“Affiliates programs” are generally donation-based financial support programs for university departments and schools. In exchange for a membership fee, an affiliate program participant gets various premiums–invitations to research reviews, open houses, and job fairs; access to university labs and faculty investigators; visibility in program materials. Most affiliates programs are in engineering, as that is where they started and largely have stayed. Frederick Terman at Stanford appears to be the creator of affiliates programs–another of his amazing run of innovations in research enterprise. Here is a great account of the history of these programs at Stanford by Carolyn Tajnai.
The primary purpose from the university side is to get companies engaged with university work, for the good of both companies and university research and instruction. A secondary purpose, always lurking, is to get more money from industry. From the perspective of companies, interest in affiliates programs ranges. One of the most insightful comments I heard regarding them, from a big company’s university liaison, was that affiliates programs were most interesting when a company is hiring, and much less so otherwise. One might think, then, that the opportunities for affiliates programs is best in industry sectors that are showing growth in general, and with a continuing need for new talent. In this setting, affiliates programs augment a fundamental method of technology transfer–getting students placed in jobs that take advantage of their specialized training.
At one point I compiled a list of such premiums in affiliates programs–everything from free parking to library privileges. Most such premiums are regarded as having no particular value–invitations and listings, for example. Some do have value, and can be treated as quid pro quo donations–just subtract the value of the premiums from the membership fee and everything is good.
The one thing you won’t find in affiliates programs, however, is a promise of access to intellectual property. One might promise an early look at new IP, or expedited access, or help navigating university red tape, but generally you won’t get a promise of IP. That is in part because affiliates programs are generally run by departments and schools with their development offices, not by a technology licensing office. I have often been surprised by the competition, and even quiet animosity between IP offices and development offices. When I have sneaked into development officers’ talks out in the field to industry, I’ve been surprised at what I’ve heard–“We will help you work around our university’s red tape and arcane policies.” Meaning, “we are your best buddies and we don’t like the licensing office any more than you do.” Perhaps that’s just rhetoric to win the favor of company officials, but overall my impression is that the IP script in development offices is remarkably different from the IP script in licensing offices.
However, there are ways to bring IP into affiliates programs without the competing scripts. There is actually a three way competition for industry attention. First, there is the development office. It works with donations, cultivates rich donors, and may even write policy that prohibits faculty from approaching rich donors. (There is a list of such folks at the University of Washington, for instance–don’t bother talking with Microsoft or Boeing–the development office has already told these companies to ignore anything that hasn’t been endorsed by development and a dean or higher). Donations via development may go to a development foundation, but for support for research and instruction, these typically go to department and school accounts. The attraction to faculty is that donations carry low overhead fees (typically under 6%) and few restrictions on their use. For companies, donations are often easier to transact than other ways of getting money into a favored department or lab, but may be more difficult to get internal authorization to donate. For a development office, that 6% administrative fee helps to fund development operations.
Then there is the sponsored projects office. Sponsored projects offices work with organized research–research with a “principal investigator,” an express detailed budget, statement of work, deliverables, and one or more external sponsors. Organized research is mediated by some form of agreement–whether a grant, contract, collaborative agreement, consortium agreement, and the like. Sponsored projects offices, like development offices, depict their “success” in terms of ever-increasing income from awards. Stanford faculty at one point a few years ago broached the idea that perhaps more and more research funding wasn’t the goal, but rather perhaps focusing on quality research–addressing the greatest challenges, for public interest–might be more important. But that’s not where things end up. No, “rankings” of universities include amount of research, not so much quality of research. For sponsored projects offices, the income also matters because the indirect cost charges–the “Facilities and Administration” fee that runs on top of direct costs–often about 50% of those costs, or adding 1/3 again as much to the total cost of the research–funds the SPO and a lot of administration as well.
There have been running battles between development offices and sponsored projects offices over whether incoming money is a restricted gift or is a grant. An audit in the UC system forced substantial changes in policy. At stake is whether a company can deduct the money as a charitable donation or has to book the transaction for a basic research tax credit or a business expense.
Finally, there is the technology licensing office, which deals in IP. The TLO works against both development and sponsored projects. For development, the TLO insists that no IP be offered in the context of donations. That would be a quid pro quo where the value of the IP might be way more than the value of the donation–one might not know that for sure, but it’s enough to assert it as a matter of principle. For the SPO, the TLO insists that no IP be offered in the context of sponsored research agreements. One reason given for this is the Tax Reform Act of 1986, which introduced the idea that private research conducted in tax-free bonded facilities cannot be offered up front to sponsors, but must be valued as if the sponsor had not funded the research at the time the “technology is ready”. This has been the subject of a couple of Revenue Procedure guidance documents which are anything but clear on the topic. Under the original Revenue Procedure, even the federal government’s support of research in tax-free bond supported facilities appeared to be a “private use”. Go figure. One might think that the problem was with a badly conceived Revenue Procedure. TLOs, however, opportunistically have used these Revenue Procedures to assert that no licenses can be considered until an invention has been made, reported, and assigned to the university for licensing.
TLOs have a number of other arguments as well–that allowing some IP go for no charge creates a “race to the bottom” that damages money positions for other IP, that taxpayers expect a “return on their investment” from industry using university research; that royalty sharing policies all but require a university to charge for IP or inventors will have a cause of action against the university (as if the inventors don’t have a cause of action for a university not licensing anything and as a result no one using their inventions!).
Recently there has been some relaxing of the TLO position in some universities, allowing a non-exclusive royalty-free license in some industrial research settings, but generally refusing to allow sponsor right of assignment, except in the case of federal funding under Bayh-Dole. The TLO position with regard to industry-sponsored research has been such a source of friction that a whole industry has arisen around trying to resolve it, led by the University-Industry Demonstration Project, which typically demonstrates how intractable the problem has become.
These three silos of industry relationship fail to work together when it comes to IP, other than to acquiesce in the TLO position that IP exists to be monetized by means of a license, not through other methods, such as hiring of students, personal consulting, or public acclaim leading to more support for the university. For many TLOs, this appears to be a matter of survival, as the TLO is funded from the proceeds of licensing, not from indirect costs of grants or heaven forbid donations.
The past thirty years have witnessed policy shakeouts among these three units, none of which has produced much that might appeal to companies by way of collaboration, relationship, or shared goals for research and development. Each unit instead has ended up with a money interest tied to its financial well being quietly shaping its policies.
Where does this leave affiliates programs? For that, consider the analogues. In the SPO, an affiliates program looks like a research consortium, adding a statement of work, a budget, and the rest of a research apparatus. In the TLO, an affiliates program looks like a non-exclusive licensing campaign, in which companies take a license to IP, and then get services of various sorts–help with explaining the invention (the *real* technology transfer), notice of improvements and updates, documentation and data of various sorts, and the like. In the development office, an affiliates program looks like a bunch of donors all contributing to the same program because they want to be associated with it, and gain access to expertise in the program, all without quite licensing anything or sponsoring anything.
There are ways to combine these programs into a single sort of relationship. Company funding for each sort of conventional program typically comes from different accounts with differing criteria. Donations are not sponsored research are not licenses-in. Consider one further kind of relationship, that typically is not recognized by American universities, but is practiced in other countries, such as the UK. This relationship is the service relationship, under which experts at the university help a company with its work in exchange for payment. In the US, this is the standard approach to faculty consulting, which generally operates outside university conventions–and often faculty are required not to use their university offices and lab equipment for such work. Universities in the US tend to downplay this sort of thing as “contract research”. But it need not be “research” at all. The closest thing to such services agreements are clinical trials with sponsor-supplied protocols. Universities justify this work on the grounds it is a public service to test new compounds under controlled conditions, to ensure public safety. Perhaps there is something to that. But a clinical trial really isn’t research, and universities routinely allow IP developed in a clinical trial to be owned by the sponsor, despite all the protests over IP in most any other sort of sponsored project.
Services are interesting, as compared to organized research. The reason for this is that services can be purchased, and purchases are something a company both understands and has rigged for. A purchase is about the easiest money transaction a company can contemplate–easier than a donation, a sponsored project, or a license-in. If an affiliates program is based on a purchase, then a relationship can be established readily, with the potential for a company to expand the relationship as it proves to be productive–motivation for university personnel to work to support company interest.
In copyright-based research licensing, services become apparent almost immediately. Unlike a patent, a copyright almost always must be licensed with a service–if nothing else, deliver of a copy of the licensed work. Thus, in a copyright relationship, any payment is distributed, necessarily, across both a grant of rights and the service of delivering a copy of the work to the recipient. The odd thing about TLO treatment of such a transaction is that the TLO fixates on the grant of rights, and assigns all of the payment to that grant, and nothing to the services, making the relationship appear to be a “license”. Perhaps that is understandable, because the L in TLO stands for “licensing”. But making a relationship into a license also tends to take it out of the purchase channel.
It is a better strategy to present an initial relationship to a company as a simple purchase of services–delivery–and along with that purchase comes a license document, which is also purchased. It is as if one purchased the piece of paper on which the license is printed. Much as one does when buying off the shelf commercial software, which despite being placed in a retail purchase channel–which is what you do when you buy a copy–the software company insists that one is licensing the software in that same transaction. And this is good, because if store clerks were required to negotiate a license with you before you purchased the delivery of the physical product, there would have to be attorneys available for both parties.
To set up an affiliates program as a purchase with regard to IP, one has to recognize that the IP will have to be part of a non-exclusive licensing program. One approach then is to bury the license in the services relationship, and place the value on the relationship, not the license. There is time to divvy up the purchase price (the membership fee) after it is received, assigning some portion to a university’s IP policy, and some portion to donation, and some portion to sponsored research, depending on how the funds are to be used. One can call such funds “designated funds”–rather like stem cells, they do not differentiate under policy until they are used for a recognized policy use.
A second approach is to grant the licenses upfront–before any affiliate relationship is formed. For instance, one can convert the lab into an “open lab,” with all IP made available on FRAND terms–fair, reasonable, and non-discriminatory. Once the terms are set for IP, then the affiliates program charges can’t be for the IP, and so the affiliates program can be operated by development, as usual. Further, any sponsored research accepted by a lab with a pre-declared open IP policy does not have to have an IP clause–there won’t be any IP to hold back, because IP is not subject to the sponsored research agreement, but to a lab policy regarding IP. By putting the license within the relationship, or putting it out in front, a lab effectively opens up all four avenues for industry relationships–donations, sponsorship, license, and services. Furthermore, one can price a basic relationship at a level that pretty much any company scientist or engineer can afford from his or her discretionary budget. Don’t believe it? We ran one project for a number of years with at one time 200 company sites paying $4K per year for a relationship with a lab under which they got software and assistance for a year. Do the math. It was an $800K/yr affiliates program, or consortium, or licensing program, or services program.
Put it another way, no sponsored projects office is interested in handling $4K sponsored research projects. They expect at least ten times this amount to be interested. Same for development. And for the TLO–they will do such deals, but grudgingly. It is only when a deal scales by the number of companies, and extends over a number of years, and has the potential for adding additional services and other assets does that first $4K transaction start to look very interesting. For the most part, however, development offices, SPOs, and TLOs do not understand transaction scale. Perhaps development offices understand this more than anyone, and are used to fund-raising campaigns that may involve lots of small donations that add up. But for the patent-heavy, exclusive license defaulting TLO, such scale looks like wasted effort for little effect. Again, we were doing $800K/yr and the companies were happy with providing the support. From their perspective, they were getting a 200 to 1 return on their investment in the lab. A pretty good deal. Call it network dynamics.
By using strategies such as these–creating a purchase relationship as the primary relationship carrying the value of the transaction, embedding licenses in that relationship or pre-declaring the licensing protocols, and allowing additional funds to come in from any company source available to those interested in engaging the lab, a university can resolve the problem of its three silos and get back to building collaborations with the technical expertise in industry, and while they are at it, generate the kind of support they need to carry on their work. [I have template documents that show how such an arrangement can work. I will try to get them posted.]