The market economy, and the scientific commons

Text-only Preview

Research Policy 33 (2004) 455–471
The market economy, and the scientific commons
Richard R. Nelson∗
Columbia University, School of International and Public International Affairs Building,
420 W. 118th Street, New York, NY 10027, USA
Received 10 March 2003; received in revised form 30 June 2003; accepted 14 September 2003
In principle there is a clear divide between science and technology. In practice there isn’t. In principle, while practical
inventions can be patented, scientific findings can’t be. In practice, increasingly scientific findings are being patented. The
argument of this paper is that this is bad for the advance of science and for the advance of technology. However, because of
the blurry lines, it will not be easy to deal with. The paper lays out a strategy that at least has some promise.
© 2003 Elsevier B.V. All rights reserved.
Keywords: Market; Commons; Capitalism
1. Introduction
tization of the scientific commons up to now has
been relatively limited, there are real dangers that
Modern Capitalism has proved a remarkably pow-
unless halted soon important portions of future sci-
erful engine of technological progress. Most of the
entific knowledge will be private property and fall
attention to its workings has focused on the business
outside the public domain, and that could be bad
firms and entrepreneurs, operating in a market set-
news for both the future progress of science, and for
ting, who are the central actors in developing and
technological progress. The erosion of the scientific
introducing new products and processes. At the same
commons will not be easy to stop. Here I want to call
time, it is widely recognized that the power of market
the alarm, and to suggest a strategy that has some
stimulated and guided invention and innovation often
is dependent on the strength of the science base from
But before I get on with this task, I need to clear
which they draw (Nelson, 1993; Mowery and Nelson,
some intellectual underbrush. A number of influen-
1999). This science base largely is the product of pub-
tial philosophers and sociologists of science have put
licly funded research, and the knowledge produced
forth a set of views, a theory, about the scientific
by that research is largely open and available for
enterprise that until recently has served well to pro-
potential innovators to use. That is, the market part
tect the scientific commons. However, this theory no
of the Capitalist engine rests on a publicly supported
longer is adequate to the task, because the way it char-
scientific commons.
acterizes the nature of the scientific enterprise does
The message of this essay is that the scientific
not fit modern perceptions and the reality. Also, under
commons is becoming privatized. While the priva-
this theory, it is hard to understand why privatization
and markets are encroaching on the commons, and if
∗ Fax: +1-212-864-4847.
they are, what is the matter with that. It is important,
E-mail address: [email protected] (R.R. Nelson).
therefore, to scrutinize that theory.
0048-7333/$ – see front matter © 2003 Elsevier B.V. All rights reserved.

R.R. Nelson / Research Policy 33 (2004) 455–471
A key element of the theory is that, outside of indus-
practical payoffs from scientific research were not pre-
try, the work of scientists is and should be motivated
dictable, but largely came about through serendipity,
by the search for understanding, and that the practi-
and that the allocation of scientific resources should
cal payoffs that often come from successful research
not be guided by anticipation of particular practical
are largely unpredictable. Bush (1945) is one among
payoffs, but rather by the informed judgments of sci-
many proponents of public support of science who put
entists regarding the most important problems to work
forth this theme, and argued that it would be a mistake
on. Keeping scientific findings in the public domain,
to look to likely practical payoffs as a guide to where
with reward to the scientist being tied to the acclaim
scientific funds should be allocated. Serendipity is the
of his or her fellows, along with public funding of re-
reason why scientific research often has practical pay-
search based on peer review of the scientific promise
off, and the chances of serendipity are greatest when
of the proposal and the scientist, then would seem an
bright and dedicated scientists are free to attack what
important part of an incentive and control system for
they see as the most challenging scientific problems
fostering productive science (for a discussion along
in the way they think most promising.
these lines, see Dasgupta and David, 1994).
For this reason, decisions regarding what questions
However, the notion that academic scientists have
to explore, and the evaluation of the performance of
no idea and do not care about the practical problems
individual scientists and broad research programs,
that their research might illuminate never has been
should mostly be in the hands of the scientists work-
fully true. In this era of biotechnology it is obvious,
ing in a field. Indeed, for the government or the mar-
if it was not before, that both the funders and the
ket to intrude too much into how scientific research
undertakers of research often have well in mind the
resources are allocated would be to kill the goose that
possible social and economic payoffs from what they
lays the golden egg. In the terms used by Polanyi
are doing. But if in fact, much of scientific research
(1967), society should appreciate and protect “The
is consciously aimed, at least broadly, at problems
Republic of Science”.
the solution to which can have major, and broadly
An associated belief or ideal is that the results
predictable, practical value, what is the case against
of scientific research are and should be published
harnessing market incentives to the undertaking of
and otherwise laid open for all to use and evaluate.
research and to the use of research results? In partic-
As Merton (1973) argued, the spirit of science is
ular, why should the privatization of these kinds of
“communitarian” regarding access to the knowledge
research results be viewed as a problem?
and technique it creates. All scientists are free to test
The case for open scientific knowledge clearly
the results of their fellows and to find them valid or
needs to be reconstructed recognizing explicitly that
not supported, and to build on these results in their
much of scientific research in fact is oriented towards
own work. Because the results of scientific research
providing knowledge useful for the solution of prac-
are laid in the public domain for testing and further de-
tical problems, that the applications of new scientific
velopment, the bulk of scientific knowledge accepted
findings often are broadly predictable, and that this
by the community is reliable (as Ziman (1976) has
is why control over scientific findings in some cases
emphasized) and scientific knowledge is cumulative.
is financially valuable property. I think there is a case
These are basic reasons why the scientific enterprise
for keeping basic scientific knowledge open, even un-
has been so effective as an engine of discovery. And
der these conditions. To privatize basic knowledge is
economists often have argued that keeping science
a danger both for the advance of science, and for the
open is the most effective policy for enabling the
advance of technology. I will develop my argument
public to draw practical benefits from it.
as follows.
My argument in this essay is that the part of the the-
Section 2 is concerned with how technological ad-
ory about good science that stresses the value of open
vance draws from science. I already have tipped my
science is basically correct, but is in danger of being
hand. Without denying the role of serendipity, I will
forgotten, or denied. A good share of the reason is
argue for the most part science is valuable as an in-
that, as originally put forth, this part seemed a natural
put to technological change these days because much
consequence of the other aspects of the theory: that the
of scientific research is in fields that are oriented to

R.R. Nelson / Research Policy 33 (2004) 455–471
providing knowledge that is of use in particular areas.
This perception of how the modern science system
These are the scientific fields that Stokes (1996) saw
actually works has eroded the notion that it is impor-
as in “Pasteur’s Quadrant”, where the research aims
tant to keep science open. My argument is that this is
for deep understanding, but the field itself is oriented
a serious mistake.
towards achieving practical objectives, like improv-
While perceptions of possible applications of re-
ing health, or achieving better understanding of the
search are not as vague as proposed in the earlier
properties of materials, or achieving a powerful the-
rhetoric about serendipity, the actual paths to applica-
ory of computing. I acknowledge that this is a some-
tion of apparently promising scientific discoveries are
what more expansive view of what science is than that
in fact very uncertain. Understandings that come from
contained in earlier characterizations of a “Republic
science seldom lead immediately or directly to the so-
of Science”. But in fact a large fraction of what is
lution of practical problems. Rather, they provide the
well recognized as science always has been undertaken
knowledge and the tools to wrestle with them more
with practical objectives in mind or not far out of mind.
effectively. I propose that for just this reason, that the
Stokes’ example of Pasteur is apt. And this fact is vital
findings of basic science set the stage for follow-on
to keep in mind when trying to understand how sci-
applications work, for society to get maximal benefit
ence operates, and the controversy this paper is about.
from its support of basic science requires that there
In Section 2, I discuss the rise and erosion of the
be open access to scientific research results. Open
idea that public support of open science is warranted
access permits many potential inventors to work with
because the expected returns are high but the areas of
new knowledge. Privatization closes off access to
return are so uncertain that market mechanisms will
only those whom the owner allows to make use of it.
not suffice. I begin by briefly reviewing the ideolog-
This is why some of the recent developments are so
ical and political debates that occurred after World
War II that led to broad consensus regarding the value
In Section 4, I discuss the current situation and the
of public support of open autonomous science. As I
dangers in more detail. Then I turn to a number of
noted that rhetoric stressed that the payoffs from sci-
measures that I believe have some promise as attacks
ence were almost completely unpredictable, and thus
on the problem.
the allocation of funds to science should not be in-
fluenced by perceptions of social needs. The publicly
supported science system that actually developed was
2. The coevolution of practice and
in fact much more oriented to facilitating making
progress on important practical problems than that
rhetoric allowed, and this is now obvious.
Virtually everybody these days appreciates that the
I do not want to argue that most academic re-
power of modern technological innovation depends
searchers working in, for example, the bio-medical
to a considerable extent on its ability to draw from
sciences define their goals as dealing with particular
modern science. But there is little general understand-
diseases. Much of the most important work in such
ing, and some quite wrong beliefs, about the nature
fields is quite fundamental in nature, in the sense that
of the science–technology links. Understanding these
it explores basic processes and phenomena, without
correctly is a precondition, I believe, for having an
a clearly defined specific practical objective in mind.
effective discussion about what public policy towards
However, the fundamental questions and appealing
science ought to be. This certainly is so regarding the
lines of research in sciences in Pasteur’s Quadrant are
current controversies about patenting in science. Thus,
strongly influenced by perceptions of what kind of
this section discusses what scholars studying techno-
knowledge is relevant to problem solving in a field.
logical advance know about these issues.
Thus, one of the reasons why cell biology now is
Technologies need to be understood as involving
such a fashionable field is belief that basic under-
both a body of practice, manifest in the artifacts
standing won here might just unlock the cancer puz-
and techniques that are produced and used, and a
zle, or enable us to understand better how receptors
body of understanding, which supports, surrounds,
and rationalizes the former. For technologies that

R.R. Nelson / Research Policy 33 (2004) 455–471
are well established, an important part of the body
occasional major leaps that radically transform best
of understanding supporting practice generally is
practice, for the most part technological advance is cu-
grounded in the empirical experience of practition-
mulative. And scholars of technological advance also
ers regarding what works and what does not, things
have generally stressed that the advanced technologies
that sometimes go wrong, reliable problem solving
of a given era almost always are the result of the work
methods, etc. However, in recent times, virtually
of many inventors and developers. Technological
all powerful technologies have strong connections
advance is a collective, cultural, evolutionary process.
with particular fields of science. These connec-
The proposition that technological advance is an
tions, of course, are central in the discussion of this
evolutionary process in the above sense in no way
denies, or plays down, the often extremely powerful
There is a widespread belief that modern fields of
body of understanding and technique used to guide
technology are, in effect, applied science, in the sense
the efforts of those who seek to advance it, at least in
that practice is directly drawn from scientific under-
modern times. A strong body of scientific understand-
standing, and that advancing technology essentially
ing of a technology serves to enlarge and extend the
is a task of applying scientific knowledge to achieve
area within which an inventor or problem solver can
better products and processes. This task requires sci-
see relatively clearly and thus make informed judg-
entific expertise, but in most cases is relatively routine
ments regarding what particular paths are promising
once the target is specified. Indeed, in his Capital-
as solutions, and which ones are likely to be dead
ism, Socialism, and Democracy, Schumpeter (1942)
ends. Also, the sciences and engineering disciplines
argued that by the mid twentieth century that was
provide powerful ways of experimenting and testing
largely the case, and the kind of competition among
new departures, so that a person or organization who
firms that had over the prior century made capitalism
commands these can explore the merit of designs
such a powerful engine of progress no longer was
without going to full-scale operational versions. Thus,
necessary. With strong science, technological advance
strong science enables the process of designing and
could be planned. Schumpeter’s views were in accord
inventing to be more productive and powerful than it
with those of many prominent scientists of his day,
would be were the science base weaker.
and today. Yet, careful studies of how technological
However, it does not change the fact that the process
advance actually proceeds in this modern era clearly
of advancing the technology remains evolutionary.
show that the process remains unplanable in any de-
Strong science provides tools for problem solving,
tail, and competitive exploration of multiple paths
but usually in itself does not solve practical problems.
remains an essential part of it (see e.g. Rosenberg,
If anything, strong science increases the advantages
1982; Nelson and Winter, 1982).
to society of having many competent actors striving
Virtually all empirically oriented scholarly accounts
to improve the art.
of how technology progresses have highlighted that the
The connections between the “body of practice” as-
process is evolutionary in the following senses (see e.g.
pect of a technology and the “body of understanding”
Basalla, 1988; Constant, 1980; Dosi, 1988; Metcalfe,
part need to be understood in this context. Virtually
1998; Mokyr, 1990; Nelson and Winter, 1982;
all modern technologies are supported by a strong
Petroski, 1992; Vincenti, 1990; Ziman, 2000). First,
body of science or science-like understanding that il-
at any time there generally are a wide variety of ef-
luminates how the artifacts and techniques employed
forts going on to improve prevailing technology, or to
work, provides insight into the factors that constrain
supersede it with something radically better. These ef-
performance and provides clues as to promising path-
forts generally are in competition with each other, and
ways toward improvement. But at the same time,
with prevailing practice. And the winners and losers
much of practice in most fields remains only partially
in this competition to a considerable extent are deter-
understood, and much of engineering design practice
mined through an ex-post selection processes. Second,
involves solutions to problems that professional engi-
today’s efforts to advance a technology to a consider-
neers have learned “work”, without any particularly
able extent are informed by and take off from the suc-
sophisticated understanding of why. Medical scien-
cesses and failures of earlier efforts. While there are
tists still lack good understanding of just why and

R.R. Nelson / Research Policy 33 (2004) 455–471
how certain effective pharmaceuticals do their work,
the development of a collection of specialized fields
and theories about that can change from time to time.
involved in agricultural research. Fields like pathol-
Technological practice and understanding tend to
ogy, immunology, and cardiology, grew up for teach-
coevolve, with sometimes advance of understanding
ing and research at medical schools.
leading to effective efforts to improve practice, and
All of these fields of science are in “Pasteur’s
sometimes advance in practice leading to effective
Quadrant”. Research done here often probes for quite
efforts to advance understanding. Thus, the germ
deep understanding. But the field as a whole, and
theory of disease developed by Pasteur and Koch,
broad programs of research in the field, are dedicated
by pointing clearly to a certain kind of cause, led to
quite explicitly to solving particular kinds of practical
successful efforts to get certain diseases (now known
problems, and advancing bodies of practical technol-
to be caused by external living agents) under control.
ogy. I have developed this story at considerable length
Maxwell’s theory of electromagnetism led to Hertz,
because in much of the writings on science, and the
Marconi, and radio. But in many cases, advances in
institutions governing science, these applied sciences
practice come first and lead to efforts to understand
tend to be ignored. However, in the US, Western Eu-
scientifically. Thus, the discovery by Shockley and
rope, and Japan, they account for the lion’s share of
his team at Bell Laboratories that a semiconduct-
the resources going into the support of science.
ing device they had built as an amplifier worked,
Popper (1989), Campbell (1974), Ziman (1976),
but not in the way they had predicted, led him to
Kitcher (1993), and other scholars of the advancement
understand that there was something wrong, or in-
of science have stressed that science is a system of
complete, about the theory in physics regarding the
knowledge. The test that guides whether new reported
electrical characteristics of semiconductors, which in
findings or theories are accepted into the corpus of
turn led to his own theoretical work, and a Nobel
accepted knowledge is “Is it valid” “Is it true?”. Pop-
Prize. Rosenberg (1996) has argued that a number of
per and his followers have argued that there can be no
the most challenging puzzles science has had to face
firm positive answer to that question. Ability to stand
have been made visible by or been created by new
up under attempts at refutation, or (probably more
technologies, and the puzzles of why they work as
commonly) for apparent implications to hold up when
they do.
they are explored, may be the best humans can do.
Much of the development of modern science should
But in any case, from this philosophical perspective,
be understood as the result of institutionalized re-
the quest in science is for understanding in its own
sponses to these challenges and opportunities. Quite
right. And there certainly is a lot of truth to this po-
often, specialized fields of applied science or engi-
sition as a characterization of the nature of scientific
neering developed out of the experience of more gen-
erally trained scientists working on the problems of
On the other hand, as Vincenti and others who have
a particular technology or industry. Thus, the field of
reflected on the similarities and differences between
metallurgy came into existence as chemists worked
technological and scientific knowledge have argued,
on problems of quality control in the rapidly grow-
the central test for technological knowledge is “is it
ing steel industry (Rosenberg, 1998). As the industries
useful”. Technological knowledge is part of a cultural
producing chemical products expanded, chemical en-
system that is concerned with achieving practical
gineering developed as a field of research, as well as
ends, rather than knowledge for its own sake. The
teaching. The physics of mechanical forces long had
objective is to get something that works, or works
been useful for civil engineers designing buildings and
better, and “understanding” is important only in so
bridges. But with the new physics of electricity and
far as it helps in that effort.
magnetism, a whole new set of science-based indus-
However, the selection criteria for new science and
tries was launched. As complex electrical “systems”
for new technology cannot be kept sharply separate for
came into place, the new field of electrical engineer-
sciences in Pasteur’s Quadrant. In these fields, an im-
ing grew up. Later on, the invention of the modern
portant and often stringent testing ground for science
computer would spawn the field of computer science.
is provided by those who think they see how it might
Stronger knowledge in chemistry and biology led to
be applied in practice. And failure to understand why

R.R. Nelson / Research Policy 33 (2004) 455–471
something works is a strong motivation for scientific
tremely important. This enables there to be at any time
a significant number of individuals and firms who pos-
By far, the lion’s share of modern scientific re-
sess and can use the scientific knowledge they need in
search, including research done at universities, is in
order to compete intelligently in this evolutionary pro-
fields where practical application is central in the def-
cess. The “commitarianism” of scientific knowledge
inition of a field. And, not surprisingly, these are the
is an important factor contributing to its productivity
fields on which efforts to advance technology mostly
in downstream efforts to advance technology.
draw. Two recent surveys (Klevorick et al., 1995;
Cohen et al., 2002) have asked industrial R&D ex-
ecutives to identify the fields of academic research
3. The governance of public science
that contributed most to their successes in R&D. The
fields they listed were exactly those discussed above.
World War II and the period just after marked
The most recent of these studies (Cohen, Nelson,
something of a watershed in broad public and po-
and Walsh) also asked about the kind of research out-
litical recognition of the important role that public
put that was most valuable to industry, and the most
science plays in technological progress, particularly
important pathways through which industry gained
in the US and the UK. To be sure, much earlier
access. Contrary to much of the current discussion,
visionaries like Francis Bacon had argued for sup-
prototype technologies were not rated an impor-
port of science as a means through which societies
tant output of academic research for most industries
could progress materially. Scholars like Price (1962),
(biotechnology is an exception), but rather general
Hart (1998), and Guston (2000) have described the
research results and research techniques (and even in
earlier history of debate about science policy in the
biotechnology these kinds of research outputs were
US. But, it was the World War II experience, where
rated as useful much more often than prototypes).
government-supported and -focused R&D was so
Relatedly, in most industries the respondents reported
successful both in the development of weapons that
that the most frequent use of university research re-
won the war, and in the development of medical ca-
sults was in problem solving in projects, rather than
pabilities that greatly reduced casualties both from
in triggering the initiation of projects.
wounds and from infectious diseases compared with
In most industries, the respondents said that the
earlier wartime experiences, that gripped the pub-
most important pathway through which people in in-
lic attention. The title of the Vannevar Bush report
dustry learned of and gained access to what was com-
(1945) advocating a major postwar program in the
ing out of public research was through publications,
US of support of science caught the spirit: “Science,
and open conferences. Put another way, today industry
the Endless Frontier”.
gets most of its benefit from academic science through
In both the US and the UK the discussion about the
open channels. In their more narrowly focused but
appropriate postwar role of public science was struc-
more detailed study of the pathways through which
tured and constrained, for the most part, by recognition
research results of the MIT departments of mechanical
of the central role of companies with their own R&D
and electrical engineering get to industry, Agrawall
capabilities in the process of technological advance;
and Henderson (2002) arrive at a similar conclusion.
the point of view there was implicitly Schumpeterian.
I want to conclude this section by again stress-
While there were exceptions, the discussion was not
ing that in all the fields of technology that have
about contesting that role. Rather, the focus was on the
been studied in any detail, including those where
system of public science, done in universities and pub-
the background science is very strong, technological
lic laboratories, that was separate from the corporate
advance remains an evolutionary process. Strong sci-
system but strongly complementary, and which needed
ence makes that process more powerful, but does not
public support. The argument of those who advocated
reduce the great advantages of having multiple paths
stronger government support was that this would make
explored by a number of different actors. From this
the overall system of innovation more powerful.
perspective, the fact that most of scientific knowledge
In both the UK and the US, the debate about the
is open, and available through open channels, is ex-
governance of public science squared off along much

R.R. Nelson / Research Policy 33 (2004) 455–471
the same lines. In the UK, J.D. Bernal (1939), a
win the war. Hence, the emphasis on serendipity, and
distinguished physicist, and a socialist, argued for a
the unpredictability of areas of potential payoff. It is
government program in which the allocation of pub-
almost certain that both men knew well that much
lic funds to science would be strongly guided by the
of scientific research was not of this kind, but rather
weighing of social needs, and the support program as a
was in fields where perceptions of practical problems
whole would be closely monitored by the government.
played a significant role in defining the broad agenda,
To this point of view Michael Polanyi, a distinguished
if not the short run priorities of resource allocation.
philosopher of science, took strong exception, advo-
However, the rhetoric of Polanyi and Bush obscured
cating a largely self-governing “Republic of Science”
the fact that most of science is in Pasteur’s Quadrant.
(1967), which would be publicly funded, but in which
It is not surprising, therefore, that in both the US
the scientific community itself would set priorities
and UK it turned out that mission-oriented agencies
and decide on what was good science.
became the primary government supporters of basic
In the US, Bush’s manifesto “Science, the Endless
research Thus in the US the Department of Defense
Frontier” argued strongly for a self-governing scien-
funded basic work in computer and materials sci-
tific community, but with national priorities playing a
ence, and electrical engineering. The Atomic Energy
role in setting broad research directions, at least in cer-
Commission (later the Department of Energy) has
tain areas. In particular, national security and health
had principal responsibility for funding high energy
were singled out as areas where the overall research
physics. The National Institutes of Health became
budget and broad research priorities needed to be
the primary funder of university research in the
made through political and governmental processes.
bio-medical sciences. The National Science Founda-
But given the funding within those broad areas, the
tion, the only significant research funding agency in
scientists themselves were to have basic discretion
the US without a mission other than support of sci-
for devising the research programs they thought most
ence, always has been a small supporter relative to
appropriate. Government non-scientists were not to
the mission-oriented agencies. The lions share of the
meddle in this. Regarding the role of public science
research done in the US, funded by government, and
in supporting economic progress more broadly, Bush
undertaken in universities and public laboratories, is
saw the government’s role as supporting basic re-
in fields in Pasteur’s Quadrant.
search, with the science system self-governing, both
This fact both removes the puzzle of why science
with respect to identification of the broad fields of
has contributed so much to technological advance,
greatest promise, and the details of allocating funds
and enables one to understand better why Vannevar
and carrying out research.
Bush (and most of his science trained followers writ-
There is no question but that, like Polanyi’s re-
ing about science policy) had such strong faith in the
sponse to Bernal, Bush’s articulation of a basically
ability of the scientific community to steer their efforts
self-governing community of science was put forth in
in socially productive directions. But this recognition
good part to counter, to block, proposals for a postwar
also signals that the lines between basic science and
publicly supported science system that would involve
applied science are fuzzy not sharp. And it raises the
much more political and government control of the
question of where the publicly supported Republic of
allocation of resources. Senator Harley Kilgore took
Science ought to leave off, and the market begin. It
much the same position as did Bernal in the UK.
is fair to say that for the most part the postwar de-
Bush believed this would destroy the creativity and
bates were somewhat ad hoc about this. Thus, Bush
power of science, and it would be far better to have
recognizes a central role for market organized and
the top scientists running the show.
induced R&D, and saw public science as providing
There also is no question but that Polanyi and Bush
inputs to that market system, but being separate. But
felt it of extreme importance that government support
he provided little in the way of coherent argument
fields like theoretical physics and mathematics, where
about where the one stopped and the other began.
perceptions of potential practical payoff have little to
Indeed, despite its obvious importance, outside of
do with the way the fields unfold, yet which provided
economics, this question has aroused little analytical
important knowledge and technique that helped to

R.R. Nelson / Research Policy 33 (2004) 455–471
Economists have grappled with the question of
technique can be withheld from certain researchers,
the appropriate spheres of government activity in the
they may be effectively barred from doing productive
science and technology system using two theoretical
R&D in a field.
concepts: externalities, and public goods. The exter-
Now the fact that something is non rivalrous in
nalities concept is about benefits (and costs) of private
use does not mean that its use cannot be restricted.
economic activity that those who make the relevant
However until relatively recently, it was broadly
decisions do not see as benefits (or costs) to them.
assumed that it was difficult to restrict access to
Here economists have highlighted the “spillovers”
scientific knowledge. Certainly scientific knowledge
from industrial R&D, information and capabilities
could not be patented. This effectively took science
created by a firm’s efforts to create better products
outside the domain where market incentives could
and processes that it cannot fully capture, and hence
work. Indeed, the presumption that the returns to
which benefit other firms, including competitors. In
scientific research could not be appropriated was a
general, the analyses by economists oriented towards
central part of the argument why public funding was
the externalities from R&D have not served as a base
for arguments for a domain of public science, but
However, over the last quarter century there have
rather for arguments that industrial R&D in some
been two key developments that have challenged this
instances should be encouraged by favorable tax
view of basic science. First, the courts have ruled
treatment, and perhaps subsidies of various kinds to
that at least some of the results of basic research
reduce private costs. Indeed, the policy discussion
can be patented. And about the same time that the
proceeding under the conception that research yields
implications of these rulings were becoming evident,
externalities naturally tends to be pulled towards de-
Congress passed the Bayh–Dole act of 1980, which
vising policies that will make the results of R&D
strongly encouraged universities to take out patents
more proprietary, less public. An important part of
on their research results where they could, on the
the current policy discussion in fact is oriented in just
basis of a (not well supported) argument that this
this way.
would facilitate firms who could make practical use
The public good concept of economists is much
of the results to do so under a protective license (for
more directly relevant to analysis of the appropri-
a detailed account, see Eisenberg, 1996). The first
ate domain of public science, or at least the range
of these developments significantly increased the in-
where “communalism of knowledge” should apply.
centives for for-profit firms to engage in the areas of
For our purposes here, the most salient aspect of
basic research where the results can be patented, and
the economists’ public good concept is that a public
to try to make their living licensing patented research
good is “non-rivalrous in use”. By that it is meant
results to other firms that can make use of them. The
that, unlike a standard economic good, like a peanut
second has brought about profound changes in the
butter sandwich, which either you or I can eat but not
way universities give access to their research results.
both (although we can split it), a public good can be
As a result, important areas of science are now much
used by all of us at the same time without eroding the
more under the sway of market mechanisms than
quality for any of us.
used to be the case. And in particular, in some im-
Knowledge is a canonical case of something that
portant fields of science important bodies of scientific
is non-rivalrous in use in this sense, and this is not
understanding and technique now are private property
a proposition conjured up by economists. The notion
rather than part of the commons.
that I can tell you what I know, and then you will
A widespread reaction is “So what is the problem
know it, and I will too, almost surely has been widely
with that?”. There is a strong presumption these days
understood by sophisticated persons for a long time.
that if market organization can and will do a job,
There is no “tragedy of the commons” for a pure pub-
that obviously is a good thing. From this point of
lic good like knowledge. And to deny access, or to
view, the main argument that needs to be made for
ration it, can result in those denied doing far less well
government support of basic research is that the long
than they could if they had access. In the case in point,
run benefits to the society are high, and that for-profit
if access to certain bodies of scientific knowledge or
firms have little incentive to do much of it because of

R.R. Nelson / Research Policy 33 (2004) 455–471
the difficulties in establishing property rights, and the
the basic research scene, and in some case for-profit
long time lags and uncertainties involved in moving
business firms have explored paths that the academic
from research results to commercial product. If these
community snubbed.
barriers to market organization are lowered for some
But on the other hand, a careful reading of impor-
reason, let the market move in.
tant scientific controversies, for example the argument
I note that knowledge of an effective product de-
about the nature of combustion at the start of the 19th
sign or a production process, what customarily is
century, or of the nature of the genetic code, or of
considered as technological knowledge, shares with
whether the expansion of the universe is decelerating
scientific knowledge the property of being non rival-
or accelerating, shows the importance and the power
rous in use. Yet society relies largely on the market
of a public science system where by and large all par-
to induce R&D aimed at creating new products and
ticipants have access to much the same facts, and the
production processes, and there is little dispute that
debates about whether new proposed facts or theories
granting patents on product and process inventions is
are valid are open to all working in a field. One cannot
reasonable social and economic policy. So why not
come away from reading Horace Judson’s (1969) The
allow patents on the stuff of basic science, if that will
Eight Day of Creation, a history of the development
induce the market to move in?
of molecular biology as a field of science, without re-
My response is that the outputs of scientific re-
specting the power of open science to progress.
search almost never themselves are final products,
This is equally true for sciences that are strongly in
or even close, but have their principal use in further
“Pasteur’s Quadrant”. Porter (1997) history of med-
research, some of it aimed to advance the science
ical knowledge and practice, The Greatest Benefit to
farther, some to follow leads that may enable a useful
Mankind (1997) gives case after case where progress
product or process to be found and developed. But in
was made through a system where researchers were
both cases, the latter as well as the former, there is
free to try to replicate or refute the arguments and
considerable uncertainly about the best paths to pur-
findings of others.
sue. Progress calls for a number being explored. My
While my argument above has focused on the ad-
concern is not with patents on the outputs of scien-
vantages of an open science for the advancement of
tific research that are directly useful or close to that,
science, much of my discussion in Section 2 was con-
so long as the scope of the patent is limited to that
cerned with developing a case why open science is
particular use. It is about not hindering the ability of
important to technological progress. These arguments
the scientific community, both that part interested in
of course are mutually reinforcing. Keeping the body
advancing the science farther, and that part interested
of scientific knowledge largely open for all to use, in
in trying to use knowledge in the search for useful
the attempts to advance science, and in the attempts
product, to work freely with and from new scientific
to advance technology, is in my view an extremely
important matter. Its importance is not recognized ad-
I do not know of a field of science where knowledge
equately in the current discussions.
has increased cumulatively and, through cumulative
I want to conclude this section by putting forth three
advance, dramatically, that has not been basically
views on what should be done about the encroach-
open. It is easy to argue that scientists never have
ment of proprietary property claims into what had been
fully followed the canons of science identified and
the domain of public science. The first is to cede the
laid out by Robert Merton: universalism, communitar-
contested turf. If research findings can be patented,
ianism, disinterestedness, and organized skepticism.
accept and embrace that. If universities can patent
Scientists are well known to keep their work secret
their results, and limit access to the highest bidder,
until they are ready to publish. There certainly is a lot
fine. And welcome the presence of private firms mo-
of self-interest, opportunism, hostility, and downright
tivated to do research by the lure of patents, and con-
deviousness and lying that one observes in the histo-
trol of subsequent work in a field, or royalty incomes.
ries of the progressive sciences. A scientific paradigm
Indeed, these developments diminish or even elimi-
held by the elite in a field can hold intellectual
nate the need for publicly funding of certain fields of
tyranny. It is valuable to bring new organizations into

R.R. Nelson / Research Policy 33 (2004) 455–471
The second is to coexist and compete on the con-
search tools” (see NRC, 1997) where research tech-
tested terrain. This is pretty much the policy that de-
niques of widespread use in a field, materials that are
veloped regarding research on the human genome. The
inputs to a wide range of research endeavors, or key
argument here is that publicly supported research, and
pathways for research (like the use of a particular re-
keeping open the results of that research, provide use-
ceptor), are patented, and the patent holder aggres-
ful competition to private research, even if some pri-
sively prosecutes unlicensed use or reserves exclusive
vate firms do not like the competition (Eisenberg and
rights to further research using the tool. The second,
Nelson, 2002).
highlighted recently by Heller and Eisenberg (1998)
A third position is to resist and try to roll back the
is focused on contexts where development of or ad-
invasion of privatization. This point of view sees that
vance towards a useful product or technique may in-
invasion not only as probably undesirable, but also
volve transgressing on several patents held by different
as something that is occurring under a given set of
policies, which can be changed. Thus if the movement
The latter problem, that of the need to assemble a
of patentability upstream into the sciences, together
large number of permissions or licenses before being
with the expectations under the Bayh–Dole act, are
able to go forward, was found by the Walsh, Arora,
leading to for profit companies engaging in research
Cohen interviews and case studies not to be particu-
to identify the genetic code, and to the patenting of
larly important, as of yet. Regarding research tools, a
that code by them and by universities operating under
number of the more important general purpose ones
public funding, maybe patent law and practice, and
are available to all who will pay the price, and while
Bayh–Dole, need to be revised.
in some cases there were complaints about the price,
Above I have given my reasons for rejecting the first
at least they were available.
position. My position on this is a combination of the
On the other hand, the study did identify a num-
second and third. I believe it important to preserve as
ber of instances where the holder of a patent on an
much of the commons as possible. However, doing so
input or a pathway (for example a receptor) that was
will not be easy.
important in a particular field of exploration did not
widely license, and in some cases sought to preserve
a monopoly on use rights. It is clear that in a number
4. The importance of protecting the scientific
of the cases, the patented finding had been achieved
through research at least partially funded by the gov-
ernment. This policy well may have been reasonable
The major expansion of patents into what used to
from the point of view of the patent holders. But the
be the realm of science is well documented I am per-
burden of this paper is that it is not good from the point
suaded that there is enough of a potential problem here
of view of society, seeking to maximize the benefits
to call the alarm. However, I confess that the evidence
of publicly funded research.
that there already is a problem, that access to scientific
The authors of the study take a cautious position re-
research results having high promise of enabling solu-
garding the implications of their findings. I find them
tion of important practical problems is being sharply
sufficient evidence to indicate that there is a real prob-
limited by patent holders, presently is very limited.
lem here, or there will be soon, and it is time to think
The most detailed study is by Walsh et al. (2002).
about what can be done to contain it.
This study involved interviews with a number of re-
There are two broad policy arenas that bear on
searchers in the biomedical field, asking about whether
this issue, to which I want to call attention here.
their research had been hindered by patent rights that
One is intellectual property rights law. The second
blocked access to certain paths they wanted to explore.
is the policies of universities and public laboratories
Scholars studying this potential problem have iden-
regarding their research findings, and government
tified at least two different kinds of situations where
policy regarding the university research it funds. My
the presence of patents can hinder research (for a gen-
discussion below is oriented to what is needed, in my
eral discussion, see Merges and Nelson, 1990). One
view at least, to preserve an appropriately wide area
of these is the problem caused by patents on “re-
of public scientific knowledge.

Document Outline

  • The market economy, and the scientific commons
    • Introduction
    • The coevolution of practice and understanding
    • The governance of public science
    • The importance of protecting the scientific commons
      • Can we protect the republic of science through patent law?
      • Will universities come to the defense of the scientific commons?
    • Acknowledgements
    • References