|dc.description.abstract||Patents offer a unique window into the often abstract world of innovation because patented inventions leave a detailed paper trail. In this dissertation, I first show that the typical patent is actually of negligible value. I examine the relationship between market value and the patent yield from R&D and find that an additional patent per million dollars of R&D results in less than one tenth of a one percent increase in market value for the average firm in the manufacturing sector. Yet some patents are extremely valuable. For firms in the drugs and medical industry, the effect on market value of an additional patent per million dollars of R&D is nearly one-hundred times as large as that for the average firm. This contrast highlights the vast differences in value across patents with different characteristics. I argue that economists who study patent data must use any available signals of patent value to inform their research.
One such signal is involvement in litigation. In the second essay, I study citations to litigated patents to shed light on geographic patterns of knowledge flows for important inventions. Compared to a group of control patents, citations to litigated patents tend to be more local to both the inventor's state and to the trial state. Inventor-state localization rates for litigated patents are large and do not fade across time. Trial-state localization effects may increase with the onset and the conclusion of litigation; inventor-state localization effects do not. Also, litigation tends to geographically "follow" citations.
In the third essay, I study patent enforcement over the course of several decades and test the responses of litigation risk, duration, and outcomes to various patent characteristics and to significant changes in patent policy and precedent. I find that litigated patents differ significantly from typical patents in several observable ways. I also find that several patent characteristics are systematically related to the duration and outcome of litigation. With respect to patent policy and precedent changes, I find that "flash of genius" era litigation is disproportionately difficult for corporate patentees in terms of duration but not necessarily in outcomes. The Patent Act of 1952 realigns the duration of litigation for corporate and individual patentees but invites a rash of submarine patenting behavior.||