Thursby Fuller Thursby (2007) - Us Faculty Patenting Inside And Outside The University

Jump to navigation Jump to search
Has bibtex key
Has article title Us Faculty Patenting Inside And Outside The University
Has author Thursby Fuller Thursby
Has year 2007
In journal
In volume
In number
Has pages
Has publisher
©, 2016


  • Thursby J., A. Fuller, & M. Thursby (2007), "US Faculty Patenting Inside and Outside the University", NBER working paper *13256. pdf


This paper examines the empirical anomaly that in a sample of 5811 patents on which US faculty are listed as inventors, 26% of the patents are assigned solely to firms rather than to the faculty member's university as is dictated by US university employment policies or the Bayh Dole Act. In this paper we estimate a series of probability models of assignment as a function of patent characteristics, university policy, and inventor fields in order to examine the extent to which outside assignment is nefarious or comes from legitimate activities, such as consulting. Patents assigned to firms (whether established or start-ups with inventor as principal) are less basic than those assigned to universities suggesting these patents result from faculty consulting. A higher inventor share increases the likelihood of university assignment as compared with assignment to a firm in which the inventor is a principal but it has no effect on consulting with established firms versus assignment to the university. Faculty in the physical sciences and engineering are more likely to assign their patents to established firms than those in biological sciences.

Summary of the paper

This summary is short because this paper is weak.

Of patents created by university researchers, only 62% were assigned to universities. This is low, since employment contracts require assignment to the university when university resources are used. In addition Bayh-Dole allowed universities to retain ownership on IP created with federal funds - historically universities have insisted on ownership. This is in comparison to Europe where "professor privilege" allowed faculty ownership of their IP - see Geuna and Nesta (2006).

Interviews suggest that the bulk of assignments outside of the university come from consulting. The paper constructs a model to try to examine whether patenting outside of the university is nefarious or not. Since Bayh-Dole there has been a dramatic growth in TTOs. TTO's surveyed believe that less than half of all inventions are disclosed to them. But TTO's also shelve inventions, and may then decide to turn over ownership to the faculty. Of course they might decide to do this anyway.

Faculty are more likely to disclose to their university if:

  • They get a higher share of the royalties
  • Non-disclosure may be accidental - the commercial potential may not be apparent to the innovator.
  • They are willing to get involved in the university licensing process (which is complicated and may be difficult to work with)

For start-up companies that are originally based on faculty research it is common for faculty :

  • to serve on the board
  • to spend time in the company
  • to legimately assign follow-on IP to the company

This leads to the following legitimate reasons for a patent not being assigned to the university:

  • The TTO turned it over to the faculty (perhaps after shelving it)
  • The research was conducted outside of the university (note that there is a limit on consulting days a faculty member can do)
    • Either with an industry firm
    • Or with a start-up (co-)founded by the faculty
  • The university may allow it (like to the University of Wisconsin used to)

Or the faculty may have circumvented university policy!

University research tends to be more basic that the applied research conducted by industry. A measure of the 'appliedness' of the research is used to determine whether the patent assignment is legitimate. This measures is based on the incrementalness of the patent - in practice the number of backward citations made is used for the appliedness and the NBER measure of patent originality is used for basicness. These measures are so far from measuring (in my opinion) the basicness or appliedness of the patent that they are meaningless, and even if they did numerous problems are still inherent and not addressed.

Other Notes

Patents that are unassigned are the property of their inventor.

Note that the author's didn't mention that firms (i.e. non-universities) would likely be uninterested in the assignment of basic research, so even if assignment were random (and accidental), self-selection would give the same results.