| This dissertation examines two related questions about the U.S. patent system: (1) Do inventors strategically withhold citations to relevant prior art? (Chapter 2) and (2) How do patent pools influence the rate and direction of innovation? (Chapters 3 and 4). Chapter 2 examines whether patent applicants strategically withhold citations to prior inventions that form the basis for their own inventions (material "prior art"). U.S. patent citation data suggest that applicants withhold between 21 and 32 percent of relevant citations to prior inventions. Variation in withholdings is explained by patent value and patent portfolio size. The data also highlight important differences across technology classes. In particular, firms are significantly less likely to withhold a citation when applying for chemical and drug patents. These results are robust to controlling for unobservable heterogeneity among patent examiners and for the nationality of inventors.;Chapters 3 and 4 use the example of the first patent pool in U.S. history, the Sewing Machine Combination (1856--1877), to examine the effects of patent pools on innovation. Contrary to theoretical predictions, the sewing machine pool appears to have discouraged patenting activity, especially for members of the pool. Objectively quantifiable measures of improvements in the performance of sewing machines indicate that this change was not limited to patents, but that instead rates of innovation slowed during the pool, and only increased again after the pool dissolved. Moreover, the existence of a patent pool seems to have encouraged, rather than discouraged innovation in substitute technologies that were not covered by the pool. Firms that shifted their R&D efforts to such substitute technologies faced higher rates of exit, suggesting that pools may discourage competition. |
