Exclusive licensing creates its own complexities, just like the Peanuts character Pigpen brings his own dust storm with him wherever he goes. If inventions are “dedicated” to the public or are licensed non-discriminatorily on standard terms, there’s not so much complexity. The “complex problems of technology transfer” that Jones refers to have to do with exclusive licensing, with licensing by nonprofits to generate profits, and with the tax consequences of doing so. Those are complex problems–but they are problems created by the practices of “commercialization” using patents as monopolies. These “complex problems” of technology transfer are not the only problems of technology transfer–they just happen to be ones that university business officers have been willing to create to get their way.
As for the idea that “the licensing process draws the research university closer to industry”–this, too, is not so true for exclusive licensing practices. Wherever a university licenses a patent exclusively, everyone else in industry has a motive to undermine that invention–to design around it, to create blocking patents on improvements and applications, to build standards that exclude it, to downplay its importance in industry roadmaps. Do such things “draw the research university close to industry”? No, not really. Non-exclusive licensing might. But so might dedication of results to the public domain, allowing graduates from a lab to practice what they have learned directly in industry, without waiting for bureaucratic approval, which may never come (unless one happens to take a job with an eventual exclusive licensee). No, the licensing process tends to distance the research university from industry, even while forming a close working relationship (as parasite to host) with the company that obtains an exclusive license.
Jones then cites various statistics showing that there’s not much money in university patent licensing and that few inventions get licensed or develop into commercial products. His numbers are of this sort: 56 patents licensed of 236 (23%)–which happens to be about what the government’s own biotech licensing rate was. But licensing rates do not translate directly into commercialization rates, and commercial licensing (if exclusive) may indicate that everyone else is being denied access to the invention in favor of whatever efforts are being made by the exclusive licensee. That is, an exclusive license counter-indicates broad access of an invention and disables parallel development, whether collaborative or competitive. At MIT, according to Jones, 10% of patents licensed, and half of those do better than break even on costs, but only 1 in 1,000 makes $1m or more (now, perhaps the figure would be $5m). Despite the arguments for universities doing a better job than the government in licensing inventions for commercial development, it appears that universities never have done better than the government, even when the universities were working through nonprofit foundations and the government was defaulting to non-exclusive licensing. This is even more true now, when universities are spending substantial amounts–multiple millions of dollars a year–on patenting and their commercialization rates appear to be no better than those Dr. Jones testified about in 1978, when university patent brokers pinned their hopes on an expanded IPA program.
After Dr. Jones finished his statement, Dr. Jones, MIT’s legal counsel Arthur A. Smith, and Clark McCartney from the University of Southern California received questions from the Senate committee. One line of questioning had to do with the amount of royalties from patent licensing and how universities used this money. Here’s Clark McCartney’s response:
That is, for tax reasons, money from patent licensing goes to education and research–so, to exempt purposes of the university. But restricting the use of licensing income over costs from public service does not make much sense. Surely public service is also an exempt purpose of a university. What’s fascinating is that best I can tell, money from university patent licensing almost never goes to education–it goes to research budgets and to administrative accounts that can only generally be called “education” in the sense that they are accounts maintained by an organization that is known for providing “education.” The money actually doesn’t get used for education. For all that, I’ve never seen anyone account for how royalty money is used, outside of supporting “research.”
One might go further and argue that as universities established their own patent licensing operations and made the work of these offices routine, comparable to any trade activity in the for-profit world of patent management, allocating profits to further research would appear to complete the cycle of producing inventions from research funding, licensing patents on those inventions, and using the money to fund more research for which the university has a claim on any further inventions. One might think that things have slipped more than a little–but federal tax auditors aren’t able or willing to recognize the change in practice.
What do we learn from all of this? First, that the restrictions in Bayh-Dole on nonprofit use of licensing income appear to arise from worries that university patenting will run afoul of federal income tax laws pertaining to nonprofit exemptions and unrelated business income. Thus, Bayh-Dole requires federal agencies to require nonprofits to pay inventors a share of royalty income, to make it clear that such payments are acceptable (even though requiring payments by federal contract does not waive tax law); and requires any balance after expenses to be allocated to nonprofit “related” exempt purposes–scientific research and education. But not public service, which in some strange way is not sufficiently “related.” Thus, Bayh-Dole ensures that universities cannot use income from patent licensing to contribute directly to public benefits–despite the requirement in federal grant regulations (also part of the federal funding agreement) that intangible assets acquired or improved with federal funds should be used for the beneficiaries of the sponsored project, which would appear to be a clear requirement for public service uses of licensing income.
Second, we learn that the concern for broad public benefit from university research inventions has shifted from a practice of making inventions broadly available to one of stating an intention to make inventions broadly available while restricting most licensing to exclusive licenses–and necessarily restricting access to those licensed inventions to everyone else, pending the development of a commercial product (which rarely happens). Thus, while there were tax concerns early on with regard to Bayh-Dole, these appear to have been unfounded and it does not matter whether a university’s patent practice licenses most inventions non-exclusively and perhaps royalty free or whether the university merely states an intention to make inventions available for public use (and meaning by that restricting licensing to exclusive licenses for commercial product development, and rejecting all requests for licenses just to make and use an invention without any obligation to develop commercial products).
Third, we gain some insight into the expectations that those working with the proposed expanded IPA program and then with Bayh-Dole had with regard to inventors and assignment. Inventors did indeed own their inventions, and the transaction by which they assigned inventions to their universities for management involved actual consideration–so that one might go so far as to think that such assignments are voluntary, the subject of negotiation, and certainly not simply compelled as a matter of law or federal contract or university compliance with a federal contract. While the IPA program master agreements did have a clause requiring nonprofits to have a patent agreement with each potential inventor under which the inventor agreed to assign inventions to the nonprofit, Bayh-Dole has no such requirement–not in the law and not in the standard patent rights clause. Some people think this is a gap or defect. I think it is a feature that has been obfuscated by university patent administrators who don’t like the idea that they should have to make a case for assignment rather than to demand assignment with impunity.
But there’s a bit more in all this. Even if the IPA program required universities to require assignment, it did not stipulate that this assignment was consideration-free. The assignment might be required, but only in the context of a negotiated agreement. As far as Sugarman and Mencino are concerned, the basis for paying inventors a share of royalty income is that the payments are consideration for the acquisition of the inventors’ property. Once university administrators moved to break the agent model and replace it with a corporate portfolio model, they also broke the rationale for how they navigate their university’s exempt tax status and allow employees to have a private interest in the business of the nonprofit. Under any number of university patent policies, the royalty sharing is a perk, not compensation. Perhaps tax laws have changed, and such perks are permitted. Otherwise, a requirement to assign is not sufficient. There still must be a negotiation regarding consideration to make the assignment both effective and subsequent royalty sharing conform to federal tax law for nonprofits.
Finally, we have a better view of the idea of “commercialization” that has haunted federal invention policy since nearly the beginning–and at least since 1947 when the federal government started subvention funding–grants-in-aid for research without a procurement rationale. A strong effort has been mobilized against the idea that the results of federal subvention research should be made available to all–dedicated to the public domain or licensed non-exclusively, royalty free; or if licensed for a royalty, then a reasonable royalty and not a monopoly royalty. In this opposition to making research results available, the idea of “commercialization” been a primary tool. Research has been depicted as the necessary start of a linear sequence of production by which inventions are made, patents are obtained, exclusive licenses issued, and commercial products developed. It is of course true that this linear sequence does happen. What is not true is that it is not the general case and not even the typical case. The typical case is this: research is conducted, inventions are made, patents are issued, and the public is prevented from the use of the inventions pending an effort to find a wealthy speculative investor willing to share the upside of a patent monopoly. This version of “commercialization” dominates all discussions of Bayh-Dole, of the role of universities as patent brokers, of criticism of government policies that favor making federally supported research at nonprofits–almost all of it subvention funding–available to all.
One cannot have patent monopolies compete with public access or even with consortia and standards. Thus, the idea of the sources of innovation from research get caught up in counts of patents, licenses, startups (meaning exclusive licenses), and royalty income–without ever accounting for any actual commercial products available to the public on reasonable (not monopoly) terms. The linear model of research to invention to institutional ownership to patent to exclusive license to monopoly commercial product then gets wrapped into an extended model that places the justification for public funding of research on the front end and economic development on the back end. The claim, then, is not just that new knowledge drives an economy but rather institutionally owned and brokered patent monopolies do. It’s just that other than these repeated, apparently seductive arguments, there’s no evidence that such practice exists in the wild or that the results of that practice have had anywhere near the effect claimed for it. The principal effect that can be documented is that “commercialization” has created a tremendous livelihood for university patent brokers, who, like Pigpen, bring licensing complexity to their work in the form of a dedication to “commercialization.”