The implications of incumbent intellectual property strategies for open source software success and commercialization.

摘要

While open source software (OSS) emphasizes open access to the source code and avoids the use of formal appropriability mechanisms, there has been little understanding of how the existence and exercise of formal intellectual property rights (IPR) such as patents influence the direction of OSS innovation. This dissertation seeks to bridge this gap in prior literature by focusing on two closely related topics. First, it investigates how OSS adoption and production are influenced by IPR enforcement exercised by proprietary incumbents. Second, it examines the impact of royalty-free patent pools contributed by OSS-friendly incumbents on OSS innovation. It particularly looks at the influence of such patent pools on OSS product entry by start-up firms.;The first chapter provides an overview on the dissertation. The second chapter investigates how intellectual property rights (IPR) enforcement against open source software (OSS) affects the adoption and production of related OSS projects. This chapter suggests that when an IPR enforcement action is filed, user interest and developer activity will be negatively affected in two types of related OSS projects—those that display technology overlap with the litigated OSS and business projects that are specific to a focal litigated platform. These hypotheses are examined using two widely publicized lawsuits—SCO v. IBM and FireStar/DataTern v. Red Hat. The empirical analyses are based on data from SourceForge.net. The difference-in-difference estimates show that in the months following the filing of SCO v. IBM, OSS projects that exhibit high technology overlap with the litigated OSS experienced a 15% greater decline in user interest and 60% greater decline in developer activity than projects in the control group; OSS projects that are intended for business and specific to a litigated OSS platform had a 34% greater decline in user interest and 86% larger decrease in developer activity than the control group. Similar results are also found following the filing of FireStar/DataTern v. Red Hat. These results are robust to a variety of robustness checks, including a falsification exercise and subsample analyses.;The third chapter examines whether open source software (OSS) patent pools contributed by large software incumbents influence new OSS product entry by start-up software firms. It argues that increases in the size of the OSS patent pool related to a software segment will facilitate OSS entry by start-up firms into the same segment; further, the marginal effect of the pool on OSS entry will be especially large in software segments where the cumulativeness of innovation is high or where patent ownership in a segment is concentrated. These hypotheses are empirically tested through examining the impacts of a major OSS patent pool—the Patent Commons, established by IBM and a few others in 2005—on OSS entry by 2,054 start-up firms from 1999 to 2009. The empirical results largely support these hypotheses and are robust to adding a variety of controls as well as to GMM instrumental variables estimation.