How Intellectual Property Rights Affect Innovation
Technological progress in health care is a double-edged sword, providing considerable benefits to patients but also contributing very substantially to rising health care costs. The laws governing intellectual property rights, such as patents and copyrights, are one lever the government has at its disposal to influence the rate of technological progress. The traditional motivation for intellectual property is to provide incentives for research on new technologies. However, new products often require several steps of invention and research, implying that intellectual property on an existing technology may affect incentives for research on subsequent applications of and improvements on the initial discovery. Relatively little is known about this question of how intellectual property rights affect cumulative innovation.
In Intellectual Property Rights and Innovation: Evidence from the Human Genome (NBER Working Paper 16213), researcher Heidi Williams explores this relationship by analyzing the sequencing of the human genome.
In the late 1990s and early 2000s, there were two major efforts to sequence the human genome, one by the public Human Genome Project and the other by the private firm Celera. The two groups took different approaches to DNA sequencing, and as a result some genes were sequenced first by the public effort and others were sequenced first by Celera. If a gene was sequenced first by Celera, Celera's intellectual property placed restrictions on how researchers at other institutions could use the gene sequence data, and required some institutions to pay substantial fees in order to access the data. Once a gene was sequenced by the public effort, it was placed in the public domain - with the stated goal of encouraging research and development. By the end of 2003, all genes were in the public domain.
Williams explores whether genes first sequenced by Celera experienced a different level of subsequent innovation than did genes first sequenced by the public effort. Innovation is measured here both in terms of published scientific research and the development of gene-based diagnostic tests.
To examine this question, Williams constructs a unique gene-level data set including the roughly 28,000 known genes on the human genome. For each gene, she records whether the gene was held with Celera's intellectual property for a period of time; this occurred in 6 percent of genes. She then measures the number of published scientific papers related to all genotype-phenotype links for each gene (that is, links between a gene and an observable trait, such as the Huntington gene and Huntington's disease). Next, she identifies whether there is any commercially available genetic test for each genotype-phenotype link. Finally, all data is aggregated to the gene level.
Three strategies are used to test how Celera's intellectual property affected subsequent innovation. The first compares innovation on genes ever held by Celera to that on genes originally sequenced by the public effort. The second traces what happens once the intellectual property protection is removed on genes held by Celera; this approach may help to address the concern that Celera genes could have differed from non-Celera genes (for example, in terms of inherent commercial potential). The third limits the sample to Celera genes, and examines variation in the length of time a gene was held with Celera's intellectual property.
Turning to the results, Williams finds that Celera genes had 35 percent fewer publications over the period 2001 to 2009 than non-Celera genes. Celera genes were also 1.5 percentage points less likely to be used in a diagnostic test by 2009; since only 3 percent of genes are so used, this is a large effect. Results from the second and third empirical approaches are largely similar. Interestingly, the lower levels of innovation on Celera genes persist more than five years after all genes had moved into the public domain.
In short, less scientific knowledge was generated on Celera genes both during and after the time they were held with Celera's intellectual property. One explanation that is consistent with these results is that there are increasing returns to research and development - that is, once scientific knowledge has accumulated in a given area, future discoveries related to that knowledge may be made at lower cost relative to discoveries in other areas where less is known.
Williams notes that her study does not evaluate the overall welfare effects of Celera's entry into the effort to sequence the human genome. Celera's ability to obtain intellectual property likely encouraged the firm to undertake its sequencing effort. To the extent that Celera's entry accelerated the public sequencing effort, Celera's intellectual property could thus have increased total innovation, even if there was reduced innovation on Celera genes. Rather, the results suggest "an alternative institutional mechanism may have had social benefits relative to Celera's chosen form of intellectual property." For example, the public sector could have paid Celera some fee to buy out Celera's property rights and place Celera genes immediately in the public domain. Most broadly, the results suggest "open access to scientific materials may encourage cumulative innovation."
Financial support from the National Institute on Aging through grant T32-AG000186 to the NBER and from the Center for American Political Studies at Harvard is gratefully acknowledged.