آیا صدور مجوز لایسنس، تاثیرات حقیقی بر روی تحقیقات دانشگاهی می گذارد؟ دیدگاه چرخه عمر

Do financial returns to licensing divert faculty from basic research? In a life cycle model in which faculty can conduct basic and/or applied research (the latter can be licensed) licensing increases applied relative to basic effort. However, leisure falls so basic research need not suffer. If applied effort also leads to publishable output, then research output and stock of knowledge are higher with licensing than without. In a tenure system licensing has a positive effect on research output unless license incentives are high. Overall results suggest a positive impact of tenure on research output over the life cycle.

The dramatic growth of entrepreneurial activity associated with university licensing in the last few decades has prompted much debate. While universities tout this as evidence of the increasing role of universities in economic growth, others question whether such activity compromises the basic research mission of universities. For example, a provocative Atlantic Monthly cover story on the “kept” university suggests the increasing trend of university industry deals (such as the Novartis–Berkeley research agreement in the late 1990s) could seriously compromise research agendas, diverting faculty toward research in corporate interests. The increasing trend of faculty to hold positions in startup and corporate boards further suggests faculty may increasingly face conflicts with the primary responsibilities in research ( Boyd et al., 2003 and Zerhouni, 2004). Scholarly analysis of these issues is limited and provides mixed results. Lach and Schankerman (2003) provide empirical support for the view that university research responds to financial incentives, showing that invention disclosures are positively related to the share of license income accruing to inventors.1 However, Thursby and Thursby, 2002 and Thursby and Thursby, in press suggest that increased disclosure activity is more reflective of an increased willingness of faculty to engage in commercial activity than a change in research profile. Their study of faculty in six major research universities shows that over the last two decades, the probability a faculty member will disclose an invention has increased tenfold, while research productivity has remained roughly constant. In essence, despite the importance of the issue, we know little about the effect of faculty involvement in licensing on the nature of research. In this paper, we construct several life cycle models of faculty behavior that allow us to examine this and related issues. In the models we consider, the faculty member faces a fixed teaching load and chooses the amount of time to devote to research (which can be either basic or applied) and the amount of time to take as leisure. We model both the puzzle solving and financial motives for the faculty member to conduct research, and we consider her behavior with and without the possibility of licensing. This allows us to examine the effect of licensing on the research mix, as well as the total amount of time working, throughout the life cycle. We also examine the effect of the tenure decision on the type of research conducted with and without the possibility of licensing. We show that, with or without licensing, and with or without a tenure system, the faculty member devotes more time to research early in her career, so that leisure rises over time. In that sense, licensing does not alter the life cycle pattern. We show that there are, nonetheless, real effects of licensing since it yields a higher ratio of applied to basic effort and lower leisure throughout the life cycle. Thus, as suggested by Lach and Schankerman, faculty respond to economic incentives. Importantly, however, this diversion does not mean that research is compromised. In our models, leisure is the activity most compromised, so that total research effort rises, and in most of the models we consider, basic effort rises with the introduction of licensing. The implications of licensing for research output and the stock of knowledge depend not only on the effect on applied and basic effort, but also on whether applied effort contributes to the stock of knowledge. We show that in the worst case scenario, the applied effort involved with licensing is pure development and adds nothing to the stock of knowledge. If, however, the applied effort involved in licensing leads to publishable output as well as licenses, then the outlook is more favorable. In this case, we show that research output and the stock of knowledge are generally higher with licensing than without. The exception to this is when a tenure system is coupled with very high incentives to license. In Section 1, we discuss prior work in this area and how this paper contributes. Section 2 presents the basic model. Section 3 presents life cycle behavior for three different scenarios: a development model in which only basic effort contributes to the stock of knowledge, a complements model in which basic and applied efforts are complements in the production of both research and licenses, and a model in which basic and applied effort are substitutes in research production. Section 4 presents results when tenure is introduced to the model, and Section 5 concludes.

An important issue in the debates over university licensing is whether the associated financial incentives compromise the research mission of the university by diverting faculty from basic research. In this paper, we argue that understanding the effects of licensing on research requires an understanding of faculty motives in conducting research and how they vary over the life cycle. We construct several life cycle models of faculty behavior that take into account both the puzzle solving and financial motives for faculty to conduct research. In the models we consider, the faculty member faces a fixed teaching load and chooses the amount of time to devote to research (which can be either basic or applied) and the amount of time to take as leisure. We consider her behavior with and without the possibility of licensing. This allows us to examine the effect throughout the life cycle of licensing on the research mix as well as the total amount of time working. We show that, with or without licensing and regardless of the research production functions considered, faculty devote more time to research early in their career, so that leisure rises over time. In that sense, licensing does not alter life cycle patterns. There are, nonetheless, real effects of licensing since it yields a higher ratio of applied to basic effort and lower leisure throughout the life cycle. Thus, as suggested by Lach and Schankerman, faculty respond to economic incentives. This is not to say, however, that licensing compromises research effort. In our models, leisure is the activity most compromised, so total research effort rises, and in most of the models we consider, basic effort rises with the introduction of licensing. The implications of licensing for research output and the stock of knowledge depend on the model specification. The worst case scenario is of course the development model without tenure because in this case applied effort adds nothing to the stock of knowledge. In this case, research output suffers from the introduction of licensing. If, however, the applied effort involved in licensing leads to publishable output as well as licenses, then the outlook is more favorable. In these cases, research output and the stock of knowledge are generally higher with licensing than without. Interestingly, this result is not dependent on the assumption that basic and applied effort are complements in production. It stems, rather, from the fact that applied effort contributes to the stock of knowledge and the complementarity of basic and applied effort with the stock of knowledge in the license production function. The complementarity between research effort and the stock of knowledge also explains the positive view of tenure in our models. Finally, the only negative effect of licensing in these cases is with a tenure system when the incentives to license are extremely high. Several limitations of the analysis should be noted because they present opportunities for future research. First, we focus on a subset of issues related to the “kept” university. Another important issue is how commercial involvement by faculty affects the dissemination of research. There is increasing empirical evidence that sponsored research and/or licensing may lead to delays in publication and/or research that is kept secret, particularly with exclusive licensing (Murray and Stern, 2005, Thursby and Thursby, 2002 and Thursby and Thursby, 2003). We abstract from this since we examine a single faculty member's research. In a more general setting with many researchers, restricted knowledge flows might well alter our conclusions regarding cumulative knowledge for the complements case. Second, we do not examine welfare implications of faculty choices. A welfare analysis is well beyond our scope and would require consideration of many other aspects, such as dissemination patterns as in Mukherjee and Stern (2005), the optimal stage for research to be transferred to industry as in Aghion et al. (2005), and other educational issues (see Stephan and Levin, 1996). Note, as well, in modeling tenure we assume that the faculty member has the freedom to choose research topics so that issues of academic freedom are ignored. Finally, by focusing on the transfer of knowledge through either publication or licensing, we have abstracted from other ways that university researchers interact with industry. For example, we do not differentiate between income from licensing to an established firm and income from a license to a start-up company where the faculty member might take a leave from the university. We have also not discussed faculty consulting, which is common with both start-ups and established firms (Murray, 2002 and Mansfield, 1995). While the model we construct says nothing about the case of the faculty member taking a leave, the results could easily be reinterpreted in terms of consulting. That is, the income the faculty member earns from applied research could be interpreted as consulting income in any of the models we present without loss of generality. If one believes that consulting contributes to the faculty member's income but not to research output, then the appropriate model would be the development model. Alternatively, the complements model is consistent with Mansfield's survey results, showing that consulting often leads to basic research ideas. We believe further analysis of this case is important, since in other work (Thursby et al., 2007 and Jensen et al., 2006) we have found substantial involvement of university faculty in industrial patenting (both in start-up and established firms) while they remain employed in the university.