AAU, APLU, and others aim to “bolster” federal technology transfer, 6

We are working through the advice that AAU, APLU, and other “higher education associations” have provided to NIST on how to improve federal laboratory technology transfer. Do you expect that these associations thoroughly vetted their draft response with all their members? Is it even a good assumption that the senior leadership of all these universities are behind this effort to help NIST? The HEAs’ position is that the way to improve federal laboratory technology transfer efforts is to ponder deeply the complaints that the HEAs have with regard to their own members’ technology transfer efforts–which of course are working out just as Bayh-Dole intends, except for the lack of a clear mission, an uncertain relationship to the public interest, problems with the patent system, the tax code, administrative overhead, confusion about software, and, oh, this:

No funding was provided for patent costs or other costs associated with the operation of university technology transfer offices established to implement Bayh-Dole…

Gracious me. Someone doesn’t understand Bayh-Dole. Bayh-Dole’s requirements are easy to comply with and easily may be accommodated with a tiny part of the administrative indirect cost payment that comes with each federal grant. Bayh-Dole implementation does not require the establishment of technology transfer offices and does not require universities to incur patent costs. Whatever university administrators decide to spend because of Bayh-Dole is because of their own choices–if not their cluelessness about Bayh-Dole and their cluelessness about how research technology comes to be used. 

Under the standard patent rights clause, a university must

(1) designate personnel responsible for patent matters.

That’s a one-time thing. A research foundation, an administrator, a patent attorney, Research Corporation. Easy.

(2) educate employees on the importance of timely disclosure of subject inventions

Subject inventions are ones that the university has an equitable ownership interest in, other than as a consequence of allowing employees to work on a federal award or use university resources that the university has obligated to that award. Again, education is easy. It can be a notice in an employment agreement or faculty appointment. It can be a notice at the time an employee signs off on agreement to draw pay from an account with federal funds. It can be a web site with the bold words:

“It’s really important that if you make an invention in which the university has equitable title and the invention was made in performance of work under a federal grant, that you promptly disclose the invention to [designated personnel for the administration of patent matters],

“sufficiently complete in technical detail to convey a clear understanding to the extent known at the time of the disclosure, of the nature, purpose, operation, and the physical, chemical, biological or electrical characteristics of the invention. The disclosure shall also identify any publication, on sale or public use of the invention and whether a manuscript describing the invention has been submitted for publication and, if so, whether it has been accepted for publication at the time of disclosure.”


(3) require each employee other than clerical and nontechnical employees, to make a written agreement to protect the government’s interest in subject inventions.

In the written agreement, potential inventors agree to disclose subject inventions–ones that the university has an equitable ownership interest in, sign papers to allow patent applications to be filed, sign papers to establish the government’s rights in these inventions, and assign these subject inventions to the university. Again, this is a simple agreement. It can be a form with wording like the above (or see the link at Research Enterprise), and a check box on the same paperwork that adds an employee to the federally funded project. Again, we are not talking a lot of work, once the form is set up. Just another checkbox in the wall.

(4) report, if requested by the agency, each invention made under the grant at the closeout of the grant.

That’s more work, but it can be delegated to the principal investigator, so it’s no big thing.

A university does not have to take ownership of any invention made in work receiving federal support. If a university does not take ownership, then it has no obligation to disclose subject inventions to the federal government because it has no such subject inventions to report. (There’s a provisio on this–but I don’t know a single university administrator who will accept the argument that the written agreement requirement turns inventors into small business subcontractors and their inventions become subject inventions to be reported through their universities, but without university ownership or any special right to acquire title). If a university doesn’t have to take ownership, then it also doesn’t have to use the patent system, and that’s the end of its financial obligations.

Everything else that happens, whether expensive or not, comes about as the result of university administrators deciding to spend money on inventions all on their own. The federal government doesn’t make them do it. They do it to themselves. If they put themselves on an expensive, complicated, ineffective course of invention management, that’s no reason to fuss that the federal government doesn’t step in and out of pity pay the administrators’ costs–the patenting, the databases, the marketing staff, the glossy brochure, the web site, the tech available for licensing list, the expos and conferences, the contract drafting, the legal reviews, the responses to the bitter faculty emails with disingenuous bureaucrat-speak, the records retention compliance. Them there are administrative choices, not federal unfunded mandates. If university administrators don’t want the expense, then don’t set up to own any inventions made in projects with federal support, or if administrators are too bumbling to avoid such ownership, then at the very least disclose the inventions and elect not to retain title and be done with them. Heck, ignore the disclosure requirement and let the federal government take title by default–that pathway is built into Bayh-Dole, so use it. Of course, if administrators are too bumbling to figure out how not to own inventions, then there’s no hope that they can nevertheless figure out how to give up owning.

There is nothing in Bayh-Dole that obligates universities to take ownership of inventions, to use the patent system, to seek to commercialization inventions, to try to make money, to stimulate the “economy”–it’s just not there. All that is made-up fantasy. Yes, even the  fantastical mischaracterization of Bayh-Dole in “Academia, Industry, and the Bayh-Dole Act: An Implied Duty to Commercialize.” What’s more impressive than the fantasy, however, is that those that dream this golden dream really want it. They do not fantasize a dystopia and argue against it. They fantasize what isn’t there and then argue for how good their fantasy is as public policy. You know, just like the HEAs are doing in their advice to NIST which is really a list of complaints about how a law that is working as intended and doesn’t need to be changed isn’t working so well at all.

Bayh-Dole states a conditional–if you take ownership of a federally supported invention, then decide whether to keep that ownership or not. If you keep it, then achieve timely practical application. Otherwise, let the government try. There’s simply not that much cost involved. If university administrators can’t do these few things with the indirect costs they harvest from every federal award, the university really isn’t qualified to receive federal funding.

To get at this “we don’t get money from the federal government so we can overspend on patenting” problem raised by the HEAs, let’s consider university technology transfer practices prior to Bayh-Dole.

First prior practice: refer inventors to an invention management organization such as Research Corporation, University Patents, Battelle, or an affiliated university research foundation. The invention management organization then determines what will benefit from a patent, pay the costs, and do the work to find licensees. The university is not out any money, doesn’t have to have a technology transfer office, and if a patent is licensed, the university can get a share of the upside by having an agreement with either the inventor or the invention management organization.

In this prior practice, universities do not take an ownership position on inventions, they do not make decisions about what to patent, they do not try to find a favorite company to license to exclusively, they do not threaten anyone with infringement, they do not have to involve their own legal advisors, who often have no understanding–I speak from experience here–of intellectual property matters, they do not create institutional conflicts of interest for themselves, and they do not have to keep on hand a staff of licensing professionals to try to navigate policies, contracts, research administration, departmental politics, and all the rest.

This prior practice had other advantages. A national invention management organization such as Research Corporation could work with the best inventions submitted to it, could build working relationships with representatives in industries wherever those industries were clustered, specialize in specific areas of inventive work, and be selective in what new work it would take on. Having multiple such invention management organizations available meant that a university inventor had choices based on organizations’ focus, management, terms, and track record. If a national invention management organization had just done a big patent deal with five chemical companies, then perhaps it would be positioned to do similar deals with these companies. If, by contrast, inventions fall to the staff at a university that hosted the research, almost every invention is a matter of first impression, and one unlicensed invention is followed by force of policy by five more unlicensed inventions. A local licensing office finds it hard to say no to management, and often front-loads the work, bringing more inventions under management than it can possibly handle and then–ahem–begging for more money since it is doing so much work.

The NRC report that the HEAs referred to earlier in their NIST advice estimated that a university had to have 200 active licenses to have a statistically likely chance to have one of those licenses generate $1m in royalties–that’s cumulative, or like $50,000 a year, which is hardly a sign that an invention has made it commercial sales. For that matter, a single patent with its foreign counterparts can run upwards of $200,000, so one fifth of that $1m could be simply reimbursements for patenting costs. A university with a small research footprint will not see enough inventions to be in a position to expect a lucrative patent licensing deal. The university will not be any better off therefore trying to capture all 63 inventions or however many may be made in a given year. By the fourth year, the inventions of the first year will be issuing as patents, but the path to licensing will have gone stale for most–the lab will have moved on, maybe even having invented new approaches that make past disclosed inventions obsolete.

A university with a research budget less than, say, $100m, won’t make it by trying to capture and review everything. It will have to use a different approach and be very selective. In a sense, until it has a lucrative patent license, it has no business opening up its own patent licensing operation. Indeed, historically, some of the major university-associated licensing operations were started with a valuable invention already in hand–Research Corporation and WARF, in particular. People who forget this part of the history of university technology transfer are unlikely to be aware enough to repeat it, however. Instead, they go off and do bureaucratic things like trying to be like Stanford or MIT, which is a very, very bad move if one isn’t already like Stanford or MIT. The alternative is to use various selection filters so that when something really important does come along, the inventors can be directed to a resource with experience that can help them. If things work out, the university will have a royalty stream that they can blow on crazy administrative schemes to make their good luck appear to be the routine output of an innovation process.

There was a second prior practice that developed at universities, in addition to the one that involved using an external invention management firm. In this practice, when a university office receives an invention disclosure, it contacts companies that might be interested in the work and asks whether the university should file a patent application, and if so, would the companies have an interest in using or developing the invention. If companies say, yes, please file and we will take a license on reasonable terms, then the university goes ahead and files a patent application and works with the companies to arrange for their access to the invention–having the university’s inventors, say, start working with their company counterparts. The companies pay the patenting costs, take a low-cost license perhaps with some sort of happiness payment if things go roaringly well, and technology gets transferred.

In this prior practice, a university does not spend any money on patenting until one or more companies indicates that patenting is something they would like to see happen. The primary university activity in this practice, then, is to talk with companies about what they would like, and provide opportunities responsive to the companies’ feedback. In short, built relationships with company contacts based on trust and good communications. Let companies guide the university’s choice of what to patent and how to license rights. Again, in this approach, a university has a positive commitment from industry to using an invention before either licensing or patenting. The sequence is, (1) listen to companies that ought to want to use the invention; (2) if companies want to use the invention, create a working relationship in which they can; (3) if the companies believe that a patent should be filed, then file the patent application on the condition the companies will reimburse the costs and will get a royalty-free or very low cost license to the invention under the patent.

That is, in brief:

  • find companies
  • make the invention available
  • license the invention
  • file a patent
  • confirm the patent license

Do you follow this approach? It is almost entirely the reverse of the approach used by the HEAs’ members. They file patent applications, then try to find companies to take licenses, then try to negotiate licenses. They end up dealing with quasi-trolls, second-tier speculators, and very occasionally with a venture-backed biotech company or a major pharmaceutical company. But mostly they end up not dealing at all and over time spending an increasing amount of their time and resource servicing their back portfolio–dealing with patent applications, cleverly trying to delay national phase patent work, paying maintenance fees, updating nonconfidential disclosures, updating inventor contact information, updating database entries or updating the database itself, producing reports, trying to budget for future costs of doing all these things, training staff to deal with all these things, trying to fix what untrained staff have done previously that looked good to them but wasn’t.

In the second form of prior practice, universities spent primarily on making connections with industry representatives, and filed patent applications only after companies had committed to use or develop an invention, had got a license, and were then willing to reimburse the university for its patenting costs. The outcome if there’s not a license or even a patent is that the university has built more relationships with companies. That’s a good thing in itself. As to the costs, these are hardly anything, once again. One person with an engaging voice, and an aptitude for presenting and listening for opportunities to collaborate.

There arose yet a third prior practice in universities, following on the WARF approach. Once a university did secure a nicely paying patent license, then it did have money to work with. A university then could dip into that revenue stream and file patent applications without using an invention management agent or asking companies what they thought it was best for the university to do. While using the university’s own licensing money to fund patenting sounded like a great idea to patent administrators, in practice it has led to a rather different approach to industry. Rather than request industry advice as to whether to patent (as Stanford did with the Cohen-Boyer gene-splicing inventions), universities would send out non-confidential descriptions of inventions that they were already in the process of patenting. If one is a dull university administrator, it all sounds just the same–an invention, a contact in industry, an offer to license. But it’s totally different. In this third practice, university licensing offices, in sending invention reports to companies, put them on notice of a given invention, and that means that if the company is caught using the invention without a license, it is exposed to a claim of willful infringement that carries the prospect of treble damages and paying the university’s legal fees.

When I started up the campus licensing office at the University of California Santa Cruz, we worked for two years to get the central licensing office to send us the files for the inventions that it had been managing for the campus. When we finally obtained them, we found the usual letters going to companies with attached descriptions of inventions that the university licensing office was patenting or had patented. In reply would be letters from companies pleading with the licensing office to stop contacting them in this way. Then, a few months later, there would be another letter from the licensing office to those same companies, with new descriptions of inventions. How to piss away goodwill by failing to understand the situation.

In this third approach, however, universities have a source of revenue to ramp up their licensing offices. And given a patent has a term of 20 years from the date of filing, once there’s licensing revenue, it may well continue for the term of the patent. Thus, for a decade or more, a university may fund its licensing office out of pocket. In one way, folks can act like they have it made. In another, however, they will be tempted to go slack and spend more time and money filing patent applications than they should and not doing enough to improve their luck–the luck that they can, in a decade, find another invention that will also generate comparable royalties. Stanford, a top research university in the heart of the top venture capital region in the world, reported at one point 16 inventions of 6,400 over 36 years had produced cumulative revenue over $5m. That’s 1 in 400 inventions–at Stanford. For a university that has the good fortune of landing such a patent license but has a research base of even $200m a year, the odds are still almost hopelessly low that through an administrative process involving filing patent applications up front and then marketing the rights to industry a second such invention will be identified, one or more licensees found, and a deal secured. “You had good fortune and you used that good fortune to fund a bureaucratic process that reduced the likelihood of further good fortune.”

It takes tremendous administrative discipline not to burn up a ready source of licensing revenue by changing one’s operating fundamentals. Obtaining more patents does not translate to better odds of licensing–especially if the patents are obtained without extensive prior interactions with industries that would use the inventions.

Case in point. University of Washington, nearing the end of a lucrative patent license for a biotech invention that included a hepatitis C vaccine, burned through about $100m in six years with a written plan to make so much money in technology transfer that the university would change its financial model. It was a ludicrous scheme but university administrators bought into it. University startup activity went down by a factor of two, expenses more than doubled, licensing revenue was flat. The licensing office was reduced to faking its metrics and producing self-congratulatory press releases. They even won an international award based in large part on their fake metrics.

They became the envy of other universities’ administrators.

But the vice provost for “commercialization” was forced out at the peak of the program’s claimed successes, the office was reorganized and in a show of how committed to change the administration was, they renamed the office yet again. Renaming a university department might purge administrative regret but it does not magically produce confidence in the companies that were shafted by the university. The university has never retracted its fake data and its fake reports to the state legislature. At the University of Washington, there’s a lost $100m under the carpet. But other university administrators often don’t look under other institutions’ carpets.

Finally, there has been a fourth university approach to invention management, even worse than the third. In this fourth approach, university administrators without a source of invention licensing revenue find themselves envying universities that do have invention licensing revenue. They set out to imitate those universities. “Do what they do!” is the command. If a university judged to be successful has a technology licensing office, then create a technology licensing office. If they have a patent policy that demands university ownership of all inventions, then copy that patent policy. If they paid top dollar to get in an industry experienced licensing professional to run the office, then do the same thing. We will find the money.

This is a cargo-cult response to the opportunity to deal in patents on research inventions. By putting up a simulacrum of the technology transfer offices that report significant licensing income, these university administrators believe they have everything ready to find big hit patent deal after big hit patent deal. But they have things backwards in 3D. They emulate the slack approach of a university with royalty income but without operational discipline–and call this adopting best practices. Now calling something a best practice is a good way to sucker other administrators into funding it–why would a university that desires to be known for excellence adopt anything other than best practices? But how does one who lacks practice experience determine what is a best practice? Certainly not by superficial observation and bureaucratic envy of partial truths presented in glossy annual reports issued by other universities’ licensing offices.

Universities without patent licensing revenue cannot afford to adopt the slack practices of offices that do have, for a time, sufficient licensing revenues to abandon operation discipline. Instead, these universities need to work with external invention management organizations, or work with industry to decide what it is valuable to patent–valuable to industry. They don’t need more money. They need to improve their chances for good fortune. And for that, they need great relationships, selectivity, and patience. And if they are wise, even when they do get research inventions into use or development, they maintain their operational discipline to continue to improve the changes of good fortune rather than attempting to reduce technology transfer to an administrative procedure endorsed by formal policy and dedicated to increasing the volume of activity while improving efficiency. That’s an approach that fails. That’s the approach adopted by many of the HEAs’ member universities. That’s the approach that costs the HEAs’ members all the money–and apparently the HEAs believe that the approach would “work” if only the federal government sent them more money. The implication for NIST, it would appear, is that NIST should consider adopting this same approach for federal technology transfer. It’s a loser’s offer, and it remains to be seen what NIST is made of on this point. Given NIST’s cluelessness about Bayh-Dole, one wonders if NIST even has a chance at this point.

In short, the current lack of funding in this area poses a fundamental challenge to universities’ ability to transfer federally funded technologies.

Read: “Give us more money to waste to badly copy what other universities put out to look good because we don’t have the sense to recognize what we ought to do ourselves. We recommend that NIST follow our lead and copy the superficial patterns of universities adopting slack but complicated, expensive, and ineffective practices.   That will solve the federal technology transfer problem!”

This entry was posted in Bayh-Dole, Bozonet and tagged , , , . Bookmark the permalink.