Dr. Dobb's Special Report December 2000
The lifeblood of new technology is research. In an ideal world, data and experimental results would be shared freely among those working toward the common goal of discovering new technologies for the benefit of mankind. In the real world, however, research is more often than not driven by profit motives -- by the quest for developing new products that can yield large profits for the principals who developed the products, and the investors who are backing those principals. So, often the public interest and the profit motive collide.
In the case of privately owned or corporate research institutions, the situation is fairly clear cut. The charter of these types of research entities is to find and develop new technologies that will produce revenue and profit for the company. Any research findings shared with other companies are generally done so through commercial licenses. Xerox PARC is a classic example of a corporate research center that has licensed many of its discoveries (sometimes to its later chagrin) to other companies (the GUI that first appeared on the Macintosh, for example). For the most part, private companies can do what they want with their research findings.
In academia, the ethics of research versus profit are more complicated. Large multidisciplinary universities depend heavily on research to recruit top experts in their fields, bring in much needed grants and other forms of external funding, and of course aid in the teaching of students. But taking that academic research and using it to produce a product and selling it via a startup company is not supposed to be part of the charter. Nevertheless, that practice has become commonplace. A good example is Stanford University, with Jim Clark and Silicon Graphics, Andy Bechtolsheim and Sun Microsystems, and John Hennessy (now the president of Stanford) and MIPS Computer, to mention a few.
The benefits for universities participating in technology transfer programs are enormous. According to the Association of University Technology Managers (http://www.autm.net/), in 1998, academic licenses generated $33.5 billion in economic activity and supported 280,000 jobs. Furthermore, 63 percent of the total 3668 new licenses were granted to companies with fewer than 500 employees. Of the small companies licensed, 364 were new companies created to develop and commercialize results of academic research. To further put this in perspective, this 1998 research activity is based on expenditures by the institutions of $24.4 billion. In comparison, the total 1998 sponsored research funded by federal government sources was $15.3 billion, while industry-funded research was $2.4 billion.
Stanford is a private university and has had experience dealing with the issue of researchers going off to start up companies. It routinely grants professors two-year leaves of absence to start commercial ventures. And it routinely requires researchers to license from the university intellectual property that was developed as a Stanford employee.
For example, Mendel Rosenblum, the founder of VMware and principal developer of its VMware software, is a professor of computer science at Stanford, specializing in operating systems -- specifically virtual machine operating systems. VMware has been a successful venture, although much of the original research that led to the VMware software was undertaken at Stanford. The difference is that VMware is a desktop tool for the x86 processor, while the research done at Stanford focused on large multiprocessing systems from SGI. According to Rosenblum, VMware does not pay a license to Stanford, but will have to do so for a new server-side virtual machine OS product.
From Rosenblum's perspective, the biggest problem with research professors starting companies is the difficulty universities have in retaining faculty. With his two-year leave expiring, Rosenblum is returning to Stanford and will be teaching a class normally taught by another professor who is starting a two-year leave to launch his company. The "brain drain" problem is confirmed in a September 19, 2000, front-page story in The New York Times about research institutions struggling to retain top researchers who are being lured away by higher salaries and stock options at venture-capital backed startups.
While there are no easy answers, guidelines are in place to help university administrators and researchers find a common ground; see the IEEE's "University Intellectual Property Policy Guidelines" at http://www.ieee.org/usab/documents/forum/library/positions/university.html. Nevertheless, profits should not interfere with teaching and the sharing of information. And that is a growing danger in academia. If the primary goal of researchers is to make a fortune with a startup company, then we will all lose in the long run.
Nicholas Baran
Senior Editor
nbaran@ddj.com