|
Richard
Nelson
One pattern, arising after the Morrill Land-Grant Acts in the mid-19th
century, is direct federal and state support for portions of the university system specifically
working on problems bearing on agriculture. At experimentation stations affiliated with land-
grant universities, faculty members worked on diseases of wheat plants, appropriate ways to
fertilize and feed crops, and similar questions; parts of the university were thus dedicated, by
intent and by funding, to the farming community and related industries such as pesticides and
food processing. A number of engineering schools, too, have offices directly concerned with
solving problems for local industry.
In a different model, individuals or groups at universities perform basic research to understand
particular phenomena, then publish an article about that understanding; after reading this
public information, companies are able to better design products. Since the 1930s, for
example, a considerable amount of university research has allowed the aerodynamics industry
to benefit from new knowledge about the shapes and behavior of propellers. The biomedical
research community and the pharmaceutical industry have had the same type of relationship;
until recently, universities had little to do with the direct creation of new pharmaceuticals, but
quite a bit to do with the understanding of biochemistry and diseases.
Another entirely different model for collaboration involves clinical trials. University medical
schools have been the place where most pharmaceuticals are tested and tried. This
relationship may be associated with the development of new techniques of dealing with
patients, or with the attempt to develop new medical devices, but these trials generally take
place in an entirely different part of the medical school than the research departments. The
relationship model associated with biotechnology--the model most commonly touted in the
press--appears when university researchers themselves develop an embryonic medical device or
technique that involves them immediately with industry (either existing firms or their own
new startups).
Basic and applied and public and private
Universities and companies differ in important respects, particularly their approach to the
public or privatized character of knowledge. As a general rule, firms do not share the academic
tradition of publishing results and sharing both information and materials with colleagues.
There are important exceptions, however, in fields where corporate labs have to keep up with
the frontiers of knowledge by employing the people actually working on those frontiers.
Scientists in the large pharmaceutical companies, or the great diversified electronic companies
like AT&T, IBM,
and General Electric, have one a considerable amount of
publishing, as have scientists in the new biotechnology firms (with careful screening for
patentable material early in the process). By and large, though, publication comes from only a
small fraction of the industry research enterprise. In addition, over the last 15 or 20 years,
industry funding of university research has cut down on the extent to which researchers
publish their work freely. Scandals involving direct suppression of findings have occurred and
need to be guarded against, though they are rare.
The Ford Foundation reported a few years ago
about university-industry contractual relationships placing strict constraints on publication,
including industry pre-clearance. Unfortunately, this study did not go on to distinguish
between terms different universities agree to in contracts. Columbia does not sign contracts
involving prohibition of publication, although lesser restrictions (e.g. delay) are permitted.
This is also the case with Harvard, Yale, Stanford,
the University of California system, and others; at
present, the premier research universities are doing a good job protecting the ability of
scientists to publish, but policies differ greatly depending on the partnership model, the
discipline, and the relation of a university to local industry. Confidentiality is important, for
example, in a state university's engineering school working closely with a local furniture
industry; in the absence of widespread interest in a field, publication isn't important. On the
other hand, people at the same engineering school working on semiconductor technology,
with significant finance from large companies, are likely to fight for the ability to publish their
results even though the companies will want privileged access to the findings, because it is
absolutely essential for the scientific community at large to know about developments in
semiconductors.
Policy change and philosophical
implications
Universities, of course, are not monolithic; they are also congeries of separate schools and
departments. The changes evoked by Bayh-Dole and by the development of technology
transfer offices provide an occasion to re-examine the nature and mission of the university
itself. The typical American university, we should remember, is not the Ivies, which have traditionally stood
relatively aloof from the mission of fomenting economic development, but the large state
universities, which have always been involved in local industry, agriculture, and engineering.
With Bayh-Dole, this mission has been highlighted and, to some extent, hyped, but it has
always been present. In the modern research university, two conceptions are intermingled,
and to some extent at odds: the ancient notion of unified human knowledge, of the university
as a community of connected scholars and students, vs. former University of California
president Clark
Kerr's idea of a "multiversity" reflecting the progressive specialization of disciplines.
Presidents and provosts have the important role of fighting to preserve the idea of the
community of scholars against fragmentation.
An important and controversial consequence of technology transfer is the notion that
universities ought to engage in establishing intellectual property rights and reaping income
from licensing inventions. Most of a university is not directly affected by this process;
Bayh-Dole has brought little or no change to the fields of physics, astronomy, evolutionary
biology, the social sciences, or the humanities. Inventing at any university is concentrated in a
small number of fields, such as medicine, engineering, chemistry, and telecommunications, but
the funds from licenses (at least here at Columbia) are allocated proportionately to the
department where the inventor or inventors reside, to the school, and to the university as a
whole. It is not at all clear that patent-holding departments are delighted to give a fraction of
what they earn to the history department; there is a natural tension between
revenue-producing departments and university administrators who decide on the allocation
schemes. Yet the prestige of a university like Columbia, viewed as an entire intellectual
community, benefits revenue-producing departments by facilitating the recruitment of the
most able researchers in their fields. Those departments have an incentive to help the
institution as a whole, because the strength of the whole in turn strengthens its parts.
Though entities such as Columbia Innovation Enterprise bring about
some organizational convergence, technology transfer remains a set of discrete processes. It is
by no means certain that these processes are converging into a single "win/win proposition," as
the rhetoric behind Bayh-Dole presumes. Divergent interests inside and outside the university
have a stake in the division of the pies. Industry in general needs a university's research, and
companies want the patent; in some arrangements there are no broader interests arguing
against one company having it. But a non-trivial fraction of the things universities are now
licensing would previously have been put into the public domain. What a university's own
patents and licenses do is to force companies to pay for techniques and material they are used
to having access to for free. The large pharmaceutical companies, in particular, have begun to
complain vociferously that since they and the public pay for this research through taxes given
to the university, it is not fair for them to pay again for access. On the other hand, the
university's position is that this is an appropriate way to finance research, through a "user
charge." Tension also appears within industry: between biotech-firm specialists wishing to pick
up exclusive licenses on university-generated technology, which they can in turn use to
develop products they can license to large pharmaceutical firms, and those large firms, who
would just as soon keep the technology in the public domain. It is no surprise that technology
transfer is generating brisk business for another profession as well: attorneys who conduct
intellectual-property litigation.
By and large, universities like Columbia have avoided a
one-shoe-fits-all philosophy, recognizing the diversity of mechanisms involved. There are
officials elsewhere who emphasize strictly proprietary concerns ahead of the mission to
generate public knowledge, narrowly interpreting the language of Bayh-Dole regarding the
university's role in economic development; that way, I believe, lies trouble. But as long as
policies accommodate these differences and universities continue to recognize that their raison
d'etre is contributing knowledge to the public, the research community can balance public and
private interests, to everyone's ultimate advantage.
There is no such thing as "technology transfer"--or at least no single thing.
Rather, different disciplines have given birth to different models for university research
structures to interact with industry. The lion's share of fundamental research in the United
States is done in universities; some of it is done in governmental labs, and a small fraction is
done in industry, generally in only a few companies. It benefits all parties for knowledge to
pass from one sector to another, but a simple model of linear transfer from basic research to
commercial applications is inadequate to account for the complex interplay of interests
involved.
There is a widespread belief among laypersons--though much of the scientific community has
also picked it up by osmosis--that basic and applied research are essentially different and
separate; in other words, that most research aimed at fundamental understanding proceeds
with little notion of what practical problems the research might eliminate. In reality, except
for a few cases like Maxwell's work on electromagnetism, proceeding with only the vaguest idea of what it might be
used for, basic research has never been divorced from applications. The scientists who
developed the theory of thermodynamics, or the whole spectrum of work in medical schools
and biology departments, have had good ideas about the applications of their fundamental
understanding. This has been known for a long time, but the knowledge tends to be repressed.
The late Donald Stokes's wonderful
book Pasteur's
Quadrant (Washington, DC: Brookings Institution Press, 1997) makes a
compelling case that a great deal of fundamental scientific work is also clearly targeted at
particular problems and occurs after a basic technological breakthrough has occurred, rather
than the other way around. The development of the transistor at Bell Laboratories in the late 1940s is a good example:
William
Shockley, one of the primary inventors, then spent about a year trying to understand it,
then wrote Electrons and Holes in Semiconductors (NY: Van Nostrand, 1950),
which revolutionized solid-state physics and won him the Nobel Prize.
The Bayh-Dole Act of 1980
represented a sea change in attitude and emphasis in this whole arena. Before Bayh-Dole, only
a few major universities were actively engaged in patenting and licensing; more (including
Columbia, until the mid-1970s) merely allowed individual faculty members to patent, or even
discouraged patenting, as Columbia did in the health
field. Culturally and in policy,
universities placed inventions in the public domain. Bayh-Dole changed this philosophy
dramatically, especially with the rise of biotechnology. In the 1970s it was unclear what
aspects of biotechnology would and would not be patentable; now, what's coming right out of
medical schools is directly feeding into patentable, potentially profitable products and
processes used by companies. In engineering schools, which had always been patenting,
Bayh-Dole produced quantitative rather than qualitative change.
Related links:
RICHARD
NELSON, Ph.D., is George Blumenthal Professor at Columbia's School of International and Public
Affairs, professor at Columbia Law School, and professor at Columbia Graduate School
of Business.