| Abstract |
Advocates of technology innovation often invoke rhetoric associated with "riding down the technology cost curve," in which technology costs fall as technology deployment increases. These assumed cost reductions, however, require a number of necessary developments to take place within a technology's innovation system. This research looks at shifts in the biofuel technology innovation system over time, and discusses the role that key government policies may have had in promoting successful technology innovation. Through the use of Natural Language Processing alongside machine learning algorithms, we assess shifts in biofuel technology innovation across several hundred firms. The full background text of over 755,000 patents from the U.S. Patent and Trademark Office patent database has been analyzed and classified using the Stanford NLP Classifier. For the case of biofuels, there have been two periods of innovation; one associated with a strong coalition of agricultural firms, and a second period marked by disparate biotechnology firms working to secure a poorly defined niche market. Data show that government policy may have encouraged and facilitated innovation activity for 1st generation biofuels, but may have been largely ineffective at encouraging knowledge development and diffusion for 2nd generation technologies. Our data indicate that 2nd generation biofuels are far from market maturity compared to 1st generation biofuels, and that new government policy approaches may be necessary to better promote knowledge development and diffusion, or use of these 2nd generation biofuels may remain limited. |