The division of innovative labor among university, entrepreneurial, and corporate R&D in an analysis of agricultural biotechnology patents

Theory would predict that R&D organizations in different sectors of the economy specialize—according to each one’s comparative advantages—in generating innovations of different qualitative economic characteristics and at different phases in the evolution of a technology over time (i.e. in the technological trajectory). This idea of specialization by R&D sector is tested using U.S. patent data on biotechnology inventions in crop agriculture, categorized into subsets estimated to capture the most significant technological trajectories as they emerged during the formative years of the agricultural biotechnology industry. Established citations-based indices proxy for patents' heterogeneous economic and technical qualities such as value, originality, generality, and appropriability. The results of a discrete random variable regression model show systematic differences in the qualities and timing of innovations across sectors, but they are much more nuanced and interesting than a simple linear model of 'basic' publicsector and 'applied' private-sector innovation. Assuming that these innovations were, at least in some sense, optimally produced by the different sectors, these results may be interpreted as 'revealing' the underlying comparative advantages characteristic of each R&D sector.