Germany has produced its own version of template agreements for research between universities and industry, called the “Berlin contracts.” There is even an English translation that makes for interesting reading. German law is also interesting because in addition to patent law they have the German Employees Invention Act (EIA). Under the EIA, inventions are classified as “service” or “free.” Here is a good explanation of the differences:
Service inventions are those made during the term of employment which either result from the employee’s activities in the business or public service, or are significantly based upon the experience or activities of the business or public service. In practice, most inventions made in the course of the employment relationship are service inventions. All other inventions are free inventions which are owned by the employee but subject to the limitations of §§ 18 and 19, namely that the employee inventor has an obligation to notify the employer of the invention and offer the employer at least a non-exclusive license, if it wishes to exploit the invention.
The EIA establishes a requirement that the employer who retains a service invention made by an employee must compensate the employee for the invention. The EIA also provides a protocol for how this compensation is determined. There are two steps. In the first, the value of the invention is determined using one of three approaches:
- by using a licensing analogy, i.e., by determining the license fee that would have to be paid for the use of a comparable invention owned by a third party,
- by calculating the benefits from the invention accruing to the employer, i.e., the difference between costs and revenues resulting from the use of the invention, or
- by estimation of the value of the invention, i.e., by determining the price which would have had to be paid by the company to buy the invention from a free inventor.
The second step involves creating a scale factor on the total value. This “share factor” takes into account such things as the inventor’s position in the company, whether the inventor was working on a problem for the company that led to the invention, and the involvement of the company in directing the invention. This determination of share factor involves the assigning of points on a 20 point scale that is then mapped into a multiplier for the total value, leading to the compensation. Here is a detailed account. The upshot is if the invention arose in an assigned task, solved a problem using company knowledge, and was senior in the company, the share factor is going to be close to 0%, and if the invention was outside of assigned work, if the invention is wildly insightful and draws on outside information, and is done by low level or unskilled employees, then the share factor will be near 100%.
The employer must pay the compensation within 3 months of the employer’s receipt of assignment if the employer is using the invention, and if the employer is not using the invention, up to 7 years to study it before making payment. A good summary of all this in the context of recent revisions (c. 2009) to the EIA is here. (The UK has a different arrangement, allowing employee inventors to seek additional compensation if the employer-owner of their inventions enjoys an “outstanding benefit.” A recent discussion is here.)
When an employer claims an invention, it is required to file a patent application or a utility model. Anywhere the employer doesn’t file, the employer must release the rights to the inventor, reserving the right to use the invention in those territories. The reservation of rights, too, requires reasonable consideration.
Another provision of the law was amended in 2002. That provision provided that university faculty inventions were “free” inventions. The law was amended in a number of ways. Faculty were required to disclose their inventions, universities were granted the same right as other employers to take ownership of these inventions, and faculty inventors were provided with at least 30% of gross licensing income. According to one article, these changes were made to make German law parallel with Bayh-Dole. Sadly, the parallel being sought appears to have been with the fingers-crossed-secret Bayh-Dole Act, the one depicted as vesting title with universities. There are, however, parallels worth exploring.
All this means that for university research contracting, an industry sponsor has to deal with two issues–first, the university, which is assumed to retain title to inventions, but has the power to release them; and second, the investigators, who have personal rights in inventions until assigned to the university, and after assigned have a right to rights everywhere that the university doesn’t pursue patenting. The industry sponsor has to deal with both situations in parallel, ensuring that regardless of how the relationship between the university and the investigators shake out, at least one of them is able to convey whatever rights to the sponsor that may have been promised.
We can look then at how the Berlin contracts address the EIA and the alternative ownership conditions, and follow with an examination of parallels with Bayh-Dole. Before we jump in, however, I should point out an insightful study of the effects of the compensation protocols on German inventors by Dietmar Harhoff and Karin Hoisl, who surveyed a number of inventors.
The Berlin contracts present four templates, one involving “Research Collaboration,” two that are variants of “Contract Research”–one involving assignment of rights, and the other licensing of rights, and one that involves Research Services–essentially doing a task to spec without the need for independent analysis. The discussion that accompanies the contracts is well worth the read. I won’t summarize it all here, but I will draw attention to some general points.
The Berlin contracts are directed primarily for the use of small and mid-sized companies that do not have access to robust contracting resources. The experts who developed the Berlin contracts emphasize the importance of drafting a careful statement of work. Much of the ownership and rights provisions in the contracts depend on the scope of work. All too often in American practice at least, university contracting officials ignore the statement of work–some don’t read it at all–and rely on a general statement of the area of the research in the text of the contract. Doing so in the context of the Berlin contracts can greatly expand the scope of IP rights in play, both “new rights” (developed in the research) and “old rights” (necessary for the exploitation of the new rights).
One might note that a narrow statement of work actually provides for greater flexibility in committing IP rights. A narrow statement, especially one that defines a specific deliverable of interest, makes it much easier to commit to assigning or licensing rights in that deliverable to the sponsor for its use. In book publishing, an author may readily pre-commit publication rights to a publisher if there is a clear statement of the work that the publisher desires to publish–a manuscript of so many words on a given topic, written for a particular audience, by a certain date, with authorial services such as proofing and assistance in marketing. As the statement of work becomes broad and indefinite, who can possibly anticipate what rights, exactly might be created and what other rights might therefore be implicated? In such circumstances, one has to be conservative in what is offered by way of rights, leading to a kind of caginess that comes off as subterfuge–“we can’t make any commitments until we see what is developed, and how valuable it is, and then you can beg us for access to it, if you have enough money.” From the university administrative perspective, it’s all reasonable–but only after folks have perhaps been lazy or indifferent with their treatment of the statement of work.
The Berlin contracts assume that administrative proprietary control of inventions is the primary objective of the agreement. If the company sponsor obtains ownership, then (so the assumption goes), the sponsor will use the invention, develop product, and life goes on. If the university retains ownership, then it has a mandate to find other commercial partners. Implicit in all this is that the inventors cannot find such partners, or will not (being academics). Left unconsidered is that inventors need not find such partners on their own, but could use other agents to do so, or could ignore proprietary approaches and release their inventive work to standards, to open platforms, or engage in a simple RAND–reasonable and non-discriminatory licensing program, such as one might find in an industry affiliates program or research consortium.
Thus, the Berlin contracts are all about dividing up the proprietary spoils of research between employer and sponsor. That is their greatest limitation, in considering the activities that go into university scholarship and how new technology becomes available for use. Once the idea is that the sponsor of such activity should own, or hold an exclusive license, as the benefit of their bargain, and that it is the employer’s obligation to see that this happens, with the expectation of compensation for the research, for old rights, and something reasonable for the new rights, then the Berlin contracts thoughtfully navigate this geography.
A few of their management points worth mention. The Berlin contracts distinguish joint ownership and interest in patenting activity (see Section 6). If in a Research Collaboration (where both university and industry personnel expect to be involved), an invention is joint, then if it is more than 50% university staff contribution, it will be owned by the university. If it is less than 50% university but more than 20%, then both the university and sponsor will be joint owners for the purposes of filing a patent application, but the university will assign its ownership interest to the sponsor 18 months after filing. If the university share in the invention is less than 20% (9.1), however (such as, 1 university inventor among 6 sponsor inventors), then the university should have no involvement in the patenting process.
The Berlin contracts must navigate the IEA, so they necessarily must involve the principal investigator and any other research personnel personally. The sponsor must have assurances from both the university and from its personnel, since it cannot rely on the university to properly claim (or as the law is now, improperly unclaim) inventions made by faculty. Thus, the principal investigator is made a party to the contract, and agrees not to involve others in the work unless they make a written declaration agreeing to the terms of the contract–specifically with regard to the obligation of their rights in inventions and know how covered by the contract.
The various versions of the Berlin contracts vary mostly in their handling of IP. In Research Collaboration, IP is owned by the university (unless solely made by the sponsor), and licensed to the sponsor. The contract provides two alternatives for handling payment. In the first alternative, the total cost of the research contract is to include payment for any rights (old and/or new) that are licensed in the course of the work, taking into account the anticipated rights that may be developed and their typical value in industry. The contract takes pressure off this determination by including a “re-opener” clause if the benefit to the sponsor appears to be disproportionate to the amount paid, either the sponsor (if it paid too much) or the university (if it received too little) can re-open the payment clauses and seek an adjustment.
The second alternative proposes a basic fee for each inventions upon the filing of a patent application (or after a set time after disclosure, whichever is sooner), with an additional payment due when the invention is used by the sponsor “for commercial purposes.” The additional payment may take any of three forms: a) lump sum upon use with a kicker if the sponsor waits too long; b) milestone payments on a schedule based on sales; c) “adequate remuneration” to be determined later “in due course.” In each alternative, the payment is tied to a patent family, not to each particular patent, which makes a lot of good sense. The kicker payment for not using commercially in alternative a) also is a nice touch, as it suggests that the institution is willing to give up a share of payment to encourage use.
The Contract Research agreements come in two flavors, one in which inventions are assigned to the sponsor and another in which they are licensed. The drafters note that small and mid-sized companies may benefit by allowing the university to file the patent applications (and pay the costs) as a preferable alternative to taking ownership and dealing with the costs and infrastructure requirements directly. Rather than seeing the issue as one of ownership, this approach considers costs and control.
In the assignment version of the Contract Research agreement, the university and the project manager (principal investigator) both make a present assignment for inventions made in the research (6.2). There is a breakout for “research and teaching activities” which amounts to a grant back by the sponsor for all inventions. If inventions are made “that do not relate to the subject matter of the Agreement,” the university will offer the sponsor a non-exclusive license “on reasonable terms.” Payment (13.1) for the primary rights–inventions made within scope–is included in the total amount for the research. As with the Collaborative Research template, there is a re-opener clause (13.2) if the benefit is disproportionate to the agreed-upon payment. Any additional payment, however, must take into account the payments under the research agreement as well as any “significant change in the inherent purpose of the Agreement.”
The license variant of the Contract Research template has ownership with the university, but the university grants an exclusive license to the sponsor upfront (6.2), with payment included in the total quoted for the research (11, 12.1), with the same re-opener for disproportionate benefit. It should be noted that disputes go to arbitration, so the effect of this approach is to put an assessment of the benefit relative to what was anticipated in the statement of work and contract in independent hands. As a strategy that makes a lot of sense for the relationship, and puts pressure on both the university and the sponsor to keep things reasonable and proportionate to what has been anticipated. I wonder how many times the arbitration clause for disproportionate payment has been invoked–it’s mere presence may be sufficient to ensure that it won’t be invoked very often.
The Agreement for Work and Services is more compact, eliminates the investigator as a party, and assumes that no IP is going to be created worth bartering over upfront. Anything that does arise is the university’s. A confidentiality agreement covers information provided by the sponsor. Presumably, if the provided information specifies the work to be done, then the work gets done, nothing unexpected happens, and the company gets the benefits of its deliverables. IP, in essence, drops out of the agreement as uninteresting. Rather than leaving a robust apparatus, the drafters reference IP, disclaim its importance, leave it with the university, and move on. Again, a sensible strategy for de-emphasizing issues that aren’t generally material to such contract work. American universities would do well to include this register of agreement. It is not the “faculty proposal” based “basic research”, nor is it merely selling services, as it may rely on distinctive expertise and equipment available at the university. However, this type of relationship focuses on services rather than on, specifically, research of the exploratory or problem-solving variety.
The Berlin agreements have some parallels with the situation under Bayh-Dole. Like the Berlin agreements, Bayh-Dole works in parallel with the investigators and the contracting university. The Standard Patent Rights Clause (37 CFR 401.14(a)) sets up a primary agreement with the university (the funding agreement–typically involving the requirements at 2 CFR 200, Circular A-110) and then incorporates a secondary agreement under which the principal investigator and any research personnel commit to offer their personal rights to the government. The key in the SPRC is that the primary agreement requires the university to require this secondary agreement of its employees. That is, the authority for the requirement is the university, not the federal government, and when the university requires this secondary agreement, it is in effect cancelling any other requirements it may have that would conflict with this agreement, as a condition of receiving the federal award.
As with the Berlin agreements, the secondary agreement (the (f)(2) agreement) under Bayh-Dole protects the government agency sponsors from whatever might take place between inventors and their employer. Indeed, Bayh-Dole is set up to allow a range of such relationships, and rather than dictating how such relationships must be managed, Bayh-Dole is satisfied so long as the funding agency has an agreement with everyone involved within the context of the primary funding agreement.
Attached to the Contract Research agreement is a nice instance of a participation agreement for others that join the research project (see Annex 3, for instance).
The Berlin contracts assume ownership of inventions as a precondition for use. A similar assumption goes into the implementations under Bayh-Dole at many American universities. Without title, and “title certainty,” so the argument goes, no invention will be used or developed. Yet we see around us many instances that demonstrate that such a claim surely cannot hold up in general. The internet and world wide web, for instance, both are built on inventive technologies that became standards and common platforms without passing through a regime of ownership, title clarification, licensing, and commercial product development. Once an ownership and proprietary development process is assumed, then templates like the Berlin contracts are excellent expressions of the issues that need to be considered. However, if one considers open strategies, or anticipates collective development, standards, platforms, and libraries of tools, then setting up industry agreements to focus on ownership as an anchor rather misses the point.
Indeed, in coming to a university for research services, a company may very well wish to establish a neutral ground on which it and its competitors, and companies with complementary technology might work towards a fundamentally new platform on which products and services might be based. That would be a distinctive role for universities to play, one that is much more challenging–and difficult–for any single company to pull off. It is native to university culture, or at least it once was, before the misinterpretation of Bayh-Dole led so many university administrations to impose an ownership model. Bayh-Dole does not impose such an ownership model. Inventions are to be reported, but there is no mandate for universities, or faculty inventors, to own them, nor for the funding agencies to own them. The idea was, to draw from the pool of those inventions those that might benefit from patenting, and use the patent system to promote their use. Nothing about making money or adding value, but for extending utility.
In the rush to “change university culture” into one that is motivated by money and hot cars, university administrators have also overrun the native university culture, the one that was generous with its scholarship, that preferred publication to exclusive rights, and did not generally attempt to put a price on each and every finding of any significance. Instead of adopting a single model–that of demanding payment for patent rights–an approach that is surely causing great damage, American universities have an opportunity to develop a diversity of approaches, including ones that create commons, platforms, libraries, and standards. For this, we need a different set of templates to manage university-industry relationships. It is not a matter of who owns an invention, but rather where the control points are. For control, what matters is the governance. When that governance is a commons, or consortium, or standards setting body, then the associated contracting can differentiate between “essential claims”–claims of a patent necessary for the operation and objectives of the commons–and “nonessential claims” that can be held and managed by whatever owner happens to hold those rights.