Speaker(s):  Kira Markiewicz Fabrizio (Berkeley) - CV

Title:   Absorptive Capacity and Innovation: Evidence from Pharmaceutical and Biotechnology Firms

Abstract:
Since the late 1970s, patenting of universities research has gone from nearly non-existent to a significant focus of many research universities, in part due to the Bayh-Dole Act of 1980. This represents a fundamental change from the norms of “open science” which promote quick, full, public disclosure of research results by faculty. Many important university research results are now protected by intellectual property rights, and use of these research results is limited to the patent holders and licensees. Since university research is an important input to the innovation process in many industries, this potential “fencing off” of upstream research has drawn interest.[1] Many researchers have expressed concerns that increased university patenting may limit access to knowledge necessary for innovation, noting restrictions to follow-on development (Mazzoleni and Nelson 1998), reductions in information sharing and collaboration (Louis et al. 2001, Campbell et al. 2002, Blumenthal et al. 1996 & 1986), changes in faculty focus (Feller 1990), and formalization of interactions that may destroy the informal organization of knowledge transfer (Rappert et al. 1999). This dissertation examines firm strategies for identifying, accessing, and exploiting public science and the implications of these strategies for the innovative and economic performance of the firm. In addition, I explore the effect of the increase in patenting by universities on industrial innovation and publication behavior of patenting university faculty members.

My research investigates of the effect of increased university patenting on the use of public science by firms and the lag with which patented knowledge is utilized in follow-on patented innovations. Increased university patenting may slow the pace of knowledge exploitation if university-based research results that are useful as an input to the industrial innovation process are unavailable (due to limited licensing or increased researcher secrecy) or available with delay (due to lengthy licensing negotiations or delayed publication of research results). In addition, limited dissemination of university research knowledge may benefit firms that do gain access relative to firms that are excluded. A first analysis of these issues is focused at the technology class level. Using patent data from the NBER database (Hall et al. 2001), I measure changes in the pace of knowledge exploitation in industrial patents within a given technology class using changes in the distribution of backward citation lags to prior patents. I proxy for exploitation of public science with the count of citations to non-patent prior art contained in these industrial patents, and explore the change in the variance of the number of such citations across firms over time.[2]

Results indicate that an increase in university patenting is associated with an increase in the variance of citation to public science across firms, suggesting the public science is being increasingly channeled to some firms relative to others. I also find that increasing university patenting is associated with a slowdown in the pace of knowledge exploitation in industrial patents, particularly in areas which rely more heavily on science as an input to the innovation process. A 1% increase in the percentage of patents issued to universities is associated with an increase of 2.4% in the mean citation lag. For example, in technology class A61K (which covers pharmaceuticals, cosmetics, and dental preparations) the cumulative percentage of university patents increased by 2.95% between 1985 and 1995. My results suggest that this would be associated with an average increase in the mean citation lag of 8.5 months relative to the 1985 mean citation lag of 9.8 years for patents in this class. This suggests a significant slowdown in the pace of knowledge exploitation in fields in which university patenting has increased since the Bayh-Dole Act.[3]

These results raise the question of what enables some firms to exploit public science more than others. Following on literature exploring the “absorptive capacity” of a firm, or the ability to identify, assimilate, and exploit knowledge from outside of the firm (Cohen and Levinthal 1990), my second paper addresses the exploitation of public science in innovation at the firm level, using panel data set covering 78 firms in the pharmaceutical and biotechnology sectors which includes firm publication, patent, corporate structure, and financial data for the 1975-95 time period. I examine the relationship between firm activities expected to enhance absorptive capacity, such as research intensity, internal basic science research, and collaborations with university scientists, and the exploitation of public science in patented innovations for firms in these sectors, which rely heavily on university science as an input to the innovation process (Mansfield 1991, 1998). I also evaluate the relationship between these firm activities and the pace of knowledge exploitation of the firm. I find evidence that firms with more internal basic science and more collaborations with university scientists cite public science more than other firms and also that patented innovations by these firms exploit existing knowledge more quickly. Finally, my results suggest that citation of public science and a faster pace of knowledge exploitation are associated with a higher market to book value ratio for the firm. Firms that invest in these absorptive capacity-building activities do appear to exploit public science more than other firms, and these firms enjoy advantages in terms of innovative and economic performance.

I follow up on these results with a researcher-level empirical investigation of the relationship between faculty patenting and publishing in order to assess the potential impact of university patenting on the open publication of research results. Using a panel data set containing a broad group of patenting faculty members and a matched control group of non-patenting faculty members constructed with patent data from the NBER database, publication data from the science citation index, and faculty personal information from internet searches, I investigate the publication activity of university researchers following the invention of a patented advance. Results of a negative binomial, researcher-level fixed effects analysis suggest that the number of annual publications by a researcher is significantly greater following patenting, and increases with the cumulative number of patents, controlling for field, number of years since the Ph.D. date, and year effects. These results suggest that publication does not decrease after patenting by a university researcher.  The publication behavior of university patenters therefore does not seem to explain the slowdown in knowledge exploitation documented in my first paper.

These papers collectively add to the literature on knowledge transfer, absorptive capacity, firm knowledge management strategy, and intellectual property rights with substantive empirical evidence of interest to firm managers and policy makers. The increase in formal property rights associated with university research appears to be channeling the flow of university-based input to the industrial innovation process and is potentially slowing the knowledge transfer process. Firm research strategies do have some consequence for the exploitation of public science in firm inventions, and these strategies provide innovative and economic performance benefits. Incentives for faculty to publish research are still strong, even if their research is also patented.

David Blumenthal, E.G. Campbell, M.S. Anderson, N. Causino, and K.S. Louis. Withholding research results in academic lifescience: Evidence from a national survey of faculty, Journal of The American Medical Association, 277(15): 1224-28, 1996.

David Blumenthal, M. Gluck, K.S. Lewis, M.A. Stoto, and D. Wise. University-industry relations in biotechnology: Implications for the university, Science, 232: 1361-1366, 1986.

Eric G. Campbell, Brian R. Clarridge, Manjusha Gokhale, Lauren Birenbaum, Stephan Hilgartner, Neil A. Holtzman, and David Blumenthal. Data withholding in academic genetics, JAMA, 287(4): 473 – 80, 2002.

Wesley M. Cohen and David A. Levinthal. Absorptive Capacity: A new perspective on learning and innovation, Administrative Science Quarterly, 397-422, 1990.

Irwin Feller. Universities as engines of R&D-based economic growth: They think they can, Research Policy, 19: 335-348, 1990.

Bronwyn Hall and Adam Jaffe and Manuel Trajtenberg. The NBER patent citation data file: Lessons, insights and methodological tools, NBER Working Paper #8498, 2001.

K.S. Louis, L.M. Jones, M.S. Anderson, D. Blumenthal, and E.G. Campbell. Entrepreneurship, secrecy, and productivity: A comparison of clinical and non-clinical faculty, Journal of Technology Transfer, 26(3): 233-45, 2001.

Edwin Mansfield. Academic research and industrial innovation, Research Policy, 20: 1-12, 1991.

Edwin Mansfield. Academic research and industrial innovation: An update of empirical findings, Research Policy, 26: 773-76, 1998.

Robert Mazzoleni and Richard R. Nelson. The benefits and costs of strong patent protection: a contribution to the current debate, Research Policy, 27: 273 – 284, 1998.

Brian Rappert, Andrew Webster, and David Charles. Making sense of diversity and reluctance: academic-industrial relations and intellectual property, Research Policy, 28: 873 – 890, 1999.

John P. Walsh, Ashish Arora, and Wesley M. Cohen. Effects of Research Tool Patents and Licensing on Biomedical Innovation, in Cohen and Merrill (eds.), Patents in the Knowledge-Based Economy. The National Academies Press, Washington, DC, 2003.



[1] Work by Mansfield (1991, 1995, 1998), Narin (1992, 1997), and Cohen, Nelson, and Walsh (2000, 2002), has documented the importance of public science, much of which resides in universities, to the industrial innovation process. In 1998, 73% of papers cited by U.S. industrial patents were from public sources, authored at academic, governmental, or other public institutions (Narin et al. 1997). Industrial R&D managers report that a significant percentage of their projects would not be possible or would be delayed significantly without the public science upon which they rely (Mansfield 1991, 1998).

[2] Non-patent prior art citation are dominated by citations to research generated at public institutions, particularly universities (Narin et al. 1997).

[3] This is consistent with comments from biotechnology firm researchers interviewed by Walsh et al. (2003), who report that negotiations with universities over intellectual property rights cause delays of firm research on the order of months.