately a 20 percent overall citation boost, a patent grant is associated with an insignificant (though negative) impact on citation in the year in which the patent is granted. However the post-grant effect is associated with a statistically significant 19 percent decline in the expected citation rate. In other words, the initially higher citation rate for patent-paper pairs is erased in the years after a patent is granted."''
*The effect is close to 10 percent in a differences-in-differences approach with fixed effects for each article.
*Tracked by year, the average citations appear to decline dramatically following patent grant (see Figure 4 of article), suggeting patent grant is cause.
*Decline in citations is driven by patent grants to public affiliated authors, suggesting that grants to private affiliated authors were expected by the research community.
*The article does not find strong evidence that other patent thicket measures impede the flow of knowledge into public domain as measured by patent citations.
:''"...there is a statistically significant (though quantitatively) modest incremental impact of longer patent lags on subsequent citations and a much more salient effect for non-research tools. In other words, though there is some evidence for interactions with patent characteristics, the findings are quite noisy; furthermore, though a key tenet of the anti-commons theory is that the effects are particularly salient for research tool patents, there is no evidence that the impact of patent grant is significant for these types of inventions."''
*Citations for articles with granted patents, interacted with patent granting lag is a small -0.0002%, but significant.
*Citations for acticles with Non-research tools (expected to be of less value commercially) declines by 19.7% significantly, but research tools term is insignificant.
===Social Welfare Consequences===
:''"there is robust evidence for a quantitatively modest but statistically significant anti-commons effect; across different specifications, the article citation rate declines by approximately 10 to 20 percent after a patent grant...With that said, the use of citation data is only a noisy indicator of the impact of any given piece of research, and our approach does not separately identify any potentially positive impact of IPR on research incentives (from the perspective of the original inventor)."''
===Dependent Variable and Model===
*The dependent variable is the number of follow-on ("forward") citations in public/peer-reviewed academic publications that cite to the original paper in a given year.
*A negative binomial model is used for analysis.
*Anti-commons/patent thicket related variables include:
**Public sector assignee of (at least one of the patent's)inventors;
**Government funded research;
**Patent complexity as measured by the number of claims in the patent and length of time (in days) between application and grant of patent;
**Number of citations to prior patents;
**Number of references to prior non-patent arts;
**Tools patent (patents in 3-digit technology classes 435 or 800 are coded as one, with remaining 9 classes as zero).
*Control variables for patent characteristics are:
**An indicator of whether a patent was in force in a given year;
**An indicator of whether the patent was granted in the year;
**In some specifications, an indicator for each year before and after the grant is used to assess when the "news" of a grant became public;
*Other controls include:
**Article fixed effects to control for paper quality;
**Age of the article (using fixed effects in some specifications and age and its square in others);
**Fixed effects for observataion year;
**Number of authors;
**US Author;
**Public sector affiliation of (at least one of the article's) authors;
**Private sector affiliation of (at least one of the article's) authors.
