We are looking at metrics for managing university-based technology transfer and policy. First, we have none. We have disconnected proxies that don’t inform either management or policy. Let’s look at what university administrators care about: money. That’s it. I have met only a handful of university administrators that care about anything else. For a few, publicity. For a few others, industry relations. But as far as I can tell, the vast majority of university administrators care about money more than anything else. Public discussions will be about innovation, public benefit, the entrepreneurial university, regional economic development, life-saving medicines, and startups. But on the back end, this is all so much idealistic rhetoric to dissuade public scrutiny of the money. So, track the money, report the money. No one does that now, except university administrators, and they aren’t talking.
1. What university administrators want most: money
- Spent
- Made
- When
- Use
The first set of metrics is what university administrators care about, and only loosely connected to technology transfer. Money. Money. Money. So, track money and report money. More money doesn’t mean more or better technology transfer, and less money might indicate the very best technology transfer, but if university administrators believe that the secret public mission of their organizations is to make money, and that’s what they insist on, then track this stuff with regard to technology transfer.
If a university administrator makes funding for technology transfer contingent on income, then track and report the entire money situation. If a university makes a technology transfer budget contingent on a share of income from licensing, then track and report the entire money situation. There are badnesses for technology transfer in making a licensing operation get a set percentage of licensing income, even as one might consider doing that for a patent licensing operation running as a stand-alone organization (or business).
There was a time, say 1960, when the idea of university technology transfer did focus on making money to supplement university research budgets. In 1960, the University of California started an in-house patent licensing program (entirely voluntary participation by faculty inventors) on the premise of making money for further research.
Even Research Corporation, started by a University of California faculty member in 1912, was premised on the idea that patent licensing would generate income that then could be used to support research–it’s just that the Research Corporation model was that industry would guide the licensing to industry (Research Corporation’s board was a who’s who of industry representatives–industry was willing to, as it were, tax itself to gain expedited access to university faculty research discoveries) and income after expenses would be passed over to the Smithsonian Institution to be spent on further research anywhere in the United States where something looked worth studying.
The Wisconsin Alumni Research Foundation started in 1925 with a similar premise–money for the university–but with a differences: the money was for the University of Wisconsin only, and patent licensing income would be used to ramp up the Foundation’s investment in stocks, and the profits from that investing would be made into an annual gift to the university. Unlike Research Corporation, WARF focused on a single university–so more research might lead to more inventions might lead to more patents and licenses, more income, and–you see the reasoning here. The more research money meant presumptively more funding for the program making the licensing money for more research. In the bureaucratic logic of the WARF approach, patenting and licensing exist at the minimum to fund the effort to patent and license. The prospect of anything extra keeps administrators and faculty ardently supporting the effort.
By the time we reach Bayh-Dole, many universities had cut deals with Research Corporation so they got a share–WARF-like– of income made from what they thought to be “their” inventions–not the inventions made by faculty and students hosted by their campuses. Administrators looked for ways to compel faculty to use their patenting and licensing arrangements. The NIH and NSF IPA programs were designed to do just that–under an IPA program, a university administration agreed with a federal agency that the university must take ownership of every invention made in work with agency support that the university administration decided to patent. In effect, a university administration could, without changing its patent policy, simply contract with federal agencies to make it appear that a federal contract required university ownership of inventions. They then set about trying to make the IPA programs government-wide (hence the need for a “uniform” federal policy) and built into regulations (so federal law would require universities to own inventions, not merely a federal contract that university administrators generally had no policy authority to agree to (but did anyway).
Bayh-Dole, as an implementation of the IPA programs according to its primary drafter, Norman Latker, patent counsel for the DHEW, offered the prospect of making the vesting of inventions with universities a matter of federal statute. Bayh-Dole did not do this, but university administrators wanted Bayh-Dole to do this, and so they claimed that Bayh-Dole did something it didn’t do–vesting ownership of inventions made in federally supported work with the universities. It took a Supreme Court decision in 2011 to reject the university “interpretation” of the law–30 years after Bayh-Dole became effective.
In that time, university administrators used the claim of vesting to change university patent policies “to conform with federal law” (which was not true, but it sounded good and was what they wanted), so when the Supreme Court rejected the vesting claim, it didn’t much matter because the universities had also implemented the vesting claim as a matter of their own policies–not that their policies were well drafted or valid. All that mattered is that most university inventors did not have the money and time to fight university administrators with seemingly endless budgets for lawyers and don’t care to risk their employment or research or even academic freedom over inventions. University administrators used employment policies, threats of ethics violations, as well as general pounding of policy positions to force assignment of inventions.
In many ways, the last forty years have been a war by university administrators on informal methods of technology transfer. Let me explain. Prior to 1968, university technology transfer involved three layers of selectivity. First, investigators decided what findings were worth identifying as inventive suitable for patenting. Stuff that wasn’t suitable, investigators published or were indifferent to it–stuff anyone could do, stuff that was so far afield that no one could do anything with it, and the like. Maybe there was only one suitable thing in 100 findings. For federally funded work–mostly NIH and NSF funding–stuff was mostly published or let go because that was what these federal agencies would do if they took ownership of inventions.
If something in all of this was thought suitable for patenting for some reason (to protect the public, to make money for research, to go off an be entrepreneurial, to engage private investment capitalists who otherwise would not be bothered and no one else would be bothered either so it was them or no one), then investigators reported such inventions to university patent committees. These committees, then, reviewed the invention and decided whether it was worth spending university resources on it (such as, to fund further development work) and whether doing so was in the public interest (delicate thoughts, here). If a committee thought a given invention was suitable, then it referred the invention for management with a contracted patent development firm–often Research Corporation but also often via an affiliated research foundation that took ownership of the invention and then assigned (or exclusively licensed) the invention to Research Corporation. A committee might accept half of the inventions submitted for possible university involvement.
By demanding disclosure and assignment of all inventions made by university personnel–even inventions outside any duties of employment– administrators destroyed the first level of selectivity that had been in operation–inventors no longer could choose which inventions to report that appeared to benefit from university patenting. In place of this selectivity the university demanded the disclosure of all inventions, and just to be safe, all non-inventions that might look like inventions or might be used to make money via licensing or non-licensing made to look like licensing.
When people start pointing to a growth in university patenting after Bayh-Dole, they are not pointing out increased inventiveness, though they may want everyone to believe just that. They are pointing out an administrative move to eliminate inventor selectivity with regard to what is suitable for patenting and replacing it with administrator non-selectivity. Administrators also moved to bring patenting in-house, and in so doing removed a second level of selectivity, that of the external patent development firms such as Research Corporation. Research Corporation, for instance, accepted only 10 to 15% of the inventions submitted to it for management. In the new administrator-run invention regime, instead of a selectivity of maybe 1 invention in 200 being brought into an institutional patent management scheme, university administrators were now taking in 150, and trying to patent 75 of those.
There is your increase in patenting after Bayh-Dole. People were not getting more inventive (though patent law opened up genes, business methods, software, and all sorts of “life” for patenting, too)–rather, university administrators got way less selective about what sorts of inventions matched university research and public service. They sought out mediocre inventions–stuff that might be patentable but really had no profile for needing the “incentives” of patent monopoly positions for use or commercialization or mass production. They took patent positions “just in case” they could make money, claiming that doing so was in the public interest, that research funding was wasted if they didn’t, and that anything not patented would just “sit on the shelf” and not get used.
The ramp up in university patenting results from the administrative destruction of selectivity to match patenting opportunities with benefits for industry adoption. In the place of selectivity, we have instead universities acting as patent trolls, forcing industry to design around or undermine or infringe just to get things done. Federal research work spread around to multiple investigators at multiple institutions then gets fragmented into tiny bits of institutional patent claims, never to be reassembled into a coherent research or development platform or standard during the term of the patents. Non-exclusive access would reassemble such fragmented bits, but once university administrators have made spent $15,000 or more on a patent application, they are not able to back down and release the invention non-exclusively, royalty-free.
The gross number of inventions patented by universities is more closely a measure of the non-selectivity of administrators with regard to the use of the patent system. They may make money from licensing–a winner every two decades pays for a lot of losers mindset–but they are doing much worse at technology transfer for all this patenting.
But it’s money that university administrators most care about. For money, the metrics that matter are how much are we spending, what are we making for that spend, and when will we get it? There’s nothing here that’s meaningful about what happens with what gets “transferred.” Sell, license, troll, speculate–it really does not matter how the money gets made, so long as it gets made in a transaction that looks like an IP license and there’s no negative publicity.
It’s pretty direct, in principle. There’s no reason to sugar-coat this idea in abstractions like “public return on investment.” It’s about money, free and clear, to do things that aren’t brought to the public for comment or approval or authorization. Call it administrative freedom. Call it going rogue. Call it greed or corruption of purpose. Call it prudent financial discipline. Call it the entrepreneurial university. Whatever. But out with it. How much have you made, what has it cost you to make it, what’s your exposure, what’s your opportunity?
While we are at it, then, let’s track the use that the technology transfer income is put to. What is it that is so important that a university persuades itself it must own, patent, and charge for access to discoveries, inventions, data, software, compositions, collections, and know-how? Or is making money–more money than others, say–the most important thing of all (for, you know, reputation)?
Track the money, then. Rough cut is easy. Getting careful about it is not so easy. University finance is a strange mess of cross-subsidies and misdirections. Who would have thought that the University of Washington’s budget for instruction includes all the football “scholarships”? One can look at the budget for the university’s technology licensing office–salaries and benefits, legal expenditures, professional development, furniture and equipment, and that gets at some of it. But then there’s lease space. Once a unit is making money, or claims it will, then administrators like to move that unit off campus into leased space, to help cover the cost of the lease. This cost is then booked against the technology transfer program, even though it’s entirely an expense of the university’s own making. For instance, at the University of Washington, my six-person team worked out of space in Fluke Hall–at no cost–but the university administration demanded that we move into leased space off campus, where we did not want to be, and we had to pay to build out the space to make it even usable–more expense charged to our budget–compulsory overhead to help the university pay for its real estate business dealings.
Then there are legal costs. Some of these are usually billed to outside counsel and so are relatively easy to track. But a university typically also has in-house counsel, and that counsel may spend time on technology licensing matters–drafting contracts, reviewing contracts, negotiating, supervising patent work (if not also doing patent work), dealing with irate university inventors, and the like. How much of that legal time gets accounted for? Unless an attorney is on staff at the technology transfer office, maybe little of the cost–time, overhead, travel–gets booked directly against the technology transfer budget.
Same thing goes for risk management, for public relations, for the office of development, information technology, and other university units that may get involved in technology transfer activities but don’t necessarily bill out all their services. Given all this, it is actually more difficult that it looks to provide a full accounting of technology transfer costs to a university–and we are looking here only at the money, not the opportunity costs.
On the income side, you would think the money would be easier. Companies taking licenses send checks, and those checks get booked by the technology transfer office. Done. But no, it is not that easy. Consider–a company sponsors research, say $100K, and insists on a royalty-free license to any results. A totally reasonable position, on the face of it. But now the poor technology transfer office has to track inventions and other work product from the research, get disclosures, deal with IP, notify the company, issue licenses–and all for nothing. No income at all.
But there’s a bigger issue yet. To track money, we often think about cash accounting. At the end of the year, look at what you spent and what you got and decide what you are going to spend on in the next year. That’s fine for a lot of things, but not for IP management based on licenses, especially exclusive commercialization patent licenses. For such things, one has to use an accrual method of accounting–stretched over twenty to thirty years: an invention made now, patented in four years, licensed in seven, with a twenty year run, with a family of related inventions with their own patent terms ending even later. One might not see payments until a product is sold or a company is purchased–ten, fifteen, twenty years down the road. What one pays out in years one to four for patenting work and “marketing” work won’t be “recovered” until there’s a big hit deal somewhere in the IP portfolio. At the elite programs, such as Stanford’s, those big hits come once a decade.
If a program is not elite, then maybe once every thirty years. The University of Washington–a leader in receipt of federal funding–got one big hit in the early 1980s and has struggled to do anything else, despite rolling in money diverted to patent licensing. About five years ago, UW went all in on a plan to transform its financial operations by starting companies and selling these off to venture investors to make millions. They spent down the licensing office’s reserve funds, diverted the last of their big hit patent revenues to the effort as well, and managed to cut the number of startups in half while doubling the budget for doing it. When it was clear that nothing was going to come of it, the folks in charge simply left–no accountability, everything swept under the proverbial rug.
Thus, the big challenge in accrual accounting for IP management is recognizing that in exclusive patent licensing, especially for speculative investment, one can spend a lot of money for a long time, and then–maybe–a decade later there’s a lucrative deal that pays for all that spending for all the inventions under management. What’s left out–and many administrators in IP licensing will not admit it–is that even when there’s a lucrative deal, that deal does not come about as a result of all the prior spending. It is not that by building a huge, non-selective, even mediocre portfolio of patents that statistically one increases the likelihood of gaining a lucrative deal. Lucrative deals may happen more by luck than by systematic bureaucratic compulsory procedures. It may well be that systematic bureaucratic compulsory procedures actually increase the “noise” of inventions, treating mediocre and significant inventions using the same procedures, distracting focus, spreading resources thin, reducing the chances to get lucky. Even when there then is a lucrative deal, one may not readily see that all that prior spending set up the deal–it could be just the opposite, that folks were even more lucky to get the deal at all because of all that spending, that bureaucratic system in place, the demand that all licenses follow a template designed to accommodate volume not meet distinctive opportunity. In other words, accrual accounting cannot without special care distinguish parasitic activity from productive activity.
If a university gets lucky with a lucrative deal, you would not be able to tell for the next twenty years whether the licensing office was being effective in its spending or not. Certainly not from AUTM licensing survey information. All those inventions in a given year, all those new patents, and all those new licenses would look like they were somehow related to that income. And yet there is almost no possible relationship other than that more income in a given year sets up more spending in the next year. It’s just that there’s nothing to indicate whether the spending next year produces anything of value–say, ten years or fifteen years down the line. For that, one would have to see office activity as case-based accounting–out of pocket, time reporting, plus an overhead for all the databasing, reporting, filing, copying, scheduling, and dithering. At best one might find that “winners pay for losers” or more like “winners, if there ever are any, pay for losers; otherwise, institutional suckers (students, administrators (not their own money, though), state economic development programs, entrepreneurship programs) pay for losers.”
Even after all this, the accounting is still directed at formal programs of technology transfer–taking ownership of inventions, filing patent applications, and then having to find ways to release control (preferably, according to administrators, for money, lots of it). But there’s all sorts of technology transfer going on that does not engage with formal programs–students learn and graduate, papers get published, presentations get made, labs get visited, people consult, people tell other people their ideas, their hunches, software gets released open source, people post stuff on their web sites.
To cut through the complications, it is a good idea to isolate technology transfer operations in a separate unit. For that, making that unit entirely independent of the university is a good idea. Then its accounting is just that. Even then, independent units can make work for the rest of the university–causing blunders or wasting people’s time or asking for in-kind help (review this technology available for licensing document, please). Still, external helps to isolate the accounting.
Of course, if accounting is not the big deal–money isn’t the most important thing, and even break-even is not a virtue (think: technology transfer as a form of library lending), then going external does not matter so much and may even get in the way of technology transfer. Since having a care about making money–and big money–from patent licensing and any other sort of licensing is not about technology transfer, then running licensing operations from an external, even independent unit, is a smart, if not necessary, move. Putting technology transfer into an arm’s length independent unit does have this additional advantage for public universities–such a unit can keep its finances and business information well away from state disclosure laws. Ah, but you see the dilemma, then, for public accounting for the money. To do the accounting right, one has to eliminate the cross-subsidies of university accounting, and for that, one ought to push patent licensing to an external unit. But an external unit will be motivated to keep everything secret, and so thwart the accounting. Pffst.
Tracking the money, then, is full of areas for gaming the accounting. Perhaps there’s nothing for it. The more technology transfer moves to informal, to non-exclusive licensing, to agent-based rather than university-based activity, the more a cash accounting approach works. Technologies pay for their administration and their administrative costs are kept low and so there’s not much that has to be carried on the books for a decade. As technology transfer moves to patenting early and often, using in-house operations, holding out for lucrative exclusive deals, the accounting has to be carried for decades. For accrual accounting, track patent families for expenditures and income over these long time periods–thirty years from start to finish. The overall money picture will then be a composite of the patent family picture. The big events will be the decision to spend anything on patenting, patenting cost recovery from a license (in the rare chance there ever is a license) and income from downstream events from a license (milestone payments, realized equity, earned royalty on sales–all even much rarer than a mere license). Thus, patent families may be distinguished as (i) patenting (ii) legal cost recovered (but not necessarily administrative cost recovered) and (iii) downstream earning. Track and report total costs and income for each patent family and in each class of patent family. That gives you the financial performance of the portfolio with enough structure to see what is going on for management purposes.
There’s plenty more to discuss–non-patented stuff, federal and non-federal, non-exclusive work, inventions tied up by sponsored research agreements. But patents are the anchor claim of university IP operations, and exclusive licensing has been the anchor justification for patents, and money is the goal. So this is how money ought to be tracked in such an operation. Once you have got a good view of money, then look at how the money is used–the use had better be worth the effort to get it. And look, too, at where the money spent has come from–IP licensing operations? donations? state allocations? tuition? reserve funds? Just what. With that, you will have a clear picture of what is going on and how the money part connects an in-house IP operation with other university programs and with the public.