Inventing in an uncertain world: biofuel patents and the anticommons

Abstract

In 1998 Heller coined the term "anticommons" as a situation in which overlapping property rights lead to underutilization of a resource. Biofuel technologies offer a unique opportunity to explore Heller's hypothesis as there are a dozen different pathways all leading to the same end: a marketable fuel. Ethanol is the most common biofuel on the market today, but this first-generation biofuel comes at a high environmental cost. Advanced biofuels use non-food crops and other waste materials to create a final fuel, which offers hope when considering the environmental impacts of traditional liquid fuels. The patent rights surrounding each fuel type span from a single owner to many owners and vary across technology pathways. These papers explore biofuel technologies over 60 years via granted patents and show that along all technology and ownership patterns the number of assignees on a patent negatively impact follow-on citations. Paper 1, "The rise and fall of innovation in biofuels" was published in Nature Biotechnology in 2016. This paper is a patent landscape of biofuel technologies from 1960 to 2013. In this paper we show an uptick in biofuel patents in the early 2000's, followed by a slow-down post 2008. These trends vary by location and ownership structure, which is explored further in paper 3. Paper 2 builds on the traditional Nordhaus innovation model in which innovation outputs are a function of inputs. The model is expanded to include the potential for fragmented property rights in the form of diffuse patent ownership. The success of an innovation depends on basic research inputs, such as resources available to the researcher and existing patents for the technology in question. A hopeful inventor must consider the time and cost of negotiating rights to gain access to existing patents. And the risk of getting tied up in litigation, before deciding to move forward with a particular idea. As patent ownership becomes more diffuse private outcomes decrease relative to the social optimum and useful ideas are abandoned. Paper 3 uses the data from paper 1 to test the hypothesis of the anticommons empirically. The model uses count of citing patents as a function of various innovation inputs, including cited patents, inventor and assignee counts, time, and other control variables. The data is divided into technological pathways to illustrate how varying amounts of patent assignees impacts follow-on inventions. In all cases the number of assignees negatively impacts the number of follow on citations, suggesting the anticommons is present.