The relationship between one's work as a scholar and
one's personal concern for--indeed, commitment to--a particular
political agenda is an issue that might well have worried many classical
liberal scholars. In this article, I seek to shed some light on this
The problem may arise in various ways. We may be struck, for
example, by the gap between certain scholars' commitment to liberty
and the liberty for which they have provided really telling arguments.
Alternatively, we ourselves might have received paternalistic advice
from a dissertation adviser, a department chairman, or a dean that to
get on in the academic world, we should put our personal agenda--the
things about which we care deeply--to one side and, to phrase the matter
bluntly, should make our mark instead by contributing to the
"normal science" of our day (Kuhn 1962). It is no consolation
to discover that those who occupy other positions in the political
spectrum face the same sort of problem (Jacoby 1987).
I am concerned here with a specific aspect of the issue: What are
we to make of a commitment to liberty given that our claims to knowledge
are fallible and that we are always hostage to empirical argument and
philosophical contestation? Must we retreat into what one might call a
cautious accountant's view of political philosophy in which a
ringing commitment to liberty and an enthusiastic espousal of commercial
society and voluntary activity are replaced by a statement that perhaps
in some circumstances human freedom might be a good idea? Obviously, if
writers such as Ludwig von Mises, F. A. Hayek, and Murray N. Rothbard
had held such a view, they probably would not have made much impact on
But what of human fallibility and, in light of it, the problems of
enthusiasm about values? One response might be to deny that we should be
concerned about fallibility. Indeed, one way of reading the work of
Mises and Ayn Rand and some of Hayek's methodological writings
might support such a response.
This approach, however, is mistaken. In very broad terms, those who
take it are beguiled still by what was historically a misunderstanding
of the epistemological status of Euclidian geometry. This
misunderstanding offered deductive argument to striking conclusions from
what seemed self-evident premises. Not only did it lead to the adoption
of the axiomatic form of proof in other areas, but it led also to a more
general quest for foundations of knowledge to serve as the basis for
such an enterprise.
I do not deny that certain things seem undeniable--for example,
that I am having certain experiences at the moment. It certainly appears
to me that I am using a word processor, and so on. Moreover, we cannot
imagine anyone's wishing seriously to contest many kinds of claims
about the world and about morality. To deny certain assumptions--about
logic, about the conditions relating to argument, and so forth--would
lead us into self-contradiction. We might wish to explore just what
claims of this kind can be made and how well they stand up to skeptical
questioning; indeed, those of us with an interest in philosophy have
doubtless spent many enjoyable hours in such exploration. Here, however,
I shall not engage in such discussion but rather shall comment about
what we can expect to get out of it.
In the light of what we know about valid inference, the discovery
of seemingly undeniable truths does not do us much good. If a claim is
problematic, we do not gain much from the knowledge that it follows
validly from something else that seems more obvious. All such a
derivation does is to inform us that, in fact, our premises were not as
obviously true as they had seemed, for our very derivation has informed
us that they turn out to contain some problematic content--indeed, the
very thing from which we started! Accordingly, although obvious truths
and certain things that we cannot deny without fear of
self-contradiction surely exist, they cannot help us much in convincing
others of the correctness of other claims they are currently contesting.
In our situation--the human situation--our views about the world
and our substantive ideals are always likely to be on the line. Our
ideas always stand open to criticism, and we may discover in the light
of dialogue with others that whenever we deal with significant issues,
ideas that we had assumed to be uncontroversial are in fact problematic.
This condition shapes our problem about values and our commitment to
them in our position as scholars in the public realm. How can we enter
the public realm of scholarship with the kind of commitment that writers
such as Mises and Hayek showed to liberty and to free markets, given
that such a commitment remains always open to challenge?
I suggest that we treat our values as constituting the core of a
research program (Shearmur 1991, 1996, chap. 1). This program typically
would consist of matters that we find morally attractive and rationally
compelling. At any time, we face open questions and issues that we
cannot immediately resolve, but our research program offers us
suggestions and theoretical tools to deal with those questions and
issues. In some cases, we will be on a roll: the sort of thing we have
to say will be clear, and showing the strength of our approach will be
exciting, challenging, and highly satisfying. In other cases, we will be
on the defensive: our critics or our own critical scrutiny of our
tradition may raise interesting and difficult problems. Our task is to
perform well over time, to show that our approach produces valuable
results and also that we can do better than those pursuing other
research programs. Along the way, we may find that the critical work of
historians, economic or social theorists, or philosophers poses
challenges to the core of our ideas, and some of us clearly should try
to meet such challenges. We can continue to work within our research
program even while admitting, at any point, that we have not yet
answered every question and not yet resolved every issue.
The big advantage of such an approach is that it offers a way of
combining commitment--indeed, passionate commitment--with an
acknowledgment of our fallibility. We are putting ourselves on the line
and explicitly opening ourselves and our program to the competition that
we regard as so necessary elsewhere. That we should take such an
approach is one of the key themes of my book Hayek and After (Shearmur
1996), in which I also apply it to an evaluation of Hayek's work.
Such an approach also offers us an important critical tool. It suggests
how we might appraise the extent to which a program or a particular
thinker within it is succeeding and then identify what would need to be
done in order to make the approach more successful.
Hayekian Liberalism as a Research Program
Hayekian liberalism as a research program has several aspects.
Spelling them out may help to clarify the approach and thereby to
promote its understanding and its criticism.
First, it involves taking a historical approach to Hayek's own
work, examining what Hayek's intellectual problems and concerns
were at any particular point, how he brought specific ideas to bear in
his attempts to solve those problems, and thus how his views changed
over time. This approach contrasts strikingly with much of what has been
written about Hayek's work, for many writers implicitly treat all
his work, although created over a long lifetime, as if it composed the
pieces of one huge jigsaw puzzle. Such writers seem to be striving to
show how all the pieces of the puzzle fit together.
Now, a scholar's views might be consistent throughout a
lifetime, but such consistency must be shown, not simply presumed. Such
consistency, however, would be odd, for it would suggest that the
scholar never had learned anything, never had discovered that he was
wrong about something. As scholars and as people interested in liberty,
we form and develop our views in dialogue with others. We encounter
different problems as we go through our lives; we must come to terms
with different theories and ideas as they are pressed upon us; and from
such grappling we acquire useful new lines of thought. Above all, the
world with which we are dealing is itself constantly changing and
offering us new challenges.
Indeed, the situation of the intellectual interested in liberty is
not unlike like that of the entrepreneur. We are concerned with truth,
to be sure, but we also have "customers" with problems and
concerns of their own, to which we need to respond. Just because one way
of solving a set of problems worked in the past does not mean that it
necessarily will suffice in solving the new problems that arise. I am
not advocating relativism but maintaining that we will be challenged
continually to demonstrate that classical liberalism is correct by
showing that we can resolve new problems and deal with new issues as
When reviewing papers for academic journals, I regularly find
Hayek's work being treated in an atemporal manner, as writers try
to integrate materials drawn from different periods and contexts.
Sometimes the outcome is ingenious. Often, however, the result seems to
me a bit beside the point and to resemble the work of old-fashioned
biblical scholars who would rather be committed to reconciling materials
that do not actually mesh rather than admit that the materials vary in
character. A simple way of putting my point would be to say that Hayek
changed his mind--that he came up with new ideas not always consistent
with the ideas he had embraced previously.
My second concern: What should we make of such changes? This topic
is important, for one mark of intellectually shoddy scholars is that
they shift the grounds on which they argue as they go along, without
acknowledging that they are doing so. If they do so, it is not enough to
discover what they are saying now. We also need to find where they
started and to track the shifts and developments in their views over
Such tracking in itself may be a devastating form of criticism. It
may reveal that without acknowledgment a scholar has shifted from his
initial views and has ended up conceding the key points that others were
pressing against him; he never concedes that he has shifted his position
at all but represents his current views as if they were the ones he had
always defended. An approach focused on research programs is concerned
not only with the historical context of a scholar's work, but also
crucially with its development over time. Identifying such changes plays
a key role in helping us to evaluate critically the views at issue.
It is not enough, however, simply to document that a scholar's
views changed--that, in effect, he did learn something. Without making
any changes in his thinking, he scarcely would have been able to respond
to changing situations.
How, then, do we distinguish between shifts that cause a problem
for the research program and those that do not? Two tasks are important
in this regard. First, we must identify fairly clearly the core elements
of our ideas and the kinds of solutions to problems that are compatible
with them. To depart from either is to make a significant concession,
which should not be made covertly. Second, we need to identify the
problems that require our attention and determine whether, when we
change our views, we can still resolve the problems we were resolving
previously on the basis of our earlier ideas.
Let me illustrate these abstract ideas in more concrete terms by
referring to a specific issue that arises in Hayek's work. It has
been argued that in his earlier writings Hayek worked within a broadly
general-equilibrium perspective (McCloughry 1984, x). Subsequently, as
is well known, he emphasized market processes and the dispersed nature
of knowledge. This move is a significant and exciting one, but if we
follow him in making it, an important question clearly arises: What can
we claim about the outcome of such market processes? Insofar as we seek
simply to understand the operation of the economy, the answer to this
question can be that the attributes of that outcome must be discovered.
If we are concerned with classical liberalism as a research program,
however, then it is not acceptable just to leave this matter open
because classical liberalism clearly takes the view that we can rely on
markets to produce certain kinds of desirable orderings in our social
affairs--that under conditions classical liberals favor for other
reasons, markets can produce coordination rather than chaos, well-being
rather than misery.
Of course, some classical liberals try to avoid recourse to such
arguments, insisting that we should concern ourselves with rights rather
than consequentialism, but in taking such a position, we encounter two
problems. First, it amounts to a retreat from the ideas with which
classical liberalism historically has been associated. Indeed, if in
such classical liberals' view, classical liberalism has nothing
substantive to say about people's well-being but can assure us only
that, whatever the outcome, people's rights would not have been
violated if they had complied with classical liberal recommendations,
then the first move that individuals with rights might make, if they
care about social coordination and material well-being, is to establish
an interventionist government to secure their well-being! Second, even
those of us who favor a rights-based approach should not overrate what
it can achieve. A number of powerful arguments can be offered for
recognizing rights as classical liberals have understood them, but it is
not clear how determinate such rights can be if we justify them by moral
argument alone. Just what, for example, do we acquire by mixing our
labor with hitherto unowned land, and exactly what constitutes such
mixing? What constitutes the imposition of an illegitimate externality
on others? Philosophy can contribute much to the resolution of such
questions, but the fine structure of such matters must arise from
conventional agreement, the choice of one convention rather than another
being made on the basis of what we expect to be the consequences of one
convention rather than another. Thus, not by accident did John Locke in
his Second Treatise mix moral and consequentialist arguments in his
discussion of rights. Classical liberalism historically flew with two
wings, the moral and the consequentialist, and we court disaster if we
try to make do with only one. Accordingly, an approach to Hayek's
work that focuses on his work in relation to classical liberalism as a
research program will note with interest any new developments, but it
also will ask: Is anything lost as these developments are being made,
and, if so, what is the significance of that loss?
Returning to Hayek, I suggest that we need to evaluate changes in
his views, such as his explicit shift to a market-process approach, in
two dimensions. In the first, we ask: Is the new move compatible with
the broad program within which he is working? Here Hayek's ideas
about market processes are clearly a valuable development of traditional
classical liberal ideas. In the second, we ask: If we follow him in this
move, to what extent can we still solve the problems that we must solve?
Here, the issues are more difficult and more interesting. They challenge
the classical liberals who follow Hayek to spell out just how the
consequentialist aspect of classical liberalism will look in the new
On this score, those who have been influenced by Hayek's ideas
have done them less than justice. What have they said about the
conditions under which we can expect coordination to result from the
activities of individuals acting on the basis of price signals in the
various institutions and situations in which they find themselves?
Hayek's ideas seem to offer opportunities for all kinds of fruitful
work, but that work must meet the challenge of identifying the specific
circumstances and institutional arrangements that facilitate the
successful operation of a free society.
Must we, however, continue to solve every problem that we started
with? Can't we drop problems as we go along without creating
difficulties for ourselves? Yes, but in doing so we must be careful.
Clearly, we can drop problems generated internally by theories that we
have discarded for other reasons. Thus, in terms of our current
theories, we may be unable to answer questions that concern our
colleagues or our earlier selves just because we now take the view that
these questions arise from assumptions that should be discarded.
Similarly, problems may become irrelevant because the social conditions
that gave rise to them change. We should not drop problems, however,
simply because we no longer can suggest how to solve them.
Consider, for example, why Hayek initially became concerned with
economics: it was, he tells us, because of the suffering that he
encountered in Vienna at the time. We ought not to allow such problems
simply to drop out of sight. We may wish to say that some such concerns
are mistaken--for example, a concern for "social justice" in
the sense that Hayek has criticized. We may urge such disregard because
the moral theories on which such ideas rest should themselves be
discarded. Or we may urge that such concerns be rejected because the
achievement of such states of affairs is incompatible with other things
of greater moral importance. But if we do so, we need to be frank about
what we are doing and why.
Simply to let a problem drop seems to me ill-advised, for two
reasons. First, it causes us to lose sight of something important in
evaluating how well we are doing. Confronting our initial aspirations
and our own subsequent history shows us where we currently stand.
Second, retaining our interest in the classic problems enhances our
ability to deal with others' concerns.
It is important for classical liberals to be able to speak to
others about their concerns, not least because the world is not a safe
place for liberty if we are a small minority whom others judge to have
nothing interesting to say about the things that matter to them.
Responding to the concerns of others does not mean that we must accept
those concerns unquestioningly. Indeed, in dialogue with them, we may
choose to take issue with the moral theories that inform their work or
with the empirical and theoretical claims they make or to argue that
resolution of the problems they emphasize is incompatible with the
attainment of other outcomes they should value more highly. We need to
be aware, however, of what others' problems and concerns are and to
consider them if we are not to condemn ourselves to irrelevance.
Here, again, Hayek is interesting because he was not initially a
classical liberal. Well might we consider the concerns that he had prior
to becoming a classical liberal and deal with those concerns in order to
have something to say to people who are not classical liberals because
they find appealing the same kind of concerns that the young Hayek found
Consider, for example, Hayek's concerns about poverty and
suffering in Vienna when he was a young man. Such matters pose a
significant issue (what is more, an issue that goes beyond the concerns
of subjectivism) because even in the wealthy Western world many people
worry about poverty and suffering, not because of their personal
exposure to such misfortunes but because they shape their political
voting and their support for certain government policies with an eye to
such matters. Although classical liberals ought to criticize the
failings of governmental programs, they also ought to show--and I here
echo a concern that Karl Popper expressed to Hayek when he read The Road
to Serfdom (Shearmur 1996, 64)--how their ideas respond to concerns such
as those of the early Hayek; otherwise, the argument is lost. I do not
mean to suggest that the classical liberal must embrace various
governmental welfare schemes, although it is worth noting that Hayek
himself consistently did so. Rather, on the one hand, we need to
explicate clearly what people can and cannot expect from a free,
market-based society in terms of well-being, and, on the other hand
(here following the path of scholars such as David Green [1986, 1993]
and David Beito ), we need to consider how nongovernmental means
can remedy problems such as poverty and insecurity. Green and Beito have
examined these topics historically and have suggested that we take
inspiration from the way in which problems were solved privately in the
past. Alternatively, we may approach such issues de novo. The vital
matter is that we do tackle them. Indeed, in my judgment, showing how we
can deal with issues of public concern by nongovernmental means
constitutes one of the most important subprograms for research in
contemporary classical liberalism.
More generally, we need to keep open dialogue with others in our
society. In part, this communication will take the form of criticism,
whether of their specific views or of their overall research programs.
In part, it will consist of learning from them about genuine problems.
The Future of Hayekian Liberalism
I start here with three difficulties, then turn, in conclusion, to
two more- positive themes.
First, with respect to the consequentialist side of a Hayekian
approach, we need to clarify our claims about the broad welfare
consequences that will flow from the operations of market-based
societies in various settings. In what circumstances will people guided
by prices and economic self-interest succeed in coordinating their
activities? In this regard, what difference do different, specific
The terms in which all this work must be done, moreover, must go
beyond the purely subjective and deal with such matters as Hayek's
concern about misery in the streets of Vienna. We will not be able to
pursue such subjects in the technical style of modern welfare
economics--no real shortcoming. Indeed, we will do ourselves and others
a favor if we write in a style as accessible as that of Smith's
Wealth of Nations or Mill's Principles of Political Economy,
speaking to the concerns of our contemporaries as they did to theirs.
Second, we need to clarify the issue of human freedom and why it
matters--in particular, why the freedom of each person should matter to
everyone else in just the way that classical liberals say that it does.
The classical liberal tradition historically was rooted in a religious
agenda. If one asked Locke why individual freedom matters and why we
should respect the rights of others, his answer went back to his
religious concerns. Further, this background shaped the details of his
argument. Broadly, the same account he gave of why we should respect
individual rights and people's property rights in the land on which
they have worked also supported what he said about an indigent's
right to welfare (in the First Treatise of Government, section 42). For
those of us who do not share Locke's religious beliefs or do not
wish to base our political arguments on religious values, it becomes
urgent to offer an alternative, secular account of the basis of rights
as they are understood in the classical liberal tradition.
Classical liberals need not only to find ways to make our ideas
compelling to our contemporaries, but also to make sure that those ideas
are coherent. If we simply talk about preferences, we must explain why a
third party should care whether or not we enjoy the freedom to get what
we want, especially if that party regards our wants as base. In this
context, it is better that we not talk about life plans and high-flown
goals if in fact all that our own argument entitles us to invoke is
bare--and perhaps to the observer, ignoble-- preferences. If we wax
eloquent about the importance of each individual as an end in himself
and about this self's having life plans and goals, we need to
explain why (and how) we resist the idea that such individuals should be
accorded positive rights, at least in terms of assistance in gaining the
capacities that would allow them to choose and to meet some minimal
goals. I am not arguing here that people should be accorded positive
rights, only that classical liberals need to make a coherent case for
resisting this idea--a case better than the one Hayek made, a case that
will be persuasive in a world in which the currency of public argument
Third, there is the problem of the state and of the limits of its
activities. In this respect, Hayek got into a mess. In his arguments
about socialist calculation, he made some devastating points against
central economic planning. He failed to clarify, however, what argument
he had against the pursuit of egalitarian values by means of state
action that respected the incentives of the marketplace and the rule of
law. Indeed, he himself favored a nonmarket welfare safety net, although
he never explained how generous its benefits should be. As a result, he
left himself vulnerable to an argument that his underlying views are
compatible with a generous form of welfare provision, such as that
favored by Raymond Plant, a British academic and current Labour member
of the British House of Lords (Plant 1984; Hoover and Plant 1989;
compare Miller 1989).
One line of argument that Hayek might have used is that welfare of
the sort Plant favored would be too costly in terms of the other things
that it rules out. In making such an argument, however, Hayek would face
difficulty because he is averse to arguing in quantitative terms. If we
accept his methodological argument that economics is limited to pattern
prediction, as opposed to the production of quantitative conclusions,
then we cannot enter the argument about what the level of welfare
provision should be.
To be sure, in some of his later work Hayek does offer pertinent
public-choice arguments about the problems of governmental provision.
Although I am broadly sympathetic to these arguments, they seem to
suggest that we should take a more minimalist view of the state than
Hayek himself would have accepted. Indeed, Hayek's own
public-choice arguments pose a challenge because in so much of the rest
of his writing he offers a positive agenda for governmental action. (On
Hayek's view of the role of the state and on some of his attempts
to deal with its relation to his ideas about liberty, see Shearmur
1997.) Here, classical liberals have a choice. We may say "away
with government." Then, however, we have to show in detail how we
would tackle the problems that classical liberals typically have granted
that government should solve (and that critics such as Stephen Holmes
 have pressed against us). Alternatively, we might offer an
account of just what the government should--and can--do, explaining how
it can accomplish those tasks but not the other tasks that welfare
liberals would have the government undertake. It is not acceptable,
however, for us to have a positive agenda for government action and at
the same time to deploy general arguments against the effectiveness of
government in order to dismiss our critics' proposals for more-
expansive government action.
Hayek placed great emphasis on the significance of the social
division of knowledge. He offered important insights into how markets
enable us to make use of socially disaggregated and even tacit
knowledge. His article "The Use of Knowledge in Society"
(1948, chap. 4) and other related writings are classics in this sphere.
However, he limited the scope of these arguments to activities in the
marketplace. He did not deal with the social division of knowledge
within the firm or within other organizations. Some interesting recent
work extends Hayekian approaches into the firm itself (Cowen and Parker
In another perspective, we may view knowledge as fallible theory
rather than as socially dispersed information. Some recent research in
management theory has proceeded along these lines. I myself recently
have been working on an extension of broadly "Austrian"
concepts into this field. Such efforts may create something of value to
complement Hayek's work, not only in terms of our understanding of
human organizations but also in its relation to human freedom (Shearmur
2000). At the same time, these efforts may suggest that some things can
be done that Hayek suggested could not be done, and thus they might be
viewed as undermining some of his arguments.
In one respect, Hayek's work on economic coordination
oversimplifies and is in an important way misleading. I have in mind the
picture he offers of individuals coordinating their activities solely by
the use of the price mechanism. Clearly, this sort of coordination has
fundamental importance, but market participants also coordinate their
actions in other ways. Companies increasingly cooperate with one
another, even with other companies that are in other respects their
competitors. In important ways, this cooperation involves the sharing of
information, not least of information about problems. Rather than
waiting to see if a particular desired product will be offered in the
marketplace, a business purchaser may go to its existing suppliers with
specific information about what the purchaser wants. Further, the
potential purchaser may work with those suppliers to help them improve
the quality of their output so that it meets the purchaser's
Important issues for classical liberals arise in such settings,
including issues of trust and reputation, some of which have been
treated in Daniel Klein's valuable collection Reputation (1997). We
need to study the interplay of the Hayek-style commodity markets, in
which buyer-seller coordination takes place purely in terms of price,
and the various arrangements for commercial cooperation in which buyers
and sellers mingle more intimately and communicate extensively in
detail. By failing to deal with such issues, we may be working with
models that in significant respects do not fit reality, and therefore we
may lay ourselves open to charges of irrelevance. If we do take note of
these issues, however, and integrate our treatment of them with
established Hayekian themes, we will be in a strong position to continue
the Hayekian research program--and this particular strand of argument
for classical liberalism--with greater confidence into the twenty-first
Acknowledgment: This article is based on the Friedrich Hayek
Lecture delivered by invitation at the Ludwig von Mises Institute 1998
Austrian Scholars Conference.
Beito, David. 2000. From Mutual Aid to the Welfare State. Chapel
Hill: University of North Carolina Press.
Cowen, Tyler, and David Parker. 1997. Markets and the Firm. London:
Institute of Economic Affairs.
Green, David G. 1986. Working Class Patients and the Medical
Establishment. London: Palgrave -MacMillan.
--. 1993. Reinventing Civil Society. London: Civitas.
Hayek, Friedrich A. 1948. Individualism and Economic Order.
Chicago: University of Chicago Press.
Holmes, Stephen. 1997. What Russia Teaches Us Now. The American
Prospect 8, no. 33 (July 1-August 1). Available at:
Hoover, Kenneth, and Raymond Plant. 1989. Conservative Capitalism
in Britain and the United States. New York: Routledge.
Jacoby, Russell. 1987. The Last Intellectuals. New York: Basic.
Klein, Daniel, ed. 1997. Reputation. Ann Arbor: University of
Kuhn, Thomas. 1962. The Structure of Scientific Revolutions.
Chicago: University of Chicago Press.
McCloughry, Roy. 1984. Editor's introduction to Money, Capital
and Fluctuations: Early Essays, by F. A. Hayek. London: Routledge and
Miller, David 1989. The Fatalistic Conceit. Critical Review 3, no.
Plant, Raymond. 1984. Equality, Markets, and the State. Fabian
Tract no. 494. London: Fabian Society.
Shearmur, Jeremy. 1991. Popper, Lakatos, and Theoretical Progress
in Economics. In Appraising Modern Economics, edited by M. Blaug and N.
de Marchi. Aldershot, U.K.: Edward Elgar.
--. 1996. Hayek and After. New York: Routledge.
--. 1997. Hayek, Keynes, and the State. History of Economics Review
26 (winter- summer).
--. 2000. The Use of Knowledge in Organizations. Knowledge,
Technology, and Policy 13, no. 3 (fall) 584.
Jeremy Shearmur teaches philosophy at the Australian National
University in Canberra.