Overcoming Barriers to Collaborative Research: Report of a Workshop (1999)

One theme from an earlier meeting on collaborative research organized by the Research Roundtable, the Industrial Research Institute, and the Council on Competitiveness was that structuring and managing partnerships that produce real gains for all partners takes experience, careful planning, and ongoing attention.⁸ At the March 1998 workshop, attendees discussed barriers to collaboration and the approaches that have been developed to overcome them. It was not possible to cover all issues comprehensively, and the discussion revealed several issues on which important work could be done in the future.

In this chapter the issues are divided into the following categories, which are often interrelated.

Partners May Lack Understanding or Trust. In some cases partners enter into agreements with an inadequate understanding of the management, internal politics, decision-making structures, and even fundamental interests of the other partner, resulting in slow decisions and insufficient resources.

Possible Solutions. The importance of building trust between partners and effective "relationship management" emerged time and again during the workshop. The right mechanisms may depend upon the scale and substantive focus of the collaboration. For example, large mega agreements may require regular meetings among senior management of the company and university, as well as regular exchange at the bench level. Eugene Slowinski of Rutgers University described several specific techniques that could help bring potential problems to light at the negotiation stage: stakeholder mapping (in which stakeholders in each organization are identified), decision-

making analysis (in which the decision-making processes of each organization is made explicit), and expectation mapping (in which the expected roles and contributions of each partner are analyzed).

Industry and Universities Often Have Different Time Horizons. Industry and universities have different time horizons for good reasons. Although some senior management officials in industry are concerned that universities are becoming too short term in pursuing specific projects and collaborations, industry is generally operating on a shorter time horizon than academia. In some cases discussed during the workshop, industry partners were seen as disruptive by universities when industry pulled out of projects on short notice or hired students in the middle of degree programs.

Possible Solutions. For broad collaboration (master agreements) and focused collaboration (clinical trials), misunderstandings of the preceding type appear to be uncommon. More commonly, problems arise with projects with less well-defined outcomes than clinical trials. Several participants noted that universities should avoid overselling in terms of potential accomplishments and timelines. Likewise, industry should understand that, in most sponsored research, the effective time unit is the time it takes to complete a Ph.D. dissertation, usually three years. Although some companies are exploring the possibility of sponsoring university research with a faster turnaround, it is unclear whether it can become a standard practice.

Institutional Reward Structures May Act as Disincentives to Collaboration. In some cases, because key staff members of the industry partner may not be rewarded financially if the project succeeds, the staff members assign collaboration a low priority. Financial rewards resulting from collaboration are more likely for university faculty members. On the other hand, collaborative work can impede young faculty members from getting the intellectual recognition necessary to gain tenure. In one example discussed at the meeting, a young faculty member spent a great deal of time and effort on a collaborative project, probably undermining her chances for tenure.

Possible Solutions. Companies have assigned liaisons with project responsibility, resulting in management of collaborative work being included

in employee performance evaluations. Often, these are researchers who are working on a similar in-house project. On the university side, participants reported that it is possible to structure tenure decisions and other promotion processes in ways that give faculty credit for being effective collaborators with industry. Industry partners can be asked for an evaluation, for example. Government can also play a role through its funding decisions. The National Science Foundation's Engineering Research Center and Science and Technology Center programs have encouraged partner universities to create mechanisms to recognize effective collaborations.

In recent years, a number of universities have sought to implement innovative incentive structures by starting or expanding programs aimed at facilitating the launch and growth of start-up companies based on university research. Such structures include university-managed incubator facilities, provision of seed funding in return for equity, and focused efforts to attract management talent and catalyze company formation. In Appendix B Teri Willey discusses ARCH Development Corporation, which has been undertaking these efforts for a number of years. A university moving in the opposite direction is the University of Arizona, which settled a patent suit brought by a start-up several years ago and now licenses technology to only established companies. ⁹

Student Time May be Misused and Conflicts of Interest May Arise. When a faculty member holds an equity stake in a company that sponsors university research and has graduate students working on that research, tensions and suspicions can arise. In one case, a faculty-owned company hired graduate students as consultants, blurring the distinction between student and employee. The university-based participants in the project later considered this arrangement a mistake.

Possible Solution. Participants reported that experienced universities had developed policies to deal realistically with these issues. Although some faculty members may wish to minimize involvement in collaborations by the university's research administration and departments, others assert that open communication with these bodies can prevent abuses, as well as suspicion and misunderstanding about such arrangements. Some universities and departments simply do not allow students to become regular or part-time employees of research sponsors.

Industry May Seek to Stretch Resources by Not Paying Indirect Costs and Faculty May Pressure the University to Agree. Given the

substantial indirect cost rates of research at universities and elsewhere, it is understandable that some persons in industry are reluctant to pay, or wish that all the funding be used only for research. If indirect costs are waived on industry research, however, they must be made up somewhere else (e.g., tuition or other research grants).

Possible Solutions. Several of the industry participants recognized the need to pay indirect costs, and creative approaches have been developed to increase incentives to do so. For example, some universities have traded current overhead recovery for a greater share of downstream royalty income or for equity. Care must be taken, however, because indirect costs are current and downstream income is uncertain, and any trade-off must not short-change another part of the university.

Another example is the state of California's Microelectronics Innovation and Computer Research Opportunities program, launched in 1981. For approved projects, California puts up a third of the funding, the company puts up a third, and the University of California campus involved waives overhead on the industry and state funding, essentially providing another third. University participants reported that this approach has been very successful in encouraging a broader range of companies to support research.

This topic deserves more examination than was possible during the workshop. A future examination may need to take the position of the federal government into account.

Are Intellectual Property Rights an Inhibiting Factor in Research Collaboration? Several of the industry participants saw some universities increasingly taking too restrictive an approach to licensing and putting too high a value on their intellectual property contributions. Industry is increasingly seeking out second-tier U.S. universities and foreign universities for collaboration when they perceive first-tier universities to be too difficult to deal with. Some university boards of trustees may see technology transfer activities more as a revenue source than as a component of the university's public responsibility to assist in commercializing research results. This attitude can raise barriers to negotiations that actually reduce revenue over the long term. Given that only a small percentage of university-generated inventions produce significant revenue, some participants likened the strong emphasis on protecting proprietary rights of some universities to "buying lottery tickets."

Most of the discussion of this topic and suggestions from both industry and university participants focused on issues related to the university side of collaborations. There was also recognition, albeit with less detail and fewer examples, that the effectiveness of industry approaches also has a major impact.

Possible Solutions. Participants expressed a broad range of views on this issue. It is important that faculty, as well as university and industry leaders, understand that the role of intellectual property in the innovation process varies by field. Approaches that make sense in the biomedical field may not make sense in engineering and computer science. Several participants suggested that universities consider forgoing all proprietary rights outside the biomedical area, essentially putting inventions in the public domain. Other participants responded that many universities do not seek patents on their inventions unless an industry licensee has been identified, and that this approach is more likely to facilitate commercialization than a blanket policy of not patenting inventions outside the life sciences. To many participants, the main issue is whether universities manage their technology transfer roles to comply with the intent of the Bayh-Dole Act by enhancing the use of university-generated inventions. Several speakers believe that a well-run technology transfer operation governed by a realistic university policy can do this more effectively than a general policy of putting inventions in the public domain.

In addition to university licensing policies, premature definition and valuation of intellectual property can become an obstacle at the initiation stage of a collaborative project. Granting the company the right of first refusal to negotiate an exclusive license is one commonly used practice to delay concrete negotiations until the commercial value of an invention is easier to assess.

Patenting of Research Tools May Discourage Beneficial Research. This topic increasingly appears to be an issue in the biomedical area. For example, different universities may hold patents to different receptor cells of the same class that influence a disease process. To start a research program it would be necessary for a company to license the rights to all these receptors. If each university demands a 1% or 2% royalty, the company may not find it feasible to go forward with the program due to high transaction costs or the prospect that the overall royalty will be too high.

In one case described in the workshop, related receptors were developed by a university with federal funding and were licensed exclusively to the

university's for-profit subsidiary. When a company inquired about licensing the receptors, they were told that they would need to sponsor research at the university as a condition. One industry participant noted that companies have been fairly receptive to the use of their research tools in the past, and wondered whether the companies would be less so in the future.

Possible Solutions. Universities and industry should both want the widest possible use of research tools. Licenses that carry low rates until an invention generates a certain level of income have been used in some cases. One industry participant urged universities not to patent partial gene sequences and research tools at all, since it might be very difficult to prove infringement. There was some discussion of universities defining a "deminimus threshold" for an invention before a university would seek a royalty-bearing license.

Publication Delays and Non-Disclosure Requirements May Impair the Openness of the University Research Environment. Companies sponsoring research often need time to evaluate whether applying for a patent is worthwhile, and ask universities to delay publication of results. Such delays can impair the open research environment or prevent junior faculty from building a strong record of publications needed to gain tenure.

Possible Solutions. There is clearly a wide range of practices among universities in this area. Some do not allow any delays in publication. Many allow publication delays of 60 to 90 days. One university allows a delay of up to two years on a case-by-case basis if no graduate students are involved in the research and non-tenured faculty members sign a statement indicating that they understand the policy. Despite these diverse approaches, a number of participants seemed confident that effective ways of handling these issues are widely available. In the area of non-disclosure, one industry participant reported that in the rare instances when there is a need to share confidential information with a faculty member, this can be done with a consulting agreement separate from the sponsored research agreement.