The Economic Theory of Costs Foundations and New Directions

Чтобы посмотреть этот PDF файл с форматированием и разметкой, скачайте его и откройте на своем компьютере.
The theory of costs is a cornerstone of economic thinking, and �gures crucially
in the study of human action and society. From the �rst day of a principles-level
course to the most advanced academic literature, costs play a vital role in virtually
all behaviors and economic outcomes. How we make choices, why we trade, and
how we build institutions and social orders are all problems that can be explained
This volume explores, develops, and critiques the rich literature on costs,
examining some of the many ways cost remains relevant in economic theory and
practice. The book especially studies costs from the perspective of the Austrian
or “causal-realist” approach to economics. The chapters integrate the history
of economic thought with contemporary research, �nding valuable crossroads
between numerous traditions in economics. They examine the role of costs in
theories of choice and opportunity costs; demand and income effects; production
and distribution; risk and interest rates; uncertainty and production; monopsony;
Post-Keynesianism; transaction costs; socialism and management; and social
Together, these papers represent an update and restatement of a central element
in the economic way of thinking. Each chapter
reveals how
the Austrian, causal-
realist approach to costs can be used to solve an important problem or debate in
economics. These chapters are not only useful for students learning these concepts
for the �rst time: they are also valuable for researchers seeking to understand the
unique Austrian perspective and those who want to apply it to new problems.
Matthew McCaffrey
is Assistant Professor of Enterprise in the Alliance Man-
chester Business School, University of Manchester, UK. His research focuses on
entrepreneurial decision making, the role of entrepreneurship in social and eco-
The Economic Theory of Costs
Political Economy as Natural Theology
Sharing Economies in T
Edited by Anthony Ince and Sarah Marie Hall
Philosophy in the T
Edited by Kenneth W. Stikkers and Krzysztof Piotr Skowroński
Public Policy and the Neo-Weberian State
The Economic Theory of Costs
Routledge Frontiers of Political Economy
For a full list of titles in this series please visit
The Economic Theory
Edited by Matthew McCaffrey
2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
711 Third Avenue, New York, NY 10017
Routledge is an imprint of the Taylor
© 2018 selection and editorial matter, Matthew McCaffrey; individual
The right of the Matthew McCaffrey to be identi�ed as the author of the
Trademark notice
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
A catalog record for this book has been requested
Typeset in Times New Roman
by Apex CoVantage, LLC
List of tables
List of contributors
Introduction: the economic theory of costs in perspective
The “income effect” in causal-realist price theory
The evolution of causal-realist production theory
From Marshallian partial equilibrium to Austrian general
s production theory
Man, economy, and state
oducer’s activity
Risk, uncertainty, and cost
The myth of the risk premium
Time and the theory of cost
Causal-realist price theory: debate and synthesis
Monopsony theory revisited
Costs and pricing: an Austr
Economic organization, entrepreneurship, and the �rm
notes on a doubtful compatibility
Management is what’s wr
Economic calculation and the limits of social

Derivation of a �rm’s demand curve for a factor of production

Total product outlay curve
Graphical illustration of production of Product P for various
Marginal unit of output in production
Production costs with different techniques in the ERE
Jones’ money returns
Factor combinations for the production of Good A
of Good A
Gross revenue in the production of Good A, Case (a)
Gross revenue in the production of Good A, Case (b1)
Gross revenue in the production of Good A, Case (b2)
Gross revenue in the production of Good A, Case (c)
production of Good A
Smith’s production decisions for Product P
Smith’s money returns for various money outlays
Smith’s money returns for various outlays, continued
Smith’s money returns for various outlays, total
Smith’s money outlays for producing Products P

Smith’s money returns for various investment decisions
Types of entrepreneurship by income and or
Per L. Bylund
is Assistant Professor of Entrepreneurship and Records-Johnston
Professor of Free Enterprise in the School of Entrepreneurship at Oklahoma
State University. He is the author of two books,
The Problem of Production:
New Theory
of the Firm
(Routledge, 2016) and
The Seen, the Unseen, and
the Unrealized: How Regulations Affect Our Everyday Lives
2016). His areas of research are entrepreneurship, strategic management, and
economic organization.
Jeffrey M. Herbener
is Chairman of the Department and Professor of Economics
at Grove City College. He serves as Associate Editor of the
Quarterly Jour
nal of Austrian Economics
and is a Senior Fellow of the Ludwig von Mises
is Professor of Economics at the University of Angers, a
Senior Fellow of the Ludwig von Mises Institute, and a member of the Euro-
pean Academy of Sciences and Arts. He is the author of
Krise der
The Ethics
of Money Production
(2008), and
�ve other books.
He has also edited
The Theory
of Money and Fiduciary Media
(2012) and
other books. His writings have been translated into twenty languages. His cur-
rent research focuses on the political economy of �nancial markets, and on
monetary theory.
is an Assistant Professor at the Institute of Economic Sciences
at the University of Wroclaw, and a Researcher at the
Faculty of
Social and
Xavier Méra
holds a PhD in Economics from the University of Angers, France.
He is a Teaching and Research Assistant at Université Rennes 2 and an Associ-
Matthew McCaffrey
is Assistant Professor of Enterprise in the Alliance Man-
chester Business School at the University of Manchester. His research focuses
on entrepreneurial decision making, the role of entrepreneurship in social and
economic development, and the institutional conditions in which enterprise
received his PhD in Economics from Auburn University in
2016. He is a Fellow of the Ludwig von Mises Institute and the Online Course
Manager for the Foundation for Economic Education. He currently teaches
economics at Auburn University.
Patrick Newman
is an Assistant Professor of Economics at Florida Southern
(1926–1995) was a leading economist of the Austrian
school and the author of dozens of books and scholarly research articles. His
treatise on economic principles,
Man, Economy, and State
, played a major role
in the revival of the Austrian tradition in the United States.
Joseph T. Salerno
is Professor of Economics at Pace University and Academic
Vice President of the Ludwig von Mises Institute. He is the Editor of the
terly Journal of Austrian Economics
and the author of
Money: Sound and
(2010) as well as of numerous articles in peer-reviewed economics
Mihai-Vladimir Topan
is Associate Professor in the Department of International
Business and Economics at the Bucharest University of Economic Studies. He
is also President of the Ludwig von Mises Institute Romania and Founder of
the Academia Privată.
appreciate the discussions with the 2016 Mises Institute Research Fellows
on this topic. I
would also
like to thank Randy Beard, Joseph Salerno, and
Michael Stern for discussing opportunity costs and other esoteric economic
riddles with me. Matthew McCaffrey is an excellent editor and I
thank him
his great advice and his patience. Of course, I
take full
responsibility for
om Marshallian partial equilibrium to Austrian general
equilibrium: the evolution of Rothbard’s production theory
The author would like to thank Peter Boettke, Peter Klein, Matthew McCaf-
frey, Ennio Piano, Joseph Salerno, and an anonymous referee for helpful
comments. In addition, he thanks the Ludwig von Mises Institute for the use
of the Murray N. Rothbard archives, as well as Barbara Pickard for archival
assistance. Any remaining errors are the author’s.
The myth of the risk premium
Jörg Guido Hülsmann
Comments from Mr
Maximilien Lambert
and Mr
Cleverly are gratefully
The author wishes to thank Renaud Fillieule, Guido Hülsmann, Tudor Smirna,
and Georges Lane for their comments on an earlier version of this chapter.
Any remaining errors are the author’s.
Mihai-Vladimir Topan
The author would like to thank Radu Muşetescu, with whom he has discussed
transaction cost economics
. Any remaining errors are the
1: Economic calculation and the limits of social entrepreneurship
Matthew McCaffrey
The author would like to thank Carmen-Elena Dorobăţ and Patrick Newman
for helpful comments on the earlier drafts of this chapter. Any remaining
errors are the author’s.
At �rst glance, the “economic theory of costs” seems like a mundane topic, even
for the dismal science. The term does not call to mind fashionable research trends
in economics, nor does it clearly hint at pressing problems in the global economy.
In fact, if anything, questions about “the theory of costs” recall the years of “high
theory” in the early twentieth century
– and
rightly so. At that time, econom-
ics was still carving out its niche in the social sciences, and therefore welcomed
deep and wide-ranging discussion of its fundamental problems. It was thus a
very different discipline from the narrowly empirical and formal profession it has
become, in which “big” questions, and earlier chapters in the history of thought
in general, are mostly irrelevant. However, although they represent a road not
taken (or, rather, a road discontinued), the years of high theory are in many ways
exemplary. They witnessed many vigorous exchanges of ideas between major
economists, debates that laid foundations for modern work that have only partly
For reasons that will become clear, many of the vital problems of this era, and
of economics in general, can be grouped under the heading “the economic theory
of costs.” The present book draws together several new and valuable contributions
to this literature. Before outlining its contents, however, I
would like
to explain
further what is meant by “the economic theory of costs,” as well as clarify how
the topic �ts within economics in general, and why it is worth studying. Doing so
will also help to describe the scope and purpose of the individual chapters, and of
Despite appearances, the theory of costs is neither an obscure nor an uninterest-
ing strand of economic research. In fact, it is a cornerstone of economic thinking
that can profoundly in�uence fundamental theory, its countless modern branches
and applications, and even other social sciences and management disciplines. Far
from being a relic of a forgotten era, then, the study of costs lies at the center of a
vibrant and ongoing research agenda in and around all �elds of human action and
The concept of cost has played a prominent role in economics for more than
two centuries. Most importantly, in the classical era, costs
– speci�cally
, the
long-run costs of production
– were
often given pride of place as the ultimate
Matthew McCaffrey
Matthew McCaffrey
prices (Klein, 2007, p.
8; Buchanan,
1969, pp.
1–7). They
thus inextricable from any discussion of the essential theorems of economics.
Nevertheless, the exact relevance of costs for value and prices was �ercely
debated, resulting eventually in the overturn of the classical approach by the
marginalist revolution that began in 1871. This sea change in economics thor
oughly revised the theory of costs. Today, costs remain inextricable from the
theory of value, but for the opposite reason than the one imagined by the clas
The rise of marginalism is mainly associated with the work of Carl Menger,
William Stanley Jevons, and Léon Walras. However, it was Menger who most
fully realized the potential of the subjective theory of value that lies at the heart
of marginalism (Menger, 2007 [1871]). Like his contemporaries, Menger helped
revolutionize the theory of cost by explaining that the value of any good is ulti-
mately determined by its marginal utility to consumers. Yet he went a step further
in his explanation of the pricing process by showing that costs of production are
not an independent cause of value; on the contrary, the prices of the factors of
production (which together form the costs of production) are determined by the
value consumers assign to the �nished goods the factors produce. The subjective
valuations of consumers thus stand in a causal relation to the objective prices that
In elaborating this theory, Menger thus identi�ed two traits of sound economic
reasoning: �rst, it seeks to understand economic phenomena
– e.g.
prices, wages,
and interest rates
as they
appear in the real world
, and second, it investigates
relationships between these phenomena. These characteristics have led
some economists to describe Menger’s unique approach as “causal-realist.”
Menger’s ideas are usually associated with the geographical Austrian school, but
his writings inspired many economists outside the narrow con�nes of Vienna,
including Philip Wicksteed, Lionel Robbins, John Bates Clark, Frank A. Fetter,
and Herbert J. Davenport, each of whom contributed to the larger Mengerian,
Menger’s insights into value and price were most notably developed by Eugen
von Böhm-Bawerk and Friedrich von Wieser. Wieser is remembered mostly
for elaborating the opportunity cost concept, but Böhm-Bawerk also played a
formative part in early Austrian research, much of which involved clarifying and
defending Menger’s subjectivist approach to value theory (e.g. Böhm-Bawerk,
1962a [1891a], 1962b [1894], 2002 [1892]). In particular, Böhm-Bawerk repeat-
edly debated the role of costs in determining prices, and one of the crowning
achievements of early Austrian economics was to show, contra Marshall and the
classicals, that all prices are ultimately determined by subjective values, rather
than being determined by consumer values on the demand side and costs on the
supply side (Böhm-Bawerk, 1962b [1894]). Unlike value, which is a starting
point of economic theory, cost is a dependent concept. As Böhm-Bawerk put it,
The question of the relation of cost to value is properly only a concrete
form of a much more general question
– the
question of the regular relations
the values of such goods as in causal interdependence contribute to
costs are connected in various ways to virtually every fundamen-
tal concept in economics, including value, choice, utility, exchange, money,
pro�t, loss, and entrepreneurship (cf. for example, Böhm-Bawerk, 1962a, p.
Böhm-Bawerk even
remarked that in regard to “the interaction of price, value and
. it
is in my opinion no exaggeration to state that to understand their con-
nection is to understand a good half of economics” (Böhm-Bawerk, 1959, p.
The present
book offers a path toward the understanding to which Böhm-
Bawerk alluded. Exploring the theory of costs was instrumental in the early
development of Menger’s subjective value theory, and it is no less important for
contemporary economists who hope to advance Menger’s causal-realist approach
through their own work. That is the purpose of this collection: to showcase a vari-
ety of research strands within the modern Mengerian tradition that relate in some
way to the theory of cost. The organization of the topics follows a logical progres-
sion, beginning from the fundamental concept of choice
– where
the idea of cost
also begins
– and
building up to discussions of pricing, production, economic
organization, and comparative economic systems.
In Chapter
1, Jonathan
Newman surveys a recent controversy over the idea
of opportunity cost. Contemporary mainstream economists are divided about the
exact meaning of this vital concept, as well as the question of how it should be
taught to students. In particular, competing de�nitions of opportunity costs have
sown confusion in both economic research and teaching. Newman explains that
this confusion can be resolved by appealing to the thoroughly subjectivist view
of opportunity cost developed in the causal-realist tradition. If we recognize that
opportunity costs refer to a subjective and ordinal preference ranking, the tension
between different views of opportunity cost evaporates. However, some critics
close to the Austrian school have disputed whether the idea of opportunity cost is
valuable at all. Newman also shows why these views are mistaken, and he defends
the continued use of opportunity cost as a foundational concept in economics.
Speci�cally, he argues that criticisms of opportunity cost depend on assuming
away the forward-looking nature of action and cost and also on falsely con�ating
different kinds of choices that actors make.
The concepts of action, preference, and choice lead naturally to a discussion of
the formation of individual demand curves. In Chapter
2, Joseph
Salerno outlines
a causal-realist method for deriving the key principles of demand analysis using
the individual’s ordinal scale of values. Value scales provide the basis for deriving
the law of marginal utility
– which
requires the assumption of the constancy of
money’s purchasing power
– and
in turn allow us to deduce individual demand
curves. Unlike mainstream economic theory, the demand curve in causal-realist
analysis is a temporal construct that refers to the individual’s personal economic
situation at the moment of his purchases. This implies that income, as a �ow of
money, has no direct role in determining the demand curve, which is based solely
Matthew McCaffrey
the individual’s value scale (a ranking of existing stocks of goods
As a result, movements along the demand curve do not produce income effects,
which are a theoretical illusion. However, the causal-realist approach does shed
new light on the theory of substitution effects. That is, focusing on value scales
reveals that all goods are at least partial substitutes for each other. In fact, substi-
tution is a necessary relationship that prevails among all goods. Finally, Salerno
argues that, even without the concept of the income effect, the causal-realist
approach still provides an explanation for the existence of a backward-bending
The causal-realist approach extends beyond the analysis of individual con-
sumer behavior, however. In Chapter
3, Patrick
Newman outlines some major
differences between the causal-realist and mainstream views of production
theory. His discussion relies on a proto-chapter
of Murray
Rothbard’s treatise
Man, Economy, and State
, which appears as Chapter
4 of
this book. As Newman
explains, Rothbard’s early work relied on many conventional mainstream eco-
nomic assumptions and concepts in order to explain producer’s activity, including
perfect competition and the isolated �rm. Newman compares Rothbard’s earlier
with his
published work in order to chart the evolution of his thinking on
these vital economic topics. In particular, he shows that after drafting the origi-
nal production theory that appears in Chapter
4, Rothbard
became increasingly
aware of its theoretical shortcomings. He became especially critical of the Mar-
shallian partial-equilibrium analysis he initially embraced. As a result, Rothbard
abandoned this analytical apparatus, choosing instead to use the works of earlier
writers in the causal-realist tradition to build a highly original “Austrian general
equilibrium” production theory that was eventually included in the �nal version
Man, Economy, and State
. However, although he eventually took a very differ-
ent path in his theorizing, Rothbard’s proto-chapter
– and
Newman’s comparative
study of it
– helps
tease out several points of contrast between the causal-realist
To take one example, a truly realist approach to economic theory cannot rely
excessively on equilibrium constructs that abstract from the passage of time, and
therefore assume away the problems of risk and uncertainty. In Chapter
5, Jör
Guido Hülsmann revisits the theory of risk and questions the current role that risk
plays in economics, especially in the theory (and real-world formation) of interest
rates. He rejects the view that interest rates can be viewed as the arithmetic sum of
several separate, identi�able components. In contrast to this view, he argues that in
a free-market setting, all known risks either are accounted for through entrepreneur
ial judgment, or are irrelevant to acting individuals. As a result, observable interest
rates cannot contain a risk premium; instead, differences in prices that appear to
re�ect compensation for increased risk are actually nothing more than re�ections
of different subjective evaluations of available investment opportunities.
Importantly, the concept of uncertainty is far more challenging for economic
analysis than risk. In Chapter
6, Jef
frey Herbener uses this fact as a starting point
for integrating the element of time into causal-realist production theory. Time
brings with it the problem of uncertainty, which in turn has a profound in�uence
the production process. In particular, uncertainty implies that capitalist-
entrepreneurs must
speculate about the future discounted marginal revenue prod
ucts of the factors they employ. Their anticipations, and the interaction of the
anticipations of all entrepreneurs in the market, cause the costs of production
to conform to output prices regardless of the technical relationships that exist
in production. The spectrum of quality in entrepreneurial foresight determines
the speed and accuracy of this process, and also explains the pro�ts earned by
entrepreneurs during the adjustment, pro�ts that are missed by their less-astute
As explained above, the theory of costs is inextricable from the body of price
theory, which in turn provides the basis for analyzing production in both free
and hampered market economies. In Chapter
7, Xavier
Méra examines the latter
case, speci�cally, the problem of monopsony. Méra explains that although early
causal-realist writings hinted at the possibility of monopsony, later writers like
Mises and Rothbard largely dismissed the idea as unimportant. The reason is that
both older and newer monopsony theories fail to adequately distinguish between
monopsonistic and competitive prices. However, Méra argues that in their writ-
ings on monopoly the Austrian critics actually laid the foundation for a theory of
“monopoly price-gap” in which monopsony and monopoly prices are part of the
same phenomenon, namely, the hampered market economy.
Austrians are not alone in their criticisms of the mainstream economic approach
to costs. In Chapter
8, Mateusz
Machaj surveys some common ground between
Austrian and Post-Keynesian theories of price formation. Like Austrians, Post-
Keynesians are critical of the conventional view that �rms operate in practice by
equalizing marginal costs and marginal bene�ts. A
growing body
of empirical
research suggests that real-world managers do not make decisions according to
this rule. Machaj argues that this “business practice” critique is nothing more than
an alternate way of expressing how the Austrian theory of “imputation” plays out
within the �rm. This common ground means that some Post-Keynesian arguments
can strengthen Austrian critiques of mainstream price theory.
However, not all strands of economic thought are compatible with the causal-
realist approach. Mihai-Vladimir Topan argues in Chapter
9 that
the transaction
costs paradigm is one such. He examines the transaction costs literature, espe-
cially the work of Ronald Coase, in light of the two �elds in which it is most
successful: the economic analysis of property rights, and the theory of the �rm.
According to Topan, neither application of transaction costs is successful. The
reason is that the concept of transaction costs is both vague and based on a faulty
distinction between production and exchange, or between the �rm and the market.
As a result, transaction costs are at best helpful as heuristic devices in those lim-
Topan’s critique of the transaction costs theory of the �rm leads logically to
the question of economic organization. In this �eld, causal-realist price theory
provides a bridge between the theory of the �rm and the study of comparative
economic systems. Speci�cally, Mises’s theory of economic calculation can be
used to explain the unhampered market economy and the �rms within it as well as
Matthew McCaffrey
conditions under a system of socialist central planning. Mises’s famous
critique of socialism showed decisively that without genuine market prices based
on private property and the entrepreneurial division of labor, it is impossible for
central planners to accurately appraise the costs of their decisions, and therefore
to allocate resources to their most urgent uses (Mises, 1990).
In Chapter
10, Per
Bylund returns to Mises’s argument and elaborates one of
its vital distinctions: the difference between entrepreneurs and managers. Entre
preneurs, in their capacity as owners and decision makers, bear the uncertainty of
investing in the market in order to create value. Managers, on the other hand, are
limited to adjusting the technological conditions of production in order to ensure the
physical ef�ciency of production methods that have already been tried and tested
by entrepreneurs. The implication is that although unhampered market economies
will tend toward constant improvement in consumer welfare, socialist societies
are managerial and lack the ability to revolutionize production in an innovative,
entrepreneurial way. Consequently, they will be at best static and at worst
– and
more likely
– will
persistently decline in terms of their ability to improve consumer
The problem of economic calculation also applies outside the extremes of
purely for-pro�t enterprise and socialist central planning: calculation is also a
vital lens through which to view alternative forms of economic organization in
the market economy. In Chapter
Matthew McCaffrey examines one such
example: the growing �eld of social entrepreneurship. Although it attracts major
interest in management studies, social entrepreneurship has received scant atten-
tion from economists. This chapter
resolves this
oversight by placing the theory
of social entrepreneurship on an economic foundation. McCaffrey outlines the
economic meaning of social behavior and shows that conventional market entre-
preneurship is deeply social, while at the same time, social ventures are inevitably
bound up with some kind of pro�t motive. This implies that the line between
social and conventional entrepreneurship is not as clear as is sometimes thought.
Importantly, social enterprises must engage in economic calculation if they want
to survive in competitive markets. This means they must rely on external prices
for the goods and services they produce, as without them they cannot estimate the
It is not an accident that this book concludes with discussions of economic cal
culation; in many ways, Mises’s contributions in this �eld unify and complete
the work begun by Menger, Böhm-Bawerk, and many others who attempted to
explain the role of costs in individual action and in the social order. The following
chapters each attempt to develop a part of the systematic body of economic theory
built by these economists over the course of more than a century. By doing so, they
show clearly that the theory of costs in the causal-realist tradition is a valuable and
indeed a vital theoretical framework for understanding a wide range of economic
problems old and new. The tradition established by Menger thus continues to grow
and thrive through these and many other published works. However, only a tiny
portion of all potentially relevant writings can be discussed in these pages: due to
the limitations involved in assembling a collection such as this, and the enormous
scope of the theory of cost, some important topics have fallen by the wayside.
These include the law of comparative advantage and the theory of externalities, as
as studies of speci�c works like James Buchanan’s
and the many important papers included in his collection
(1981 [1973]). These may seem like unforgivable omissions; however, we can sat
isfy ourselves with the knowledge that these works already receive attention from
scholars in and around the Mengerian tradition, and hope that this trend continues.
The chapters included here offer an antidote to some of the worries plaguing
mainstream economic thought. The causal-realist approach they embody is espe
cially vital at a time when the economics profession is under attack for its lack of
realism and inability to address urgent problems at all levels of the economy, and
even to explain them to people outside its ranks. Fortunately, alternative frameworks
can and do encourage progress toward addressing each of these criticisms, and their
current success in doing so is a cause for optimism. To take only one example, it is
notable that in addition to their academic responsibilities, many of the contributors
to this book are also founders or leaders of thriving private organizations that take
the public teaching of economics as their mission, a task in which mainstream eco
nomics has little interest. In any case, it is our hope that this collection will further
develop research and teaching in the Austrian, causal-realist tradition, its engage
Matthew McCaffrey
May, 2017
Böhm-Bawerk, Eugen von. 1959.
Positive Theory of Capital
. Trans. George D. Huncke
and Hans F. Sennholz. South Holland, IL: Libertarian Press.
[1891]. “The Austrian Economists.” In
[1894]. “The Ultimate Standard of Value.” In
[1892]. “Value, Cost, and Marginal Utility.”
Quarterly Journal of Austrian
Buchanan, J. M. 1969.
Cost and Choice: An Inquiry in Economic Theory
. Chicago: Uni-
. New York: New York University Press.
Klein, Peter G. 2007. “Foreword.” In
Principles of Economics
. Auburn, AL: Ludwig von
“The Mundane Economics of the Austrian School.”
11 (3–4): 165–187.
Menger, Carl. 2007 [1871].
Principles of Economics
. Auburn, AL: Ludwig von Mises
Mises, Ludwig von. 1990.
Economic Calculation in the Socialist Commonwealth
. Auburn,
Salerno, Joseph T. 1999a. “Carl Menger: The Founder of the Austrian School.” In
Great Austrian Economists
. Ed. Randall G. Holcombe. Auburn, AL: Ludwig von Mises
“The Place of Mises’s Human Action in the Development of Modern Eco-
nomic Thought.”
Quarterly Journal of Austrian Economics
Confusion over the opportunity cost concept came to a head in the economics
profession with the publication of Ferraro and Taylor’s (2005) �nding that only
21.6% of economists could correctly answer an introductory-level question on
opportunity cost. Their study was based on a sample of PhD holders and PhD
students at an academic conference. The ensuing debate over the de�nition and
exposition of opportunity cost continues even today, including through a special
In the present chapter, we compare the ways the opportunity cost concept is
presented in mainstream and Austrian principles textbooks. We then review the
lines of argument in the instructive
Journal of Economic Education
which is taken as representative of broader debate in the economics profession.
We show that the causal-realist ordinal and subjective conception of opportu-
nity cost provides the clarity and consistency the symposium participants desire.
Moreover, the “value speci�cation” of opportunity cost they reject in favor of a
quantity speci�cation is not a true value speci�cation, but another quantity speci-
�cation. Finally, we address recent debate in Austrian literature on opportunity
costs. The aim of these discussions is twofold: (1) to show that the causal-realist
theory of opportunity cost does not suffer from the problems now being exposed
in the mainstream, and (2) to critique alternative views presented in the main-
stream, as well as some eclectic views close to Austrian literature.
The opportunity cost concept is fundamental in economics, and it is one of the
�rst principles taught in introductory-level economics courses. The starting point
Individuals choose how scarce resources are to be used as means toward the
attainment of ends. All action aims at attaining higher-valued ends by forsaking
lower-valued ends because the use of means toward one end implies that other
on opportunity cost theory
ends must
be forgone (Menger, 2007, p.
95). The
opportunity cost of any action,
then, is the value of the highest-ranked end forgone because of the action.
such, the opportunity cost concept is inseparable from the concepts of value and
action. The ordinality and subjectivity of preferences applies to both value and
Since value and cost only exist in human action, all costs are opportunity costs.
Accountants and economists may refer to different types of costs, like operating
costs or transaction costs, but these are only convenient terms for certain kinds
of opportunity costs, and they only make sense in view of a proper understand-
ing of opportunity costs. So-called transaction costs, for example, are ends an
actor might forgo just to participate in a transaction. Transaction costs are not
weighed in a categorically different way in action, because individuals ultimately
are exchanging a less desired state of the world for a preferred state of the world.
Actors consider everything they deem relevant, including “transaction costs,” and
�nally, a choice is made.
When an actor chooses, he does not simply select one good over others, but
over other ends. More appropriately, when he makes a choice, the actor
exchanges one state of the world for another.
Rothbard (2009) explains that these
states of the world are ranked on the actor’s scale of preferences:
all action involves exchange
– an
exchange of one state of affairs,
X, for Y, which the actor anticipates will be a more satisfactory one (and
therefore higher on his value scale). If his expectation turns out to be correct,
the value of Y on his preference scale will be higher than the value of X, and
state of the world includes speci�c consequences that extend beyond the
mere immediate and direct satisfaction of consumption, such as “the quenching of
thirst.” The actor anticipates some of these consequences, and, since each action
is forward-looking, anticipates some consequences that only
occur. Both the
satisfactions and other consequences are bound up in the state of the world he seeks
to attain in action. By consuming a beverage, for example, an actor anticipates
the removal of thirst, but also forgoes the removal of thirst that could have been
achieved by the same beverage consumed later, along with any other use of the
beverage. Also, he may accept the possibility that he could spill the drink or that
the caffeine in the beverage will make it more dif�cult for him to sleep that night.
These consequences and possibilities, including the possibilities that some of them
may occur jointly, are taken into the actor’s consideration of the action, and all of
, these consequences include, for the actor, knowledge of forgone
opportunities. This poses no problem for the logic of action and opportunity costs.
If an actor knows that pursuing one end, A, involves the sacri�ce of his next-
best alternative, B, then he also knows that if he had pursued B, then that too
would have involved the sacri�ce of A. Put another way, the knowledge of for-
gone opportunities is applied to each of the actor’s options, and therefore such
knowledge does
not necessarily increase or decrease the psychic pro�t for the
actor in the potential states of the world he may pursue. The same applies to
scenarios in which options are added to an actor’s choice set. Therefore, it cannot
be said that the knowledge of more options necessarily makes an actor worse off,
which is one of the criticisms of the opportunity cost concept made by Reisman
(1998) and Braun (2014), explored below.
Of course, actors may overlook some consequences. In fact, exchanging states
of the world is impossibly complex if
consequences are to be considered. As
a result, acting man only considers those consequences he deems relevant and
important enough to give him pause before he initiates the action and commits to
the outcome. Considering an action is also an action because an actor’s attention
and mental capacity to weigh all the various consequences of an action are lim-
ited. He must choose which consequences are relevant and important and forgo
dwelling on consequences that are not. In hindsight, the actor may regret that he
did not consider a particular consequence, but this is an inevitable fact for �nite
beings acting under uncertainty. The existence of regret does not pose any trouble
for the opportunity cost concept. Regret only exists in hindsight, while opportu-
The above causal-realist view of opportunity costs has been lost in many leading
textbooks in favor of a cardinal and objective conception. While a popular criti-
cism of the mainstream approach is that utility is treated in a cardinal and objec-
tive way,
this is not the direct cause of the confusion over opportunity costs in
textbooks. In fact, neoclassical utility functions are not even mentioned in the �rst
pages of principles-level textbooks. Instead, the confusion has a different source,
namely, that opportunity costs are presented alongside
production tradeoffs
the two terms are used interchangeably. This gives students a muddled under-
standing of the critical and fundamental concepts of value and opportunity cost.
Moreover, this presentation and false con�ation are maintained throughout the
same textbooks. The result is that, bereft of a correct understanding of opportunity
costs, students are ill-equipped to think critically as they learn about a wide range
of other important topics, including justi�cations for government intervention or
Consider two examples. In Mankiw (2014, p.
52), the
production possibilities
of a farmer and a rancher producing meat and potatoes are presented �rst and then
the concept of absolute advantage is introduced and applied. Next, as a stepping
Time spent producing potatoes, therefore, takes away from time available for
producing meat. When reallocating time between the two goods, Rose and
Frank give up units of one good to produce units of the other, thereby moving
along the production possibilities frontier. The opportunity cost measures the
trade-off between the two goods that each producer faces.
The natural
conclusion students reach is that opportunity costs are a part of the
physical relationships of production, and are determined by the producer’s tech-
Similarly, Krugman and Wells (2009) �rst introduce the concept of opportunity
cost using an example with a student deciding between two college courses (p.
Opportunity cost
is de�ned as “what you must give up in order to get an item you
want” (p.
7). The
de�nition does not clarify whether the “what” is the actor’s for-
gone satisfaction or a quantity of goods. Indeed, only ten pages later opportunity
cost is reintroduced along with production tradeoffs: “the slope of a straight-line
production possibility frontier is equal to the opportunity cost” (p.
27). Mankiw’
(2014) and Krugman and Wells’s (2009) treatments of opportunity costs are typi-
cal among popular textbooks. A
de�nition of
opportunity cost is offered in the �rst
few pages, where the most fundamental concepts of economics are introduced to
get students “thinking like economists.” This �rst de�nition is typically accom-
panied by an example of a choice a college student might make. Later, however
often in
the very next chapter, in fact
– when
production possibilities frontiers are
introduced, a new de�nition and application of opportunity cost is introduced as
The second de�nition of opportunity cost describes a physical production
In contrast to the typical mainstream approach, consider two examples from
principles textbooks by Austrian authors. Ritenour’s (2010) exposition leaves no
room for confusion. He presents a section of Helena’s value scale, in which �our-
less chocolate cake is preferred to buttermilk pancakes, which in turn is preferred
to orange roughy. Helena must choose how to use her stick of butter, which may
be employed in the preparation of any of those three dishes, but only one. She
chooses to allocate the butter toward the preparation of �ourless chocolate cake,
and Helena’s forgone enjoyment of the buttermilk pancakes is identi�ed as the
Economists refer to this
doing without
as a cost. In fact, they have a special
name for it:
opportunity cost
. Opportunity cost is the value of the alternative
that must be forgone as the result of choosing to achieve a certain end.
Murphy’s (2012) textbook, the opportunity cost concept is presented using
The cost of a particular decision is the value that Crusoe places on the most
important goal that he
be able to achieve, because of the decision.
Economists often drive home the point by using the longer term
, which they de�ne as
the subjective value placed on the next-best alter
textbook presents opportunity costs alongside production tradeoffs.
While Ritenour’s example involves Helena producing �ourless chocolate cake
with butter
as an input, it is clear that the opportunity cost is a consequence of her
choice, and that her opportunity cost is
the physical relationship between the
butter and all of Helena’s other production possibilities.
Mainstream confusion
A) Ferraro and Taylor spark debate
Confusion in the mainstream led to the embarrassing results of Ferraro and Tay-
lor’s (2005) survey, in which only 21.6% of economists could correctly answer a
question about opportunity costs from a principles-level textbook. The question
You won a free ticket to see an Eric Clapton concert (which has no resale
value). Bob Dylan is performing on the same night and is your next-best alter-
native activity. Tickets to see Dylan cost $40. On any given day, you would
be willing to pay up to $50 to see Dylan. Assume there are no other costs of
seeing either performer. Based on this information, what is the opportunity
A $0
The correct answer was supposed to be B, $10, though the most popular answer
was D, $50. The value of the Dylan concert is taken to be your willingness to
pay to see the Dylan concert, $50. However, by going to the Clapton concert,
you avoid paying the $40 to see Dylan, and “an avoided bene�t is a cost, and an
avoided cost is a bene�t” (p.
3). Therefore,
the $40 avoided cost is subtracted
from the $50 willingness to pay to arrive at $10, the opportunity cost of seeing
An Austrian economist would have to write in the correct answer because she
would not �nd it among the choices given. With the information given in the
question, a preference ranking could be constructed to �nd the actor’s next-best
Enjoying the
Clapton concert, knowing that you will not have to pay for the
Clapton concert and that you would miss
the Dylan
concert, tickets to which
Enjoying the
Dylan concert, knowing that you will have to pay $40 for a ticket
and that you would miss
the Clapton
concert, the ticket to which was a gift to you
In this
scenario, the ends are bundles of speci�c consequences in each “state of the
world” the actor is considering, including the knowledge of forgone opportunities.
The ends
are not simply “the enjoyment of the Clapton concert” and “the enjoy-
ment of the Dylan concert” because the actor attaches to the end that might be
attained all the characteristics and consequences he deems relevant. Mises (1998)
makes this clear in his famous analogy about the choice between capitalism and
socialism: “A
man who
chooses between drinking a glass of milk and a glass
of a solution of potassium cyanide does not choose between two beverages; he
chooses between life and death” (p.
676). Therefore,
the opportunity cost of see-
ing Clapton is the actor’s subjective value of the counterfactual course of events:
Ferraro and Taylor’s (2005) result sparked debate among economists, including
a special symposium in the
Journal of Economic Education
(2016). The sympo-
sium features an article by textbook author Michael Parkin (2016a) and commen-
taries by three critics. Parkin offers his preliminary thoughts on the issue, the three
critics respond, and then Parkin responds in turn. Parkin �rst suggests that there
are two ways to present opportunity cost: “a value speci�cation and a quantity
speci�cation” (Colander, 2016). In his response to the critics, Parkin concludes
that the confusion between the two speci�cations of opportunity cost is due to two
speci�cations of value. In the end, he unfortunately settles on the “quantity speci-
�cation,” which is in line with the confusing mainstream textbook presentation
outlined above, in which opportunity costs and production tradeoffs are presented
as the same concept. However, we show below that Parkin’s “value speci�cation”
is just another quantity speci�cation rather than a version of the more correct
B) Parkin’s “reexamination”
The �rst question Parkin (2016a) takes on is this: “Is opportunity cost an ambigu-
ous and arbitrary concept or a simple, straightforward, and fruitful one?” (p.
Parkin answers
the question as follows: “regrettably, opportunity cost is an
ambiguous concept, [.
.] because
two de�nitions are in widespread use. One of
the de�nitions is indeed simple, fruitful, and one that students can learn. The other
has the potential to be ambiguous” (p.
12). This
does not directly answer the ques-
– Parkin
appears to be claiming that opportunity cost is ambiguous because
there are two versions of opportunity cost, and one of them is ambiguous.
involve a forgone next-best alternative, but one is “a physical quantity forgone”
while the other is “a value forgone” (p.
13). Parkin
decides that the two speci�ca-
tions are equivalent in some cases but con�ict in others, depending on the problem
o compare the two de�nitions, Parkin constructs a mathematical model to
show that in competitive equilibrium, an opportunity cost in quantity terms is the
same in value terms. Parkin uses three ways of measuring value: a marginal rate
of substitution, prices, and marginal utilities (as �rst derivatives of utility func-
tions). This is an important departure from the casual-realist value speci�cation of
opportunity cost as a forgone end. The implication is that Parkin is not comparing
two meaningfully different de�nitions of opportunity cost. Both are objective,
– (1)
a quantity of means and (2) a willingness to pay,
which is a quantity of dollars, or in the case of a MRS, a quantity of the other
good. A
true value
speci�cation would involve two competing ends in relation to
each other in an ordinal preference ranking. Nevertheless, Parkin shows how the
“two” speci�cations under scrutiny are equivalent in competitive equilibrium but
Parkin then compares the capabilities of the physical quantity speci�cation and
the poorly-de�ned value speci�cation (measured in dollars) in satisfying various
purposes of the opportunity cost concept. For him, the opportunity cost concept
should be able to “do” the following: (1) help explain the fundamental economic
problem of scarcity, (2) convince students that cost is something more than dol-
lars of expenditure, (3) easily distinguish itself as the
alternative and
the forgone alternatives, and (4) provide a suitable input in rational choice
This exercise is an unfortunate example of a deeper problem in mainstream
economics, with roots in Friedmanite positivism: judging economic concepts by
their tractability in models (rational choice in this case) and their ability to pre-
dict human behavior instead of their truth and logical consistency with the idea
of choice. Parkin’s faulty view of value and choice notably clouds his view of a
value speci�cation’s ability to explain the fundamental economic problem:
Lesson one in the principles course is the insight that scarcity is the source of
all economic questions and problems, that faced with scarcity we must make
choices, and that in choosing, available alternatives are rejected or forgone.
The insight that cost is a forgone alternative can be gained without any notion
of value except in the vague sense that wanting something is synonymous with
valuing it. The quantity-forgone version of opportunity cost does a good job
of deepening this insight. How does the value-forgone version perform in this
task? First, it brings an added layer of complexity that obscures the insight.
Second, it requires a lengthy detour to de�ne and explain the concept of value.
The quantity-forgone version wins on this �rst purpose of opportunity cost.
Parkin would like to avoid a “lengthy detour” when teaching the funda-
mental concept of value, we have shown that, pedagogically, a �rm grasp of value
The next purpose of opportunity cost in Parkin’s list is that it should help con-
vince students that cost is not simply dollars of expenditure. Since Parkin and the
majority of the economics profession express value in dollar terms, Parkin again
selects the physical quantity speci�cation as the clearer approach: “The value-
forgone version again obscures the insight, and it especially obscures it when
value is expressed in dollars rather than other goods willingly forgone” (2016a,
20). Here,
again, a true value speci�cation of opportunity cost survives Parkin’s
criticism. In causal-realist value theory, value is never expressed in dollar terms.
Opportunity costs are forgone ends, which are immaterial and subjective.
true value speci�cation would also satisfy Parkin’s third purpose of the
opportunity cost concept – that it is easily distinguishable as only the next-best
alternative and not all the forgone alternatives. He explains that it is sometimes
dif�cult to single out the one forgone alternative and that deciphering the value of
that forgone alternative merely adds extra steps to the process. Parkin once again
favors the quantity speci�cation: “Figuring out
is forgone is the key step
and is better not sidetracked by the harder task of �guring out
is forgone. Again, the physical quantity version wins” (2016a, p.
21, emphasis
original). It is clear that the extra layer of dif�culty added by his value speci�ca
tion derives from the way value is measured and not the fact that it is a value
per se.
For example, given that in causal-realist explanations of
value and choice, the ends that are ranked by an actor are often represented in
shorthand by the means that the actor would employ to attain those ends, Parkin’s
“quantity speci�cation” is actually closer to the causal-realist value speci�cation
than his own dollar-denominated “value speci�cation.” The only difference in
this case is an understanding that the physical goods as stated represent ranked
ends for the actor.
Parkin’s �nal purpose suggests that opportunity costs should be easily inputted
into cost-bene�t models. He �nds that both the quantity and value speci�cations
are suited for this task, as long as the costs and bene�ts are denominated in the
same way (units of a good, dollars, utils, etc.). He adds that if value is measured
in dollars, the quantity speci�cation still wins on the grounds that “it cuts through
the veil of money” (2016a, p.
21). While
the causal-realist conception of oppor-
tunity cost is not tractable in neoclassical rational choice models, the problem
may be with the model itself and not with the subjective and ordinal conception
of opportunity costs. Here it becomes apparent that the mainstream and Austrian
economists work in opposite directions when faced with dilemmas over concepts
like opportunity cost. Mainstream economists like Parkin use their models to put
a menu of opportunity speci�cation options through trials to see which one per-
forms best. In contrast, Austrian economists, when faced with a similar dilemma,
work from the ground up such that any new or revised concepts still comport with
The �rst critic, Daniel Arce (2016), generally agrees with Parkin and suggests that
economics textbooks simply need more examples of the opportunity cost concept
in application. The third critic, Daniel Stone, like Arce, is not critical at all, and
even regrets that he began by thinking to doubt such an authority on the issue as
Only the second critic, Rod O’Donnell (2016), offers meaningful criti-
cism of Parkin’s arguments for a quantity speci�cation of opportunity cost.
O’Donnell (2016) arrives at a similar conclusion to the one outlined above
regarding textbook treatments of opportunity cost: “In my reading of textbooks,
it is not uncommon to �nd value de�nitions of OC [Opportunity Cost] followed
by physical quantity examples and applications” (p.
27). O’Donnell
(2016) also
comes close
to recognizing that willingness to pay is just another objective, quan-
tity measure: “Bene�t is then measured by the maximum amount the agent is
willing to give up to get X, so linking bene�t to willingness to pay (WTP), an
idea aimed at rendering objective that which is subjective” (p.
28). O’Donnell
gues that Parkin, by using two- and three-good models, assumes what he needs
to prove, especially given that Parkin’s three-good models reduce to two-good
models because the third good is never an object of choice and so becomes a unit
of account for the other two goods. The exercise is therefore fruitless because the
opportunity costs in a two-good world are always a quantity of the good not cho-
sen (O’Donnell, 2016, p.
29). Finally
, O’Donnell disagrees with each of Parkin’s
conclusions about the quantity speci�cation’s ability to satisfy the various pur-
poses of the opportunity cost concept (O’Donnell, 2016, pp.
29–31). In
his reply,
Parkin (2016b) attributes the bulk of their disagreements to miscommunication
about how value is measured: willingness to pay or market prices. After emphati-
cally declaring for the former (“
Value is willingness to pay
,” p.
35, his
italics), he
doubles down on his original conclusion that the quantity speci�cation of oppor-
tunity cost should be adopted by all economists and that “there is no theoretical
The symposium authors do not mention any problems with presenting opportu-
nity costs along with production tradeoffs as the same fundamental concept. Yet,
pedagogically at least, the two should not be introduced together. However, it
should be noted that there is a way to appropriately identify opportunity costs in
an example involving production possibilities. Consider the classic guns and but-
ter tradeoff. Producing more of one good means less of the other can be produced,
and vice versa. Production tradeoffs present a set of options for the actor, such as
(A) �ve guns and zero pounds of butter, (B) four guns and three pounds of butter,
(C) three guns and �ve pounds of butter, etc. One can only claim, “the opportunity
cost of three pounds of butter is one gun,” with important caveats: (1) referring to
the goods chosen and forgone is shorthand for the ends attained by the actor with
those goods (in their respective bundles), and (2) the actor must have chosen B,
four guns and three pounds of butter, over A, �ve guns and zero pounds of butter.
The actor’s preferences must be:
That is,
the next highest ranked alternative must be the choice that is consistent
with the opportunity cost as stated. Going from A
to B,
the actor gains three pounds
of butter, but loses one gun, and this is consistent with the actor’s preferences.
To be clear, it would make more sense pedagogically to �rst present the actor’s
preferences and his options before identifying opportunity costs. This is not com
mon practice in mainstream textbooks, though, and the caveats are also withheld.
Unfortunately, students are left with two separate but falsely con�ated ideas.
Critical views in Austrian literature
Other misconceptions about opportunity costs have emerged in literature close to
Austrian economics. The most notable example is found in the work of George
Reisman (1998). Reisman, along with several of his students, denies what he calls
the “opportunity-cost doctrine.” He presents multiple scenarios involving missed
opportunities for monetary gain to show how allegedly absurd opportunity cost
is. Below, we will see that his critique is based on a misunderstanding and misap-
plication of the concept. We will also cover a scenario offered more recently by
Eduard Braun in his book,
Finance Behind the Veil of Money
(2014), which cites
Reisman on the same issue. Brie�y, the answers to their criticisms are that oppor-
tunity costs cannot be identi�ed in hindsight and that opportunity costs may only
, Reisman begins his critique of the “opportunity-cost
doctrine” by describing the version he is targeting, namely, the one used by Samu-
An opportunity cost is an imputed cost
– a
cost which does not actually exist
in the sense of an expenditure of money being made, or having been made,
but which is treated as though it existed. An opportunity cost is said to exist
by virtue of the failure to earn a revenue or income that otherwise might be
earned or might have been earned. It represents the absence of a revenue or
continues to point out absurdities in using opportunity costs to analyze
various hypothetical scenarios in which an actor makes a choice.
However, his
scenarios are analyzed, and the opportunity costs identi�ed, only in hindsight.
Viewing opportunity costs in hindsight is categorically different from weighing
opportunity costs before an action is taken. A
similar point
is made in Howden’s
(2016b) rejoinder to Braun (2016), as part of their back-and-forth in the
Journal of Austrian Economics
. Opportunity costs viewed in hindsight have no
bearing on the logic of action
– only
ex ante opportunity costs matter in economic
Of course, hindsight can play a role in an actor’s future decision-making,
but this
simply takes past information as a basis for making a new, forward-looking
We cannot use what could have happened, given knowledge of the course
of events and the present state of the world, to describe opportunity costs, else
every dollar spent would carry the “opportunity cost” of the winning lottery
ticket or the best-performing stock. Yet this is how Reisman treats the opportu
nity cost concept in his �rst example, in which the owner of the neighborhood
hardware store earns $50,000 in pro�t over the previous year. A
nosy economics
textbook author tells him he could have sold the store and earned $15,000 from
the invested funds and $45,000 in wages working elsewhere. Reisman says that
the “opportunity-cost doctrine” must then say that the hardware store owner,
far from earning $50,000, has actually lost $10,000.
However, this missed
opportunity is only seen in hindsight, and cannot have been the store owner’s
forward-looking opportunity cost (assuming the store owner only cares about
money pro�t). Even if the store owner knew about the $60,000 opportunity that
would have involved him selling his business and still passed on it, there must
have been nonmonetary reasons for his choice. The $10,000 difference might
be considered his willingness to pay for the state of world in which he retains
ownership of the store. Either way, there is not a legitimate way to use the oppor
tunity cost concept to arrive at Reisman’s conclusion that the store owner has
incurred a loss of $10,000. The costs of production are reckoned by the entre
preneur at the time of his �nal decision to produce, as factors are committed to
the production process.
Hindsight may reveal to the entrepreneur that he could
have earned more had he sold his inputs, but this information can only be used
The opportunity cost of the hardware store owner’s decision to continue his
business for another year was the next-best alternative in his mind at the time he
made the choice. Afterwards, of course, he may look back and determine whether
he made a “good choice”
– perhaps
it turns out later that what he thought was
next-best actually would have been better than the option he chose.
And, of
course, he can also expand his consideration of the choice beyond the options
he thought he had at the time of the decision. This is where the lottery ticket or
the great stock pick would come into play. Importantly, though, none of the store
owner’s reminiscing, new information, or regret changes the opportunity cost
. Action is always forward-looking, and so opportunity costs
(as they are used and de�ned in economic theory) may only be identi�ed in the
same manner.
Let us consider another one of Reisman’s (1998) examples, this time involving
Yet another manifestation of the absurdity of this doctrine can be seen if it is
applied to the stock market. Imagine that an individual is considering invest-
ing a
million dollars,
and must decide between two stocks, A
and B.
stocks are currently selling at $10
per share.
The individual decides on stock
A. It goes to $20
per share.
In the same period, however, it turns out that
stock B
goes to $30
per share.
If one believes the opportunity cost doctrine,
this is grounds for leaping from the nearest skyscraper
– one
has lost a
should be immediately clear that Reisman has once again incorrectly identi�ed
the opportunity cost by viewing the problem in hindsight. The investor may real-
ize after the fact that she could have made more money by investing in B, but this
does not change her opportunity costs at the time she made the choice. Reisman
takes Samuelson and Nordhaus’s (1989) version of opportunity cost as represent-
ative of that of all proponents of the opportunity cost concept, and he criticizes
it in full. Yet, from the perspective of the causal-realist approach to opportunity
costs, this is simply tilting at windmills. The opportunity cost concept on Reis-
man’s chopping block is not the causal-realist subjectivist conception. Rothbard
(2009) makes this explicit when he explains that “[opportunity] costs are subjec-
tive and cannot be precisely determined by outside observers or be gauged ex post
over Reisman’s controversial position on opportunity costs emerged in the
Quarterly Journal of Austrian Economics
with David Howden’s review (2015) of
Eduard Braun’s book,
Finance Behind the Veil of Money
(2014). Braun maintains
Reisman’s line of thinking on the opportunity cost concept and, much like Reisman,
offers a scenario with the purpose of showing how absurd it is that an actor would
suffer more or higher costs when faced with more options. The example involves
a hiker who is gifted an apple by a friend. The friend allows the hungry hiker to
choose between two identical apples, which, according to Braun, means that choos
ing one involves incurring the opportunity cost of the other. Thus, the net bene�t for
uridan’s ass issue.
Howden (2016a, 2016b) presents valid objections to Braun’s argument based
on a solution to the Buridan’s ass problem (namely, �nding the missing alterna-
tive) along with the proper application of opportunity cost as a forward-looking
concept in action. However, another objection may be raised regarding the hiker’s
supposed dilemma. Braun presents two choices as one, which prohibits any clear
identi�cation of opportunity cost. The two choices are (1) to accept or reject the
gift of the apple and (2) to choose one apple of the two for consumption. The hiker
clearly bene�ts by the gift of the apple, even though the giver is not clear about
which apple is the gift. We can consider the gift as an apple coupon, redeem-
able for one apple. The hiker demonstrates a preference for an apple over no
apple by accepting his friend’s gift. The second choice (which apple to take) is
the same type of choice an individual makes whenever he chooses one unit of a
homogenous stock
– a
kind of low-stakes choice the practically everybody makes
all the time. I
do not
toil over which edge of my coffee mug from which to take
a sip or what particular M&M happens to fall into my hand when I
pour out
the bag.
Attention is scarce and valuable, so we voluntarily allow some random-
ness, impulse, and spontaneity to affect which units of a homogeneous stock are
selected for consumption and which are saved for later. To the extent that these
However, to the extent that a unit of an apparently homogeneous stock is con-
sciously chosen over other units, it means that the actor must have noticed some
heterogeneity among the units. An actor may consciously choose one unit simply
because it is on top of the heap, or nearest to the actor.
These characteristics may
enter the actor’s choice without posing a problem for the fundamental concepts
of value, action, and opportunity cost. This is the case even when all the units are
perfectly identical from the perspective of an outside observer. In economics, the
physical characteristics of any good must �rst pass through the actor’s subjective
interpretation and judgement before a choice is made and analyzed, including
those characteristics an outside observer may deem negligible or does not notice
Recent literature has brought many misconceptions regarding opportunity costs
to light, both in mainstream and Austrian circles. The current debate over the
opportunity cost concept in the mainstream is mired in a false choice between
two quantity speci�cations, yet the larger problem is that mainstream principles
textbooks present opportunity costs and production tradeoffs as one and the same.
Students, teachers, and textbook authors would do well to consider a more intui-
tive approach to value, choice, and action from the causal-realist Austrian school,
such as the one adopted by Ritenour (2010) and Murphy (2012). At the same
time, though, George Reisman and others close to the Austrian literature have
criticized the opportunity cost concept. Yet it has been shown that their objections
do not stand up to scrutiny, though the debate sparked by Reisman and continued
by Braun has been a healthy exercise that hopefully will lead to greater clarity for
Mises (1998) provides a complete exposition of the central place of human action in
Mises (1998):
“Costs are equal to the value attached to the satisfaction which one must
forego in order to attain the end aimed at” (p.
97). Rothbard
(2009): “
is simply
the utility of the next best alternative that must be forgone in any action, and it is there-
fore part and parcel of utility on the individual’s value scale” (p.
This conclusio
n is surprisingly uncontroversial. Consider, for example, Krugman and
Wells (2009): “The concept of opportunity cost is crucial to understanding individual
As Mises
(1998) explains: “Acting man is eager to substitute a more satisfactory state
of affairs for a less satisfactory” (p.
13). This
insight is particularly useful in solving
apparent conundrums in the logic of action. For example, McCaffrey (2015) uses it to
address a separate criticism of causal-realist theory regarding love and gifts: “in the
universal, praxeological sense, it is not material goods that are exchanged, but rather
(1956) “Toward a Reconstruction of Utility and Welfare Economics” is a
See Boyes
(2014) for an example of how opportunity costs may be applied in a critique
of the Keynesian multiplier.
For other
examples, see Acemoglu, Laibson, and List (2015, pp.
9, 12,
173), Chiang
(2017, pp.
9, 36),
Coppock and Mateer (2017, pp.
13, 35),
Cowen and Tabarrok (2015,
4, 17),
Hubbard and O’Brien (2015, pp.
8, 39),
Parkin (2015, pp.
9, 33),
and Taylor
for example, the two de�nitions offered in Chiang (2017): “Opportunity cost
measures the
of the next best alternative use of your time and money, or what you
give up when you make an economic decision” (p.
9, emphasis
mine); and “Oppor-
tunity cost
– The
cost paid for one product in terms of the
(or consumption)
another product
that must be forgone” (p.
Perhaps he
means that
the state of opportunity cost in the economics profession
ambiguous because there are two commonly used versions of opportunity cost, and
Herbener (201
1) summarizes the relationship between means and ends in the mind
of an actor: “Since attaining the end is the purpose of an action, the value a person
attaches to the attainment of the end is primary. A
person attaches
only derivative value
to the means used in action since they are merely aids to the attainment of the end.
Means have no value independent of the value a person attaches to the end they help
attain. The human mind imputes value to the means according to the aid they render in
Stone explains:
admit it
was my instinct to defend the value de�nition (i.e., my own
position, or at least that of my earlier paper), searching for �aws in Parkin’s piece and
arguments against it. As I
read and
thought about his article more carefully, however,
realized what
should have been obvious right away: Parkin is certainly knowledge-
able on this topic, and his arguments are well-founded” (Stone, 2016, p.
Some of
the scenarios Reisman brings up are quite comical, and intentionally so: “An
analogy to this procedure would be the following. One gains ten pounds, but might
have gained twenty pounds. This is then taken to mean that one has lost ten pounds.
When one’s alleged loss of weight cannot be reconciled with the fact that one is now
ten pounds too large for one’s clothes, one’s oversize is explained on the grounds that
one’s clothes have shrunk the equivalent of twenty pounds” (p.
Howden (2016b)
suggests that these are two “uses” for the opportunity cost concept,
“The second
use of opportunity costs is an ex post facto assessment to determine
if the chosen option was the correct one” (p.
184). While
agree with
for the sake of clarity I
hesitate to
adopt the same language. Missed opportuni-
ties in hindsight are categorically different from forward-looking, anticipated opportu-
nity costs, and only the latter are relevant to economic theory.
Reisman (1998),
459: “These
forgone opportunities or passed-up alternatives, the
textbook author then argues, must be counted as costs of the owner’s business, just as
much as the store’s payment for merchandise and the labor of hired help, if its actual
For more
on the entrepreneur’s anticipations of the costs of production through the pro-
duction process, see Jeffrey M. Herbener’s (2018) contribution in the present volume.
Once again,
Rothbard (2009) is indispensable: “It is convenient to distinguish the two
vantage points by which an actor judges his action as
ex post
his position when he must decide on a course of action;
it is the relevant and dominant
consideration for human action
. It is the actor considering his alternative courses and
the consequences of each.
Ex post
is his recorded observation of the results of his
past action. It is the judging of his past actions and their results.
Ex ante
, then, he will
always take the most advantageous course of action, and will always have a psychic
pro�t, with revenue exceeding cost.
Ex post
, he may have pro�ted or lost from a course
This is
why, in any telling of the Buridan’s ass problem, the oases or hay bales are
Acemoglu, Daron, David Laibson, and John List. 2015.
. New York: Pear-
Arce, Daniel G. 2016. “Opportunity Cost and the Intelligence of Economists: A
William J. 2014. “The Keynesian Multiplier Concept Ignores Crucial Opportunity
Quarterly Journal of Austrian Economics
Braun, Eduard. 2014. Finance Behind the Veil of Money: An Essay on the Economics of
Capital, Interest and the Financial Market.
“Reply to Dr.
Howden on
Opportunity Costs.”
Quarterly Journal of Aus
Chiang, Eric. 2017.
Economics: Principles for a Changing World
. 4th ed. New York:
Worth Publishers.
Colander, David. 2016. “Introduction to Symposium on Opportunity Cost.”
The Journal of
47 (1): 11.
Coppock, Lee, and Dirk Mateer. 2017.
Principles of Macroeconomics
. 2nd ed. New York:
W.W. Norton
Tyler, and Alex Tabarrok. 2015.
Modern Principles of Macroeconomics
. 3rd ed.
New York: Worth Publishers.
Ferraro, Paul J., and Laura O. Taylor. 2005. “Do Economists Recognize an Opportunity
Cost When They See One? A
Dismal Performance
from the Dismal Science.”
tions to Economic Analysis
, Jeffrey M. (ed.). 2011.
The Pure Time-Preference Theory of Interest
. Auburn,
“Time and the Theory of Cost.” In Matthew McCaffrey, ed.,
The Eco
nomic Theory of Costs: Foundations and New Directions
. Abingdon, UK: Routledge,
David. 2015. “Review of ‘Finance Behind the Veil of Money: An Essay on the
Economics of Capital, Interest, and the Financial Market, by Eduard Braun’.”
Journal of Austrian Economics
Finance Behind the Veil of Money
: Response to Dr.
Quarterly Journal of Austrian Economics
“Finance Behind the Veil of Money: A
Quarterly Journal of
Hubbard, R. Glenn, and Anthony P. O’Brien. 2015.
. 5th ed. New York:
Krugman, Paul, and Robin Wells. 2009.
. 2nd ed. New York: Worth
Mankiw, N. Gregory. 2014.
Principles of Macroeconomics
. Mason, OH: Cengage
McCaffrey, Matthew. 2015. “What’s Love Got to Do with It? Action, Exchange, and Gifts
in Economic Theory.”
Quarterly Journal of Austrian Economics
Menger, Carl. 2007 [1871].
Principles of Economics
. Trans. J. Dingwall and B. F. Hoselitz.
Auburn, AL: Ludwig von Mises Institute.
Mises, Ludwig von. 1998 [1949].
Human Action: Scholar’s Edition
. Auburn, AL: Ludwig
Murphy, Robert P. 2012.
Lessons for the Young Economist
. Auburn, AL: Ludwig von Mises
O’Donnell, Rod. 2016. “Complexities in the Examination of Opportunity Cost.”
The Jour
. 12th ed. New York: Pearson Higher Education.
“Opportunity Cost: A
The Journal
of Economic Educa
“Opportunity Cost: A
Journal of Economic Education
47 (1):
Reisman, George. 1998.
Capitalism: A
eatise on Economics
. Chicago, IL: Jameson
Ritenour, Shawn. 2010.
Foundations of Economics: A
Christian V
. Eugene, OR: Wipf
Rothbard, Murray N. 1956. “Toward a Reconstruction of Utility and Welfare Economics.”
In Mary Sennholz, ed.,
On Freedom and Free Enterprise: Essays in Honor of Ludwig
. Princeton, NJ: D. Van Nostrand, pp. 224–262.
Rothbard, Murray N. 2009.
Reexamination’: A
Case in
aylor, John. 2007.
Principles of Macroeconomics
. 5th ed. Boston, MA: Houghton Mif�in.

The distinction between the substitution and income effects introduced into neo-
classical price theory by Hicks (1946) in 1938 has been a �xture in mainstream
This distinction has been conspicuously absent from the causal-realist tradi-
tion in price theory, which is a variant of neoclassical economics that traces back
to Carl Menger.
The causal-realist analysis of the individual demand curve pro-
ceeds exclusively in terms of the elaboration of the law of marginal utility against
Unfortunately, there has never been an explicit comparison of the causal-realist
and mainstream neoclassical approaches to deriving the demand curve. In his
, Mises (1998) never mentioned the Hicksian approach.
In a memo written in 1961, Rothbard (Salerno, 2011, p. 14) brushed aside “the
alleged division between the ‘income effect’ and the ‘substitution effect,’ and the
consequent ‘Giffen Paradox.’
” In
his treatise, published a year later, Rothbard
(2009, p.
915) gave
only a cursory mention of the two effects in a discussion
of income taxation. More recently, there has been an awakening of interest in
the question of the existence of the income effect in causal-realist price theory.
The recent treatments of this topic are widely scattered in time and broadly dis-
parate in depth and focus (Caplan, 1999; Salin, 1996; Gonzalez, 2000; Rajsic,
2010; Hudik, 2011). More important, while of some value, these treatments never
achieved resolution of the issue, because they failed to recognize the profound
differences that have developed since the 1930s between Mengerian causal-real-
ist price theory and the Marshallian-Walrasian-Hicksian synthesis at the heart of
In this paper I
will ar
gue from an explicitly causal-realist perspective that the
income effect is illusory and that the individual demand curve can be fully deduced
from the law of marginal utility, which is the logical implication of the existence
of individual value scales (Mises, 1998, pp.
94–96, 1
19–127). In addition, I
attempt to
show that the demand curve deduced in this way naturally integrates
money into the analysis and explains the determination of the moment-to-moment
structure of prices actually paid in real-world transactions. Causal-realist demand
The “income effect” in
causal-realist price theory
Joseph T. Salerno
Joseph T. Salerno
is thus able also to provide an integral explanation of the purchasing
power of money that does not invoke the (neo)classical dichotomy between real
and nominal variables that arti�cially separates “relative” prices from the “abso-
In section
2, I
brie�y review
the controversy over the nature and assumptions of
the Marshallian demand curve that was carried on during the 1950s among neo-
classical economists and was never satisfactorily resolved. Section
3 addresses
centrality of ordinal scales of value or preference to demand analysis. In sec-
4, I
that, for the causal-realist, the stock of money assets, and not the
�ow of money income, is the relevant concept for demand theory. The case is
made in section
5 that
constancy of the purchasing power of money is an indis-
pensable assumption for using the law of marginal utility, which is derived from
the ordinal value scale, in deducing the individual demand curve. Section
6 draws
distinction between the “instantaneous” demand curve yielded by causal-realist
analysis and the temporally ambiguous conventional demand curve. The nature of
the substitution effect in causal-realist demand theory is clari�ed in section
7. Sec-
8 presents
some concluding remarks on the income effect and the backward-
bending supply curve of labor.

The neoclassical controversy over
The income effect was the central problem in what Leland Yeager (1960) referred
– Friedman
calls them “unrelated goods”
– must
inversely as price changes along the demand curve. In other words, at each point
on the demand curve buyers face a different con�guration of prices of the goods
The “income effect” in causal-realist price theory
unrelated to
the good in question. These con�gurations are conceived to vary in
such a way that the purchasing power of money and, therefore, its marginal utility
to buyers, remain unchanged along the entire demand curve. Thus, for example, a
fall of the price of a good and the accompanying increase in its quantity demanded
re�ect exclusively the decline of its own marginal utility relative to the marginal
utilities of all other goods. The altered price-quantity position on the Friedmanite
demand curve is therefore completely unaffected by forces that would increase
“real income” by raising the purchasing power and thereby lowering the marginal
Marshall (1920, p.
62) indeed
emphasized the central role played by �xity of
If the purchasing power of money rises with regard to some things, and at the
same time falls equally with regard to equally important things, its general
purchasing power
. has
remained stationary
. [W]e
may throughout the
volume neglect possible changes in the general purchasing power of money.
Thus the price of anything will be taken as representative of its exchange
Marshall explicitly incorporated these assumptions about money into his state-
ment of the “law of marginal demand price,” thus banishing the income effect
Friedman, however, preferred an alternative formulation of the
ceteris pari
assumptions of the demand curve ([1949] 1970, pp.
51–52), which
shall himself adopted in the Mathematical Appendix to the
838–858). In
this variant, Marshall replaced the assumptions about the con
stancy of money income and the purchasing power of money with the assump
tion of a constant real income while additionally assuming a constant
of all other prices. To maintain real income unchanged under this set of assump
tions implies compensating variations in buyers’ money income that offset the
income effect of the variation of the price of the good. Friedman opts for this
variant of maintaining constant real income because “it is somewhat more con
venient mathematically” while “essentially equivalent mathematically” to the
constant purchasing power of money/constant money income variant ([1949]
e do not have to probe very deeply into Friedman’s methodological rationale
for preferring the constant real income demand curve to the Hicksian constant
money income demand curve. Suf�ce it to say that the Friedmanite interpretation
Joseph T. Salerno
the demand curve re�ected his unique brand of positivism. As Yeager per-
ceptively pointed out, for Friedman and the positivists, the income compensated
demand curve was an attempt to deal only with “actually or conceptually attain-
able positions of equilibrium,” and to derive “comparative static predictions”
about actual price-quantity equilibria that will appear in the market (Yeager, 1960,
In general, demand curves are useful only because they allow some pre-
dictions to be made regarding the effects of changes in supply. Hence each
point on the demand curve must represent an attainable equilibrium between
demand and supply. It is extremely dif�cult to imagine any other possible use
eager rejected the positivist view of the demand curve and argued that demand
curves are not to be regarded as “objectively existing things” but rather as “a
pedagogical device for helping students grasp the inverse price-quantity relation
asserted by the Law of Demand” (1960, pp.
63, 64).
In addition, Yeager (1960,
58) pointed
out that the positivist conception of the demand curve precludes
consideration of disequilibrium positions and the equilibrating processes that they

Unlike Friedman, Buchanan, and the Chicago school, the causal-realist rejection
of the income effect does not invoke positivist considerations. Indeed, quite the
contrary: in formulating the demand curve, the causal-realist approach seeks to
illustrate the logical implications of the existence of subjective value scales and
the law of marginal utility for price formation. Thus, like Yeager, economists fol-
lowing this approach conceive the demand curve not as a real thing “out there” in
the world, but as a hypothetical construct that assists in grasping the reality of the
What further distinguishes the causal-realist approach from the Hicksian and
Friedmanite approaches to the demand curve and to price theory in general is its
paramount aim of explaining the determination of
prices actually paid on
the market. All of the concepts, techniques, and constructs used in causal-realist
price theory are shaped by this overriding goal. In a market economy, relative
prices are embedded in the structure of money prices, a structure that emerges
anew on the market at every moment. Thus, money prices that are actually (or
anticipated to be) paid are the prices relevant for the economic calculations of
capitalist-entrepreneurs seeking to appraise and erect a complex structure of
heterogeneous capital goods. Analyzing a �ctional economy without money or
where money is neutral is merely a preliminary step
– albeit
a necessary one
– in
The “income effect” in causal-realist price theory
the development
of a realistic theory of price (Mises, 1998, pp.
202–206). Causal-
price theory thus treats money as an integral part of the valuation process
The analysis of the demand curve begins with the individual’s scale of values or
preferences, which is the cause of all economic phenomena in a real and profound
Like “choosing” and “preferring,” the value scale is part of what we may
call the “ordinary language” of action. It is not a psychological category whose
existence is established by introspection. Without the concept of a scale of values,
it would be impossible to even describe an acting being or speculate meaningfully
The scale of values is a strictly personal and ordinal ranking of goods and
money that determines and reveals itself in every individual’s choices and actions
on the market. The interaction of individual value scales thus determines the
“equilibrium” prices and quantities exchanged at every instant in all markets.
This momentary equilibrium position denotes a state in which all consumers allo-
cate expenditures across goods so that the marginal utility of the last unit of each
good purchased just exceeds the marginal utility of the sum of money expended
for its price (Wicksteed [1933] 1957, 1: pp.
141–145, 212–233;
Rothbard, 2009,
way of expressing the same situation is to say that the price and quan
tity exchanged of each good is determined at the intersection of its demand and
supply curves. Although the language of demand and supply is more abstract and
less precise than that of value scales, it is useful in simplifying and abbreviating
the analysis and in highlighting the implications of the operation of the law of
marginal utility in the pricing process. This important point has been emphasized
by causal-realist price theorists. For example, Böhm-Bawerk (1959, p.
still �nd it questionable whether, with its resulting unavoidable suppres-
sion of any personal point of view, this method of presentation [by demand
and supply curves] is really capable of supplanting and making super�uous
a description by running commentary of the determination of price [i.e., in
terms of the marginal pairs]. [I]t has been my personal opinion that the run-
Mises also recognized the primacy of the ordinary language of value scales for
The ultimate source of the determination of prices is the value judgments of
the consumers. Prices are the outcome of the valuation preferring
If one
says that prices tend toward a point at which total demand equals total
supply, one resorts to another mode of expressing the same concatenation of
Joseph T. Salerno

Money income or money stock?
Because it seeks to explain prices as the outcome of a unitary valuation process
that includes money, causal-realist price theory holds the following data constant
in deriving the individual demand curve: 1. the buyer’s value scale; 2. his money
balances; 3. all other prices; and 4. the (anticipated) purchasing power of money.
“Value scales” are used in preference to the more usual “tastes and preferences”
because the former expression is richer and encompasses the utility rankings of
money as well as goods, in contrast to the conventional expression that assumes
the absence of money or its use purely as an unvalued
. Assumed con-
stancy of individual money holdings thus supplants the assumption of constant
money income (Hicksian) or real income (Friedmanite). This revision is neces-
At the moment before any set of exchanges is consummated all that objectively
exists are given individual stocks of goods and money.
These stocks were accu-
mulated by their owners from a series of discrete exchanges in past factor and
product markets. The current marginal utilities ascribed to these goods and money
balances are
determined by their relative positions on individual value
scales. Thus, neither money income nor real income is a direct determinant of the
demand curve. Of course, this is not to deny that expectations of future “income”
or, more properly, the anticipated pattern of future prices indirectly in�uence the
value rankings and marginal utilities of current stocks of money and goods. But
“income” is not an observable and concrete historical outcome of the market pro-
cess like a money price or the de�nite quantity of goods for which it is paid. All
action takes place in the present moment and all economic action refers to the
deployment of existing
of goods and money according to instantaneously
The notion of (net) income as a “�ow” is the outcome of the individual entre-
preneur’s judgment of a recorded sequence of concrete transactions during a de�-
nite period of the past; or it may refer to his summary appraisement of quantities
of goods or money that will accrue from discrete acts of exchange expected to
take place during a relevant future time period. In either case, it is the product of
a subjective judgment, because income in economic theory exists on a different
plane of abstraction from realized prices and present stocks of goods and money.
The concept of a stock of goods or a realized price is a �rst-order abstraction
of an observable phenomenon referring to an objective result of valuation and
action. The twin concepts of income and capital are, in contrast, derived abstrac-
tions referring to unobservable mental categories used by the actor in calculating
the costs and returns of alternative uses of the objective means of action. These
categories are employed in the intellectual process of economic calculation to
establish a quantitative distinction that enables capitalist-entrepreneurs and factor
owners to net out the consumable product from the gross revenues of their pro-
ductive activities and to thereby maintain intact the capital value of their resources
The “income effect” in causal-realist price theory
Mises emphasize
d the crucial conceptual distinction between prices and income
A market price is a real historical phenomenon, the quantitative ratio at which
at a de�nite place and at a de�nite date two individuals exchanged de�nite
quantities of two de�nite goods. It refers to the special conditions of a con-
crete act of exchange. It is ultimately determined by the value judgments
of the individuals involved
The market
does not create or determine
incomes. It is not a process of income formation
. There
is in nature no such
thing as a stream of income. Income is a category of action; it is the outcome
In his
comprehensive treatment of price theory in
Man, Economy, and State
Rothbard repeatedly identi�ed the individual’s “money assets” or “money stock”
rather than “money income” as the direct co-determinant of the utility rankings
of money and goods. Like Mises, Rothbard maintained that a person’s current
money assets are derived from past market exchanges, “from his and his ances-
tors’ abilities in serving consumers on the market” (Rothbard, 2009, pp.
He explicitly
conceived of money income as derivative of concrete exchanges
that feed the reservoir of the directly valued element of money assets: “The
distribution of
money income (
. was
the necessary
consequence of a market allocation of prices and production” (Rothbard, 2009,
Rothbard emphatically rejected money income as a determinant of
demand. In an evaluation of a textbook on price theory, he critically remarked:
[W]hat in the world is the justi�cation for the totally illicit leap
. from
ing about marginal utility to talking about a consumer’s “money-income.”
What has happened to the concept of the marginal utility of money? It is
completely illegitimate to
. discuss
the utilities of [goods]
. without
mentioning, or ranking, the marginal utility of the money-price along
according to causal realist-theory, an individual always establishes the
marginal utility rankings of money and goods, which underlie his demand curves,
on the basis of his presently available money stock, not on estimations of past or
[E]very unit of the money commodity in a man’s stock
. is
always being
allocated to the three categories of use in accordance with his value scale
[i.e., to consumption, saving/investment, and cash balance]
. [E]ach
Joseph T. Salerno
be used for the most useful end not yet achieved. It is in accordance with
these principles
– the
maximization of his psychic income
– that
each man
will allocate his money stock. In accordance with his value scale, each man
will judge the respective marginal utilities to be obtained by each monetary
unit in each use, and his allocation of money expenditures
. will
be deter-

The purchasing power of money and the demand curve
In the causal-realist derivation of the individual demand curve, then, units of
various goods and of money are ranked and compared with one another by the
individual. But in order to intermingle units of money with units of goods on
a unitary value scale and judge their relative utilities, a
power of money must be assumed. The derivation of the demand curve in causal
realist theory therefore necessarily involves reference to the time element, a dis-
tinction between “yesterday” and “today.” The marginal utilities of goods today
are derived directly from the varying importance of the wants they are expected
to satisfy today. However, judging the subjective marginal utility of money today
necessarily entails knowing yesterday’s objective purchasing power of money,
that is, the inverse of the structure of money prices in all their particularity. This
means that before an individual can formulate his value scale in anticipation of
today’s exchanges, he must refer back to the purchasing power of money that
emerged in the immediately previous round of exchanges. In other words, an
individual’s value rankings and marginal utilities of goods and money, which are
operative in determining today’s structure of prices, are based on today’s valua-
tions of goods and money. But the valuation of money today must refer back to
purchasing power of money, because it is the only means by which its
prospective purchasing power in today’s market can be anticipated and its mar-
ginal utility set. If money did not have a pre-existing purchasing power
– that
if money never exchanged against goods in the past
– market
participants would
lack the knowledge needed to assign a value ranking to it and, consequently, no
one would accept it in exchange for goods today.
Every money price therefore
There is, therefore, no contradiction in assuming that the purchasing power of
money is constant
that the price of the good whose demand curve is being
analyzed is permitted to vary. For the purchasing power of money that is held con-
stant and on the basis of which the individual establishes his demand curve today
is the purchasing power of money expected to prevail today, which refers back to
yesterday’s structure of prices as the starting point for the forecast. As Rothbard
summed it up: “[T]oday’s purchasing power of the monetary unit is determined
by today’s marginal utilities of money and goods, expressed in demand schedules,
while today’s marginal utility of money is directly dependent on yesterday’s pur-
The “income effect” in causal-realist price theory

We now turn to the analysis of the income effect in causal-realist demand theory.
As we have argued, the role of the
ceteris paribus
assumption in demand analysis
is to permit us to abstract from the complexities of the value scale and to trace out
a curve that isolates the relationship between the price and quantity demanded of
a single good. This curve illustrates the operation of the law of marginal utility
in determining the inverse relationship between price and quantity demanded. As
the individual gives away additional units of money (of �xed purchasing power)
in exchange for additional units of the good, his “real” money stock declines,
causing an ordinal rise in its marginal utility ranking, while his stock of the good
increases, producing a fall in its marginal utility.
The fall in the marginal util-
ity of the good in relation to money (and to all other goods) as his stock of the
good increases and his stock of money declines implies that the demand price
of the good will decline and that therefore quantity demanded of the good will
increase as its money price declines (Rothbard, 2009, pp.
238–240; W
argued above, money income plays no part in demand analysis. The shape
and position of the demand curve are fully explained by the structure of an indi-
vidual’s value scale, upon which are ranged
stocks of goods and money.
An increase in an individual’s stock of money or a decrease in his demand for
ceteris paribus
, will lower the marginal utility of money and shift the
demand curve to the right. A
decrease in
his stock of money or an increase in his
demand for money,
ceteris paribus
, will have the reverse effect. If income, con-
ceived as a “�ow” of money (or goods), plays no direct role in the determination
of the demand curve, then, strictly speaking, a movement along the demand curve
cannot generate an income effect.
The contrast between causal-realist and mainstream demand analysis cuts far
deeper, however. Causal-realist analysis distinguishes temporally and logically
between cause and effect, in contrast to the timeless nature of the mutual deter-
minism that mars the standard analysis.
In the causal-realist view, the relevant
curves are the “instantaneous” demand and supply curves that give rise to the
prices realized in actual market transactions.
These demand curves are derived
from the individual’s personal economic situation
– his
value scale, stocks of
goods, real money stock, etc.
– as
it exists at the moment of his impending pur-
chases. But the focus on the present moment does not mean that the in�uence
of the individual’s past or future is disregarded. His current stocks of goods and
money are the resultant of his whole
of market transactions, including his
experience of the exchange value of money in his last round of transactions. His
scale of values is shaped by these stocks as well as his
of the prices
The individual demand curve is thus an
construct. It is derived from
value scales formed on the basis of money prices experienced in the present,
Joseph T. Salerno
becomes the immediate past once market activities are initiated.
another way, the demand curve is based on a person’s overall economic position
and his expectations prevailing in the moment preceding action.
If this were not
the case, if the demand curve did not refer to a period temporally and logically
antecedent to action, it would be impossible for individuals to formulate a coher
ent value scale, because the purchasing power of money would be unknown and
units of money could not be meaningfully ranked against units of goods. The
very existence of money prices thus logically implies the absence of an income
effect or, more properly, a “purchasing power effect.” That is, in causal-realist
analysis, the individual’s ex ante real money stock cannot vary with movements
along the demand curve, because the curve can only be derived based on an
already existing and “known”
– or
rather, de�nitely anticipated
– purchasing
power of money.

The nature of the substitution effect
Our causal-realist analysis of the income effect has the happy consequence of
shedding new light on the nature of the substitution effect. In analyzing the indi-
vidual value scale, it becomes clear that the substitution effect is an immediate
inference from the law of marginal utility. Since the value scale involves a relative
or ordinal ranking of all goods and money, all consumer goods are partial substi-
tutes for one another. As Rothbard points out:
consumer’s goods are
substitutes for one another. When a
man ranks in his value scale the myriad of goods available and balances the
diminishing utilities of each, he is treating them all as partial substitutes for
one another. A
change in
ranking for one good by necessity changes the rank-
in the rank-order or the money price of any one good will produce
changes in the relative positions of all other goods that will upset the equilibrium
among the marginal utilities of goods and their money prices. A
reallocation of
money assets among the various lines of consumption expenditure will be
required to realign the marginal utilities of all goods with their money prices.
For example, if the price of good A
increases along
an elastic individual
demand curve, the marginal utility of A
will fall
below the marginal utility of its
new money price and also relative to the marginal utilities of the increased quan-
tities of goods B, C, D, etc., which can now be had in exchange for A’s higher
money price. The more favorable terms on which the alternatives to A
can be
will also tend to raise their marginal utilities relative to their respective
(constant) money prices. Or, in other words, the fact that less of the individual’s
money stock is spent on good A,
ceteris paribus
, logically implies a decline of the
marginal utility of money in relation to substitute goods. The result will be that the
individual will readjust the marginal utilities of goods and money by substituting
The “income effect” in causal-realist price theory
additional units
of some or all of these goods for units of good A, shifting his
If the individual’s demand curve for A
is inelastic,
then a rise in the price of
good A
will cause
a greater proportion of his money stock to be used to purchase
and less
money to be spent on all other goods combined. A’s position on the
value scale will once again fall relative to its new, higher money price and the
law of marginal utility will dictate a restriction in the number of units purchased.
However, the increased allocation of money to the purchase of good A
rise of the marginal utility of money relative to some or all non-A goods and the
marginal utilities of these substitute goods will decline below their �xed money
prices. To re-establish equilibrium in accordance with the law of marginal utility,
the individual would reduce his purchases of these goods. In the new equilibrium
position: 1. the marginal utility of the stock of each good in his possession would
again just exceed the marginal utility of the sum of money representing its price;
and 2. expanding the purchase of any good beyond this margin would result in a
disequilibrating reversal of the marginal utilities of the good and its money price.
In the case of an inelastic demand for good A, the demand curves for substitute
It is important to note that alterations in marginal utilities that we have been
discussing in the last two paragraphs should not be construed as changes in the
individual’s underlying value scale, which is assumed �xed in demand analysis.
Rather, they are expressions of the structure of a given value scale in response
to external changes in the market situation, i.e., prices. The con�guration of the
value scale is necessarily “price dependent” because a change in the external
structure of money prices revolutionizes the internal terms on which alternative
goods can be substituted for one another.
Given that the values of goods are
inherently interdependent because they are ranked on a unitary value scale, the
law of marginal utility implies that all goods are at least partial substitutes for one
another. In contrast to mainstream consumer demand theory, causal-realist analy-
sis treats the concepts of substitutability and complementarity as derived from
distinct epistemological bases. Substitution is a necessary and universal relation-
ship among goods that is directly deduced from the existence of value scales,
although the degree of substitutability among different goods may be greater or
lesser depending on concrete historical circumstances. Complementarity is funda-
mentally a historically contingent relationship that depends on the concrete struc-
ture of individual value scales. Furthermore, complementarity among consumers’
goods, where it exists, will always be mixed with the all-pervasive relation of
substitutability, and it will only be potentially observable where it is stronger than
the substitution effect.

The income effect and the backward-bending labor
It may clarify the argument of this paper to critically analyze the recent con-
troversy among critics and defenders of causal-realist price theory regarding the
Joseph T. Salerno
between the income effect and the backward-bending supply curve
of labor. The controversy focused on Rothbard (2009), who, in his treatise
Economy, and State
, ignored the income effect in deriving the demand curve
while conceding the possibility of a backward-bending supply curve of labor.
Caplan (1999, pp.
828–829) criticizes
Rothbard for being inconsistent in ban-
ishing the income effect from demand analysis while admitting that a backward
bend in the labor supply curve was conceivable. Caplan (1999, p.
829) also
that Rothbard goes “so far as to mention a substitution and income effect” in
discussing the increase in labor supplied in response to an increase in taxation of
wages. This does not, however, imply, as Caplan claims, that Rothbard recognizes
the existence of a Hicksian income effect and “borrows [it] on an ad hoc basis”
(Caplan, 1999, p. 829). First, Rothbard places the terms “income effect” and “sub-
stitution effect” in scare quotes in the text (Rothbard, 2009, p.
in a later section of his book, Rothbard provides a more elaborate analysis of
income taxation and the backward-bending labor supply curve strictly in terms of
variations in the marginal utilities of leisure and money assets (Rothbard, 2009,
165). And, third, as noted above, there is the memo Rothbard (2011)
wrote explicitly dismissing “the alleged division between the ‘income effect’ and
the ‘substitution effect’
(1996) presents an immanent critique of the income effect using the Hick-
sian analytical apparatus. He concludes that “the income effect does not exist as
a general phenomenon; it is a mathematical illusion in a badly speci�ed world”
97). The
details of Salin’s argument do not concern us here. What is important
is that in the appendix to his article, he (1996, pp.
104–106) ar
gues for a position
Caplan takes for granted, viz., the existence of the income effect is the necessary
condition of a backward-bending supply curve. However, since Salin maintains
that the income effect is illusory, he concludes that the backward-bending supply
Gonzalez (2000) explicitly rejects Salin’s conclusion and defends Rothbard’s
position that the backward-bending labor supply curve is compatible with the
derivation of the individual demand curve using a purely ordinal value scale and
the law of marginal utility. Although his analysis marks a substantive advance of
the debate, Gonzalez unfortunately muddies the waters by super�uously import-
ing the terminology, if not the conventional concept, of the income effect into
his analysis (Gonzalez, 2000, p.
56). Thus,
he concludes, “the possibility of the
income effect of a price change is implied by the Misesian pure logic of choice”
(Gonzalez, 2000, p.
57). This
is particularly jarring in light of Mises’s and Roth-
bard’s views on the concept of income discussed above.
The conundrum presented by Rothbard eschewing the income effect in his
demand analysis while recognizing the possibility of a backward-bending sup-
ply curve was actually neatly resolved once and for all in a brilliant and concise
piece of analysis by Lionel Robbins ([1930] 1997) in 1930. Robbins developed
his analysis to rebut the arguments of Frank Knight and A.
C. Pigou
that the law
of diminishing marginal utility implied that the quantity of labor supplied always
varied inversely to changes in the after-tax wage rate, or, in the words of Robbins
The “income effect” in causal-realist price theory
([1930] 1997,
81), “that
the imposition of a tax will always have the effect
of making a man work more, and a rise in his wage rates will always make him
work less.” Robbins demonstrated that a forward-sloping, as well as a backward-
bending, labor supply curve is consistent with the law of diminishing marginal
Forty years later, James Buchanan (1971, p. 383) resurrected Robbins’s
forgotten analysis, which he deems an “essentially correct explanation.” Buchanan
(1971, p.
383) uses
the analysis to demonstrate that the “Hicksian income effect-
substitution effect apparatus” was “wholly unnecessary” to derive “the backbend-
ing supply curve of labor.” Indeed, Buchanan (1971, p.
283) characterizes
view among modern economists that the income effect is necessary to deriving
the backward-bending supply curve as “an example of doctrinal retrogression.”
Robbins’s analysis is straightforward, and it can be presented, as it was originally,
with the help of a few simple diagrams. While two of the diagrams are similar
to the two used by Robbins, I
have added
a third suggested by Buchanan (1971,
383). For
clarity of exposition, I
have also
added tables of hypothetical data,
presents an individual labor supply schedule, which generates the
backward-bending supply curve shown in Figure
as A-B-C. (We ignore for
the moment the �gure
in parenth
eses.) Between the wage rates of $5 and $20
Table 2.1

Wage rate
ge Rate ($/hour)
Figure 2.1

Joseph T. Salerno
, the quantity supplied of labor on a weekly basis varies positively with the
wage rate; as the wage rate rises from $20 to $40
per hour
, the quantity of labor
hours supplied declines and the supply curve bends backwards. Now, as Rob-
bins ([1930] 1997, p.
79n1) demonstrated,
this result is easily derived from the
law of marginal utility by following Wicksteed and “exhibiting all psychological
variables as phenomena of demand.” Thus, we portray the laborer as exchanging
hours of leisure, which he values and demands as a consumer’s good, for real
money assets (units of real income in Robbins’s example).
In order to transform the supplier of labor hours into a demander of money
assets, we translate the labor supply curve into a demand curve for money assets
with money on the quantity axis and the reciprocal of the wage rate, that is, labor
or for
gone leisure
– hours
per dollar
purchased on the price axis.
These results
are shown in Table
where, for example, a wage rate of $20
hour is
equivalent to a labor price of money assets of .05
dollar and the
total quantity demanded is equal to $1,000 (= $20/hour x 50
hours). Note
that the
demand curve for money assets, A-B-C, slopes downward throughout its entire
length, despite the fact that it was derived from a backward-bending supply curve.
This indicates that as the labor price of money assets falls,
ceteris paribus
Table 2.2

Labor price of money,
Figure 2.2

The “income effect” in causal-realist price theory
quantity demanded
increases, so that the existence of a backward-bending supply
curve does not necessarily contradict the law of marginal utility. In other words,
as the “pre-Hicksian” Hicks approvingly summed up Robbins’s contribution:
“[t]he only natural deduction from the law of diminishing marginal utility is, not
that the supply curve of labor must slope downwards [the Knight-Pigou position],
o avoid confusion, a few comments on this analytical construction are in
order. First, note that the backward-bending segment of the supply curve, A-B
corresponds to an inelastic demand for money assets in terms of
labor, segment B-C in Figure
This can be seen from Figure
which is the
graph Robbins derived from the demand curve for money assets by plotting the
total expenditure of labor hours for each quantity of money assets purchased as
A similar “reciprocal demand” or “offer curve,” �rst formulated by Marshall
and Edgeworth, is used in international trade theory to show the total exports that
a country is willing to expend for each given quantity of total imports as the terms
of trade vary.
In this case, the “exports” are labor hours and the “imports” are
money assets. The reciprocal demand curve is, in effect, a total expenditure curve.
As in the case of conventional demand-curve analysis, a decline in total expendi-
ture as price decreases indicates that the associated segment of the demand curve
is inelastic. Thus, a backward-bending segment of the labor supply curve implies
an absolute value of the elasticity of demand for money assets in terms of labor
Second, the law of marginal utility rules out a supply curve that bends back
so sharply that it yields a lower quantity of money assets at a higher wage rate.
For example, let us substitute a quantity supplied of 20
hours (in
parentheses) for
hours at
a wage rate of $40
per hour
in Table
The labor supply curve in
Figure 2.3

Joseph T. Salerno
changes shape, with the backward-bending segment now represented
by Aꞌ-B, which is �atter and more elastic than segment A-B. The total quantity of
money demanded now
from $1,000 at a wage rate of $40
per hour
to only
$800. But this implies that fewer money assets are purchased as their labor price
falls from .05
hours/$ to
hours/$ and
the demand curve for money assets over
This example thus violates the law of marginal utility, which dictates that a lower
labor price of money assets must always lead to a greater quantity demanded. For
if a laborer responds to an increase in the wage rate by working fewer hours and
consuming more leisure, his marginal utility of leisure falls. In order to balance
this fall in the marginal utility of leisure, he must allocate his additional resources
so that his marginal utility of money also declines, which is inconsistent with a
Although Rothbard does not cite Robbins’s analysis in his own discussion of
the backward-bending labor supply curve, he explains the phenomenon in similar
terms, referring to the labor price of money income or assets. In his analysis of the
Income taxation reduces every taxpayer’s money income and real income
His income
from working is more expensive, and leisure cheaper, so that he
will tend to work less
[M]uch has
been made of the fact that every man’s
marginal utility of money rises as his money assets fall and, therefore, that
there may be a rise in the marginal utility of the reduced income obtainable
from his current expenditure of labor. It is true, in other words, that the same
labor now earns every man less money but this very reduction in money
income may also raise the marginal utility of a unit of money to the extent
that the marginal utility of his total income will be
, and he will be
induced to work
as a result of the income tax. This may very well be
true in some cases and there is nothing
. contrary
to economic analysis in
164–1165; emphases in original)
Rothbard concisely speci�es the structure of an individual’s value scale that
must prevail in order for a change in the wage rate to produce a backward-bending
There will be such a backward supply curve if the marginal utility of money
falls rapidly enough and the marginal disutility of leisure forgone rises rap-
point can be expressed in terms of our example. At an hourly wage rate
of $20, the individual chooses to sacri�ce 50
hours of
leisure for $1,000. On
his underlying value scale, then, the marginal increment of $20 is ranked just
The “income effect” in causal-realist price theory
above the
50th hour of forgone leisure and, therefore, above infra-marginal hours
like the 31st hour. When the wage rate increases to $40
per hour
, the individual
decides to forgo 30
hours of
leisure in exchange for $1,200. The additional earn-
ings of $200 causes the marginal utility of money to fall so rapidly relative to the
marginal utility of leisure that at the new equilibrium point on his value scale the
previously infra-marginal hours of leisure forgone from the 49th down to 31st that
were ranked
$20 at the lower wage rate are now ranked
$40. This
reinforces the point that, although the labor supply curve may bend backwards,
its bend cannot be so extreme that a rise in the wage rate yields not only fewer
yours of labor but also lower money earnings, since this would result in a rise and
not the required rapid decline in the marginal utility of money relative to that of
leisure. In technical terms, “the elasticity of individual supply of labor must be
to Caplan’s claim, therefore, Rothbard did not “borrow” the income
effect from post-Hicksian neoclassical economics in order to derive the backward-
bending labor supply curve, let alone the conventional demand curve. Rather, in
ignoring the income effect, Rothbard adopted and consistently applied the causal-
realist approach to demand analysis developed in the writings of pre-Hicksian
price theorists running back through Mises, Robbins, Wicksteed, Fetter, and Dav-
enport to Böhm-Bawerk and Menger.
The most
notable contributors to this tradition include Menger’s students Eugen
von Böhm-Bawerk and Friedrich von Wieser, as well as prominent Anglo-American
economists such as Philip Wicksteed, Lionel Robbins, J. B. Clark, Frank A. Fetter,
and Herbert J. Davenport. After World War Two, causal-realist price theory was fur-
ther elaborated by Ludwig von Mises and, especially, Murray Rothbard. See Salerno
E. Mason (1996) demonstrates conclusively that the so-called classical dichotomy
was actually an “inversion” of classical monetary theory perpetrated by neoclassical
quantity theorists such as Irving Fisher, Alfred Marshall, and Edwin Kemmerer.
comprehensive survey of the debate that also includes a collection of some of the
more important contributions can be found in Ekelund et
al. (1972).
Yeager (1960) pro-
Ekelund et
al. (1972,
42) ar
gue that Hicks was cognizant of Marshall’s assumption
of the constant purchasing power of money but viewed it as a “simplifying assumption
Thus Ekelund
al. (1972,
43) conclude:
“In the theoretical formulation of the
demand curve, Marshall’s formulation �ts the constant real income classi�cation and
the Friedman interpretation appears valid.” The authors contend, however, that Mar-
shall dispenses with the assumption of constant purchasing power of money, while
retaining that of constant money income in practical applications of the demand curve.
In his exhaustive textual analysis of Marshall’s derivation and application of the
demand curve, Alford (1956) comes to a more or less similar conclusion.
Friedman ([1949] 1970, pp.
56, 84–86)
interprets the demand curve so as
to rule out Giffen goods.
Joseph T. Salerno
At the end of the day, Yeager (1960, p.
63, fn.
32) remains a Hicksian price theorist
who accepts that Giffen’s Paradox “could conceivably occur” in the real world. Cf.
also Yeager (1999).
[W]e do
not have any knowledge or experience concerning the shape of such [sup
ply and demand] curves. Always, what we know is only market prices
– that
is, not
the curves but only a point which we interpret as the intersection of two hypothetical
curves. The drawing of such curves may prove expedient in visualizing the prob
lems for undergraduates. But for the real tasks of catallactics they are mere byplay.
Economics is
mainly concerned with the analysis of the determination of money
prices of goods and services exchanged on the market
. Catallactics
is the analy-
sis of those actions which are conducted on the basis of monetary calculation. Mar-
ket exchange and monetary calculation are inseparably linked together. A
market in
For example,
for Wicksteed ([1933] 1957, p.
36), “The
conception of the scale of
preferences” is “quite fundamental” and underlies “all investigations” of economic
phenomena; and for Rothbard, “[T]he individual’s value scale provides the key to the
determination of all events on the market” (Rothbard, 2009, p.
379). According
Kirzner: “The fundamental premise the theory of demand (and also market theory in
its entirety) is built upon is that
men do not consider all of their desires to be of equal
. [W]e
these inclinations or desires as either more or less urgent”
([1963] 2011, p.
Also, W
icksteed, in the �rst volume of his treatise, presented exhaustive partial and
general equilibrium analyses of price determination using preference scales exclusively.
He did not introduce supply and demand curves until the second volume of his work
(Wicksteed, [1933] 1957, 1: 126–157, 212–265). And even here, Wicksteed emphasized
that these curves present “many problems both of interpretation and construction,” are
“purely abstract,” and are necessarily “isolated” such that a system of such curves cannot
be constructed “as to be valid simultaneously” (Wicksteed, [1933] 1957, 2: 439, 474).
This felicitou
s term was coined by Arthur Marget ([1938–42] 1966). According to
Marget: “[A] realized price represents the passage of money for an article sold for
. [A]ny
given realized price is what it is as the result of the conformation
and position of the market demand curve and market supply curve prevailing at the
moment the price is realized” (Marget [1938–42] 1966, 2: 240).
The emphasis
on stocks rather than �ows in supply and demand analysis is distinc-
tively causal-realist and non-Marshallian. As Creedy notes, it traces back through
Wicksteed and Edgeworth to Jevons and Böhm-Bawerk (Creedy, 1991, pp.
On the
meaning and nature of maintaining, accumulating or consuming capital, see
The fundamental
notion of economic calculation is the notion of capital and its cor-
relative income
. It
is a product of reasoning and its place is in the human mind.
It is a mode of looking at the problems of acting, a method of appraising them from
the point of view of a de�nite plan. It determines the course of human action and is,
in this sense only, a real factor.
The “income effect” in causal-realist price theory
This is the essence of the regression theorem formulated by Mises (1971, pp.
The historical
component of any one particular money price, say, the price of corn
today, is not yesterday’s price of corn, but the entire structure of prices obtaining on
yesterday’s market, the reciprocal of which constitutes yesterday’s purchasing power
of money. See Rothbard (2009, p.
In analyzing
the individual’s scale of preferences, Wicksteed viewed the money price
as an “index” of the terms on which alternatives to the good are available, which is
[W]e will begin by taking for granted
. the
purchasing power of money and the
existence of market or current prices as facts
. [W]e
do not know the actual alter-
natives represented by the price of any one commodity until we know the price of
(Wicksteed [1933] 1957, 1: 18, 22)
Davenport also recognized that analysis of the demand curve assumes “an existing
system of prices upon goods in general and an established price relation for these
All valuations
involving money, of course, refer to the real money stock and not to the
See the
attempt by Patinkin (1965) to explain the purchasing power of money while
eschewing temporal, causal analysis in favor of simultaneous mutual determina-
tion. For a critique of Patinkin from a causal-realist perspective, see Salerno (2010,
According to
Marget ([1938–42] 1966, 2: 235 fn. 31), writing in the late 1930s, the
term “instantaneous” was “commonly” used to designate such demand and supply
See the
insightful discussion by Rothbard (2009, pp.
279–281) of
ex ante
nature of
As Marget perceptively stated:
essential element
. is
summed up by the statement that the demand sched-
ules of the general Theory of Value are concerned with what has
. been
“the pre-formation of market prices”
. [T]he
market demand schedules
. are
” curves representing the
of possible purchasers with respect to the
market situation, as that situation is evaluated in the light of his own present
economic situation.
. [S]uch
an evaluation and therefore the “plans” based upon
it, would take into account (1) the ways in which the purchaser’s own present eco-
nomic position and the general market situation have been affected by past events;
and (2) the purchaser’s expectations with respect to his own future position and the
(Marget [1938–42] 1966, 2: 177–178; emphases in original)
On this point see the classic discussion by W
Rothbard presents
the clearest and most comprehensive account of the causal-
realist approach to substitutable and complementary goods, which highlights
the important epistemological distinction between the two kinds of related
goods(Rothbard, 2009, pp.
280–288). Mises
distinguishes the general praxeolog
ical relation of substitution among goods, which he calls the “the general con
nexity of the prices of all goods and services,” from the historically contingent
“particular connexities of the prices of a limited number of commodities” (Mises,
Joseph T. Salerno
in an unpublished draft of his article, Caplan (1997) placed quotation
marks around the two terms in his own text, but still did not unambiguously indicate
that they originally appeared in scare quotes in Rothbard’s text.
Rajsic (2010)
offers a critique of Caplan that purports to derive Hicksian substitution
and income effects using Rothbard’s ordinal value-scale approach. Needless to say, the
attempt, although certainly ingenious, fails. We do not discuss it in the text because it
It is
telling that economists such as Knight, Pigou, and Robbins, who were trained
in the analytical techniques of the pre-Hicksian era, had no problem grasping that a
backward-bending supply curve did not violate the law of marginal utility.
Steeped as
he was in Austro-Wicksteedian theory, Robbins evidently did not think it
necessary to draw the labor supply curve and perform this transformation, but imme-
diately began his analysis with the demand curve for “income in terms of effort.” This
On the
development of reciprocal demand analysis from Mill to Marshall and Mar-
shall’s formulation of the offer curve, see Allen (1965, pp.
and Wu ([1939]
2007, pp.
163–175). For
the derivation and use of the offer curve in modern trade
theory, see, for example, Thompson (1993, pp.
Alford, R. F. G. 1956. “Marshall’s Demand Curve.”
Allen, W. B. (ed.). 1965.
International Trade Theory: Hume to Ohlin
. New York: Random
Böhm-Bawerk, E. V. 1959.
Capital and Interest
, Vol. II:
Positive Theory of Capital
. Trans.
George D. Huncke. South Holland, IL: Libertarian Press.
Buchanan, J. M. 1958. “
“The Backbending Supply Curve of Labor: An Example of Doctrinal Retro-
Caplan, B. 1997. “Why I
Am Not
an Austrian Economist,”
bcaplan/whyaust.htm, retrieved April
“The Austrian Search for Realistic Foundations.”
Creedy, J. 1991. “The Role of Stocks in Supply and Demand: Wicksteed’s Problem.”
Oxford Economic Papers
Davenport, H. J. 1968 [1913].
The Economics of Enterprise
. New York: Augustus M.
Edwards, J. R. 1985. The Economist of the Country: Ludwig von Mises in the History of
Monetary Thought. New York: Carlton Press.
Ekelund, R. B., E. G. Furubotn, and W. P. Gramm. 1972. “Demand Theory in Marshall’s
Tradition.” In R. B. Ekelund, E. G. Furubotn, and W. P. Gramm, eds.,
The Evolution
of Modern Demand Theory: A
Collection of
. Lexington, MA: D.C. Heath and
Company, pp.
M. 1954. “The Marshallian Demand Curve: A
of Political
[1949]. “The Marshallian Demand Curve.” In M. Friedman, ed.,
R. 2000. “Misesian Economics and the Response to a Price Change.”
Journal of Austrian Economics
The “income effect” in causal-realist price theory
Hayek, F. A. 1975 [1939]. “The Maintenance of Capital.” In Friedrich August Hayek, ed.,
Pro�ts, Interest and Investment and Other Essays on the Theory of Industrial Fluctua
. Clifton, NJ: Augustus M. Kelley, pp.
“Capital Consumption.” In Roy McCloughry, ed.,
Money, Capital and Fluc
The Theory of Wages
. New York: Palgrave Macmillan.
alue and Capital: An Inquiry Into Some Fundamental Principles of Eco
Hudik, M. 2011. “Rothbardian Demand: A
Review of

24 (3): 311–318.
Kirzner, I. M. 2011 [1963].
Market Theory and the Price System
. Ed. P. J. Boettke and
F. Sautet. Indianapolis, IN: Liberty Fund.
Klein, P. G. 2008. “The Mundane Economics of the Austrian School.”
of Austrian Economics
11 (3–4): 165–187.
Marget, A. W. 1966 [1938–1942].
The Theory of Prices: A
Re-Examination of
the Central
Problems of Monetary Theory
. 2 vols. New York: August M. Kelley.
Marshall, A. 1920.
The Principles of Economics: An Introductory Volume
. 8th ed. London:
Mason, W. E. 1996.
Classical Versus Neoclassical Monetary Theories: The Roots, Ruts,
. Ed. W. N. Butos. Boston: Kluwer
Mises, L. V. 1971.
The Theory of Money and Credit
. Trans. H. E. Batson. 2nd ed. Irvington-
on-Hudson, NY: Foundation for Economic Education.
Mises, L. V. 1998.
Human Action: A
eatise on Economics
. 2nd Scholar’s ed. Auburn, AL:
Patinkin, D. 1965.
Money, Interest, and Prices: An Integration of Monetary and Value
. 2nd ed. New York: Harper
Rajsic, P. April
8, 2010.
“Did Rothbard ‘Borrow’ the Income and Substitution Effects?”
Mises Daily Articles
tution-effects, retrieved March
L. 1997 [1930]. “On the Elasticity of Demand for Income in Terms of Effort.” In
S. Howson, ed.,
Economic Science and Political Economy: Selected Articles
. New York:
New York University Press, pp.
M. N. 2009.
Man, Economy and State: A
eatise on Economic Principles With
Power and Market: Government and the Economy
. 2nd Scholar’s ed. Auburn, AL: Lud-
Salerno, J. T. 1999. “The Place of Mises’s Human Action in the Development of Modern
Economic Thought.”
Quarterly Journal of Austrian Economics
“Ludwig von Mises’s Monetary Theory in Light of Modern Monetary
Thought.” In J. T. Salerno,
Money, Sound and Unsound
. Auburn, AL: Ludwig von Mises
1. “Varieties of Austrian Price Theories: Rothbard Reviews Kirzner.”
Salin, P. 1996. “The Myth of the Income Effect.”
The Review of Austrian Economics
9 (1):
Stigler, G. J. 1952.
. 2nd ed. New York: Palgrave Macmillan.
Stonier, A. W., and D. C. Hague. 1953.
A Textbook of Economic Theory
. New York: Long-
Joseph T. Salerno
Thompson, Henry. 1993.
International Economics: A
. White
Plains, NY: Longman.
Wicksteed, P. H. 1957 [1933].
The Common Sense of Political Economy and Selected
Papers and Reviews on Economic Theory
. 2 vols. Ed. L. Robbins. London: Routledge
Wu, C.-Y. 2007 [1939].
An Outline of International Price Theories
. Auburn, AL: Ludwig
Yeager, L. 1960. “
“Should Austrians Scorn General-Equilibrium Theory?”
Review of Austrian
11 (1): 19–30.
of causal-realist
production theory
[C]oncentration on a single �rm and the reaction of its owner is not the appropriate
route to the theory of production; on the contrary, it is likely to be misleading.
In the
current literature, this preoccupation with the single �rm rather than with the
interrelatedness of �rms in the economy has led to the erection of a vastly compli-
cated and largely valueless edi�ce of production theory
Murray Rothbard’s
Man, Economy, and State
(2009 [1962]) is a landmark book
in Austrian economics. In the tradition of earlier writers, especially Ludwig von
Mises (2008 [1949]), it is written in the form of a treatise that derives the general
body of economic theorems from the ground up, starting with isolated individual
action, moving on to various forms of interpersonal exchange, and ending with
government intervention. When developing this economic organon Rothbard syn-
thesized the ideas of many economists working in the Austrian tradition, includ-
ing Carl Menger, Eugen von Böhm-Bawerk, Frank Fetter, Phillip H. Wicksteed,
Ludwig von Mises, and F. A
Hayek. The
Austrian, or “causal-realist,” approach
adopted by these writers concentrates on issues such as real-world price formation,
entrepreneurship and the market process, and the relationship between time and
the production structure. It uses the praxeological method, deduction grounded in
the fact that humans behave purposively, along with a series of realistic empiri-
cal assumptions, such as that there exist a variety of natural resources and that
of Rothbard’s monumental contributions in this book was the construc-
tion of a systematic production theory that integrated various strands of thought
developed by earlier writers working in the same tradition, including theories
of capital and interest, the structure of production, rent and factor pricing, and
entrepreneurship (Salerno, 2009, p.
xxvi). One
especially notable achievement of
Rothbard’s was his integration of the Mises-Fetter pure time preference theory of
interest with the Hayek-Wicksell structure of production analysis (Salerno, 2009,
xxvii; Rothbard,
2009 [1962], p.
lvii). In
general, his production theory inte-
grated all of the interrelations of the production structure and set out to actually
From Marshallian partial
The evolution of Rothbard’s
explain the
formation of output and input prices throughout the economy. This
synthesized Austrian production theory is different from the more well-known
Marshallian partial equilibrium approach. The latter is best represented in mod-
ern economics by Chicago production theory, which was mainly developed by
George Stigler and Milton Friedman, who each built on the works of Alfred Mar-
shall and Frank Knight (Salerno, 2011, pp.
1–2). This
theory analyzes equilibrium
production decisions from the viewpoint of an isolated �rm with given input and
In contrast, Austrian production theory is the halfway house, or middle ground,
between excessive microeconomic analysis
– the
Marshallian partial equilibrium
view that concentrates on a single �rm facing �xed prices
– and
excessive macro-
economic analysis, exempli�ed by the Keynesian aggregative approach that her-
metically seals off sectors of the economy from each other. The Austrian theory
shows that a change in any sector of the economy must always impart its in�u-
ence through repercussions in the structure of prices and production in other sec-
tors. This Austrian general equilibrium is starkly different from Walrasian general
equilibrium for three important reasons. The �rst is that it is dynamic, not static,
because it emphasizes the importance of uncertainty and change and reveals the
equilibrating processes between equilibrium states that are driven by pro�t-seeking
capitalist-entrepreneurs. The second is that it recognizes the intertemporal hetero-
geneous capital structure. The third is that it expresses its theorems using verbal
logic rather than non-causal, mutually determined mathematical equations (Roth-
bard, 2009 [1962], p.
361; 2008
[1963], pp.
65–66). At
the beginning of his trea-
One “radical” feature of our analysis of production is a complete break with
the currently fashionable “short-run” theory of the �rm, substituting for this
a general theory of marginal value productivity and capitalization.
“general equilibrium” analysis in the dynamic Austrian sense, and not in the
static, currently popular Walrasian sense
However, many readers of Rothbard’s are unaware that the original drafts of
his treatise contained a production theory grounded in Marshallian partial equi-
librium theory, and that only after writing extensively did Rothbard realize much
of it was untenable. Only in a 1990 interview did Rothbard brie�y mention the
lost chapter, stating that “I
took Chapter
5 out
Man, Economy, and State
, which
included the usual cost-curve analysis. I
wrote the
whole chapter
before I
that the
approach I
was taking
was nonsense. So I
started over”
(Rothbard, 1990).
Tentatively titled “Chapter
’s Activity” (2018 [1953]), the manuscript
was recently reconstructed by the present author from the materials available in
the Rothbard archives at the Ludwig von Mises Institute. Among other things,
what is particularly interesting is that Rothbard constructed the chapter
the point
of view of an individual �rm, and based the analysis on four points
about which he later changed his mind: the distinction between a free-market,
Evolution of Rothbard’s production theory
competitive price
and a monopoly price; the model of perfect competition and
the price taker assumption for output prices; using the price taker assumption for
input prices and the isoquant-isocost framework to derive factor demand curves;
and using the isolated �rm as a unit of analysis to understand optimal production
The competitive-monopoly price distinction is inappropriate for analyzing free
market situations because it arbitrarily assumes that a certain price is competitive
and a higher price that increases revenue is monopolistic. Both prices still are mar-
ket prices that are consonant with consumer wants, especially when one realizes
the restriction of output that increases revenue releases factors of production that
can be used elsewhere. The price taker assumption in perfect competition is unre-
alistic because every �rm contributes to the total market supply and commands
some in�uence on its output price. As a result, all �rms are price searchers and
engage in so-called imperfect competition, and the traditional ef�ciency bench-
mark of perfect competition is a poor standard because it is impossible to attain.
The factor demand curve derived from the isoquant-isocost framework does not
actually explain the formation of the input price it sets out to explain, nor does it
show the causal in�uence of the output price on the input price. Lastly, the correct
unit of analysis for understanding optimal production and investment in a �rm is
not that of the manager of a �rm, but of the capitalist-entrepreneurs who invest
in the �rm. This is because, unlike the manager of a �rm, capitalist-entrepreneurs
can invest in multiple �rms. The most important implication of this is that, as
opposed to a static general equilibrium where all investment returns are equal, in
a dynamic world of multiple investment opportunities for capitalist-entrepreneurs
with varying degrees of pro�tability, pro�t may not be maximized in a given �rm
where marginal revenue equals marginal cost (MR
paper traces the evolution of Rothbard’s thinking on these issues. It is
important to realize that this is not an arcane exercise in the history of economic
thought, as juxtaposing Rothbard’s divergent production theories highlights
important differences and shows the weaknesses of modern Marshallian partial
equilibrium production theory using Austrian general equilibrium theory. This
reinforces the view of Klein (2010b [2008], p.
149) that
the “mundane econom-
ics” of the Austrian school is distinct from the neoclassical paradigm.
The rest of the paper is outlined as follows: Section II provides a brief history of
Rothbard (2018 [1953]) and his changing ideas on production theory; Section III
juxtaposes Rothbard’s analyses of monopoly prices and perfect competition; Sec-
tion IV compares Rothbard’s analyses of the derivation of factor demand curves;
Section V discusses the pro�t maximizing output level of a �rm in a dynamic
world, and Section VI concludes.

s production theory
In the fall of 1949, the Volker Fund asked Rothbard to write a “college-style”
economics textbook modeled after Mises’s
Human Action
(1949). After Mises
reviewed and approved a sample chapter
on money
, Rothbard began working on
the project.
What was originally supposed to be a principles-level textbook, how-
ever, developed into an advanced economic treatise that would occupy a large
part of Rothbard’s intellectual efforts in the 1950s, eventually becoming
Economy, and State
The order in which he initially wrote the textbook mirrored
the structure of his later treatise. By April
1953 he
had written rough drafts of
what can be considered Chapters
1–4 of
this, Rothbard moved on to writing a large chapter
on production
ory, tentatively titled “Chapter
5: Producer
’s Activity” (2018 [1953]). Its main focus
was the production decisions of an individual �rm and an analysis of input and
output pricing. It also contained a competitive-versus-monopoly price framework
which included perfect competition. When discussing the production decision of a
�rm, he derived constant outlay (isocost) and constant product (isoquant) schedules
as well as rates of constant outlay and constant product substitution in order to show
that the cost-minimizing level of output is where these two rates are equal. Rothbard
he producer.
One important feature of the chapter
is that
it lacks a Misesian or Austrian
“feel.” As the original plan was for Rothbard to write a textbook version of Mises
(1949), this is surprising and puts a unique perspective on the chapter. His �rst
chapters had followed Mises and earlier Austrians more closely, as they directly
dealt with topics on which those writers concentrated. For instance, the �rst
tried to
clarify Mises’s analysis of the fundamental laws of praxeology,
i.e., the means-end relationship and the laws of marginal utility, time preference,
and returns (Stromberg, 2004, p.
xxxii). His
other chapters, particularly those on
exchange and pricing, built signi�cantly on writers other than Mises, not because
their views were different, but because Mises had assumed his audience already
knew the material, and as a result had not covered supply and demand analytics
in depth (Stromberg, 2004, p.
xxxv). For
example, on the topic of basic price ana-
lytics, Mises simply assumed his readers understood the “marginal pairs” frame-
work, and brie�y cited Böhm-Bawerk (1959 [1889], pp.
207–256) for
those who
wanted more information (Mises, 1949, pp.
202, 324).
Aside from citing Fetter
on time preference
– a
point on which Rothbard planned to elaborate in a later
– and
Mises and others on monopoly price theory, for the bulk of his
Rothbard utilized
the standard tools of price theory, relying in particular
on Stigler (1947 [1946]) and Weiler (1952). This is interesting because Rothbard
(2009 [1962]) would later use these works as models for criticism of contempo-
rary production theory. His initial use of these writers, and not Mises, most likely
stems from the fact that Mises did not talk about the topic Rothbard wanted to
discuss, namely, the optimal production decisions of an individual �rm.
Ultimately, Rothbard decided that his approach was totally in error (for rea-
sons to be described in later sections), and as a result proceeded to completely
rewrite his production theory. In particular, he realized he would have to forge
a new path into areas that Mises did not explicitly develop, and in some cases
such as
the competitive-versus-monopoly price distinction
– correct
his views.
This decision was mainly responsible for his proposal in February
1954 to
Evolution of Rothbard’s production theory
from writing
a textbook based on Mises (1949) to a full-blown economic treatise.
Writes Rothbard:
The original concept of this project
. was
as a step-by-step, spelled-out
version of Mises’s
Human Action
. However, as I
have been
proceeding, the
necessary elaborations on the sometimes sparse framework of Mises has led
inevitably to new and original presentations. Now that I
have been
ing to the theory of production where the whole cost-curve situation has to be
faced, Mises is not much of a guide in this area. It is an area which encom-
passes a large part of present-day textbooks, and therefore must be met, in
one way or another. Mises, in his treatise, deals only tangentially with the
and really with good reason, but a more detailed treatise, or one that
attempts to be a textbook, must tackle this issue. After much thought about
the problem, and many false writing starts, I
have come
to the conclusion
that the whole complex of cost curves
. and
the whole emphasis on the size
of �rm, cost curves to plant, etc.
. is
all erroneous speculation on techno-
logical irrelevancies .
. [T]his
whole line of approach [is] now glori�ed in
the texts as the “theory of the �rm”
. [T]he
�rst draft of Chapter
. is
being completely rewritten to omit “the theory of the �rm”
. [It]
become evident from my work on the book, that the result cannot be a text-
1955, Rothbard
decided to split his work on production into multiple
chapters, what are now Chapters
5–10 of
Rothbard (2009 [1962]) (Stromberg,
2004, p.
liv). Major
work on the book apparently stopped around 1956, and by
1959 Rothbard had completed it (Stromberg, 2004, p.
lxiii). As
explained below,
Rothbard’s revised production theory did not suffer from the pitfalls of the partial
equilibrium �rm analysis of his early work. We now turn to a discussion of the
evolutions of Rothbard’s thought on some of these critical matters.

Kirzner (2013
[1973], pp.
15–18, 70–107)
provides the most recent defense and
elaboration of Mises’s views. During the Monopolistic Competition Revolution
in the 1930s, this approach was superseded by the familiar perfect-imperfect
competition framework (Salerno, 2004; Rothbard, 2009 [1962], p.
720). Modern
economics now
analyzes markets as situations in which �rms engage in either
perfect or imperfect competition. Under perfect competition, the individual �rm
is so small relative to the total market that it has no in�uence on the price of its
product and takes the market price as “given,” i.e., it can increase or decrease its
output without changing the price. There are no barriers to entry, all �rms in the
given market produce a homogenous product, and all �rms in the market have
perfect knowledge (Stigler, 1946, p.
Under imperfect competition, the �rm
has an in�uence on the price of its product, and when it increases or decreases
its output, the price must change.
The important assumption for Rothbard (2018
Rothbard’s earlier production theory employed both frameworks in a hybrid
fashion. In the �rst two sections of Chapter
5, titled
1: The
Demand for
a Firm’s Product” and “Section
2: Competitive
Price and Monopoly Price,” he
discusses the various production possibilities a �rm can use to make Good A,
and explains the scenario under different pricing situations. Rothbard presents
possible combinations of inputs X and Y that can produce varying amounts of
Good A
at a
constant cost, assuming the prices of X and Y are �xed (Rothbard,
2018 [1953], pp.
77–79). He
then analyzes the �rm’s output decision in various
pricing scenarios (a), (b), and (c) for Good A. In (a), the �rm’s individual demand
curve is horizontal, or perfectly elastic, as opposed to sloping downward as in (b)
and (c). In (b) and (c), the �rm faces downward-sloping demand curves, but in
(b) the point of maximum output from a given amount of money invested in the
factors is the point of maximum revenue, while in (c) the point of maximum rev
enue is no longer at maximum output, i.e., the demand curve is inelastic above
the point of maximum output. It therefore pays for the �rm in (c) to restrict out
put relative to its counterfactual production level under more competitive condi
tions (Rothbard, 2018 [1953], pp.
79–83). Rothbard
argues that a competitive
price will result in both (a), which is traditionally described as perfect competi
tion, and
(b). However
, in case (c) a monopoly price will result (Rothbard, 2018
grudgingly adopts the terms “competitive price” and “monopoly
price,” noting that their use in economics is unfortunate and misleading, but also
that they “must be used for traditional reasons” (Rothbard, 2018 [1953], pp.
88). Much
of the analysis in Section
2, on
monopoly prices and monopoly, are
later reproduced as whole paragraphs in Rothbard (2009 [1962]), “Chapter
Monopoly and
Competition.” For example, Rothbard notes that monopoly was
originally de�ned as a grant of state privilege to produce a good or service, that
de�ning a monopolist as the producer of a single good is a poor de�nition, and
that there is a built-in tendency for cartels to break down. He even notes that a
monopoly price does not defy the wishes of consumers and is not immune from
competition, since all goods compete for consumers’ money. Exchanges are
still voluntary because the consumers still voluntarily form their demand curves
(Rothbard, 2018 [1953], pp.
88–90). Importantly
, Rothbard pins the unfortunate
monopoly-competitive price distinction on Marshallian partial equilibrium the-
ory, writing that, “the terminology is the result of an old neoclassical preoccupa-
Evolution of Rothbard’s production theory
Rothbard later
scrapped the entire competitive-monopoly price distinction and
trenchantly critiqued it (Rothbard, 2009 [1962], pp.
672–704). He
wrote that on
the free market the entire distinction is spurious because it is impossible to de�ne
a competitive price. There is no way to look at the demand curve a �rm faces and
decide which is a competitive price and which is a monopoly price (Rothbard,
2009 [1962], p.
689). In
order to show that a �rm obtains a monopoly price by
restricting output along an inelastic demand curve to increase revenue, one has
to assume the original price was a competitive price. Yet this assumption is com-
pletely arbitrary because a competitive price cannot be identi�ed, since in both
cases the �rm is trying to produce at the most pro�table level of output. Rothbard
notes that there is no way to distinguish the monopoly price situation from a
situation in which the capitalist-entrepreneur has overestimated the demand for
a given stock of a good and realizes he can earn more by producing less. Moreo-
ver, this analysis of a “restriction of production” is spurious in general once it is
realized that decreased production releases factors of production that can be used
elsewhere in the economy. Those factors will go to more pro�table uses that better
satisfy consumer wants (Rothbard, 2009 [1962], pp.
638, 690).
Even the require-
ment that the demand curve be inelastic above the competitive price is arbitrary.
As Méra (2010, pp.
51–55; 2015)
points out, the demand curve can be elastic
above the competitive price and yet net revenue still increase because costs of
production fall by even more. The competitive-monopoly price distinction that
tries to show a violation of consumer sovereignty is a misleading partial equilib-
rium framework that narrowly concentrates on one market instead of looking at
the entire economy.
What of case (a), a situation of perfect competition where the demand curve is
perfectly elastic? Rothbard seems to use it only grudgingly and even contradicts
himself at times when defending its applicability. He describes the perfectly com-
petitive case (a) as a situation where, regardless of how much the �rm produces,
“the market-supply curve will not be affected suf�ciently to lower the price”
(Rothbard, 2018 [1953], pp.
78–79, 81–84).
One could extend Rothbard’s logic
to the traditional neoclassical juxtaposition of the �rm’s perfectly elastic demand
curve and the industry’s downward-sloping demand curve (Stigler, 1946, p.
Individual Firm’s Demand Curve
Figure 3.1

This comparison
highlights the fact that the �rm’s marginal contribution to
output is “so” small it cannot affect the price. Only large enough increases in sup-
ply can accomplish that. Rothbard writes that even in case (a), with a change in
output, there has to be “some effect” on the supply curve that will “tend to affect
the price,” but that the overall change would be “too small to alter the point of
intersection” (Rothbard, 2018 [1953], p.
79). When
he later explains that the mar-
ket supply is always affected, he states that, “It may well be, of course, that, within
the relevant range, the action of the �rm is not large enough in relation to the
Rothbard’s defense of the model is confusing. How is it possible that every
change in supply must tend to affect the price, but in some scenarios, when the
change is small enough, price does not change? Either every change in�uences
prices, even in extremely small ways, or some marginal amounts of supply have
no impact on prices. Basic economic analysis shows that Rothbard’s original logic
is correct: every nonzero change in supply must change the market price. Total
market supply is the sum of the individual supplies produced by the �rms; an
The error derives from approximating a very small number as zero. For example,
Stigler uses the example of a market with 10,000 sellers, where each produces an
equal quantity. If one �rm increases its sales by 100
percent, then
the total quantity
increases by 0.0001
percent, an
amount so imperceptible that it can be treated as
zero (Stigler, 1946, pp.
91–92). But
percent is
not equal to 0
percent. Stigler
(1957, p.
8) recognizes
this and writes that only when there are an in�nitely large
number of �rms does every �rm have zero in�uence on price and a perfectly elas
tic demand curve. Otherwise, every �rm has an (albeit extremely small) nonzero
in�uence on price and faces a downward sloping demand curve. However, he also
maintains that as the number of �rms increase, the market approximates perfect
Yet in�nity is not a real number and can never be “reached,” so it is
approximating a
very small number as zero carries enormous implications that drastically change
how the market is analyzed. Every �rm, no matter how small, must have an in�u
The constant price assumption does seem plausible when taken from the real
world perspective of an individual �rm with many competitors. The �rm, when
entering the market, “looks around” at the going market price for a given homog-
enous product. It uses this as an estimate and appraisement of the future sell-
ing price of its product. It may in fact sell all its stock at the ruling price. But
this does not mean that the demand curve is horizontal; all it means is that the
demand curve shifted slightly outward to accommodate the increased supply, or
that some producers now have a surplus of unsold goods (which means not all of
the supply was sold at the market price). If all else could be held equal in the real
world when the producer sells his additional supply, then the price would have to
fall (Armentano, 1999 [1982], p.
23). The
isolated �rm approach is completely
unhelpful and misleading when analyzing the formation of the output price. Only
when one understands this and abstracts from the �rm to look at the market as
a whole is the fallaciousness of the approach revealed. Even though the output
Evolution of Rothbard’s production theory
price is
determined by the industry supply and demand curves, each individual
�rm within the industry confronts a downward-sloping demand curve that gives
it some in�uence over price. Rothbard later realized this point and as a result
concluded that the perfectly elastic demand curve was a deceptive illusion and
addition to his critique of the competitive-monopoly price distinction, Roth
bard’s argument that the perfectly competitive framework is invalid because it is
impossible for a �rm not to have any in�uence on its output price is a unique criti
cism in the Austrian tradition. The most well-known Austrian work in this �eld is
Hayek (2009 [1946]), which does not attack the price-taker assumption, but rather
the assumption of perfect knowledge. Hayek emphasizes that competition is not
an end state but rather a dynamic process involving entrepreneurship and uncer
tainty. The capitalist-entrepreneur is not given the relevant knowledge, such as the
demand curve, but instead must estimate and discover it in the market. Similar
Austrian criticisms include Ludwig Lachmann (1977 [1954]) and Israel Kirzner
(2011 [1963], pp.
312–315; 1973,
71–74, 90–95).
Real world competition is
not an optimization problem with given demand and cost curves, but rather requires
uncertainty-bearing capitalist-entrepreneurs trying to estimate consumer wants and
allocate resources using economic calculation, i.e., through pro�t and loss account
ing (Mises, 1949, pp.
349–354). This
is not to say that Rothbard disagreed with
this argument. In fact, he embraced and emphasized it in his analysis.
He simply
added additional arguments regarding competitive prices and perfect competition.
Once we realize that every �rm’s demand curve slopes downward, no matter
how slightly, it becomes clear that all �rms engage in imperfect competition, and
the differences between them are of degree, not of kind. In this case, the degree
refers to the relative slope of the demand curve, while a difference of kind would be
between a downward-sloping demand curve and an impossible horizontally sloping
one. Yet there is no difference between the kind of demand curve faced by the small
wheat farm and the one faced by the Hershey Chocolate Company, as both are
downward-sloping (Rothbard, 2009 [1962], pp.
721–722). This
carries important
implications for some neoclassical ef�ciency analyses of competition. It is some
times held that imperfect competition is less ef�cient than perfection competition
because the former’s most pro�table level of output is exists only where MR
= MC
and not where P
MC. However
, �rms with downward-sloping demand curves are
not necessarily “inef�cient,” nor do they “restrict output” or necessarily “misallo
petition to which
they are compared is an impossible standard that cannot obtain in the real world.
All �rms face a downward sloping demand curve of some kind and are imperfect,
contrast between Rothbard (2018 [1953]) and Rothbard (2009 [1962]) on
competitive and monopoly prices and perfect competition is stark. In the former,
the concepts are used to analyze actual markets in the unhampered economy. In
the latter, in addition to emphasizing the older Mises-Hayek position that competi
tion is a rivalrous process that involves ef�cient entrepreneurs earning pro�ts and
inef�cient entrepreneurs sustaining losses, Rothbard emphasized that there are no
differences in kind between various �rms and their output prices. Every price is
a market
price based on entrepreneurs’ estimations of the wants of consumers.
Every �rm exercises some in�uence on its output price and cannot take it as given.

The factor demand curve and the causal in�uence
Like �xed output prices, �xed input prices are a tool used in much of neoclassi-
cal production theory. For example, they are prevalent in isocosts, cost curves,
and perfect competition in factor markets. Rothbard (2018 [1953]) frequently
employed this assumption, but later realized it was highly misleading and dis-
carded it in exchange for a framework that explained the formation of input prices
without taking them as given. This section surveys Rothbard’s analysis of iso-
quants and isocosts in preparation for tracing out factor demand curves, and com-
pares it with his later derivation of the factor demand curve, as well as his remarks
After �nishing Section
2 with
a discussion of the competitive-versus-monopoly
price distinction and some possible de�nitions of monopoly, Rothbard returns
to the individual production decisions of a �rm in “Section
3: The
Product and
Outlay Schedules of the Firm.” Aside from unique terminology, which will not
be used here for ease of exposition, the section provides a fairly familiar exercise
in production theory. Rothbard derives isoquants and isocosts, and shows that
the slope of the isoquant is the marginal rate of technical substitution, and the
slope of the isocost is the ratio of the �xed prices, and that the cost minimizing
combination of factor inputs for a given level of output is where the slopes of the
isoquant and isocost are equal (Rothbard, 2018 [1953], p.
96). Later
, in a subsec-
tion, Rothbard engages in a mathematical and graphical formulation of the above
Figure 3.2

Evolution of Rothbard’s production theory
P1, P2
. P7
are the �rm’s isoquants for given levels of production, while O1,
. O7
are the �rm’s isocosts for given levels of expenditure based on the �xed
input prices for X and Y. A, B
. G
represent the cost minimizing combinations
Aside from illustrating a �rm’s optimal production decisions, Rothbard seems
to undertake this analysis in order to draw out a �rm’s demand curve for a factor,
analysis which apparently was to take place in another chapter
that appears
not to
have been written. Rothbard initially writes that the isoquant-isocost apparatus is
“essential to an analysis of the prices of factors of production” and describes it as
one that “will be handy in later analyses of the pricing of factors of production”
(Rothbard, 2018 [1953], pp.
96, 103).
The demand curve Rothbard appears to
have wanted to derive would have been taken from Weiler (1952, pp.
The modern analysis based on this approach, described in Benjamin, Gunder-
son, and Riddell (2002, pp.
incorporates both scale and substitution
effects, which are similar to the familiar income and substitution effects in con-
sumer demand curves, and is shown in Figure
Scale Effect
Substitution Effect
Firm’s Demand Curve
Figure 3.3

Derivation of a �rm’s demand curve for a factor of production
At prices Px and Py, the �rm originally demands Q1 of X. To show the down-
ward sloping nature of a factor demand curve with the isoquant/isocost approach,
let the price of factor X rise relative to the price of factor Y from Px to Px*. This
pivots the isocost inward to re�ect the higher price of factor X, which is then
shifted rightwards until it is tangent with the original isoquant at a new cost mini-
mizing combination B, and as a result, the �rm demands only Q2 of X. The dif-
ference between Q2 and Q1 is the substitution effect. There is also a scale effect
that occurs because the increase in the relative price of X increases the �rm’s
marginal costs, which causes it to produce at a lower level of output and raise its
price. Thus, the �rm’s new isocost shifts leftward from where it was tangent with
the original isoquant to a point where it is tangent to the new pro�t-maximizing
isoquant. The cost minimizing combination is now at A
and the
�rm demands
only Q3 of X. The difference between Q3 and Q2 is the scale effect. Overall, at
the higher
price, the �rm demands less of factor X and more of factor Y, and so
its demand curve for factor X slopes downwards (Benjamin, Gunderson, and Rid-
is important to realize that this process takes place from the perspective of
an individual �rm facing given input prices. Nowhere does Rothbard explain how
this given input price was originally determined, or why it changed. Furthermore,
there is no recognition of the fact that input prices are imputed from output prices.
In fact, in this framework, the analysis implies the reverse. A
rise in
the price (for
some reason) of input X leads to a decrease in the production of the output and
a rise in its price. Rothbard realized the weaknesses of this approach and argued
that “the chief error is that of basing a causal explanation of factor pricing on the
assumption of given factor prices
” (Rothbard, 2009 [1962], pp.
454–455; empha-
Rothbard, the correct method of deriving a demand curve for a factor of
production is through marginal productivity analysis that does not assume the
price of the factor is already given. Moreover, this procedure does not start from
the perspective of a �rm but rather from the general demand for the factor and
its interrelations. In the Austrian static general equilibrium known as the Evenly
Rotating Economy (ERE), the price of any given factor is its discounted marginal
revenue product (DMRP). It must be noted that there is an important yet neglected
difference between Austrian “Böhm-Bawerkian” input price theory and the neo-
classical “Knightian” input price theory. Namely, for the Austrians, a factor price
is equal to its DMRP and not simply its MRP, because the capitalist-entrepreneur,
due to time preference, receives an interest return on his investment from supply-
To be fair, the MRP in these approaches refers to different things. The neo-
classical view argues that the factors are paid their MRP if the MRP is taken to
be the revenue from selling the immediate semi-�nished product. However, this
semi-�nished product does not mean anything to the capitalist-entrepreneur; what
he cares about is selling the
product for expected
money. And since
future money is discounted due to time preference, he will only supply to fac-
tors in the present a smaller amount of present money. This difference is interest
(Block, 1990; Rothbard, 2009 [1962], pp.
504–507; 201
1a [1987], pp.
The Austrian
approach thus emphasizes the importance of time and futurity in
production, as opposed to the Knightian view. The Knightian view is present
in the works of Stigler and contemporary price theorists, and is traceable to the
works of Clark, Marshall, and Walras. It neglects the temporal structure of pro-
duction analysis involving heterogeneous capital goods, and as a result its capital
DMRP is the marginal physical product (MPP) of a factor times the mar-
ginal revenue earned from its employment discounted by the pure rate of interest.
The DMRP of a factor in its general uses (among different production processes
and in a single production process) is decreasing as its supply increases because
both the MPP and the output price fall as output increases. The DMRP of a factor
in a particular process is the point where the total stock of the factor intersects the
Evolution of Rothbard’s production theory
general DMRP
curve, and, through the entrepreneurial process, the price of the
factor is brought into alignment with its marginal use (Rothbard, 2009 [1962],
456–476). If
the prices of a given factor are unequal, then entrepreneurs shift
factors from lower-priced lines of production into higher-priced ones to try to earn
pro�ts, thereby bidding up the price of the factor in the former and lowering it in
the latter until the uniform DMRP is established.
The prices of the factors of production are ultimately determined by the output
prices of the goods they produce. The causal formation of prices is as follows:
anticipated future output prices determine present input prices, or the costs of
production, not the reverse. This is Böhm-Bawerk’s (1889 [1959], pp.
1962 [1894])
“Law of Costs” that Rothbard emphasized in his analysis of the
�rm (Rothbard, 2009 [1962], pp.
361, 588–589).
The �xed input prices a �rm
sees are not costs determined beyond its control. They are the prevailing prices of
factors based on other capitalist-entrepreneurs’ estimations of their marginal uses
elsewhere as determined by consumer demand. In short, these prices re�ect the
opportunity cost of using the factors in other lines of production. By entering into
this factor market, the �rm directly in�uences the formation of the new prices by
bidding them up, which would increase them if other intervening processes did
not occur during the interim. This is similar to the �rm’s in�uence on its output
– the
output price appears given, but in reality, by producing for the market
the �rm is directly contributing to the formation of the price and does exert an
in�uence. As opposed to the original factor pricing theory that Rothbard planned
to write, his revised theory is one that explains the formation of the factor price
without assuming it is given, and clearly shows the causal in�uence of output
above factor demand curve derivation is similar to the derivation of the
short-run demand curve for a factor where marginal productivity analysis is also
used (Stigler, 1946, pp.
175–178). This
is opposed to the long-run demand curve
for a factor, which is the isoquant-isocost method critiqued above. The short run
refers to a period where some factors are �xed and cannot be changed, while in
the long run all factors are variable. While Rothbard and Stigler’s approaches are
largely similar for the short-run curve compared to the long-run demand curve,
there are important differences. Aside from the fact that the factor in Rothbard’s
analysis earns its DMRP while in Stigler’s it earns its MRP, the main difference
is that Stigler’s analysis begins from the vantage point of a �rm facing a given
input price (as determined in the general market). The �rm then hires the factor
along its diminishing MRP schedule until the MRP is equal to the price of the
input, because that is the point where its MR
MC. It
is true that in the ERE, the
�rm will produce where MR
MC, as
any other output level will lead to negative
pro�ts (as opposed to the point at which MR
MC, where
the �rm will earn zero).
However, as will be explained below, in the dynamic world, the pro�t-maximizing
level of output for the capitalist-entrepreneur investing in the �rm is generally
where the �rm’s MR
is greater
than its MC. When applied to a non-general equi-
librium world, however, this factor demand curve derivation is incorrect, as the
�rm’s MR
gence does not always occur.

The capitalist-entrepr
eneur and the optimal level
This section is framed differently from the prior two because it presents an impor-
tant critique of the Marshallian partial equilibrium �rm theorizing, a critique that
Rothbard (2018 [1953]) describes but does not fully emphasize in Rothbard (2009
In the �nal elongated section that the present author pieced together,
tively titled “Section
4: The
Output and Investment Decision of the Producer,”
Rothbard strives to develop an optimal theory of investment of the capitalist-
entrepreneur. The implication of this analysis is that in the dynamic world where
there is a mélange of �rms with varying degrees of pro�tability for their capitalist-
entrepreneur investors, there can be no theory of optimal investment formulated
for a single �rm in isolation. This implies that a given �rm’s optimal production
may not be where its MR
MC. This
momentous realization undoubtedly con-
tributed to Rothbard’s later decision to discard the isolated �rm analysis and com-
pletely rewrite his production theory.
Although in many ways Rothbard (2018 [1953]) is quite similar to traditional
neoclassical production theory, one major difference between the two is that the
“producer” Rothbard concentrates on is a capitalist-entrepreneur investing his own
money, while the neoclassical producer is a propertyless manager who can bor-
row an unlimited amount of money at a given interest rate and
given �rm
. When analyzing the �rm, Rothbard views the capitalist-entrepreneur
as the controlling factor who earns an interest rate of return on his money invested,
while the manager is a hired factor of production whose income is a money cost.
The capitalist-entrepreneur can also earn a pro�t when his total rate of return is
greater than the interest return (Rothbard, 2009 [1962], p.
510). In
the neoclassi-
cal framework, the entrepreneur is considered to be a propertyless manager and
the controlling factor in the �rm who pays an interest return on money borrowed
from capitalists, which is counted as a cost of production. The manager still earns
a management wage, which is also counted as a cost of production, but also can
earn a pro�t, which is the difference between his total revenue and the principal
and interest payments on his borrowed money. The former has been explicitly
called the “Austro-Wicksellian” theory of the �rm, which views interest as rev-
enue, as opposed to the neoclassical theory, which views it as a cost (Gabor and
difference is present even in the beginning of Rothbard’s chapter. In Sec
1, before
discussing the various production decisions and possible demand
situations a single-product �rm might face, Rothbard brie�y looks at the opti
mal investment decision of a representative capitalist-entrepreneur who can invest
in multiple lines of production. The capitalist-entrepreneur has a given stock of
money that he can spend on consumption, investment, or keep in his cash bal
ances. He will only invest if the rate of return is greater than or equal to his rate
of time preference, the speci�c premium on present money over future money
that represents the minimum return the capitalist-entrepreneur requires in order
Evolution of Rothbard’s production theory
to invest.
For a given amount of money invested, the capitalist-entrepreneur will
choose the line of production that maximizes his expected monetary return (Roth
bard, 2018 [1953], pp.
75–77). Rothbard
then postpones further analysis of the
investment theory, and for the rest of Section
1 through
3 turns
to analyz
ing competitive prices, monopoly prices, perfect competition, and the isoquant-
isocost apparatus that shows the cost minimizing level of output. This is described
above in Sections III and IV of the present chapter. After this analysis, Rothbard
then returns to investigating investment decisions in the �nal section of his chapter.
In Section
4, Rothbard
considers a capitalist-entrepreneur who has chosen
to invest in a given �rm that produces Product P. Rothbard describes various
amounts of expenditure, or total costs, which lead to various total revenues the
capitalist-entrepreneur can earn by investing in the given �rm. For ease of exposi-
depicts Rothbard’s (2018 [1953], p.
results with the familiar
tal Cost
tal Cost
Figure 3.4

Graphical illustration
of money revenue and outlay from production of Product P
Rothbard calls the difference between total revenue and total cost “net income,”
Now, Rothbard asks, what level of output will be chosen? How much will the
capitalist-entrepreneur invest in the given �rm, and, consequently, how much
Product P output will the �rm produce? What is the optimal level of output? The
traditional answer is that the optimal level of investment and output is where net
money income is maximized, i.e., the greatest distance between total revenue and
total cost. In familiar terminology, using continuous curves, this is the point where
MC. As
explained above, the usual perspective taken is that of the prop-
ertyless manager, with a �xed interest return to the investing capitalist included
in the cost curve. To maximize his pro�t amount, the manager of the �rm should
borrow from
the capitalist-entrepreneurs at a given interest rate and invest until
, the crucial problem with this analysis is that it neglects the ques-
tion of whether or not the capitalist-entrepreneurs can invest their money in other
�rms that earn higher than the rate of return earnable in this �rm. The output level
that maximizes the �rm’s pro�t may not be the output level where the capitalist-
entrepreneurs who invest in the �rm maximize their pro�ts. This is because
capitalist-entrepreneurs are not restricted solely to investing in a given �rm but
can also invest in others where they could potentially earn a higher rate of return
on their marginal money invested. The rate of return that can be earned always
varies, because in the real world there is uncertainty and, consequently, pro�ts
and losses. Only in a static general equilibrium devoid of uncertainty are all rates
of return uniform (and equal to the interest rate). The problem does not go away
if we postulate that the given �rm can produce multiple products, since the unit
of analysis is still “a �rm” rather than the capitalist-entrepreneurs who can invest
among multiple �rms producing multiple products. To maximize their total prof-
its, capitalist-entrepreneurs may spend only a certain amount of money in one �rm
such that it produces where its MR > MC because they can invest in other �rms
to increase their total pro�ts more than they could if they were to solely invest in
the original �rm up to the point where its MR
MC. As
a result, one cannot look
at a �rm in isolation in the partial equilibrium approach and �gure
out how
will be produced or invested in it (Gabor and Pearce, 1952, 1958; Rothbard, 1961,
7–8, 18;
Rothbard, 1993; Klein, 2010c [1999], pp.
38–39; Klein
and Foss,
2012, p.
238). Although
Rothbard (2018 [1953]) does not frame the problem in
terms of MR
and MC,
this is clearly what he is getting at, as he trenchantly writes
that “there is no precise theory of the determination of the investment in, and out-
To clarify, this is not to say that in an isolated �rm the maximization of pro�t
for the manager does not occur at an output level where its MR
Given a
total revenue and total cost curve for the �rm that includes interest payments,
the optimum is clearly where MR
What it does say is that from the van
tage point of capitalist-entrepreneurs who supply funds to the �rm, and once the
range of investment opportunities is broadened beyond the individual �rm to the
entire production structure, in a dynamic world where lines of production earn
different pro�table (i.e., above interest) rates of return, pro�t maximization may
occur where the �rm’s MR > MC, since capitalist-entrepreneurs can invest in
other industries where they can reap potentially greater economic pro�ts. Only
in the ERE, where all pro�ts are wiped out and all lines of production earn the
same uniform interest return will the optimal level of output be at the point where
the �rm’s MR
MC. This
is because the capitalists cannot invest their funds in
another �rm to earn a higher-than-normal rate, as such opportunities do not exist
(Rothbard, 2009 [1962], pp.
600, 695,
734–736). At such an output level, since
pro�t is zero, total revenue will be equal to total cost, assuming the capitalists’
interest return is included as an opportunity cost. At any other output level, there
will be negative pro�ts. In order for the �rm to exist in general equilibrium, it
Evolution of Rothbard’s production theory
must not
earn negative pro�ts, so it will have to produce where its MR
problem appears when the economy is out of general equilibrium, and not
all lines of production earn the same uniform interest return. Now the produc
tion of a �rm may not be where its MR
MC, since
the capitalist-entrepreneurs
who invest their money in it can invest in other, more remunerative �rms. The
Austrian general equilibrium approach stresses this dynamic world because it is
the dynamic world in which we live and that consequently we try to understand.
On this entire issue, we �nd much on which to agree with Gabor and Pearce
(1952, 1958), whose articles
– which
heavily in�uenced Rothbard
– emphasize
this important distinction between the Austrian and neoclassical approaches with
regards to optimal production in a given �rm in disequilibrium. It is worth quot
[T]here is much to suggest that a great deal has been lost by the failure to
produce a more adequate synthesis of all that is best in the work of both the
Austrian and the neoclassical schools. In the �rst place, the fact that two theo-
ries of pro�t lead to the same general equilibrium is not suf�cient to make
them the same theory.
The route by which equilibrium is attained is often as
important as the equilibrium itself
. We have shown elsewhere [Gabor and
Pearce, 1952] that, if general equilibrium has not been attained, and the fact
that an investment is being contemplated in any industry implies that it has
not, then the two theories of pro�t, the [Austro]-Wicksellian and the neoclas-
sical, lead to different conclusions
of trying to look at the optimal level of investment in and output of a
compartmentalized �rm, Rothbard argues that the correct approach is to develop
an optimal theory of investment of the capitalist-entrepreneurs who face a gamut
of various �rms in which they can invest (Rothbard, 2018 [1953], pp.
This is what Rothbard seeks to do throughout the rest of Section
4. Rothbard
that a theory of optimal investment decisions needs to focus not only
on maximizing the rate of return on a given amount of money invested, but on
weighing the rate of return with the capitalist-entrepreneur’s rate of time prefer-
ence. Rothbard calls this the Law of Investment Decision (Rothbard, 2018 [1953],
121). The
capitalist-entrepreneur will invest his own money
in general
up to the
last discrete point where the average and marginal rate of return are greater than or
equal to his average and marginal rate of time preference.
In each individual �rm
the capitalist-entrepreneur invests up to the last discrete point where his marginal
rate of return in that �rm is greater than or equal to the marginal rate of return of
s revised production theory (2009 [1962]) unfortunately does not
explicitly mention any Law of Investment Decision for the capitalist-entrepreneur.
However, it does start with an analysis of time preference, interest rates, and the
important function of the capitalist who invests in a temporal structure of pro-
duction in the ERE (Rothbard, 2009 [1962], pp.
319–451). The
ERE is essential
because, in
Rothbard’s presentation, it is the point toward which the economy
always tends, and would reach if the data remained constant. It is necessary in
order to isolate the differences between pro�t and loss, on the one hand, and inter-
est on the other. In addition, the ERE is indispensable for deducing economic
theorems, as it allows the economist to mentally hold constant all changes in the
data except one in order to isolate the effects of that change. However, Rothbard
extended the edi�ce in order to describe the
processes to the ERE
, the processes
that describe the dynamic world in which we live. Capitalist-entrepreneurs take
center stage by investing in various pro�table �rms with different periods of pro-
duction and by engaging in a rivalrous process of ef�cient competition with each
other that consequently distributes scarce resources according to the intertemporal
preferences of consumers. In the end, Rothbard substituted an Austrian general

This paper provides a comparison of Rothbard’s earlier production framework
with his �nal system. The earlier theory was closer to the Marshallian partial equi-
librium theory. It analyzed production from the perspective of an isolated �rm and
employed many standard tools such as the competitive-monopoly price distinc-
tion, perfect competition theory, the isoquant-isocost framework used to derive
factor demand curves, and the isolated �rm. Rothbard’s later revised theory, and
in some cases parts of his earlier theory, criticized these tools.
It is shown that Rothbard argued that the competitive-monopoly price distinc-
tion is inappropriate for analyzing free market situations because it arbitrarily
assumes the existence of a competitive and a monopoly price. In addition, all
�rms are price searchers that exercise some in�uence on their output price and
consequently engage in “imperfect” competition. Thus, the traditional ef�ciency
benchmark of perfect competition is a poor standard because it is impossible
to attain in the real world. The factor demand curve derived from the isoquant-
isocost framework does not show that expected output prices determine input
prices, and assumes the input price it tries to explain. Finally, and most impor-
tantly, in order to understand optimal production in a �rm, the correct perspective
is not that of the borrowing manager in the isolated �rm who seeks to maximize
his pro�t, but rather that of the capitalist-entrepreneurs who supply money for the
�rm while seeking to maximize the rate of return on their total capital invested.
This is because
– unlike
the manager
– the
capitalist-entrepreneur can invest in
multiple �rms, which means that in a dynamic world, he may not maximize his
, the evolution of Rothbard’s production theory leads from the
Marshallian partial equilibrium approach to the Austrian general equilibrium
approach. This is because Rothbard shows the causal in�uence of input and out-
put prices and actually explains their formation, and his mature theory does not
analyze production from a single isolated �rm that can treat prices as �xed in a
Evolution of Rothbard’s production theory
static world,
but rather from the decisions of competing pro�t-seeking capitalist-
In this paper, the “Chicago theory of the �rm” refers to the “black box” production
view, where the �rm is given a set of underlying data and chooses its out-
put level through mathematical optimization (Klein, 2010a [1996], pp.
3–4; Foss
Klein, 2012, pp.
136–137). It
does not refer to Coasean theories that analyze the rea-
sons for the existence, organization, and limits of the �rm. With regard to the latter,
Rothbard was heavily in�uenced by Coase and was one of the �rst economists to
Mark Blaug
has similarly contrasted what he calls Austrian “total equilibrium analy-
sis” with both Marshallian partial equilibrium and Walrasian general equilibrium anal-
Mises asked
Rothbard on several occasions throughout this period to present some of
the chapters at his New York University seminar (Hülsmann, 2007, pp.
The price
assumption also applies to input prices, which are treated in the next section.
This is
not to deny, as Rothbard emphasized, that the distinction is unimportant when a
government intervention allows the �rm to restrict output and attain a monopoly price.
Here, the element of coercion has tampered with the voluntary actions of consumers,
and now the restriction exists in de�ance of their choices. For the consequences of fac-
See also Hirschleifer
al. (2007, p.
165), who ar
gue along similar lines: “While [per-
fect competition is] never literally true, this may approximate reality if the �rm pro-
duces only a small fraction of the output in its industry.”
For more,
see Keen (2011, pp.
76–77, 85–90,
95), who presents a critique very similar
to Rothbard’s (2009 [1962]), although from a more mathematical perspective.
Kirzner notes
that Mises thought the model was confusing and unrealistic, and inten-
tionally avoided discussing it (Kirzner, 2001, pp.
While agreein
g with Kirzner’s dynamic critique of the model, Rothbard did criticize
Kirzner’s use of the perfectly elastic demand curve (Kirzner, 1963 [2011], pp.
Rothbard, 201
1b [1961], pp.
14–15). See
Mises (1949, pp.
356–357) for
his limited
Rothbard (2018
[1953]) cited Weiler (1952, pp.
in his discussion of factor ratios
and production coef�cients, and Weiler (1952, pp.
141–161) was
also cited later in
Rothbard (2009 [1962], pp.
589–590), in
his critique of cost curves. The above citation
is to a Weiler chapter
devoted entirely
to the analysis of the demand for the factors of
production. Although Rothbard also cited Stigler (1946) in both places, his demand
curve is derived differently and is discussed below. For an explanation similar to
Weiler, see Hicks (1979 [1946], pp.
89–98). Kirzner
(1963 [2011], pp.
194, 215–216)
See the editor’
Armentano, Dominick T. 1999 [1982].
Antitrust and Monopoly: Anatomy of a Policy Fail
. Oakland, CA: The Independent Institute.
Benjamin, Dwayne, Morley Gunderson, and W. Craig Riddell. 2002.
” In
Bruce J. Caldwell, ed.,
. Durham,
Block, Walter. 1990. “The DMVP-MVP Controversy: A
Review of
Austrian Eco
Böhm-Bawerk, Eugen von. 1959 [1889].
Capital and Interest
. South Holland, IL: Libertar-
[1894]. “The Ultimate Standard of Value.” In Eugen von Böhm-Bawerk, ed.,
Foss, Nicolai, and Peter Klein. 2012.
Organizing Entrepreneurial Judgment: A
Approach to the Firm
York: Cambridge University Press.
Gabor, Andre, and I. F. Pearce. 1952. “A
New Approach
to the Theory of the Firm.”
“The Place of Money Capital in the Theory of Production.”
Hayek, Friedrich August. 2009 [1946]. “The Meaning of Competition.” In
and Economic Order
. Auburn, AL: Ludwig von Mises Institute.
Value and Capital
Hirschleifer, Jack, Amihai Glazer, and David Hirschleifer. 2007.
Price Theory and Appli
. New York: Cambridge University Press.
Huerta de Soto, Jesús. 2006 [1998].
Money, Bank Credit, and Economic Cycles
. Auburn,
Hülsmann, J. Guido. 2007.
Mises: The Last Knight of Liberalism
. Auburn, AL: Ludwig
Keen, Steve. 2011.
. New York: Palgrave Macmillan.
Kirzner, Israel. 2001.
. Wilmington, De: ISI Books.
1 [1963].
Market Theory and the Price System
. Indianapolis, IN: Liberty Fund.
Competition and Entrepreneurship
. Indianapolis, IN: Liberty Fund.
Klein, Peter G. 2010a. “Economic Calculation and the Limits of Organization.” In
Capitalist and the Entrepreneur
. Auburn, AL: Ludwig von Mises Institute.
“The Mundane Economics of the Austrian School.” In Peter G. Klein, ed.
The Capitalist and the Entrepreneur
. Auburn, AL: Ludwig von Mises Institute.
“Entrepreneurship and Corporate Governance.” In Peter G. Klein,
Capitalist and The Entrepreneur
. Auburn, AL: Ludwig von Mises Institute.
Lachmann, Ludwig. 1977 [1954]. “Some Notes on Economic Thought, 1933–1953.” In
Ludwig Lachmann, ed.,
Capital, Expectations, and the Market Process
. Menlo Park,
McGraw Hill. 2002. “Chapter
Notes.” McGraw
Hill Online Learning Center. http://high
Méra, Xavier. 2010. “Factor Prices Under Monopoly.”
Quarterly Journal of Austrian Eco
Note on
the Limits to Monopoly Pricing.” In Per Bylund and David
Howden, eds.,
The Next Generation of Austrian Economics
. Auburn, AL: Ludwig von
Mises, Ludwig von. 2008 [1949].
. Auburn, AL: Ludwig von Mises Institute.
Moroney, John R. 1972. “The Current State of Money and Production Theory.”
Evolution of Rothbard’s production theory
1954. Letter to R. Cornuelle, “Textbook or Treatise?”
“The Science of Liberty: An Interview With Murray N. Rothbard.”
America’s Great Depression
. Auburn, AL: Ludwig von Mises
[1993]. “Letter to Sandra K. Johnson.” From Peter G. Klein, 1999. “Entre-
preneurship and Corporate Governance.” In Peter G. Klein, ed.,
The Capitalist and The
. Auburn, AL: Ludwig von Mises Institute.
1a [1987]. “Breaking Out of the Walrasian Box: Schumpeter and Hansen.”
In Murray N. Rothbard, ed.,
Economic Controversies
. Auburn, AL: Ludwig von Mises
1b [1961]. “Comments on Israel M. Kirzner’s MS.”
Libertarian Papers
3 (25): 7–23.
[1953]. “Chapter
5: Producer
’s Activity.” In Matthew McCaffrey, ed.
Economic Theory of Costs: Foundations and New Directions
. Abingdon, UK: Rout-
Salerno, Joseph T. 2004. “Menger’s Theory of Monopoly Price in the Years of High The-
ory: The Contribution of Vernon A. Mund.”
“Introduction to the Second Edition.” In Murray N. Rothbard, ed.,
1. “Varieties of Austrian Price Theory: Rothbard Reviews Kirzner.”
Stigler, George. 1947 [1946].
. New York: Palgrave Macmillan.
“Perfect Competition, Historically Contemplated.”
Journal of Political
Stromberg, Joseph. 2004. “Introduction to Man, Economy, and State with Power and Mar-
ket.” In Murray N. Rothbard, ed.,
Editor’s foreword
The following manuscript is an unpublished chapter
of Murray
N. Rothbard’s
Man, Economy, and State
(hereafter, MES) (2009 [1962]). As the tentative title,
5: Producer
’s Activity,” indicates, it was meant to be the �fth chapter
the book
and the �rst on production theory. In it, Rothbard discusses the optimal
production and investment decision of the producer, and uses familiar analyti-
cal tools such as perfect competition, the isoquant-isocost framework, and the
competitive-versus-monopoly price distinction. This chapter
was drafted
Rothbard still planned to write a textbook version of Ludwig von Mises’s
(1999 [1949]), before he decided to write a full blown treatise. Rothbard’s
decision to change course was heavily in�uenced by his concurrent decision to
abandon this chapter
and rewrite
his production theory, as he thought that the
new material would be unsuitable for an introductory textbook. In particular, the
above-mentioned analytical tools were all subjected to trenchant criticism by
Rothbard in his �nal production theory.
The chapter
is meant
to serve as a companion to Newman (2018), a concur-
rent paper written by the present author to discuss the evolution of Rothbard’s
approach to production theory and its implications for modern theory. However,
this chapter
is also
a valuable resource for scholars interested in Rothbard and
Austrian economics, as well as for historians of production theory in their own
The chapter
was found
and reorganized by the present writer in the Murray
N. Rothbard archives at the Ludwig von Mises Institute in Auburn, Alabama.
Over the years, Rothbard saved many of his draft pages for MES on various top-
ics, including both those that would eventually appear in the book and those that
would not. Rothbard did not neatly organize his draft pages, so many of the pages
that could be found next to each other in the archive boxes concerned completely
different topics. Fortunately, Rothbard did number his pages, so the present writer
was able to reconstruct the chapter
by sifting
through the archive boxes and link-
ing up pages based on their pagination and also on the practical matter of whether
the sentences ran from one page to the next. The document that follows pieces
Man, economy, and state
Producer’s activity
Murray N. Rothbard
Murray N. Rothbard
as much as possible the missing chapter
based on
the available surviving
The chapter
is in
a rough stage, as Rothbard appears to have written only one
draft before deciding to revise his production theory and remove the chapter
his planned
work. However, it is written very clearly and is easy to understand,
so in terms of editing the paragraphs, for the most part I
have only
had to make a
few grammatical and stylistic changes regarding his numerical examples. In some
cases, exclusively in the last section, I
had to
add a few words and sentences in
order to clarify Rothbard’s argument. The largest of such additions occur toward
the end of the chapter, where several of Rothbard’s draft pages could not be found.
Consequently, I
had to
�ll in the gap with some summary transitional sentences,
and in one case, a paragraph detailing what I
believe Rothbard
would have dis-
cussed in these pages, based on some references contained later in the chapter. All
additions I
have made
are in brackets, and the reader will see that I
have faithfully
only what can be inferred from the rest of the chapter, and in a style con-
sistent with it. In addition, I
have provided
information in endnotes (prefaced with
Editor’s endnote
) about various references Rothbard makes to either previously
I have also had to make some minor changes to the structure of the chapter.
The three sections I
located were
1, 2,
and 4, and there was a missing
third section of the chapter
that I
could not
�nd. Furthermore, Section
4 included
of the material that is now Section
3, which
made it quite large and unwieldy.
Therefore, I
have split
4 into
two parts based on their respective topics
and given Section
3 an
appropriate title so that there are now four well-organized
In the �rst section, titled “The Demand for a Firm’s Product,” Rothbard con-
centrates on the production function of an individual producer for a given good.
With �xed prices for inputs, Rothbard investigates the optimal production choices
in situations where the �rm either has or does not have an in�uence on the output
price. In Section
2, “Competiti
ve Price and Monopoly Price,” Rothbard intro-
duces the terms “perfect competition,” “competitive price,” and “monopoly
price,” and de�nes a monopolist as someone who receives a grant of state privi-
lege. In Section
3, titled
by the present author “The Product and Outlay Schedules
of the Firm,” Rothbard returns to the �rm’s production decisions and analyzes
factor ratios and production coef�cients. Rothbard derives constant cost (isocost)
and constant product (isoquant) schedules as well as rates of constant outlay and
constant product substitution. Rothbard shows that the cost minimizing level of
output is the point at which these two rates are equal. In a subsection, he presents
a mathematical and graphical formulation of the above theories and brie�y men-
tions their relation to the determination of factor pricing. Rothbard �nishes the
with Section
4, titled
“The Output and Investment Decisions of the Pro-
ducer,” by constructing “The Law of Investment Decision” using the concepts of
rates of net income, marginal, and average rates of return. Rothbard argues that
the investor will not produce at the outlay where either his pro�t amount or
centage rate
of return is maximized, but rather up to the last outlay where the
Man, economy, & state
average and
marginal rates of return are greater than or equal to his average and
marginal rates of time preference. This theory may be useful for those scholars
interested in an Austrian “Theory of Investment,” as it is a portfolio theory of how
the capitalist-entrepreneur allocates his money across various enterprises. This is
linked with a brief criticism of �rm analysis, which undoubtedly in�uenced Roth-
In conclusion, this chapter
will be
a fertile source for both historians of thought
and contemporary theorists interested in Austrian economics and production
1: the demand for a �rm’s product
We have seen that the money prices of goods on the market are set at the intersec-
tion of the demand and supply curves. Setting aside the relatively simple problem
of the market for old stock, the market price and quantity exchanged are deter-
mined by the intersection of the market supply curves of the producers, and the
demand curve. We have seen above that the stock thrown on the market in any
given period is largely determined by
anticipations of market conditions
in this period. This notion has been presented in terms of the “�nal supply curve”
of producers. In other words, if the selling price of a certain line of washing
machines is expected to be 20 ounces of gold next September, how many washing
machines will Smith begin to invest in
so that the �nal product will emerge
next September? We have described this �nal supply in terms of the
calling forth a
investment for a
production; strictly, of course, it is
expected future price
that calls forth investment now for future production. All
Thus, all producers’ activity
– the
central nexus of the economy
– is
based on
certain anticipations of future selling prices. We have analyzed above the deter-
minants of market price for consumer goods, durable and nondurable, and now
we must analyze the “�nal supply curve,” and the process of producers’ activity.
Consumers’ goods, the end of human activity, must be produced by producers,
and the overwhelming number of
goods must be produced through
the monetary exchange process outlined in Chapter
Therefore, analysis of
producer activity is vital; not only will it provide the �nal clue to the analysis of
consumer goods’ prices
– through
discussing the determination of the size of the
stock thrown on the market
– it
is also the key to the analysis of the determina-
tion of the money prices of factors of production. All other goods but consumer
goods are factors, and all of these are demanded and bought solely by producers.
It is producers that purchase with money, capital goods, land, and labor, and it is
producers that use other factors to produce the capital goods. It is only through
a more detailed analysis of producer activity, therefore, that the prices of factors
To analyze the actions of a producer let us take a hypothetical case, Mr.
Jones. Jones, like all others, must decide on the allocation of his money assets to
Murray N. Rothbard
expenditures, consumption expenditures, and to his cash balance. Let
us postpone discussion of changes in cash balance to a later chapter
on the
for money and its utility.
Jones must allocate his expenses between consumption
and investment expenditure. The motive that impels him to spend money on pre-
sent consumption is the grati�cation of his desires through present consumption.
What is the motive that impels him to
a certain amount of money by restrict-
ing his possible consumption, and invest that money in expenditure on various
factors of production? This motive must be the expectation of
income in the future. We have already seen that every man prefers a satisfaction
of a desire earlier than later, and therefore that a given amount of present money
is always preferred to the same amount of money in the future.
At any given
point, he will have a certain rate of time preference, a rate by which he will prefer
present money
to the
present prospect of money at some date in the future
have seen that the more he allocates to investment, the greater will be the marginal
utility forgone of present consumption, and the less will be the marginal utility of
“Investment opportunities” for a greater supply of future consumer goods are
always open to man, because investment in capital goods adds to the capital struc-
ture, and increases future product of consumer goods. On the other hand, man
must satisfy his present needs �rst. Thus, men must always balance their prospect
of future gain as against their rate of time preference for present as against future
satisfactions. The primary activity in deciding whether or not to be a producer is
the weighing of the anticipated future gain against the person’s rate of time pref-
erence. In the words of Professor Fetter, “The different time-periods, present and
future, and their different economic situations are brought into comparison
. by
choice between the thing actually present and the future good more or
We must postpone detailed consideration of time preference and its effects to
later chapters of this work.
Here it suf�ces to point out that each individual has
his own rate of time preference, expressed as a
percentage premium
of present
over future goods, and that the more he saves at any time, the greater his subjec-
Let us suppose now that Jones is considering whether or not to invest 1000
ounces of gold, or spend this money in consumption. Let us say his rate of time
preference for these 1000 ounces is 6%
per annum.
In other words, if he antici-
pates a return on this investment of 6% or less for the following year (assum-
ing for simplicity that only one year is taken into account), he will not make
the investment. Thus, in deciding on productive investment or not, his minimum
return for the year will be an anticipated 1060 ounces. If he anticipates this or less,
Now Jones surveys the prevailing conditions, and estimates that he has avail-
able four different lines of investment. For the sake of simplicity, we will now
assume that all of these lines are in the production of consumer goods (since we
have not yet explained the determination of any capital goods prices), and we will
also assume that the
period of production
for each of these processes is exactly
Man, economy, & state
year. This period of production, as explained in Chapter
I, is
the length of
time from the beginning of the action
– the
– to
the reaping of the
�nal product.
It should be clear that it is a simple task to make the necessary
adjustments in calculation if one or other of the processes takes more or less time
to complete. The four lines of investment open to Jones he estimates will net
him, in the year to be considered, net money returns of 10%, 8%, 7%, and 5%
In other words, with an investment in factors of 1000 ounces now
and in the near future, Jones will be able to reap the following returns for lines of
investment A, B, C, and D, as illustrated in Table
Table 4.1

Jones’ money returns
Gross money return
percentage net
money return is to be
found. Here, we must remember the quali�cation that he will only choose such a
course if other psychic factors are neutral. Thus, if he has a special fondness for
the production of Good B, or a special hostility toward the production of Good
A, the 8% money return may be worth more to him on his value scale than the
10% return to be made in Good A. Noting this quali�cation, however, it will be
convenient for us to set it aside, and assume that psychic factors are neutral in our
example, in which case the investor will always choose the greatest prospects for
We must now investigate the line of production more closely. Suppose that
we con�ne our attention to the line of production that produces Good A. Jones is
eager to maximize the
percentage return
from his investment. What are the factors
that will determine the size of his return? These factors are: a) the money prices
of the factors purchased, b) the selling price of his product, and c) the physical
productivity of the factors in their transformation into the product. It is obvious
that, other things being equal, the
the prices he must pay for the factors, the
greater will be his return; the
the price of his product, the greater his return;
his physical productivity, the greater his return. Let us assume
for the moment that the prices of the factors are given. Jones also discovers that
there is available to him a range of technical possibilities in the production of the
particular good. For the sake of simplicity, let us suppose that only two factors,
Murray N. Rothbard
and Y, are required in the production of Good A. The money price of X and the
money price of Y are �xed on the market
– say
it is 4 ounces
per unit
of X, and
10 ounces
per unit
of Y. Jones knows (or believes) that there are several possible
proportions of X and Y that he can buy with his 1000 ounces in order to produce
Good A. These may be the following:
Table 4.2

Factor combinations for the production of Good A
1)40X plus 84Y
2)50X plus 80Y
3)60X plus 76Y
With prices at 4 and 10, these combinations will all add up to expenditures of
1000 ounces. As Table
depicts, in the �rst combination, Jones spends 160
ounces on X and 840 on Y; in the second, he spends 200 ounces on X and 800
ounces on Y, etc. Now the question arises: which combination does Jones choose
to adopt? First, this depends on the physical productivity of each combination.
This physical productivity is the effect of the production
, a recipe which is
known to the producer in making his decision. The relationship between physical
input and product is sometimes known as the “production function.”
Let us say
Table 4.3

Factor combinations and output for the production of Good A
Resulting product
110 units
It would certainly seem that Jones will pick that combination which will yield
maximum physical output
. In this case, it would be Combination 3, by
which we can produce 110 units from 1000 ounces’ worth of factors. There
is one quali�cation to this course of action, however, and that would be if his
increase in units produced would so lower the market price of the product as to
his gross revenue from the sale of the produced stock. In other words,
suppose as Case (a), that the price of his product will be 10 ounces
per unit,
and that he correctly estimates it as such. Furthermore, suppose that regardless
which production process he chooses, the market price will continue to be 10
ounces. In other words, whether he chooses to produce 100 or 110 or 96, etc.
units, the market supply curve will not be affected suf�ciently to lower the
price. In this case, the gross revenue from the various combinations will be as
shown in Table
Man, economy, & state
Table 4.4

Gross revenue in the production of Good A, Case (a)
Combinations of inputResulting productPrice of product
Cross revenue
1)40X plus 84Y
2)50X plus 80Y
10 oz./unit1000 oz.
3)60X plus 76Y
110 units
1100 oz.
Jones will choose Combination 3, yielding the largest gross revenue and hence
the largest net revenue with a given investment (1000 oz.) and the largest
age net
revenue on the investment. It is evident that, regardless of the number of
alternative combinations available, where the price is constant, the combination
chosen will be the one that maximizes the physical product from a given amount
Now suppose Case (b), where Jones’ production is important enough in the
market supply of his product so that a change from one combination to another
affect the market price at which the product will be sold.
Within the range
of choice of combinations, a larger output will increase the market supply curve
enough to lower the price of the product. It is evident that this is the usual rule
on the market. Strictly, indeed, even in Case (a) there must have been
on the market supply curve from the change in output, however small, and this
minute change will tend to affect the price. In Case (a), however, the change was
too small to alter the point of intersection. In Case (b), the price is affected by
the change in quantity, but not so much as to lower the gross revenue with an
increased output. Thus, a typical situation might be that shown in Table
Table 4.5

Gross revenue in the production of Good A, Case (b1)
Combinations of inputResulting productPrice of productGross revenue
1)40X plus 84Y
10.5 oz./unit1008 oz.
2)50X plus 80Y
10.4 oz./unit1040 oz.
3)60X plus 76Y
110 units
1100 oz.
In this case, the increase in product and supply of the producer lowered the
market price, but not in any case enough to lower revenue. Strictly, this condition
only need prevail, in Case (b),
at and above the point of maximum output
. Thus,
it would have been possible for the price, at a supply of 96 units, to have been
11 ounces
per unit,
and the gross revenue therefore to have been 1056 ounces.
Table 4.6

Gross revenue in the production of Good A, Case (b2)
Combinations of inputResulting productPrice or productGross revenue
1)40X plus 84Y
11.0 oz./unit
2)50X plus 80Y
10.4 oz./unit1040 oz.
3)60X plus 76Y
110 units
10.0 oz./unit1100 oz.
Murray N. Rothbard
it is true that as the supply increases from 96 to 100 units, the con�guration
of the demand curve and the market price is such that the revenue is lowered. How
ever, the important consideration is that
the point of maximum output is also the point
of maximum revenue
. Should Jones shift to another than the maximum combination
in order to restrict the product, the higher price will not be suf�cient to compensate
for the loss of revenue. In both Case (b1) and Case (b2), the producer will choose the
This data can be translated into terms of the
. The individual producer, after all, is not concerned with what the mar-
ket demand curve will turn out to be
– he
is concerned what the price will be for
his particular product. He must ask himself the question: if I
produce so
units, what will the selling price be; if I
produce so
many more units, what will
be the effect on the selling price? In other words, he in effect is estimating what
price the buyers will pay for different possible supplies of
his particular product.
This analysis applies whether or not the producer is one of hundreds producing
the same product, or whether he is the only one producing that good
. In any case,
For Case (b1) above, the
demand curve to the individual producer
can be con-
structed as seen in Figure
110 units
Figure 4.1

Man, economy, & state
When the
supply of Jones is 96, the market price will be 10.5
– in
other words,
consumers will be prepared to demand 96 of Jones’ units at a price of 10.5. This
gives Jones one of the points, 1, on the
demand curve for Jones’ product
. The
price and supply at 10.4 and 100 respectively, and the various other items on the
schedule, yield the other points on this demand curve (such as 2 and 3). These
points are drawn together in one line for convenience. The schedule above also
tells Jones how much of his product will be demanded at any particular price.
Thus, it is clear that the producer knows that if he produces 96 units, they will be
sold for 10.5, and 100 units will be sold for 10.4, etc. He
knows that, regard-
less of the size of his stock, if he sets the price for his product at 10.5 he will be
able to sell only 96 units; if he sets the price at 10.4 he will be able to sell 100
units, etc. Thus, the supply and estimated market price yield him an estimate of a
demand curve for his individual product
. Not only will he know that a sup-
ply of 110 units will provide him with the maximum revenue, he will also know
that, once the 110 units are produced,
it will not pay for him to destroy or withhold
some units in order to raise the price on the remainder
. Thus, with this type of
demand curve for his own individual product, it is to his interest to produce his
maximum physical product, and not to deliberately restrict or withhold his prod-
uct to obtain a higher price. Even if he can obtain a higher price, restriction will
This property of the demand curve for the individual producer, determining
whether decreased production will raise or lower revenue, is called its
We remember from Chapter
II that
a demand curve is termed “elastic” over any
given range if the total outlay of the sale will be greater at a lower than at a higher
In the money economy, this means that a demand curve is
between a
range of two prices
if the amount of money spent at the lower price is greater than
the amount of money spent at the higher price
. In the case of the
to the individual producer
, the money outlay by the consumers constitutes his
gross money revenue at that price. Thus, in Case (b1), the gross revenue obtained
by the producer at a price of 10.5 and supply of 96 is 1008 ounces; at a price of
10.4 and supply of 100 units is 1040 ounces, etc. What we are concerned with in
this problem is the elasticity of the demand curve for the individual producer at
and above the point of maximum output. We compare the revenue at that point
with the revenue at possible lower outputs. In the case of (b1), the gross revenue
at the point of maximum output
– the
price of 10.0
is greater than
any revenue
that could be obtained from restricting Jones’ production to sell at a higher price.
Thus, Jones will sell at a point of maximum output when the demand curve for his
What of Case (b2)? Here, the demand curve for Jones’ product is inelastic,
if we compare the price of 10.4 and supply of 100 and the price of 11.0 and the
supply of 96.
Between these two points on the curve, the demand is inelastic, and
it would be more pro�table for Jones to restrict his production from 100 to 96 in
order to take advantage of the greater money revenue. However, this is irrelevant
for Jones’ action, because the demand curve is still elastic relative to the point
of maximum output
. The point of maximum output yields the point of maximum
Murray N. Rothbard
and hence with respect to this point, the demand curve for Jones’ product
is elastic throughout its range. The choice will still be Combination 3, the supply
of Jones will still be 110 units, and the market price will still be 10.0.
If Jones were in the situation of Case (a), the analysis would be even simpler.
It is obvious that if the price were 10 regardless of Jones’ product in the relevant
range, the demand curve for his product is completely elastic, and it would always
pay for him to be at his most productive, and produce the maximum physical out-
put with a given monetary investment on factors. In this case, too, the producer
strives for maximum physical productivity, and maximum output coincides with
Another conceivable case is Case (c), where the demand curve for the individ-
ual producer is inelastic at the point of maximum output. Suppose, for example,
Table 4.7

Gross revenue in the production of Good A, Case (c)
Combinations of inputResulting productPrice of product
Cross revenue
1)40X plus 84 Y
11.6 oz./unit
1114 oz.
2)50X plus 80Y
11.5 oz./unit1150 oz.
3)60X plus 76Y
110 units
10.0 oz./unit1100 oz.
Figure 4.2

With this sort of demand curve facing him, it pays the producer best to supply
to the market 100 units instead of the 110 units which he could supply. With a
Or, diagramming the data in Table
shows, in the form of the
Man, economy, & state
price of
per unit
instead of 10.0, the result is a larger gross revenue of 1150,
Jones can restrict his production in either of two ways, and it does not matter
which course he takes. He may use the less productive combination of factors,
Combination 2 instead of Combination 3, thus reducing his physical productiv-
ity; or, he may produce the maximum amount (Combination 3) and destroy the
difference (the 10 units). Economically, it doesn’t matter which course he takes,
since the result is to supply less for the market than he could have done with the
We see that when a demand curve confronting the individual producer is inelas-
tic as in Case (c), there are two major points of differentiation from the Cases (a,
b1, and b2), where this curve is elastic. First, in the other cases, physical produc-
tivity (output on a given investment in factors) is at a maximum, and all of this
output is supplied on the market. In Case (c), there is a restriction of productiv-
ity by the producer to obtain greater revenue. Secondly, the �nal market price is
always lower in the other cases, other things being equal. The effect of the action
in Case (c) is
always to raise the price to the buyer
. The effect of the restrictive
action is always to raise the price of the individual �rm’s product
Murray N. Rothbard
“monopolistically competitive” prices that could be set in between.
There is
Whichever price is set, whether competitive or monopoly price, the determina-
tion of the price takes place in the way we have analyzed above, via the supply
and demand schedules. The difference comes through the determination of the
quantity of stock produced. Under competitive price the producer estimates what
his selling price will be, or rather, what price he will be able to sell his stock for,
and produces the maximum stock that he can from his investment. But if the
demand curve to the producer is inelastic at that price, he can restrict his produc-
tion somewhat, produce less stock, and increase his monetary revenue. The mar-
The extra revenue which the producer obtains from the monopoly price as com-
pared to his revenue at the competitive price is a
monopoly gain
, and this concept,
along with further details of the monopoly question, will be studied further in a
It is most unfortunate that traditional terminology in economics makes it neces
sary to use such terms as “competitive price” and “monopoly price.” The terms are
highly misleading and can lead to serious errors in analysis, and they are highly
charged emotionally
– they
are “loaded terms” to most people. “Competition” is
usually regarded as �ne and praiseworthy, while “monopoly” is somehow sinister
and tyrannical. There was good reason for the sinister attachments to the word
“monopoly” in the public mind. The original meaning of monopoly was
a grant of
special privilege by the State
to a person or group of persons to produce a good to
the exclusion of other producers. As the great jurist Lord Coke de�ned monopoly:
A monopoly is an institution or allowance by the king, by his grant, commis-
sion, or otherwise
. to
any person or persons, bodies politic or corporate,
for the
buying, selling, making, working, or using of anything, whereby
any person or persons, bodies politic or corporate, are sought to be restrained
of any freedom or liberty that they had before, or hindered in their lawful
The original meaning of monopoly therefore was a grant or exclusive trade in
some area, conferred by the State to the hindering of the “lawful trade” of other
would-be traders, or “competitors,” in the same �eld. Such monopoly grants were
historically important in the Western world, and it is not surprising that, with
the growth of the spirit of liberty and of the libertarian movement, monopolies
Many present day writers have changed the original meaning of the word
“monopoly,” and the result is an unwarranted transference of this acquired hostil-
ity toward entirely different conditions. Some de�ne “monopoly” as any producer
who is alone in the production and sale of any particular product, or “monopo-
listic” as the exertion of any perceptible in�uence over the market price. These
conditions are far removed from privileged grants of monopoly. On such de�ni-
tions, any individual producer of a good that the consumers regard as unique, and
Man, economy, & state
from other goods, is a “monopolist.” Ford has a monopoly over the
sale of Ford cars; John Williams, lawyer, has a monopoly over the sale of the legal
services of John Williams, etc. In this interpretation, every seller of an individual-
Labels for concepts are basically immaterial, the main requirement being that
the original meaning continue in force to avoid confusion and error. In view of its
historic origins, and emotional connotations, such a use of the term “monopolist”
is highly inexpedient, and should be rejected. Similarly, to classify
for individual products as grants of monopoly is an illegitimate
use of the term. For the government to protect any individual in the use of his
own trademark is identical with protection against Jack Smith calling himself
“John Williams” and selling his own legal services in the guise of forgery. In other
words, it is equivalent to the governmental function of defending an individual’s
freely obtained property against violence and fraudulent theft.
Each individual,
in a free economy, has the right to his own self, to
, and to the exclu-
sive use of his own property. He is no more a “monopolist” over his own name,
than he is over his own will or his own property. The governmental function of
defense of person and property, so vital to the existence of a free economy and
a voluntary society, necessarily involves the defense of each person’s particu-
lar name or trademark against the fraud of forgery. It is absurd to use the term
“monopoly” or “monopolistic” with respect to the consumers’ differentiation of
various individual’s products and services. If the consumers consider Williams’
and Smith’s legal services as different in quality and therefore as different goods,
then they are different goods. To allow Smith to pass himself off as Williams,
because of the latter’s greater reputation for quality, is to permit violation of each
person’s ownership over his name and product.
To de�ne “monopolist” as the exclusive seller of any given product is thus
highly inexpedient. We shall employ the original de�nition of monopoly as a
grant of special privilege by the State, con�ning a �eld of trade of produce to
one individual or group, to the exclusion of others who would be eligible to enter
such production in a purely free economy.
We shall de�ne that voluntary society
where there are no grants of monopoly privilege as a society of
free competition
i.e., one where anyone may enter any �eld of production that he desired (so long
as he does not usurp the name of another individual). His
to do so in any
case depends of course on the capital he can invest or borrow, and on his entre-
preneurial ability in forecasting future conditions, but this of course is his
responsibility. He is
to compete, not only when he has the ability to do so, but
It should be clear by this time that there is a great distinction between the con-
cept of “monopoly” and of “monopoly price,” and hence the misfortune of the
same word applying to different concepts. The two are entirely different. The
monopolist, in our sense, may or may not be able to achieve a monopoly price.
The demand curve for his product may be elastic, or there may not be even any
consumer demand for his product at all, in which case he could make no net
return in producing the good. Thus, the State may grant Hiram Jones an exclusive
Murray N. Rothbard
privilege for the manufacture of kerosene lamps, but if so few people
wish to buy these lamps as to make the production unpro�table, the monopolist is
not able to achieve a monopoly price or a monopoly gain. On the other hand, the
production may be pro�table, but the demand curve elastic, so that the monopolist
does not restrict production and sells at what would have been the competitive
price. Similarly, the “monopolist” in the faulty sense of a single seller of any
product, may not be able to achieve a monopoly price for his sale. A
lawyer will
not be able to gain more revenue by restricting his hours of legal service
in order to raise the market price; a producer of a particular brand of breakfast
cereal may not be able to make gains by restricting his production in order to raise
Thus, it is perfectly possible for a “monopolist,” either in the sense of a privi-
leged seller or as the sole seller of an individualized commodity, not to be in the
position of charging a monopoly price for his product. The result depends on the
demand curve for his individual product. On the other hand, it is possible to be
able to charge a monopoly price without being a “monopolist” in either of the two
senses. Thus, let us suppose that there are several sellers of the same product, and
that therefore there is no monopoly. For each of the producers, the demand curve
for his individual product is elastic at the competitive price, and therefore there is
no way to achieve an extra monopoly gain by restricting production and raising
price. On the other hand, the demand curve for the product as a whole, the total
market demand curve, might be decidedly inelastic at the market price. In such a
case, there might well be a tendency for the various producers to get together and
decide production and price policy as if they were one �rm only. If they could
make such an agreement, they could act as one �rm, and the market demand curve
would then be identical with the demand curve for that “�rm,” and the inelasticity
would permit a general restriction of production and a rise to a monopoly price.
Such an agreement by many producers to act as one �rm in the market is known as
. A
arrangement can permit numerous �rms to act as “monopolists”
in the sense of sellers of an individualized commodity.
There are many stumbling blocks in the paths of �rms attempting to form such
a cartel, however. Although the demand for the whole product may be inelastic,
the demand for each �rm will be elastic. Therefore, each �rm will agree that the
total product and sale should be restricted in order to raise the price, but each
producer will be reluctant to restrict
product and sales. For if the other
�rms restrict their sales, each �rm can gain considerably by expanding his own
and taking advantage of the higher price. Hence, it is necessary for each cartel
member to agree on a certain quota of the aggregate product and sales, and restrict
himself to that quota. It is quite clear that the dif�culties to the establishment, and
the maintenance, of such a cartel are well-nigh insuperable. In the �rst place, there
is likely to be a great deal of bickering about the assignment of quotas since each
�rm will try to acquire a larger quota. Whichever basis quotas are assigned are
arbitrary, and will always be subject to challenge. As Professor Benham states:
Firms which have produced a relatively large share of output in the past will
demand the same share in the future. Firms which are expanding
– owing,
Man, economy, & state
example, to
an unusually ef�cient management
– will
demand a larger share
than they obtained in the past. Firms with a greater “capacity” for producing,
as measured by the size of their
. plant
will demand a correspondingly
Particularly likely to be restive under a cartel system are the more ef�cient
producers, those who are making larger pro�ts, and who are eager to expand their
business. These �rms will be eager to take advantage of the elastic demand curve
to their own sales, and to test their own mettle against the less ef�cient �rms pro-
tected by the assured cartel’s quota. It is obvious that the cartel, increasingly as it
persists, tends to protect the sales and earnings of the
as compared to
The successful maintenance of a combination, once it is formed, is threat
ened both from within and without. Conditions will change as time goes
on, and will make it dif�cult for the combination to retain the adherence
or “loyalty” of some of its members. Some �rms will �nd that consumers
demand more of their particular products than before and will resent having
to pass on orders (in excess of their quota) to be executed by other members
of the combination. Again, some �rms will outstrip others in taking advan
tage of the progress of technical knowledge, and will conclude that they
have more to gain by expanding their sales at lower prices than by continu
ing their membership of the combination. If the demand for the products
of the industry falls considerably, the proportion of “unused capacity” will
increase, and this will strengthen the desire of some �rms to break away
and make fuller use of their plants, thus increasing their receipts, by selling
The ever present temptation to each producer, particularly a venturesome and
ef�cient one, is to defy the cartel, either secretly or openly, and expand his own
sales. The great instability of the cartel stems from the fact that once the �rm steps
out of line, the others must do so as well. For with A, B, C, etc. restricting their
output to maintain the monopoly price, if competitor D expands his output, and
cuts the price slightly, he tends to take a great deal of business away from the other
producers. Even if price is not affected a great deal, D’s expansion earns revenues
while the others must limit theirs.
The result is a speedy breakup of the cartel and
Just as great a menace to the existence of a cartel is the threat of outside compe-
tition from newcomers. As a matter of fact, the greater the success of the cartel in
maintaining its internal cohesion, and earning monopoly gains which are appor-
tioned to the members, the greater will be the temptation for new �rms to enter
the �eld. These new �rms, unhampered by cartel agreements, can expand their
production and sales to take business away from the cartel, and may cut the price
of the product as well. This factor is a powerful one in causing the dissolution of
the cartel agreements. As a result of these factors, it is not an exaggeration to state
Murray N. Rothbard
almost no cartel agreement, unaided by special privileges from governments,
has been able to survive more than a very short period of time.
The type of State
privilege is varied, and will be dealt with in the chapters on State intervention and
the Hampered Market.
One such measure is compulsory cartelization, another
is the imposition of arti�cial restrictions in the freedom of entry of potential com-
Another important factor tending to prevent the rise of cartels is that, in a free
economy, an agreement to form a cartel
is not enforceable
in the courts. In other
words, if Jones signs an agreement to join a cartel and only process 10% of the
output of certain other �rms, he may violate the agreement at any time without
suffering governmental penalties, such as payment of damages of compulsion
to abide by the contract. This is due to the particular scope which governmental
enforcement of contracts has in a free economy. It was seen in Chapter
II that
governmental agency, in a voluntary society, enforces contracts, not simply
because they are contracts or promises
per se
, but because they represent
ished exchanges of property
Suppose, however, that a monopoly price has been established on the free mar-
ket, either by an individual �rm or by a remarkably stable cartel. Are the conse-
quences necessarily sinister, as has often been assumed? In the �rst place, it must
be realized again that the term “monopoly price,” used in contrast to “competitive
price,” is really a misnomer, although the terms must be used for traditional rea-
sons. The monopoly seller or sellers are not immune from, or beyond the pale, of
competition. Quite the contrary. The terminology is the result of an old neoclassi-
cal preoccupation with single “industries.” Every monopoly seller competes with
every other seller for the money of the consumer. Every consumer allocates his
money expenditure among all the available uses, and therefore this fundamental
competition obtains between all sellers of all the goods and services. Producers
compete for wide groups of laborers of various types, of lands and capital goods.
Thus, Ford does not only compete with General Motors; it competes with the sell-
ers of washing machines, of television sets, of houses, of caviar, of concert music,
etc. Everyone on the free market is a mutual competitor. Thus, the monopoly
seller who obtains a monopoly price is not beyond competition. He does not dic-
But even if the monopoly seller is subject to competition, isn’t the consumer
worse off when a monopoly price and restricted production obtains? Can we not
say that there is a loss of consumer welfare in a monopoly price situation? Isn’t
this an important exception of the harmony of interests that prevails on the vol-
untary market? To answer these questions, let us recall the exchange situations
detailed in Chapter
Jackson and Smith are in isolated exchange, the former
has a horse and the latter has �sh, and they bargain to make an exchange. Let’s
say the agreed upon terms of exchange are 90 barrels of �sh for the horse. Now,
critics could charge that Jackson is worse off than he would have been if the price
had been set at 95 or higher, while Smith is worse off than he would have been if
the agreed price were less than 90. Such charges, however, miss
the point
of the
analysis. The point is that both voluntarily agreed on the price, that both believed
Man, economy, & state
that there
were no better alternatives available. The same is true for every price
in every exchange, regardless of the number of exchanges. The purchase or the
sale of the unit of the good at the agreed upon price is considered the best possible
alternative action by each party. Thus each is the best off, has the highest welfare,
that he can obtain, consistent with the maximum welfare of everyone else. Smith
could force Jackson at the point of a weapon to make the exchange for 80 or 70 or
60 or no �sh at all. But in that case, it is obvious that the use of coercion has made
Jackson worse off, and that Jackson is being exploited by Smith. Furthermore,
this action brings up all the problems of violence and an exploitative society,
which have been mentioned previously and will be discussed fully in later parts
of this book.
Within the framework of a voluntary society, the market price is the
best price that either the seller or the buyer can get, and therefore comparing the
welfare of either one with some impossible ideal is vain. In the same way, buyers
and sellers on the market are “included” or “excluded” from exchange by their
But what of the case of a monopoly price? When it is set in the framework of
the free market, again all parties to the exchange bene�t. A
coerced lower
price or
greater product could only exploit the sellers for the immediate bene�t of the buy-
ers. Monopoly pricing, on the other hand, is not the exploitation of the consum-
ers, because the payment is voluntary. This conclusion is con�rmed by a closer
look at the inelastic demand curve, which must obtain in all cases of monopoly
price. Thus, suppose that a �rm’s maximum productivity would yield a product
of 100 units at the competitive price of 10 ounces. Its inelastic demand curve is
such that a stock of 50 units raises the market price to 30 ounces, the monopoly
price. In the former case, the �rm’s revenue is 1000 ounces from its investment;
in the latter case, it is 1500 ounces. This means that consumers have
paid more money for the product in the monopoly price situation. How can it
be deduced from this that the consumers are worse off under a monopoly price?
After all, the inelasticity of the demand curve is not �xed in Heaven; it is the result
of the voluntary action of the consumers in paying more money for the product
at a monopoly price. If the consumers really felt that they were worse off than
they could be because of the monopoly price, they could, individually or jointly,
the product and refuse to buy at the higher price. Such action, would, of
course, render the demand curve for the good elastic, and force the �rm or the
cartel to increase its output and lower the price to the competitive one. The money
withheld in the boycott could be added to cash balances, spent on the products of
competitors, or used to invest in a competitor to a cartel. There is therefore never
any need to worry about the situation of the consumers in a free market. The shape
of their demand curve, and therefore the �nal market price, is purely the result of
It should be clear from the above discussion that there is nothing particularly
reprehensible, or frustrating of consumer freedom, in the establishment of a
“monopoly price” or in a cartel action. A
cartel action,
if it is a voluntary one,
cannot injure freedom of competition or, if is pro�table, cannot injure consum-
ers. On the contrary, they are, as are all other actions on the free market, perfectly
Murray N. Rothbard
with a free society, with individual self-sovereignty, and the earning of
As Benjamin R. Tucker brilliantly concluded in dealing with the problem of
That the right to cooperate is as unquestionable as the right to compete; the
right to compete involves the right to refrain from competition; cooperation
is often a method of competition, and competition is always, in the larger
view, a method of cooperation
. each
is a legitimate, orderly, non-invasive
Viewed in the light of these irrefutable propositions, the trust, then, like
every other industrial combination endeavoring to do collectively nothing
but what each member of the combination might fully endeavor to do indi-
vidually, is,
per se
, an unimpeachable institution. To assail or control or deny
this form of cooperation on the ground that it is itself a denial of competition
is an absurdity. It is an absurdity, because it proves too much.
The trust is a
denial of competition in no other sense than that in which competition itself
is a denial of competition
. (Italics ours.) The trust denies competition only
by producing and selling more cheaply than those outside of the trust can
produce and sell; but in that sense every successful individual competitor
also denies competition.
. The
fact is that there is one denial of competition
which is the right of all, and that there is another denial of competition which
is the right of none. All of us, whether out of a trust or in it, have a right to
deny competition by competing, but none of us, whether in a trust or out of
it, have a right to deny competition by arbitrary decree, by interference with
voluntary effort, by forcible suppression of initiative.
This is not to say, of course, that joint co-operation or combination is necessar-
ily “better than” competition among �rms. We simply conclude that the relative
extent of areas within or between �rms on the free market will be precisely that
proportion most conducive to the well-being of consumers and producers alike.
This is the same as saying that the size of a �rm will tend to be established at the
3: the product and outlay schedules of the �rm
Let us now return to the activity of the �rm and its
production function
. We will
assume now that the �rm is competitive, and produces for a competitive price,
so that its situation either �ts Cases (a) or (b) above. In the production schedule
drawn up for Jones shown in Table
the ratios between the quantities of the
factors differ for the various technical alternatives available. Thus, 50X combined
with 80Y produces 100 units of product, and 60X combined with 76Y produces
110 units. The ratios between the quantities of factors: 50/80, 60/76, etc. may vary
considerably. The list of technological alternatives varies according to the speci�c
“engineering” data of the product in question. In very rare cases, there might be
Man, economy, & state
cases where
only one ratio, or one set of “production coef�cients,” is permissible.
In such cases, for example, the product could only be produced with a combina-
tion of 5X to 8Y, in that ratio. In almost all cases, however, it is possible to vary
the ratios of the factors. Thus, some might assume that the factor ratios in a �rm
producing, say, chemical dyes are inalterably �xed by the chemical formula of the
dyes. This is a complete misconception of the problem, however. The point is that
the variations can take place among the number of workers, the number of vats,
the amount of land, management, etc., that will be used. The greater the develop-
ment of the economy, the advance of technological knowledge, and the amount
and variety of factors, the greater the opportunity for variability of factor ratios. It
is doubtful, indeed, if there are any instances of production where the factor ratios
In Jones’ case, given the factor prices, and the production functions available,
it is clear that he will choose the combination 60X plus 76Y in order to attain the
maximum output, and hence maximum revenue, from the original investment. In
order to analyze more fully the problem of production combinations, the �rms’
production, and factor prices, we will assume a far greater range of production
alternatives by extending Table
Suppose, for example, that with the price of
Factor X at 4 ounces
per unit,
and the price of Factor Y at 10 ounces
per unit,
ounces will purchase the following alternative combinations of factors yielding
Table 4.8

alternative combinationsUnits of product
These are the technological alternatives that can be accomplished with 1000
ounces’ worth of factors. The maximum productivity is still at 60X plus 76Y, and
Now, simply from the given factor prices, we can deduce the
, i.e., the rate at which one factor must be subtracted to compensate
for the addition of another factor, so as to have a
(in this case,
1000 ounces). In the present case, 2 fewer units of Y have to be compensated by
5 additional units of X in order to arrive at the “constant outlay combination” of
Murray N. Rothbard
ounces. For example, starting from the �rst line, we know that 40 times 4
equals 160; 84 times 100 equals 840, and the sum equals 1000. If we add 5 units
of X and subtract 2 units of Y to move to the second line, we know that 45 times
4 equals 180, 82 units of Y times 10 will give 820, to sum to 1000. It will be seen
below algebraically below that the
rate of outlay substitution of one factor for
another is equal to the ratio of the prices of the two factors
. Therefore, the rate
of substitution of factor X for factor Y is 2/5, while the ratio of the money price
X to
the money price of Y is 4/10, or 2/5. This ratio of 2/5 obtains regardless of
As yet, we have not progressed far beyond the conclusion that Jones will pro-
duce at the (60X; 76Y) combination. However, this line of approach permits fur-
ther insight into the activity of the �rm, and the interplay of technological and
�nancial factors. Let us now shift the focus of attention and consider this type of
question: assuming for the moment that Jones wishes to produce, say, 105 units;
what are the alternative combinations of factors which can produce them? The
answer is a purely technological one, and in accordance with the technological
knowledge available, Jones can draw up a list of alternative physical combina-
tions that would yield this result.
, in this sort of problem, no �nancial or
monetary considerations have yet entered. We already know that 105 units can
be produced by the combinations: (55X and 78Y) and (70X and 72Y). Let us say
that the following, shown in Table
are the combinations of the two factors that
Table 4.9

Combinations of producing 105 units of product
It is obvious that in investigating any
constant product
combinations, an addi-
tion in the amount of one factor must be offset by a decrease in the quantity of the
other for the �nal product to be the same.
This can be deduced from the mere
fact of these factors as instruments of production. It is
deducible from the
very fact of the existence of factors. As more and more of one factor is added,
and another factor is diminished, the added quantities must compensate less and
less for losses in the other factor. Conversely, the more a factor is diminished,
the greater will be the need to compensate by adding to another factor, to pro-
duce the same product. This is called the
imperfect substitutability of factors
. This
Man, economy, & state
imperfect substitutability
is deducible from the very existence of human action.
The very fact that consumer goods are scarce implies that factors of production
are scarce, and the very fact that there
factors implies that there is more than
one factor, since if there were only one factor it would be a consumer good and
not a producers’ good. The very fact that there is more than one factor, in turn,
implies that the different factors are
perfectly substitutable for each other;
otherwise, they would not be separate factors at all. The common example of such
imperfect substitutability is that if labor were perfectly substitutable for land on
a farm, constant production could be insured with a constantly diminishing area
simply by adding to the number of workers, so that 100,000 workers in the space
of a thimbleful of land could produce as much wheat as 100 workers on a hundred
acres of land. The imperfect substitutability, however, applies to
We may de�ne the
marginal rate of production substitution
of one factor for
another as the ratio of the amount of the second factor that can be diminished as a
result of an increase in the �rst factor in order to yield a constant product. It is clear
that the marginal rate is
as the factor continues to be added. When
the combinations change from (40X; 100Y) to (45X; 90Y), the marginal rate of
substitution of X for Y is 10/5, equal to 2; but later on in the proceedings, when the
combination changes from (65X; 73Y) to (70X; 72Y) the marginal rate of substi
tution is 1/5. What the actual rates are depend on the speci�c technological data,
but economics does tell us that the
marginal rates of product substitution diminish
Suppose that Jones decided to produce 105 units of product; he could affect the
production in each of the above different ways. Which alternative would he choose?
Obviously, he could choose the alternative that involved the least expense in money,
and that would depend on the prices of the factors.
, he would have
no way to choose between the various combinations, because technologically all of
them are equally effective. It is only the existence of factor money prices that permits
the producer to choose among these combinations. With the original factor prices of
4 ounces of gold
per unit
for X, and 10 ounces for Y, the necessary money expenses
Table 4.10

Combinations producing 105 units of product
Money outlay necessary
Murray N. Rothbard
this particular example, Jones will choose either (60X; 75Y) or (65X; 73Y)
either of which minimizes his required money outlay at 990 ounces. Given the
amount of production at 105 units, the
Some writers discuss the activity of the �rm as if this were the most appropriate
manner of analysis, as if a quantity of product is arbitrarily set, and the producer
looks for the minimum outlay combination of factors to produce it. In reality,
however, it is clear that the beginning point is the decision to invest a certain
amount of money in factors, and the attempt to choose a combination so as to
maximize the productivity of the factors, as we have seen above. The present
analysis is subsidiary and supplementary to the previous one, but it is useful to
Reverting back to the 1000 ounces’ worth of combinations depicted in Table
saw that Jones chose that combination which maximized production for 1000
ounces, at 110 units of product (60X and 76Y). We shall now demonstrate that this
combination is also the minimum outlay combination of all the factor combina-
tions that could produce 110 units of product. The demonstration of this truth is
simple. In the �rst place, we may rule out those combinations which require less
factor, such as (55X; 74Y). We have seen above that obviously wasteful
combinations are discarded immediately; therefore, if (60X; 76Y) are
to produce 110 units, there could not be another constant product combination
with less of each factor that could also produce 110 units. This follows from the
very nature of scarce goods and scarce factors. Therefore, the possible combina-
tion which might be able to produce 110 units for less outlay would have to be a
constant product combination schedule such as listed above in Table
for 105
units, with more of one factor compensating for the subtraction of another. Now
suppose that this supposed minimum outlay combination for 110 units has a quan-
tity of X of more than 60, and a quantity of X of less than 76. But to be cheaper,
the combination would have to have less of one factor
– given
the other
– than
combination on the 1000-ounce constant outlay schedule. But for each addition of
X (X is assumed for convenience to only change in blocks of 5 units, but this does
not alter the fundamental result), the constant outlay combination produces
units of product: 107, 105, 100, etc. In order to be cheaper for any given X, the
units of Y would have to be even fewer; and it is manifestly impossible for such
a combination to produce as much as these amounts, let alone 110 units. Sym-
metrically, the same is true for combinations with less X and more Y. For constant
outlay, each of the possible alternative combinations produces less than 110 units;
to be cheaper than each of these, any other combination could only produce still
less, and could not produce 110 units.
It is therefore universally true that the maximum product combination for any
given outlay of money is also the minimum outlay combination for that particular
Thus, we see that, on the free market, each �rm, in maximizing the product
that can be produced from any given outlay, is also engaged in reducing the
money outlay required for each product. Given the prices of the factors, there is
only one way to increase his money income from the investment: to �nd a factor
Man, economy, & state
combination that
will be the most productive of physical product, and that, in
consequence, will be the cheapest method of producing that amount. This analysis
enables us to see clearly the different roles played in production by technological
and by economic considerations. Technological considerations yield knowledge
of the various series of constant product schedules that would be available. At any
given product that could possibly be considered, the prospective producer could
command a series of tabulations that would yield him the production functions
and combinations that could produce it. This would be the contribution of tech-
nology. But this knowledge by itself would tell the entrepreneur next to nothing
about the crucial questions in the whole problem of producers’ activity: should
he enter the business at all? How much should he invest? Which of the alterna-
tive constant product combinations should he choose? The answers to these vital
questions can only be provided by
, by
, as opposed to
, considerations. Speci�cally, it is the establishment of money on the
market which enables the businessman to make these decisions in a rational and
intelligible manner. The prospective producer will invest in that line of business,
in that particular �rm, which will maximize his expected money income, over any
period of time that he chooses. This rule, as we have explained before, is modi�ed
when psychic nonmonetary matters intervene, thus obeying the general, universal
rule that in
action the actor maximizes his expected psychic income. Setting
aside cases of con�ict between money and psychic income, which have already
been noted, investors drive to maximize their money income. They will enter that
line of business which promises the greatest return on their investment, they will
invest in accordance with their expected return balanced by their time prefer-
ence, and they will produce that combination which requires the least monetary
expenditure for the particular product. And to accomplish this they will sell their
products for as much as they can
– which
we have seen will quickly tend to be
the competitive market price; will try to buy their factors for as little as they can
which we
will see below will be the competitive price; and will try to increase
the physical productivity which can be obtained from any given set of factors,
i.e., increase their
productive ef�ciency
to the utmost. But it is clear that none of
these decisions could be made if the investor did not have the various price data
and estimates to guide him in his choices. And it is only because the money com-
modity has become the general medium of exchange that such markets, and such
price and income comparisons and estimates, are possible.
And these price and
income calculations and estimates are most emphatically
estimates; they
We have already demonstrated that the maximum product combination for any
given outlay of money is also the minimum outlay combination for that particular
physical product. It is therefore also true that every minimum outlay combination
is the maximum product for that outlay. Let us then take the case of an inves-
tor with 990 ounces of gold to invest. His maximum product combination will
produce 105 units, at either the combination (65X; 73Y) or (70X; 72Y), which
are also the minimum outlay combinations for 105 units. We may see above
the behavior of the rate of product substitution as the number of units of factors
change; the rate of product substitution of X for Y changes from 2, to 6/5, to 3/5,
Murray N. Rothbard
We notice that the minimum outlay combination is reached at the approximate
point where the rate of product substitution is equal to 2/5; i.e. is equal to the rate
of outlay substitution, which, given the prices, is constant throughout at 2/5. If the
rate of product substitution is appreciably less than or more than the rate of outlay
substitution, it will pay for the producer to shift to other alternatives until the two
Thus, there is a tendency for the �rm to produce at such a rate and such a way
that the
rate of product substitution between factors is equal to the rate of outlay
per outlay
(which will be the minimum outlay for that
product). On the contrary, the two ratios will not by any means always be equal,
because the range of production alternatives available may not be suf�cient. If
there are only a few production alternatives, then there cannot be the small steps
which are necessary to allow equality of rates, or meaningful discussion of such
rates. Thus, if only two combinations can produce 105 units of product: namely,
(45X; 90Y), and (65X; 73Y), Jones will choose the minimum outlay combina-
tion, but the “rate of product substitution” between such distant combinations
will be 17/20. However, the rate will still be the nearest approach possible to 2/5,
and in that sense, we may still say that the tendency will be to approach that rate.
The value of the concepts of rate of substitution will fully emerge as essential to
an analysis of the prices of factors of production, and, speci�cally, the demand
At this point, it is now time to turn to an algebraic and geometric presentation of
The de�nition of a constant outlay schedule is that the total sum of money
expended be constant, whatever that sum may be. In other words, for two factors,
the sum of the amount of money spent on factor X plus the sum of the amount
spent on factor Y is always equal. The amount of money spent on each factor, in
turn, is always equal to the price of that factor times the total quantity of the factor
that is purchased. Thus, if the price is 10 ounces
per unit,
and 5 units are bought,
the total sum of money expended is 50 ounces. Therefore, for a constant outlay
schedule, if
is the money-price of factor X;
is the money-price of factor Y,
is the number of units of X bought at any given point;
is the number of units
of Y bought at any given point; and
Man, economy, & state
This equation
de�nes any given point on any constant outlay curve for two fac-
tors. Now, suppose that we wish to move from this point to any other point on the
constant outlay curve. The amount of X then becomes
, while the amount of
Y, which diminishes in compensation, becomes
. At this point then:
we may multiply out in equation
, and substitute from equation
This gives
us proof of the statement in the text that the
rate of outlay substitu
tion between two factors is equal to the ratio of the prices of the factors
. As X
increases, the ratio of the decline in Y due to the increase in X needed to maintain
Returning to equation
, let us solve for
, the quantity of Y at any given
let us solve equation
for those points where
is equal to zero, i.e.,
there are zero quantities of X. Then:
Now, we may substitute (
we can see that equation
is directly applicable to the case of Jones’
1000 ounces.
refers to the values of Y at each point, and therefore may be writ-
ten as Y. Similarly,
refers to the values of X and can be written as X. The ratio of
is equal to 4/10 or 2/5.
is the value of Y when X is zero; it is equal to the
constant outlay (1000) divided by the price of Y (10)
for Jones’ condition of 1000 ounces and the given prices of the factors:
Y = 100 – (2/5)X
Murray N. Rothbard
All constant outlay curves for two factors have the shape of a straight line. The
slope of the line is negative, and is the ratio of the prices of the two factors, which
is also equal to the rate of outlay substitution between them. When X is zero (even
though such a choice will never arise in practice), Y is equal to the constant outlay
sum divided by the price of X; and when Y is zero, it is easily seen that the value
of X is the constant outlay divided by the price of Y.
This algebraic analysis enables us to establish a whole series of constant outlay
curves for different values of k for different constant outlays. Whatever the con-
stant outlay, the curve can be determined: it again will be of the same slope as the
other curves, while the difference will be in its position. Thus, say the constant
outlay is 800 ounces of gold. In this equation, when X is zero, Y will be equal to
800/10, or 80. When Y is zero, X will be equal to 800/4, or 200. And the constant
In this way, we can establish a whole
of constant outlay curves. All that
is needed is the knowledge of the prices of the two factors, which are assumed
to be given; and then for each possible constant outlay, the combinations of the
factors can be determined. Some of the members of the family of constant outlay
curves in Jones’ case are as in Figure
k = 1000
Figure 4.3

Now, it is important to realize that the prices of the factors are the sole determi-
nants of the family of constant outlay curves. These prices are always approach-
ing uniformity on the market. Therefore, the constant outlay curves are not only
applicable to Jones;
the very same ones are applicable to all producers who use
. Thus, the given set of constant outlay curves and the given rates
Man, economy, & state
of outlay
substitution are the same for all the �rms producing with these factors,
not just for one �rm alone. At any one time, then, the family of constant outlay
curves for any two factors is the same for all producers on the market. This fam-
ily of constant outlay curves is a series of regular, similarly sloped lines, easily
production function
a given data to all producers. The
production function is the estimate of the maximum quantity that could be produced
from each combination of factors. Although this is technological rather than catallac
tic knowledge, it by no means follows that it is “given” to all prospective producers.
This knowledge is not simply of engineering formulae; it involves numerous minute
details of individual skills, correctness of estimates, judgment of materials and loca
tion, etc.
It is far more likely that each individual’s production function differs than
that it is the same, even with the same product and the same factors. As we will see
below, this likelihood is made a certainty when there are many more than two fac
tors of production, and when, as is almost always the case, some of these factors are
unique (
), in some ways to the individual �rm. Production functions, there
fore, are irregular and differ from one producer to another. Furthermore, they are not
“objectively” given; they are only estimates in men’s minds.
What is the shape of the production function? Some might be of �xed propor-
tions, i.e., only one combination of factors can produce each possible quantity
of output. We have seen in the text that this is practically never the case, but if it
were, a diagram would be as follows: the quantity of one factor on the horizontal
Figure 4.4

Murray N. Rothbard
this combination (1X; 10Y), there is very little of X and a great deal of Y.
Now suppose that X is increased to 2; what will be the loss in Y to compensate
and maintain production at 10 units? We cannot know the answer except for the
concrete case, but it is clear that since the two factors are imperfect substitutes for
each other by their very nature, where the quantity of X is low a slight addition
of it will compensate for a big loss in Y to maintain constant production. Let us
say that the constant production combination is (2X; 6Y). In the diagram, we may
connect the two points for the sake of convenience. Now, what if X is increased to
3 units? Since X has been increased and Y has diminished, it will now take a lesser
loss of Y to compensate for an increase of X. Thus, the point (3X; 4Y) might be on
the constant product curve. Between the �rst and second points, the loss of Y was
As Figure
shows, the numbers designate the quantity of output yielded at
the various points. These quantities can be of any amount, but they must increase
as the quantities of X and Y increase, by the nature of production.
With the existence of varying proportions of factors, so that there are alterna-
tive factor combinations for each quantity of product, we can draw up
product schedules
, and therefore
constant product curves
. If we assume that there
are many possible combinations for each possible product, then we may ask the
question: suppose, for example, that 1 unit of X and 10 units of Y combine to
produce 10 units of product, as in Figure
Figure 4.5

Man, economy, & state
4 and
the gain of X was 1 unit; the ratio of the two is 4/1, or 4. From the second to
the third point, Y lost 2 and X gained 1; the ratio was 2. This ratio is the
rate of product substitution
between the factors, or the rate of substitution of X
for Y. It is evident that as X increases, this rate
. As X increases and
Y diminishes, more and more gain of X is needed to substitute for less and less
loss of Y. Thus, the succeeding points on the constant product curve above may
be (4X; 3Y), (7X; 1.5Y), with marginal rates of substitution at those points 1 and
0.5 respectively.
We have arrived at one constant product curve. At each constant product, it is
evident that there will be a similar shape, in that the marginal rate of substitution
diminishes throughout. However, it is obvious from the nature of production that
the larger product calls forth a larger quantity of both factors at each point. Thus,
suppose that we are interested in a constant product curve at 20 units. Suppose
X is 1 unit; it is obvious that Y will have to be more than 10 in order to produce
these 20 units. What amount this will be we do not know; we only know it will be
greater. Let us suppose that the point will be (1X; 15Y). We can now draw in a set
of succeeding points, assuming only a diminishing marginal rate of substitution.
It is clear that all these points will be above, or to the right of, the corresponding
Thus, we see that there is a
of curves for each constant product. The
higher products are above (to the right of) the lower ones. The property of dimin-
ishing rates of marginal substitution make these curves tend to be convex to the
origin. As the product gets lower and lower, the curves get closer to the origin,
�nally reaching that point itself at zero product, since zero quantities of factors
yields zero product. On the other hand, the curves never cross the X or Y axes.
Since both factors are assumed to be necessary ones for the production of the
product, and hence the imperfect substitutability of the factors, no increase in the
one factor, however great, can compensate for the loss of the whole supply of
the other. A
common classical
example is the case of a wheat farm, where no
amount of labor, however great, can produce wheat when there is no land avail-
able; on the other hand, no amount of acreage can produce wheat without any
labor. The point applies, however, to all types of production.
The point has come when this information can be consolidated. For any process
of production using two factors, there are two families of curves: constant outlay
curves, and constant product curves. Constant outlay curves hold for all producers
who use the two factors, since they depend solely on the market prices of the fac-
tors. Constant product curves are estimates by the enterprising producers, and will
differ from �rm to �rm. While the former are regular straight lines determined by
the ratio of prices and total outlay in view, the latter are irregularly spaced, their
only condition being the diminishing rate of substitution between the factors. The
two families of curves will be somewhat as in Figure
we have seen in the text, at any given outlay, the actor will produce at the
maximum product. What does this mean in graphic terms? Let us take, as in Fig-
Figure 4.6

Figure 4.7

Man, economy, & state
This diagram
has seven constant product curves, marked 1 to 7, in ascending
order of the size of the product. As the constant outlay curve begins at the top,
it intersects constant product curve 1 at point A. At point A, that combination
of factors X and Y yield a total product of order 1. Proceeding further along the
constant outlay line, (further in the sense of increasing X and decreasing Y), we
intersect point B, at which point X the factors will produce products of size 2. So
as we proceed along the constant outlay line, we arrive at higher and higher prod-
– at
curves further and further to the right. Finally, we arrive at the point with
the highest size product, and the point of production that will be chosen with this
outlay. This is point E of size 5, the point of tangency between the constant outlay
line and the highest constant product curve obtainable with that outlay. Beyond
this point, the constant outlay line again intersects the lower-sized product curves.
For any constant outlay line then, the entrepreneur will strive to act so that his
combination of factors will be at a point tangent to the constant product curve. Of
course, the entrepreneur in practice does not need to know about such tangencies
and curves; he is only concerned with maximizing his output for the given outlay.
But we have seen that mathematically this is implied by such maximum output.
It must be cautioned that in practice, the constant production curves are a series
of dots, of discrete points, rather than continuous lines. A
curved line
implies that the distance between the points of decision by the actor are in�nitely
small; actually, this can never be the case
– human
action of necessity deals with
discrete objects and distances. However, in the realistic case, the choice of the
maximum product is the closest approximation to such tangency that could be, or
It is clear that this elaborate analysis of families of curves and tangencies is of
no particular aid in this problem; however, it provides analytic tools that will be
handy in later analyses of the pricing of factors of production.
For one thing, we
know geometrically that the marginal rate of product substitution, which is always
diminishing, is equal to the
of the constant product curve, when the latter
is a continuous curve. At a point such as E, of tangency with the constant outlay
line, elementary geometry tells us that the slopes of the curve and the line are
equal. The slope of the line equals the marginal rate of outlay substitution, which
is constant throughout and equal to the ratio of the factor prices, and therefore,
the point of tangency, the marginal rate of outlay substitution equals the marginal
rate of production substitution
. Under real conditions, this is only an approxima-
tion rather than an actual fact, but this proves the assertion in the text that the
producer sets his production so that these two marginal rates tend to be equal.
And this means, furthermore that, for each producer’s decision,
the marginal rate
of product substitution between the two factors tends to equal the ratio of their
This equality is only an approximation, since for the universal case of more or
less discrete points, the point of decision will only be the nearest approach to such
equality. However, because of the divisibility of money, the constant outlay curve
tends to be (although never will be) a
line, while the more advanced
Murray N. Rothbard
this �gure, we depict constant product curves, P1, P2,
. P7,
and constant
outlay lines, O1, O2,
. O7.
They have points of tangency at A, B, C, D, E, F, and
G. The zero point is also a point of tangency, at zero input of factors. The points
of tangency enable the producer to determine his
maximum product outlay curve
For at any given outlay, the tangency points will yield the size of the maximum
constant product curve. Thus, O1 will be tangent to P1 at point A. The same is
true to every other alternative. Thus, the decision points A, B, C, etc. reveal to the
producer: 1) the maximum product for each outlay, and 2) the best combination
the production structure and the more complex the alternative combinations, the
nearer will the constant production schedules approach being continuous curves.
The more highly developed the market economy, therefore, the greater will be the
tendency to approach equality between the ratio of the prices of factors and the
marginal rates of product substitution between them.
At each possible constant outlay line, therefore, the producer will pick his pre-
ferred combination of factors at the point of maximum output, or approximate
tangency to a constant product curve. The higher the amount of money to be
spent, and therefore the higher the constant outlay line, the higher and the further
to the right will be the constant product curve, and the various points of tangency.
Thus, a typical family of constant product and outlay curves may have points of
Figure 4.8

Man, economy, & state
shows that, at an outlay of 1000 ounces of money, different alterna-
tive combinations could yield various amounts of product, namely 110, 107, 105,
100, 97, and 96, as listed in Table
above. The highest production, or the top
dot on the line, will be the one that is chosen, and the combination of factors will
be picked accordingly. This dot is crossed to represent the product of the combina-
tion that will be chosen. The same sort of process will be undertaken regardless of
the amount that the producer has to invest. Thus, if he has 990 ounces to invest,
he will choose the combination yielding him the maximum product, at 105 units.
At each possible investment of money outlay, the producer will choose that factor
combination which yields him the maximum product. Thus, the diagram of such
For each straight line, the top crossed dot will be selected. Thus, we see a series
of possible vertical straight lines, representing the constant outlay, with units of
product on the vertical axis and money outlay on the horizontal axis. Each vertical
4: the output and investment decision of the producer
We must now return to Jones and his outlay of 1000 ounces. We have already
seen that, given an investment of 1000 ounces, Jones will select one combination
which will yield him a maximum product. Out of a group of alternative combi-
nations, he will select the best combination. We could diagram this situation as
Money Outlay
Figure 4.9

Murray N. Rothbard
line is a constant outlay line, and the crossed top dot is the maximum
product that would be selected in each case. As Figure
shows, the crossed
dots can be joined for convenience to give us a connected line of potential prod-
Money Outlay
Figure 4.11

9001000Money Outlay
Figure 4.10

Man, economy, & state
notice that we may conveniently omit the crossed dots from the �nal con-
nected line. From the line, we may read off the maximum product which would
be yielded by the expenditure of any given outlay.
Without discussing at this moment when the curve is likely to be horizontal, it
is obvious that no producer knowing the situation will pick any outlay along the
Each producer will try to determine the various points on this product outlay
curve. As we have seen, he estimates the various alternative factor combinations
for producing each particular quantity of product, and, using these and the prices
of the factors, the producer will be able to judge his constant outlay combinations,
and which combination will yield him the maximum product for each outlay. This
will give him the series of crossed top dots for each outlay, and yield him the
maximum product schedule
for each outlay.
What can economics say about the shape of this important curve? In the �rst
place, it is obvious that a
greater outlay can never produce a lower maximum
. We have seen above that the 1000 ounces will yield a maximum product
of 110 units. A
greater outlay
, say 1050 ounces, cannot produce a maximum prod-
uct of
than 110 units. This is obvious from the very nature of production and
of factors. At the very least, the 110 units could be produced, even if the excess
factors purchased with the other 50 ounces cannot be used. Thus, the maximum
product schedule
slopes upward or remains horizontal when the money
Another characteristic of the maximum product outlay curve is an obvious one:
it must pass through the zero point, since no expenditures will obviously result in
no production. A
Money Outlay
Figure 4.12

Total product outlay curve
Murray N. Rothbard
except the cheapest: i.e., the point on the extreme left of each horizon-
tal line. Thus, if 1000 ounces of outlay will produce 110 units maximum and 1050
ounces of outlay will also produce 110 units maximum, it is clear that there will
be no hesitation in choosing the 1000 ounces, and not the more expensive outlays.
Any other decision would be a pure waste of money by the producer. Therefore,
without yet fully answering how much money the producer will decide to invest,
we can immediately answer that he will never decide to invest that amount which
lies along a horizontal line. Thus, if 1000 ounces will produce 110 units, and all
greater expenditures up to 1100 ounces will only produce 110 units (with expen-
ditures of over 1100 ounces yielding more units), we can be sure that Jones will
not decide to invest a sum of between 1001 and 1100 ounces. He will either invest
more or less. In Figure
above, we cross the horizontal lines with vertical
’s decision.
So far, from Figure
we know two de�nite points on Jones’ maximum prod-
uct outlay curve: 1000 ounces netting him 110 units of product and therefore 1100
ounces of money revenue; 990 ounces netting him 105 units of product and there-
fore 1050 units of revenue (selling prices are assumed to be 10 ounces
per unit).
the former case, he makes a net money income of 100 ounces, equaling 10%
of his outlay; in the latter case, he makes 60 ounces net, equaling about 6% of
his outlay. Now, we must directly pursue the question of how much Jones, or any
other producer, will decide to invest in any particular line of production, and how
much he will decide to produce. It is clear that the determining in�uences are the
expected net income, its amount and its
percentage. Their
exact nature, however,
must wait on a more elaborate explanation of the relation between outlay, product,
Before �nally analyzing which point on the maximum product outlay curve
will be chosen, it is necessary to extend the analysis to remove the restrictive
assumption of two factors. What will be the situation with
number of factors?
This is a vital consideration, since it is very rare to �nd an actual case where only
number of factors, with market prices assumed to be given, the
producer’s investment decision turns out to be almost identical with the case of
two factors. The situation may not be diagrammed as in the case of two factors,
but the greater mathematical dif�culties in the description of the case of
does not by any means signify dif�culty for the producer. The producer is, again,
confronted with a complex of technological alternatives for producing various
amounts of output. Now, the production functions will be combinations of vari-
ous quantities of factors X, Y, Z, etc. Once again, a constant outlay will enable
a certain set of factors to be chosen, in accordance with their market prices. The
producer may draw up the list of alternative factor combinations and correspond-
ing outputs, plus a list of factor combinations that can possibly be bought at each
given outlay. And, once again, the producer will choose the maximum product
combination for each outlay. The fact that there are now many factors does not
change the desire of the producer to maximize his product for each possible out-
lay. The shape of the maximum product curve does not change; it is still true
Man, economy, & state
that a
greater outlay cannot yield a lower product, and that those greater outlays
which will not increase product will not be chosen. It is evident that the analysis
based on the maximum product curve is not changed by permitting any number
What of the interrelationships between the factors and the factor combinations
that will be chosen as points on the maximum product curve? Here, it is clear that
the situation, with
factors, is more complicated. It is, however, essentially the
same, and does not materially alter the analysis. It is still true that we can represent
the producer as adjusting, and substituting, all of his factors for each other. Each
factor is an imperfect substitute for each other factor, the degrees of imperfection
varying with the data of each concrete case. There can be no perfect substitutes
for different factors, and there are few or no cases of absolute �xed proportions
between all factors, so that, within limits, more of one factor can be substituted
for less of the others. The marginal rate of substitution
Smith estimates the future selling price of his product. It is quite possible
that, as Smith’s prospective product decreases, his selling price will rise. This
estimate depends on his idea of the market demand schedule for his individual
Murray N. Rothbard
this point, we must broaden slightly our application of the concept of
monopoly and competitive price. A
monopoly price
situation will occur
if less produced from a given money investment yields a greater pro�t, but also if
a lower money outlay, and its lower product, yields a greater pro�t because of the
higher selling price. It is clear, however, that this does not materially change our
analysis of competitive and monopoly price. In the previous section we assumed
a given investment and a lower than maximum product; here, a lower outlay can
also yield the same goal of a lower product, and without the waste of the former.
This, then, is the actual case. If the demand for the �rm’s product is inelastic, so
that a lower product, thrown as stock on the market, will so raise the price that
money revenue is increased, the �rm acts as a “monopolist” to cut back produc-
tion and outlay to the lower �gure. Thus, suppose that at a money outlay of 60
ounces, and at a maximum product of 50 units, as in Table
the price of the
per unit
is 2 ounces. The money revenue, then, will be 100 ounces, for a
net income of 40 ounces. If the demand schedule for the �rm’s product is inelastic
above this range, then, for example, a sale of 10 units will raise the price to 20
ounces, and a total revenue of 200 ounces. Now, obviously, Smith will not invest
60 ounces, produce 50 units, and then throw 40 of these units away in order to
acquire 200 ounces. We assumed this above, because we were dealing with the
assumption that money outlay is �xed at a certain amount. Obviously, he will
rather choose the minimum money outlay required to produce 10 units, i.e., 20
ounces. There will therefore be no need for him to throw away 40 units, and he
There is therefore no change in our analysis of the demand curve for the �rm,
and its relation to the incidence of monopoly price. This curve depends only on
quantity sold
, and bears no relation to how this quantity is produced. The
change in our analysis of the monopolist is that even
will choose the maximum
product for the money outlay that he spends. Even the monopolist will choose a
Table 4.11

Smith’s production decisions for Product P
– Product P
Total money outlay (gold ounces)
otal maximum product
Man, economy, & state
point on
his maximum product outlay schedule, and therefore even he strives to
gain further pro�ts producing whatever units he makes as ef�ciently and as pro-
ductively as possible. If his demand curve is inelastic, he will simply reduce his
money outlay from the amount that he would have invested under a competitive
price. The reduction of his outlay will reduce his product to the most pro�table
On the other hand, there is no reason to restrict the de�nition of competitive
price to a situation where the amount the �rm produces has absolutely no effect
on the price. It is clear that a change in the amount a �rm produces always
change the market stock of the product, and therefore tends to affect the price. It
may well be, of course, that, within the relevant range, the action of the �rm is
not large enough in relation to the product as a whole to change the market price.
There is no need, however, to restrict the discussion of competition to this limited
case. The only criterion is that the demand curve is not such as to raise revenue for
The following is a tabulation of Smith’s productive situation [
and the �rm pro
ducing P that he can invest in
], with the above total outlay and total product
schedules, plus an
expected selling price
schedule for each quantity produced
and sold of P. The selling price declines as the stock increases, but is not such as
to yield a monopoly price situation (i.e., an increased total product for the �rm
does not lower its gross revenue). From these three columns we can deduce three
others, which are also presented:
expected total money revenue
(which equals
expected selling price times product);
percentage of
money outlay). These three schedules are
Table 4.12

Smith’s money returns for various money outlays
(0, 0)(10, 0)
(30, 22.4)
(40, 47.6)
(50, 64)
(60, 75)
(90, 91)
(100, 98)
(70, 83)
(80, 83)
Money Outlay
Figure 4.14

Graphical illustration
of money revenue and outlay from production of
(10, 0)2040
Figure 4.13

Graphical illustration of production of Product P for various money outlays
Man, economy, & state
and 4.14 illustrate Table
In Figure
total units of product
are plotted on the vertical axis, as against corresponding money outlay on the hor-
izontal axis. The �gure
reveals the
amount of maximum total product that could
and would be produced at different amounts of monetary outlay. The result is the
product outlay curve, which is read vertically. There is a dotted line bypassing the
point at the money outlay of 80, because here the product curve is horizontal, and
no producer would consider such a waste of his resources as to produce at such
the product schedule is multiplied by the expected selling price
at each quantity of product, to yield the expected total revenue for each point of
outlay. This yields the total revenue schedule of Column 4. In this �gure, money
revenue is plotted on the vertical axis, and money outlay on the horizontal axis,
the result yielding a
revenue outlay curve
, which expresses the expected revenues
for each amount of invested money outlay.
It is clear that there is a direct resemblance between the shape of the revenue
and product curves, since the former is derived from the latter. At a 45 degree
angle between the two axes, there is a diagonal straight line. Since the units on
each axis of Figure
are exactly the same (money in gold ounces), with the
same distances, such a 45 degree line can also (vertically) represent
on the diagram. Thus, let us take a money outlay of 60 ounces. This is given by
the distance 0A on the horizontal axis. However, if we read vertically upwards
from point A, we �nd that the distance between A
and the
intersection point B on
is also precisely 60 ounces. Therefore, AB, and other such
This device makes �gure
a very easy task. At the outlay of 60 ounces,
the money outlay equals AB. What is the money revenue? This can be read off
from the revenue curve, and will equal AC, or 75 ounces. This permits a clear
Murray N. Rothbard
fer monetary loss from such investments. He will not invest 20 ounces, where
there would be no income from his investment. Which alternative will he choose
Most writers on this important subject have gone astray in their answers to
this question. They look at the schedules and simply assume that every producer
is interested in “maximum money pro�ts,” or, in better terminology, “maximum
net income.” Almost invariably, they would conclude in Smith’s case that Smith
would choose a money outlay of 60, and the expected money revenue of 75, since
this yields the highest expected net income, i.e., 15 ounces. This is greater than
any of the other alternatives. At �rst sight, this assumption seems plausible. Fur-
ther analysis, however, reveals the unsoundness of such a simple assumption. It
is true that if Smith invests 60 ounces, he expects a return of 75, and a net income
of 15. Yet compare this with the alternative of investing 50 ounces and obtaining
a net income of 14. In the former case, his
percentage net income
percentage net
income of only 10% on
these last 10 ounces. If, as seems plausible, Smith can �nd a greater
Man, economy, & state
�rm, a few in another, and several in a third, the investors hiring managers to
supervise the actual production.
In all of his actions, psychic factors being equal,
he will attempt to maximize the rate of net income from each unit of money that
he invests, thereby maximizing his total net income from his entire investment in
all branches. To pursue this approach will lead us to a theory of the savings and
investment of the investor, rather than of the output of the �rm, and thence to the
theory of the savings and investment of all the investors, indeed all the individu-
als, in the economy. This will be inextricably connected with the problem of
, which we have already seen in Chapter
to play
a determining role in
the decision of the individual as to how much he will save and invest compared to
This will be discussed in a later chapter.
It is evident that, in the pursuit of the maximum possible rates of net return, the
investors will invest each sum of money, large of small, in that �rm or in those
�rms where the rate of net return, for each size of money invested, will be at
its maximum. Investors will spurn 2% return projects to invest in expected 20%
At this point, we must make a crucial distinction in our analysis of investment
and production
– the
distinction between the investor or investors considering
investment in
�rms, and those contemplating the extension or continuance of
investment in
�rms. New �rms are those which are starting from the begin-
ning. If Smith is a new investor, he will decide as follows: [
with a given 60 ounces
], he will invest 50 ounces so as to produce 40 units [
in this �rm for Prod
], and earn an expected 28% net income [
and invest 10 ounces elsewhere to
try to earn more than a 10% marginal rate of net income
]. However, if he cannot
earn [
more than
] 10%, or 1 ounce, on 10 ounces elsewhere [
in another �rm
], he
will invest 60 ounces to produce 50 units [
of Product P
], and earn 25% on the
It is clear that there prevails on the market a tendency toward equalization of
expected net income rates on
�rm investments. Suppose that in one �rm or
product, the rate of net return is expected to be unusually high compared to other
investments, say 28%. It is clear that the new investors will �ock to invest in this
�rm, or in competing �rms producing the same product. If the data on the market
remain the same, then this �ood of investments will tend to lower the price of the
product, and raise the price of the factors, particularly those speci�c to that prod-
uct, until the expected rate of return will be drastically lowered. Furthermore, in
unusually unpro�table �rms, such as those earning 2%, the
investors, given
enough time, will allow their capital goods to wear out, and shift their investments
to the more pro�table investments. Suppose we postulate, then, an
evenly rotating
, such that the data never change, i.e., on each day consumer demand,
saving and investment, tastes and resources, and technological knowledge, will be
the same.
In this case, given enough time, the rate of net return will be equalized
in every �rm and every branch of production
. This will be an economy of cer-
tainty, since there will be no uncertainty of future price, demand, or supply. In this
rate of return will invariably be the realized rate of return, and
this will be equalized for every �rm and investment. This rate of return is called
Murray N. Rothbard
e rate of interest
. What rate will it be, and how will it be determined, we
must leave to further chapters.
In the evenly rotating economy, then, every �rm
will earn the same net return, say 5%. Since there is no uncertainty, every �rm will
[Returning to the individual investor, Smith, in the above example we assumed
he was going to invest 60 ounces in one or more �rms. But how does Smith choose
the amount of money that he is going to invest at all? We have shown above that
we cannot simply concentrate on maximum net income from an investment, but
must also pay attention to its rate of net income.] Can we then say that Smith will
invest that sum which will yield him the largest
percentage, or
rate of net income?
No, we cannot simply make such a plausible statement, either. Suppose, for exam-
ple, we consider the investment of 40 ounces, yielding a
net income of
19%. An additional investment of 10 ounces would yield an additional net income
of 14 minus 7.6 ounces, which equals 6.4 ounces [for a rate of net income of 28%
on his 50 ounces]. This is a return of 6.4 ounces on an outlay of 10 ounces, a mar-
ginal rate of return, or marginal rate of net income, of 64%. Yet, circumstances
are conceivable when Smith would not make the additional investment. We must
never forget, as we pointed out in Chapter
III above,
that every individual is
always engaged in balancing his various consumption, and his various investment
expenditures, and additions or subtractions from his cash balances. Suppose, now,
that Smith has a money stock of 200 ounces, which he is in the process of allocat-
ing. It is entirely possible that, while he may choose to invest 40 ounces in factors
of production yielding him a 19% net income, even so high an additional return
of 64% on the next 10 ounces will not induce him to restrict his consumption
further. In such a case, Smith prefers present consumption spending with these 10
ounces to the 64% rate of income; therefore, his marginal rate of time preference
for these 10 ounces is higher than 64%, and he does not make the investment. His
investment in the product will then be 40 ounces and his level of output will be 28,
every case, therefore, the amount of money investment by the producer,
and consequently the amount of product made, depends on the interrelation-
ship between the
expected rate of net income
and the individual’s
rate of time
This interrelationship, speci�cally, is most important in its
The reader is referred again to Chapter
I, the
basic foundation for the later analy-
There we saw how man allocates his stock of goods in accordance with their
marginal utility in the various uses.
We also saw how man allocates his labor
in accordance with the marginal utility of the expected products in the various
uses, and with the marginal disutility of the forgone leisure.
This is particularly
relevant. We recall that each man allocates his labor in units, say hours, to that
particular use which provides the greatest value of marginal product on his value
This analysis, in its essence, is applicable to the present problem. Smith is
choosing, not between the utility of labor and its product versus leisure forgone,
but between the
utility of an expected future net money income
, and between the
Man, economy, & state
expected net money income
are taken from
other columns require extended explanation. The purpose of the added columns
is to better analyze Smith’s �nal investment decision in production. Column 7
–7.6, equaling
15.2, since there is no possibility that Smith would
ever consider an outlay of 30 ounces, yielding a negative return. The
disutility of present consumer goods forgone
, by investing in factors of produc-
tion. Again, his decision in every case is
, i.e., he deals with divisible
units of a good. In this case, he is dealing with units of a money commodity used
to purchase factors. He knows, or believes that he knows, the various technologi-
cal alternatives by means of which certain quantities of factors will yield him cer-
tain quantities of product, and from this, he estimates the expected money revenue
Thus, in Table
let us consider an expansion of Smith’s choices [
for the �rm
producing Product P
] as shown in Table
Table 4.13

Smith’s money returns for various outlays, continued
Exp. net
Exp. marginal
Murray N. Rothbard
between 0 and 10, 10 and 20, etc., but between 0 and 40 only. The marginal
net income at 40, then, equals 7.6 minus 0, which equals 7.6. From then on, the
margin occurs every 10 ounces, for that is the decision unit, so to speak. Smith
estimates that the
10 ounces of investment will increase his net income from
7.6 ounces to 14 ounces
– giving
him a marginal net income by these 10 ounces
of 6.4. From 50 to 60, the 10 new ounces only increase the net income from 14
to 15 ounces, a marginal net income of 1 ounce. After this point, the net income
declines; therefore, the marginal net income is
. Thus, after 60 ounces,
an additional 10 ounces will lower the net income to 13; thus, its marginal net
Immediately, we have learned something more about Smith’s eventual invest-
ment production decision. It is obvious that
no one will knowingly invest addi
– the
two suc-
cessive points of decisions. The marginal outlay at 60 is also
10 ounces
. After that,
there is no need to apply the concept, because these decisions have been ruled
out. Column 9 lists the
of net
income which each additional investment of units of money will earn. At
40, an addition of 40 ounces earns 7.6 ounces net; this is a
percentage return
. At 50, an addition of 10 ounces earns 6.4 more ounces of revenue
– a
percentage return
. At 60, the additional 10 ounces earns only one
The alternatives that remain for Smith’s consideration are condensed in
, taken from Tables
Table 4.14

Smith’s money returns for various outlays, total
Exp. marg.
Man, economy, & state
o summarize how we obtained these columns: from technological knowl
edge, Smith could calculate the maximum physical product that could be
obtained from each combination of factors, and this with the prices of factors,
which we have taken as given, determines the maximum total product schedule
for each possible alternative outlay of money investment. Horizontal spaces
in the schedule were eliminated, i.e., where the marginal product is zero for
each increase in outlay (it can never be negative). For each possible product,
Smith estimates the selling price for which he could sell the product, and this
times the quantity produced yields him the revenue schedule for each outlay.
The net income is then easily calculated, and points where this absolute net
money income is expected to be zero or negative are immediately eliminated
from consideration. The
percentage that
the net income
bears to the money outlay at each point.
percentage return
from the total
investment, and the
return on
each successive dose of monetary investment? The answer is de�ni-
tively yes; in fact, at any point, the
rate of net income is equal to the weighted
average of the rates of marginal net income at that and preceding points
weights being the size of the marginal outlay at each point. Thus, at an outlay of
50, the rate of net income is 28. This is equal to the average of the rates of mar-
ginal net income at that and preceding points, namely 64 and 19. However, it is
This would be an
unweighted average
of the two numbers. Each number is mul-
tiplied by the
marginal outlay
at that point, and the sums are divided by the sums
of the marginal outlays, which is total outlay at the �nal point. Thus, 19 times
Murray N. Rothbard
, at the money outlay of 60, the rate of net income equals 40 times 19, plus 64
Furthermore, at the
feasible marginal step, whatever it may be (in this
case, it is from 0 to 40 ounces), the rate of net income equals the rate of marginal
net income, the net income equals the marginal net income, and the total outlay
equals the marginal outlay. This is because the starting point is always zero
– no
– and
the total of something after the �rst step is the same as the dif-
Thus, we see that the average rate of return is the weighted average of the pre-
ceding marginal rates of return, and that at the �rst step, the two rates of return are
equal. This indicates another important truth:
that the average rate at any point is
equal to the marginal rate, if the distance between that point and zero is taken as
. Thus, if Smith is considering the investment of 60 ounces, his expected
average rate of net income is equal to the marginal rate of net income, if the “mar-
gin” is taken as a unit of 60 ounces. Thus, the decision on an investment of a sum
of money is a “marginal” one in two senses: a) in the sense of the last small unit
of money and its return, and b) in the sense of the return to a marginal unit taken
as the size of the sum itself. Both sizes of marginal chunks are discrete steps, and
both are taken into consideration by the actor.
[Now we must return to the important concept of the rate of time preference
and integrate our analysis of the rate of net income.] Any man, in deciding upon
the allocating of any given sum of money between consumption and investment
purposes, estimates the expected yield of net money income to be derived from
his investment (modi�ed where necessary by other psychic considerations) and
compares it with his minimum required monetary return from that sum of money,
taking into consideration his total stock, and his value scale. This minimum rate
of return is his rate of time preference: any investment which he expects will yield
him a lesser return will not be made. [Thus Smith and his investment decisions
in the �rm producing Product P, as shown in Table
are compared with his
rate of time preference.] This rate of time preference is set by his relative valu-
ations of present and future satisfaction; it is his “minimum supply price”
– the
“price” at which he will part with his present money in order to invest in a
prospect of a higher income at some time in the future. As an individual allocates
more money to investment and less to consumption at any time, his marginal
rate of time preference increases, until it �nally becomes prohibitively high for
any investment. This fact is set by man’s necessity to consume in any given pre-
sent, before making investments for the future. The entire schedule of a man’s
time preference rate, therefore, increases as the invested outlay increases, �nally
nearing verticality. [It can be calculated in marginal and average form like net
income.] If the rate of net income from the investment outlay is greater than the
rate of time preference, he will make the investment; if not, he will abstain from
Man, economy, & state
The investor
Smith, in sum, does not simply try to maximize his expected net
money income. He, like every actor in every situation and every choice, tries to
maximize his psychic revenue
psychic pro�t
. He cannot only con-
sider money income from the investment. He must weigh this against his psychic
time preference rates. His maximization of psychic revenue, therefore, impels his
investing so long as the rate of average and marginal net income exceeds his aver-
age and marginal rates of time preference.
Investment decisions in a �rm, then,
or B
or C, but in his income from all
of these investments as a whole. Therefore, he weighs his average and marginal
rates of time preference against the gross revenue that can be achieved from all
of his investments at the given outlay. Thus, at any total outlay, say 120 ounces,
he determines what distribution of money among the alternative investments will
yield him the maximum total gross revenue, and hence the maximum net income
and maximum rate of net income for the given outlay. At each point of outlay, he
decided on the distribution that will accord him the maximum gross revenue, and
therefore he is able to deduce the maximum average and marginal rates of net
income for each outlay. He invests his money up till the largest amount at which
the maximum average and marginal rates of net income are larger than his average
and marginal rate of time preference, respectively. At this amount, he distributes
his outlay among the various enterprises in accordance with the “maximum rev-
enue distribution” at that outlay.
In the �nal form of the Law of Investment Decision, then, there is not the
previous direct and complete link between investment outlay of the individual
producer and the output of the individual product as there is when the individ-
ual producer invests in only one line of production. It is still true that the actor
invests in production
– in
general up to the last point that his expected average and
Murray N. Rothbard
net income schedules in Table
reveal what net income Smith expects
to enjoy when investing a certain outlay in any given line of production. But these
schedules permit combination into

Smith’s money outlays for producing Products P
Man, economy, & state
13 plus
8, or 21 net ounces. At each hypothetical outlay, the investor picks what
appears to be that combination that will yield the highest net income. Table

Smith’s money returns for various investment decisions
weighted average
of these two yields by the
respective outlays is: 30 times 43, plus 20 times 40, divided by 50. This equals
42%, the weighted average, which also equals the rate of maximum net income
(amount of maximum net income divided by money outlay). Thus, the best dis-
tribution can be determined from schedules of rates of net income for each of the
various outlays in the various lines of production. In this case, the distribution is
not con�ned to producing just Q, even though producing Q is more pro�table than
As Table
shows, from the maximum net income schedule, there can be
deduced schedules of rates of maximum net income, marginal outlay, marginal
maximum net income, rates of marginal maximum net income, etc. Thus:
Murray N. Rothbard
or any investor, then proceeds analogously with the case of one product,
investing money outlay (in the best distributions) up to the largest amount that
– in
usual case, this line of investment will be the one that is expected to yield the great
est net return from the outlay. Exceptions are cases where other psychic factors,
such as particular like, or dislike, for the production process or the product itself,
alters the decision from a pure consideration of monetary return. Otherwise, a man
We have thus seen what determines the amount of stock of any good that will
be produced in any particular period
– it
will be the amount that the producer had
invested in a previous period in order to aim at such production. The amount of
previous investment depends on the producer’s anticipated net monetary return. It
is clear that an increase in anticipated rate of net income in any line of production
will tend to increase the investor’s outlay in that product, and that on the other
hand a decrease in the anticipated return will tend to diminish his investment in
that process. If we interpret the concept of “increase in rate of net income” as
meaning an increase in the entire rate of net income schedule, so that at each
outlay of product, net income is expected to increase, it is obvious that the rate
of net income schedule will intersect the investor’s time preference rate schedule
at a further point, so that an increase in the expected net income schedule will
increase the amount of investment outlay in that product, and contrary for the
decrease. Furthermore, an increase in expected return for producing P will tend to
shift more of the investment outlay to this �rm from competing �rms Q, R, etc.,
Man, economy, & state
As a
matter of fact, changes in anticipated rate of net income are most likely to
take place throughout the entire range of the schedule. The factors that can change
the rate of return are: a) expected future selling price, b) the prices of the factors,
and c) the producers’ production function
– the
physical ef�ciency in converting
quantities of factors into quantities of product. It is evident that, with factor prices
here assumed to be given and known, the producer’s anticipations of future income
are governed by his anticipations of selling price and of his production function. It is
clear that a rise in expected selling price for any good will,
ceteris paribus
, increase
the amount of investment outlay in its production; and that an increase in physical
productivity for any good will,
ceteris paribus
, increase the amount of investment
outlay. Conversely, decreases in expected selling price and/or decreases in physical
We have learned, therefore, that consumers’ goods prices are determined by
consumers’ demand schedule and by the stock produced and sold; that the sales
of produced stock depend on anticipated future price; that the amount of stock
produced depends on previous investment in production; that the previous invest-
ment in production depends on the net money income that the investor anticipated
receiving, and the amount of investment will be up to the last amount at which
the anticipated rate of return exceeds the rate of time preference; and that the
anticipated rate of return depends on expected future selling price and production
technique (given factor prices). In the last analysis, then, consumers’ goods prices
depend on consumers’ demand schedules and general time preferences, produc-
ers’ anticipations of prices, and productive techniques.
Many questions remain to be answered. Among them is the discussion in Chap-
on the �nal supply curve of the producers as compared to the stock on the
Figure 4.15

Murray N. Rothbard
The “�nal supply curve” is the amount that will be called forth in supply
in the future by certain prices. The discussion in Chapter
implicitly assumed
that the present ruling prices would be the ones that would be anticipated in the
future. Thus, Figure
This implicitly assumes that the present price of P1 is assumed to be the future
price, and will call for the equivalent amount on the SF curve, which will tend
to lower the �nal market price to P2. However, we may alter this restriction and
make the necessary mental allowances for any anticipated change in price. The
main point of the diagram still obtains
– that
the present market price is not neces-
sarily the “�nal” one toward which the market forces are tending. The question
then remains: what principles determine the “�nal” equilibrium market prices?
Even though this price is never attained in practice, it is important because it is
the point (though always shifting) toward which prices tend to move. And a �nal
selling price, given the productive technique, and given factor prices tend to set
net entrepreneurial income. On what basis does entrepreneurial net income, the
driving force in the money economy, tend to be determined? This problem, along
Chapter 4
of this volume is a lightly revised version of a paper published in the
terly Journal of Austrian Economics
, vol. 18, no. 4 (Winter 2015): 456–486.
: Rothbard’s reference to his earlier presentation of the “�nal sup-
ply curve” is absent from MES. Rothbard’s discussion of supply and demand for an
already produced stock of goods and his introductory analysis of entrepreneurship and
production can be found in Rothbard (1962, 153–161, 249–257). See also (pp.
126) Rothbard’s further discussion of the �nal supply curve.
“In every
case the choice made is, at the moment when made, a present choice. We
have no future desires though we may have a present forecast of a future desire. ‘Future
desires’ means desires that will be present at some future time. Present desires are all
those desires now being weighed in choice. Present desires may be either desires for
present uses or for future uses (either in the same or in different goods). A
desire for
future uses is but the anticipation of a future desire, though the two may be
of unequal magnitude. It appears therefore that all time-choices are, in the last analysis,
reducible to choices between present desires for psychic incomes occurring at different
: See the editor’s endnote below on p.
It is,
of course, likely that Jones will weigh his decision on the basis of expected returns
over a much longer period, say a decade, in which these returns may be considered to
take place for a ten year period. We can adjust his calculations to cover any desired
Man, economy, & state
Here it must be noted that the constancy of price assumed in Case (a) did not neces-
follow for all possible decisions of Jones. Thus, if he decided not to produce
the good at all, the price might well be affected, and be, say, 12 ounces instead of the
ounces if
he did go into production. But the constancy of price is only assumed for
relevant range of choice
– in
this case between the three different combinations.
Case (a) only needed to assume that, between a product of 96 and 110 units, market
supply would not be affected enough to change the price.
On competitive
price and monopoly price, see Fetter (1915, pp.
77–84, 381–385);
(1949, pp.
273–279, 354–376),
Mises (1951 [1922], pp.
385–392), Menger
Wieser (1927 [1914], pp.
: Rothbard slightly modi�es his de�nitions of monopoly and competi-
For example,
see Chamberlain (1942). Recently, however, Professor Chamberlin has
repudiated the implications drawn by his followers that the “pure competition” situa-
: This later section, whether or not it was intended to be included in the
in a later one, was not found by the editor in the Rothbard archives.
See Rothbard (1962, pp.
677–680) for
his mature theory of monopoly gains on the
free market. It is important to note that here he no longer uses the competitive-versus-
See Ely
(1917, pp.
190–191). The
famous Blackstone gave almost the same de�nition,
The battle
of the equal-liberty movement against monopoly has had a long history in
England. In 1603, the British courts decided, with respect to one of Queen Elizabeth’s
numerous grants of privilege: “That it is a monopoly and against the common law. All
. are
pro�table for the Commonwealth, and therefore the grant to have the sole
making of them is against the common law and the bene�t and liberty of the subject.”
In 1624, Parliament declared that “all monopolies are altogether contrary to the laws of
this realm and are and shall be void.” In the American states, the Declaration of Rights
of the Maryland Constitution asserted: “monopolies are odious, contrary to the spirit
of a free government and the principles of commerce” Ely (1917, pp.
191–192). See
alker (1911, pp.
: In this footnote, Rothbard refers the reader to later chapters on the
hampered market regarding various monopoly grants. Rothbard originally wrote mul
tiple chapters on the hampered market before the publisher required that he cut down
the length of the book and remove controversial parts of the manuscript. Rothbard
then had to write a summary chapter
of his
analysis (Rothbard, 1962, pp.
s chapters on government intervention were eventually published as Roth
bard (2009 [1970]). See Rothbard (1970, pp.
for his analysis of various
grants of monopolistic privilege. Rothbard also mentions in this footnote that copy
rights and patents would be discussed below: see Rothbard (1962, pp.
745–754) for
That such
was the original de�nition of monopoly in economics as well as law is
demonstrated by the de�nition of the economist Arthur Latham Perry: “A
the derivation of the word implies, is a restriction imposed by a government upon
the sale of certain services” (Perry, 1892, p.
190). Still
earlier, Adam Smith discussed
monopoly in similar terms, and pointed out how monopolists may use the government
privileges to restrict sales and raise selling prices: “Such enhancements of the mar-
ket price may last as long as the regulations of police which give occasion to them”
Murray N. Rothbard
Benham (1941,
pp. 232–233).
On the rapid breakup of even a relatively successful
cartel, see Fairchild et
al. (1926,
54–55). Also
see Molinari (1904, pp.
In many
cases, fear of possible outside competition prevents any formation of a cartel,
even when other conditions seem favorable. This is known as the in�uence of
See T
ucker (1926, pp.
248–257). For
a defense of voluntary combinations from a juris-
tic point of view, see Cooley (1878, pp.
270–271). Also
see Flint (1902) and Croly
Does our
discussion imply, as Dorfman (1949, p.
247) has
charged, that “whatever is,
is right”? We cannot enter into a discussion of the relation of economics to ethics at this
point, but we can state brie�y that our answer, pertaining to the free market, is a quali-
�ed Yes. Speci�cally, our statement would be: Given the ends on the value scales of
individuals, as revealed by their real actions, the maximum satisfaction of those ends
for every person is achieved only on the free market. Whether individuals have the
“proper” ends or not is another question entirely and cannot be decided by economics.
11–112) and Weiler (1952, p.
It is
obvious that, for each of these combinations, more of both factors will produce
at least as much as, and probably more than, the particular product. Thus, if (40X;
100Y) can produce 105 units of product, so can (45X; 105Y). This follows from the
nature of scarce goods and scarce factors. The use of the latter combination to produce
105 units, however, would clearly be senseless. The latter, obviously more expensive,
combination would either produce more and the surplus thrown away
– which
be a ridiculous procedure; or else would produce just as much, in which case the fac-
tors would still be wasted and needless money expended. In describing constant outlay
combinations, therefore, we assume that those combinations which are obviously more
expensive for each product
– using
– will
be discarded at once.
The only question then comes from the partial substitutability of one factor for another.
The absurdity
of the “technocratic fallacy” here becomes obvious. The technocratic
charge is that business conducts “production for pro�t” instead of “production for
use,” and that the latter would prevail if engineers were granted dictatorial control over
the productive system. It is clear from the discussion that technology cannot solve the
production problem, and that therefore “production for (money) pro�t”
possible method of production
beyond the very primitive level. Technology by itself
could neither provide a guide to “maximizing production” nor to determining
should be produced. And it is also evident that business on the market takes account of
the technological factor as much as is necessarily possible. It should also be clear that
production for pro�t is necessarily production for “use.” There is no reason to produce
any good except to supply the demand for its use by consumers, whether the consumer
is other persons or the producer himself (in the more primitive production situations).
: See also (pp.
108–109) Rothbard’
s analysis when more than two fac-
On the
vital importance of knowledge of “particular circumstances of time and place”
: This analysis of factor pricing was planned to be in a later section,
however it was never written because Rothbard changed his mind on the usefulness of
Man, economy, & state
This statement will be surprising only to those who have been misled by the use of the
ferential calculus in economics. In calculus, the steps between points are treated as
in�nitely small, and therefore the marginal is thought to be the in�nitesimal. In that
case, “small” sized units will be recognized as approximations to some “ideal” mar-
ginal unit, but a “big” unit will not be thought of as marginal. Actually, the size of a
marginal unit can be any amount, depending on the decision to be made. There is noth-
ing ideal about in�nitesimally small units, and they are not relevant to the real world
: Strictly speaking, it must be greater than
equal to. An investor
would still invest if the rate of return is equal to the rate of time preference, since his
rate of time preference is the
he would need to earn in order to forgo the
present money and invest. In the Evenly Rotating Economy, each investor only earns
: Although not discussed in terms of a “�nal supply curve,” a similar
. New York: Pitman Publishing.
Economic Analysis
. 1st ed. New York: Harper and Bros.
Brown, Harry Gunnison. 1908. “Competitive and Monopolistic Price-Making.”
Chamberlain, Edward H. 1942.
Theory of Monopolistic Competition
. Cambridge, MA:
“Product Heterogeneity and Public Policy.”
Cooley, Thomas McIntyre. March
1878. “Limits
to State Control of Private Business.”
In Smith’
s particular case, marginal net income is only negative in the early and later
stages. In some cases, there may well be points where the marginal net income is nega-
Murray N. Rothbard
Ely, Richard T., T. S. Adams, Max Otto Lorenz, and Allyn Abbott Young. 1917.
. New York: Palgrave Macmillan.
Fairchild, Fred R., Edgar S. Furniss, and Norman S. Buck. 1926.
Elementary Economics
New York: Palgrave Macmillan.
Fay, Charles Norman. 1912.
. New York: Doubleday.
Too Much Government
. New York: Doubleday.
Fetter, Frank A. 1915.
. New York: The Century Co.
Flint, Charles R. 1902. “Centralization and Natural Law.”
Friedrich August. 1948 [1945]. “The Use of Knowledge in Society.” In
ism and Economic Order
Menger, Carl. 1871.
Mises, Ludwig von. 1949 [1999].
. Auburn, AL: Ludwig von Mises Institute.
. 2nd ed. New Haven, CT: Yale University Press.
The Society of Tomorrow
. New York: G.
P. Putnam’
Newman, Patrick. 2018. “From Marshallian Partial Equilibrium to Austrian General Equi-
librium: The Evolution of Rothbard’s Production Theory.” In Matthew McCaffrey, ed.,
The Economic Theory of Costs: Foundations and New Directions
. Abingdon, UK: Rout-
Perry, Arthur Latham. 1892.
Political Economy
. 21st ed. New York: Charles Scribner’s
. Auburn, AL: Ludwig von Mises Institute.
Risk, uncertainty, and cost
In economic analysis, the word “cost” is used with two very different meanings.
On the one hand, it refers to the subjective value of the most important foregone
choice alternative (opportunity cost). On the other hand, it is also used to desig-
nate the monetary value of the factors of production that are being consumed in
a business venture (production costs). The present chapter
deals with
costs in the
Most present-day economists consider observable interest rates to be the arith-
metic sum of three main components, each of which is held to result from a dis-
tinct cause. First, there is a pure or real or risk-free interest rate component, which
is typically believed to spring from time preference. Second, there is a risk pre-
mium that compensates the investor for market risks. Third, there is a price pre-
mium that compensates the investor for losses in the purchasing power of money.
If we denominate the observable gross market rate with the letter
, the real or pure
interest rate with the letter
, the risk premium with the letter
(as in chance), and
the price premium with
, then something like the following equation [1] is sup-
This equation
can then be applied to calculate the risk-free interest rate; with
being derived from observation; variable
supposed to be equal to
some calculated price-in�ation rate; and variable
supposed to be equal to some
calculated risk premium, typically a standard deviation around some average
value. The equation can also be used to determine the value of risky assets by
discounting their future cash �ows, etc. Whatever the variant of this approach,
such as the capital-asset pricing model, the basic idea is always the same: observ-
able interest rates are held to be the arithmetic sum of different components, each
of which can be determined in separation from the others. The basic problem of
this conception is that it is disconnected from ordinary supply-and-demand price
theory. Human choice and human action either do not enter the picture at all, or
they enter the picture under highly contrived assumptions, such as in the capital-
The myth of the risk premium
Jörg Guido Hülsmann
Jörg Guido Hülsmann
In the present paper, we present an alternative realist approach to the study
of risk, based on Ludwig von Mises’s distinction between case probability and
class probability. In light of this realist approach, it will appear that the prevailing
conception of risk as related to the gross rate of interest is ill-founded. It is wrong
to conceive of the gross interest rate as the sum of separate components. A
analysis reveals
that the whole idea of a risk premium within the gross rate of
The chapter
is or
ganised as follows. In section
1, we
present the distinction
between case-probable and class-probable judgements. Section
2 contains
a real-
ist approach to the analysis of human action under uncertainty. In section
3, we
this approach to study the impact of risk on the return on capital. In sec-

Case probability and class probability
Frank Knight (1971 [1921], pp.
and 198f) revolutionised the economic analy-
sis of uncertainty, pro�t, and loss by stressing the crucial difference between two
types of uncertainty, namely, quanti�able uncertainty or risk; and unquanti�able
uncertainty or, simply, uncertainty. Knight highlighted the crucial fact that risk,
strictly speaking, entails no uncertainty at all. It can be anticipated in advance.
Entrepreneurs can protect themselves against it through suitable provisions in
their balance sheets or by insurance contracts. Risk cannot therefore be the origin
Ludwig von Mises (1949) later elaborated on this distinction by stripping it
to its logical core. Most notably, Mises dissociated the analysis of probability
from the analysis of risk. The theory of probability exclusively concerns
questions (the truth of a judgement), while the analysis of risk also concerns
judgements (risk is an
consequence of action). Mises stressed that the
Knightian categories of risk and uncertainty were rooted in two completely differ-
ent types of probability, which shared only one basic characteristic:
A statement is probable if our knowledge concerning its content is de�cient.
We do not know everything which would be required for a de�nite decision
between true and not true. But, on the other hand, we do know something
about it; we are in a position to say more than simply
non liquet
, Mises distinguished class probability from case probability. The former
is applied in the natural sciences, while the second is applied in the sciences of
We know or assume to know, with regard to the problem concerned, every-
thing about the behavior of a whole class of events or phenomena; but about
The myth of the risk premium
the actual singular events or phenomena we know nothing but that they are
crucial feature of a class-probable judgement is that the person in question
ignores the causal sequence that brings about a concrete event. For example, he
this bottle breaks rather than another. He ignores
this barn burns
rather than another. But he knows from experience that of all the bottles that are
�lled in that factory, .08% will break on any given day; and that of all the barns
in his county .03% will burn down each year. He knows that this concrete bottle
is one of the bottles �lled in that factory, and he knows that this barn stands in his
county. Therefore, even though he is ignorant of the exact causes that will prompt
this bottle to break and this barn to burn, he can make a class-probable judgement
Knowledge and ignorance are combined quite differently when it comes to
case-probable judgements. This is how Mises de�nes case probability: “We know,
with regard to a particular event, some of the factors which determine its outcome;
but there are other determining factors about which we know nothing” (Mises,
1949, p.
He immediately adds: “Case probability has nothing in common
with class probability but the incompleteness of our knowledge. In every other
regard the two are entirely different” (Mises, 1949, p.
Indeed, the person
who makes a case-probable judgement knows this and that exact causal sequence.
For example, he knows that the revenue he will earn with his bakery depends
on the number of other bakeries within walking distance. He knows that he can
produce computer screens of type X with technique A
and also
with technique B.
He knows that that the tomato output of his farm will be at maximum with 250
sunny days and 80 days of rain. He knows the laws of mathematics, of physics,
Knowledge in all these cases is exact and sometimes even universal. But it is
de�cient in two regards. On the one hand, it is incomplete. The person in ques-
tion knows the in�uence that this and that factor will have on his revenue, on his
physical output, etc. But there are other factors that also might come into play
and about which he knows nothing. On the other hand, when confronted with the
choices of other people, he is also ignorant of their future value judgements. That
is, he ignores how several factors in�uencing these judgements will combine in
Consider again the bakery example. Our would-be entrepreneur knows that
the revenue of the bakery depends on a multitude of concrete causes, such as the
number of other bakeries within walking distance, the number of families with
children, the revenue of these families, the effort he puts into merchandising his
croissants and breads, the unit prices at which he sells, etc. But he does not know
exactly the relative impact of each of these factors on his income. That is, he does
not know how much the customers will value a nice presentation and how much
their decisions will depend on price. He might have some rough idea about the
Jörg Guido Hülsmann
relative importance of each of these factors in the past. But he cannot extrapolate
this knowledge into the future. He needs to speculate or, in Mises’s words, he

ealist approach to human action under uncertainty
The conventional way to integrate risk into economic analysis is fatally �awed in the
very way it conceives of the problem. The implicit assumption is that risk is some
thing ‘out there’ which can be studied by economists and other scholars, and which
sooner or later will also be discovered by all other rational decision-makers. The
risk-that-is-out-there can be included in the utility functions of all economic agents
and thereby determine demand and supply schedules. However, this conception is
untenable because risk (more generally speaking: probability) is an epistemic, not
an ontic category. Probable judgements are relevant for economics only to the extent
Subjective value is the �lter through which all probably true judgements, and
therefore all assessments of risk, have to pass in order to become relevant for
human action. And only the judgements that pass through that �lter are therefore
relevant for economics. Moreover, and most importantly, probable judgements
that are considered to be important lead to action. Economic goods that are con-
sidered to be important for the realisation of one’s projects are being intentionally
brought into existence (that is, produced), whereas all factors opposed to that
realisation are eliminated as far as possible. Let us explain these considerations
Probability theory makes propositions about
about the world,
not about the world as such or about human action in particular. The world as
such and the transformation of the world as such are not probable. They are what
they are, irrespective of how much human beings know about them. All things
that happen in our world are completely determined. They are all subject to the
inexorable laws of cause and effect. But human beings can only gain a very partial
knowledge of these laws. It is our knowledge about the world that is more or less
probable, not the world as such (see Poincaré, 1912, p.
2; Fisher
, 1906, p.
, probability theory as such has no
relevance for economics.
Human action is guided by judgements that are subjectively perceived to be
probably true. But such judgements of truth are always mediated through value
judgements. There is no direct connection between the (probable) truth of any
proposition on the one hand, and human action on the other hand. Human beings
always and everywhere need to weigh different (and often con�icting) judge-
ments about future states of affairs in order to act.
Economics deals with human action. It deals with the causes and consequences
of decision-making in a context of scarcity. It does not concern itself with what
The myth of the risk premium
the decision-makers
about themselves or the world around them. Econo-
mists take due account of the fact that the acting persons must have de�nite ideas
about themselves and the world around them. They are not primarily interested in
the truth of these ideas. They analyse the (intentional) consequences of true ideas
no less than the (unintentional) consequences of wrong ones. But the �rst step of
analysis always consists of adopting the point of view of the acting person. Risk,
then, becomes relevant for economics to the extent that it relates to the value
judgements of that person. The realist method of analysing risk adopts the point
of view of subjective value, not the point of view of any “objective” probabilities
The realist approach leads to us to stress two fundamental points that are at
odds with the conventional approach. The �rst is that not all things that are risky
(that is, probable and undesirable) from
point of view (for example, from the
point of view of a scholar) are also relevant from the point of view of the acting
person. The second is that the incompleteness of knowledge, which characterises
all case-probable judgements, is not risky
per se
and does not necessarily have an
From a subjectivist point of view, there is very often no risk associated with the
chosen course of action. This is particularly clear in the case of consumer choice.
Most of the risks that an external observer might identify in human action simply
disappear from the point of view of the protagonist. Smith thinks that drinking
beer makes him happy. His mother, the external observer, disagrees. She thinks
his beer drinking is risky. But for Smith this is irrelevant. He believes to know that
drinking beer makes him happy. It is true that he ‘only’ believes to know, but for
economics this is all that counts. Smith’s opinion might be wrong. But from an
economic point of view, this would be irrelevant because it would not in�uence
the way Smith values beer relative to money.
Similar considerations can be brought to bear on the problem of incomplete
knowledge. The more Smith knows about the multifarious causal connections into
which A
and B
are embedded, the better informed are his value judgements. But
more information does not necessarily
his value judgements. He may know
only one thing about an economic good and be perfectly happy with this very par
tial knowledge because, for him, it is the only relevant consideration. For example,
in choosing a vacation hotel, Smith might only be interested in the distance between
his hotel room and the beach. There are an unlimited number of other circum
stances that would also in�uence his vacation experience. But
only the distance between the bed and the beach. We might call him foolish, but if
Even if we admit that most consumer decisions are made with more circum-
spection, the fact remains that very often there is no signi�cant (subjective) risk
involved. Consumers’ preferences very often depend on a relatively small number
of characteristics of the goods that they consider, and they are (or rather, believe
to be) perfectly well-informed about these characteristics. If they hesitate, it is
not because of the risks associated with the factual assessment, but because of the
Jörg Guido Hülsmann
The main risk of consumer choices pertains to durable goods. The risk is that
one’s future valuation will be different from one’s present valuation. Future valu-
ations may be different because other factors than the ones that count now will
become primordial. For example, in buying an apartment, a young family will
not necessarily have a lot of appreciation for single-�oor units, whereas a retired
couple typically will. Future valuations may change as well because some of the
known factors that count will deteriorate. For example, the subjective value of
the apartment might sink when very obnoxious neighbours move into the other
But consumers are typically aware of this problem. Precisely because future
conditions might be different from present ones they have an incentive (a) to
investigate as much as possible the factors that might in�uence future conditions,
including their own valuations; and (b)
to act strategically in the present
, in such
a way as to eliminate as far as possible the in�uence of factors that are likely to
have a negative impact, that is, case-probable risks. Indeed, human action is the
most important means through which the probability of desirable in�uences is
This strategic approach is likely to be even more pronounced in the case of pro-
ducers. Brown is an entrepreneur fabricating headphones. The essential purpose
of his activity is to gain his living, that is, to earn monetary revenue. He does this
through the network of exchanges. He buys factors of production and sells head-
phones. For him the physical characteristics of the factors that he buys and of the
headphones that he sells are only remotely connected to the immediate causes of
his success, which are market prices, that is, the conditions of demand and supply.
In assessing market conditions, Brown is very much concerned with the prob-
lem of incomplete knowledge, much more so than Smith the consumer. Brown’s
success depends on the valuations of
people. Their valuations are driven
by a great number of circumstances, which our producer does not and cannot
completely know. Moreover, for Brown the quality of his own knowledge is a
crucial factor. Smith might believe that beer drinking makes him happy even if,
in fact, it ruins his physical and mental health. Then beer drinking would be the
right thing to do from his subjective point of view. Smith would be a “successful
consumer.” By contrast, for Brown, it would be fatal to use a production technique
that impairs the quality of his products as perceived by the customers. He can-
not be successful by producing goods and services that are to his own liking. He
must meet the demands of the customers. It follows that for entrepreneurs such as
Brown there are very strong incentives to �ll the voids of his incomplete knowl-
Moreover, and crucially, entrepreneurs such as Brown do not so much produce
material goods and services as they “produce success.” They act strategically, in
whatever way is necessary, within the limits of reasonable cost, to bring all factors
into play that favour their success, and to eliminate or diminish the in�uence of all
The myth of the risk premium
factors that prevent or endanger their success (case-probable risks). They do this
most notably by making suitable arrangements for the physical production of the
goods and services from which they intend to derive revenue: they must hire the
right executives, choose the right location, determine the appropriate amount of
capital to be dedicated to this and that venture, etc. But they also try to stimulate
all factors that are likely to increase their sales, and to eliminate or reduce the
The bottom-line is that the very nature of entrepreneurial action is to eliminate
the in�uence of case-probable risks and to bring into play the in�uence of desired
factors. One might say that entrepreneurship tends to reduce the known (case-
probable) risks and multiply the known factors of success. However, as we have
seen, the risks are not reduced in the sense of a reduction of the stochastic prob-
ability of their impact on the overall result; they are reduced in the sense that fac-
tors that have a known negative impact will be eliminated. For example, a tomato
planter might reduce the risk of insuf�cient irrigation and insuf�cient temperature
It is clear that not all risks can and will be eliminated. But the known negative
in�uences will tend to be cut back to the extent that marginal cost is still covered
by marginal expected revenue. And the presently unknown negative in�uences
are researched as far as possible, within the same constraints of reasonable cost.
Let us also notice that there is no one-to-one relation between the incomplete-
ness of knowledge and the value of the activity (or the asset) about which one is
ignorant. One may know a lot about an asset, but these bits of information might
all point in a negative direction. And, inversely, one might know just a few things
about an asset, but these bits of information might make it appear very desirable.
Finally, there is no way to value the unknown. Brown might be conscientious of
the fact that, despite all his research and business intelligence, there might still be
factors around that bear on the success of his endeavour and which he has not con-
sidered at all. But he cannot evaluate these factors, precisely because he ignores
them. His evaluations can only be based on the factors that he knows to have a
positive or a negative impact on his project. He will then rely on the positive fac-
tors and try to contain the negative ones as far as possible. That is what production
is all about. But he cannot evaluate what he does not know.

The mirage of the risk premium
Let us now apply our realist approach to study the question of the risk premium
within gross interest rates. Above we stated that the conventional approach is
based on the idea that the observable interest rates are an arithmetic sum of differ-
ent components, each of which can be determined in separation from the others.
As far as the component of the risk premium is concerned, this approach involves
three related challenges: (1) to de�ne a risk, (2) to measure that risk, and (3) to
explain how that risk so measured relates to demand and supply schedules. It is
fair to say that there is today no general agreement on how these problems should
best be solved.
In what follows, we will brie�y discuss the most widespread
Jörg Guido Hülsmann
approach, at the heart of which is the capital-asset pricing model, and then con-
45), “the
market will price risky assets in such a way that the return on a
risky asset will be equal to the risk-free rate of return plus a fraction (or multiple)
of the whole market risk premium.” This idea is commonly expressed in equa-
The capital-asset
pricing model is ingenious because it explains and justi�es
the standard view, which holds that observable returns on capital are the sum
of different components. It responds to all three challenges that we highlighted
above. It de�nes risk, it measures risk, and it explains how risk in�uences the
demand and supply of any asset. Indeed, each market participant would be ready
to pay a price for any asset A
such that its return R
would correspond to equation
[2]. And any current owner of A
However, the problems of this approach are numerous and well-known. Let
us just mention two very important ones. First, the de�nition of risk as market
risk might be plausible for certain groups of investors, for example, for those
with very short-term time horizons or for those who are heavily indebted. It is
implausible for all others. Long-term investors are likely to consider market risk
a boon rather than a bane.
Second, the linear relationship between market risk
and return (between the risk premium and the gross interest rate) is premised on
the notion that the economy has reached general equilibrium and that all market
participants share the same perception of returns and risks. This assumption is
highly problematic in that it implies a completely different risk structure than in
any real-world market economy. Asset-price �uctuations would be much lower
than in the real world, but liquidity risk might possibly be greater. People who
agree in their assessments of risks and returns are much less likely to exchange
Moreover, and most importantly, if the explanation of risk premia concerns an
equilibrium world, then at best it could “shed some light” on risk premia as they
The myth of the risk premium
exist in the real world, in which disequilibrium is permanent and ubiquitous. But
what explains the rest? In other words, while the model does
explain how
case-probable knowledge affects demand and supply schedules in the real world,
there can be no doubt that the real world exists and that market participants buy
and sell economic goods based on imperfect knowledge. What, then, are the real
B) A
– its
marginal value prod-
uct (MVP
For example, the investor might overestimate it relative to another
known contributing factor Y. In this case, MVP
would be too high and MVP
low. The investor also runs a risk because he might ignore the in�uence of a fac-
However, these risks are irrelevant for the discounting of the MVP of X
. What-
ever the investor’s subjective assessment of MVP
, he will discount it by the
same personal discount rate. If he estimates MVP
to be relatively high, then the
associated DMVP
will also be relatively high. If he judges MVP
to be relatively
low, then the same follows for DMVP
. In short, whatever the case-probable risks
that might come into play, they do not
per se
affect the subject value difference
between any factor of production and its contribution to the monetary value of the
product. Whatever case-probable risks the investor confronts, they do not show up
in a risk premium within the discount rate that the investor uses to assess the pre-
sent value of future cash �ows. Again, this is implied in the very nature of the type
of risk that an investor confronts, which is case-probable risk. A
risk premium,
because it would be included in the discount rate, would compensate
him for risks
as far as the causal connection
between a factor X and the monetary
Jörg Guido Hülsmann
value of the product of X is concerned. But in his subjective judgement this con-
The implication of the foregoing considerations is that differences in observ-
able interest rates cannot be explained as compensations for risk. So how can they
be explained? In what follows, we argue that they result from different subjec-
tive appreciations of available investment opportunities. Different people value
different causes differently. Some may think that cause A
will entail
a rise in
the market price of asset X, others might believe that A
brings this
about only in
conjunction with B. Still others might consider that C is the relevant factor, etc.
As a consequence, the demand schedule for X will be composed of a continuum
of individual demands with very different motivations. If all popular causes are
present, the demand will be at its maximum. If C is absent, the demand will be
lower, and lower still if B, too, is not given.
Now, from a microeconomic point of view, this might create the impression
that the price of X depends on a risk premium. If there are a lot of favourable
circumstances, then the risk of owning X seems to be relatively low, and this goes
in hand with a relatively high price. And inversely, if there are but a few such
favourable circumstances, then the risk of owning X seems to be relatively high,
and this goes in hand with a relatively low price. The difference in the two prices
But this interpretation is unwarranted. It is not the case that all market par-
ticipants evaluate the pros and cons of an asset X in the same way. It is not the
case that they share the same view about which negative factors could impact its
yield, and how much they might affect its market price. Each of the persons who
wishes to own X is motivated by different considerations. There is no agreement
between the different contenders. What appears to be a key factor of success in
the eyes of one of them is irrelevant in the eyes of the other. They do not agree
on the risks involved. Each of them evaluates X in light of the circumstances that
he himself, and maybe only himself, considers to be relevant. And each is willing
to make a bid for X in light of his personal assessment if and when the relevant
No entrepreneur gambles with his capital.
Each entrepreneur, when he buys
an asset, is convinced that this purchase will permit him to preserve and increase
his capital. Otherwise, he simply would not buy it. The reason why some are will-
ing to bid a higher price, and others a lower one, for the same asset is that they
have different visions of what the relevant investment alternatives are. Suppose
Mr. Black is convinced he can make a 10
percent return
on that capital by buying
the asset X at the current market price. This conviction implies a maximum price
he would be willing to pay for the alternative asset Y. Things would be very dif-
ferent for someone with a different vision of his own investment opportunities.
The ordinary difference in price between a 2-year bond and a 5-year bond
results from the fact that more people are willing to buy the former at relatively
high prices. It does not result from any “discounting” of the 5-year bond. There
is in this respect not the least difference between assets that allegedly are subject
to a risk premium and the economic goods that are not. Always and everywhere,
The myth of the risk premium
different goods attract different people who wish to own them at different prices,
most notably because they see different alternative results from owning them.
But is it not true that some people discount the price they would otherwise be
willing to pay by a risk component? Johnson would pay 100 dollars for a share
of XY stock if it were free of risk. But because there are risks associated with
it, he is only willing to bid 80 dollars. What is wrong with this representation of
the investment process? There is no way of knowing what the price of the share
would be if it were risk-free. The share can only be bought as is. It is impossible
to compare an actual market price to something that does not exist. Therefore,
even if it were true that Johnson fancies himself to be discounting the price with
reference to some number that he has made up, or that somebody else has made
up for him, this is irrelevant for the economic analysis of what he does. From the
economic point of view, there are only two relevant questions. The �rst is whether
Smith is convinced that buying an XY share at this or that price will suit his ends,
such as earn a pro�t, preserve his capital, control the company, etc. The second
concerns the available alternatives. If Johnson were not convinced of being suc-

Risks as costs
The idea that risks are part and parcel of business costs is fundamental for busi-
ness accounting and reporting. In particular, risk may affect various business costs
through the risk component of the internal discount rate. The higher
that risk component, the lower is the demand for factors of production, and thus
cost expenditure. At the same time, risk may also affect business costs
through the income statement. Firms may create various provisions that enter
the annual income statement as business costs. For example, bad-debt provisions
can be made for the probable default of debtors and contingent-liability provi-
sions for potential legal costs, etc. All of these provisions enter the annual income
statement as business costs. Similarly, in setting up its balance sheet, a �rm may
estimate the monetary value of its �xed assets by discounting the associated cash
�ows using an interest rate that may include various risk components. Changes of
valuations resulting from changes in the risk component enter the income state-
ment, too, under the rubric of “other comprehensive income,” and thereby enter
From a microeconomic point of view, it is acceptable to think that risks are
a “given” part of the �rm’s environment. However, economic analysis needs to
overcome this narrow perspective and explain the ultimate causes of the realities
that are given to the immediate experience of any single market participant. This
is what Böhm-Bawerk achieved with respect to the general theory of costs of
production. He explained how factor prices ultimately result from the prices of
their least important products. They are formed by discounting the prices of their
As far as risk is concerned, we need to emphasise again that probabilities lead
an epistemic existence, not an ontic one; and that risks are subjective elements of
Jörg Guido Hülsmann
human action, not qualities of its objective environment. Provisions for risk are a
misnomer. In fact, such provisions are made
the investment. They are an
acknowledgment of an investment error. Errors are without doubt costly; thus
it is pertinent to account for them in the income statement as soon as the error is
discovered. But errors are not “risky” at all. Risk is by its very nature a quality of
judgements about the (imperfectly known) world, not an
ex post
Similarly, as we have argued, the risk component in the gross interest rate is a sort
of optical illusion. Different prices for different assets result from the fact that buy
ers and sellers appreciate them subjectively. From a microeconomic perspective, the
implied differences in yield might be called risk premia. And one might use such
premia in computations with an internal interest rate, to distinguish more interesting
ventures from less interesting ones. But this does not alter the fact that the idea of a
Despite the importance of risk in business accounting and in virtually all other
areas of economic life, the economic analysis of risk is not in a satisfactory state.
The reason is that economists have too light-heartedly adopted analytical tools
forged by mathematicians and statisticians. Most present-day economists con-
sider observable interest rates to be the arithmetic sum of a pure interest rate, a
risk premium, and a price premium, each of which can be determined in separa-
tion from the others. We have argued that this conception is problematic in that it
As a solution, we present an alternative realist approach to the study of risk,
based on Ludwig von Mises’s distinction between case probability and class prob-
ability, and on the principle of subjective value. This approach leads us to appreci-
ate that, in a free-market setting, known risks that pertain to business tend to be
eliminated through entrepreneurial activity. It also leads us to the conclusion that
case-probable risks pertaining to business decisions are not re�ected in the dis-
counting process
per se
, but rather in the assessment of the marginal value product
of factors of production. The implication is that differences in observable interest
rates cannot be explained as compensations for risk, but result from different sub-
Chapter 5
of this volume is a reprint of an article published in the
, vol. 18, no. 4 (Winter 2015): 487–561.
This was
recognised early on by Bernoulli, who stressed that “no valid measurement
of the value of a risk can be obtained without consideration being given to its
The economic
point of view also makes it irrelevant to consider the (psychological)
phenomenon that certain beliefs are strongly held, while others are not. For an econo-
mist, strongly-held beliefs are precisely those that are valued more than others. It is not
The myth of the risk premium
This crucial consideration is prominent in some of the recent literature on “one-reason
We owe this simple, crucial insight to Dr
“The spread
between the interest rates on bonds with default risk and default-free
bonds, both of the same maturity, called the risk premium, indicates how much addi-
tional interest people must earn to be willing to hold that risky bond” (Mishkin, 2013,
161). This
is also the usual way of presenting the risk premium among Austrian
school economists. See Mises (1940, pp.
490–508; 1998
[1949], pp.
543–556) and
(1993 [1962], p.
497). The
latter states “In the real world there is an addi-
entrepreneurial (or “risk”) component
, which
to the interest rate in par-
ticularly risky ventures, and in accordance with the degree of risk” (Rothbard, 1993
Nature of Capital and Income
, Fisher (1906, p.
279) distinguishes
between three
types of values: riskless, mathematical, and commercial. He proposes that to obtain the
mathematical value of an interest rate, “we simply add to the riskless value the value of
Also, notice
that the application of this conception of risk (as a standard deviation) puts
That is,
unless he gambles, in which case he would by de�nition not invest (see Mises,
On the
basic mechanisms of the determination of marginal value products (MVPs)
and discounted marginal value products (DMVPs), see Rothbard (1993 [1962],
Again, this
is a tautological statement. Somebody who gambles with his savings is by
de�nition not an entrepreneur, but a gambler.
in French accounting practice, �rms may establish provisions for risks and
costs such as anticipated law suits (provision pour risques et charges) and provi
sions for doubtful claims (provisions pour créances douteuses), as well as provi
sions for exceptional use of �xed assets (provisions exceptionnelles). Firms may
assess the value of their �xed assets as the present value of discounted expected
future cash �ows, based on the CAPM model. The standard procedure is to use a
discount rate that results from the addition of three components: a riskless inter
est rate, a market-risk premium, and a premium for �rm-speci�c risks (Ferdjallah-
He thereby
con�rmed the Ricardian insight that the prices of factors were intimately
tied up with the price of capital. Ricardo held that factor revenues and capital revenues
were caught up in a zero-sum game. The higher the return on capital, the lower must
be the aggregate revenue of factors of production, and the other way round. Böhm-
Bawerk added nuance to this result by taking account of the time structure of pro-
duction. In light of his analysis, we can see that there is not necessarily an inverse
relationship between the return on capital and aggregate factor income. For example,
if, as a consequence of a decrease in the return on capital, the structure of production
lengthens, and if the factors are now employed more than before in the higher stages
of production (where their prices are discounted more than in the lower stages), then
aggregate factor incomes are likely to shrink even though the return on capital dimin-
Bernoulli, D. 1954 [1788]. “Exposition of a New Theory of the Measurement of Risk.”
Jörg Guido Hülsmann
Ferdjallah-Cherel, E. (ed.). 2014.
Mission d’évaluation: Guide Pratique
. 2nd ed. Paris:
Fisher, I. 1906.
Nature of Capital and Income
. New York: Palgrave Macmillan.
Holton, G. A. 2004. “De�ning Risk.”
Financial Analysts Journal
Howells, P., and K. Bain. 2000.
Financial Markets and Institutions
. 3rd ed. New York:
Katsikopoulos, K. K., and G. Gigerenzer. 2008. “One-Reason Decision-Making: Modeling
Violations of Expected Utility Theory.”
Knight, F. H. 1971 [1921].
Risk, Uncertainty, and Pro�t
. Chicago: University of Chicago
. Auburn, AL: Mises Institute.
. Auburn, AL: Mises Institute.
Mishkin, F. S. 2013.
Money, Banking, and Financial Markets
. 10th ed. New York: Pearson.
Calcul des probabilités
. 2nd ed. Paris: Gauthier-Villars.
Rosa, E. A. 1998. “Metatheoretical Foundations for Post-Normal Risk.”
Man, Economy, and State
. Auburn, AL: Mises Institute.
Sharpe, W. F. 1964. “A
Theory of
Market Equilibrium Under Conditions of Risk.”
Although similarities exist between various elements of the Austrian and neo
classical theories of the market economy, on examination they often prove to be
super�cial, sometimes obscuring signi�cant underlying differences between the
two approaches. In other cases, elements of one approach make manifest dif
ferences it has with the other. Time is one such element. As Ludwig von Mises
Economists often erred in neglecting the element of time. Take for instance
the controversy concerning the effects of changes in the quantity of money.
Some people were only concerned with its long run effects, i.e., with the �nal
prices and the �nal state of rest. Others saw only the short-run effects, i.e.,
the prices of the instant following the change in the data. Both were mistaken
and their conclusions were consequently vitiated. Many more examples of
paper contends that the theory of cost is another example of the mistreat-
ment of time in economic analysis. While great strides have been taken in the
causal-realist tradition toward integrating time into cost analysis, in particular the
recognition of the principle of discounting in determining the prices of factors of
production, more insights concerning time have lain dormant. As Murray Roth-
bard wrote in an assessment of a book by Israel Kirzner,
The abstention from money is unfortunate but not fatal, but the abstention
from time and capital analysis
, and this cannot be remedied by an appendix
that Kirzner promises us on time. Problems of time, capital, [and] interest
(Rothbard, 2011, p.
built the conceptual structure of causal-realist economic theory in his
magnum opus
. In his own treatise on economics,
Man, Economy,
Time and the theory of cost
Jeffrey M. Herbener
Jeffrey M. Herbener
, Rothbard developed the analytical apparatus of Misesian economic
theory, and through his analytics, wove production theory into the structure con-
tained within
. While embracing the device of production curves,
however, Rothbard stopped short of adopting the analytics of cost curves. He
As an explanation of the pricing of factors and the allocation of output it is
obvious that cost curves add nothing new to the discussion in terms of mar-
ginal productivity. At best, the two are reversible
. But,
in addition, the shift
cited, in particular, the error in neoclassical theory of assuming �xed
prices for the factors of production in constructing cost curves from production
relationships, while production itself must occur over a de�nite time period in
which factor prices can change. In the Austrian construction, however, such
ceteris paribus
assumptions can be placed within a dynamic analysis in actual
time, moment to moment. This technique is used, for example, in constructing
a demand curve. The downward-to-the-right shape of the demand curve follows
ceteris paribus
, holding the buyer’s preference scale constant and considering the
quantity demanded by the buyer at prices both higher and lower than the actual
price at which he purchased an actual amount of the good. The position of the
demand curve is given only at the moment the buyer makes his purchase, which
becomes the relevant moment for economic analysis. As time passes, the demand
curve can shift outward or inward as the buyer’s preferences and circumstances
change. A
treatment can be given to production decisions and cost curves.
Marginal costs and production decisions
Rothbard accepted the standard account of production relationships (Rothbard,
2009, pp.
453–478). The
marginal physical product (MPP) and average physical
product (APP) to increasing amounts of a variable factor used in combination
with a given set of complementary factors generate ∩-shaped curves because of
�rst increasing and then diminishing returns. An entrepreneur’s demand for the
variable factor depends on its MPP and the marginal revenue generated by the sale
of the increment of output produced by the variable factor in the entrepreneur’s
line of production or, in short, its marginal revenue product (MRP). The price of
the variable factor is determined, in turn, at the point where the demand by all
entrepreneurs and supply by all factor owners clear the market for the factor of
production. An entrepreneur’s production decision in any line of production is
made by comparing all factor prices from buying inputs, i.e., costs, with the price
As noted in the quote above, Rothbard pointed out that the standard short-
run cost curves of neoclassical economics add nothing to this argument concern-
ing production decisions.
They are merely the reverse of production curves,
Time and the theory of cost
constructed by monetizing the production curves via multiplication by given
factor prices. The ∩-shape of the per-input-unit production curves generates the
U-shape of the per-output-unit cost curves. As illustrated in Figure
ginal cost (MC) curve is downward-sloping-to-the-right over the range of increas-
ing returns and upward-sloping-to-the-right over the range of decreasing returns.
When MC stays below average variable costs (AVC) and then average total costs
(ATC), they decline. When MC moves above AVC and then ATC, they rise.
The assumption of given factor prices necessary to generate such a cost struc-
ture, however, is untenable in a causal-realist framework, which takes account of
actual time. As Rothbard wrote:
The mathematical bent toward replacing the concepts of cause and effect by
mutual determination has contributed to the willingness to engage in circular
a neoclassical model of general equilibrium, all prices are instantaneously
determined as the solution of the model’s system of equations. The timeless char
acter of the model neatly sidesteps the logical inconsistency, noted by Rothbard
in the quote above, of the model’s implication that the price of output is a deter
minate of the prices of inputs (through their MRPs) and that the prices of inputs
(through costs of production) are a determinate of the price of output. In the
real world, input prices incurred to produce a given output arise before the price
for that particular output arises. An entrepreneur’s decision to produce output,
and concomitant decision to demand inputs, comes before his decision to sell
that particular output to buyers who will be demanding it at that later point in
time when the sale is offered. Production takes time. Input prices and output
prices, therefore, cannot be brought into consistency with each other by assum
ing they occur synchronously. In the real world, the element that binds input
prices now together systematically with output prices in the future is entrepre
Figure 6.1

Jeffrey M. Herbener
Moreover, in the neoclassical model of perfect competition, an entrepreneur
considers prices for factors of production given because his decision to alter his
demand for inputs has no effect on the overall market demand, and, therefore, on
the prices of the factors of production. Whether or not such a condition exists in
actual markets, the model omits capitalists who invest in various lines of produc-
tion by purchasing the assets or claims to assets used in production. In the real
world, the prices of the assets owned and used by the entrepreneur will vary as
conditions change regardless of what happens to prices for inputs purchased by
the entrepreneur. Investors-at-large will increase their bids for such assets when
pro�t arises and reduce their bids for such assets when losses ensue. The conse-
quent change in asset prices will change “�xed costs” and therefore shift the cost
As Rothbard emphasized, pro�ts and losses will be capitalized into the prices of
assets speci�c to an entrepreneur’s enterprise. Investors-at-large bring this about
by bidding more heavily for assets in lines of production anticipated to be more
pro�table and bidding less heavily for assets in lines of production anticipated to
be less pro�table. Ignoring this variability of the prices of capital capacity leads to
Further, the result of abstention from capital leads to all the crucial errors
of the “cost curve” analysis. For example
. that
a �rm will invest funds in
production up to the point where “marginal revenue” equals “marginal cost.”
(2011, pp.
from ignoring the variability of asset prices in calculating “marginal
cost,” the neoclassical approach errs by assuming that the technical units of the
good in the functions underlying marginal revenue and marginal cost are the only
relevant units for decision-making by the entrepreneur. As Menger stressed, how-
ever, a person who is acting chooses the amount of the good he considers relevant
for the action at hand. For economic analysis, the chosen amount of the good is
the relevant marginal unit of the good. A
consumer who
wants to buy enough
gasoline to �ll the tank in his lawnmower, for example, may purchase 2½ gallons
For him, in this action, the unit of gasoline is 2½ gallons and the price to buy
that unit is $7.50. The price of gasoline posted on the gas pump of $3/gallon is
merely a technical standard chosen by the entrepreneur because consumers and
producers consider it convenient. In similar fashion, an entrepreneur chooses the
amount of the good to produce and sell along with the myriad of other aspects of
producing and selling he chooses. Because it has fundamental importance for his
action, the selected amount is the “unit” of the good, not the technical standard.
Mises refers to the period of production or what might be called the production
run as relevant for assessing the “marginal unit” of the good in production. In
considering the pro�tability of producing and selling the iPhone SE, for example,
Apple Inc. may have considered a period of production of several months with a
Time and the theory of cost
In making this decision, the marginal revenue anticipated by the entrepreneurs
at Apple must have exceeded the marginal cost they anticipated by an amount
suf�cient to justify Apple’s investment in this line of production. Over the time
span of producing and selling the 10
million phones,
marginal revenue could vary
due to larger or smaller demand than anticipated, and the marginal cost could vary
due to higher or lower factor prices than anticipated. Marginal costs are not suf-
fered merely for “variable” or purchased inputs, but also for the prices of “�xed”
or owned inputs over the period of production. The existence of markets for all
factors of production, separately or in combinations, implies that the opportunity
cost of production must include both inputs acquired by an entrepreneur through
purchases in the factor markets and inputs retained by an entrepreneur instead
As Rothbard argues:
[T]here is no one simple, determinate “marginal cost,” because, as we have
. there
is no one identi�able “short run” period, such as is assumed
by current theory. The �rm faces a gamut of variable periods of time for the
investment and use of factors, and its pricing and output decisions depend on
illustrates the analysis suggested by Rothbard. The revenue and cost
anticipated for each production-run quantity, and corresponding period of produc-
tion, are shown as the vertical distances R and C, respectively. The entrepreneur
will select the economizing combination of inputs for each level of output produc-
tion. The marginal costs of production (C
, C
, or C
) vary with each production
run (Q
, Q
, and Q
) and include the market prices of all inputs used in production,
Figure 6.2

Marginal unit of output in production
Jeffrey M. Herbener
both those purchased and those owned by the entrepreneur. In some cases, an
entrepreneur may choose the same con�guration of assets and increase produc-
tion by purchasing more materials and labor. In other cases, an entrepreneur may
choose a different con�guration of assets as well as amounts of materials and
labor when increasing production. Q
may represent the production and sale of
million iPhone
SEs produced in a factory in China, for example, while Q
represent 4
million iPhones
produced in a factory in the United States, and Q
may represent 16
million iPhones
produced in three different factories in China
and Taiwan. Of course, making such changes in production processes takes time,
which introduces other considerations for an entrepreneur in selecting an input
combination to produce a given amount of output. But as long as all factors of
production are saleable, their use by an entrepreneur in his line of production,
Some factors are best used in a certain quantity over a certain range of output,
while others yield best results over other ranges of output. The result is not a
dichotomy into “�xed” and “variable” costs, but a condition of many degrees
analytical apparatus need not, however, be arrested at this step. It is possible
to show the implications of the analysis in terms of the technical units of the fac-
tors of production instead of merely for the production run as a whole. As noted
above, the construction of demand curves makes such a transition. Each buyer of
gasoline, for example, has a personal marginal unit of gasoline, i.e., an amount
he anticipates will be best suited to the attainment of his end. And yet, demand
and supply analysis is conducted in terms of the technical unit, e.g.,
gallons, in
the price is denominated in the market. The transition is accomplished by
conjecturing what must logically happen, given diminishing marginal utility, to
the quantity demanded of a good at lower and at higher prices. In similar fashion,
it is possible to construct revenue curves and cost curves per-technical-unit as
long as the marginal adjustment used as the basis of per-technical-unit revenue
and cost computations is the entire production run. For each asset combination
chosen by the entrepreneur as best suited for a given production run, a cost curve
can be constructed by conjecturing what must logically happen, given increasing
and diminishing returns, to per-technical-unit costs at lower and higher levels of
production. Given the resulting cost structure, then, corresponding shifts in the
In contrast to this causal-realist analysis, the neoclassical treatment of long-run
costs displays an array of different production techniques associated with each
short-run production function. With input prices given, the production relation-
ships generate a long-run average cost curve (LRAC) that envelops the set of
short-run cost curves, as illustrated in Figure
The U-shape of the long-run
cost curves comes from the assumption of initial economies, which pushes down
long-run marginal cost (LMC) and eventually diseconomies of scale, which
Time and the theory of cost
pushes up LMC. As with their short-run cost analysis, neoclassical economists
can reconcile the assumption of given factor prices across the entire array of pro-
duction possibilities, capital capacity embodying various technologies and labor
and materials, with the further assumption of full, instantaneous adjustment of
entrepreneurs to the production technique which renders minimum LRAC. Even
if this procedure sidesteps the problem of different prices for the factors of pro-
duction purchased by entrepreneurs at each level of production for the various
production techniques, it cannot account for different prices for assets owned by
entrepreneurs across the various production techniques. As illustrated in the corn
production example outlined below, a variety of production techniques for a good
can coexist as long as the greater physical productivity of more heavily capital-
ized production techniques is balanced by higher asset prices compared to lower
asset prices for production techniques of smaller physical productivity. Such a
result depends, in turn, on the existence of a spectrum of speci�city of assets
across the different production techniques, a condition which is not uncommon
in the real economy. For example, older, less physically productive techniques
of steel production compete against newly introduced, more physically produc-
tive techniques as investors increase demand for the new asset con�guration and
decrease demand for the old con�guration. The value of the superior productivity
of the new assets will be capitalized into a higher price of the new asset combi-
nation, which will raise the structure of costs of production. And the decreased
investor demand for the old, inferior-productivity con�guration will lower its
capital value, and the lower price for the old asset con�guration will lower the
Time: discounting and costs
Instead of the neoclassical distinction between short-run and long-run costs,
which is based solely on the technical circumstances of production, Rothbard
Figure 6.3

Jeffrey M. Herbener
suggested two relevant time periods for production analysis: the immediate run
and the Evenly Rotating Economy (ERE). As he explains,
[T]here are in production theory two important and interesting concepts
involving periods of time. One is what we may call the “immediate run”
the market
prices of commodities and factors on the basis of given stocks and
speculative demands and given consumer valuations. The other important
concept is that of the “�nal price,” or the long-run equilibrium price, i.e., the
question remains: what is the dynamic analysis that links the two together?
The immediate-run analysis is depicted in Figure
At a moment in time, the
total stock (TS) of a good possessed by people is given. The price achieves the
level at which the total demand (TD) people have to possess the good is satis�ed
by the total stock of the good. Some people may desire to enlarge their stock,
which they do by purchasing some of the good from others who desire to reduce
their stock, which they do by selling some of the good. The former persons gener-
ate demand for the good and latter persons render supply.
The analysis of demand and supply depicted in Figure
views market
exchange from an equivalent, but different, perspective, as the analysis of total
demand and total stock. For a newly produced good, say the iPhone SE, consum-
ers have exchange demand (D) and the entrepreneur has reservation demand or
supply (S). The willingness of a consumer to purchase an iPhone SE, or any other
good, depends on his anticipation of the realization of the ends to which he will
put it and his anticipation of the sacri�ce of the best alternative end his purchase
entails. Likewise, an entrepreneur’s decision to supply or retain a good depends
upon his anticipation of the realization of the end attained by selling it and his
anticipation of the sacri�ce of the best alternative end, which can be categorized
as either personal use of the good or sale to another consumer at a later time,
Figure 6.4

Time and the theory of cost
which his sale now entails. The actual price of a good at any moment is based
upon speculation, which provides the link between prices now and prices in the
Immediate-run analysis of the price of a factor of production is depicted in
The actual wage of a labor service, say a tech worker hired by Apple
Inc., is also based upon speculation. The worker anticipates the realization of
the end attained by selling his labor services and his anticipation of alternatives,
categorized as either personal use or sale to another entrepreneur at a later time,
which his sale now entails. The entrepreneur anticipates the realization of the
end attained by earning the marginal revenue product of the labor service and
anticipation of
the sacri�ce from attainment of the best ends foregone. Like
the actual price of a good, the actual price of a factor of production at any moment
Immediate-run analysis highlights the fact that current prices of factors of pro-
duction do not directly correspond to current prices of goods. The prices of fac-
tors of production are bound to prices of goods by speculation, not by the logic
of models under the assumption of full, instantaneous adjustments. Moreover,
entrepreneurs must pay the costs to acquire or retain factors of production sooner
and receive the revenues from selling to customers later. Their demand for factors
of production, therefore, will discount the factor’s marginal revenue product.
ERE analysis preserves the insights of immediate-run analysis, namely, the
time sequence of production and the discounting of payments to factors of pro-
duction. Although both the neoclassical conception of general equilibrium and
the ERE abstract from uncertainty, the neoclassical construct contains neither the
time sequence of production nor the discounting of factor payments. As Rothbard
In the ERE, of course, all costs and investments will be adjusted, and irrevo-
cably incurred costs for all �rms will equal the price of the product minus
pure interest return to the capitalist-entrepreneurs, and also, as we shall see,
Supply – Entrepreneurs’ Opportunity Costs
Demand – Consumers’ Subjective
Personal Use
Figure 6.5

Jeffrey M. Herbener
minus the return to the “discounted marginal productivity of the owner,” a
factor which does not enter into the �rm’s money costs.
entrepreneurs pay DMRPs for factors of production, actual cost struc-
tures lie below the cost structures of neoclassical models. Rothbard uses this point
to rebut the neoclassical conclusion that an entrepreneur with monopoly power
would operate with excess capital capacity. As depicted in Figure
if the pro�t
maximizing output for an entrepreneur in a monopolistically competitive market
is Q
, then he will not select a cost structure tangent to demand at point A, as the
neoclassical model concludes. Instead, he will select the production technique
that minimizes average cost at point B. Rothbard maintains that this interest return
gap (P
– AC
) between the actual cost structure of an enterprise and demand for
Once we bring investment interest return into the picture, we see that the
whole elaborate cost-curve structure is totally faulty and should be tossed
Rothbard’s lead, however, cost-curve analysis can be reconstructed
on solid ground. Doing so sharpens the contrast between causal-realist and neo-
classical economic theories and permits further integration of fundamental insights
into economic analysis and, therefore, exposes more fully the cause-and-effect
relationships of the real economy. To achieve this goal, one need only follow-up
on Rothbard’s key point that costs adjust to prices of output through changes in
demand by capitalist-entrepreneurs. According to Rothbard,
Costs are not �xed by some Invisible Hand, but are determined precisely by
the total force of entrepreneurial demand for factors of production. Basically,
rkers’ Opportunity Costs
Demand – Entrepreneurs’ Judgments of DMRPs
Personal Use
Figure 6.6

Time and the theory of cost
Figure 6.7

as Böhm-Bawerk and the Austrians point out,
costs conform to prices
particular, prices of more-speci�c factors adjust through capitalization to
the price of the good they help produce. Consider the production of a good in
the Evenly Rotating Economy in which entrepreneurs have adopted a variety
of combinations of factors of production. For example, corn production by an
entrepreneur in eastern Nebraska would employ a larger land area and a more
capital-intensive process than corn production by an entrepreneur in western
Pennsylvania. And yet, in the ERE the price of each bushel of corn is the same as
every other bushel, the price of each unit of a less-speci�c factor of production,
like farmer labor or gasoline, is also the same, and the interest rate of return is
uniform across the various processes for producing corn. It follows that the prices
of more-speci�c factors of production, like land sites and specialized equipment
such as pivoting-irrigation systems used in Nebraska farming, must adjust to make
the average cost of production the same across the various production processes.
As depicted in Figure
the average cost curve for production in western Penn-
sylvania hits its minimum point at an output (Q
) lower than that for production in
eastern Nebraska (Q
). The greater physical productivity of Nebraska farm land
attracts capitalist-entrepreneurs who bid more for such land, generating a higher
s view,
[T]he various “costs,” i.e., prices of factors, determined by their various
DMVPs in alternative uses, are ultimately determined solely by consumers’
demand for all uses. It must not be forgotten, furthermore, that changes in
demand and selling price will change the prices and incomes of
in the same direction. The “cost curves” so fashionable in current
Jeffrey M. Herbener
economics assume �xed factor prices, thereby ignoring their variability, even
factors of production will be the locus of capitalization and
therefore of shifting cost curves as they adjust to changes in demand for outputs.
Then, in turn, as prices rise for the more-speci�c factors of production, the addi-
tional pro�t from producing them will attract the capitalist-entrepreneurs who are
doing so to increase their production, which necessitates increasing demand for
the factors of production used to produce them. Greater demand by consumers
for corn bread will increase the pro�t of corn production, which will increase the
pro�t of the production of pivoting-irrigation systems, and so on. The changing
pattern of demand will arc across the entire production structure in the economy,
bringing with it an economizing reallocation of resources throughout the stages of
production. The reallocation ceases when costs have adjusted to revenues in each
and every affected line of production.
To illustrate the impact on cost structures of capitalization, consider a �nal
state of rest analysis depicted in Figure
Starting in an equilibrium in the
economy in which the price for the good depicted, say taxi service in NYC, is
and the cost structure is AC
, suppose demand declines due to the entry of
Uber into the market. As a consequence, the price of taxi service declines to P
capitalist-entrepreneurs to
reduce their demands for inputs, which in turn
pushes their prices down. The downward adjustment of factor prices will array
according to the speci�city of the factors. Wages and prices for taxi cabs may
not decline at all, but the price of speci�c assets, such as taxi medallions, will
fall signi�cantly. In fact, the prices of speci�c factors will fall enough to push the
cost structure down to the point (AC
) at which investment in taxi service renders
the going interest rate of return, as it did in the original equilibrium. As investors
Figure 6.8

Production costs with different techniques in the ERE
Time and the theory of cost
shift demand away from taxi service and into other lines of capital investment,
capitalist-entrepreneurs increase the production of capital goods across the capital
structure whose prices are rising and reduce the production of capital goods across
the capital structure whose prices are falling. Eventually, the economy returns to
a new equilibrium in which cost structures have been brought into conformity
with the new pattern of prices which was set in motion by the switch of consumer
demand away from taxis and toward Uber.
In actual markets, this adjustment process is rarely, if ever, completed, because
the underlying causal factors are continuously changing. Mises emphasized that
The notion of change implies the notion of temporal sequence. A
rigid eter
nally immutable universe would be out of time, but it would be dead. The
concepts of time and change are inseparably linked together. Action aims at
change and is therefore in the temporal order.
in turn, implies uncertainty. “Every action refers to an unknown
future,” Mises wrote, and “It is in this sense always a risky speculation” (Mises,
1998 [1949], p.
106). Rothbard
in particular noted the relevance of speculation in
the face of uncertainty for prices. In evaluating Kirzner’s book, he asked, “Gen-
eral Comment for Book: why has there been no discussion of the important in�u-
ence of ‘speculation’ on price determination?” (Rothbard, 2011, p.
main role of speculation that Rothbard explains in his price theory con-
cerns the adjustment process in the market (Rothbard, 2009, pp.
130–137). Accu-
speculation hastens the movement of prices toward market-clearing and
makes both demand and supply more sensitive to changes in price. Mises, how-
ever, focused on the speculative character of action in the face of uncertainty, and
Figure 6.9

Jeffrey M. Herbener
thereby provided the insight essential to incorporating uncertainty into the theory
It might be helpful to summarize before taking the �nal step. The general
equilibrium analysis of neoclassical economics is timeless. The solution to the
system of equations generates prices and quantities traded for all goods and fac
tors of production simultaneously. The array of equilibrium prices and quanti
ties are known to the market participants who act in a manner that brings about
those equilibrium prices and quantities. The immediate-run analysis of the Aus
trians refers to moments of time in actual markets. The prices of outputs and
inputs at a moment in time cohere together in a system through speculative
anticipations. For example, input prices at the moment are consistent with the
output prices anticipated by capitalist-entrepreneurs who will be using inputs to
produce output, which they will sell at a later moment in time. The ERE analy
sis of Austrian economics incorporates the realistic time sequence embedded in
immediate-run analysis. Because uncertainty is absent in the ERE, input prices
at the moment are consistent with actual output prices that emerge in the future.
Unlike the neoclassical analysis, having a realistic time sequence permits the
integration of capitalization into cost analysis through the discounting of pay
ments to factors of production.
In reality, the passage of time involves not only
a time structure or sequence but uncertainty, an element that has not yet been
fully integrated into cost analysis, even the Austrian variant. The step necessary
depends on their
depicts contrasting views of the market for a labor service. The
far left panel shows the neoclassical equilibrium in which the wage of the labor
service conforms to its MRP. The middle panel illustrates the ERE, in which the
wage conforms to DMRP. The far right panel, then, adds uncertainty to the time
sequence incorporated into the ERE. In the real economy, entrepreneurial demand
for labor depends on the anticipated DMRP or ADMRP.
me & Uncertainty
Figure 6.10

Time and the theory of cost
Time: uncertainty and cost
Time implies that entrepreneurs discount factor payments because of the inevita-
ble time sequence of production, i.e., costs occur sooner and revenues occur later,
and earn an interest return from production as a consequence. However, time also
implies uncertainty and therefore that entrepreneurs base their factor payments
on anticipations of the future, and earn pro�t or suffer losses as a consequence of
their superior or inferior foresight, respectively. In Mises’s view,
Attempts to establish cost accounts on an “impartial” basis are doomed to
fail. Calculating cost is a mental tool of action, the purposive design to make
the best of the available means for an improvement of future conditions. It is
wo further implications follow from the fact that “calculating cost is a mental
tool of action.” First, cost will adjust as the accuracy of entrepreneurial foresight
improves. Increasingly accurate foresight by capitalist-entrepreneurs will result
in raising (lowering) the capital value of speci�c assets in pro�table (unpro�t-
able) lines of production. Second, the calculation of cost will be personal to each
entrepreneur. Those with superior foresight will anticipate the realizable price of
factors of production more accurately than entrepreneurs with inferior foresight.
Prices of speci�c assets, in particular, will adjust to the realizable price of output
more fully as entrepreneurs with less accurate foresight adjust their anticipation
To integrate these insights concerning uncertainty and entrepreneurial antici-
pations into cost analysis, consider �rst the immediate-run price analysis. Fig-
depicts the inclusion of anticipations into the demand for and supply of a
consumer good. Demand depends upon the anticipated satisfaction the good will
help bring about for the consumer compared to the anticipated satisfaction associ-
ated with the best alternative foregone. Buyers of the iPhone SE, for example, for-
mulate expectations about its serviceableness in the future as the grounds for their
purchases. Likewise, the entrepreneur supplies his good based on his anticipation
Supply – Entrepreneurs’ Anticipated Opportunity Cost
Demand – Consumers’ Anticipated Subjective
Figure 6.11

Jeffrey M. Herbener
of satisfaction associated with the revenue received compared to his anticipation
of the satisfaction associated with the best alternative foregone. Both sides of the
market are speculative and therefore re�ect differences in the speculative acumen
One important analytical implication of this emphasis on anticipations is a rein
forcement of the ef�cacy of Eugen von Böhm-Bawerk’s marginal-pairs analysis of
the market price of a good (Böhm-Bawerk, 1959 [1889], pp.
207–247). In
the neo
classical analysis, different initial marginal utilities between two traders are based
on nothing more than different initial endowments of the traded goods. Once all
mutually-advantageous trading has been accomplished, the marginal utility (MU)
received by each trader is the same. This conclusion is extended to a market with
numerous buyers and sellers. Each buyer continues to purchase additional units
of a good until he equates the MU of the last, i.e., least-valued, unit acquired
with the price, and each seller disposes of additional units of his endowment of
a good until he equates the MU of the last, i.e., least-valued, unit retained with
the price. Each buyer, then, is the marginal buyer and each seller is the marginal
seller. Böhm-Bawerk’s analysis, in contrast, highlights the underlying differences
between persons. Supra-marginal buyers enjoy a higher marginal utility, relative
to that of a given amount of money, than that received by marginal buyers, even
after all trading has occurred. A
buyer of the iPhone SE, for exam
ple, may have been willing to pay $500 (while unwilling to pay the market price
of $400 for a second iPhone), but the marginal buyer is the one willing to pay
the smallest premium above the market price of $400. In Böhm-Bawerk’s causal-
realist conception of the economy, the market does not bring every person into a
condition of equal MUs. In addition to the fundamental differences between per
sons in their preferences that Böhm-Bawerk notes, persons differ fundamentally
in the quality of their foresight. Some persons more accurately anticipate the reali
zation of the ends they are striving to achieve via exchange, while other persons
anticipate less accurately. Even if this were the only source of fundamental differ
ences among persons, a spectrum of valuation would still emerge. For example,
investors trading shares of stock in a company buy and sell on the basis of their
divergent foresight in anticipating the stock price in the future. Those possessing
superior foresight into the higher (lower) price in the future are able to �nd sellers
In similar fashion, the demand for and supply of factors of production are spec-
ulative. As illustrated in Figure
an entrepreneur demands a factor based on
his anticipation of the MRP that using its services will generate in the future.
and other factor owners, supplies his factor of production based on his
anticipation of the value of the income earned by doing so compared to the antici-
pation of the value of the best alternative uses to which it can be put. As with the
spectrum of consumers generating a downward sloping demand curve for a good,
the downward sloping demand curve for a factor of production is generated, in
part, by the range of foresight among entrepreneurs. Unlike the typical neoclassi-
cal model in which a single cost-minimizing method of production exists and is
selected by every entrepreneur who produces the product, the causal-realist view
Time and the theory of cost
of the process accepts the underlying differences in the physical productivities
among various land sites, among various capital goods, and among various per-
sons. As illustrated in the corn growing example above, in the ERE this condition
generates a spectrum of production methods, all of which have the same mini-
mum AC point in equilibrium but otherwise generate different average costs for
each level of output. This spectrum of costs is reinforced by the range of foresight
among entrepreneurs producing the good and the spectrum of foresight among
Finally, consider the issue of the dynamic adjustment process in the market.
There is nothing automatic or mechanical in the operation of the market.
The entrepreneurs, eager to earn pro�ts, appear as bidders at an auction, as it
were, in which the owners of the factors of production put up for sale land,
capital goods, and labor. The entrepreneurs are eager to outdo one another
by bidding higher prices than their rivals. Their offers are limited on the one
hand by their anticipation of future prices of the products and on the other by
the necessity to snatch the factors of production away from the hands of other
o illustrate the relevance of integrating anticipations into the adjustment pro-
cess of the market, consider the exposition in Figure
Suppose a current pro-
duction process generates revenue from the price P
with the accompanying cost
structure that renders AC
. At the same time a different entrepreneur imagines an
alternative production process with anticipated revenues from P
and associated
costs at AC
. The pro�t anticipated in this production process occurs because this
entrepreneur anticipates a capital value for the assets associated with the process
that is higher than the value of the same assets anticipated by other entrepreneurs.
This condition gives room for the entrepreneur with superior foresight to pay
lower prices for these assets, incur a lower cost structure, and earn the correspond-
ing pro�t. When other entrepreneurs come to anticipate the higher realizable value
rkers’ Anticipated Opportunity Cost
Demand – Entrepreneurs’ Anticipated DMRPs
Figure 6.12

Jeffrey M. Herbener
of the assets, then they too will increase their demand for them and consequently
bid up their prices. The higher prices of the more-speci�c assets will raise the cost
structure and eliminate the pro�t that has been earned by the entrepreneurs with
superior foresight. The additional pro�t is progressively capitalized into the prices
of the more-speci�c assets as more and more capitalist-entrepreneurs awaken to
the pro�tability of the new technique of production and bid more intensely for
such assets. The capitalized value of the �rst-mover advantage is earned by entre-
preneurs with superior foresight regardless of the speed of adjustment as the addi-
tional pro�t from increased production is progressively eliminated. The speed and
completeness of the adjustment of prices of more-speci�c assets and the concomi-
tant rise in cost structures depends on the improving accuracy of anticipations by
As capitalist-entrepreneurs draw resources away from less-valuable and into
more-valuable uses in this process, the disproportionately lower prices of more-
speci�c assets, compared to less-speci�c assets, in the shrinking lines of produc-
tion and disproportionately higher prices of the more-speci�c assets, compared to
less-speci�c assets, in the expanding lines of production result in larger losses and
pro�ts, respectively, for the production in the more-speci�c assets in shrinking
lines and expanding lines of production. Production is realigned throughout the
capital structure in accordance with the adjustment of cost structures conforming
to prices for each of the various assets. As with the shifting cost structures in the
lines of production that occur �rst, the speed and completeness of adjustment in
shifting cost structures in the other related lines depends on the improving accu-
Mises also understood the variation across entrepreneurs of the accuracy of
[The promoting and speculating entrepreneurs] are people intent upon pro�t-
ing by taking advantage of difference in prices. Quicker of apprehension and
Figure 6.13

Time and the theory of cost
farther-sighted than other men, they look around for sources of pro�t. They
buy where and when they deem prices too low, and they sell where and when
movement of prices of more-speci�c assets, the corresponding shifting of
cost structures, and the resulting earning of capital gains, depend therefore on the
spectrum of foresight of the various capitalist-entrepreneurs. For example, the
price of lithium mines has risen with the realization of greater demand for lithium
as an input in battery production. The capitalist-entrepreneurs with superior fore-
sight purchased lithium mines years ago at lower prices and have been earning
pro�ts from lithium production over the last several years. The growing recogni-
tion by less-astute entrepreneurs of the realizable price of the lithium mines has
led them to invest in these mines also. As more and more entrepreneurs awake to
the realizable price of the mines and bid for them as well, the pro�ts of lithium
production are capitalized into the price of the mines. Even as the pro�tability of
additional lithium production dwindles from the rising cost structure pushed up
by higher prices for lithium mines, the entrepreneurs with superior foresight reap
The passage of time has two main implications for production costs, only one
of which has been incorporated into the theory of cost in the causal-realist tradi-
tion. Production has a time structure as well as a physical structure. The stages of
production, from the extraction of natural resources to the construction of higher-
stage capital goods down through the construction of lower-stage capital goods
and �nally the making of consumer goods, are bound together not only by the
physical laws relevant to the production techniques adopted in each production
process but also by the time sequence of the stages of production. Neoclassical
models of general equilibrium incorporate the physical relationships among all of
the producer goods and their corresponding consumer goods. Such models con-
clude that each factor of production commands a price determined by its MRP.
The causal-realist theory of cost incorporates the discounting implied by the time
sequence of production. Each production process takes time from the payments
made to acquire inputs to the receipts received from the sale of outputs. Entre-
preneurs will, therefore, discount the MRP of each factor when paying to acquir-
ing its services. Recognition of the interest rate of return as the gap between an
entrepreneur’s revenue structure and cost structure invalidates some claims of the
neoclassical models and leads to insights beyond those gained from such models.
The other implication of the passage of time for production costs is the uncer-
tainty of the future. In making production decisions, an entrepreneur discounts the
future revenue to be generated by employing the services of a factor of produc-
tion, but can only anticipate its MRP and the appropriate discount to apply. Factor
prices are, therefore, determined by ADMRPs of entrepreneurs. Being speculative,
Jeffrey M. Herbener
prices of factors of production and the corresponding cost structures conform to
output prices regardless of the technical aspects of production relationships. The
speed and accuracy of the process of costs conforming to revenues depends on
the spectrum of foresight possessed by various entrepreneurs. Those with superior
foresight move earlier into what prove to be pro�table lines of production and
earn pro�ts which will then be capitalized into the prices of assets more speci�c
to that line of production as the less-astute entrepreneurs follow suit. Even when
the adjustment process reaches its climax and no additional pro�t can be earned
from a further expansion of production because cost structures have been pushed
up by rising prices for the more-speci�c assets used, the entrepreneurs with supe-
rior foresight will have earned capital gains by buying the more-speci�c assets
earlier in the process than less-astute entrepreneurs. Just as the incorporation of
entrepreneurial discounting of factor prices into cost theory has led to advances
in understanding the working of the market economy, even deeper understanding
For an example of the development and use of neoclassical cost curves, see Varian (2014).
On the
importance of discounting in the determination of prices of factors of production,
Block, Walter. 1990. “The DMVP-MVP Controversy: A
Review of
Austrian Eco
Böhm-Bawerk, Eugen von. 1959 [1889].
Positive Theory of Capital
. Trans. George D.
Huncke and Hans F. Sennholz. South Holland, IL: Libertarian Press.
Mises, Ludwig von. 1998 [1949].
Human Action: Scholar’s Edition
. Auburn, AL: Ludwig
Rothbard, Murray N. 2009.
1. “Comments on Israel M. Kirzner’s MS,
Market Theory and the Price Sys
Varian, Hal. 2014.
Intermediate Microeconomics
. 9th ed. New York: W.W. Norton.
Causal-realist price theory

The core idea of monopsony theory is the possibility of “pro�ts at the expense
of wages” (Machlup, 1952, p.
362) or
, more generally, pro�ts through an arti�-
cial lowering of money costs (expenses). Monopsony has long been a theoretical
curiosity in the academic literature, or at least, so it was before the 1990s. To be
sure, then as well as now, most microeconomics textbooks included some sections
dedicated to monopsony theory, showing how the price and the quantity sold of
a good or service
– usually
labor services
– are
determined when there is only
one buyer. Yet monopsony was seen more as an interesting thought experiment,
however impeccably carried out, than as an empirically relevant piece of theoriz-
ing (Friedman, 1976, p.
193). The
reason is that scholars had dif�culty identifying
The situation has changed in the past two decades as a new generation of schol-
ars (Boal and Ransom, 1997; Manning, 2011) has enthusiastically rediscovered
and built upon the original insights of Robinson (1933, pp.
The new
thinking holds that mere “frictions” on the labor markets are considered suf�cient
for “monopsony power” to emerge. Since in this case there is no longer any need
for a sole buyer or even a few buyers (“oligopsony”) to dominate the market in
order for a monopsony price to emerge, the distortions resulting from monopsony,
compared to perfect competition, are now taken to be a problem of far higher
Obviously monopsony theory has policy implications, such as helping make a
case for minimum wage laws, which are otherwise viewed by economists with the
utmost suspicion as far as their ef�cacy in ful�lling the goal of raising wage rates
without adverse effects on the labor force is concerned. Unsurprisingly, then, the
renewed legitimacy of monopsony theory in the academic �eld is spilling over
to the �eld of policy-making as well, and no less a group than the White House
Council of Economic Advisers has recently released an issue brief relying on
the new literature to promote various government interventions in labor markets
(Irwin, 2016; Council of Economic Advisers, 2016). Last but not least, if proof
was required that monopsony theorizing makes an impact outside the academic
world, Blair and Harrison (2010, pp.
xiii–14) have
shown that since the 1990s
antitrust court cases in the US have increasingly dealt with the behavior of buyers.
Monopsony theory revisited
“The most important idea in the analysis of monopsonistic labor markets,”
writes Manning (2003, p.
41), is
that “the labor supply to an individual �rm is
increasing in the wage paid so that the labor supply curve facing an individual
�rm is not in�nitely elastic as is assumed in perfect competition.” Reading the
recent literature, one would not guess that the relevance of this dichotomy between
monopsonistic and competitive markets has been challenged. In particular, long-
standing criticisms of the underlying imperfect-versus-perfect competition frame-
work from some “Austrian” scholars have generally been overlooked. Yet this
is of tremendous importance: if the distinction is arti�cial, perfect competition
cannot serve as a benchmark against which to contrast imperfect competition out-
older Austrians, such as Böhm-Bawerk (1962, pp.
162–167), and
fellow travelers
included room in their price theories for monopsony prices,
as a consequence of rejecting the perfect-versus-imperfect competition paradigm,
Mises (1949, pp.
380–381, 591–595)
and Rothbard (1962, pp.
717–718) ended
almost scrapping the entire idea. To them, monopsony theory essentially belongs
within the analysis of socialism (with one and only one possible employer, the
State), and believed that even here the concept of a monopsony price did not make
sense. This also seems to be the view adopted by contemporary Austrians (Bel-
However, I
want to
show here that Mises, Rothbard, and their followers may
have gone too far, and that their predecessors were onto something. In other
words, there was no need for the perfect-versus-imperfect competition framework
to produce a theory of monopsony prices in a market economy, in the same way
that Austrians did not need it to endorse a theory of monopoly prices (Mises,
1949, pp.
354–385; Rothbard,
1962, pp.
903–907). Expanding
on Méra (2010),
which explores the relationship between monopoly prices for products and the
prices of their factors of production, I
show that
the Austrian case against the
imperfect-versus-perfect competition framework and Eugen von Böhm-Bawerk’s
distinctively Austrian law of costs (Böhm-Bawerk, 1930, pp.
223–234) provide
foundations to revisit the issue and build what might be called a theory of
“monopoly price-gap,” with monopsony prices and monopoly prices arising
2 is
a brief overview of classic and new monopsony literature. Since
the Austrian case against their common framework remains largely overlooked,
and since it highlights some fundamental principles necessary to our more con-
structive task, section
3 restates
it to show how both the classic and new theo-
ries are built on questionable grounds. Section
4 outlines
Mises and Rothbard’s
alternative to the dominant paradigm and how it led them in particular to dismiss
monopsony-related concerns. Section
5 questions
part of their �ndings and lays
out the foundations of the theory of monopoly price-gap, based on the idea that
monopsony and monopoly
– de�ned
as grants of privilege in buying and selling
(Rothbard, 1962, pp.
– go
hand in hand in the context of production
decision-making. Sections
6 and
7 outline this theory for cases where the fac-
tors of production involved are speci�c to one production process and when they
Monopsony theory revisited
are not, respectively. Section
7 stresses
in particular that non-speci�city does not
shield factor owners from downward pressure on their prices, as a consequence of
the law of costs, despite mainstream and Austrian claims to the contrary. Finally,
the conclusion asks how this recasting of monopsony theory within a larger frame-
work affects the question of its empirical relevance, i.e. the potential pervasive-
ness of monopoly-monopsony distortions, and provides a non-exhaustive answer.

It is common to distinguish between “classic” and “new” monopsony theory
(Manning, 2008). Classic monopsony is the typical textbook version. In short, one
buyer faces many sellers, so that the quantity supplied to this “monopsonist” is an
increasing function of price. This stands in contrast with the perfectly competitive
case in which the supply schedule confronting the buyer is perfectly elastic (the
corresponding supply curve being perfectly �at) at the market equilibrium price.
The buyer is one among many such that any attempt on his part to lower the price
would result in the hiring of all units by its competitors at the current price. In
other words, the buyer is a “price-maker” in the �rst case and a “price-taker” in
Since the model typically deals with the market for a factor of production
usually labor
factors facing the demand of a pro�t-maximizing employer
– the
are as follows: the employer will hire or buy units of the factor until
the monetary value of the marginal product of the factor equals its marginal cost.
In perfect competition, this means that the employer hires up to the point where
the marginal revenue product equals its price. Under less than perfectly competi-
tive conditions, however, marginal cost is above the price for any quantity hired
but the smallest, since the price rises for all units hired as the quantity of factors
hired grows (except when dealing with the case of price discrimination). There-
fore, the pro�t maximizing point will be reached with a lower amount of factors
hired than would have occurred under perfect competition, and there will be a gap
between the value of the marginal product and the price of the factor, the marginal
revenue product being higher and the price being lower than they would be under
This gap is the extra earning or “pro�t”
per factor
hired for the employer that
he could not obtain under perfect competition, and also a measure of his “market
power.” In this setting, a legally mandated minimum price above this monop-
sony price can conceivably improve the situation of all sellers, as well as improve
ef�ciency in the resource allocation process, as long as it is not set above the
market-clearing price. Indeed, if the minimum wage happened to be the exact
market-clearing wage, the pro�t-maximizing quantity of factors hired would be
the same as under perfect competition (which usually serves as the benchmark for
Classic monopsony also refers to oligopsony, where the market structure is
such that a few buyers compete. The theory becomes more complicated because
one must take into account the strategic interactions between buyers, since the
supply schedules facing each buyer are affected by their competitor’s deci
sions. Nevertheless, the heart of the matter remains that the buyers face less-
than-perfectly-elastic supply schedules. In the same way that “monopoly” now
often refers to any situation in which a seller has some market power (that is,
faces a less-than-perfectly-elastic demand schedule), “monopsony” nowadays
often refers to any situation in which a buyer faces a less-than-perfectly-elastic-
supply schedule (Manning, 2008), including monopsony in the narrow sense,
as well as oligopsony or indeed any other possible con�guration within the
framework of imperfect (Robinson, 1933) or monopolistic competition (Cham
The “new monopsony” literature still concerns imperfect competition, forward
sloping supply curves, etc. The main difference with the older literature is, as
Manning (2008) puts it, that, “modern theories of monopsony do not generally
argue that employer market power over their workers derives from there being a
small number of employers. They tend to emphasize the role of frictions in the
labour market.” It is not surprising then that these authors �nd more room for
monopsonistic distortions in the economy than their predecessors. For as long as
workers do not immediately quit a �rm
en masse
whenever the employer “cuts
wages by one cent” (Manning, 2008), which we can con�dently assume never
happens, employers have some market power.
Manning (2003, p.
3) merely
follows this thread of thought consistently when
he asserts that the simple monopsony model is a better �rst approximation of
the situation for any �rm than the perfectly competitive model. However, he is
a leading author in developing a more sophisticated approach drawing on the
“search costs” literature in particular (Manning, 2006), pursuing Burdett and
Mortensen’s (1998) initial line of thought. In a nutshell, the idea is that looking
for a job is a costly endeavor, so that workers do not automatically �nd it bene�
cial to switch jobs when a better-paying alternative is available. This leaves room
for monopsonistic behavior on the part of employers. As Borjas (2013, p.
it: “In effect, mobility costs help generate an upward-sloping supply
As a result, a large part of the new monopsony literature is dedicated to esti-
mating the elasticity of supply of various groups of laborers and the extent to
which employers actually take advantage of their market power. Manning (2003,
2011) and Ashenfelter, Farber, and Ransom (2010) review most of the relevant
papers. In the latter’s view, they show that “the allocative problems associated
with monopsonistic exploitation are far from trivial” (Ashenfelter, Farber, and
Ransom, 2010, p.
209). The
new approach has its critics too, of course, such as
Kuhn (2004), who �nds Manning’s (2003) evidence for pervasive monopsony
We do not need to assess this controversy here, however, since what is of inter-
est in the literature for the purpose of this paper is the theoretical framework that
the terms of the debate take for granted. To our knowledge, only Bellante (2007,
22) has
weighed in on the current debate to reject those terms based on Roth-
bard’s case against the perfect-versus-imperfect competition dichotomy.
Monopsony theory revisited

It should be clear that the whole point of debating by how much the elasticity of
real world supply schedules deviates from perfection, as a proof for the existence
and as a measure of monopsony power, must assume that the dichotomy between
perfect and imperfect competition is relevant in the �rst place. The distinction
between a monopsony price and a competitive price for factors of production (or
for any good), as well as the distinction between a monopoly price and a competi-
tive price for their products, can be relevant only if competitive prices are within
the realm of possibility. It must be possible for supply of the schedules individual
buyers face (and the demand schedules individual sellers face) to be perfectly
elastic, or any discussion contrasting this situation with another one has no rela-
tionship to anything real. Can supply schedules (or demand schedules) ever be
As Rothbard suggests, it is dif�cult to see how this could ever be considered
a possibility were it not for the habit of thinking of action in terms of in�nitely
small steps as a consequence of the perceived need to present economics in terms
of mathematical models. For it is only in such a case that the decision of any
buyer can be considered to be without any impact whatsoever on the total market
demand and therefore on the market price. The idea of a perfectly elastic supply
schedule precisely assumes that the individual contribution of any buyer to total
demand can be neglected, since there are “many” buyers, so that the contribution
of each is very small. But the total quantity demanded of the factor on the market
at any price is made up of the quantities demanded by each individual buyer. Each
one is a contributor, and the individual decision to demand x instead of y units of
a factor cannot but affect the total, and therefore the equilibrium price.
In order for the contribution of each to be actually negligible, total demand
would have to be in�nite: n+x
n only
if n equals in�nity. But there never can
be such a thing as in�nite demand. Demand must be limited by the scarcity of
whatever is offered in exchange for the good under consideration. It follows that,
depending on the speci�cs of the case, a supply schedule may or may not be very
calculus may be convenient in mathematical modeling, but it can-
not have any economic meaning and counterpart in the real world. The problem,
as praxeologists such as Mises have argued, is that what we are interested in
– must
occur in terms of discrete steps. After all, altering the state of the
world, one way or another, presupposes that the individual actor uses means com-
bined in a de�nite quantitative relationship which he believes can help achieve
some end. However, size-less means can hardly produce anything but size-less
results, which for practical purposes means that nothing is produced at all. Only a
non-producing producer
– a
contradiction in terms
– can
make an in�nitesimally
small difference, which is to say, no practical difference at all.
Manning is correct then in assuming that supply schedules are normally imper-
fectly elastic, but not that this is because of the failure of the market to conform
to some realistic standard. Since the essential feature of perfect competition is
impossible, “deviations” from it, including monopsony and other “imperfec-
tions,” are not signs of a failure to conform to a normal or superior state of affairs.
Welfare comparisons based on this criterion imply a nirvana fallacy (Demsetz,
1969): since such perfection is beyond anybody’s reach, the standard of com-
parison is irrelevant.
It follows that equilibrium in no case can be characterized
as a situation in which the marginal revenue product and the price of the factor
of production are strictly equal. Instead, the marginal revenue product equals the
marginal cost of hiring the factor in all cases. And while no one can ever unilater-
ally decide what the price in a voluntary exchange will be, no one ever is a pure
price-taker, either.
In addition, Rothbard (1962, p.
718) also
argues that as long as anyone is
allowed to enter the market, pro�t can be arbitraged away, regardless of the degree
of elasticity of a supply schedule at any given time. As a result, there is no valid
criterion to differentiate monopsony or oligopsony prices from competitive prices
in the standard framework. If the idea of market power for a buyer is to be saved,

Monopsony and monopoly theory without
354–385) or
price (Rothbard, 1962, pp.
903–907). Rothbard’
s version of this theory states that
if the market demand schedule for a good is inelastic above its free-market price,
threats of punishment for entering the �eld might alter the demand schedules fac-
ing the remaining �rms from a greater than unitary elastic position to an inelastic
position, so that they might restrict their sales, thereby charging a monopoly price,
to earn a higher income.
Now, we might expect Mises and Rothbard to likewise
defend a monopsony price theory which does not rely on the perfect competition
benchmark. Yet this is not the case, because the symmetry between imperfect
competition in selling and in buying (monopoly and monopsony prices as a con-
sequence of less than perfectly elastic demand and supply schedules faced by
sellers and buyers, respectively) is a consequence of the imperfect-versus-perfect
Indeed, once we drop this framework and uses the elastic-versus-inelastic
demand schedule criterion for monopoly prices, no equivalent criterion appears
available on the other side of the monopoly coin, monopsony. Writes Mises,
Monopoly prices can emerge only from a monopoly of supply. A
demand does not bring about a market situation different from that under
not monopolized demand. The monopolistic buyer
– whether
he is an indi
vidual or a group of individuals acting in concert
– cannot
reap a speci�c
Monopsony theory revisited
gain corresponding to the monopoly gains of monopolistic sellers. If he
restricts demand, he will buy at a lower price. But then the quantity bought
is certainly not meant to imply that restrictions on buying do not make
any difference in the pricing process, or that such restrictions cannot bene�t the
remaining buyers in any way.
Instead, the claim is that there is no analytical
criterion allowing for a special category called “monopoly of demand price” or
“monopsony price.” This is because price and quantity vary in the same direction
along a particular supply schedule, so that a lower price is always associated with
lower expenses, whereas price and quantity vary in the opposite direction along
a demand schedule so that a higher price may be associated with higher or lower
expenses depending on the elasticity of demand. The old monopoly price theory
relies on this speci�c feature of a demand schedule, and no equivalent exists when
one discusses monopsony. In this sense, there can be no such a thing as a monop-
sony price. Consequently, there can be no monopsony gain corresponding to the
monopoly gain of monopolistic sellers. But the question remains then if a monop-
sony or oligopsony might take advantage of the exclusion of competitors to push
prices downward in order to improve its lot, even if the resulting ratios may not
In the case of labor factors, Rothbard, drawing on Mises (1949, pp.
It is
often alleged that the buyers of labor
– the
– have
some sort
of monopoly and earn a monopoly gain, and that therefore there is room for
unions to raise wage rates without injuring other laborers. However, such
a “monopsony” for the purchase of labor would have to encompass all the
entrepreneurs in the society. If it did not, then labor, a nonspeci�c factor,
other words, only socialism with its unique employer could provide the suf-
�cient conditions for lower wage rates.
Absent socialism, non-speci�city of the
factor appears to be what protects its owners from the prospect of monopsonistic
pressure when government interventions hamper the bidding of employers. The
implication seems to be that, for a market economy, lower prices for the factors
of production and the corresponding gains for the buyers are possible, but only
when violent threats are made against would-be buyers of the factor
– such
when licenses to buy are given to privileged buyers only
– and
when those factors
This point is not explicitly stated by Mises or Rothbard, and they make no
attempt to systematically analyze pricing under such conditions. Both pay lip
service to the limited possibility of lower prices under monopoly of demand
conditions under interventionism, but they only deny that this is possible for
non-speci�c factors, without discussing other cases. Their followers on this ques-
tion do the same (Block Barnett, 2009, p.
80; Bellante,
2007, p.
17). The
ing analysis both renders explicit what is implicit, in the case of speci�c factors
and, drawing on Méra (2010), reveals that the above conclusion regarding the
case of non-speci�c factors, including labor, is unwarranted, such that monopso-
nistic pressure on non-speci�c factor prices can be a widespread phenomenon. In
order to better grasp this point, we need �rst to take a
and establish in broad
terms what sort of relationship exists between monopoly and monopsony, if any,

Monopoly and monopsony: toward a theory
It is useful to introduce here a distinction between a “short run” and a “long run”
analysis, or better stated, between an analysis of choice with already-produced
goods on the one hand and an analysis of choice regarding production decisions
on the other. We begin with the �rst. This seems all the more relevant given Mises
and Rothbard’s way of approaching monopoly price theory. Indeed, even though
they pay some attention to factor allocation in their analysis of the conditions
for the emergence of monopoly prices, it is mainly an afterthought (Méra, 2015,
127–131). Their
analysis, as described above, tends to run in terms of already-
produced goods, ruling out production costs as a decisive consideration: the basic
requirement for a monopoly price to emerge is that demand schedules to remain-
ing sellers are made inelastic above the free-market price (or “competitive” price,
for Mises), allowing for the speci�c monopoly gain Mises alludes to in the above
quote. The “monopoly price” here is what Fetter (1915, pp.
80–81) referred
to as a
“crude monopoly price,” the quali�er emphasizing the lack of deliberations about
production in this setting (Fetter referred to “monopoly price” in the context of
If one contemplates monopsony through the same lens, the following picture
emerges. Absent any production decision, the only choice to be dealt with, as
far as a buyer is concerned, is the decision of a consumer. The �rst thing which
needs to be stressed is that although it is noticed in the literature that a monopson-
ist often is a monopolist, this is not the case here. The position of someone as a
consumer-monopsonist is unrelated to any monopoly position she might have as
a seller. Second, restricting the quantity she buys does not bring about any gain,
even if a lower price can thereby be paid. No rationale comes into play that would
explain why the buyer would be interested in restricting her exchanges in some
way similar to monopoly price theory. If this was advantageous to her, her demand
schedule would have been lower in the �rst place. Therefore, in perhaps a stronger
sense than Mises intended in the above quote (1949, p. 380), when one speaks of
a demand for consumption goods, “a monopoly of demand does not bring about
a market situation different from that under not monopolized demand” (Mises,
1949, p.
380). There
is no symmetry beyond the fact that competitors are excluded
Monopsony theory revisited
in both cases: monopoly can bring about monopoly prices, but no monopsony
price emerges from a monopsony situation.
So far, Mises’s case against the symmetric impact of monopoly and monopsony
is unobjectionable, but does it imply that there can never be such a thing as lower
factor prices speci�cally resulting from a monopsonist’s choices? Does this mean
most economists have taken the wrong track by endorsing one variant or another
of monopsony theory, including Mises’s predecessor, the master of Austrian price
theory, Eugen von Böhm-Bawerk? Is there a fatal �aw in the latter’s account of
“wage determination under employer’s monopoly,” for instance? In this discus-
sion, although he does not describe an economy-wide monopoly but rather a car-
tel of employers in an industry, he nevertheless states that:
The rate of wages would be �xed according to the general formula applying
to a purely sel�sh monopoly, already mentioned before in another connec-
tion: they would be �xed at that point which promises the largest returns,
after a careful consideration of all circumstances, and with due regard to the
inevitable fact that with changing prices, the amount of goods to be disposed
of pro�tably will change, only that in the case of a buyers’ monopoly the
results are exactly opposite to that of a sellers’ monopoly.
Or stated concretely: the lower is the wage rate �xed by the monopolist,
the smaller will be the number of workers available, and from a correspond-
ingly smaller number of workers will the entrepreneurs be able to collect
that increased return which might accrue from pushing the wage scale down
below the value of the product of the marginal laborer
. in
fact, this value
might even increase through a reduction in the output, which would cause a
This situation is different from the one the consumer-monopolist �nds her-
self in, however, since Böhm-Bawerk deals with capitalist-entrepreneurs hiring
factors of production. This is the realm of production decisions, more speci�-
cally, the production decisions of those who invest their money in a productive
process. In their capacity as capitalist-entrepreneurs, they act as buyers and sell-
ers, sequentially, with or without physically transforming what they have bought
in the meantime. They buy in order to sell and earn a net income based on a
hoped-for price differential. This has profound implications for the issue at hand,
implications that neither Böhm-Bawerk nor Mises and Rothbard appear to have
completely drawn: monopsony implies a different market situation when buyers
are producers, although the corresponding gain might not be said to be speci�c to
the buyers’ position as monopsonists only.
Indeed, the �rst implication to be stressed is that, whether a threat of infring-
ing upon the property rights of some would-be capitalist-entrepreneurs is made
in their capacity as sellers (granting a monopoly privilege to other capitalist-
entrepreneurs) or in their capacity as buyers (granting a monopsony privilege),
or both, a restriction on one is a
de facto
restriction on the other (Wieser, 1927,
219; Méra,
2010, p.
55). This
is because buying and selling are parts of the
same production plan, so if someone is barred from entering the market as a
producer-seller, he will not be able compete on the market for the required factors
If it is often noticed that a monopoly is a monopsony or a monopsony is a
monopoly, this is rarely considered a necessity. And it is true that, with an exclu-
sive grant of monopoly privilege on the sale of a good, one may be its sole seller
while still one among many buyers of its non-speci�c factors of production. How-
ever, even in this case competition is hampered on the factors’ markets since no
competitor is allowed to hire them for the production of the monopolized good.
With an exclusive grant of monopsony privilege, one may be the sole buyer of a
factor of production while still one among many sellers of a good it helps to pro-
duce, provided this factor is not indispensable to its production. Yet even in this
case competition is hampered in the product market, because competitors are not
allowed to produce the product using this factor.
The only way the capitalist’s positions as a buyer and as a seller could be con-
sidered as independent from each other is by introducing the perfect competition
hypothesis in one market or the other. Then we could claim that the exclusion of
some competitors from the perfectly competitive market has a negligible impact
on the price (as long as there are still many participants competing, that is). A
could then be a price taker in one market and a price maker in the other. Yet
we have already ruled out perfect competition as a point of reference. Outside per-
fect competition, then, the forced exclusion of some competitors will make a dif-
ference, albeit a very small one in the above case. No one’s position in the product
market is independent from one’s position in the factors’ markets. The impact of
hampering competition in one market must reverberate throughout the others. In
the large sense of being shielded from the competition of some other capitalists, a
Second, when viewed in this light, it is immediately plausible that monopsony-
monopoly is a market situation in which prices are affected in both product mar-
kets and factor markets, or at least product markets and divisible factor markets.
All that seems really necessary for such an outcome to emerge is that monopoly-
monopsony hampers the arbitrage process otherwise occurring across the whole
social structure of production in all its dimensions in such a way that the price-gap
between products and factors cannot be reduced to the height determined by time
Before examining in more detail the required conditions for such a result, let
us at this stage emphasize that this situation is simply another aspect of the pro-
cess described by Böhm-Bawerk in the above quote, and that it contradicts Mises
and Rothbard’s idea that no downward pressure on wages exerted by a cartel of
employers can ever succeed in establishing a permanently lower wage rate for
some workers. On the other hand, when realizing that product prices will rise
when capitalists-monopsonists “�x” a lower wage rate, Böhm-Bawerk hints at,
but nevertheless seems to miss, the idea that monopoly and monopsony go hand
Monopsony theory revisited
in hand for capitalists. As a consequence, he sees these situations as symmet-
ric but unnecessarily intertwined, and suggests there can be a speci�c gain for a
buyer’s monopoly and a speci�c gain for a seller’s monopoly. Yet if monopolistic
grants of privilege to capitalists grant both kinds of monopoly to them as sellers of
their products and as buyers of the required factors
– no
matter how the restriction
was nominally introduced
that is, if monopoly and monopsony are two sides of
the same coin, then there is no speci�c gain to each. There is no speci�c gain for
the monopsonist corresponding to the gain of a monopolist in the same sense that
there is no separate interest revenue for a capitalist as a buyer of factors’ services
and as a seller of their products. The capitalist-monopolist bene�ts from a speci�c
Now, if it is problematic even to speak of a gain associated with the capitalist
as a buyer only
– since
the alterations of factor and product prices accounting for
it are certainly not independent
– a
special emphasis on factor prices is required
when reassessing monopsony theory and its features. In order to explain the con-
ditions under which the monopoly price-gap can arise, and also the contribution
of lower factor prices to such an outcome, it is convenient to begin with the sim-

The monopoly price-gap with speci�c factors of production
Under (general) equilibrium in the free market, i.e. when all expectations errors
have been avoided and no privilege in buying or selling exists (free competi
tion), all pro�ts and losses have been arbitraged away so that the remaining
gap between the average expense on factors of production and the price of the
product remunerates the capitalist
capitalist (with what Mises calls the
“originary interest rate”).
For any remaining �rm, any attempt to increase net
income would be self-defeating as long as preferences, resources, and all the
determinants of this equilibrium position remain the same. Expansion would
be such that gross income would fall while expenses would rise, or, that gross
income would rise at a slower pace than expenses. On the other hand, if a �rm
tries to increase net revenue by buying and selling less, it would fail because its
income would fall at a higher pace than its expenses. Gross income would fall
and not rise, because the demand schedule of the �rm for its product must be
elastic above the equilibrium price (otherwise the initial situation would not be
properly characterized as general equilibrium) and it would fall at a higher pace
than expenses (for the same reason). In other words, and the demand schedule
that each �rm faces for its product must be elastic enough in the voisinage of
their equilibrium prices that no other choice would bring about a higher net
Therefore a grant of monopoly-monopsony privilege can increase the gap if
and insofar as the forced exclusion of competitors decreases the elasticity of the
demand schedule for the product and/or the elasticity of the supply schedules
of the factors in the relevant ranges (above the free-market equilibrium price of
the product and below the free-market equilibrium prices of the factors). The net
incomes of the remaining �rms are maximized through contraction in this branch
(compared to the free-market level of activity). In this market situation, a monop-
oly-price gap emerges and the corresponding monopoly gain under the new equi-
librium position cannot be arbitraged away.
Let us assume that a grant of monopolistic privilege has been conferred upon
some �rms in the chocolate industry (as sellers of chocolate).
The more exclu-
sive the grant is, the less elastic the demand schedule becomes to each remaining
�rm, the more drastic the contraction, and the higher the new product price will
be. If only one seller is authorized, it faces the whole market demand schedule,
which is necessarily less elastic than the demand schedules all the competing
�rms would have faced under less restrictive conditions. As a consequence, the
net income-maximizing quantities of output and of factors hired are bound to be
Four remarks are in order at this stage. First, and directly relevant to our cen-
tral concern about monopsony theory: if it is true, as explained above, that there
can be no such thing as a perfectly elastic factor supply schedule, the monopoly
price-gap is partially accounted for by lower factor prices. Clearly, lower produc-
tion requires fewer units of at least one divisible factor, a market situation under
which no upward pressure on its price can be felt (Méra, 2010, pp.
57–59), and
can then be implemented by “moving along” the supply schedule to
a position where the price is lower (with lower quantities bringing about a higher
Second, how low the price is depends on the reservation demand of the factor
owners. With a speci�c factor, the only alternative consists in being employed
by another of the remaining employers, if there are any, or in being unemployed.
If the monopoly grant is exclusive to one �rm, this �rm is the sole buyer of the
services of the factor. It faces the market supply schedule of the factor, which
must be less elastic than any supply schedule any one �rm faces under less restric-
tive conditions. The more exclusive the grant is, the more drastic the contraction
and the fall in the speci�c factor prices are. It is not simply that prices must fall
because there is never such a thing as a perfectly elastic supply schedule that faces
buyers. It is that hampering competition results in those supply schedules to each
remaining buyer becoming less elastic, which must then account for a part of the
monopoly price-gap. This must be the case if it is true, as argued above, that the
situation of capitalists as buyers is not independent from their role as sellers, such
that hampering competition in the chocolate market
hampers competition
Third, this does not mean that a grant of privilege makes no difference, whether
it is a monopoly grant on the sale of chocolate or a monopsony grant on the pur-
chase of one or several of its speci�c factors. In the second case, it could be that
other ways to produce the same chocolate product are known that do not require
the use of this speci�c factor. We would then expect the elasticity of the demand
schedule for chocolate faced by any remaining sellers to be altered less relative to
the elasticity of the supply schedules of this factor that they face as buyers. As a
consequence, the monopoly price-gap would arise more as a consequence of a fall
Monopsony theory revisited
in the speci�c factors’ prices than as a consequence of a rise in the product price,
while both would nevertheless occur.
Fourth, to the extent that a lowered factor price is related to a change in the
elasticity of the supply schedules faced by the remaining buyers, one may refer to
the resulting price as a “monopsony price.” Recall that Mises rules out the concept
of a monopsony price or “monopoly of demand” price because no equivalent to
the inelasticity of demand criterion for monopoly prices exists in monopsony.
However, this inelasticity criterion turns out not to be decisive once we shift focus
away from “crude monopoly price” theory and toward an analysis of the require-
ments for the emergence of a mere “monopoly price” in the context of production
decisions. Indeed, as Méra (2015, pp.
130–131) ar
gues, Mises (1998, p.
7) implic
itly admits this when he allows himself to discuss the emergence of monopoly
prices in relation to production expenses by devising an instance of monopoly
price reached through a move along an elastic demand schedule (with the move
being pro�t-enhancing since the example is devised in such a way that the fall in
expenses is higher than the fall in gross income). And as Rothbard (1970, p.
– contradicting
his previous presentations of the theory
– the
circumstance for a monopoly price to arise is that a demand schedule facing a �rm
is made “suf�ciently less elastic” through the forced exclusion of competitors, not
necessarily that it is made inelastic. If it is legitimate to refer to such an outcome
as a monopoly price, it is certainly legitimate to refer to its counterpart in the fac-
Semantics aside, the most important result is that a monopoly price-gap can
emerge with higher product prices and lower speci�c factor prices, speci�cally as
a result of a monopolist’s choice to contract its activity in order to take advantage
of demand and supply schedules made less elastic through the threats of �nes,

What difference do non-speci�c factors make? Is it true that non-speci�city
shields them, and especially the labor factors among them, from monopsonistic
pressure? Or, as in the framework of the monopoly price-gap theory above, could
it be that only increases of product prices (as well as decreases of speci�c factor
First, in regard to labor, the sense in which Rothbard speaks of it as non-
speci�c, when he claims that only an all-encompassing monopsony could extract
a gain from it, does not appear to be a strict one. He must be alluding to “labor-
in-general” (Rothbard, 1962, pp.
572–575) as
a class of factors. Otherwise, the
universal monopsony clause would hardly be required. Strictly speaking, a non-
speci�c factor performs a task which can be used in several production processes
(Rothbard, 1962, p.
39), but
is nevertheless “homogeneous in its supply” (Roth-
bard, 1962, p.
562). This
is certainly true of many tasks that laborers are able to
perform. But labor-in-general is non-speci�c in the larger sense that each person
is able to perform various labor tasks. It would be more accurate to say, then,
if one sticks to the strict de�nition, that a person normally can embody several
labor factors and that she can switch from being a supplier of one to a supplier
of another. The reason it is important to keep this in mind is that there will be
a different marginal revenue product schedule for each task, and the next most
remunerative task an employee will be able to perform may pay far less than
the current one. Furthermore, this particular factor they embody may be spe-
ci�c to the process they participate in. As a consequence, privileged employers
monopolists-monopsonists) may
be in a position to pay a lower than free-market
wage rate without such workers �eeing
to another �eld, although the
lower limit below which workers will quit is higher than if they could not perform
What about a genuinely non-speci�c factor, be it a labor factor, a land factor, or
a capital good? Rothbard claims, following Mises, that a non-speci�c factor will
simply go elsewhere if the employer attempts to push its price below the free-
market level. One problem, however, is that Rothbard’s point proves too much.
If it really followed from the non-speci�city of a factor that a monopsony could
not lower factor prices, would it not follow too from the marketability of money
that no monopoly price of the sort Rothbard envisages can emerge from a monop-
oly of supply? After all, one could say that the buyers’ money “can move else-
where” when the monopolist tries to charge a monopoly price, especially given
that money is the non-speci�c asset
par excellence
, by de�nition. Yet the theory
of monopoly price, as understood by Rothbard and others, implies that such non-
This tension suggests that there is something wrong about Rothbard’s view,
but it does not pinpoint what the issue is exactly. The essential weakness in the
thesis of the necessity of socialism to lower non-speci�c factor prices should be
clear, however. This thesis invites the question: for what price could non-speci�c
factors move into other �rms and industries under partial monopsony? The claim
suggests that the prices the factor owners can charge elsewhere are a given, “inde-
pendent variable” that has nothing to do with the pattern of monopolistic action
under examination. And yet the well-known insights about the price relationship
between substitute goods (Rothbard, 1962, pp.
280–288), combined
with the dis-
tinctively Austrian “law of costs,” imply that this is not the case.
Böhm-Bawerk (1930, pp.
223–234) developed
the view that costs, in the sense
of money expenses, are ultimately determined by consumers’ preferences so that
the determination of relative prices has to be entirely driven by consumers in the
market. In the absence of obstacles, arbitrage aligns the prices of the products
with their costs throughout the economy: this is the common knowledge shared
by classical and pre-classical economists. But where do the costs come from?
For Böhm-Bawerk as well as later Austrians, there is no such thing as a “natural
price” derived from a somewhat exogenous cost of production. Capitalists who
want to keep operations going in whatever industry must advance payment to
factor owners at prices high enough that they do not go elsewhere. The price
the factor owners can obtain elsewhere re�ects the bids of other capitalists. And
Monopsony theory revisited
the most that each capitalist can pro�tably pay in a particular stage of a process
is determined by what the capitalists at the next stage are ready to pay, which is
in turn determined ultimately by what consumers are ready to pay for the �nal
product of these factors in this branch. Capitalists everywhere compete for the
use of these factors and the whole arbitrage process then determines costs of
production as imputed backward from consumption goods prices, not the other
way around. Hence, the height of the bids for factors of production anywhere at
the beginning of a stage depends on the expected income from the sale of their
products received at the end of the stage, which is ultimately determined by con
What is missing when Rothbard tells us that non-speci�c factors can simply go
elsewhere when a monopsonist attempts to bid down their prices is the following
insight: since a grant of monopoly-monopsony may distort the pattern of spending
on various products
beyond the one produced by the monopolist, the upper limit
on the capitalists’ bids for the factors of production “elsewhere” may be a differ-
Indeed, what happens on the side of the buyers of a product when a monopo-
list produces and sells less than he could have if he had been happy to earn a
“normal” income only? Let us assume �rst that the demand schedule above the
free-market price is inelastic, for even though this is not a necessary condition
for such a decision to pay off, it certainly cannot hurt. In such a case, the pattern
of spending by the �rm’s clients is altered in such a way that they are going to
spend more on the product they have to pay a monopoly price for and less on
other products than they would have otherwise. This means that their demand
schedules for those other goods are lower. Therefore, their prices tend to be lower
and capitalist-entrepreneurs have fewer reasons to invest there than they would
have had otherwise. In other words, the marginal revenue product schedules of
the factors employed there were lower than otherwise. Now, if a factor used in the
production of the product charged a monopoly price is also used in one or several
of those sectors, its lower marginal revenue product schedules there affected its
reservation demand here, in the �rst process (Méra, 2010, pp.
59–64). What
can earn “elsewhere” can be negatively in�uenced by the monopolistic pattern
Insofar as there exists a tendency for the factors to command a lower price
elsewhere because of these shifts, their reservation demand here is affected. Their
supply schedule here becomes even less elastic below the free market price, with
more room for a monopolistic-monopsonistic contraction and the establishment
of a monopsony price than otherwise, re�ecting the formation of a larger phenom-
enon, a monopoly price-gap. The fact that under monopoly-monopsony here they
can earn less than they otherwise could here and elsewhere are two sides of the
same pattern of action originating in the monopolistic grant. The less speci�c such
a factor is, the higher the likelihood it will be in use in a branch where its marginal
revenue product schedule must be lower because of the aforementioned shifts.
This is why the non-speci�city of a factor does not automatically shield its own-
ers from the adverse effects of monopolistic-monopsonistic pressure. Depending
on the direction of shifts in product spending, it can actually expose the owners
to such effects.
On the other hand, what if the demand schedule for the good a monopolist pro-
duces remains elastic above the free-market price? A
monopoly price
then results
in lower spending on the product (the monopolist, if successful, will correctly
anticipate that this lower income is more than compensated by lower expenses).
This means that the demand schedules for other goods are higher. Therefore, their
prices tend to be higher and capitalist-entrepreneurs were justi�ed in investing
more there than they would have otherwise, which implies the marginal revenue
product schedules of the factors used there are higher. If a factor used in the pro-
duction of the product charged a monopoly price is also used in some or all of
those sectors, its higher marginal revenue product schedules there affected its res-
ervation demand here. This is a factor in the increase of the elasticity of its supply
schedule to the monopsonist-monopolist, although the exclusion of competitors
inherent in this monopsony position has the opposite effect.
This means is there is less room for a monopoly price-gap in such a case, although
we cannot rule out this possibility, since the elasticity of the demand schedule for
the monopolized product could still have decreased relative to the elasticity of
the supply schedules of its factors of production in the relevant ranges. And there
is even less room for the establishment of a monopsony price for the factor (a
price which would be lower as a consequence of employers taking advantage of
an absolute decrease of the elasticity of its supply schedule), although it remains
possible. Here, non-speci�city is a mitigating or even a shielding in�uence, but
whether it is neutral or whether it increases monopolistic-monopsonistic pressure
on the factors’ prices depends on the direction of the shifts in spending on the

he possibility of lower non-speci�c factor prices as a consequence of grants of
monopolistic privileges, deduced from Böhm-Bawerk’s law of costs and the law of
substitutes, suggests an anomaly in the history of economic thought that this paper
was intended to identify and correct. The anomaly is that the authors who did not
follow Böhm-Bawerk’s lead in some important matters have insisted there exists
signi�cant room for monopsony distortions in the economy, while the authors
who placed Böhm-Bawerk on a pedestal tend to dismiss concerns about monop-
sony, even when it is conceived as a manifestation of interventionism. This is an
anomaly insofar as the �rst group tended to emanate from the Marshallian par-
tial equilibrium tradition (Robinson), which naturally lead to taking factor prices
outside a particular branch of production as given, while the second endorsed an
insight that could have led them to think of those factor prices as dependent upon
actions within that same branch. In other words, taking Böhm-Bawerk’s law of
costs seriously could have led them to realize that there is more room for monop-
sony distortions in a market economy than a Marshallian approach suggests. And
yet, by and large, the opposite occurred.
Monopsony theory revisited
If monopsony theorists, and especially the new wave of scholars among them,
have been able to advance monopsony theory as empirically relevant, it is because
they took a completely different approach. Pure and perfect competition is used as
a benchmark status for welfare comparisons, so any deviation from it originating
from any source can be considered an issue in need of corrective action. A
than-perfectly-elastic supply
schedule for a factor of production, for instance, can
be taken as a manifestation of an employer’s market power and a sign of inef-
What has been argued above is that this way of conceiving of the issue is a
dead end, in that it compares real world situations with an impossible ideal, so
that mainstream monopsony theory, old and new, lacks a realistic criterion to dis-
tinguish between monopsony and its absence, i.e. between monopsony prices and
competitive prices. This is the application to monopsony of the fundamental Aus-
trian criticism of the perfect-versus-imperfect competition paradigm, which lies
at the heart of mainstream monopsony theories. On the other hand, this paper has
also challenged the Austrian view that any monopsony theory should therefore be
considered at best of small empirical relevance and at worst as an absurd piece of
On the contrary, once we endorse a causal-realist outlook in the Mises-
– the
view that human action occurs in discrete steps, that the supply
schedules of factors of production must always be less than perfectly elastic, and
that they can be made less elastic as a result of coercion
– we
naturally arrive
at an Austrian theory of a monopoly price-gap, with monopoly and monopsony
prices as two features of the same phenomenon. A
monopoly price-gap
and a
corresponding monopoly gain can emerge when the everyday arbitrage activi-
ties of capitalist-entrepreneurs throughout society’s structure of production are
hampered in such a way that the demand schedules for products and the supply
schedules of their factors of production are made less elastic as a consequence of
The possibilities of such distortions are more widespread in a market econ
omy than Mises and Rothbard’s writings on monopoly price theory suggest,
for two reasons. First, the requirement they put forward for the formation of
a monopoly price
– demand
schedules for products made inelastic
– is
only in the context of decisions over an already-produced stock of a good.
In the realm of production decisions, there is no need for demand schedules
to become inelastic for contraction to pay. This is because maximizing gross
income is not the relevant consideration. Expenses are compared with gross
incomes, so that an expected falling gross income is an insuf�cient reason
not to contract activity. Contraction and its outcome
– lower
factor prices and
higher product prices
– can
therefore be a more widespread phenomenon than
the inelasticity-of-demand-based theory suggests. Second, non-speci�city
of the factors of production does not necessarily protect their owners from
monopolistic-monopsonistic pressure, since the amount they can earn outside
of the monopolized branch of industry can be negatively affected as an indirect
Finally, if coercion-induced changes in the elasticity of demand schedules for
products and supply schedules for their factors are the decisive requirements
for monopsony prices and a monopoly price-gap to emerge, it immediately fol-
lows that the removal of those grants allows wage rates and the marginal revenue
product of labor factors to be as close as possible at any given time. A
analysis of
minimum wage legislation is beyond the scope of this paper. Yet while
such price controls can conceivably mitigate the impact of monopoly-monopsony
on factor prices
– if
they are not set above and remain below (prone to change)
market-clearing prices
– their
lack of �exibility and adaptability to various factor
market and situational changes render them rather inadequate, as compared with
See, for example, von Wieser (1927, pp.
218–220), Clark
(1915, p.
76), and
Applying this
insight to Manning’s approach, which explains less-than-perfect elas-
ticity through search costs, we only need to realize that there is no possible world in
which they can disappear. Search costs exist because no one is omniscient. If we argue
that welfare losses result from this lack of omniscience, we might as well claim that
having to pay any price at all for anything entails tremendous welfare losses as com-
See Rothbard
(1962, pp.
661–704) for
the justi�cation of his thesis that only aggressive
violence or the threat thereof implies a market situation that is distinguishable from a
As Mises
(1949; p.
380) adds:
“In the same way in which governments restrict compe-
tition in order to improve the position of privileged sellers, they can also restrict com-
petition for the bene�t of privileged buyers. Again and again governments have put an
embargo on the export of certain commodities. Thus by excluding foreign buyers they
have aimed at lowering the domestic price. But such a lower price is not a counterpart
be sure, a monopsonist can take advantage of the exclusion of its competitors in
the sense that any buyer, given a market supply schedule, is better off when the market
demand schedule and equilibrium price are lower. But the gain here is not speci�c to
its situation as a monopsonist. By the same token, the monopolist can bene�t from the
exclusion of competitors in the sense that any seller, given a market demand schedule,
is better off when the market supply schedule is lower and the equilibrium price con-
sequently higher. But this is not speci�c to its situation as a monopolist, and occurs
Even here,
speaking of a monopsony price is problematic because in socialism, there are
Again, it
is true that the equilibrium price will be lower as a consequence of the exclu-
sion of competing would-be buyers. But the gain involved is not speci�cally a monop-
sonistic gain. In fact, far from aiding in somehow restricting demand, a simple supply
and demand analysis shows that the monopsonist buys higher quantities than it would
avoid any misunderstanding, at the time Böhm-Bawerk was writing, the word
“monopsony” was not yet in use. Economists were only speaking of a “monopoly of
demand,” hence the expression “monopolist” here, not to be confused with the monop-
olist as a seller.
Monopsony theory revisited
One could then wonder if it makes any sense to speak of “monopoly” and “monopsony”
simply using “monopoly” for all related examples. However, in the structure of
production, not everyone can be both a buyer and a seller in his capacity as a producer,
and not everyone can be a capitalist. Some will have to be sellers of original factor
services since production processes must use some original factors of production (land
and labor). As a consequence, it still makes sense to maintain the distinction between
buyers’ monopolies and sellers’ monopolies because some people can only be granted
a monopoly of supply. The symmetry is with the situation of a consumer-monopsonist
who cannot be a monopolist. They have one thing in common, which is that they can-
not be both monopolists and monopsonists, while capitalists are both insofar as at least
some of their would-be competitors are forcefully excluded from the relevant markets.
Or whatever
cause actually results in a “normal” price-gap. See Mises (1949,
521–533) for
the view that such an “originary” interest rate is determined by time
Because error
is an ever-present possibility in the realm of human action (Rothbard,
1962, p.
7), there
is no reason that equilibrium, especially general equilibrium, must
occur. The assumption is made only as a way to analytically disentangle various
sources of discrepancies in the rates of return we can observe in an actual market
The assumption
is that one or several sellers have been given a license to sell choco-
late, so that no one else is allowed to. This is an explicit monopoly grant. However, it
should be clear that any other regulation which, in effect, drives competitors out of the
market is not fundamentally different.
this means that the standard distinction between the case in which labor
allocation and pricing is affected by monopsony (Borjas, 2013, pp.
187–194) and
case in which it is affected by monopoly in the market for the product (Borjas, 2013,
It follows
that, for the analysis of production decisions, the Mises-Rothbard-inspired
theory of monopoly price-gap, with monopoly and monopsony prices as its symmetric
components, is actually more similar to the mainstream one than Mises and Rothbard’s
explicit discussions of monopoly prices and monopsony would have seemed to imply,
with the important difference that these discussions have no room for a perfect compe-
In this
respect, one would expect that the possibility of lowered wages under
monopsony-monopoly pressure would mostly concern very speci�c, highly quali�ed
As Rothbard
(1962, p.
905) puts
it, “the consumers are only fully responsible for their
demand curve on the free market; and only this demand curve can be fully treated as
In this
paper I
essentially focus
on factor prices and production in a branch where a
monopoly-monopsony grant exists to explain how a monopoly price gap can arise.
Other industries are dealt with only insofar as the grant indirectly affects the marginal
revenue product schedules of the same factors there, leaving more or less room for a
monopoly price gap (and therefore a monopsony price) in the �rst. However, other
indirect consequences in the pricing process are to be expected and can be felt virtually
anywhere in the economy. For a more complete analysis, see Méra (2010). In short,
grants of monopoly-monopsony privileges to capitalists imply a prevailing tendency
for lower land and labor prices and aggregate income than under free competition
even though
some of them may command higher prices
– the
overall distribution of
incomes being altered to the disadvantage of those classes of factors. In addition, over-
all output is reduced insofar as some units of those factors leave production altogether
in the face of declining prices for their services, while the allocation of the remaining
Ashenfelter, Orley C., Henry Farber, and Michael R. Ransom. 2010. “Labor Market
Bellante, Don. 2007. “The Non Sequitur in the Revival of Monopsony Theory.”
Journal of Austrian Economics
Blair, Roger D., and Jeffrey L. Harrison. 2010.
Monopsony in Law and Economics
. Cam-
Block, Walter E., and William Barnett. 2009. “Monopsony Theory.”
American Review of
Boal, William M., and Michael R. Ransom. 1930.
The Positive Theory of Capital
. Trans.
William Smart. New York: G. E. Stechert
“Monopsony in the Labor Market.”
Journal of Economic Literature
35 (1):
Böhm-Bawerk, Eugen von. 1962. “Control of Economic Law?” In
Eugen von Böhm-Bawerk, Vol. 1
. 6th ed. New York: McGraw-Hill Education.
Burdett, Kenneth, and Dale T. Mortensen. 1998. “Wage Differentials, Employer Size, and
Chamberlin, Edward. 1933.
The Theory of Monopolistic Competition
. Cambridge, MA:
. New York: Palgrave MacMillan.
Council of Economic Advisers. 2016. “Labor Market Monopsony: Trends, Consequences,
and Policy Responses.” White House Issue Brief, October
2016. https://obamawhite�les/page/�les/20161025_monopsony_labor_mrkt_
Davenport, Herbert J. 1913.
The Economics of Enterprise
. New York: Palgrave Macmillan.
Demsetz, Harold. 1969. “Information and Ef�ciency: Another Viewpoint.”
Fetter, Frank A. 1915.
. New York: The Century Co.
. Chicago: Aldine Publishing Company.
Irwin, Neil. 2016. “A
New Movement
in Liberal Economics That Could Shape
Hillary Clinton’s Agenda.”
The New York Times
, November
4. www
Kuhn, Peter. 2004. “Is Monopsony the Right Way to Model Labor Markets? A
of Alan
Manning’s Monopsony in Motion.”
International Journal of the Economics of
11 (3): 369–378.
Machlup, Fritz. 1952. The Political Economy of Monopoly: Business, Labor, and Govern-
ment Policies. Baltimore: The John Hopkins Press.
Manning, Alan. 2003.
Monopsony in Motion: Imperfect Competition in Labor Markets
Generalised Model
of Monopsony.”
The Economic Journal
116 (508):
“Monopsony.” In
The New Palgrave Dictionary of Economics
, Sec-
ond Edition, eds. Steven N. Durlauf and Lawrence E. Blume. Palgrave Macmillan.
1. “Imperfect Competition in the Labor Market.” In David Card and Orley
Ashenfelter, eds.,
Handbook of Labor Economics
, Vol. 4B. Amsterdam: Elsevier,
Monopsony theory revisited
Méra, Xavier. 2010. “Factor Prices Under Monopoly.”
Quarterly Journal of Austrian Eco
Note on
the Limits to Monopoly Pricing.” In
The Next Generation of
Austrian Economics: Essays in Honor of Joseph T. Salerno
, eds. Per Bylund and David
Howden. Auburn, AL: Mises Institute, pp.
Ludwig von. 1998. “Monopoly Prices.”
Quarterly Journal of Austrian Economics
Human Action: A
eatise on Economics
. Scholar’ ed. Auburn, AL:
Robinson, Joan. 1969 [1933].
The Economics of Imperfect Competition
. 2nd ed. New York:
Rothbard, Murray N. 1993 [1962].
Man, Economy, and State: A
eatise on Economic
. Auburn, AL: Ludwig von Mises Institute.
Power and Market: Government and the Economy
. Auburn, AL:
Wieser, Friedrich von. 1927.
. Trans. A. Ford Hinrichs. New York: Adel-
phi Company.
The fundamental feature of neoclassical price theory is the U-shaped average cost
curve and its twin curve, the U-shaped marginal cost curve. Firms and producers
seem to automatically respond to the existing marginal revenue curve crossing the
marginal cost curve at a certain point. The “optimal” choice for the individual �rm
may be established under the famous MR=MC equation. The reason why the com-
pany does not boost production further is the disappearance of falling marginal
cost, and also of average cost. As costs start to rise, the L-shaped curve reaches its
lowest point and becomes the U-shaped curve, which stops the �rm from expand-
ing further. The boundary is set and the �rm loses previously bene�cial econom-
A further feature mainstream price theory is revealed in the mostly arbitrary
distinction between long- and short-run production. Fixed costs are not really rel-
evant for immediate price formation, since they have already been covered. The
marginal cost curve crossing the marginal revenue curve seems to allow for the
highest possible pro�t (or lowest possible loss). Often this point is presented using
mathematical language, where derivatives allow for local maxima in accordance
with graphical presentations. In the short run, only variable costs matter. Appar-
ently, �xed costs seem to be at the same time sunk costs, borne in the past and with
Another side feature of this approach is its treatment of prices as parameters
resulting from the interaction of other underlying factors. In the choice of various
theoretical frameworks, prices (which always respond to changing conditions)
can either be central in the decision making process, or be more like end results
of the market process. Despite their recognized importance, however, prices are
seen in the neoclassical framework as almost completely passive �nal outcomes.
In other words, prices are determinate rather than determinates. One questionable
consequence of this view is the high �exibility of prices
– which
no doubt exists
in some markets, but is not a universal phenomenon. Prices are not perfectly �ex-
ibly or adjustable once surrounding factors change. As a result, it is suggested
in the Post-Keynesian literature that economic theory could start with a division
into �ex-price markets as opposed to �x-price markets, which also makes for a
An Austro-Post-Keynesian synthesis?
distinction between markets corresponding to the traditional theory and markets
core of the challenge to the above version of marginalism comes from
various empirical studies related to assessments of business and management
practices. These studies show that managers working in different industries do
not act according to marginal principles. Naturally, in and of itself this may be
a completely irrelevant argument, because the responses of the people polled in
these studies may not be accurate. Just as people may deny the in�uence of grav-
ity, so it may be with the case of accurately articulating the in�uence of marginal
thinking on economic decisions. Denial may be an equivocal quibble about some
unknown factor, or it may be unclearly and imprecisely de�ned by the question-
naire, yet contribute signi�cantly to the results. Nevertheless, detailed inquiries
into business practices and the workings of price setting are actually instructive
for seeing how prices work from the perspective of a single enterprise (on cost
studies, see Lee, 1986, pp.
An individual company appears not to obey
neoclassical principles because its actual curves, based on real data, are shaped a
certain way, rather than because of free-�oating beliefs expressed in a vacuum. In
many of the �rms that were questioned, the curves were not in a typical marginal
central observation of these studies is that prices are “administered” by
�rms. They are “�xed” in their business plans based on the costs incurred in pro-
duction. This is not to say, though, that companies are acting in a “monopolistic”
manner in the neoclassical sense. Rather, the whole neoclassical notion of monop-
oly and monopolization loses its value and becomes of marginal importance (Hall
and Hitch, 1939, pp.
30–31). What
matters in business calculations are average
total costs, which include, besides variable costs, �xed costs that do not vary
with the amount of product created. Firms apply a markup over those total costs
(Fabiani et
al., 2006,
41). In
the neoclassical approach, the marginal cost curve
dictates the decision of how much to produce and which price to set, whereas
�xed costs play no part in the marginal cost curve.
Yet the reality of pricing seems to run against the practice of ignoring �xed
costs. In practice, marginal costs
– costs
incurred because of increased output
are not
the decisive element. This is not because entrepreneurs do not know what
“marginal cost” means, but because they price their products based on the rel-
evant factors of production. Various empirical surveys con�rm that total costs are
considered in pricing, but not variable costs, and therefore the whole notion of
marginal cost seems to be disconnected from the pricing process (Shim and Sudit,
does pricing happen? A
�rm starts
off with an accounting calculation of
the necessary expenses to produce a product, say, running shoes. The variable
costs are measured (leather, rubber, etc.), but also costs which must be covered
no matter what the level of production is. These costs come to be known as �xed
because they do not vary with output. Next, assuming a particular level of produc-
tion, average cost is assessed, say at 50 dollars
per pair
of shoes at current capac-
ity. On top of this there is an additional markup to be earned by the company,
for example, an overhead of 100%
percent, so
that the shoes are sold at 100 dol-
per pair
. Fifty dollars of this is a money surplus that stays within the company
in return for adding something productive to the market (organizing the factors,
allocating them through the passage of time, etc.). As an aside, there are some
differences between Post-Keynesian approaches to pricing, especially involving
the concepts of: mark-up pricing, full-cost pricing, normal-cost pricing, historic
normal-cost pricing and target return pricing. Yet as Lavoie argues (2014, p.
there are
no substantial differences between these �ve variants. All rely on the
idea that companies are focused on their total (not marginal) costs of produc-
tion. The variances come from dissimilarities in the accounting de�nitions used in
The notion of cost-based pricing should not be confused with a deterministic
classical doctrine in which prices are thought to be formed almost automatically
based on the past costs of production (actually, we should be careful about assess-
ing the classical school this way
– it
simply did not go into these issues beyond
the general intuitive level). Even though costs are part of business projects, prices
are not simply cost determined. There is no mechanical transmission of costs into
prices. Rather, they are “strategically” determined in the decision process of the
company, which is cost-focused in managing allocations and price setting (Shap-
iro and Sawyer, 2003, pp.
One of the main points of the Post-Keynesian critique of neoclassicism is to
emphasize the fact that prices are not constantly readjusting, and that sellers of
�nal goods adjust inventories �rst and prices second (Eiteman, 1947, pp.
917). Once
goods are purchased in higher quantities, the decision is made to
increase inventories and prepare additional stocks to offer to consumers. In the
case of decreased demand for products, unsold inventories will pile up, and fewer
goods will be ordered by the seller. If this state of affairs continues, prices will
also adjust. Prices can fall to clear the inventories, and the same is true of goods
ordered from producers. Both costs and prices will eventually adjust, with spe-
ci�c lags along the way. The typical business practice is to periodically review
prices, and after such a review (which is not done on an everyday basis, but only
rarely), sometimes adjust them (Melmies, 2010, pp.
450–453). Reviews
and read-
justments can be spread throughout a whole year, but some empirical studies sug-
gest that the adjustments are concentrated around one month, such as January, in
which most of the contract reassessments are done (see e.g. Martins, 2005, p.
Firms in
general are not operating at what we might call “full capacity.” They
make decisions alongside generally downward-sloping average cost curves, and
marginal costs that are roughly constant in the spectrum of decisions. There is
always room for reserved extra capacity in case inventory adjustments must be
triggered. The main argument for rejecting the typical U-shaped neoclassical
= MR
analysis is
the observation that �rms do not operate in the face of ris-
ing marginal costs (Lavoie, 2014, pp.
argue that prices are �xed, though it would be more accurate
to say that prices are expected to be �xed for a chosen amount of time at some
future date. We may slightly clarify this approach by pointing out that producers
undertaking an investment process need not assume that current prices will pre-
vail under future circumstances. They do not even have to assume that particular
cost levels will stay the same. The vital point is to keep the business plan in
– or
actually, to use a much better expression, to keep the business canvas
in sight (Blank 2013). A
business plan
is a �xed and equilibrated approach to
decision making. It assumes prior knowledge of a particular product with known
average costs of production and revenues at particular turnover levels. In practice,
business plans are never executed, however. The reality of uncertainty quickly
veri�es the plan, and the ef�ciency of the market allows for adjustments to take
place. Sellers adjust both products and monetary prices. Instead of talking about
business plans, it is much better then to talk about the business canvas and the
ability of the enterprise to respond properly both to consumer preferences about
the product (its functioning included) and the monetary surroundings of the �rm:
A neoclassical response?
Neoclassical economists are well aware of Post-Keynesian criticism. Their gen
eral response is to claim that there is no inherent con�ict between how businesses
function and how economic theory describes the tendencies of market activity. In
other words, it is always possible to use the supply and demand framework. The
Post-Keynesian argument is that if customers decrease purchases at the same time,
�rms will probably increase inventories. There is nothing in the supply and demand
framework to prevent us from agreeing with this point, which can be explained
either as a form of reservation demand or as a consciously chosen optimal market
surplus. Also, the long-run change can be shown through supply and demand curves
Supply and demand schedules are one way to conceptualize parties’ prefer-
ences about potential trades and bargains. There is nothing speci�cally neoclassi-
cal about this. Things become a little more problematic once we try to answer the
If �rms appear to price their products based on average costs
of production, can the scheme of marginal costs and revenues be defended at all?
It appears so, but the argument may be somewhat surprising
– the
defense is that the marginalist description leads to the exact same results as the
ones presented in the markup pricing approach. In order to arrive at these results
various rede�nitions are necessary. Langlois makes a point persuasively while
developing a marginalist model with inventories included:
The model developed below [in Langlois’s paper] transforms the neoclassical
vision of the �rm to integrate the holding of inventory as a strategic response
of the manufacturing �rm. This approach relies on the assumption that con-
sumers respond to the availability of commodities, and �rms hold inventory
to endow their output with this characteristic. Thus production and pricing
decisions will take explicit account of a �rm’s desired or target level of inven-
tory. The model developed in this paper is descriptive of the particular level
that a �rm could choose for target inventory, but it does not make explicit
the dynamic process of adjustment of production and price that must take
place to maintain inventory at its target level. The hypothesis presented and
empirically tested in what follows is that the level of target inventory is cho-
sen together with its selling price so that pro�t is maximized over the time
it will take to sell the inventory. The �rm achieves this goal by maximizing
per unit of time
, optimization which generates a formula for optimal
markup over the
of goods in inventory, where average cost is
the average production cost of the goods themselves. Thus, by transforming
the neoclassical vision of the �rm to integrate inventory, direct cost pricing is
predicted using marginalist terms.
counterargument could begin with Occam’s Razor
– if
the end result of
the revised approach is the same, why take a longer route? Besides, as Langlois
explains, the model excludes the explicit elements of the adjustment process, the
heart of everyday market adaptation. However, we will not judge here whether
it is possible to defend the neoclassical position. Instead, we turn to the Austrian
An Austrian take on markup pricing
Even though the Post-Keynesian criticism may create potential problems for the
neoclassical version of marginalism, this need not be the case with respect to
Austrian theory, which at times is poles apart from neoclassical analysis. Actu-
ally, in some cases the Post-Keynesian contribution to price theory strengthens
Austrian arguments about the market process, especially in those aspects where
First, consider the question market equilibration and stabilizing tendencies. Post-
Keynesians argue that markets do not clear in the neoclassical sense: prices are
not constantly responding to changing supply and demand conditions (or actually,
as neoclassicals would say, within the framework supply and demand curves have
speci�c elasticities). Perfect market-clearing equilibrium would be the case with
consistent movements towards equilibrium in the non-monetary sense, where
valuations of consumer preferences are almost perfectly imputed back onto the
factors of production. Two forces working from opposite sides of the market
subjective utility
and physical limitations on supply (marginal rates of substi-
tution and marginal rates of transformation)
– would
create constantly shifting
prices as adjustable parameters. With all the data frozen, a �nal equilibrium could
Since the data change all the time, the various adjustments performed by non-
omniscient humans are imperfect. Completely �exible prices could still miss
target, which under uncertain conditions may generate more chaos than more
stability. We could never be sure that automatic and instantaneous adjustments
happened in the proper direction and did not result in the piling up of mistakes on
mistakes. The rational method for market participants is to create optimal price
stickiness and price rigidity with rational idle capacity (Salerno, 2010, pp.
193). For
every market the variable of elasticity is unique and is only discovered
as the process goes on. In case of a stock market, prices are very �exible. In case
of retail markets, prices are more �xed and the appropriate responses are revealed
over longer periods of time, since the immediate “clearing” of the market
– that
emptying the shelf
– may be
a poor economizing decision. The discovery pro-
All this relates to the concept of the plain state of rest
– a
moment when market
transactions cease and people are no longer interested in trading. This concept of
equilibration was developed by Mises (1966, p.
245). In
no sense should this point
be seen as either a full equilibrium or a �nal state of rest with complete adjustment
of costs to prices. There is no imaginary construct involved. “Plain state” means
that under current circumstances there is simply no consensus between the trading
parties that an exchange should take place. Either the market is closed, or there are
no more customers willing to purchase the product at the current price. In order
to attract more customers, the price would have to fall. This is not guaranteed,
though: the only potential buyers and sellers interested in trading were dissatis�ed
with the price and believed it to be either too high or too low at its previous point.
Hence, the preferences of sellers who decide not to boost sales further by lower-
ing their prices �t well with the notion of a temporary plain state of rest (Mises,
o relate the notion of “rest” to the above discussion: marginalist neoclassicals
seem to talk about the �nal state of rest
– when
future prices are in full accord-
ance with the costs existing throughout all industries, including interest payments.
Post-Keynesian reservations are the result of a plain state of rest perspective. In
the immediate run, in everyday transactions, sellers make inventory adjustments
based on their capacities and their business canvas. As decisions become more
incorrect, further mistakes are capitalized into the project. The adjustment toward
a (never attainable) �nal state of rest is a constant process that includes temporary
optimal lags. Therefore, despite the fact that changing supply and demand con-
ditions do not produce equilibrium prices as conventionally de�ned, economic
reasoning can rely on the realistic momentary equilibrium of the plain state of rest
A second point, related to the �rst, has to do with the economic calculation debate.
Mises’s challenge was centered on the concept of a common denominator prop-
erly employed in order to rationally assess production projects (Mises, 1990).
This occurs ex ante, when speculative business decisions are made, and ex post,
when pro�ts and losses are capitalized into the current market value of capital.
Mises’s critique focuses on the lack of a properly functioning price system in
which appraisements are at the heart of economic decision making (Salerno,
1990, p.
44). Post-Keynesian
descriptions of how prices are formed, instituted,
and “�xed” by �rms are stories about the important role these organizations play
in setting monetary values. The “administration” of prices in relation to their costs
of production and overall market demand is an essential part of the market process.
It is subjected to entrepreneurial decision making, with each entrepreneur con-
stantly trying to hit the “right” prices with suf�cient markups; this requires active
judgment as opposed to passive alertness (McCaffrey, 2015). Prices in themselves
are not automatically produced by underlying conditions. They are “set” by com-
peting entrepreneurs, and further subjected to critical consumer judgment and to
the competitive pressures of rivals. The inclusion of the costs of production in
business calculations is actually the inclusion of the whole social appraisement
process involving other competitive bidders and other possible ways of employ-
ing the factors.
Part of the empirical research on pricing by managers actually
con�rms this view: managers give reasons for not raising prices that include fear
discussing the impossibility of socialism, Mises nowhere attempts to
argue that the main problem with market socialism is the insuf�cient speed of
price adjustments because of changing supply and demand conditions (as argued
by Hayek, 1940, p.
135). Rather
, the crucial argument for Mises is about the
lack of competitive appraisement due to the abolition of �nancial and manage-
rial markets. Markets cannot be “played,” not because prices adjust over a long
time (Mises, 1966, pp.
706–709), but
because under central ownership there is
no true competition among entrepreneurial assessments of resources. Competi-
tive planning
– actually
, competitive canvasing
– is
substituted for bureaucratic
planning. Part of this story may have something to do with optimal price rigidity
and price adjustments based on entrepreneurial expectations, but this is never
more than part of the story. Competition is also not homogenous. New ways of
producing things do not appear uniformly in the market and affect all �rms to the
same extent. There is much heterogeneity in markets, both in the ability to reduce
costs and in the �eld of product diversity. The point is further strengthened by the
Post-Keynesian reference to the “Schumpeterian” type of competition (Lavoie,
are not phenomena straightforwardly reducible to other underlying
and determining factors, for example, physical factors (Huerta de Soto, 2010,
206–207). Surpluses
and shortages are not simply calculated by measuring
physical quantities being added to a pile, or that are removed from one. Instead,
they are assessed in comparison to already-existing prices and estimations of the
prices that could have been formed if the market was not centrally controlled.
Therefore, price-setting and price-formation on the part of entrepreneurs are the
driving forces of the markets. They have to be placed within the scope of eco-
nomic reasoning, and cannot be reduced to alternative other forms of imputation,
for example, “value imputation” or imputation based on physical characteristics
(Herbener, 1996).
Another theme similar to questions of equilibration relates to the problem of
price coordination as opposed to plan coordination. As Salerno argues, market
developments should be seen in the spirit of W.
H. Hutt’
s analysis, which empha-
sizes price coordination (Salerno, 2010, p.
182). In
contrast, plan coordination
is a concept mainly developed in the Walrasian tradition, where the data them-
selves equilibrate the market. Price coordination, however, emphasizes the entre-
preneur’s role, in that this view does not see prices as intermediate steps toward
an equilibrium fully reducible to other variables (such as utility, scarcity, etc.).
This approach to prices inherently leads to Lange’s perspective, in which prices
are pure parameters equalizing and balancing out other, more relevant and fun-
damental elements (Lange, 1956, p.
89). However
, contrary to this view, prices
are themselves fundamentals, and are not reducible to anything less than human
Waters articulates the problem this way:
Suppose that we de�ne an essential element in any �rm called “entrepre-
neurship.” This cannot be bought on the market, but is a specialised and
personal attribute of each individual in the community. A
�rm exists
an individual employs his entrepreneurial ability. But since the amount of
entrepreneurial ability possessed by a man is an attribute speci�c to him, the
production function will vary from one person to another.
(Walters, 1963, p.
other words, as Post-Keynesians argue, prices are not set by “invisible forces”
or �ctitious omniscient auctioneers (Lavoie, 2014, p.
156; Melmies,
2010, p.
Frederic Lee,
one of the most prominent Post-Keynesians, uses the concept of
“sequential acts of production,” which �rms use to �nance plans for growth and
expansion. Part of these plans has to do with coordinating prices and avoiding
complete elasticity, which would complicate things (Lavoie, 2016, pp.
3–4). In
s words, “competition is pervasive but not pernicious or destructive” (Lee,
2011, p.
18). Entrepreneurs
do not equate marginal costs and revenues, at least,
not unless we allow for a suf�cient amount of rede�nition and restatement in
order to make these terms conform to reality.
A third point involves tracing the challenges of the imputation process. The cen-
tral but broad idea behind marginal economics is that prices are determined by
marginal utilities of goods. During the development of marginal economics it has
sometimes been argued that people often pay prices for goods that are even lower
than their marginal utilities would suggest (indicating a consumer surplus). In this
view, marginalism works perfectly in the case of closed markets with no supply
responses (such as markets for rare paintings, an example that was mentioned
already in the work of David Ricardo). Yet once the analysis shifts to markets
for reproducible goods, prices can fall below the limits set by marginal utility.
Demand becomes less decisive and important, and becomes only a limiting �nal
Examples can be given for various goods. Take the hypothetical case of blue
shirts, demand for which is much higher than for green shirts. Despite signi�cant
differences in demand, the prices of these two goods can generally be assumed
to be the same. Adjustments mostly occur on the inventories side
– there
more blue shirts in the warehouse, because higher consumer demand is expected.
According to the overall market marginal utilities, the price of blue shirts could be
higher (as higher marginal utility of shirts indicates that people would be inclined
to pay a higher price than for green shirts).
Instead, sellers adjust the quantity
available, not prices. In the standard case they could increase the prices of blue
shirts, but if they do so they also create a pro�t opportunity for other sellers to step
in and sell blue shirts for a lower price. Rivals can compete with entrepreneurs
who raise prices precisely because costs are low enough to be comparable with
green shirts. Hence, costs bring prices down to their levels. The same story can
be told about sports jerseys and other types of products with similar price labels
but signi�cant differences in demand (we abstract here from branding that could
The Post-Keynesian story seems to �t well in these examples. Yet do these
cases prove that there are limitations to the marginalist approach? No, at least,
not in light of Böhm-Bawerk’s analysis of the law of costs and marginal utility
(Böhm-Bawerk, 1962; Böhm-Bawerk, 2002). Marginal utility is a broad concept
applicable to a wide variety of the products in the market. The basic examples of
marginal utility suggest that a loss of the �rst unit does not entail the loss of the
highest goal in the preference ranking. It simply means that the last unit must be
used for that purpose, so that only the least important goal is lost. We “reproduce”
a lost unit by reallocating the least valuable unit to the most valuable goal. As
Böhm-Bawerk brilliantly demonstrated, the same applies to cases of reproducible
goods. Once we lose a unit of some �nal good, we do not necessarily lose the
particular utility currently attached to it. We simply reallocate other units
case of reproducible goods, we employ complementary factors of production to
recreate the lost utility.
According to Böhm-Bawerk, the law of costs is actually an idea about mar-
ginal utility in disguise. In the shirts example, for instance, it does not matter that
demand (and marginal utility) for blue shirts is higher relative to green shirts. What
matters are the marginal utilities of other goods and services that would have to be
given up in order to reproduce blue shirts. And since green and blue shirts require
basically the same sacri�ce, virtually the same marginal utility would have to be
lost. If we lose the last-produced blue shirt, we only have to give up the produc-
tion of the last green shirt and switch green dye for blue (just as when we lose
the most important blue shirt we only have to use the marginal shirt as the �rst).
Therefore we have a perfect explanation of why the costs of both shirts are the
It may be sensibly argued that Böhm-Bawerk’s work anticipates some elements
of the Post-Keynesian critique of neoclassicism. He elegantly demonstrated the
connection between �nal prices and the costs of production expended during pro-
duction. From the perspective of an individual producer, it may seem that sellers
practice cost-based pricing. Yet at the same time, this fact in no way validates
the broad marginalist point that costs themselves result from other potential
investment avenues that could be undertaken. Once we look at the economy as
a whole, we see price-based costing despite the fact that �rms attempt to engage
in cost-based pricing. A
markup is
simply another price for a productive service
performed by a �rm. Its value depends on the usefulness of organizing produc-
tion subjected to the diminishing marginal utility. After all, markup levels depend
on the quality of entrepreneurship and the amount of competition in the sector
fourth point has to do with the appraisement process permeating all the factors
of production, even the ones already devoted to production processes. So-called
�xed costs do not automatically fall outside business calculations simply because
they were already covered in the past. Fixed costs are not a synonym for sunk
costs (Wang and Yang, 2001). They matter just as much as marginal costs. If only
additional marginal costs mattered for pricing, then in most circumstances sellers
would give away most goods they possess for free, since giving away existing
supplies does not generate any costs (and the materials have already been paid for
in the past, just as in the case of �xed costs!). The reason sellers never do this is
that they know or expect that their goods will be more highly valued. They econo-
mize the goods they have by increasing their reservation demand, because they
expect future values to increase beyond their momentary current levels. That is
why their calculations include �xed costs also. Only parts of these costs covered
are sunk and cannot be recovered. These are labeled as money wasted, and their
current prices represent capitalized losses. Only once this occurs do costs already
Consequently, as long as there is potential economic value in existing, already-
paid �xed costs, they are part of the pricing process (Machaj, 2013, p.
Potential econom
ic value exists as long as there are valuable opportunities for
factor employments. For example, an already-rented place of work is a �xed cost,
since it does not vary with output. Yet it is not a loss simply because it was paid
for in the past (simply put, it is not a sunk cost because it was covered in the
past). If work is canceled the cost, or at least parts of it, can be recovered, say,
by using the place for some other purpose (as a consumer good or by re-renting).
Remember Böhm-Bawerk’s law of costs. If the place can be rented by someone
else, then it has an alternate utility. In order to justify not using it in its alternative
capacity, the chosen use has to generate suf�cient employment value. In this way
the �xed costs of renting are part of normal everyday business calculations, even
if the costs have already been paid. The past is de�nitely not irrelevant, so long as
Mises does not mention the concepts of marginal revenue and marginal
cost while discussing the basic functions of pro�ts and losses (Mises, 1966,
289–294). Instead,
he focuses on total costs and total revenues from produc-
tion. Yet at one point he does adopt the neoclassical view:
If the entrepreneur is still free with regard to the project in question, because
appears to have fallen into the confusion that �xed costs (and past costs)
are always sunk costs. Nevertheless, as we have shown above, past and commit-
ted costs are not yet what Mises calls in this paragraph
inconvertible investments
at least, not until the losses are realized and capitalized. While making a deci-
sion about future output, entrepreneurs surely include the values of previously
purchased capital equipment as long as there is potential value in them. This also
includes past �xed costs, since they can be reallocated and economic value could
be extracted from of them in one way or another.
Post-Keynesian teachings on price formation are quite relevant for empirical
descriptions of entrepreneurial choices. As we have seen, Post-Keynesians raise
challenges to existing neoclassical versions of marginalism. At the same time,
these criticisms do not pose a threat to the Austrian version of marginalism; rather,
they strengthen some Austrian arguments about price formation, especially in
regard to concepts such as coordination, calculation, and price setting. The Aus-
trian version of marginalism can easily be enhanced by the real-world business
practice of markup pricing. This practice neither undermines the backwards impu-
tation of value nor changes the Böhm-Bawerkian notion of the law of costs work-
ing as an implication of marginal utility in disguise. Additionally, the concept of
markup pricing enriches calculation arguments and the entrepreneurial approach
Actually, �xed costs may be included in the marginal cost curve, but only at the begin-
Post-Keynesians also emphasize the process nature of the market, although
from a somewhat different perspective than Austrians. For example, “Competition in
classical economic thought from Adam Smith to Karl Marx is thus a process, not an
end-state. As re�ected in investment and growth policies, competition involves the pro-
cess by which resources are allocated
– and,
ultimately, income distributed
– between
classes over time rather than just their allocation among individuals at a point in
time. This emphasis re�ects the preoccupation of the classical economists (particularly
Ricardo and Marx) with the concept of capital and the process of capital accumulation”
This would
be the case if the supplies of both types of shirts were �xed and nonadjust-
As Mises
(somewhat) con�rms in a discussion of �xed capital in the accounting books:
“What is alone decisive is whether, after covering all current operating costs and after
paying interest on the circulating capital, there is still so much left over from the gross
revenue that something more can be reaped than an adequate return on the value which,
after discontinuation of the enterprise,
the �xed capital would have in view of the possi
bility of using it for other production
(occasionally this will be only the scrap value of the
machines and bricks). In that case the continuation of the enterprise is more pro�table
than its discontinuation. If the �xed capital has a higher book value than corresponds to
its present and probable future earning capacity, then the book value must be lowered
to that extent” (2003, p.
245, emphasis
added). While discussing “inconvertible capital”
Mises surely did not mean “all past capital,” but only capital that was malinvested. The
above quote con�rms that �xed costs are part of investment (and pricing) considerations.
They cannot be brushed aside in favor of a narrow focus on marginal costs.
Blank, Steven. 2013. “Why the Lean Start-Up Changes Everything.”
Harvard Business
Eugen. 1962. “The Ultimate Standard of Value.” In
“Value, Cost, and Marginal Utility.”
Quarterly Journal of Austrian Econom
Eiteman, Wilford J. 1947. “Factors Determining the Location of the Least Cost Point.”
Eiteman, Wilford J., and Glenn E. Guthrie. 1952. “The Shape of the Average Cost Curve.”
Fabiani, Silvia, Martine Druant, Ignacio Hernando, Claudia Kwapil, Bettina Landau,
Claire Loupias, Fernando Martins, Thomas Matha, Roberto Sabbatini, Harald Stahl, and
Ad Stokman. 2006. “What Firms’ Surveys Tell Us About Price-Setting Behavior in the
Euro Area.” MPRA Paper No.
R. L., and C. J. Hitch. 1939. “Price Theory and Business Behaviour.”
Oxford Eco
Hayek, Friedrich August. 1940. “Socialist Calculation: The Competitive ‘Solution.’
Herbener, Jeffrey. 1996. “Calculation and the Question of Arithmetic.”
Review of Austrian
“Time and the Theory of Cost.” In Matthew McCaffrey, ed.,
The Economic
Theory of Costs: Foundations and New Directions
. Abingdon, UK: Routledge, pp.
Huerta de Soto, Jesus. 2010.
Socialism, Economic Calculation and Entrepreneurship
Cheltenham, UK: Edward Elgar.
Kenyon, Peter. 1978. “Pricing in Post-Keynesian Economics.”
Lange, Oskar. 1956.
. New York: McGraw-Hill.
Langlois, Catherine. 1989. “Markup Pricing Versus Marginalism: A
Controversy Revis-
Marc. 2014.
Post-Keynesian Economics: New Foundations
. Cheltenham, UK:
Edward Elgar.
“Frederic Lee and Post-Keynesian Pricing Theory.”
Lee, Fred. 1986. “Post Keynesian View of Average Direct Costs: A
Critical Evaluation
the Theory and the Empirical Evidence.”
8 (3):
1. “Old Controversy Revisited: Pricing, Market Structure, and Competition.”
MPRA Working Paper.
Machaj, Mateusz. 2013. “Sunk Costs and Contestable Markets.”
Quarterly Journal of Aus
Machlup, Fritz. 1946. “Marginal Analysis and Empirical Research.”
Martins, Fernando. 2005. “The Price Setting Behaviour of Portuguese Firms: Evidence
From Survey Data.” European Central Bank Working Papers Series no.
frey, Matthew. 2015. “Economic Policy and Entrepreneurship: Alertness or Judg-
ment?” In P. Bylund and D. Howden, eds.,
The Next Generation of Austrian Econom
ics: Essays in Honor of Joseph T. Salerno
. Auburn, AL: Ludwig von Mises Institute,
Jordan. 2010. “New Keynesians Versus Post Keynesians on the Theory of
Mises, Ludwig. 1966.
Human Action. A
eatise on Economics
. Chicago: Yale University
Calculation in the Socialist Commonwealth
. Auburn, AL: Lud-
Problems of Economics
. Auburn, AL: Ludwig von Mises
Salerno, Joseph. 1990. “Ludwig von Mises as Social Rationalist.”
Review of Austrian Eco
“The Concept of Coordination in Austrian Macroeconomics.” In
. Auburn, AL: Ludwig von Mises Institute, pp.
Nina, and Malcolm Sawyer. 2003. “Post Keynesian Price Theory.”
Shim, Eunsup, and Ephraim Sudit. 1995. “How Manufacturers Price Products.”
Walters, A. A. 1963. “Production and Cost Functions: An Econometric Survey.”
The transaction cost economics paradigm was born, in our opinion, as a byproduct
of the perfect competition revolution.
When a highly abstract
– and
– picture
of the economic process occupies the central place in the toolbox of
economic analysis, it is no wonder that every now and then a reverse tendency
toward reconnecting real-world phenomena and processes makes itself felt. It is
a rather obvious feature of the transaction costs paradigm that it depends, both
in its content and implications, on the background of the functioning of the price
mechanism (Coase, 1937). Coase himself considered the perfect competition con-
ceptualization of the price mechanism to be a valid starting point of analysis. Even
though he considers it to be incomplete, his goal is to perfect the model using the
scalpel of transaction costs rather than to replace it entirely. This seems to avoid
the main problem, however, because if something more (or rather, something
else) than transaction costs has been lost on the road from the real market process
to the Neverland of perfect competition and general equilibrium, then restoring
transaction costs to the system amounts to little more than hoping to rescue a
In what follows, we ask a few fundamental questions concerning the transaction
costs paradigm, its validity, internal consistency, and degree of compatibility with
Austrian economics. If the impression is that we have undertaken a critique
– even
harsh one, here and there
– we
will not consider that our intentions have been
misconstrued or misunderstood. Transaction costs economics has acquired much
fame and many followers, has been directly and positively sanctioned through
the Nobel Prizes of Ronald H. Coase (1991) and Oliver E. Williamson (2009),
and informs a great deal of research in various �elds within economics, includ-
ing theories of the �rm (and economic organization generally), property rights,
institutions and contracts, the economic analysis of law, etc. Especially given this
prominence, an effort to set the record straight is in order. Obvious forgeries do
not pose a danger to a genuine art market, but rather skillful fakes. Likewise, the
most dangerous heresies are not the ones evidently at odds with orthodoxy, but
those which differ only slightly
e open the discussion on Coase and transaction cost economics by brie�y dis-
cussing three fundamental issues which we believe shed important general light
Mihai-Vladimir Topan
Mihai-Vladimir Topan
on the nature of this paradigm: the production-exchange (markets versus �rms)
dialectic, the nature of transactions, and the nature of transaction costs. We then
attempt to show that the transaction costs perspective raises more problems than
it solves in both �elds where Coase put it to work: the problem of social cost and
property rights, and the theory of the �rm and economic organization. We end by
suggesting that the concept has at most a heuristic validity, namely, in analyses
A false dialectic: production versus exchange
One of the most important economic legacies left by Coase is a strong dialectic
Outside the �rm, price movements direct production, which is co-ordinated
through a series of exchange transactions on the market. Within a �rm, these
market transactions are eliminated and in place of the complicated market
structure with exchange transactions is substituted the entrepreneur-co
, who directs production.
It is clear that these are alternative methods of
this dichotomy is analytically useful in the limited context of a Crusoe
model in order to properly understand the two fundamental concepts of produc-
tion and exchange, in a complex economy based on the extended division of labor
and the use of money it does not hold. It has nevertheless
– inadequately
in our
– morphed
into the so-called make or buy problem. Otherwise stated,
this problem contrasts “in-house production versus acquisition on the market,”
or �rms versus markets, or management versus transactions, or “markets versus
hierarchies.” Austrian economics, especially in the Misesian tradition, is well-
Ludwig von Mises was probably the �rst economist to understand the fact that
a complex economy based on the extended division of labor is impossible without
private property, exchange, and money (Mises, 1920, 1951 [1922]; Salerno, 1990,
49). His
contribution has been retained as part of the famous socialist calcula-
tion debate, in which he argued that in the absence of monetary prices for the
factors of production (i.e. the situation that obtains as a consequence of the aboli-
tion of private property over the means of production), socialism is impossible as
a viable economic system. There is, under socialism, no method for calculating
whether a certain (centrally planned) allocation of resources is
– retrospectively
– more
pro�table or economic than others. Nevertheless, this
Misesian impossibility theorem actually consists of three overlapping impossibil-
ity theorems: (1) the impossibility of a socialist (extended or complex) economy;
(2) the impossibility of a complex barter economy, even if based on private (dis-
persed) property;
and (3) the impossibility for the isolated individual to economi-
cally manage (construct, maintain, and expand) a large structure of heterogeneous
resources held in the form of undispersed private property.
One important implication of these theorems is that in the unhampered, mod-
ern, complex market economy, production and exchange cannot be meaningfully
disentangled or separated: neither theoretically, nor operationally. Production in
the market economy involves exchanges, and exchanges involve or presuppose
production, albeit in various degrees. A
business �rm
consists in the actions of
the entrepreneurs who own and run it, employing means to achieve certain ends,
buying inputs to produce goods and services for actual and potential customers,
and selling them. Business activity consists of numerous actions, some of which
involve (interpersonal) exchanges, and some of which do not. And the actions
considered to be part “of the �rm” are constantly judged in terms of monetary
calculation using market prices. Moreover, whatever seemingly happens “inside
�rms” that cannot in any way be related to the rest of the market via monetary
Thus, the �rm is part and parcel of the market (Mathews, 1998). What happens
in the �rm, happens in the market, and for any �rm, what happens “inside” it, or
at least
as importantly
– “inside”
other �rms, is at the same time something that
happens “in (or on) the market.”
It is not, as Coase says, that the �rm “replaces”
or “supersedes” the market; rather, the �rm is the other side of the market, or the
market seen not only from the point of view of its main product/outcome (the
structure of prices), but more broadly in term of its structure and components.
We belabor the above in order to shed light on the starting point of Coase’s
1937 analysis. He intends �rst to establish a sharp dichotomy between “the price
mechanism” and “the �rm.” Thus, he accepts and treats as valid the idea that
the price system “works itself” (“Itself”? To who or what exactly does this term
refer?) and can allocate resources quite well
– an
implication being that the sphere
of this misty “price mechanism” contains no �rms.
Finally, he raises the ques-
tion of why there are �rms if this is the case. All this is intended to suggest that
there must be some “cost of using the price mechanism” that the �rm could avoid
or economize. Thus, the transaction cost paradigm is open for business. If this
procedure is contested, namely if we do not concede that the price system works
itself without �rms, then a more realistic
– and
– question
can be raised: in
what way are �rms involved in, or necessary for, the explanation of the function-
ing of the price system itself? Pursuing this route means explaining the �rm as a
constitutive part of the market. And transaction costs (as de�ned by Coase) can
play no part in this story.
To sum up, this false production-exchange dichotomy can be shown to be inad-
equate within the framework of Austrian economics by pointing to two theoretical
results. First, categorially, there is no fundamental difference between production
and exchange. Praxeologically, as Mises would say, any human action has the
structure of an exchange
– autistic
exchange or interpersonal (direct or indirect)
– involving
the giving up of a certain state of affairs in favor of another
that is expected to be more satisfactory. Both Mises and Rothbard speaks of “action
as an exchange” (Rothbard, 2009, p.
70; Mises,
2008, p.
97). Thus,
the general
category of costs, understood as opportunity costs of the actions undertaken by
human agents, cannot
– production
Mihai-Vladimir Topan
costs and exchange (or transactions) costs. They are simply part of the same
general category of cost with no substantive difference to set them apart.
ond, human action in complex modern economies is dependent on (dispersed)
private property, market exchanges, and monetary prices, all of which constitute
the preconditions for economic calculation. Any calculated action in this context
is exchange-dependent (even indirect or monetary exchange). Thus, no “outside
the market” sphere is available to make economically meaningful choices. This is
valid for factors of production in particular. Let us now move on to the transaction
What are “transactions” and “transaction costs”?
Ronald Coase never clearly de�ned or conceptualized the idea of a “transaction.”
At �rst glance, the “transaction” of which transaction costs economics speaks is
the same thing as interpersonal exchange. Nevertheless, closer inspection reveals
a more peculiar use of the concept. In passages such as “a �rm will tend to expand
until the costs of
organizing an extra transaction within the �rm
become equal
to the costs of carrying out the same transaction by means of an exchange on the
open market” (Coase, 1937, p.
395; emphasis
added), or “[as] more
are organized by an entrepreneur
, it would appear that the transactions would tend
to be either different in kind or in different places” (Coase, 1937, p.
397; emphasis
added), it becomes clear that the transactions Coase speaks of are not only inter
personal exchanges. They might appear in this form outside the �rm, but inside it
they are simply operations or actions performed by the entrepreneur or by some
one under him at his request. We are confronted here with an almost technological
view of transactions, or at least a hybrid view
– here
economical, there technologi
cal. Transactions are somehow objective bits of action or operations or tasks that
could be done inside or outside the �rm.
Later, this becomes explicit in the work
of Williamson. In a fragment titled “The Technology of Transacting” he adopts
Commons’ proposal that the transaction be made the basic unit of analy-
. attention
is focused on economizing efforts that attend the organiza-
tion of transactions
– where
a transaction occurs when a good or service is
transferred across a technologically separable interface. One stage of activity
(Williamson, 1996, p.
non-catallactic view of “transacting” renders the name of the paradigm
itself (“transaction cost economics”) a misnomer from the perspective of the tra-
ditional, economic view of exchange. This is even truer from the point of view of
Austrian economics. It opens up a technological, objectivist, mechanical, “real”-
cost view of the economic system which is impossible to treat in terms of human
As for the transaction costs concept, it was not coined as such by Coase in
the “The Nature of the Firm.” There, the speci�c idea is rendered as “the costs
of using the price mechanism” (Coase, 1937, p.
390; at
page 391 he uses inter-
changeably “costs” with “disadvantages” of “using the price mechanism”). Or,
sometimes more succinctly, the idea is expressed in terms of “marketing costs.”
In order to carry out a market transaction it is necessary to discover who it is
that one wishes to deal with, to inform people that one wishes to deal and on
what terms, to conduct negotiations leading up to a bargain, to draw up the
contract, to undertake the inspection needed to make sure that the terms of the
, he accepts the crystallization provided by Dahlman, for whom
transaction costs are “search and information costs, bargaining and decision
costs, [and] policing and enforcement costs” (Coase, 1988, p.
6). Gradually
, the
established view of transaction costs has come to analogize them with frictions
In mechanical systems we look for frictions: do the gears mesh, are the parts
lubricated, is there needless slippage or other loss of energy? The economic
counterpart of friction is transaction cost: for that subset of transactions
where it is important to elicit cooperation, do the parties to exchange operate
harmoniously, or are there frequent misunderstandings and con�icts that lead
to delays, breakdowns, and other malfunctions? Transaction cost analysis
entails an examination of the comparative costs of planning, adapting, and
(Williamson, 1996, p.
a seemingly speci�c and narrow category, transaction costs become
generalized. Thus, later interpreters and commentators are left with the task of
�guring out for themselves what to consider “transactions” and/or “transaction
costs.” In fact, almost anything can be interpreted as a transaction cost: asym-
metric information, incomplete contracts, bounded rationality, opportunism and
shirking, team organization, frequency of interaction, asset speci�city, bargaining
and searching, contract enforcement, administration or management, monitoring
At this point, it is useful to recapitulate Mises’s view on mechanical metaphors
The economist is often prone to look to mechanics as a model for his own
work. Instead of treating the problems posed by his science with the means
appropriate to them, he fetches a metaphor from mechanics, which he puts
in place of a solution. In this way the idea arose that the laws of catallactics
hold true only ideally, i.e. on the assumption that men act in a vacuum, as it
were. But, of course, in life everything happens quite differently. In life there
Mihai-Vladimir Topan
are “frictional resistances” of all kinds, which are responsible for the fact that
the outcome of our action is different from what the laws would lead one
to expect. From the very outset no way was seen in which these resistances
could be exactly measured or, indeed, fully comprehended even qualitatively.
So one had to resign oneself to admitting that economics has but slight value
both for the cognition of the relationships of our life in society and for actual
practice. And, of course, all those who rejected economic science for political
and related reasons
– all
the etatists, the socialists, and the interventionists
joyfully agreed
. Once
the distinction between economic and noneconomic
action is abandoned, it is not dif�cult to see that in all cases of “resistance”
what was involved is the concrete data of economizing, which the theory
comprehends fully.
risk of such a comprehensive view of transaction costs is that it is vulnerable
to reductions to the absurd. For instance, Hülsmann (2004, p.
49) identi�es
tion costs” as a friction in the economic system on which the economist’s attention
might be focused, and on the basis of which he could explain the nature of all sorts
Lastly, at this point, one must ask: what does the “zero transaction costs world”
really mean? We propose thinking of it in two forms: a strong version and a weak
version. The strong version of the zero transaction costs world is one devoid of any
“friction,” where everything works smoothly, un-problematically, even instantly.
In Coase’s own words, “where there are no costs of making transactions, it costs
nothing to speed them up, so that eternity can be experienced in a split second”
(Coase, 1988, p.
15). Insisting
upon using this version of the theory is not likely
On the other hand, if we try to construct the weaker version of the zero transac-
tion costs world, in which we try to abstract only from the categories that Coase
explicitly includes in the transaction costs category (searching and information,
bargaining and deciding, policing and enforcing), we may obtain more interesting
results. Thus, we argue below that property rights can also be proven necessary in
the zero-transaction world
– given
the weak interpretation
– and
that, therefore,
transaction costs are not as essential for understanding property rights as Coase
and his followers suggest. Likewise, if rationales for the existence of �rms can
be squared with the (weak) zero-transaction costs assumption, then the paradigm
also fails to be important for the theory of the �rm. This will be shown below.
Undermining property rights while explaining
Coase’s later works (1959, 1960) raise the problem of the best (most socially
productive or least socially costly) allocation of resources at the social level. In
his view, transaction costs seem to be just the right theoretical instrument to shed
light on the matter. Starting from the puzzling question of what is to be done in
those situations where the actions of a business �rm have harmful effects on oth-
ers, Coase departs from the traditional (Pigovian) manner of looking at things in
terms of an externality that must be internalized. Instead of judging the matter in
terms of aggressor and victim, where, for instance, smoke from a factory reaching
neighboring properties is either stopped, or moved, or �ned, or taxed, for Coase,
the problem should be seen more as “reciprocal” in nature, like a clash of equally
valid concerns. The problem, in Coase’s opinion, is not “who is right,” but rather,
The question is commonly thought of as one in which A
in�icts harm
on B
and what has to be decided is: how should we restrain A? But this is wrong.
We are dealing with a problem of a reciprocal nature. To avoid harm to B
would in�ict harm on A. The real question that has to be decided is: should
be allowed
to harm B or should B be allowed to harm A? The problem is to
taking the example of the cattle that stray into neighboring crops, Coase
frames the problem as one of social choice. If the cattle are allowed to destroy the
crops, there is more meat and fewer crops. If they are not allowed, there will be
less meat and more crops. This imposes a social choice of meat versus crops, a
choice that presupposes that “we know the value of what is obtained as well as the
value of what is sacri�ced to obtain it” (Coase, 1960, p.
2). The
so-called Coase
Theorem is then brought to bear on the discussion. With zero transaction costs, it
does not matter what the initial allocation of property rights is (as long as there is
one), for the �nal allocation of resources will be the same (the strong version of
the theorem). Or, even if it is not the same, it will maximize the value of produc-
tion (the weaker version of the theorem).
Thus, in the world of zero transaction
costs the free interaction of people on the market, whatever the initial allocation
However, if transaction costs are positive, or high, the free interaction of peo-
ple on the market, given a certain initial allocation of property rights, does not
automatically maximize production. Institutions other than the free market are
required if wealth is to be maximized. As Hülsmann (2004) observes, these insti-
tutions might be �rms, or governmental institutions like courts and legislation. If
the ideal is to approximate
– with
the help of transaction cost-economizing institu-
tional instruments
– as
closely as possible the outcomes of a zero transaction cost
world, then it follows that the role of these institutions is precisely (and sometimes
it just seems to be so positively) to put resources in the hands of those who would
This perspective raises a number of important problems. First, in judging inter-
personally whether meat is more important “socially” than crops, Coase adopts an
objectivist perspective, or at least one based on an objective theory of value. The
value of what is lost and the value of what is gained in such cases can be com-
pared and balanced against one another only if they are the same in the eyes of
Mihai-Vladimir Topan
all parties (the two parties involved and any outside observers). James Buchanan
It is unfortunate that Coase presented his argument (through the examples)
largely in terms of presumably objectively-measurable and independently-
determined harm and bene�t relationships. In his formulation, these rela-
tionships become
in the perception of all parties to any potential
exchange of rights. Hence, the unique “ef�cient” (bene�t maximizing or loss
minimizing) allocation of resources exists and becomes determinate concep-
tually to any external observer. The ef�cacy of free exchange of rights in
attaining the objectively-determined “ef�cient” outcome becomes subject to
testing by observation. The exchange process, in this perspective, is itself
evaluated in terms of criteria applied to the outcomes that the process is
observed to produce. There are values inherent in allocations that exist quite
critics of Coase (Block, 1977; Rothbard, 1982; North, 1992, 2002)
have also argued this point forcefully, emphasizing the subjectivity of costs and,
especially, indicating possible psychic or non-monetary income or bene�ts which
cannot be taken into account by third parties, including courts (and that thus would
be unjusti�ably excluded from social welfare). Nevertheless, we feel that these
If we adopt the weak version of the zero-transaction cost economy described
above, which aims at a modicum of relevance for the real world, it can be shown
that the initial allocation of property rights matters even in this world. Bearing
in mind all the above considerations, if zero transaction costs were interpreted
to mean not that the parties will instantly and effortlessly reach an agreement
(which is tautological: parties get along in a zero transaction costs world because
zero transaction costs would mean precisely that parties get along), but only that
they will instantly �nd out whether their views and judgments concerning certain
resources coincide or are compatible,
then a situation is conceivable in which,
with zero transaction costs, one person would use a resource one way, and another
person, another way. Thus, given the
different judgments
and the subsequent
mutually exclusive uses
for a certain piece of property, the initial allocation of
property rights over the said resource becomes essential. Crucially, transaction
costs play no part. Differences in judgments between persons in our scenario are
not necessarily con�ned strictly to non-monetary considerations. The differences
may contain the whole range of elements
– monetary
, non-monetary; short-term,
The special character of monetary entries into acting man’s economic calcula-
tions prevents us from even setting up the tables and numbers by means of which
the Coase theorem is usually discussed. A
the exchange ratios which we have to deal with are permanently �uc
tuating. There is nothing constant and invariable in them. They defy any
attempt to measure them. They are not facts in the sense in which a physicist
calls the establishment of the weight of a quantity of copper a fact. They
are historical events expressive of what happened once at a de�nite instant
and under de�nite circumstances. The same numerical exchange ratios may
appear again, but it is by no means certain whether this will really happen
and, if it happens, the question is open whether this identical result was the
outcome of preservation of the same circumstances or of a return to them
rather than the outcome of the interplay of a very different constellation of
price-determining factors. Numbers applied by acting man in economic cal
culation do not refer to quantities measured to but to exchange ratios as they
are expected
– on
the basis of understanding
– to
be realized on the markets
of the future to which alone all acting is directed and which alone counts for
Again, if the above is true, then it is not surprising to �nd that Mises is abso-
The height of conceptual confusion is reached when one tries to bring calcu-
lation to bear upon the problem of what is called the “social maximization of
pro�t.” Here the connection with the individual’s calculation of pro�tability
is intentionally abandoned in order to go beyond the “individualistic” and
“atomistic” and arrive at “social” �ndings. And again one fails to see and
will not see that the system of calculation is inseparably connected with the
individual’s calculation of pro�tability.
if we reach such a conclusion, then it means the essential element for explain-
ing property rights and related institutions is once again the scarcity of resources
and the impossibility for two agents, or two wills, to make use of the same
resource for mutually exclusive purposes. As we have explained in the discussion
above, transaction costs play no necessary role in this story (though in one case or
another, they could even be the most important issue).
Transaction cost-based explanations of the �rm: neither
necessary, nor suf�cient
Let us now turn to the transaction cost-based theory of the �rm. Given the elu-
siveness of “transactions” and “transaction costs,” it is not surprising that �rm
activity very often relates to them. But the claim of transaction cost economics
especially of
– is
not a weak one, of the “there are certain connections
between transaction costs and �rms” sort. Rather, it is a strong thesis: transaction
costs themselves and the need to economize on them fundamentally
�rm, revealing its
, or essence. The implication is that the phenomenon
cannot otherwise be understood. “It was the avoidance of carrying out transac-
tions through the market that could explain the existence of the �rm in which
Mihai-Vladimir Topan
the allocation of factors came about as a result of administrative decisions (and
think will
be considered in the future to have been the most important
contribution of this article is the explicit introduction of transaction costs into
economic analysis. I
The Nature of the Firm
that the existence of
transaction costs leads to the emergence of �rms.
e will argue in the next few pages that there is no such thing as a valid or coher-
The discussion is important because the �eld has grown enormously follow-
ing the contributions of Oliver Williamson. Explanations of the multinational,
or transnational, or international �rm have also been coined, in part or in full, in
terms of transaction costs (Hymer, 1990[1968]; Hennart, 1977; Teece, 1985). In
hindsight, we can look back at these efforts with feelings similar to those sug-
gested by C.
S. Lewis
The Screwtape Letters
, in which he mocked the academic
habit of asking irrelevant or inconsequential questions. In fact, any question at all
can be asked, other than the one that matters: whether or not a theory is true. In
the present case, almost the whole transaction costs research program has ended
up asking whether this or that explanation is one in terms of transaction costs, and
if so what speci�c type thereof, whether transaction cost explanations were coined
earlier or later, and so on. The only question that is not asked anymore is whether
the transaction costs theory of the �rm makes sense, or otherwise said, if it is true.
Let us now proceed to give a few reasons why the Coasean transaction-based
It is the essence of serfdom that the price mechanism is not allowed to oper-
ate. Therefore, there has to be direction from some organizer. When, how-
ever, serfdom passed, the price mechanism was allowed to operate. It was not
until machinery drew workers into one locality that it paid to supersede the
price mechanism and the �rm again emerged.
Coase, the relations inside a �rm are somewhat ambiguous. Even though
sometimes the idea is that they must be completely voluntary and contractual
(albeit of a special contractual nature), quite often the same relations are presented
as equivalent to relations of power or authority. The serfdom reference is a case in
point, as is the example of the master-servant relation. Moreover, all the explicit
identi�cations of planning in the �rm with planning at the national level (e.g. “a
�rm, that little planned society”; Coase, 1992, p.
716) fall
into the same category.
Coase does not have a theory of property rights and therefore cannot properly set
or identify the boundaries of the market economy. He therefore ends up searching
(and often �nding) the �rm ambiguously on both sides of that frontier. As Bylund
rightly points out, the context of the socialist calculation debate and Coase’s early
If the avoidance of costs of using the price mechanism is, in a strong sense,
realized only in non-voluntary settings (e.g. slavery, serfdom, socialism), then
what Coase has obtained is not a transaction costs-based theory of the �rm, but a
transaction cost-based theory of these aggressive or coercive institutions. The vol-
untary cooperation-based business �rm, which cannot so strongly supersede the
price mechanism
– if
it can at all
– remains
unexplained. Moreover, even if both
phenomena are explained in terms of transaction costs, it still remains for Coase
to provide us with a theory of the regular, non-coercive, business �rm. One might
also suspect at this point that the supersession of the price mechanism, far from
The same problem can be found in the works of Williamson, whose view
evolves from a market-versus-hierarchy dichotomy to the idea of a spectrum of
governance, as suggested in both
The Economic Institutions of Capitalism
liamson, 1985) and
Mechanisms of Governance
(Williamson, 1996). Yet even in
the latter theory Williamson includes in the spectrum of governance everything
from spot contracts to hybrid ones, to �rms, and �nally, to legislation and bureaus
(his term for the state sector).
All of these are explainable in terms of transaction
costs economies. Yet if this is correct, what explains the business �rm in particu-
lar? We are of the opinion that no adequate answer has been provided.
Apart from the ethical problems involved, the difference between voluntary
and non-voluntary structures is that in voluntary ones (e.g. regular business
�rms) all parties involved participate through the speci�c contracts and prices
they negotiate. Through this process they create what Mises calls the “intellectual
division of labor,” which makes possible the formation, maintenance, and con
stant revolution of the structure of monetary prices on the market. That is why
all the components of the �rm and all the participants in it are unavoidably con
nected via monetary calculation to the market environment outside the �rm. The
�rm does not supersede this market, but rather is a part of it. Serfs or slaves, on
the other hand, are not connected to the price mechanism as they cannot negotiate
the terms of their employment, and therefore do not participate in the abovemen
tioned intellectual division of labor. Their masters exercise authority or command
over them. But the position of the employee in the �rm puts him at the disposition
of the managers or bosses no more than a plumber, or a dentist, is at the disposi
Transaction costs prove too much
Another reason for considering that a transaction cost explanation of the �rm
has not actually been provided by Coase is the fact that he (and also Williamson)
gradually developed what might be called transaction cost imperialism. Not only
Mihai-Vladimir Topan
�rms, but the market itself, along with money and all the institutions of capital-
Markets are institutions that exist to facilitate exchange, that is, they exist
in order to reduce the cost of carrying out exchange transactions. In an eco-
nomic theory which assumes that transaction costs are nonexistent, markets
In fact, a large part of what we think of as economic activity is designed to
accomplish what high transaction costs would otherwise prevent or to reduce
know of only one part of economics in which transaction costs have been
used to explain a major feature of the economic system, and that relates to
the evolution and use of money
. The
bene�ts brought about by the use of
suggests that the effects of transaction costs are “pervasive in the econ-
omy.” The problem is that if transaction costs explain everything, they end up
Firms in the zero-transaction cost world
As in the previous section, we will end by pointing out the important possibility of
the existence of �rms in a zero transaction costs world. If this world is not simply
tautologically interpreted to mean that any two parties taken together will have the
same view or judgments concerning all relevant aspects concerning the immediate
or more remote future, then �rms make sense. Various potential business partners
in a �rm
– entrepreneurs
– might
discover (effortlessly, or instantly, as imposed by
the exigency of zero transaction costs; once more, the weak version) that their busi
ness plans and judgments are irreconcilable. Thus, the only possible path consists of
establishing different �rms. Either both partners establish �rms, or one of them does
while the other refrains. Thus, if we corroborate this insight with the praxeological
analysis of entrepreneurship as speci�c business project uncertainty-bearing, which
can implicitly manifest itself only by means of �rms, we have arrived at a severe
verdict concerning the importance of transaction costs for the theory of the �rm: they
The missed opportunity for a better explanation of the �rm
Ironically enough, Coase provided what, from the point of view of Austrian
economics, could be an adequate theory of the �rm. Thus, he claims that “[a]
�rm, therefore, consists of the system of relationships which comes into exist
ence when the direction of resources is dependent on an entrepreneur” (Coase,
1937, p.
393). Of
course, the resemblance with the Austrian theory of entrepre
neurship is only super�cial. For Coase, entrepreneurship is nothing more than
another factor of production, namely, management or administration, and he
even strives to take this factor out of the normal functioning of the market and
instead point to its similarities to socialist central planning. Here, again, we
join Per Bylund’s interpretation of Coase, which views his discussion of the
nature of the �rm in the context of the socialist calculation debate, in particular
the search for arguments in favor of the possibility of rational central planning
(Bylund, 2014). Coase even touched on the Knightian view in which entrepre
neurship is functionally tied to uncertainty, much as in Mises’s theory. Sadly,
Coase dismissively collapsed the complex problem of uncertainty into one of
The question of uncertainty is one often considered to be very relevant to the
study of the equilibrium of the �rm. It seems improbable that a �rm would
emerge without the existence of uncertainty. But those, for instance, Profes-
sor Knight, who make the
mode of payment
the distinguishing mark of the
– �xed
incomes being guaranteed to some of those engaged in produc-
tion by a person who takes the residual, and �uctuating, income
– would
to be introducing a point which is irrelevant to the problem we are
considering. One entrepreneur may sell his services to another for a certain
sum of money, while the payment to his employees may be mainly or wholly
e should point out immediately that this dismissal of the uncertainty view is
obviously inadequate. First of all, the idea that an entrepreneur (who must bear
the uncertainty of a business unit) could “sell” entrepreneurship does not hold.
Either he sells the whole business and stops being an entrepreneur at all (the
buyer now assuming this role); or he simply offers some consultancy services or
advice to another entrepreneur, in which case he remains an entrepreneur in his
�rm, while the consultancy client remains the true (and sole) entrepreneur in the
consulted �rm. Moreover, the key to this issue is not so much the idea of the �xity
or variability of the income share (the “mode of payment”) as it is the incidence
of losses if they occur. Thus, even if an entrepreneur pays his employees with
variable sums (“shares in pro�t”), the crux of the matter is whether they also par-
ticipate in losses. If not, the mode of payment or the type of wages they receive
remains irrelevant, and does not make them entrepreneurs. But if the correct way
to pinpoint the role of uncertainty in business is to discover who bears the losses
rather than who receives income by means of a variable “mode of payment,”
the uncertainty theory of entrepreneurship and the �rm has not been adequately
criticized. Moreover, such a theory can be laid out with no reference whatsoever
Mihai-Vladimir Topan
– after
ting very close to wondering whether Coase simply rediscovered the concept of
cost (see note 8)
– locates
another delicate spot in transaction cost-based theories
of the �rm. Namely, if the idea of a “transaction” or “transactions” presupposes
that previously we have dealt adequately with the possible agents who engage in
them, then “transactions” require de�nitions in which the �rm concept is already
involved. Thus, the �rm is a more fundamental concept than a transaction, and is
presupposed by it. Explaining �rms in terms of transactions, therefore, amounts
One person phones another and directs him to purchase speci�c assets by
certain time if they can be acquired for less than a stipulated price. Is this
activity transacting or managing?
. Since
the call might be from an owner/
manager of a �rm to his employee in the purchasing department or from a
customer/investor to the brokerage house whose services he purchases, it is
hard to know whether we are dealing with a transaction or management cost
until we
already know
whether we are discussing a �rm or a market
. The
dif�culty is that the same organizing activities often characterize
As a theoretical concept, we believe “transaction costs” remain elusive, while
transaction cost economics is a chameleonic instrument which raises more ques-
tions than it solves. In situations where transaction costs can be meaningfully
– as
what Mises, echoing Weber, called ideal types
– reasoning
terms of transaction costs is possible and can yield relevant results. Thus, if in a
certain context, “transaction costs” are unambiguously interpreted to mean “bar-
gaining costs,” then any solution that implies, other things equal, a reduction in
those costs, is bene�cial. If, for instance, in an internally divided village the priest
makes the effort to convince all the villagers to participate in a certain project
(building a school, rebuilding the destroyed house of a widow, etc.) and succeeds,
we cannot say this justi�es developing a transaction cost-based theoretical expla-
As for Austrian economics and possible cross-fertilizations with transaction
cost economics, we believe these raise more problems than they solve. Based
on the praxeological paradigm, which looks at economic phenomena in terms of
human action, takes entrepreneurship (seen mainly in terms of uncertainty-bearing
by property-owning capitalist-entrepreneurs who undertake speci�c investment
projects) as its core, being subjectivist and employing the all-important concept of
economic calculation for understanding the major aspects of the complex modern
monetary economy based on the extended social division of labor, Austrian eco-
nomics can handle adequately on its own terms whatever puzzles or conundrums
the transaction cost paradigms raises. On the other hand, the occasions for imports
from the latter are as many occasions for error. For, in the end, is there such a thing
as a “cost of using the price mechanism”? Perhaps the question is meaningless
For a study of the impact of perfect competition theory on economics in general, see
Coase include
s the following footnote at this point: “In the rest of this paper I
shall use
term entrepreneur to refer to the person or persons who, in a competitive system,
Even though
traces of the idea are present, for instance, in
The Theory of Money and
(1912) and
Epistemological Problems of Economics
(1933), it is in
(1949) that Mises speaks explicitly of “the barter �ction,” harshly criticizing
the idea that the phenomena of the market economy (mainly the formation of prices)
can all be explained in terms of direct exchange, with the subsequent addition of mon-
etary aspects re�ecting only minor complications that do not alter the discussion in a
fundamental way. For the important implications of the barter �ction and its corollary,
“neutral money,” see Salerno (2016).
Options (1)
and (3) are included by Mises under the heading of “the autistic economy”
Thus, properly
understood, the “make or buy” decision is not a Coasean one of market
versus �rm, but a more complex choice between one mixed alternative (involving
some exchanges and some other operations) and another (involving different oper-
ations and different exchanges). The decision of a �rm to buy a piece of furniture
instead of making it “in house” does not involve a more enhanced use of the market
and a less intensive use of management or non-interpersonal exchange operations. It
means that some combination of transactions and operations (buying the factors of
production needed to build the furniture and then handling both them and the furniture)
is replaced by another combination of transactions and operations (buying ready-made
furniture and handling it). “Make or buy” is actually more akin to a choice between
“(make plus buy) 1” and “(make plus buy) 2,” both options being made comparable
For the
view that Coase, and Williamson after him, did not mean to create a strong
�rm-market dichotomy, see Klein (2010), p.
Even if
someone were to claim that in a market without �rms only individuals would
trade and engage in productive activities, in our opinion, the correct framing of the
situation is not “here is the market without �rms,” but “here is a market with exclu-
sively unipersonal �rms.” Then, a possible puzzle of the Coase type would be: why are
This is
echoed by Oliver Williamson in his well-known saying, “in the beginning
there were markets.” The idea of the “market without �rms” (or we could say, with a
Misesian twist, “the imaginary construction of the �rm-less market”) has to a certain
extent been entertained by authors in the Austrian tradition, such as (most recently) Per
The Theory of Production. A
Harold Demsetz
is close to deriving such a conclusion when
– in
his attempt to rig-
orously make sense of the alternative implied in Coase’s reasoning in terms of the
combined bargaining and management costs of both �rms and markets
– he
Mihai-Vladimir Topan
somewhat puzzled: “Have we come to the point of saying that �rms are used when
they are cheaper, all costs considered, but not when markets are cheaper?” (Demsetz,
Thus, a
little later in his article Coase writes: “[at] the margin, the costs of organizing
within the �rm will be equal either to the costs of organizing in another �rm or to the
cost involved in leaving the transaction to be ‘organized’ by the price mechanism”
(Coase, 1937, p.
404). Thus,
not only do we have “transactions in the �rm”; we have
“organization outside the �rm,” on the market. In what sense do �rms/entrepreneurs
do not offer any numerical examples here, as these have been more adequately
In terms
of the Dahlman enumeration, for instance: the parties would almost instantly
�nd each other and all relevant information about them; they could instantly bargain
and decide (we resist the temptation to equate costless bargaining with successful cost-
less bargaining; costless bargaining can occur only for the parties who bargain to dis-
cover their preferences are incompatible). Should an agreement be reached, the parties
could very easily or quickly police and enforce the contract. As stated above, we want
to avoid the tautological stipulation of the zero transaction costs economy as a world in
which parties instantaneously agree on everything. Perhaps some elements of human-
ity, individuality, subjectivity, and entrepreneurship can be retained. In this respect,
our weak version of the zero transaction cost economy must be understood as differing
from perfect foresight, certainty, and even the Misesian Evenly Rotating Economy.
This is
the way we believe Rothbard’s discussion of the limits that economic calcula-
tion imposes on �rm size is to be interpreted. This argument is anything but Coasean in
spirit. Where the �rm begins for Coase, for Rothbard it actually ends: “The difference
between the State and the private case is that our economic laws debar people from
ever establishing such a system in a free-market society. Far lesser evils prevent entre-
preneurs from establishing even islands of incalculability, let alone in�nitely com-
pounding such errors by eliminating calculability altogether” (Rothbard, 2009, p.
By 1960,
Coase also mentions governments, or “super-�rms
. of
a very special
Given this
discussion, we think that Klein’s interpretation
– that
Coase and Williamson
should be read not as opposing the �rm to the market, but rather the
to the
In a
sense, what has just been said is a piece of mental gymnastics: if the use of the
price mechanism is costly and �rms are meant precisely to avoid these costs, saying
that markets exist to reduce transaction costs is tantamount to saying that they exist to
reduce their own costs! The only way out of this problem is to consider that the price
mechanism is something abstract, independent from the market. Using this mechanism
is costly in an absolute sense. A
�rst step
toward reducing its cost is through markets;
another is through �rms; and a �nal one through governments. This could make sense
only with the crudest objective theory of value. Bylund (2014) argues that Coase has
in mind allocations of resources that would be identical, irrespective of where they
would �nally be undertaken (markets, �rms, governments, etc.). He only posits that
the option chosen must include economies in terms of transaction costs. This approach,
once again, has a very objectivistic �avor. It could also explain the puzzling fact that
Coase rarely, if ever, speaks of
, the true compass of economic activity (not costs
It is
also somewhat ironic that Coase undertakes this analysis of Knight’s views in
order to show that the latter, focusing on a seemingly
mode of payment, can-
not “give a reason why the price mechanism should be superseded” (Coase, 1937,
401), and
therefore has not, presumably, succeeded in establishing the need for the
�rm. Again, this suggests the market-�rm dichotomy Coase has in mind, but which is
stipulated or simply postulated more than proved. This whole discussion should speak
in favor of Knight, who provides a rationale for the �rm based on uncertainty-bearing
entrepreneurship and which, in addition, is fully integrated in the market and in the
Block, Walter. 1977. “Coase and Demsetz on Private Property Rights.”
1 (2): 111–115.
Buchanan, James. 1999 [1984]. “Rights, Ef�ciency, and Exchange: The Irrelevance of
Transactions Cost.” In
The Logical Foundations of Constitutional Liberty
, Collected
Works of James M. Buchanan, Vol. I. Indianapolis, IN: Liberty Fund, pp.
Per L. 2014. “Ronald Coase’s “Nature of the Firm” and the Argument for Eco-
Problem of Production: A
New Theory
of the Firm
. London: Routledge.
“The Federal Communications Commission.”
Journal of Law and Economics
The Firm, the Market and the Law
. Chicago: University of Chicago Press.
“The Institutional Structure of Production.”
Demsetz, Harold. 1988. “The Theory of the Firm Revisited.”
Journal of Law, Economics,
and Organization
Foss, Nicolai J., and Klein, Peter G. 2012.
Organizing Entrepreneurial Judgment: A
Approach to the Firm
Jean-François. 1977.
A Theory of Foreign Direct Investment
. PhD thesis. https://�les/749290/THEORYOF.PDF.
Hülsmann, J. G. 2004. “The A
Priori Foundations
of Property Economics.”
nal of Austrian Economics
Hymer, Stephen H. 1990 [1968]. “The Large Multinational ‘Corporation.’
” In
Mark Cas-
. Hants: Edward Elgar, pp.
Peter G. 2010. The Capitalist and the Entrepreneur: Essays on Organization and
Markets. Auburn, AL: Ludwig von Mises Institute.
Machovec, Frank. 1995. Perfect Competition and the Transformation of Economics. Lon-
Mathews, Don. 1998. “Management vs. the Market: An Exaggerated Distinction.”
terly Journal of Austrian Economics
Mises, Ludwig von. 1951 [1922].
Socialism: An Economic and Sociological Analysis
. New
Haven, NJ: Yale University Press.
. Auburn, AL:
Epistemological Problems of Economics
. 3rd ed. Auburn, AL: Lud-
Action: A
eatise on Economics
. Scholar’s ed. Auburn, AL: Lud-
North, Gary. 1992.
The Coase Theorem: A
Study in
Economic Epistemology
. Institute for
“Undermining Property Rights: Coase and Becker.”
Rothbard, Murray N. 1982. “Law, Property Rights and Air Pollution.”
2 (1):
Mihai-Vladimir Topan
Man, Economy, and State
“The Development of the Neutral Money Concept: The Austrian School
Further Dehomogenized.” January
30, 2016.
Teece, David J. 1985. “Multinational Enterprise, Internal Governance, and Industrial
Organization.” American Economic Review: Papers and Proceedings of the Ninety-
Seventh Annual Meeting of the American Economic Association 75 (2): 233–238.
Topan, Mihai Vladimir. 2013.
Întreprizătorul în �rma internaţională. O teoretizare în
tradiţia Şcolii Austriece
(The Entrepreneur in the International Firm: A
Theorization in
Austrian School Tradition). Editura ASE, Bucharest.
Outside the Extent of the Market: A
Review of
Per Bylund’s ‘
Problem of Production: A
Oliver E. 1985.
The Economic Institutions of Capitalism: Firms, Markets
. New York: The Free Press.
. New York: Oxford University Press.
After Ludwig von Mises published his essay on economic calculation and the
impossibility of planning in a socialist economy (Mises, [1920] 1935), econo-
mists engaged with and debated his ideas for nearly three decades. The socialist
calculation debate, as it is often referred to, was “one of the most signi�cant con-
troversies in modern economics” (Blaug, 1997, p.
557). It
was also a controversy
in which Austrian economists played a signi�cant role, both as instigators and as
consistent defenders of the market process perspective in economics. Ultimately,
however, the Austrians were not victorious; instead, many of the assumptions and
views held by their opponents have been integrated into the formal models of
Austrians stick to their guns, however, and claim their argument remains the
better one. In fact, there is an argument to be made that the opponents of Mises
and the other Austrians failed to understand their arguments. Rather than view-
ing economic calculation as an issue of computing power, or of the dif�culty of
collecting and aggregating dispersed information, and consequently the limited
ability of the state to
– a
single owner
(the state) with dispersed control. Under such a system, the economy would have
one principal but a myriad of agents acting independently, and would therefore
act in the same way as modern-day corporate capitalism with its passive owners
Management is what’s wr
(shareholders) and active management (Berle and Means, 1932). Despite hav-
ing common ownership through the state, Denis’s system would function like a
market and determine real market prices through individual competing producers’
While Denis’s approach is clever, I
maintain that
it offers no real challenge
to Mises’s argument. Instead, I
that Denis makes the same fundamental
mistake as other critics of Mises’s argument, as well as of the economic positions
found in the Austrian tradition more generally. This mistake, which I
on in
this chapter, is to focus on
– and
indeed even theorize from the point of
view of
– cost,
and therefore from the perspective of a closed or, at a minimum,
a non-expanding economic system. While cost is an important
– if
not the most
– aspect
of effective management and of efforts to improve business
processes and pro�tability, as well as a core concept in economic theorizing, my
claim is that it is nevertheless of secondary importance for market-creating
argument suggests that the proper way of understanding the Misesian argu-
ment, and therefore why a socialist economy is truly impossible and not simply a
computational dif�culty or a practical obstacle, is to focus on the value-creating
nature of entrepreneurship and, consequently, how the function of entrepreneur-
ship treats cost. I
will show that
cost is of secondary importance in a free market,
whereas in a planned economy it is and must be the main, if not the only, vari-
able considered. In what follows, I
discuss the
characteristics of entrepreneurial
decision-making, contrast entrepreneurship and management, and note the differ-
ence between open-endedness and closed-endedness in coordinating economic
activity. I
in line with Mises ([1949] 1998; cf. Salerno, 1990a), that the only
possible outcomes for an economic system are growth or decline, and that these
follow directly from the prevailing “driving force” of the market: entrepreneur-
Entrepreneurial decision-making
There are many de�nitions of entrepreneurship, but a common theme is the under-
taking of uncertain endeavors with the potential to create value in the market
(Cantillon, [1755] 2010; Knight, [1921] 1985). As such, entrepreneurs coordi-
nate efforts that aim to exploit a speci�c situation that they imagine can (or will)
emerge at some point in the future. Entrepreneurial decision-making is conse-
quently directed toward the creation of a speci�c value, often in the form of a new
product or service for which they believe there will be suf�cient market demand.
What is notable about entrepreneurship is that neither the state of supply nor of
demand in that aimed-for situation can be known in the decision-making present,
and therefore both are uncertain but
– or
so the entrepreneur believes
– imaginable
When deciding to pursue an opportunity, the entrepreneur is ignorant of how
the undertakings of other entrepreneurs, as well as other factors endogenous to the
The management problem of socialism
market and purely exogenous effects, may in�uence the imagined market situa-
tion. Therefore, the entrepreneur cannot know that he will be able to sell what is
hoped or planned. In fact, the imagined opportunity may turn out to be exploited
by other entrepreneurs while “our” entrepreneur is still procuring, coordinating,
and combining the means of production. It may also be the case that the imagined
opportunity is undermined by other and seemingly unrelated innovations in the
market that change consumer behavior or affect their preference rankings in other
ways. Furthermore, consumers’ real interest in the offered product will be affected
by fashions, fads, and temporary hype, as well as their choice to consume or save
Nevertheless, the entrepreneur aims for this situation because he estimates that
the value of the product or service to consumers will be high enough to warrant
a selling price that makes the undertaking worthwhile for him. The actual price
at which the good can be offered is not ultimately chosen or set by the entrepre-
neur, but is ultimately determined by consumers’ subjective expectation-based
valuation of the good in question as compared to other goods (Menger, [1871]
2007; Mises, [1949] 1998, pp.
328–329). The
entrepreneur must speculate about
what he will be able to charge for the anticipated good. Indeed, to be success-
ful the entrepreneur must estimate a revenue-maximizing price
– a
price that is
not too high, but that the market (that is, consumers) will bear. This price offers
the best chance of succeeding in the undertaking, since it combines a price level
that leaves enough consumer surplus to attract customers in suf�cient numbers to
make the opportunity worthwhile to the entrepreneur.
After estimating the revenue potential of his opportunity, and therefore its
potential “value,” the entrepreneur next chooses how best to exploit it. In other
words, he chooses a cost structure for production that offers suf�cient �exibil-
ity and scalability for the nascent enterprise to make the entrepreneur and any
investors comfortable with the chances of success. This choice is limited by the
estimated selling price but guided by the available market prices of the factors
of production, which in a market
– through
entrepreneurial bidding
– tend
approximate the social opportunity cost of the available varieties of the means of
production. There is in this situation no way for the entrepreneur to “maximize”
the outcome with precision, as his calculations are based on many unknown and
uncertain variables, the exact values of which are at best imaginable and thus
subject to the entrepreneur’s business judgment (Knight, [1921] 1985). Indeed, it
is even unknown ex ante which variables should be considered important in the
undertaking and therefore identi�ed and assessed using the entrepreneur’s judg-
The entrepreneur can only rely on existing factor market prices for guid-
ance in his choice of cost structure, which, along with his own revenue estimates,
It should be clear from the brief overview above that entrepreneurial decision-
making is in many ways better described as an art than an exact science. The
potential pro�tability of the different alternatives is subject to approximations
and conjectures that are ultimately based on the entrepreneur’s judgment; any
calculations are therefore not exact but only estimates. The best information avail-
able to the entrepreneur in estimating the value of alternative courses of action is
Value as a basis for factor prices
Without market prices for the factors of production, the entrepreneur is blind
regarding the tradeoffs between different uses of productive resources and how
they contribute to consumer value. Whereas the entrepreneur must carry out the
task of comparing different cost structures in order for his intended product to
be offered to consumers, he is unable to compare the social value of different
types of goods and services potentially offered in the market by entrepreneurs
collectively. Part of the reason is the dispersed, fragmented, and tacit knowledge
of “the particular circumstances of time and place” (Hayek, 1945, p.
521), which
decentralized decision-making or, at a minimum, decentralized collec-
tion of data for aggregation. Yet another part relates to the entrepreneur’s speci�c
personal expertise and thus, his inability to comprehend the needs and possibili-
ties available in the market. These possibilities are better understood
– and
will in
fact be revealed and discovered by individual entrepreneurs
– as
an “intellectual
division of labor” among entrepreneurs pursuing pro�t as they understand it, a
division that allows their appraisements to be assessed through the working of the
competitive market process (Mises, [1920] 1935, p.
102; Salerno,
1990b, p.
The social
opportunity cost of different courses of action becomes available to
the entrepreneur through this process, but would be unavailable without factor
These prices emerge from entrepreneurs bidding for resources they need to
complete their planned production projects. As each of the individual entrepre
neurs has only one or a set of very few speci�c value targets he is aiming for,
each can estimate, using economic calculation, the cost burden that their speci�c
undertakings can support without generating losses. Calculations are still uncer
tain, as they are based on the entrepreneur’s “guesstimates,” based at least in
part on tacit information in the form of experiential knowledge, impressions, and
beliefs about how the market will progress in the future. The expected revenue
of a project provides some guidance as to what may be the maximum cost pos
sible while still generating suf�cient pro�ts, and present prices provide important
information about factors’ relative uses. These present prices result from other
entrepreneurs’ actions, which in turn were taken based on those entrepreneurs’
anticipations of future exchange ratios (Mises, [1949] 1998, p.
therefore re�ect the market’s approximated opportunity costs. These prices are
taken as inputs by budding entrepreneurs and are an “auxiliary for entrepreneur
ial understanding of the future course of prices,” but because they are historic
they are also “by no means indispensable” for economic calculation (Salerno,
the bidding process is always ongoing, prices are subject to change as entre-
preneurs revise their judgments or exit, and new entrepreneurs enter to place bids
The management problem of socialism
based on their estimates. Each entrant entrepreneur, with a speci�c project in
mind, seeks to exploit his anticipated future exchange ratios, and he thus chooses
the cost structure most promising for the planned endeavor
– and
as a result, bids
for those resources. The entrepreneur’s bids are limited by the revenue he expects
to receive from the output of the new production process, adjusted for his pro�t
requirement and compensation for felt uncertainty in the estimates he devised. It
follows that if the market prices he needs to pay to acquire the necessary resources
are too high, or are expected to rise to a level that would make the project rela-
tively unpro�table, the entrepreneur will abandon it. If present prices are high but
the entrepreneur expects pro�tability nevertheless, he will bid for the resources
he subjectively considers undervalued. He will also steer his bidding toward those
resources that he considers most undervalued, from the point of view of the pro-
The result of this behavior by numerous entrepreneurs is a dynamic pricing
mechanism of the means of production where determined prices in the present
re�ect entrepreneurs’ joint anticipations of the value that each resource can con-
tribute to creating �nal consumption goods. Entrepreneurs thus cooperate to “dis-
What has been discussed in this section is not novel and does not differ from
Mises’s original calculation argument (Mises, [1920] 1935), which relies on
decentralized bidding by future-oriented pro�t-seeking entrepreneurs looking for
the least-cost means that will take them toward their preferred imagined product.
As entrepreneurs outbid each other to secure necessary resources, prices are deter-
mined at levels where those entrepreneurs who pursue the, relatively speaking,
most highly valued production projects, combined with relative certainty about
their value, will be able to outbid those with undertakings perceived as lower-
value or higher-uncertainty. In this sense, the individual entrepreneur’s “
factors is largely determined by forces outside himself and his own sales” (Roth-
bard, [1962] 2004, p.
356). For
our purposes, it is suf�cient to note that the market
prices of the means of production that any one entrepreneur faces in his aims to
produce are ultimately their valuations in projects pursued by
In other words, value precedes and determines cost in the individual production
process and, more generally, the market prices of factors. Time cannot therefore
be excluded from the analysis of production but, as Jeffrey Herbener (2018)
matters for entrepreneurs, and consequently for the entrepreneurially
driven market process, is not therefore cost minimization as it is generally under-
stood, but a leveraging of already-pursued value-generating production projects
to create even greater value for consumers. Factor prices determine which pro-
duction projects are undertaken in the market overall, but are leveraged by entre-
preneurs in those endeavors
– the
cost is chosen after the value of a project has
already been estimated. In fact, it can be argued that entrepreneurs are not even
able to decide how to minimize or “cut” costs at this stage. They can only choose
a cost structure for their production undertaking that, based on their expecta-
tions of future prices and, to the degree they �nd it useful, using information of
present prices in the factor markets, appears as suf�ciently pro�table to make the
whole endeavor worth their while. The entrepreneur’s choice involves appraising
Entrepreneurship versus management
The discussion in the previous section suggests that the role of entrepreneurship
is to bear the uncertainty of imagined production projects and, by acting on them
and thus bidding for resources in factor markets, entrepreneurs individually and
collectively contribute to
– if
– and
who are relatively
As noted by Schumpeter ([1911] 1934), innovative entrepreneurs do not only
produce new goods but also �nd new and better ways to produce goods already
offered in the market. They also imagine and establish new types of organization
that are more effective in production and thereby further contribute to the mar-
ket’s overall value creation. As also noted by Schumpeter ([1911] 1934, p.
these innovation
s are often introduced to the market through forming new �rms.
There is good reason for this, because, as I
elsewhere (Bylund, 2016; see
also Bylund, 2015), the economic function of �rms is to establish more inten-
sively specialized production than is currently supported by the market through
exchange contracts. In other words, the economic function of the �rm is to pro-
vide entrepreneurs with an effective means to implement innovations that cannot
be implemented using market means. The �rm is how entrepreneurs create and
establish a new supply function in the market; it is how entrepreneurs leverage the
This conception of the economic �rm suggests that entrepreneurship, as I
discussed it
here, is more than simple arbitrage (Kirzner, 1973). Arbitrage, after
all, can take place without a �rm, and, indeed, it can be a proper means to mini-
mize costs in an already-established production process. Yet this is not how we
have examined entrepreneurship above, where we instead treated the entrepreneur
as the bearer of uncertainty in creating a new supply function that he imagines is
“superior.” The entrepreneurial function is here one that provides value creation
Seeing entrepreneurship this way reveals that it is the true driving force of the
market (Mises, [1949] 2008) in that it not only pushes the boundaries of the mar-
ket’s present production capabilities, but also changes and re�nes its capital struc-
ture and thus the production apparatus that constitutes the economic “organism.”
The management problem of socialism
It is then indeed the case, as Lachmann ([1956] 1978, p.
13) noted,
that “capital
combinations, and with them the capital structure, will be ever changing, will be
dissolved and re-formed” and that it is “[i]n this activity we �nd the real function
of the entrepreneur.” This “real function” is not purely responsive to exogenous
events and therefore equilibrating within the market’s already existing bounda-
ries, but is imaginative and innovative in �nding out how to leverage resources
already used in valuable ways to produce even
value. Entrepreneurship,
from this perspective, contributes to and improves the market’s overall value crea-
tion, creates economic growth, and, through establishing innovative supply func-
tions in new �rms, increases the extent of the market (Bylund, 2016; cf. Smith,
This creative function is different from the management function that runs the
already-established �rm’s supply function. When a new supply function has been
established and the entrepreneur therefore has implemented his imagined produc-
tion process and con�rmed that it satis�es real consumer wants, the subsequent
phase is not (in fact, it cannot be) intended to create new value but rather to
exploit the value already created and thus “maximize” the effectiveness of the
entrepreneur’s new supply function. This is achieved by minimizing the costs of
production and thereby attempting to maintain the pro�tability of the �rm at an
above-normal level. The objective of management, following the entrepreneur’s
act of creation, is therefore to improve productivity within the supply function
to increase pro�tability and, furthermore, to extend the venture’s pro�table life
cycle. Management is thus tasked by the uncertainty-bearing entrepreneur, who
owns and has established the new supply function, with maximizing the function
by reducing input use and waste to push the cost
per unit
down and to use this
increased effectiveness to exploit the slope of the demand curve.
In other words,
the manager is employed to increase margins and reduce prices, which can attract
customers in larger numbers.
When the management function assumes decision-making power, the value cre-
ated for consumers is neither new nor uncertain, but revealed through actual sales
in the market. The new value was already created as the entrepreneur established
the supply function and bore the related uncertainty. The role of the manager, fol-
lowing and acting on behalf of the entrepreneur, is to maintain this function. As
a consequence, the proper focus of the pro�t-seeking endeavor shifts from new
value creation to cost minimization. The value produced using the existing sup-
ply function can only marginally be increased by tweaking the product to make it
more usable, re�ne its design, and perhaps reposition it in the eyes of consumers,
but the potential for cost-cutting and streamlining the production process offers
great possibilities for increasing pro�tability. Indeed, the creation of a new supply
function is the task of entrepreneurship and not of management. What matters
to us here are not the exact actions taken by management, but the idea that the
management function differs from the entrepreneurial function in a fundamental
way: whereas the entrepreneur bears the uncertainty of an undertaking aiming to
create new future value, illustrated by the creation of a new supply function in
a new �rm, the manager takes this supply function as a starting point and seeks
to maximize the effectiveness by which this value is created. In contrast to the
entrepreneur, management holds value-creation relatively constant, but varies and
This observation is instructive for understanding the Austrian view of the mar-
ket process and therefore the Misesian argument that economic calculation is
impossible under socialism. With a single owner of society’s resources, there can
be no entrepreneurship as discussed above: there is no uncertainty-bearing under
the threat of loss because loss is not personal, and there is also no individual ben-
e�t from pro�ting. In other words, there is no intellectual division of labor within
value creation, which makes calculation impossible, which in turn makes entre-
preneurial value creation impossible. As a result, the extent of a socialist economy
does not
– and
– expand,
as there is no value creation through the
creation of new supply functions in competition with other imagined and already-
existing ventures: there is only management of existing production structures and
the resources allocated between projects already underway.
This conclusion applies not only to the market socialist critique of Mises’s
argument but also to Denis’s (2015) more recent challenge in which several
control replaces private property. Under several control of public property there
will indeed be a factor market determining prices for the means of production as
assigned leaders of society’s production units place bids for resources to use in
their respective processes. These bids are, as Denis points out, real market bids
placed by decision-makers in a decentralized production structure; the managers
of production units do not simply “play market as children play war, railroad, or
school,” as Mises ([1949] 2008, p.
703) put
it, just like managers of corporations
do not play market. Nevertheless, this factor market will be severely crippled
because it remains unaffected by value-creating entrepreneurship: the economy’s
capital structure will not be “ever changing
. dissolved
and re-formed” (Lach-
mann, [1956] 1978, p.
13), but
will instead remain structurally the same. What
changes are made to the capital structure are at best changes in response to chang-
ing consumer preferences and exogenous in�uences in order to adjust the exist-
ing production processes. Managers may even act as arbitrageurs as they shift
their bids to cheaper inputs to be used in their processes, and thus respond to
changing market data. As a result, their combined efforts are reallocated between
production processes (cf. Kirzner, 1973). Yet new supply functions that disrupt
the market and discover previously unknown demands are different: they require
new uncertainty-bearing and are consequently entrepreneurial. Indeed, the ulti-
mate decision to shift a �rm’s capital to creation of a new supply function entails
the withdrawal of capital from its existing use and the subsequent investment in
the new endeavor, which requires ownership (Hülsmann, 1997; Machaj, 2007;
The conclusion is that a common basis for critiques of Mises’s argument against
socialism is the perception of a market economy as shaped primarily through the
function and therefore characterized by cost minimization within
the existing capital structure. In contrast, Mises and the Austrians view the market
as an open-ended process driven by the
function characterized
The management problem of socialism
by new value creation that disrupts the existing structure as well as the extent of
the market. In other words, the solutions offered to the problem Mises raised are
based on different assumptions about the very nature of the economy.
The management function is impotent with respect to reshaping an economy’s
capital structure and production apparatus, because it is limited to adjustments to,
and within, already established supply functions. This in�uences factor prices as
the manager of a venture bids on and shifts between alternate resources, but fac-
tor prices determined only by managers cannot re�ect potential new production,
as this is the function of entrepreneurship, which requires bearing uncertainty
through the ownership of invested property. Managers act within the boundaries
of the supply functions created by entrepreneurs, and they therefore cannot pro-
Importantly, the argument formulated above is applicable to more general prob-
lems than the socialist economy, because the distinction between the entrepre-
neurship function as value-creating and the management function as cost-focused
is universal and should apply to all cases of similar structure: where economic
action relates to the coordination of efforts without open-ended creation and thus
is limited to varying the inputs used in production, the undertaking is limited to
This argument is also stronger, because where value creation is the focus, what
matters to decision-makers and actors is the relative creation of value
– not
cost involved to carry out the action. Indeed, we brie�y noted above that cost
should be of little concern
– and
only of secondary import
– if
the expected out-
come is of greater value. Cost, in fact, is only a means toward producing that
value and is therefore always chosen
as a function of the value potential
project. In contrast, where cost is the primary concern the struggle is to keep exist-
ing value from falling. This is, as I
shall now
illustrate, a battle that cannot be won.
How cost destroys value
Although their functions are distinct, management and entrepreneurship tend
to co-exist in the real market. Private businesses aim both for sustained and
increasing pro�tability through process improvements and cost minimization
(management) and create new value by investing funds in novel supply functions
(entrepreneurship). Whereas the management function follows and replaces the
entrepreneurship function when the supply function passes the market test, man
agers are commonly used at least in part as a means for further entrepreneurship
aiming to improve the venture’s position in the competitive market. Businesses
that survive over longer periods of time generally engage in repeated new product
development and invest large amounts in research and development: they stay
in business by reinventing themselves. The original supply function established
through forming the �rm may therefore come to be replaced, and sometimes
combined with or augmented by, a new supply function of greater anticipated
value as a result of successful entrepreneurship efforts “internally”
– that
within the
�rm’s boundaries (cf. Jensen and Meckling, 1976, pp.
Such legal �rms can increase and extend pro�tability by continually assessing
their situation and entering new markets. They repeatedly consider their options
and thereby establish an internal value tradeoff between possible supply func
tions. Existing functions are rejected (the invested capital withdrawn) when they
no longer contribute suf�ciently relative to the potential value to the �rm’s bot
tom line of new (entrepreneurial) supply functions under consideration (new
investments). Yet whereas shifting from one opportunity to another is a task that
may be delegated to employed managers, it is still the owner of the capital in the
�rm who is the ultimate decision-maker: he is the one to suffer loss from bad
investments and thus bears the uncertainty, and necessarily has the veto right
The case is clearer when considering
�rms, which are de�ned by
and limited to their unique production undertakings (Bylund, 2016). They do not
attempt to recreate production and enter new markets through entrepreneurship,
which would generate a new �rm; this means they rely on the value created in the
original supply function throughout the �rm’s life cycle. By thereby holding the
�rm’s value creation relatively constant,
they are fully dependent on the manage-
ment function of cost minimization. In other words, the possibility of increasing
and sustaining pro�tability is limited to
reducing cost
. Consequently, the pro�t-
ability of the undertaking can increase or be sustained only for as long as the
market situation does not radically change in such a way that it undermines the
organization’s position and thus its ability to satisfy demand in a valuable way.
During a period of relative market stability, achieved pro�tability depends on the
�rm’s ability to extend or entice new demand and thereby expand the possibility
Cost minimization efforts initially allow the �rm to improve on and exploit
the full potential of the entrepreneur’s implemented innovation. This is done by
implementing tweaks for increased effectiveness and reduced resource use in
production, streamlining the production processes, and considering other materi-
als available at lower prices in the market. Technologically speaking, the sup-
ply function may be improved both by increasing the quality of the product and
reducing waste. Economically, production is shifted to utilize relatively lower-
cost resources while the value produced changes minimally as consumers are not
offered a different product but “more of the same”
– yet
perhaps offered at a lower
To illustrate, consider a �rm producing a certain product that has value to con
sumers and is thus demanded in the market. The �rm does not engage in entre
preneurship and therefore neither attempts to replace the product with a new one
or to develop products with which to enter new markets. We thus hold the supply
function constant for this �rm, by which is meant the original function remains
intact but in revised and perhaps improved form, to illustrate the effects of market
action based on management only. This allows us to work through the logic and
determine how generally applicable the thesis drafted above is, without resorting
to the speci�c conditions under which public sector services that are not held to
The management problem of socialism
pro�tability requirements or competitive service offerings in the open market
operate. The manager within this �rm is consequently tasked with increasing
and sustaining the �rm’s pro�tability, but is not allowed to act entrepreneurially
by abandoning the supply function or shifting the �rm’s production efforts into
new markets. The invested capital, in other words, remains in the original supply
As should be evident from the discussion above, holding the supply function
constant does not entail �xed output volumes, which the manager can adjust as
needed, but only a �xed
of output. For instance, a producer of light bulbs
continues to produce light bulbs, but the quantity produced, the materials used
for inputs, and the factors used to aid production are variable. In other words, we
are speaking about running the business, not creating a new one. The example
considered here begins as the newly created supply function (for the new good,
light bulbs) has passed the market test (that is, suf�cient quantities were sold at
cost-covering prices), which means there is suf�cient demand to support the busi-
ness, at least for the moment. This is why the task of management is to increase
and sustain pro�ts, which can be accomplished by cutting costs in production,
tweaking the end product, or varying the selling price to effectuate changes in
Cost-cutting can be accomplished by shifting production to favor inputs that
are offered at relatively cheaper prices in factor markets, the knowledge of which
may be limited by, for example, transaction costs (Coase, 1937).
This amounts
to what Farrell (1957) refers to as price ef�ciency, or of “choosing an optimal
set of inputs” (Farrell, 1957, p.
259), and
requires that the manager is respon-
sive to changes in prices. Within the �rm’s production process, the manager can
improve its technical ef�ciency (Farrell, 1957) or the effectiveness of the already-
established production process by reducing waste and lead times, and consequently
increasing overall resource utilization. Both of these activities suggest reduction in
The product can also be re�ned in its functionality, features, and quality, par-
ticularly as the �rm learns about its customers’ speci�c wants and can therefore
better target those most highly valued. As the product is tweaked to better �t the
real wants of consumers, the selling price may be increased, at least in the short
term. Yet whereas a “better” product is potentially demanded in greater quantities,
More importantly, and applicable in the longer term, the �rm can exploit the
shape of the demand curve to maximize revenues. This can be done by adopt-
ing techniques for implementing different forms of price discrimination that
allow the �rm to capture a greater part of the value produced and thus replace
some consumer surplus with producer surplus. However, the possibility of using
such techniques diminishes with the ef�ciency of the market, leaving the �rm to
lower prices in response to competitive pressures and as demand is satiated. For
instance, a lower price point can increase the �rm’s sales and therefore, assuming
elastic demand, its revenues by increasing the overall quantity demanded and, in
a competitive situation, its market share. Depending on the �rm’s speci�c cost
structure, increased sales can contribute to higher pro�ts or sustain pro�tability
Management needs to consider cost-cutting, product improvement, and lower
prices to maximize performance, and the �rm will consequently experience fall-
ing relative prices for input
output. Whereas lower selling prices increase
volume and thus allow the �rm to exploit the full extent of the former, at some
point the �rm will no longer be able to reduce input prices at the same rate as the
reduction necessary in output prices to increase sales. Indeed, cutting prices as a
seller is easier than reducing prices as a buyer. With the actions of entrepreneurs
in the market where the �rm operates, factor prices will be bid up as more valu-
able production is pursued, which allows them to assume more highly-priced cost
structures, thereby pushing the �rm’s cost of inputs up, while the relative value of
the �rm’s product diminishes.
It follows that without the addition of value-creative entrepreneurship, cost
minimization is an effective strategy only for a limited time period. At some point,
in order to further cut prices with the intent to increase sales, the manager will be
forced to consider a reduction in input or production costs that could also reduce
the market value of the product offered. This decision should be made whenever
the expected effect is a net contribution to the �rm’s bottom line, which suggests
that any given supply function will tend to produce
to consumers over
The benchmark for consumers considering purchasing a product is the opportu-
nity cost of that decision and, consequently, the most highly valued opportunities
that are foregone by making the purchase. If the �rm we are analyzing acts in a
competitive market with new entrants or where other incumbent �rms engage
in entrepreneurship, the opportunity cost will tend to increase, perhaps rapidly.
Indeed, consumers’ relative costs of buying the �rm’s product will increase by the
value increase in alternative opportunities in the market
the value decrease in
the cost-reduced product that the �rm offers. This suggests consumers will tend
to abandon the only-management �rm’s product as they can obtain higher value
If the �rm acts in an economy where entrepreneurship is not common or
expected (or perhaps not even permitted), the period during which the �rm can
sustain pro�tability should be comparatively much longer. The opportunity cost
of consumers will not increase as a result of value creation, which suggests the
value of competing goods (new and old) does not increase (opportunity costs may
not increase), and the relative value of the good in question will therefore not be
undermined. If the absolute value of the product offered by the �rm would remain
the same, which is unlikely for more than a short period of time, its relative value
should increase as the value of other products diminish with cost-cutting efforts.
However, the absolute value of “our” �rm’s product will also diminish as manage-
ment attempts to increase and sustain pro�tability, and thus prolong the life cycle,
by minimizing production costs. The �rm’s pro�tability may then be sustained as
The management problem of socialism
long as the absolute value of its product is higher and the rate of reduction in value
does not exceed that of products offered by other �rms.
Implications for socialist calculation
The above discussions have implications for arguments on both sides of the
socialist calculation debate. I
showed that
the Austrian conception of the market
process is one of consistent (but nevertheless �uctuating) economic progression
by means of entrepreneurial value creation. This invalidates all relevant critiques
to date that assume an economy with unchanging or static boundaries. More
importantly, I
also showed
that the assumption that challengers of Mises’s argu-
ment make about the state of the market must be false: the alternative to a pro-
gressing economy, in which the extent of the market expands through innovative
entrepreneurship (Bylund, 2016), is not a �xed-pie economy but a retrogressive,
This observation is not based on a very strict interpretation of socialism (which
has been a source of criticism of Mises, [1920] 1935), but is an implication of the
economic state of affairs that obtains where management is the predominant force
on resource allocation. Mises suggests a similar argument when stating that “the
crucial and only problem of socialism, is a purely economic problem, and as such
refers merely to means and not to ultimate ends” (Mises, [1949] 2008, p.
This is indeed what management, limited to within-system adjustments, does,
and it does so in stark contrast to entrepreneurship’s value creation. Coase (1988,
8) noted
the relevance of this connection when speculating that the planning of
Soviet Russia was “essentially the same puzzle” as the planning undertaken by
managers within �rms. The major difference, to Coase, is that a market has many
But the greater difference between a market and non-market economy is, as
we established above, entrepreneurship. Within the �rm, the manager is but the
“junior partner” of the entrepreneur (Mises, [1949] 2008, p.
301) whose
task is to
minimize costs and increase pro�tability in the given supply function created by
the entrepreneur. While managers act as bidders in the market for the means of
production, their effect on the market system is solely allocative
productive endeavors, because the market data that are available for economic
decision-making are constantly changed and challenged by entrepreneurial action.
In other words, the issue of economic calculation is not one of how to maximize
resource allocation and usage within a �xed system, but one of calculation within
boundaries where the scope and extent of economic action are either
expanding (market system) or contracting (non-market system). In an expand-
ing market economy the anticipated value created by entrepreneurial undertak-
ing takes precedence over the cost necessary to realize it. Cost is ultimately an
implication of the true choice variable: value creation. Cost is thus a lower bound
for value creation efforts as the prices determined in factor markets approximate
social opportunity costs. This lower bound is consistently pushed upward as fac-
tor market prices re�ect anticipated value creation, thereby pressuring entrepre-
neurs to create even more value. But there is no upper bound for value creation
other than that imposed by the imagination of entrepreneurs and consumers, and
consequently, entrepreneurs try to �nd the best way to leverage cost.
In a contracting economy, in contrast, the aim of decision-makers is manage-
ment: to minimize cost in order to sustain pro�tability for an already established
supply function that generates a product of �xed or falling market value. The
question in this scenario is whether decision-makers can be guided by prices,
and whether their combined bids for resources in the factor markets can deter-
mine prices that approximate their social opportunity costs. In an outright socialist
economy, as Mises showed, there are no proper prices and producers are therefore
blind as to where they can contribute to consumer welfare. In our model of the
management economy, which is not a full-scale socialist economy but simply an
economy lacking entrepreneurship (that is, with no new entrepreneurial entry),
factor prices are relevant for decision-making in everyday tasks but unimportant
for long-term welfare. Bids for resources are not intended to create new value,
which requires new supply functions, but to extend the life cycles of already-
established supply functions. Thus, factor prices may here approximate the social
opportunity cost of resources in existing production,
but this is of little use to
decision-makers in �rms as the economy is neither progressing nor stable but
in fact retrogressing and contracting. Without entrepreneurship and the crea-
tion of new supply functions, the market will be unable to respond to new and
previously-unaddressed demand
of any sort
– it
can only address changes to the
extent they relate to existing types of production. The economy should therefore
at any point in time be at its maximum in terms of value creation; the only possi-
ble development is a reduction in total value produced, either through managerial
Both entrepreneurial and managerial economies are characterized by change
and thus uncertainty, which means the solutions offered to the calculation prob-
lem cannot be limited to a �xed-boundary system. The question that Mises posed
is whether calculation is possible in the non-progressing economy, which we
now understand is necessarily a
economy. In a socialist contracting
economy, Mises found economic calculation to be impossible; our analysis above
The management problem of socialism
suggests that in a non-socialist economy that lacks entrepreneurship economic
calculation may be possible, but that it is meaningless because it applies purely
to the management of existing types of production and is not directed toward the
Concluding remarks
This chapter
to reformulate and extend Mises’s original argument that
calculation is impossible in a socialist economy. We found that without entre-
preneurship, even in a setting where economic action is taken to earn pro�ts
is decentralized, the economy tends to contract rather than
expand. In an economy where the “driving force” is economic management rather
than entrepreneurship, value will not only not be created but will be actively
destroyed. In contrast, a market with entrepreneurship as its driving force is dis-
tinguished by its progression in terms of new value creation. There appears to be
This conclusion follows from the distinction between two economic functions,
both of which may be necessary for a properly functioning market: management
and entrepreneurship. Socialist economies are characterized by management but no
entrepreneurship, hierarchically structured production, and a central plan; market
economies rely on both innovative entrepreneurship to increase the extent of the
market and management to improve already-established production processes coop
erating in a process of creative destruction. Both systems grapple with uncertainty,
but there is an important difference between them: uncertainty-bearing in a progress
ing system of entrepreneurship is offensive and focused on value creation, whereas
uncertainty-bearing in a contracting economic order is limited to management of
For a market order under entrepreneurship, the role of cost is to set a lower bound
to value creating endeavors and thus force some (the least successful) entrepreneurs
to exit. This lower bound is pushed ever higher as entrepreneurs challenge the sta
tus quo by attempting to create new and even greater value, and thus undermine the
rationale for existing production. Entrepreneurs engender the market economy’s
process of “creative destruction,” as Schumpeter ([1942] 1950, pp.
81–86) noted.
In a pure management system, in contrast, the value created through existing pro
duction processes constitutes an
bound, and the chief struggle is therefore
to maintain this level of satisfaction if at all possible. In short, economic theorists
siding with socialist ideals tend to assume socialism is not only possible, but that it
Knight’s concept of entrepreneurial judgment, which is different from Kirzner’s entre-
alertness (High, 1982), is further analyzed in Foss and Klein (2012) and
Value creation is limited to the already-established supply function and variations
which exclude creation of new value. Firms can improve their production pro-
cess as well as their product offerings, but cannot innovate new products or shift produc-
For an
elaboration of the relevance of Coase’s theory of transaction costs and manage-
This point
is made even clearer by Salerno, who summarizes Mises’s view of the prob-
lem of socialism as “the problem purely of Robbinsian maximizing, of deciding how
means are to be allocated in light of a
structure of ends” (Salerno, 1990a,
46; emphasis
in original). In other words, it is a problem of economic actions taken
during the maintenance phase of an existing production structure (or, within the �rm, an
Factor prices
may re�ect future prices as well, but as the expectations of managers relate
only to “business as usual,” or to failure, the only options without new entrepreneurship,
Berle, A. A., and G. C. Means. 1932.
The Modern Corporation and Private Property
. New
Brunswick, NJ: Transaction Publishers.
Blaug, M. 1997.
Economic Theory in Retrospect
. Cambridge: Cambridge University Press.
Bylund, P. L. 2014. “Ronald Coase’s ‘Nature of the Firm’ and the Argument for Economic
“Explaining Firm Emergence: Specialization, Transaction Costs, and the
The Problem of Production: A
New Theory
of the Firm. Abingdon:
Bylund, P. L., and G. P. Manish (forthcoming). “Private Property and Economic Calcula-
tion: A
Cantillon, R. 2010 [1755].
An Essay on Economic Theory
. Trans. C. Saucier. Auburn, AL:
“The Nature of the Firm: Origin.”
Journal of Law, Economics
& Or
Denis, A. 2015. “Economic Calculation: Private Property or Several Control.”
Review of
Farrell, M. J. 1957. “The Measurement of Productive Ef�ciency.”
Journal of the Royal
New Appr
Hayek, F. A. 1945. “The Use of Knowledge in Society.”
Hayek, F. A. V. 1978. “Competition as a Discovery Process.” In
New Studies in Philosophy,
Politics, Economics, and the History of Ideas
. Chicago: University of Chicago Press,
, J. M. 2018. “Time and the Theory of Cost.” In Matthew McCaffrey, ed.,
Economic Theory of Costs: Foundations and New Directions
. Abingdon, UK: Rout-
High, J. 1982. “Alertness and Judgment: Comment on Kirzner.” In I. M. Kirzner, ed.,
Method, Process, and Austrian Economics: Essays in Honor of Ludwig Von Mises
. Lex-
The management problem of socialism
Hülsmann, J. G. 1997. “Knowledge, Judgment, and the Use of Property.”
Review of Aus
Jensen, M. C., and W. H. Meckling. 1976. “Theory of the Firm: Managerial Behav-
ior, Agency Costs and Ownership Structure.”
Journal of Financial Economics
3 (4):
Kirzner, I. M. 1973.
Competition and Entrepreneurship
. Chicago: University of Chicago
Knight, F. H. 1985 [1921].
Risk, Uncertainty and Pro�t
. Chicago: University of Chicago
Lachmann, L. M. 1978 [1956].
Capital and Its Structure
. Kansas City, MO: Sheed Andrews
Lange, O. 1936. “On the Economic Theory of Socialism: Part One.”
“On the Economic Theory of Socialism: Part Two.”
Machaj, M. 2007. “Market Socialism and the Property Problem: Different Perspective
of the Socialist Calculation Debate.”
Quarterly Journal of Austrian Economics
10 (4):
McMullen, J. S. 2015. “Entrepreneurial Judgment as Empathic Accuracy: A
Decision-Making Approach
to Entrepreneurial Action.”
11 (3): 651–681.
Menger, C. 2007 [1871].
Principles of Economics
. Trans. J. Dingwall and B. F. Hoselitz.
Auburn, AL: Ludwig von Mises Institute.
Mises, L. V. 1935 [1920]. “Economic Calculation in the Socialist Commonwealth.” In
Friedrich August Hayek, ed.,
Collectivist Economic Planning
. London: George Rout-
Human Action: A
eatise on Economics
. Scholar’s ed. Auburn, AL:
Rothbard, M. N. 2004 [1962].
“Postscript: Why a Socialism Economy Is ‘Impossible’.” In L. V. Mises,
Economic Calculation in the Socialist Commonwealth
. Auburn, AL: Ludwig von
, J. A. 1934 [1911]. The Theory of Economic Development: An Inquiry Into
Pro�ts, Capital, Credit, Interest, and the Business Cycle. Cambridge, MA: Harvard Uni-
Capitalism, Socialism, and Democracy
. 3rd ed. New York: Harper
Smith, A.
1976 [1776].
An Inquiry Into the Nature and Causes of the Wealth of Nations
Taylor, F. M. 1929. “The Guidance of Production in a Socialist State.”
Ludwig von Mises’s theory of economic calculation is a strong contender for
the title of greatest contribution to twentieth-century economics. Mises’s writings
revolutionized economic thinking in at least two ways: �rst, they explained the
indispensable role calculation plays in the market process, and second, they dem-
onstrated the impossibility of calculation under a system of socialist central plan-
ning. Socialist organization inevitably fails because it lacks the necessary means
to appraise the costs and bene�ts of any planned allocation of resources. Without
this basis for decision making, socialism descends into “planned chaos,” leading
Mises’s contributions to the economic analysis of socialism are widely
acknowledged. However, the fact is sometimes overlooked that his work on cal-
culation offers much more than a critique: it is also a positive theory of the market
economy. In other words, it is a comprehensive account of the process by which
individual exchanges produce a vast and complex system of market prices that are
continuously responsible for the rational allocation of scarce resources in society.
Furthermore, his theory “provides the rationale for the price system, purely free
markets, the security of private property against all encroachments, and sound
money” (Salerno, 1993, p.
69). Economic
calculation is thus inextricably linked
with the most essential social institutions. More speci�cally, it is an indispensable
tool for entrepreneurial decision making, which Mises called the “driving force”
of the market and of economic progress. The price system allows pro�t-seeking
entrepreneurs to assess the costs and bene�ts of their production decisions, and
thus to make good use of society’s scarce resources. Moreover, calculation estab-
lishes the boundaries of pro�t-seeking �rms and places limits on their horizontal
However, economic calculation is also a useful and necessary framework for
analyzing a wide range of organizations that fall outside strictly for-pro�t enter-
prise or socialist central planning. These include organizations in the voluntary
market economy that are not founded with a primary goal of earning monetary
pro�ts, such as charities, cooperatives, trusts, mutual societies, social enterprises,
and other forms of economic activity that might be considered “not-for-pro�t.”
limits of social entrepr
Matthew McCaffrey
Matthew McCaffrey
Understanding calculation
– or
the lack thereof
– within
these organizations is
vital for evaluating their true costs and bene�ts, the role they play in the economy,
and their ultimate impact on society. Mises’s theory thus holds enormous potential
for economists hoping to expand their analysis to non-traditional areas of inquiry.
In fact, some scholars argue that “Mises’s concept of economic calculation
belongs at
the forefront of
. research
into the nature and design of organiza-
focuses speci�cally
on the role of economic calculation in “social
entrepreneurship.” Social enterprises are business organizations that are not
motivated by the desire to generate monetary pro�ts for traditional shareholders.
Instead, the pro�ts of social enterprise are used to solve “social” problems, often
by addressing the same kinds of needs as charitable organizations. Social enter-
prises are special, however, because they support their missions through success-
ful commercial ventures rather than through donations. This type of business is
increasingly prevalent around the world, and in its many variations is the subject
of much entrepreneurship and management research. However, little work has
been done to integrate the social entrepreneurship literature with the economic
approach to cost and bene�t, i.e. the theory of economic calculation. Yet this is a
vital task for economics, which already contributes in�uential ideas to entrepre-
neurship research. Applying the theory of calculation outside traditional market
entrepreneurship is a natural extension of previous work, and is even crucial if
emerging research �elds in entrepreneurship are to be grounded in sound eco-
The following chapter
takes a
�rst step in this direction by integrating the theory
of economic calculation with some current research on social entrepreneurship.
�rst explain
the meaning of social action in economics and in entrepreneur-
ship studies, and use the contrast between disciplines to highlight the distinctive
domain of social entrepreneurship. I
then discuss
the pro�t motive in traditional
and in social enterprises, and clarify its in�uence on entrepreneurial decision
making. This leads �nally to a discussion of calculation within social enterprises.
explore two
different models of social entrepreneurship, and explain the role of
calculation in each. I
that the price system plays an indispensable role in
placing boundaries on the size and scope of social enterprises. I
conclude with
The economic foundations of social entrepreneurship
Mises demonstrated that the de�ning characteristic of the market process is entre-
preneurial action using economic calculation in terms of money prices (Mises,
1990a, 1998). Given the overarching importance of his theory, it is no surprise
that calculation is relevant for research in social entrepreneurship as well as for
its more traditional pro�t-focused counterpart. However, although calculation
is indispensable to a functioning economy, it does not play the same role in all
organizations, and there are many institutions in which the extent of calculation
is ambiguous or unde�ned. This is especially true of
social enterprises
, which
The limits of social entrepreneurship
employ a number of non-traditional business methods and objectives that are dif-
�cult to analyze using conventional economic tools. De�ning the “social” aspects
of social enterprise
– and
explaining how social entrepreneurship is distinct from
“mundane” market entrepreneurship
– is
particularly challenging (Zahra et
2009). The
purpose of the present section, then, is to explain the economic basis
of social action by showing that pro-social behavior is at the root of both mundane
As the name implies, the signi�cance of social entrepreneurship lies in its
“social” meaning. Understanding how social enterprises work therefore requires
explaining their uniquely social facets. However, explaining these implications
and what they mean in practice for the economic structure of social enterprises
is dif�cult. Social entrepreneurship is variously and sometimes vaguely de�ned,
and its use in social science can be inconsistent with its legal meaning or with the
ways it is understood by many real-world social entrepreneurs (Agafonow, 2015).
The �rst step toward addressing this problem is to unpack the economic assump-
tions of social enterprise. Typically, the key economic difference between mun-
dane entrepreneurship and social entrepreneurship is said to lie in their respective
ends: in general, social enterprises are de�ned as “mission-driven,” while ordi-
nary enterprises are pro�t-driven (Dees, 2001; Abu-Saifan, 2012). However, this
distinction is inaccurate for two reasons, both of which make it dif�cult to draw a
simple line between the two types of entrepreneurship: �rst, pro�t-seeking entre-
preneurship is always to some extent social, and second, social entrepreneurship
is always to some extent pro�t-seeking (Austin, Stevenson, and Wei-Skillern,
2006). The following two subsections explore each of these claims in turn and lay
the foundation for a discussion of economic calculation in social entrepreneur-
ship. I
explain that
social entrepreneurship is a subset of a larger social process
that occurs through the division of labor. Furthermore, the pursuit of monetary
Human beings act socially when they cooperate and collaborate. In fact, mutual
understanding between individuals is the basis of economic and social order
(Mises, 1998, pp.
143–145). Social
bonds emerge when individuals begin to
understand the inherent advantages of concerted action over isolated or violent
actions, especially the idea that violence precludes or destroys social bonds.
Speci�cally, society is founded on the joint recognition by individuals of a vital
economic truth: concerted action is more productive than isolated action (Mises,
1998, pp.
157–160). Cooperation
is an acknowledgement of the universal bene�ts
, which Mises describes as “the ultimate social phenom-
enon” (Mises, 1998, p.
157). In
his view, action is “social” to the extent it fosters
One vital implication of this approach is that the entire market economy is a
profoundly social process of mutual want satisfaction. It is comprised of many
individual exchanges that each contribute to the formation of the prevailing
Matthew McCaffrey
structure of money prices. Economic affairs are organized by the price system,
which re�ects from moment to moment the value of goods and services in soci-
ety (Salerno, 1990a; Salerno, 2018). Pro�t-seeking entrepreneurs use prices to
estimate the future value of the factors of production and to compare the costs
and bene�ts of different available combinations of these factors (Salerno, 1990b).
Based on their appraisals, they make judgments about how best to allocate scarce
resources among innumerable competing uses (Foss and Klein, 2012, 2015). Good
judgments yield pro�ts and increased market shares; poor judgments yield losses
and, eventually, bankruptcy. Competition thus functions as a selection mechanism
that eliminates any entrepreneurs who fail to satisfy consumers, or who do so at
This entrepreneurial process of appraising the future value of factors of produc-
tion using money prices is the essence of the market economy. In the absence of
this social appraisement process
– for
example, in a socialist economy in which
there are no factor markets and hence no factor prices
– entrepreneurs
are left
without the ability to compare the costs and bene�ts of alternative production
plans. The result is the systematic waste and destruction of resources by central
planning committees and the collapse of the economy’s capital structure. Ironi-
cally, socialism is in this sense inherently
-social in that it eliminates economic
calculation and ultimately undermines the division of labor and human coopera-
tion. The inevitable outcome of socialist central planning is social disintegration.
Conversely, the entrepreneurial process underlying the market economy is
thoroughly social in that it embodies the cooperation of many individuals work-
ing for mutual bene�t within the division of labor. This notion of social action is
more inclusive than those typically used in entrepreneurship studies; most impor-
tantly, it includes a range of pro�t-seeking actions and business organizations that
are not typically described as social and that do not possess speci�c social goals.
However, focusing on collaboration and specialization reveals that even mundane
businesses motivated mainly by the desire to generate pro�t for shareholders are
deeply social. It is therefore inaccurate to contrast social with non-social enter-
prises as such; as far as the market is concerned, “there is no such thing as ‘non-
social’ entrepreneurship” (Seelos and Mair, 2005, p.
question for economists, then, is not
entrepreneurship is social, but
it is (Tan, Williams, and Tan, 2005). Pro�t-driven enterprise is only one of many
different kinds of cooperative action that are compatible with the division of
labor. Social entrepreneurship is another. Social enterprises are typically de�ned
as business organizations that do not narrowly pursue monetary pro�t and returns
for shareholders, but instead aim to provide solutions to “social” problems (Hard-
ing, 2004; Martin and Osberg, 2007). Social problems are in turn de�ned as basic
human needs that are unmet by prevailing market and political institutions (Seelos
and Mair, 2005). In practice, social entrepreneurs use business ventures to directly
and indirectly transform disadvantaged or marginalized communities, especially
those affected by low levels of education and high levels of poverty, unemploy-
ment, homelessness, or even injustice (Alvord, Brown, and Letts, 2004; Martin
The limits of social entrepreneurship
and Osberg, 2007; Abu-Saifan, 2012). The positive outcomes created by social
organizations are referred to as “social value” (Seelos and Mair, 2005; Haugh,
2006; Peredo and McLean, 2006; Dacin, Dacin, and Matear, 2010). Social value
can emerge in the ways the products or services of the venture function;
the ways they are distributed and delivered; the advocacy approach of the
venture; the ways the venture accumulates and deploys �nancial, human,
and other resources; or the ways the venture’s networks or partnerships are
key element in this approach is that social enterprises are fundamentally
business organizations. That is, they aspire to be independent of charitable and
governmental support (Dees, 1998). Ideally, they earn revenue through the sale
of goods and services just as ordinary businesses do, rather than relying on dona-
tions or public funds (Emerson and Twersky, 1996). Like traditional entrepre-
neurs, social entrepreneurs use judgment to allocate scarce resources in the face
of uncertainty (Mort, Weerawardena, and Carnegie, 2003; Peredo and McLean,
Focusing on the business at the heart of social enterprise avoids confusion
about the scope of social ventures by concentrating on a speci�c kind of economic
organization. This approach is not always standard practice in the literature,
however. For example, early empirical work on social enterprises de�ned them
essentially as innovative organizations outside the for-pro�t and governmental
sectors (Leadbeater, 1997). As a result, previous research includes a wide range of
organizations under the label of “social entrepreneurship,” including institutions
that span or blur the lines between business, charity, and government (Leadbeater,
1997; Dees, 2001; Thompson, 2002). These hybrid organizations and cross-sector
collaborations are worth studying, but they are not the main subject of this chap
ter, which focuses on social enterprise as it has come to be understood in more
recent research, as a distinct form of business venture. Of course, real-world social
enterprises can seek out non-market sources of revenue or make strategic alliances
with governmental organizations or NGOs. Yet when social entrepreneurs choose
to rely on donations or public funds rather than earning revenue through trade,
they change the economic structure of their ventures, thereby creating organi
zations that require different kinds of analysis. Importantly though, the theory
The pro�t motive is a key assumption in contemporary economic theory. It is
especially important for the theory of the �rm, which usually takes the pro�t-
maximizing business as a starting point. It is also frequently invoked in de�ni-
tions of market and social entrepreneurship. Yet like many concepts in economics
Matthew McCaffrey
that are taken for granted, the idea of the pro�t motive deserves closer attention.
The reason is simple: contrary to its name, “the” pro�t motive does not refer to a
static, homogeneous incentive that confronts all entrepreneurs in the same way.
Instead, it re�ects a dynamic, diverse range of human purposes that may or may
not exist for particular entrepreneurs. Given that social entrepreneurship is often
thought to avoid traditional pro�t-seeking, it is important to ask whether pro�t
motives actually play a necessary role in all business organizations. In this sub-
section, I
explain how
a realistic, inclusive theory of pro�t motives reveals that
pro�t-seeking in a non-trivial sense does in�uence the decision making process
All action
– de�ned
as purposeful behavior
– aims
at the achievement of an
end. Ends are valuable because of their perceived ability to provide bene�ts to
actors. The bene�ts of action are weighed against the costs, i.e. opportunity costs,
and individuals choose the alternative that yields the greatest perceived net bene�t
(Newman, 2018). However, bene�ts and costs are personal and subjective. They
can only be understood as a kind of feeling or psychological state that individuals
experience. When this feeling is positive enough to inspire action
– when
e�ts are greater than costs
– an
actor receives “psychic income” (Fetter, 1915,
26–29). This
surplus of psychic bene�t over psychic cost is known as “psychic
pro�t” (Mises, 1949, p.
287; Rothbard,
2009, pp.
71–72). The
pursuit of psychic
pro�t is a universal feature of human action: every action aims at bringing about
conditions that are more desirable from the point of view of the actor. Psychic
pro�t is present even when individuals act charitably out of loving care for others
(McCaffrey, 2015b).
The idea of psychic pro�t carries important implications for social ventures.
If all action involves the pursuit of psychic pro�t, then it is inaccurate to classify
certain kinds of action as for-pro�t and others as not-for-pro�t (Mair and Marti,
2006). Likewise, all organizations embody the actions and values of the people
who direct them, and in this sense there are no such things as completely not-
for-pro�t organizations. Even the directors of charities attempt to use them to
improve conditions from their own points of view. What is needed, then, is a more
speci�c way to distinguish between different pro�t motives. Pro�t in the univer-
sal, psychic sense is not the only kind of pro�t that entrepreneurs pursue, and it is
Economics and entrepreneurship research do not usually refer to psychic pro�t.
Instead, they focus on the pursuit of
monetary pro�t
by entrepreneurs, which is a
major criterion used to distinguish the domain of social entrepreneurship. Unlike
psychic pro�t, individuals only pursue monetary pro�t under certain conditions.
This point has led to some confusion in economics and management research.
Speci�cally, in order to explain entrepreneurial motivation economists have some-
times appealed to the concept of an “economic man” interested only in the self-
ish maximization of monetary income (Mises, 1998, p.
241, 2003,
Rothbard, 2009,
217n15). Unfortunately
, the use of such unrealistic assump-
tions has led entrepreneurship researchers to believe that economics is simply a
The limits of social entrepreneurship
method for studying sel�sh or materialistic behavior (Kirzner, 2011). In this view,
economics has nothing to offer research into “non-economic” or “social” motiva-
However, the concept of economic man (and related models) is unrealistic and
provides a misleading picture of the true scope of economic analysis. The subjec-
tive theory of value that underlies modern economics reveals that human ends are
more complex and varied than textbook analysis implies. In fact, the subjective
theory of value accounts for all possible human motivations. It therefore under-
mines the claim that economics can only focus on monetary pro�t or other narrow
goals. As a result, “[Economics] must not restrict its investigations to those modes
of action which in mundane speech are called ‘economic’ actions, but must deal
also with actions which are in a loose manner of speech called ‘uneconomic’
course, it could be argued that the prospect of money income has historically
been an important motivation for entrepreneurs. This is likely true, yet “While
the pro�t motive [i.e. income-earning] might be ‘a central engine’ of entrepre-
neurship, it does not preclude other motivations” (Mair and Marti, 2006, p.
There is
nothing to imply that the pursuit of greater personal income is necessar-
ily related to entrepreneurs’ decision making process. This point is made in both
theoretical and empirical literature. For example, in his much-cited work on “cre-
ative destruction,” Joseph Schumpeter argued that entrepreneurs are driven more
by the desire to create than by narrow considerations of income (Schumpeter,
1942). And, more recently, a growing body of research has argued that real-world
entrepreneurs are motivated by a variety of psychological and economic factors,
which include but are not limited to the desire for income and wealth (Shane,
Yet, although it is not universal, monetary pro�t is relevant in many real-world
situations. Money’s role as a general medium of exchange gives it a special signif-
icance for individuals who participate in the division of labor. In a monetary econ-
omy, money makes up one side of every exchange, and must be used whenever
goods and services are purchased. Indirect exchange allows consumers to pur-
chase consumer goods that would not otherwise be available. Therefore, in order
to increase their welfare through consumption, individuals must also increase their
money incomes. It can even be argued that “in the money economy, other things
being equal, [all actors] will attempt to attain the highest possible money income”
(Rothbard, 2009, p.
213). This
rule holds true for both individuals and �rms under
these conditions. More importantly, it sometimes holds true even for individuals
who wish to spend their incomes on charitable or “social” causes. For example,
giving alms and buying food for the homeless each require income to donate
or to spend on consumer goods (actions that yield psychic pro�t to the giver).
Anyone who wishes to give more, thereby creating social value, must acquire
more income. Other things equal, each individual and organization in a monetary
economy is bound to pursue money income in order to obtain psychic pro�t. This
applies to pro-social organizations just as it does to mundane businesses.
Matthew McCaffrey
The pro�t motive is a constraint on entrepreneurs’
rather than their
. As Mises argues,
the entrepreneur is actuated by the pro�t motive. This enjoins upon him the
urge to prefer the most economical solution, i.e. that solution which avoids
employing factors of production whose employment would impair the satis-
price system serves as a guide for entrepreneurs navigating the uncertainty of
the marketplace. Yet no matter how entrepreneurs ultimately decide to spend their
pro�ts, it is clear that they will attempt to earn as much pro�t as possible, because
in a monetary economy larger pro�ts enable greater spending on consumer goods,
Yet there are important exceptions to the above rule, as we shall see. We have
assumed thus far that “other psychic ends are neutral,” and that individuals are
primarily able to increase their welfare by purchasing consumer goods (Rothbard,
2009, p.
441). These
assumptions provide the basis for the initial discussion of
economic calculation in the following section, in which social entrepreneurs are
said to maximize monetary pro�ts. However, they are relaxed in a later discussion
of social entrepreneurship in cases where other psychic ends come into play that
Economic calculation and social entrepreneurship
We are now ready to examine the economic organization of social enterprises
and determine the role economic calculation plays within them. Social enterprises
face essentially the same key decisions as mundane businesses: “just what activi-
ties ought to be undertaken? What investments should be made? Which product
lines expanded and which ones contracted?” (Klein, 1996, p.
12). In
other words,
what goods and services should social enterprises produce, and how should they
produce them? These fundamental questions are answered through the process
of economic calculation. And although some researchers do not realize it, many
studies of social entrepreneurship are also searching for these answers, and for
something akin to calculation. For example, Thompson (2002, p.
427) observes
“Strong leadership and good management of socially entrepreneurial initia-
tives is important. There is always an opportunity cost for the resources being
utilised. Achievement below that which could be achieved is a lost opportunity, an
unmet need.” That is, social entrepreneurs need to know the value of the alterna-
Which “social needs” should have priority? Without an overarching objec-
tive, it is impossible to decide whether using resources to help the homeless
in Paris creates as much social value as feeding hungry children in Kabul.
The limits of social entrepreneurship
Table 11.1

Types of entrepreneurship by income and or
economic calculation
Income r
explains both the comparative costs and the boundaries of social
entrepreneurship. Explaining just how it works requires discussing one more
aspect of the pro�t motive: the question of entrepreneurs’ income and spending.
In common speech, the term “pro�t motive” refers to two separate questions:
�rst, do entrepreneurs use monetary pro�ts as the test of their success or failure?
Second, how do entrepreneurs spend any pro�ts they have earned? The answers to
these questions de�ne the major kinds of organizations that entrepreneurs create:
mundane, social, political, and charitable. The different possible combinations of
income/spending and organizational method are listed in Table
We have already seen that psychic pro�t is a universal feature of action, but
that money pro�t only motivates individuals under certain conditions. There are,
of course, organizations that do not use the pro�t-and-loss test as a guide for
their activities. Charities are one example. They refuse to compete commercially
because they view pro�t-seeking as inappropriate or immoral. This refusal pre-
cludes the guidance of the price system and economic calculation, which do not
place the same limits on charities as they do on businesses. Unlike businesses,
charities are not responsible to shareholders expecting to receive a return on their
investments. Instead, they spend their income according to bureaucratic rules laid
down by their directors (Mises, 1944, pp.
47–48; 1998,
307, 721–722).
donors to the organization are the true consumers of the charity’s services (Roth-
bard, 2011, pp.
453–454). A
second example
is government or political entrepre-
neurship, which is also bureaucratic and lacks access to economic calculation
(Mises, 1944; McCaffrey and Salerno, 2011). Government organizations are dif-
ferent from charities, however, in that they are not directed toward genuinely
social goals, but rather the goals of political decision makers (Salerno, 1993;
McCaffrey, 2011).
Matthew McCaffrey
Returning to the �rst question posed above: in general, the pursuit of mone
tary pro�ts goes hand-in-hand with economic calculation and the pro�t-and-loss
test. This is obvious in mundane entrepreneurship, but is less clear for social
entrepreneurship. I
below that social enterprises, as business organiza
tions, cannot ignore the price system when making their production decisions.
This leads to the second question posed above: the problem of how entrepre
neurs use their pro�ts. All forms of enterprise must maintain their capital, and
all businesses have the option of investing pro�ts back into the organization
in order to increase production (Mises, 1990b, p.
269). In
this sense there is
no difference between mundane and social enterprise. However, conventional
entrepreneurs make an additional choice between reinvesting pro�ts and either
retaining them (in an individual-owned �rm) or paying them out to shareholders
as dividends (in a joint-stock company). These options do not exist for social
enterprises because their economic and sometimes legal organization prohib
its retaining pro�t as income, requiring it instead to be spent creating social
value. Social enterprises thus provide a combination of method and spending
that results in a form of organization different from mundane entrepreneurship,
charity, and government.
We now arrive at a key point in the argument: the question of whether social
entrepreneurs engage in economic calculation. Based on the above discussions,
that they do. Given that social enterprises are business organizations that
use their own scarce resources to serve consumers in the marketplace, and given
that they earn pro�ts or losses according to their ability to do so, it follows natu-
rally that in order to survive they can and must engage in economic calculation.
The special missions that de�ne social enterprises do not negate their ability to
engage in economic calculation, because calculation does not require that entre-
preneurs use pro�ts to increase their personal income or wealth. What matters is
that entrepreneurs subject their �rms to the discipline of the market when mak-
ing their production decisions. Any organization that hopes to succeed based on
successful production and exchange must submit to the price system. Calculation
then reveals whether a �rm’s social mission is compatible with a rational alloca-
Entrepreneurs are not necessarily motivated by the sel�sh pursuit of pro�t.
Nevertheless, it is important that their decisions, whatever they might be, are
not arti�cially restricted by the institutional environment. Calculation is an
institutional process that rests on the security of private property and the divi
sion of labor. Among other things, this implies that strong property rights and
freedom from expropriation are necessary for entrepreneurs to calculate effec
tively (Salerno, 1993; Machaj, 2007; McCaffrey, 2015a). Yet these institutional
requirements do not imply that entrepreneurs are only concerned with increas
ing their personal incomes, or that they cannot be concerned with solving social
problems. Rather, the issue is that entrepreneurs must be able to use pro�ts
in ways that allow them to reach the highest points on their individual value
scales. Ultimately, entrepreneurs’ speci�c motivations are less important than
The limits of social entrepreneurship
Economic calculation and the limits of the social
A previous section argued that pro-social goals are compatible with economic
calculation. Yet their relationship consists of more than simple compatibility:
pursuing such goals actually
social entrepreneurs to focus on earning
pro�ts and avoiding losses. Nevertheless, the question remains how economic
calculation speci�cally affects the structure of social enterprises. The present sec-
tion shows that for social ventures to compete they must use the price system to
appraise the costs and bene�ts of their production decisions. Furthermore, eco-
nomic calculation through the price system acts as a limit on the size and scope of
So far this chapter
has described
social enterprise in general terms. I
now provide
more detailed analysis by examining two different types of social organizations.
The �rst and simpler case involves
complementary social enterprises
. These are
social ventures in which business activities are completely separate from the
enterprise’s social mission (Fowler, 2000). The most common examples are social
businesses that donate the entirety of their pro�ts to causes such as charitable
organizations. These charities may or may not also be owned and operated by
the social enterprise: what matters for this discussion is that the social cause does
not directly in�uence the decisions of the business. Instead, it receives operating
income to the extent that the business is successful at earning pro�ts. In this case,
For an enterprise to remain pro�table, entrepreneurs must accurately price
their inputs and correctly estimate the value of their outputs. Complementary
social enterprises approach this problem in much the same way as conventional
businesses: even though social enterprises are organized around distinct goals,
consumers still determine their viability. In other words, if social entrepreneurs
are subject to the pro�t and loss test in competitive markets, they must allow
consumer satisfaction to guide their decisions. Funding a social cause requires
resources, and social entrepreneurs will therefore attempt to maximize monetary
pro�ts. This is a necessary implication of the market process, and does not under-
To buy in the cheapest market and to sell in the dearest market is, other things
being equal, not conduct which would presuppose any special assumptions
concerning the actor’s motives and morality. It is merely the necessary off-
entrepreneurs strive to accurately appraise the future prices of the factors
of production and to purchase them, in the present, at prices that are lower than
Matthew McCaffrey
the selling prices of the goods they will eventually produce. Competition between
social enterprises, and between social enterprises and ordinary businesses, will
still be strong.
Their success depends ultimately on consumers, who assign value
to goods based on the psychic bene�ts they expect to receive from them. In social
enterprises, the value of consumer goods has two general sources: the bene�t
these goods provide through use (direct consumption) and the bene�t provided
by the knowledge that buying helps to advance a valuable social mission (indirect
consumption). These bene�ts are not mutually exclusive, and often exist side by
side. For example, consider a social enterprise restaurant that sells sandwiches
and donates the pro�ts to the homeless. Consumers might value the sandwiches
for their particular characteristics (e.g. their taste) or for their social signi�cance
(e.g. the feeling of contributing to a worthy cause), or for some combination of
the two. These different sources of value determine how consumers de�ne the
good being sold, i.e. the ends it serves. They therefore also in�uence the pricing
of social goods and explain how consumers ultimately determine the prices of the
In order to determine whether production is pro�table, entrepreneurs use the
price system to make judgments about future consumer values. Yet production
takes time, and entrepreneurs can err in their appraisals. As a result, there is
always an element of uncertainty in the pricing process (Herbener, 2018). In the
case of social enterprise, there is a special uncertainty involving the two sources
of value of social goods. Entrepreneurs must anticipate whether the value attached
to direct and indirect consumption will be suf�cient to justify charging market
prices for their outputs. If not, the value of the productive factors in the enterprise
Charities and competitive social enterprises are obliged to buy or rent factors at
prevailing market prices (Rothbard, 2011, pp.
454–455). If
entrepreneurs cannot
afford to pay these prices because consumers do not suf�ciently value their prod-
ucts, the factors will be bid away to other more productive uses and their enter-
prises will eventually close. On the other hand, if entrepreneurs correctly appraise
factor prices in light of future consumer demand, they can maintain and increase
the value of the capital invested. Factors of production are priced competitively
according to their discounted marginal productivity, which is imputed to them by
consumers through the price system. Whether an enterprise is productive enough
Complementary enterprises charge market prices for their products, prices that
result from the interaction of the two sources of value mentioned above. Impor-
tantly, it makes a difference which one dominates in the minds of consumers. If
the social mission is not a source of value for consumers, social entrepreneurs will
have to compete for patronage based solely on their ability to produce valuable
products for direct consumption. In this case the visibility of the social mission
shrinks or even vanishes in the minds of customers. However, social entrepre-
neurs whose missions are considered valuable in their own right may be able to
charge a premium for their products. If so, social goods take on a different kind
of goods-character (to use Menger’s term) than products that do not serve social
The limits of social entrepreneurship
ends. They become similar to branded goods that may be physically similar to
lower-priced goods available on the market, but are different in the eyes of con-
sumers because they serve different ends.
Entrepreneurs know that they must price their inputs and outputs competitively.
Yet to do so they must have recourse to economic calculation. They must be able
to estimate the opportunity cost of each factor employed in production, and this
means external markets for each factor must exist that can be used to establish its
implicit price (Rothbard, 2009, pp.
Once an external price is available,
the economic calculation of pro�t and loss becomes feasible. If, however, there
is no external price, an “island of noncalculable chaos” appears (Rothbard, 2009,
613–614). The
�rm becomes unable to accurately appraise a part of its busi-
ness, and as a result its overall performance suffers. All mundane �rms avoid this
situation wherever possible in order to minimize losses. Consequently, they never
This result explains why it is impossible for One Big Firm to control all
production in the economy: it would suffer from the same calculation prob
lem as a socialist system. Importantly, it also means that no economy could
ever consist entirely of
producers’ cooperatives
, because such organizations
would also be unable to separate the prices paid for different factor contribu
tions (Rothbard, 2009, pp.
608–609). Likewise,
no social enterprise can grow
so large that it becomes the “unique producer and user of [a] capital good”
(Klein, 1996, p.
16), or
else its business division would be unable to success
fully allocate resources and generate income. Its performance would suffer,
and eventually funding for the social mission would disappear. Economic cal
culation thus places an upper bound on the size of both social enterprises and
Importantly, social enterprises may have a harder time �nding external prices
more than some conventional businesses do. The reason is that many social enter-
prises rely on innovative business models that do not have counterparts among
traditional enterprises. In particular, social ventures employ unique and previ-
ously unpriced factors of production, especially labor (Dacin, Dacin, and Matear,
2010). That is, they are more prone to calculation errors to the extent that they
allocate resources without the price system to guide them. As Klein observes,
“innovation carries with its bene�ts the cost of more severe internal distortions”
(1996, p.
17n16). The
most reliable cost estimates come from mundane businesses
because these organizations strive the most to appraise prices correctly. Social
enterprises are therefore most ef�cient in sectors where ordinary entrepreneurs
have established a well-developed network of prices that they can use to estimate
their own implicit costs. However, it is not always possible to �nd such a network
in practice because some social business customers �nd it dif�cult to pay even the
minimum required prices (Mair and Marti, 2006). As a result, social enterprises
often exist on the boundaries of economic calculation. Frank Fetter remarked
that entrepreneurs take “the more exposed frontier of risk” (Fetter, 1915, p.
Along similar
lines, we might say that social entrepreneurs take the more exposed
Matthew McCaffrey
We conclude that successful complementary social enterprises will adapt their
behavior to the same standards used by conventional businesses. That is, they will
encourage ef�cient production and avoid waste as much as possible. A
more dif-
question involves exactly how effective their social investments are without
prices to guide them. When the business is independent from the social mission,
the latter is effectively reduced to a charity. In turn, the lack of external prices
for goods and services means that charities will suffer from problems relating to
the inef�cient use of physical and human resources. Only when the social mis-
sion is more closely connected with the business venture do charities begin to
overcome these organizational problems. External social enterprises are the most
straightforward examples of social entrepreneurship at work. Consequently, they
provide the clearest picture of the role of economic calculation in social ventures.
However, they are also simple organizations that are ultimately quite similar to
conventional �rms. Relaxing some of our previous assumptions about them will
allow us to account for more complex forms of social entrepreneurship, as dem-
There are many ways that the business and social aspects of an enterprise
can be connected. This section examines the case of
, wherein the business venture and the social mission overlap. Unlike
complementary ventures, integrated social enterprises are not set up primar
ily to fund other organizations. Instead, they advance their social missions
directly through the operations of their businesses (Fowler, 2000). Typically,
this involves social �rms adopting organizational forms or special business
methods that are used as bases for entrepreneurial decision making. Consider
again the example of the sandwich producer. An integrated social enterprise
might maintain the same general business while also specifying that at least
25% of the company’s workforce be homeless persons. It thus serves con
sumers through its ordinary business, but also provides special consumption
bene�ts to any workers it employs who would not otherwise be able to �nd
employment (or who could not �nd it at the same wage rate). As mentioned
above, besides their speci�c social elements, such ventures are run like typical
In this scenario, we can no longer hold equal other values besides pro�t-
seeking. In this case, social value creation through business operates directly and
indirectly. As a result, the assumption of monetary pro�t-maximizing no longer
applies. Instead, integrated social enterprises attempt to reconcile different and
possibly con�icting objectives sometimes referred to as the “double bottom-line”
(Emerson and Twersky, 1996, p.
12). I
that this con�ict represents a trade-
off between ef�cient decision making and
To the extent that they insist on making unpro�table decisions, entrepreneurs
step outside the sphere of economic calculation: they strategically choose cer-
tain divisions in their businesses that are not subject to the pro�t and loss test.
The limits of social entrepreneurship
Lacking this guide, some divisions will tend to produce inef�ciently. In conven-
tional business, poor performance encourages entrepreneurs to scale back failing
departments and increase production in more pro�table lines; in social enterprise,
however, keeping an inef�cient division alive is often a key requirement of the
core mission. Social entrepreneurs have no choice but to subsidize failing divi-
sions using the pro�ts from more successful ones. This is consistent with empiri-
cal research showing that social enterprises tend to sacri�ce monetary pro�ts in
order to expand consumption among their target groups (Agafonow, 2015). It is
also consistent with the idea that all people, including social entrepreneurs, con-
stantly try to maximize psychic pro�t, but not necessarily monetary pro�t (Mises,
the restaurant example, suppose one of the homeless workers earns wages
greater than his discounted marginal productivity.
In a conventional business, the
worker’s wages will be bid down. However, what if the entrepreneur is committed
to keeping the worker employed? One consequence is that any wage the employee
receives above his marginal productivity is actually a gift from the entrepreneur
that is subtracted from the value of some other part of the business. The source of
the gift may be the entrepreneur’s pro�ts, the capital of the enterprise, the land of
the enterprise, or the wages of other employees if they are willing to forego part
of their potential earnings, as in the case of volunteers for a charitable cause.
The gift reduces the pro�tability of the organization, but maximizes the social
entrepreneur’s psychic pro�t (Rothbard, 2011, pp.
neurs can avoid monetary losses by convincing consumers to pay higher prices
for social goods in order to compensate for what would otherwise be arti�cially
It is not always easy in practice to separate factor payments from gifts. As
The boundaries between buying goods and services needed and giving alms
are sometimes dif�cult to discern. He who buys at a charity sale usually com-
bines a purchase with a donation for a charitable purpose
. Man
in acting is
a unity. The businessman who owns the whole �rm may sometimes efface the
boundaries between business and charity. If he wants to relieve a distressed
friend, delicacy of feeling may prompt him to resort to a procedure which
spares the latter the embarrassment of living on alms. He gives the friend a
job in his of�ce although he does not need his help or could hire an equivalent
helper at a lower salary. Then the salary granted appears formally as a part of
business outlays. In fact it is the spending of a fraction of the businessman’s
income. It is, from a correct point of view, consumption and not an expendi-
ture designed to increase the �rm’s pro�t.
again, economic calculation acts as a boundary on social entrepreneurs’
choices. The greater the number of decisions that entrepreneurs make outside the
sphere of economic calculation, the more inef�cient their businesses will become.
Matthew McCaffrey
In the most extreme case, every decision is treated as social, and there is no scope
for economic calculation at all. Essentially, the social enterprise becomes a con-
ventional charity. We can infer from this that if entrepreneurs want to effectively
deliver on their promises to alleviate social problems, they will strive wherever
possible to use the price system so as to limit inef�ciencies as much as possible.
They can then focus their attention on a small number of non-calculable decisions
that deliver the bulk of the social value generated by the enterprise. But the over-
arching problem is that integrated social enterprises cannot expand their social
The above discussions focused on two speci�c types of social enterprise organ-
ization. However, real-world social businesses are not limited only to these forms:
there is a wide range of combinations of goals that social entrepreneurs might
choose. Furthermore, social organizations can change in response to internal and
external stimuli. Uncertainty and complexity thus make social organizations dif-
�cult to describe in terms of universal principles. Yet despite this fact, we can still
First, there is a spectrum of involvement between the business and social
aspects of social enterprises (Dees, 1998; Thompson, 2002; Tan, Williams, and
Tan, 2005; Peredo and McLean, 2006). The distinct role of the business, and of
economic calculation, becomes clearer the more the two sides are separated. Con-
versely, the greater the number of factors of production that entrepreneurs hold
outside the pricing process, the smaller the scope for economic calculation, effec-
tive resource allocation, and ultimately, delivery of social value. As a result, social
organizations that want to pursue their goals in the most effective way will use the
This argument also applies to the increasing number of pro�t-seeking �rms
that include social elements in their business models without fundamentally alter-
ing their objective of delivering pro�ts to shareholders. For example, Corporate
Social Responsibility (CSR) guidelines offer �rms the chance to contribute to
social causes outside their typical business operations.
These self-imposed rules
can help to increase business through generating publicity, but they also involve
�rms making gifts to speci�c producers. For example, in order to support a devel-
oping economy or encourage the ethical sourcing of products, a �rm can buy from
a supplier at prices higher than could be obtained elsewhere. Or it could commit
to donating a small
percentage of
its pro�ts to charity, thus making a gift out of
what might otherwise have been entrepreneurial income. In any case, if CSR or
similar guidelines crowd out the price system, the scope for ef�cient decision
Second, social enterprises are in�uenced in many ways by government regula
tion and monopoly privileges. The more the system of government intervention
– especially
government’s role as a major �nancier of social organiza
– the
more economic calculation is hampered and distorted. It remains an
open question, though, exactly how these distortions will affect social enterprises.
Does intervention encourage arti�cial growth in the number or size of social ven
tures? Or does it have the opposite effect by monopolizing traditional industries and
The limits of social entrepreneurship
excluding new, alternative kinds of �rms? Furthermore, what is the major cause
of entrepreneurs’ interest in alternative forms of economic organization? Is social
enterprise, as sometimes argued in the literature, a response to market failures? Or
is it a response to public policy failures that create and institutionalize the kind of
marginalized groups that social enterprises seek to help, or who might be likely to
found a social enterprise? A
full analysis
of the relationship between social enter
prises and government intervention is beyond the scope of this paper. However, the
arguments made above lay the foundation for answering these and other questions.
The theory presented in this chapter
addresses important
problems in both eco-
nomics and entrepreneurship research. From an economic perspective, it offers
a novel way to explain the organization of social enterprises, which have until
now been studied mostly by the management disciplines. Economics provides
wide-ranging theories of social interaction, value, calculation, pro�t, and pric-
ing that can be used to rigorously de�ne the domain of social entrepreneurship.
Applying these theories expands the scope of economic reasoning and can pro-
vide a realistic account of many types of economic organization in the voluntary
and public sectors. Doing so improves our understanding of these sectors while
reinforcing the universality
– or
– of
several essential principles of
economics. It also undermines the criticism that economics is unconcerned with
or unable to explain behaviors and organizations that fall outside the sphere of
maximization (McCaffrey
The economic approach outlined above also helps to resolve some of the
most controversial questions in social entrepreneurship research (e.g. Austin,
Stevenson, and Wei-Skillern, 2006). For example, the idea of economic calcu
lation explains the tension between two general theories of social enterprise:
those focusing on pro�t-maximization and those that emphasize avoiding mis
sion drift (Agafonow, 2015). As explained in this chapter, this tension actually
re�ects the boundary between economic calculation and non-calculation. Prices
and gifts play an endless game of tug-of-war in social enterprise. If one side
wins, social businesses end up behaving like traditional �rms or voluntary sector
The question of how to measure the ef�cacy of social enterprises is also ubiq-
uitous in the literature (Dees, 2001; Zahra et
al., 2009;
Dacin, Dacin, and Matear,
2010). Fortunately, economic calculation provides a basis for estimating the true
social value of social enterprises, which is demonstrated from moment to moment
through the actions of consumers and the reactions of entrepreneurs and the price
system (Salerno, 1993, pp.
52–53). Of
course, calculation cannot be used to esti-
mate every possible conception of value, but it can at least be used to ensure
that social businesses support their social missions in the most effective ways,
for example by supporting “sustainable” practices. This is a noteworthy advance,
because resources are often stretched thin in social organizations, which urgently
Matthew McCaffrey
Social entrepreneurship is only one of many innovative forms of economic
organization to emerge in the last few decades. Its increasing presence in com-
mercial society makes it an important area of study for economists, who have
sometimes limited their research to more traditional forms of business. However,
this chapter
has shown
that economic theory is well-suited to the study of social
organizations, which present a valuable opportunity for economists to improve
their ideas and extend them to related �elds like management studies. At the same
time, economic arguments can also serve to encourage entrepreneurship research-
Cf. Klein (2008) for an overview of the “mundane” economic theorizing of the causal-
Cf. Zahra
al. (2009)
and Dacin, Dacin, and Matear (2010) for surveys of different
de�nitions of social entrepreneurship. Although the term is often used loosely and
inconsistently, the general approach taken in this chapter
is consistent
with many cur-
rent de�nitions. Furthermore, as I
explain, by
relaxing some assumptions and expand-
As Mises
explains: “Even an action directly aiming at the improvement of other peo-
ple’s conditions is sel�sh [in the sense of pursuing psychic pro�t]. The actor considers
it as more satisfactory for himself to make other people eat than to eat himself. His
uneasiness is caused by the awareness of the fact that other people are in want” (Mises,
This chapter
uses the
term “entrepreneurship” in a broader sense than some economic
literature, to refer to judgmental decision making about the use of scarce, heterogene-
ous resources under conditions of uncertainty (Foss and Klein, 2012). Such decisions
take place in many different contexts, as illustrated in Table
However, it can be
argued that true entrepreneurship is inextricable from economic calculation, and that
terms such as “political entrepreneurship” or “charitable entrepreneurship” are there-
fore contradictory. However, this approach does not alter the underlying arguments
made in this chapter, which could be rephrased without changing any of the conclu-
Kirzner (201
1) imagines a scenario where all businesses in the economy are essen
tially complementary social enterprises. He concludes that such a situation would
produce the same “ruthless” competition as an economy �lled with “sel�sh”
This fact
explains why social enterprise is sometimes less attractive to prospective
entrepreneurs than conventional business: social enterprises are more prone to failure
due to calculation problems, and this danger is not necessarily offset by the prospect of
In practice,
social enterprises often include more than one social element in their busi-
ness models. However, this fact does not alter the arguments in this section, which are
valid even if there are several “islands of noncalculable chaos” within a social �rm.
This case of multiple social elements actually helps emphasize the importance of exter-
Throughout, I
assume that
the lack of calculation leads to losses. It is possible, though,
that wages set by a social entrepreneur happen to be equal to the worker’s marginal
productivity. Yet if this were the case, there would be no need for a special rule about
hiring particular kinds of employees: entrepreneurs could simply hire those workers
who are expected to be the most productive. Presumably, such rules exist because
The limits of social entrepreneurship
social entrepreneurs recognize that their target employees will be relatively inef�cient.
Furthermore, coincidences in labor pricing are unlikely to occur systematically: labor
markets tend to be ef�cient precisely because entrepreneurs can use trial and error
via the price system to guide their negotiations with workers. If this process is not
used by social ventures, there is no reason to think wages will tend toward marginal
Cf. the
discussions of psychic and monetary income in labor markets in Rothbard
Baron (2007)
actually de�nes social entrepreneurship as the willingness to found a
Abu-Saifan, S. 2012. “Social Entrepreneurship: De�nitions and Boundaries.”
Agafonow, A. 2015. “Value Creation, Value Capture, and Value Devolution: Where Do
S. H., L. D. Brown, and Christine W. Letts. 2004. “Social Entrepreneurship and
Societal Transformation an Exploratory Study.”
Journal of Applied Behavioral Science
Austin, J., H. Stevenson, and J. Wei-Skillern. 2006. “Social and Commercial Entrepreneur-
ship: Same, Different, or Both?”
Entrepreneurship Theory and Practice
Baron, D. P. 2007. “Corporate Social Responsibility and Social Entrepreneurship.”
P. N. 2012. “Introduction to the Special Section on Social Entrepreneurship.”
Dacin, P. A., M. T. Dacin, and M. Matear. 2010. “Social Entrepreneurship: Why We Don’t
Need a New Theory and How We Move Forward From Here.”
Harvard Business Review
“The Meaning of ‘Social Entrepreneurship.’
Center for the Advancement
Social Entrepreneurship
Emerson, Jed, and Fay Twersky (eds.). 1996.
New Social Entrepreneurs: The Success,
Challenge and Lessons of Non-Pro�t Enterprise Creation
. San Francisco: The Roberts
Fetter, F. A. 1915.
. New York: The Century Co.
Foss, N. J., and P. G. Klein. 2012.
Organizing Entrepreneurial Judgment: A
New Appr
Foss, N. J., and P. G. Klein. 2015. “Introduction to a Forum on the Judgment-Based
Approach to Entrepreneurship: Accomplishments, Challenges, New Directions.”
11 (3): 585–599.
Fowler, A. 2000. “NGDOs as a Moment in History: Beyond Aid to Social Entrepreneurship
Third World Quarterly
Harding, R., 2004. “Social Enterprise: The New Economic Engine?”
Haugh, H. 2006. “Social Enterprise: Beyond Economic Outcomes and Individual Returns.”
In J. Mair, J. Robinson, and K. Hockerts, eds.,
Social Entrepreneurship
. Basingstoke,
Matthew McCaffrey
Herbener, J. M. 2018. “Time and the Theory of Cost.” In Matthew McCaffrey, ed.,
Economic Theory of Costs: Foundations and New Directions
. Abingdon, UK: Rout-
Kirzner, I. 2011. “The Economics of Greed or the Economics of Purpose.”
The Annual
Proceedings of the Wealth and Well Being of Nations
Klein, P.G. 2008. “The Mundane Economics of the Austrian School.”
11 (3 and 4): 165–187.
———. “Economic Calculation and the Limits of Organization.”
Review of Austrian Eco
Leadbeater, C. 1997. The Rise of the Social Entrepreneur. London: Demos.
McCaffrey, M. 2011. “On Government Investment and Consumption.”
New Perspectives
“Economic Policy and Entrepreneurship: Alertness or Judgment?” In
P. Bylund and D. Howden, eds.,
The Next Generation of Austrian Economics: Essays in
Honor of Joseph T. Salerno
. Auburn, AL: Ludwig von Mises Institute.
“What’s Love Got to Do With It? Action, Exchange, and Gifts in Eco-
nomic Theory.”
Quarterly Journal of Austrian Economics
“Review of ‘Austrian Theory and Economic Organization: Reaching
Beyond Free Market Boundaries’.”
Quarterly Journal of Austrian Economics
18 (3):
McCaffrey, M. and J. T. Salerno. 2011. “A
Theory of
Political Entrepreneurship.”
Machaj, M. 2007. “Market Socialism and the Property Problem: Different Perspec-
tive of the Socialist Calculation Debate.”
Quarterly Journal of Austrian Economics

Mair, J. and I. Marti. 2006. “Social Entrepreneurship Research: A
Source of
Journal of World Business
Martin, R. L., and S. Osberg. 2007. “Social Enterprise: The Case for De�nition.”
. New Haven, CT: Yale University Press.
———. 1990a.
Economic Calculation in the Socialist Commonwealth
. Auburn, AL: Lud-
“Observations on the Cooperative Movement.” In
Human Action: Scholar’s Edition
. Auburn, AL: Ludwig von Mises
Problems of Economics
. 3rd ed. Auburn, AL: Ludwig von
Mort, G. S., J. Weerawardena, and K. Carnegie. 2003. “Social Entrepreneurship: Towards
International Journal of Nonpro�t and Voluntary Sector Marketing
Newman, J. 2018. “Contemporary Debates on Opportunity Cost Theory and Pedagogy.” In
Matthew McCaffrey, ed.,
The Economic Theory of Costs: Foundations and New Direc
. Abingdon, UK: Routledge, pp. 11–26.
Peredo, A. M., and M. McLean. 2006. “Social Entrepreneurship: A
Critical Review
of the
Journal of World Business
Rothbard, M. N. 2009.
1. “The Myth of Neutral Taxation.” In
Economic Controversies
. Auburn, AL:
The limits of social entrepreneurship
Salerno, J. T. 1990a. “Ludwig von Mises as Social Rationalist.”
Review of Austrian Eco
“Postscript: Why a Socialist Economy is ‘Impossible’.” In
. Auburn, AL: The Ludwig von Mises Institute.
“Mises and Hayek Dehomogenized.”
Review of Austrian Economics
6 (2):
———. 2018. “The ‘Income Effect’ in Causal-Realist Price Theory.” In
The Economic
Theory of Costs: Foundation and New Directions
. Ed. Matthew McCaffrey. Abingdon,
Schumpeter, J. A. 1942.
Capitalism, Socialism, and Democracy
. New York: Harper and
Seelos, C. and J. Mair. 2005. “Social Entrepreneurship: Creating New Business Models to
Serve the Poor.”
Shane, S., E. A. Locke, and C. J. Collins. 2003. “Entrepreneurial Motivation.”
Resource Management Review
Tan, W. L., J. Williams, and T. M. Tan. 2005. “De�ning the ‘Social’ in ‘Social Entrepre-
neurship’: Altruism and Entrepreneurship.”
International Entrepreneurship and Man
Thompson, J. L. 2002. “The World of the Social Entrepreneur.”
Zahra, S. A., E. Gedajlovic, D. O. Neubaum, and J. M. Shulman. 2009. “A
Social Entrepreneurs: Motives, Search Processes and Ethical Challenges.”
Business Venturing
marginal revenue products
Alford, R. F. G. 43n6
anticipated discounted marginal revenue
Ashenfelter, Orley C. 172
Austro-Wicksellian theory 46n31, 54
Baron, D. P. 261n10
Blair, Roger D. 169
Block, Walter 166n2, 187n15, 222n11
Boyes, William J. 24n6
Veil of Money
causal-realist theory 2, 4 – 5, 6, 7, 11 – 13,
costs and pricing; income effect” in
of the Firm” 210 – 211, 216, 219
competitive prices 5, 53, 55, 56, 57, 65, 74,
95, 111, 174, 176; monopoly 56, 83 – 90,
110, 127n15, 127n17, 173, 185
66, 67 – 68, 115, 116, 129n49, 154, 155,
Farber, Henry 172
Ferraro, Paul J. 11, 15 – 16
Fetter, Frank A. 2, 43, 43n2, 54, 55,
Fisher, Irving 43n3, 140;
The Nature of
forgone opportunities 12, 15, 24n14
Giffen Paradox 27, 43n7, 44n8
111, 121, 202n5
Harrison, Jeffrey L. 169
Hayek, Friedrich August 44n15, 51, 52,
Herbener, Jeffrey M. 6, 24n10, 24n15, 229
Hirschleifer, David 69n8
Hülsmann, Jörg Guido 4, 212, 213
69n11, 83, 172, 173, 174; competitive-
costs and pricing 191 – 201; Austrian take
challenges to neoclassical marginalism
Dacin, M. T. 260n2
Dacin, P. A. 260n2
Dahlman, 211, 222n12
Denis, Andy 225 – 226, 232
discounted marginal revenue
discounted marginal revenue products
so anticipated discounted marginal
enterprise �rm 253 – 259; marginal
Marget, Arthur 44n13, 45n22
marginal costs 5, 53, 61, 139, 171, 174,
marginal net income 117 – 120, 121, 122,
marginal rate of production substitution
marginal revenue product (MRP) 62, 63,
marginal unit of output in production
marginal utility 2, 3, 27, 28, 29, 30, 31, 33,
46n30, 54, 76, 116, 162, 198, 199, 201;
marginal pairs 31, 54, 162
Marshall, Alfred 2, 29, 41, 43n3, 43n6,
Mason, Will E. 43n3
Matear, M. 260n2
McCaffrey, Matthew 6, 23n4
McGraw, Hill 69n13
Menger, Carl 2, 3, 6, 27, 43n2, 44n14, 51,
221n3; calculation argument 229;
difference between entrepreneurs and
Epistemological Problems of Economics
“income effect” in causal-realist price
substitution effect 4, 27, 28, 36 – 37, 38,
Jevons, William Stanley 2, 44n14
11, 16
Kirzner, L. M. 44n11, 55, 59, 69nn10 – 11,
The Screwtape Letters
long-run marginal cost (LMC) 152 – 153
Mankiw, N. Gregory 13, 14
Manning, Alan 169, 170, 172, 173, 186n2
110, 127n15, 127n17, 173, 185
monopsony theory 5, 169 – 186, 186n6,
186n8, 187n9, 187nn13 – 15, 187n17;
classic and new theory 171 – 172;
gap with non-speci�c factors and the
gap with speci�c factors of production
179 – 181; monopoly theory without
imperfect/perfect competition dichotomy
Mortensen, Dale T. 172
Murphy, Robert P. 14, 23
prices 148; income effect 43; interest
long-run costs 152; marginalism 150,
162, 191 – 194, 195, 196; MRP 62;
net income 32, 65, 74, 108, 113 – 126,
, 129n47, 177, 179, 180; rate of 113,
114, 115, 116, 119 – 120, 121, 124 – 125
Nordhaus, W. D. 20, 22
Occam’s Razor 195
offer curve 41, 46n32
opportunity cost theory 11 – 23; Braun
follows Reisman 22 – 23; causal-realist
theory 11 – 13; critical views in Austrian
literature 20 – 23; critics 18 – 19; de�nition
11 – 15; Ferraro and Taylor spark debate
15 – 16; mainstream confusion 15 – 20;
Parkin’s “reexamination” 16 – 18;
production tradeoffs 19 – 20; textbook
208; income effect 38; “intellectual
175; marginal revenue 200 – 201;
mechanical metaphors 211 – 212; money
assets 33; monopoly price-gap 176 – 177,
69n11, 170, 174; praxeology 45n26,
substitution effect 23n4, 45n26; supply
of Money and Credit
113; net 108, 111, 116, 117, 119, 120,
68, 69n7, 74, 110, 111, 127n15, 173,
Ritenour, Shawn 14 – 15, 23
‘Cost’ is simply the utility of the next
curve 69n11; entrepreneurship 145n5,
151, 156; equilibrium 187n11; ERE
dichotomy 174; income effect 43;
income taxation 27, 38, 42; individual’s
value scale 44n11; labor factors 175;
Man, Economy, and State
marginal unit of output in production
producer’s activity 73 – 126; product
Parkin, Michael 16 – 19, 24n11
Perry, Arthur Latham 127n24
Pigou, A. C. 38, 41, 46n30
199, 201, 201n3; Austro- 191 – 195
production tradeoffs 13, 14, 16;
cost 248; ends 250; factors 77, 115, 124;
purchasing power of money 28 – 29, 32, 34,
Quarterly Journal of Austrian Economics
Rajsic, P. 46n29
Reisman, George 13, 23, 24n12, 24n14;
true value speci�cation 11, 17 – 18
Tucker, Benjamin R. 90
price 115, 160; immediate-run price,
system 250, 254; rate of return 66, 116;
32, 33, 35, 36, 42, 43, 44n11, 77, 89,
116, 120, 128n32, 252; ordinal 28, 38,
value speci�cation 11, 16 – 18
Volker Fund 53
Walras, Léon 2, 27, 52, 62, 69n2, 198
Weiler, Emanuel T. 54, 61, 69n12
Wells, Robin 14, 23n3
Wicksteed, Philip 2, 40, 43, 43n2,
44nn11 – 12, 44n14, 45n19, 45n25,
Williamson, Oliver E. 207, 216, 221n6,
Technology of Transacting” 210 – 211
Yeager, Leland 28, 30, 43n4, 44n8
153 – 154; “Toward a Reconstruction of
Utility and Welfare Economics” 24n5;
Salerno, Joseph T. 3, 4, 198, 240n5
Salin, P. 38
Samuelson, Paul A. 20, 22
Smith, Adam 127n24
Stigler, George 52, 54, 58, 62, 63, 69n12
Stone, Daniel 18, 24n11
substitution effect 4, 27, 28, 36 – 37, 38,
Taylor, Laura O. 11, 15 – 16
technology 6, 14, 55, 90 – 95, 99, 108, 115,
117, 119, 128n35, 153, 210, 234
theory of investment 64, 67, 75, 114
153 – 160; marginal costs and production
Topan, Mihai-Vladimir 5

Приложенные файлы

  • pdf 18415772
    Размер файла: 4 MB Загрузок: 0

Добавить комментарий