Patents are bad for follow-on innovation

Nordhaus-style models of optimal patent policy design have traditionally modeled innovations as isolated discoveries. However, in practice, many or most innovations are cumulative, in the sense that a given discovery may be an input into later follow-on discoveries. When innovation is cumulative, optimal patent policy design also depends on how patents on existing technologies affect follow-on innovation. One prominent example in recent policy debates is a patented human gene sequence. Sequenced genetic data are a research input into subsequent follow-on discoveries: by analyzing sequenced genetic data, scientists may discover links between genetic variations and diseases, and such knowledge can be applied to commercialize medical technologies such as pharmaceutical treatments and diagnostic tests. If patents on discoveries such as human genes affect follow-on innovation, that effect is a key additional input into optimal patent policy design. In this paper, we investigate whether patents on human genes have affected follow-on scientific research and product development. Our broad goal is to inform whether the Nordhaus-style tradeoff between ex ante incentives and deadweight loss is sufficient for optimal patent policy design, or whether, at least in this context, the effects of patents on follow-on innovation need to be considered. Investigating how patents on existing technologies affect follow-on innovation requires addressing two key challenges. First, in most markets it is difficult or impossible to take a given set of technologies which are claimed as intellectual property in patents and measure follow-on innovation on those technologies. Follow-on innovations which infringe the original patent legally require a license from the original innovator. But because licenses are very rarely publicly disclosed, constructing appropriate measures of follow-on innovation for any given patented invention is quite difficult. Second, a priori we expect a selection bias problem to arise if inventors are more likely to file for or to obtain patents on technologies that are inherently more valuable. This type of selection raises the concern that any measured differences in follow-on innovation across patented and non-patented technologies may reflect the selection of which technologies are patented, rather than a causal effect of patents on follow-on innovation.