تجاری سازی آزمایشگاه: اختراعات دانشکده و محیط علمی باز

This paper investigates the relationship between patenting and publication of research results by university faculty members. Our study adds to the limited evidence on this topic with an empirical investigation based on a panel data set for a broad sample of university researchers. Results suggest that publication and patenting are complementary, not substitute, activities for faculty members. This is not consistent with recent concerns regarding deleterious effects of patenting on the research output of faculty members. Average citations to publications, however, appear to decline for repeat patenters, suggesting either a decrease in quality or restrictions on use associated in patent protection.

University research has traditionally been associated with norms of rapid disclosure of research results and an environment of knowledge sharing, co-authorship, and joint projects that contribute to cumulative learning and innovation (Dasgupta and David, 1994). The reward system for faculty provides incentives for being first to discover and disclose new research results, and valuable reputations are built on diffusion of these results among the scientific community. Faculty members therefore have the incentive to pursue impactful research problems and disclose research results openly and quickly (David, 1998 and David et al., 1992). These activities stimulate the cumulative research process by facilitating the generating of fundamental discoveries and the transfer of knowledge among researchers. These norms and incentives are being challenged by the increasing patenting of university research results. In this study, we examine the relationship between a faculty member's involvement in patenting research results and the quantity and quality of publications generated. Our goal is to evaluate the recent concerns that patenting negatively affects the generation and publication of academic research. This is an important question because academic research results contribute substantially to the industrial R&D process (Mansfield, 1991, Mansfield, 1995, Mansfield, 1998 and Narin et al., 1997). To varying degrees, respondents in all industries report that a significant percentage of their product and process innovations could not have been developed without academic research, or would have been substantially delayed. University research was reported as an important source of new projects for 31% of respondents and an important source of information for 36% of respondents (Cohen et al., 2002). In their survey of R&D managers across many industries, Cohen et al. (2002) find that the most important channel for knowledge transfer from universities or government labs is publication of the research, followed by informal exchange, public meetings or conferences, and consulting. Patents were ranked as one of the least important transfer channels.1 Therefore, it is important both that academics continue to pursue fundamental research question and that they continue to publish their research findings. There is reason to believe that the influence of patenting and commercialization activity on university researchers has grown in recent years. Since the late 1970s, the university research environment has changed dramatically. Policy changes, such as the Bayh-Dole Act of 1980,2 encouraged commercialization of university research results by standardizing policies granting intellectual property rights to university researchers for outcomes of federally under research.3 At the time, federally funded research made up approximately 70% of university research.4 In addition, sources of academic research funding shifted substantially from the 1970s to the 1980s to include more industry funding and less government funding (Mansfield, 1995).5 Universities underwent organizational changes to attempt to better promote the commercialization of university research results (Argyres and Leibeskind, 1998). In response to these policy, institutional, and market changes, patenting by university researchers increased dramatically in the last two decades (Henderson et al., 1998a and Henderson et al., 1998b). This increase in patent volume was accompanied by a significant increase in the number of universities that were patenting innovations Henderson et al. (1998a).6 Providing researchers with the incentive to patent their findings may encourage commercialization of research findings, as intended by the Bayh-Dole Act. However, the increased focus on patenting may also lead to increased secrecy, delayed publication, and a shift in faculty research effort and time away from scientific research and toward commercialization activities, thereby decreasing the production and/or quality of scientific publications. The existing empirical studies that attempt to uncover the relationship between academic patenting and publishing activities have presented contradictory pictures. Several surveys of academic researchers have documented the secrecy, delay of publication, and re-focusing of research activities that accompany involvement in patenting and commercialization activities (Blumenthal et al., 1996a, Blumenthal et al., 1996b and Campbell et al., 2002). However, the few existing studies of individual research publishing and patenting activities have failed to find such a negative impact. Agrawal and Henderson (2002) examine the patenting and publishing of faculty members in two departments at MIT (Mechanical and Electrical Engineering), and find no relationship between patenting and the generation of publications, but do find that the number of citations to publications is positively related to the level of patenting, which they interpret as evidence that patenting is positively correlated with the impact of research. Azoulay et al. (2004) examine the publication activity of faculty members in the life sciences field and find a positive relationship between patenting and the generation of publication, but do not examine citations to the publications. Positive correlations between publishing and patenting can be found in recent empirical investigations of non-U.S. context (Breschi et al., 2005, Buenstorf, 2006 and Looy et al., 2005). Interestingly, Buenstorf's investigation of publishing behavior of Max Plant Institutes’ Directors show that the positive correlation between publishing and patenting is somehow weakened in the few cases where the disclosed inventions were commercialized. This study extends the existing literature in several ways. First, we draw from a broad sample of institutions and fields in which patenting is occurring, rather than focusing on one institution or one academic area. We develop a matched sample of non-patenting faculty researchers, and employ researcher-level fixed effects to estimate the relationship between patenting and annual publications or average citations to publications. This generates results that are more generalizable than some of the other studies. Second, we investigate the differential impacts of university, industry, and unassigned patents. Third, we examine the possibility that the number of patents held has a curvilinear relationship with the number of publications generated. Finally, our analysis controls for current researcher quality differently than this existing work and results suggest that these controls are important. Our findings are consistent with most of these studies, importantly contradict some results, and extend this body of literature. Results presented here indicate that publication and patenting are complementary activities, but the positive relationship between them declines with an increasing number of cumulative patents held by the faculty researcher. Although we caution against causal interpretation of these results, this evidence suggests that faculty researchers are not generating less scientific research or limiting the open publication of research results as they patent, with the possible exception of a few very patent-intensive faculty members. Results of the analysis of citations to faculty publications indicate that the number of citations to a faculty members’ publications is related to the productivity of the researchers, but not substantially different following the first patent by the faculty member. In fact, the citation intensity falls as a faculty member repeatedly patents his or her research results. This is contradictory to the limited previous evidence. Although we are focused on average quality and impact of publications rather than the citation patterns associated with an individual paper, our results are consistent with the anti-commons hypothesis put forth and tested by Murray and Stern (2005). The remainder of the paper is organized as follows. The next section considers the possible relationship between faculty publishing and patenting activities and develops the hypotheses to be tested. Section 3 describes our sample and data, summary statistics of which are reported in Section 4. Section 5 presents our empirical strategy and results with respect to the production of publications by faculty members. Section 6 continues by exploring the relationship between patenting and the number of citations to these publications. Limitations of the study are discussed in Section 7, and conclusions are provided in Section 8.

The Bayh-Dole Act and increasing collaboration between industry and university researchers have served to dramatically increase patenting by university faculty members. As patenting has grown, so has concern about the possible negative effects on the generation and publication of fundamental research results by faculty members. How the change in norms has impacted the open publication and use of university research results is a matter of concern for policy makers and firm managers in need of accessing university research results. This research contributes to the existing literature by providing an empirical investigation of faculty patenting and publishing with a data set containing inventors from multiple fields at many institutions in many fields across many years. Our results confirm the results in some existing work, contradict others, and extend the existing literature. The results presented here are not consistent with the concern that the publication activity of faculty researchers would fall with increasing patenting activity. The annual number of publications by a faculty member is in fact higher following application for a successful patent, controlling for field, year, and time profile of publications by matched non-inventors. Again, we stress that this should be interpreted as a correlation, as we cannot test or assert causation from this analysis. This result is consistent with other recent studies of faculty patenting. Interestingly, results here demonstrate that university patents (and not corporate or unassigned patents) drive the positive relationship. This suggests that the positive relationship is not attributable to academic researchers gaining new insights and ideas from working with industry researchers. If that were the case, we should expect to see at least an equal positive relationship between industry-assigned patents and faculty publications. Since the same research often underlies publications and patents, this result may indicate that the research embodied in university-assigned patents is closer to academic research, which yields subsequent publications. The results may also be driven by university technology licensing offices electing to patent the most promising research results, which are also associated with more publications, and passing on less promising results, which are then either patented without an assignee or picked up by a corporation that appears as the assignee on the granted patent. Further research into the mechanisms responsible for the differences across types of patents is needed to examine these possibilities. The relationship between patenting and citations to publications suggests that there is not a large-scale shift to lower quality or less impactful research following an inventor's first patent. However, faculty members that repeatedly patent their research results generate publications that receive fewer citations. This may be an indication that these faculty members are in fact re-focusing their research on more applied or commercializable research, at the expense of fundamental science. It may also be a result of increasing intellectual property rights inhibiting the use of their published research results in follow-on studies. This evidence is contradictory to the limited prior evidence concerning the relationship between patenting and the quality or impact of faculty publications, but is consistent with the results of Murray and Stern (2005), who present an analysis of matched publication and patents covering the same research result. Their results suggest that citations to papers decrease when a patent related to the same research is granted. This is consistent with the interpretation that increasing patent protection to a faculty member's research results is associated with fewer citations to that individual's publications. Comparing these results with Agrawal and Henderson (2002) suggests that the relationship between faculty patenting and publishing may be different for faculty members in general than it is for faculty at elite institutions, such as MIT. Perhaps one reason is that patenting, licensing, and commercialization activities may allow a faculty member access to more industry funding. MIT faculty may not be as constrained by limited funding as researchers at other schools, and they therefore benefit less from added access to funding. There are likely to be other institutional differences driving the variation in results as well. In a comparison between Stanford and Berkeley Electrical Engineering and Computer Science departments, Kenney and Goe (2003) find that a different attitude towards commercialization partially explains the more intense technology transfer activities at Stanford. In Europe, Debackere and Veugelers (2005) find convincing evidence that universities allocating a higher percentage of royalties to inventors are more effective in technology transfer activities. Further research will consider the alternative explanations for the noted increase in publications following a patent application, as well as an evaluation of possible differences across institutions of varying quality. Based on the evidence presented here, increasing faculty members’ incentives to patent research results does not seem to be overwhelming the incentives for faculty members to publish their research.