Journal of Economic Pmpectives-Volume 11, Number 4-Fall 1997-Pages 139-152
Austrian and Neoclassical Economics:
Any Gains From Trade?
Sherwin Rosen
Austrian economics has been important to the development of modern economics,
but its role in current practice is much diminished. The neoclassical
approach dominates today’s thinking. Many Austrians bemoan this
state of affairs; most neoclassical economists just ignore it. But Austrian and neoclassical
economics aren’t mutually exclusive. Each represents a distinctive point of
view, although with little meaningful intellectual exchange between them these
days. This is unfortunate.
Many different concepts are associated with both Austrian and neoclassical
economics. To concentrate on where the most exchange of ideas might
occur, I am concerned about those aspects of Austrian economics primarily dealing
with the process of competition, and that portion of neoclassical economics primarily
dealing with the determination of economic equilibrium. This, in a nutshell,
is the main intellectual difference between them. Some consequences of this difference
are discussed in what follows.
I hope to show that mutually advantageous gains from trade exist. Austrian
economics offers a valuable perspective of the economy as an evolutionary process,
akin to biological evolution, with entrepreneurial activities representing the main
instruments of change. It is primarily a macroeconomic theory of the economy as
a whole, with keen insights on the nature of decentralization and the process of
competition that deserve more attention. Neoclassical economics offers a microeconomic
theory of economic behavior that is especially useful for analyzing specific
economic problems and for orienting analysis around more sharply described empirical
phenomena and data.
w Sherwin Rosen is Bergman Distinguished Service Professsor of Economics, University of
Chicago, Chicago, Illinois, and Senior Fellow, Hoover Institution, Stanford, Califmia. His
e-mail address is (rosen@cicero.spc.uchicago.edu).
140 Journal of Economic Perspectives
Process and Equilibrium
On the terms I have chosen, the evolutionary theme at the heart of Austrian
economics is most interesting when, in a very general sense, knowledge and information
are highly decentralized, imperfect, and widely dispersed. Economic systems
are seen to evolve as the amalgamation and interactions of trials and errors among
economic agents. Entrepreneurial ventures and experiments, arbitrage activities,
and survival of the fittest play crucial roles in this process. Not only do resources
gradually move to their (perceived) highest valued uses, but values themselves are
discovered and slowly revealed along the way.
In marked contrast, the methods of neoclassical economics mainly are concerned
with the establishment of economic equilibrium under fully known or (in
Marshallian terminology) given conditions of resource availability, technology and
preferences. Perhaps it is revealing that the old-fashioned and somewhat more general
terms “supply” and “demand” have gradually given way to the new and more
specific terms “tastes” and “technology.” Both must be fully specified before analysis
begins. Whatever the language, the central approach is to analyze the costs andbenefits
of alternative courses of action of consumers and producers, and find an
overall solution where individual decisions are mutually compatible and can be
implemented by everyone.
In neoclassical economics, individual behavior is described as the outcome of
rational choices-doing the best one can under the circumstances, and pursuing
those specific actions from available alternatives that maximize the difference between
personal benefits and costs. The choice set is fully specified, technology of
sellers and tastes of buyers are given, as are the number and varieties of goods.
There is a well-defined solution to the resource allocation problem. General conditions
can be found for which a market equilibrium “exists” in the sense that all
individual agents can independently fulfill the plans that best serve their selfinterests,
given the model assumptions.
Solving the existence problem is not a detail in the neoclassical scheme. It is
a logical necessity. For if the individual plans chosen to fulfill self-interest turn out
to be infeasible, the problem must have been incorrectly specified in the first instance.
The pieces cannot fit together and it is not possible for individuals to behave
in the postulated manner. The microeconomics won’t aggregate to the economy
at large, nor to the specific problem that the analysis purports to solve. It is important
that the conditions under which an equilibrium exists are fairly minimal.
Otherwise neoclassical models of empirical phenomena could not be generally ap
plied to a wide variety of circumstances.
There are serious questions of whether “disequilibrium” analysis is possible in
the neoclassical scheme. In my view it isn’t. Consider the crucial role of the “auctioneer”
and tatonnemont in the Walrasian system: the postulated conditions on
individual choice only hold when markets clear, at equilibrium prices. Years of
analysis of “out of equilibrium” trades in that framework have gotten hardly anywhere.
To be sure, many neoclassical models put restrictions on prices and on the
Sherwin Rosen 141
existence of some markets. Sometimes these give rise to such phenomena as rationing
and queues, but those too are part of the equilibrium process and must be
specified in the underlying microeconomic decision structure.
The Austrian approach is entirely different. The economy is in a perpetual
state of disequilibrium: things are always changing and in a state of flux. The economy
is ever evolving, creating unforeseen profit opportunities that agents are constantly
trying to find and exploit. This is not to say that neoclassical analysis rules
out mistakes and errors in economic decisions. Rational decision-making under
uncertainty implies that people might well regret their previous decisions. Extensions
of neoclassical theory since the 1950s address stochastic environments and
work out the logic of how markets and other social arrangements deal with risk and
risk-bearing. The market equilibrium concept is general enough to apply to the
sharing and trading of certain kinds of risks in a relatively straightforward way,
though problems of incomplete risk markets loom fairly large at the research frontier
today.
Coordination and the Invisible Hand
The overall point I am attempting to make is well-expressed by the “invisible
hand” welfare theorems of neoclassical economics. If there are enough markets to
internalize all transactions (no external effects) and decreasing returns (no nonconvexities),
then the welfare theorems hold that a Pareto optimal allocation is
achieved by a decentralized, competitive market price system. This is a consequence
of the rationality of individual decisions and the fact that market prices, common
to all participants, exhaust all gains from trade. It is routine in graduate economic
courses today to identify these results with Adam Smith’s invisible hand and the
virtues of markets.
Austrians have found this notion of the invisible hand misleading, if not entirely
off the point. They occupy some high ground here. Much of modern neoclassical
analysis goes back and forth freely between the Pareto optimum or “central
planning” allocation problem and the decentralized market implementation of the
optimum. In the Austrian approach, all the intellectual traffic goes only in one
direction, from individual behavior to social order. The central planning problem
cannot even be defined!
This approach begins with the premise that there is an enormous amount of
ignorance in the system. No one knows or can ever know what is being maximized
overall. Decentralization is fundamental because specialization is extreme. Consider
that out of the totality of what is known in the economy at large, any single
person knows essentially nothing. Not only are people highly specialized in their
work activities-each of us is an extremely small cog in an extremely large wheelbut
none of us consumes or produces more than a trivial fraction of the truly huge
set of goods that are actually traded.
If we can’t even identify most of the goods and services that form the basis of
1 42 Journal of Economic Perspectives
economic affairs, how can we know our full opportunity sets? How can we define
preferences over such goods or over those that might appear on the market at some
future time but are unknown today? And if we do not know them ourselves, how
can anyone aggregate what is known by all? The central question is how all this
knowledge and dispersed activity combines into a meaningful social allocation
mechanism. The reverse question of decentralizing some notion of aggregate welfare
or even the concept of Pareto optimality never arises.
This is not an academic debating point of pure theory or doctrine. There are
sound economic reasons for the compartmentalization and specialization of knowledge
in society and for the incomplete dispersion of information. Economies of
scale in learning and acquiring information make it socially advantageous for people
to specialize in their skills and information and trade with each other. This is
what accounts for the elaborate division of labor and the lengthy roundabout chains
of production in modern economies. We do much better by learning a narrow
range of specific skills very well, using them as intensively as possible, and trading
with others to share the fruits of their specialized skills. Scale economies themselves
create gains from trade and provide the social basis for extensive economic interactions
among people (Rosen, 1983a). They give rise to the “propensity to truck,
barter and trade” that is unique among the human species.
What must make it all possible is that knowledge itself is somehow divisible
among people. For example, a person doesn’t need to know much about science
to construct a chair. An elementary and rather superficial knowledge of the rigidity
of materials is sufficient. The rest can be bought or transmitted by materials manufacturers.
The output of specific knowledge-the chair itself-is what has eco
nomic value, not the details of its construction. Buyers need know nothing of atoms,
glue, nails and upholstery, nor the accountancy, securities laws and other manufacturing
regulations to sit on a chair and decide whether to purchase it. Someone
has to know those things, of course. But that knowledge largely is transmitted instrumentally,
in the intermediate products that combine to produce the object
itself. It is important to understand the mechanism by which all this divided knowledge
aggregates for the social benefits of all.
Scale economies are also present in acquiring economic information. It is not
worthwhile to collect expensive information that will be seldom used. Why should
a consumer be well-informed about prices and availabilities of goods that probably
will not be chosen? Why should producers learn about opportunities they probably
will not take? Certain kinds of ignorance are rational. The costs of knowing often
exceed the benefits.
Evolutionary Processes
The idea of decentralizing the problems of a social optimum through markets
is turned on its head in the Austrian scheme. The fundamental issue becomes one
of assessing how all the individual pieces fit together and how to make sense of the
Austrian and Neoclassical Economics: Any Gains From Trade? 143
whole. However, by not spelling out any potential empirical criteria for assessing
the performance of the economy as a whole (no one can know this), it is not entirely
clear what Austrian economics has to say about the workings of the economic systern.
This is why process is paramount in their scheme. External criteria imposed on
the rules of the game-the rule of law, openness, freedom of action, and libertyplay
important roles because incomplete knowledge puts limits on how outcomes
can be evaluated. No stone must be left unturned to allow individuals to investigate
new possibilities and improve on the always imperfect existing ways of doing things.
The analytical precision of neoclassical economics as a logical system in principle
is hermetically sealed from philosophical, ethical and moral issues, though of
course in practice it is not. Neoclassical welfare economics is greatly constrained by
the necessity of respecting known or given preferences and technologies. Austrians
are not unwilling to confront questions of what kinds of social institutions and rules
of the game make for a good society.
In many ways these aspects of Austrian economics parallel the theory of evolution
and natural selection. An extraordinary image presented by Richard Dawkins
(1986) in one of his books on evolution is remarkably similar to the Austrian eco
nomic idea of a spontaneous order, an elaboration of Adam Smith’s invisible hand
stated by the Austrian economist Earl Menger (1871) and eloquently exposited later
by Friedrich von Hayek (1937, 1945, 1960).1 Dawkins recasts Thomas Paley’s criticism
of Darwin by way of the example of the construction of the human eye. How
could such a complex and wonderful object be constructed by other than a Supreme
Designer? To bring the idea closer to home, imagine a complicated machine,
say a Swiss watch or a modern automobile. Whatever example is chosen, the story
serves as a metaphor for the complex and well-oiled machine that is the economy
at large.
Buyers of the machine know little or nothing of its inner workings. They usually
value the unit only for the work it can accomplish, not particularly for the circumstances
of its manufacture. Extraordinary synchronization and coordination among
thousands of intricate components are required for smooth performance of the
machine as a whole. Consumers are ignorant of these details, and play no direct
part in this coordination. How does it come about?
The Austrian view is built up from the spontaneous activities of myriad expert
specialists single-mindedly pursuing and perfecting their own component businesses,
including intermediaries who buy and sell from the most economical sup
pliers and assemble and market the final product. There is no overall plan. The
watchmaker is blind, but the watch happens anyway. It evolves to an ever-higher
state as individual specialists and assemblers act in their own self-interests to improve
their products and to gain a competitive edge over rivals. Local conditions
If Dawins (1986) is more “Austrian ” in its discussion, then Dawkins (1983) is more “neoclassical.”
See the brief discussion by Hayek (1960, ch. 4) on Darwin’s knowledge of economics, especially of Smith
and Malthus.
144 Journal of Economic Perspectives
might encourage different sellers to use different production methods or to produce
alternative varieties that serve different types of customers.
In fact, the economic history of the development of the watch clearly reveals
its spontaneous and piecemeal evolution. Increasingly accurate measurement of
time (measurement without theory!) was instrumental for new ways of thinking
about science and led to great advancements. And the increasing sophistication of
the manufacture of precision timepieces enabled widespread use of machine tools
that spilled over to manufactures as a whole and made mass production possible
(Landes, 1983; Boorstin, 1983).
Markets, Socialism and Central Planning
It is not uncommon for neoclassical economists to push the welfare theorems
to the limit and cast markets and price mechanisms in a seemingly secondary role
for decentralizing and implementing well-defined mathematical optimization prob
lems. After all, the “central planning problem” is equivalent to a market solution,
given the specification of technology and tastes. If the conditions of the welfare
theorem hold, the answer can be applied directly to the data without studying
individual maximizing decisions at all, That’s the way markets “must do it.“’ This
method has its virtues. It allows the analyst to abstract from many complicated
details and investigate the key features of the economic environment-specific aspects
of preferences or technologies- needed to capture the main empirical features
of a problem. It also has its defects. Sometimes components of individual
behavior from which the phenomena are built up are ignored and some of the data
to which individuals are supposed to react are not examined. This method is unthinkable
in the Austrian approach. The debate over central planning earlier in
this century best illustrates the essential logic of the Austrian approach and one of
its most outstanding successes.
In the years before World War II it became popular to point out that a central
economic planner, using information on available technology, could redistribute
and set prices in a way that would produce the socially desired outcome (Lange,
1938). This “market socialism,” it was further argued, would be superior to an
uncontrolled market because the central planner could make adjustments to compensate
for monopoly power, costs of dislocation and unemployment, and so on.
In what was perhaps their finest hour, the Austrians, led by Mises and Hayek, argued
that this vision of market socialism was impossible, and that it was based on a fundamentally
misguided vision of markets and prices.”
That the central planner could successfully mimic the competitive market so-
Kydland and Prescott (1982) is an outstanding and influential example.
In this journal, Persky (1991) gives a succinct summary of some aspects of the role of von Mises and
Lange in the debate.
Sherwin Rosen 145
lution by choosing the “correct” prices is true only if information and knowledge
available to the center is complete and, even if that is granted, when the center’s
legal power to manipulate economic life is not tainted by self-serving actions.
Though some market socialists recognized that the center would have to experiment
and ask lots of questions to discover the necessary information, the Austrians
insisted that the problem was not only that the information had to be collected at
enormous expense, but also that information about technology, tastes, and products
literally existed only when the market called it into being. Even then it existed
only in an extremely decentralized form, continuously changing and evolving
within the specific circumstances of time and place. Economists have also come to
a better understanding that individuals at the center also act selfishly, and not
necessarily in the public interest.
Surely these are the reasons why the highly regulated, centrally controlled
economies have fared so poorly. The failure of planning and central control and
market reforms occurring all over the world today is one of the most important
economic events of our age. Isn’t it is odd that so few neoclassical economists could
use their theory to take much of a stand on such matters? But the logical basis of
neoclassical theory is not well-equipped for that task. The collapse of central planning
in the past decade has come as a surprise to most of us. Economists who early
on questioned the reports of the economic successes of socialist and communist
economies were ignored, if not ridiculed, by many in the economics establishment.
Mechanism Design
The challenge laid down by Hayek and the market socialist debates in part
stimulated economic theorists to develop the theory of mechanism design, the systematic
study of social devices and institutions that might efficiently communicate
private information known by one agent to others. The main difficulty is clear
enough. People with private knowledge and information have incentives to communicate
only what is in their self-interest. Their willingness to reveal what they
know depends on how it will be used and how it will affect them. Voting on a public
good is the prototypical example. Voters have strong incentives to exaggerate their
preferences because the expression of value to them is isolated from their direct
willingness to pay for it.
Truthful elicitation of information requires incentive-compatible mechanisms.
These are feedback rules from the reported information to the payoffs received by
the reporter that always make it in the reporter’s self-interest to tell the truth. A
noteworthy recent result is the “revelation principle, ” which establishes that truthful
mechanisms usually must use elicited information in socially inefficient ways
(Myerson, 1979; Townsend, 1979). Otherwise there are incentives for the reporter
to manipulate the situation by lying. Of course, lying and attempted manipulation
also are inefficient. Inefficiency can be minimized. It can’t be avoided.
While this development is interesting and firmly in the neoclassical tradition,
146 Journal of Economic Perspectives
Austrians may not be too impressed with it. Achieving incentive compatibility often
requires that the mechanism designer know an enormous amount about the conditions
of the problem-almost as much as a central planner would have to know.
This includes expert and detailed knowledge of the preferences, budgets, and technology
available to economic actors, information that is properly considered as
private knowledge in many (most?) division-of-labor contexts. Thus, economic
forces put sharp limits on how knowledge and information can be elicited for economic
planning and social welfare purposes.
There is no getting around the fact that decentralized markets generally utilize
divided, specialized knowledge rather efficiently. Furthermore, many social and
legal arrangements, including competition itself, greatly constrain specialists from
using their private knowledge to cheat others, though of course such arrangements
do not entirely eliminate cheating and deceit. In fact, it seems as if an open competitive
process itself deters opportunism and cheating by increasing the range of
choices available to economic agents. Much of the recent work on agency problems
concentrates only on contractual microeconomic solutions and ignores the important
influence of markets in disciplining economic activities. That is an important
legacy of Austrian economics that deserves more attention.
Dynamic Analysis, Expectations, and the Austrian Perspective
Some old-timers used to say that Austrian economics pursued a “dynamic”
approach while the neoclassical approach was “static.” If this claim was ever true
in the past, it is false today. Much neoclassical analysis in the past three decades has
been devoted to explicitly dynamic, intertemporal resource allocation problems,
while following an equilibrium approach. An essential aspect of these problems is
that current decisions depend on the plans that agents expect to follow in the
future. The way expectations are formed becomes an essential part of the economic
structure, one of the given conditions of analysis.
The concept of rational expectations imposes a kind of intertemporal consistency
requirement on average values of future equilibria. It postulates that agents
are sufficiently well-informed that their expectations turn out to be true on average.
It follows that, on average, previous anticipated plans can be carried out. The thinking
behind such models runs along Austrian lines; inconsistent expectations imply
unexploited profit opportunities, and create scope for entrepreneurial activities to
eliminate them.
However, since actual realizations never turn out as anticipated even in rational
expectations models (only the averages do) it stretches terminology to think that
nonrational expectations represent “disequilibrium” models in any more fundamental
sense. In this way, neither approach really addresses Austrian concerns.
Indeed, Austrians are as hostile toward rational dynamic models as are many other
camps.
Austrian and Neoclassical Economics: Any Gains From Trade? 14 7
Empirical Content and Justifying Methodology
Neoclassical economics gains much of its power and utility from the “as if’
principle: We construct economic models as if rational economic actors optimize
some objective subject to constraints. Most neoclassical economists, at least those
of my generation and older, judge the value of this enterprise by whether the relatively
simple rules of behavior implied by this form of rationality have useful descriptive
content and empirical predictive power. If rational optimizing models
work well in this sense, then theory implies that it is often possible, in principle, to
infer the underlying structure (supply and demand, for instance) from actual behavior.
Here the back-and-forth logic between equilibrium choices and given conditions
are crucial. The theory of revealed preference is the canonical example.
Knowing a person’s tastes and income, we can predict how market prices affect
choices; but having observed choices in different price and income configurations,
we can invert the process and infer what those underlying preferences must have
been, as long as preferences are reasonably stable and the source of variation is
sufficient to achieve identification.
Many Austrians hold to the view that quantitative empirical work in economics
is infeasible or uninteresting because the world is changing so much that “behavioral
relationships” inherently are unstable and it is fruitless to estimate them. An
unwillingness to pursue the consequences of “given conditions” greatly limits the
empirical scope and consequences of Austrian economic theory. The paucity of
quantitative empirical work in the Austrian tradition accounts for why so few Austrians
are found in the professional economics community today. Their approach
basically excludes most of the things that most economists do. Austrians tend to
disavow what they consider to be “routine” mathematical optimization problems
that underlie much of empirical economics.
Yet some problems are more routine than others. Consider how difficult it has
been for centrally directed economies to manage their industries efficiently and to
make the best use of their resources under known technological conditions. Empirical
work is difficult under almost all circumstances: the neoclassical inference
problem of going from observed behavior to underlying structure has proven especially
hard to solve in practice.
Less precise quantitative evidence in economics often is much more secure
than precise structural estimates; for example, most of us are much more confident
that demand curves slope downward than about the precise values of demand and
supply elasticities for specific goods. Large scale, gross evidence on economic effects,
like the general effects of OPEC on the use of oil or the way in which the
voluntary army affected the supply of recruits, may be difficult to quantify with
journalquality econometric precision, but count very strongly to inform profes
sional judgements about economic behavior. Perhaps the frustration of estimating
structural parameters has turned many non-Austrian economists away from the
neoclassical modeling approach at the moment. Many are attracted by a looser and
more institutional style. Others seek out “natural” and laboratory experiments to
148 Journal of Economic Perspectives
minimize the need to model all the intervening “nuisance” influences that always
affect a problem. Still others are attracted by the complexities of game theory and
the elaboration of strategic behavior it affords, because the market equilibrium
paradigm is too benign for their world view.
For some Austrians, lurking in the background is the view that quantification
and intellectual conceit might lead to inappropriate central control. It is mildly
interesting here that the root word for “statistics” is “state.” The historical development
of social and economic statistics in this century largely parallels growth of
the state in economic life. But the causal connections aren’t very clear. Perhaps the
development of national income statistics affected some aspects of macroeconomic
policy over the past 75 years, but social events created the need to pursue those
matters in the first instance. The more distressing fact is that the enormous growth
and influence of the state in the 20th century, surely the most important littleunderstood
economic fact of our era, cannot be explained by either Austrian or
neoclassical economics. The economics of the state, Austrian or otherwise, fills the
pages of a very small (though growing) book.
Although the methodology of positive economics and the task of making falsifiable
predictions about data from well-specified theoretical models has certain
limitations, it can also be realistic and useful, as indicated by many successful examples
over the years. Taxes, regulations and shifts in supply always occur and we
might as well try to understand them. Surely it is not without interest to know how
a tax will affect the price and quantity of a commodity; how a specific policy, such
as licensing or price controls, will quantitatively affect a market; or how a rise in
the price of some resource will affect its use. By downplaying the possibilities for
empirical inference, Austrian mainly confine their empirical investigations to historical
case studies. However, the uniqueness of each study and the difficulties of
drawing out common elements capable of wider generalization have limited their
usefulness in economics. A more quantitative approach has dominated empirical
research and will continue to do so. That kind of research is impossible without
relatively sharp definitions and quantifiable concepts.
The Role of the Entrepreneur
There is a joke circulating today among economists in Poland. Question:
“How many people does it take to change a light bulb?” Answer: “None, the market
will do it” The joke tweaks the noses of neoclassical economists, but doesn’t faze members
of the Austrian school. In Austrian economics the world isn’t exclusively populated
by optimizing automatons, passively consuming or producing their marketequilibrating
quotas of goods and services at market equilibrium prices. The real movers and shakers
in the economy are entrepreneurs. These are the people who take action, compete with
each other, and perfect markets. They are the instruments of economic energy and
change. They make the competitive, evolutionary process what it is.
Entrepreneurs are not to be found in neoclassical economics. The term does
Sherwin Rosen 149
not appear in the indexes of the main graduate texts on economic theory, nor is
the concept mentioned in any context or under a different name. The fact is that
there is no role for entrepreneurs when economic conditions are “given,” when
the list of goods to be traded is cut and dried, when consumers and producers are
clearly identified, and when resource availabilities are known. Entrepreneurial activities
are only possible when the nature of the world is not fully known, when
knowledge and information are incomplete and dispersed, precisely the conditions
postulated by the Austrian approach. Entrepreneurship is a disequilibrium phenomenon.
The entrepreneur is a person who exploits heretofore unrecognized opportunities.
4 In equilibrium neoclassical economics, “given conditions” means that
there is nothing for the entrepreneur to do.
Rather too much, in my judgement, is made in Austrian theory about defining
and interpreting entrepreneurial activities (von Mises, 1949; Schumpeter, 1934,
1962) No matter how fine the distinctions or the degree of hair-splitting, it remains
an elusive concept that lacks an operational definition and cannot be quantitatively
measured. Though all economists recognize its significance, there exist no quantitative
measures of the scale or scope of entrepreneurial activities in the economy.
Entrepreneurship is the kind of thing that can be recognized after one sees it,
but is hard to describe in the abstract. Interesting and useful case studies of important
entrepreneurs abound. However, case studies of entrepreneurial failures
are infrequent, even though it seems likely that many, perhaps even most entrepreneurial
ventures are unsuccessful. If we cannot measure the total volume of
entrepreneurial activity, there is no way to assess its economic importance and rate
of return, nor to evaluate the social and legal environments that nurture it or
suppress it. What is the current supply price and elasticity of supply of entrepreneurs?
There’s no way of knowing.
Nonetheless, totally ignoring the concept of entrepreneurship has a very constraining
effect on the neoclassical view of competition. It is precisely here where
potential gains from intellectual trade are largest in my judgement. The Austrian
view of competition as evolutionary struggle is a very compelling idea that plays no role
in neoclassical economics, which is constrained to look at the final outcomes of the
competitive process after all competitive opportunities in the Austrian sense have
been exploited.
The outwardly simple notion of competition as a comparison among alternatives
is in reality an immensely sophisticated statistical selection scheme for aggregating
incomplete and specialized knowledge, for achieving consensus among disparate
views, and for disciplining market participants to align their self-interests
with social interests. Entrepreneurial ventures are statistical experiments among
alternative ways of doing things. Those that are better tend to prosper and survive.
They pass the market test. Those that are worse tend to recede and vanish. The
role of legal institutions, government, and property rights establish the rules of the
4 See especially the comprehensive account of Kirzner (1973); my own views appear in Rosen (1983b).
150 Journal of Economic Perspectives
game and affect “fitness” of the system overall. That is why they are so important
in the Austrian scheme. Neoclassical literature on this kind of competition has
appeared in recent years, but little of it has been inspired by Austrian economics.’
Neoclassical economics undoubtedly would be enriched by a more fully articulated
view of competition as a selection device, as an economical processor of
information, and as a generator of economic change. Yet it is hard to assess exactly
what is lost without it. The neoclassical approach does not ignore change. Rather,
it analyzes change in a more empirically manageable way. The whole point of describing
how an economic equilibrium depends on given conditions is to infer how
changing conditions affect the data. This forces the analyst to pursue a research
strategy of looking for the fundamental sources of change. Neoclassical dynamics
is basically represented as a moving equilibrium process rather than as an Austrianstyle
perpetually disturbed disequilibrium. The theory of economic growth is a leading
example.
In many ways, controversies over which phenomena are in equilibrium and
which are not can’t go anywhere. The data don’t care one way or the other. The
relevant question is which way of thinking works best in practice, and neoclassical
economics has had its share of successes. The Austrian approach has had its share
too, but the spheres in which these successes have occurred are substantially
different.
The Market Test?
Debates about methodology have opportunity costs. Most of us prefer to assess
serious attempts to do economics, rather than to spend time arguing over which
methods should be used in those attempts. Instead of seeking a sure-fire method for
ascertaining economic truth, which experience shows is an impossible task, the
more practical stance is to use whatever methods work best in practice. Different
methods coexist because their usefulness differs from problem to problem. The
theory of natural selection applies to methods as well as to other things.
Karl Popper (1968) pointed out years ago that there is an enormous amount
of evolutionary Austrian competition in the marketplace for ideas. Only a few economists
have tackled this fascinating idea (Director, 1964; Coase, 1974), but the
argument goes something like this: Virtually all the a priori conditions for competition
in the industrial organization sense apply to the market for ideas. There
are large numbers of independent purveyors; the personal gains in money and
’ The classic statement of evolution as justification for neoclassical analysis is Alchian (1986). Nelson and
Winter (1982) pursue this analysis and often reach diierent conclusions. There has been substantial
interest in patent and other “racing” problems for intellectual property rights. See for instance Kamien
and Schwartz (1983). A related literature on competitive tournaments mainly examines incentive issues
(Lazear and Rosen, 1982). Some of the information revealing aspects of contests are discussed in Rosen
(1986)) but little systematic work has been done on the problem.
Austrian and Neoclassical Economics: Any Gains From Trade? 151
status are big enough to make competition interesting and attract entrants, and
entry is about as free as it gets in any business. From the Austrian point of view,
there is no dearth of intellectual entrepreneurs looking for heretofore unrecognized
opportunities to peddle their wares. Considering the common property aspects
of intellectual life, maybe there are too many! What is the fact that neoclassical
economics has scored higher than Austrian economics on the evolutionary/survival
test telling us?
Many economists and other intellectuals get nervous about the thought of
applying the market test to ideas. No doubt it has to be carefully applied (especially
to one’s own work). For instance, the equivalent of monopolistic elements, in the
form of fashion and peer pressure, appear in the intellectual marketplace from
time to time. Frictions and adjustment costs in changing professional consensus
naturally arise from specialization within the profession. It isn’t possible to be on
the frontier in all subjects of a discipline. It is rational to remain ignorant of some
topics and rely on the opinions of the “establishment.” Some fashions persist
longer than others, but as some Austrians have put it, why aren’t monopolies in
ideas just as temporary as monopolies in goods? Eventually superior new entrants
will break through. The strong grip that Keynesians held on macroeconomics and
the Marxists on history, or since we are speaking of the Viennese, the straightjacket
in which atonality bound modern musical composers for so many years, serve as
examples of why we must look to the long run in assessing the market test. It often
takes time for new ideas to gain a foothold against prevailing views.
The great Austrian economists were recognized as such in their day. The Austrian
approach dominated American economics at the turn of the century. They
are not forgotten and many of their ideas have been incorporated into the canon.
However, others have been forgotten. It would be a pity if economists lost the
enduring legacy of the economic systems aspects of their views, of how the process
of open competition and survival of the fittest work to aggregate highly decentralized
knowledge and information into an amazingly well-constructed, mostly
smoothly operating, but rudderless social organization.
n Thanks go to Kenneth Arrow, Gary Becker, Stanley Engerman, William Grampp and Robert
Lucas for discussions and criticism of an early draft. Alan Kreuger and Timothy Taylor made
helpful editorial suggestions. None of these people agree w i t h all that remains.
References
Alchian, Armen, “Uncertainty, Evolution, and Ouchi, eds., Organizational Economics. San Fran-
Economic Theory,” in Jay Barney and William cisco and London: Jossey Bass, 1986.
1 52 Journal of Economic Perspectives
Boorstin, Daniel, The Discoverers. New York:
Random House, 1983.
Coase, Ronald, “The Market for Goods and
the Market for Ideas,” American Economic Review
1974, 64:2,384-91.
Dawkins, Richard, The selfish Gene. Oxford &
New York: Oxford University Press, 1976.
Dawkins, Richard, The Blind Watchmaker. New
York: Norton, 1986.
Director, Aaron, “The Parity of the Economic
Marketplace,” Journal of Law and Economics 1964,
7, l-10.
Hayek, Friedrich A., “Economics and Knowledge,”
Economica, 1937, 4, 33-54.
Hayek, Friedrich A., “The Use of Knowledge
in Society,” American Economic Review, 1945, 35,
519-30.
Hayek, Friedrich A., The Constitution of Liberty.
Chicago: University of Chicago Press, 1960.
Kamlen, Morton I. and Nancy Schwartz, Market
Structure and Innovation, Cambridge; New York:
Cambridge University Press, 1983.
Kirzner, Israel M., Competition and Entrepreneurship.
Chicago: University of Chicago Press, 1973.
Kydland, Finn and Edward Prescott, “Time to
Build and Aggregate Fluctuations,” Econometrica,
1982,506, 1345-70.
Landes, David S., Revolution in Time. Cambridge,
MA Belknap Press of Harvard University
Press, 1983.
Lange, Oskar and Fred M. Taylor, On The Eco
nomic Theory of Socialism. Minneapolis: University
of Minnesota Press, 1938.
Lazear, Edward P. and Sherwin Rosen, “Rank-
Order Tournaments as Optimum Labor Contracts,”
Journal of Political Economy, 1981, 895,
841-64.
Menger, Carl, principles of Economics, First, general
part. Trans. and ed. J. Dingwell and Bert F.
Hoselitz, Glencoe IL: Free Press, 1950.
Myerson, Roger, “Incentive Compatibility and
the Bargaining Problem,” Econometrica, 1979,
47~1, 61-73.
Nelson, Ralph and Sidney G. Winter, An Evolutionaryy
Theory of Economic Change. Cambridge,
MA: Belknap of Harvard University Press, 1982.
Persky, Joseph, “Lange and von Mises, Large-
Scale Enterprises, and the Economic Case for Socialism,”
Journal of Economic Perspectives, Fall 1991,
5:4, 229-36.
Popper, Karl Raimond, The Logic of Scientific
Discovery. Third ed, revised. London: Hutchinson,
1968.
Rosen, Sherwin, “Specialization and Human
Capital,” Journal of Labor Economics, 1983a, l:l,
43-9.
Rosen, Sherwin, “Economics and Entrepreneurs,”
in Joshua Ronen, ed. Entrepreneurship.
Lexington, MA: LexingtonBooks, 1983b.
Rosen, Sherwin, “Prizes and Incentives in
Elimination Tournaments,” American Economic
Review, September 1986, 76:4, 701-15.
Schumpeter, Joseph A. The Theory of Economic
Development. Cambridge, MA: Harvard University
Press, 1934.
Schumpeter, Joseph A. Capitalism, Socialism
and Democracy. London: Unwin University Books,
1965.
Townsend, Robert, “Optimum Contracts and
Competitive Markets with Costly State Verification,”
Journal ofEconomic Theory, 1979, 21:2,265-
93.
von Mises, Ludwig, Human Action. New Haven:
Yale University Press, 1949.