1. University Patenting: Estimating the Diminishing Breadth of Knowledge Diffusion and Consumption Saturday, September 30, 2006 EPFL Lausanne, Switzerland Ajay Agrawal University of Toronto

2. Introduction

3. A Striking Feature … and Three Concerns Rapid rise of university patenting since 1980 Concerns associated with patenting: 1.A shift in focus from “basic” to “applied” research 2.A decline in the quality of university inventions 3.A decline in the dissemination of knowledge associated with university inventions due to the anti- commons problem

4. The First Two Concerns Empirical evidence to date offers little support for the first two of these concerns –Empirical studies that examine whether professors substitute patenting for publishing, a rough proxy for changes in research focus, do not provide evidence of such substitution Van Looy et al (2005), Buenstorf (2005), Carayol (2005), Breschi et al (2005); Agrawal and Henderson (2002); Markiewicz and DiMinin (2005); Goldfarb et al (2006) –Although the quality of inventions did decline after 1980 (Henderson et al, 1998), this was due to the entry of universities with little patenting experience; it was not due to a general decline in quality of inventions patented by all universities Mowery et al (2004)

5. The Third Concern: Retarding the Flow of Knowledge Evidence does support the third concern –Murray and Stern (2005) The authors employ a difference-in-differences identification based on patent-paper pairs Although publications linked to patents are associated with a higher overall citation rate, after the patent actually issues the rate declines substantially (by 9-17%) The authors note that the decline is particularly salient for articles authored by researchers with public-sector affiliations, such as university professors They interpret their findings as evidence of an anti-commons effect that results from moving intellectual property from the public into the private domain

6. Our Contribution (part 1) We build on this prior work and further investigate the third concern: retarding the widespread flow of knowledge associated with university inventions However, where Murray and Stern focus on the decline in the level of knowledge flows, we focus on the narrowing of knowledge flows to a smaller set of recipients Specifically, we examine whether, over time and conditional on being patented, university inventions are more likely to be cited by a more concentrated set of subsequent patent owners Such a finding would reflect the outcome of a change over time in the management objectives of university intellectual property, reflecting less emphasis on the broad dissemination of new knowledge and more towards limiting access, perhaps to maximize private returns to licensees.

7. Our Findings (part 1) Method –Herfindahl-type measure of patent assignee concentration –Difference-in-differences estimation (taking the difference of the change in concentrations over time between university versus firm patents) University diffusion premium (the degree to which knowledge flows from patented university inventions are more widely distributed than those of firms) declined by over half between the early and late 1980s Unlike the decline in invention quality that occurred during the 1980s, the increase in knowledge flow concentration we discover is driven by experienced universities This finding suggests that the phenomenon we identify is unlikely to disappear with time but may actually increase as inexperienced universities become more like their experienced counterparts with respect to the manner in which they manage their intellectual property

8. Our Contribution (part 2) We also study the pattern of flows into university inventions While firms may consciously conduct R&D in a manner to minimize exposure to the anti-commons problem, we expect university researchers are largely insulated for two reasons: 1.Universities have traditionally been shielded from patent infringement liability by the doctrine of ``experimental use exemption'‘ 2.To the extent that university researchers choose their research projects to advance knowledge and only concern themselves with patenting ex post, their project selection and prior art decisions will not be influenced by potential anti-commons problems –We examine whether there has been a decrease in the breadth of sources of knowledge inputs

9. Our Findings (part 2) Method –Herfindahl-type measure of patent assignee concentration –Difference-in-differences estimation (taking the difference of the change in concentrations over time between university versus firm patents) University diversity premium (the degree to which knowledge inflows used to develop patented university inventions are drawn from a less concentrated set of prior art holders than those used by firms) declined by over half between the early 1980s and early 1990s The estimated increase in knowledge inflow concentration is also driven by experienced universities, again suggesting that this phenomenon is not likely to dissipate with experience, but may actually increase over time

10. Methodology

11. Empirical Objective Our empirical objective is to test whether knowledge flows associated with patented university inventions become more concentrated over time Therefore, most importantly, we need –an estimation technique that enables us to identify the change in concentration over time that is university- specific –a good measure of knowledge flow concentration

12. Estimation Criteria Estimate university specific changes in concentration of knowledge flows between Periods 1 and 2 Control for general changes in concentration of knowledge flows between Periods 1 and 2 Since it is plausible that universities patent different kinds of inventions than firms, control for potential systematic differences (field, importance, generality, originality)

13. Estimation

14. Dependent Variable

15. Control Variables Importance –Citation count Generality –Dispersion of citations received from patents in different technology fields Originality –Dispersion of citations made to patents in different technology fields University science –Share of citations made to or received from university patents Technology field –NBER 2-digit field classification (robust to USPTO 3-digit)

16. Data

17. Sample Construction Sources –NBER patent database –US Colleges and Universities – Utility Patent Grants, Calendar Years 1969-2000 Two distinct samples: outflows and inflows Outflows –Focal patents Utility patents issued to US non-gov: 1980-83 and 1986-89 Must receive at least two citations –Citations to focal patents Exclude self-citations Applied for after focal patent is issued Applied for within 10 years of focal patent issue date

18. Data Problems Unidentified owners –18.4% of patents in NBER database do not identify assignees –By construction, all focal patents have identified assignees –12.0% (13.3%) of citing (cited) patents are unassigned –When calculating our fragmentation measure, we assume unassigned patents are not self-citations and that each belongs to a different assignee Robustness to this assumption –Assuming that all unassigned patents belong to a single assignee does not change our conclusions –Performing our estimation on only those focal patents for which all citing (cited) patents are assigned does not change our main findings Transfer of ownership –This would pose a problem if the likelihood of transfer of ownership (specifically the type of transfer that would cause a change in fragmentation) increased (or decreased) at a different rate for universities than for firms –The literature on this topic does not indicate whether this is the case and we do not have access to the ownership transfer data to confirm; thus, we note this as a caveat for interpreting our results and an issue warranting further research

19. Results

25. Interpretation For simplicity, we have discussed changes in university knowledge flow concentration in relative terms (i.e., the change in the university premium rather than the absolute change in the concentration of university knowledge flows) However, although the relative change seems large (>50%), the absolute change seems small (<3%) Ultimately, we are interested in whether the change is economically important

26. Interpretation Example No. of Citations Assignee AAssignee BAssignee CAssignee DAssignee EAssignee FFrag Period 12211110.929 Period 22221100.893 delta Frag =0.036

27. Implication for Outflows Changes in TLO behavior? Objective function shift from dissemination maximization (leading to predominantly non-exclusive, widely licensed patents) to profit maximization (leading to predominantly narrowly licensed patents) Changes in inventor behavior? To the extent that university inventors become more commercially oriented regarding the management of their intellectual property, their tendency to share tacit knowledge with others who are not licensees may diminish

28. Implication for Inflows If inventors become more commercially oriented and business savvy over time, they may increasingly look forward and anticipate that, to the extent that future licensees are exposed to anti-commons problems associated with access to complementary technologies, the value of their inventions will be diminished As such, inventors reason back and plan their research program in a manner that minimizes anti-commons exposure by reducing the breadth of prior art citations

29. Welfare Implications?

30. Thank you

31. Research Question Knowledge Outflows –Was the rise in university patenting accompanied by a decrease in the breadth of knowledge diffusion with respect to that knowledge associated with patented inventions? If so, to what extent? –To what extent was the estimated decrease in breadth of diffusion the result of entry by inexperienced universities, similar to the decrease in importance and generality as measured by Mowery et al (2004)? Knowledge Inflows –Was the rise in university patenting accompanied by a decrease in the breadth of sources of knowledge inputs with respect to that knowledge associated with patented inventions? If so, to what extent? –To what extent was the estimated decrease in breadth of sources the result of entry by inexperienced universities, similar to the decrease in importance and generality as measured by Mowery et al (2004)?