Research Highlights Featured Chart

January 16, 2019

Encouraging innovation

Bigstock/Chris Fleisher

A key question facing competitive markets is how to encourage innovation while protecting researchers’ rights to their work.

It’s a tricky balance. If a patent is overly restrictive, then it will be a disincentive to other researchers who might want to build off of it to invent other technologies that might benefit society. But if patents do too little to protect researchers, then they will be less inclined to develop new work.

These are important trade-offs, since most innovations do not happen in isolation — they build on a collection of discoveries that came before. It’s worth considering, then, whether the patents themselves affect the discoveries that follow them.

A paper in the January issue of the American Economic Review collects new data on mapping gene sequences to determine whether patents affected future scientific innovations.

Authors Bhaven Sampat  and Heidi Williams did find that patented gene sequences had more innovations than non-patented genes. But this has nothing to do with the patent, itself. Rather, it’s selection bias. If a gene seems like it might lead to valuable discoveries, then of course a researcher is going to try to patent it.

 

Figure 2 from Sampat et al. (2019)

 

The panels above from Figure 2 in their paper shows trends in patenting genes and the innovations that build upon those discoveries, as measured by when they’re mentioned in publications and used in clinical trials. Patented genes (represented by the blue triangles) followed the same basic pattern as those in which the researcher applied for a patent, but was not granted one (red circles). Meanwhile, genes that were never claimed in a patent application (green squares) were far less likely to be cited in journals and used in clinical trials.

The paper highlights how it’s not the existence of a patent itself that has a bearing on downstream innovations. Rather it’s more related to the expected value of the discovery.