A Comprehensive Data Set of Published US Patent Litigation Decisions: 1929-2006

点击翻译按钮获取中文摘要

In the United States we have recently seen significant changes in how patents are used and contested. For example, we have seen substantial increases in the number of patents issued and in litigation of new patents. These trends have led economists and legal scholars to argue that the US patent system is failing to achieve the Constitutional mandate 'to promote the progress of science and useful arts', and to propose policy changes. Their main argument is that too many patents are issued with too strong a presumption of validity, resulting in frequent and costly legal disputes that impose unnecessary costs on innovation.The court system is a key institution for determining how patents are enforced, through changes in policy and in legal precedent. It seems obvious that if we want to consider changes in patent policy that will reduce the number of legal disputes, we should first ask how past changes in patent law have affected actual legal outcomes. Numerous changes to patent law since the 1930s have affected patentability, damages, antitrust implications of patents, judicial procedure and other features. However, we know very little about how these changes have affected actual rates of enforcement. The major reason is the lack of a comprehensive data set of US patent litigation.This project fills that gap. The new dataset covers the time period between 1929 and 2006. It includes all decisions published by a district and/or appellate court in patent cases about patent validity or infringement. It also includes a matched set of random patents that were not litigated, which will allow researchers to compare the characteristics of litigated and unlitigated patents. The data is gathered from primary sources that are not machine-readable. The data will be publicly available to all interested researchers via the internet.These data will allow the many researchers interested in patent policy to conduct advanced statistical inquiries. The final result will be precise statistical estimates of rates of patent validity and patent infringement, conditional on specific characteristics of the patent and the litigation. It will be possible to study differences in litigated and non-litigated patents, test predictions about what kinds of patents end up in litigation, and study strategic legal decisions made during the actual litigations. Most importantly, the data span a long period of time. As a result, scholars will be able to measure the effect of the numerous legal changes from 1929 to 2006 on actual legal outcomes. The broader impact of this kind of research is key insights into the likely effects of contemporary proposals for patent reform.