===== PAGE 1 =====

COLLECTED WORKS OF CARL MENGER

Since 1930, when the first of its Reprints of Scarce Tracts in
Economic and Political Science appeared, the London School
of Economics has made one important contribution after
another to filling in the gaps in contemporary economics
libraries. The works selected by the editors for reprinting
have not only been of outstanding importance in the history
of economic thought, but have also been specially welcome
because of their inaccessibility to the majority of students in
the field. Menger’s collected works, the latest addition to the
serics, rank high in both respects. Menger’s importance in
the history of economic thought is too well recognized to
require special comment. As to the inaccessibility of his
works, the present writer can testify from personal experience
that even in Vienna the first edition of the Grundsédtze and
the two books on methodeclogy' are the greatest of rarities.
Professor Hayek appears fully justified in attributing to this
scarcity the fact that theories originally developed by Menger
himself are known to the world chiefly through the writings
of his disciples Bohm-Bawerk and Wieser.? Economists are
bound to feel genuinely grateful to the London School for
meeting so real a need.

I

To Menger, more than to any other single writer, belongs
the credit — and the responsibility — for the subjective value
departure in economics. The appearance of his works in a
form which, for the first time in fifty years, makes possible
their widespread distribution furnishes a good occasion for a
review of what that departure has meant. Such a review can

1. Untersuchungen u.d. Methode der Sozialwissenschaften und der
Politischen Oekonomie insbesondere and Irrtiimer des Historismus.

2. Introduction to the Collected Works, 1, xxv. Menger’s Grund-
sitze had already been out of print for some time by the nineties. He

refused, however, to allow a reprint, in the expectation — unrealized in
his lifetime — of bringing out a second, completely revised edition.

719


===== PAGE 2 =====

720 QUARTERLY JOURNAL OF ECONOMICS

furthermore be undertaken today with a much better prospect
of attaining both clarity and agreement than at any previous
time in the last half century.

Subjective value theory has, since its inception, been one
of the major storm centers of economics. Proponents have
made extravagant and dogmatic claims as to its importance,
while opponents, sensing something wrong and yet failing to
put their fingers on the precise spot, have denounced the
whole subjective analysis as a tangle of verbal duplications
and terminological ambiguities. Careful analysis of the prob-
lems involved shows that there should be little cause for sur-
prise at this situation. Many of the concepts involved are
slippery in the extreme, and distinctions which are essential
to clarity are easily overlooked.

That the prospect for mutual understanding has greatly
improved in the last few years is attributable basically to the
advances that have been made in the general field of scien-
tific methodology, the effects of which are now being felt in
the social sciences. A number of the post-war generation of
sociologists and economists have, in particular, been doing
excellent work in recent years in applying the lessons of
modern scientific criticism to the problems of their own
special fields. As examples of this development such widely
separated and diverse writers as Professor Gunnar Myrdal
of Stockholm, Dr. Ralph W. Souter, formerly of Columbia,
and Dr. Felix Kaufmann of Vienna, may be mentioned.?

The last named scholar is particularly important for pur-
poses of the present discussion. For one thing, his recently
published Methodenlehre der Sozialwissenschaften offers

3. Myrdal: Das politische Element in der Nationalokonomischen

Doktrinbildung. Berlin. 1932.
“Das Zweck-Mittel Denken in der Nationalokon-
omie,”’ Zeitschrift fiir National6konomie, vol. 4.

Souter: Prolegomena to a Relativity Economics. New York.
1933.

Kaufmann: Methodenlehre der Sozialwissenschaften. Vienna.
1936. “On the Subject-Matter and Method of Eco-
nomic Science,”’ Economica, vol. 13.
“The Concept of Law in Economic Science,” Review
of Economic Studies, Feb. 1934.


===== PAGE 3 =====

COLLECTED WORKS OF CARL MENGER 721

probably the most comprehensive survey of the methodo-
logical problems of the social sciences as yet undertaken in
the light of modern logical and scientific advances. In addi-
tion to this Dr. Kaufmann is an Austrian who grew up in
close personal contact with the members of the Austrian
school of economists. His analysis of utility theory is conse-
quently developed with an intimate and sympathetic under-
standing of what the utility theorists were trying to do.

In what follows I do not propose to undertake anything
like a general survey of Menger’s works. Within the limits
of economic theory, Menger applied himself to a considerable
diversity of problems, cach one of which he handled with
characteristic firmness and originality. But all these parts
of his work are of secondary importance compared with his
introduction of the subjective value approach into economics.
It is accordingly with the latter and the problems arising out
of it that the following discussion will be concerned.

II

We may start with a summary of the major assumptions
involved in the subjective value approach to the problems of
economics: (1) prices (or exchange values) can best be
explained by reference to the choice acts of individual con-
sumers; (2) consumers by and large act in accordance with
coherently formulated plans, and hence prediction as to what
they will do is possible, at least in principle;® (3) certain
general features of the internal structure of consumers’ plans
are sufficiently obvious to provide the basis for a skeleton

4. Most important are his treatment of capital theory, his contri-
butions to the theory of money, and his remarks on the problem of
imputation (Grundsiitze, pp. 138-142). All of these are discussed in
Professor Hayek’s introduction to the Collected Works.

5. This is the assumption of rational conduct, i.e., conduct in accord-
ance with a plan, which will be discussed in more detail below. It is
interesting to compare Max Weber's often cited formulation of this
assumption: “Marginal utility theory, for its own particular purposes,
treats human conduct as though it ran from A to Z under the control
of commercial calculation; i.e., calculation based on the knowledge of
all relevant conditions.” Gesammelte Aufsétze zur Wissenschaftslehre,

p- 370.


===== PAGE 4 =====

722 QUARTERLY JOURNAL OF ECONOMICS

theory of price formation; (4) detailed study of consumer
habits may well yield information about the way certain
groups or classes of consumers are likely to react in various
concrete situations and thus to prove useful for such pur-
poses as, for instance, predicting how the demand for a certain
commodity will change with changes in its price or in its
buyers’ incomes.

It will be convenient to leave the first assumption for later
consideration and to begin with an analysis of the second.
The meaning of conduct in accordance with a plan is formu-
lated for the purposes of subjective value theory in the mar-
ginal utility principle. Put in the most neutral way possible,
the marginal utility principle relates the relative importance
attached to different goods by the individual consumer to
the relative importance of possible uses for the goods, which
the consumer already has worked out in his head.® The
uscs, with their relative importance, constitute a system of
ends the means to the attainment of which are the available
goods. In so far as he acts rationally, the consumer will
estimate the relative importance to himself of any unit of a
good by the relative importance of the use he would have to
sacrifice if he were to lose the unit in question. He will,
Jurthermore, prefer any unit of a good to any other unit of
the same cr a different good if the importance he attaches to
the use that would have to be sacrificed without the first
unit is greater than that attached to the use that is depend-
ent on the second unit.

Now the important thing to be noted here is that the
marginal utility principle as stated can be interpreted in two
quite different ways, failure to distinguish between which
has proved a source of endless confusion. Looked at in one
way, it is merely a definition of rational conduct. What the
theorist is in effect saying is: “When I speak of rational
evaluation of goods on the part of a consumer, I shall mean

6. The basis for the individuals decisions, i.e., whether in response
to physiological need, aesthetic taste, or what not, as to which uses are

more and which are less important is not an essential part of the utility
principle itself. Cf. the discussion of needs below.


===== PAGE 5 =====

COLLECTED WORKS OF CARL MENGER 723

evaluation in accordance with a hierarchy of uses which is
characterized by such and such internal relationships.””
Since the marginal utility principle, so regarded, merely
defines rational conduct there can be no question of its veri-
fication in practice. In fact it may be entirely unrelated to
anything in reality, since our ability to conceive of action in
accordance with an internally consistent plan is quite inde-
pendent of the way people actually do act.

But the marginal utility principle may also be intended as
a hypothetical assertion of fact. Its meaning in this case is:
“If people act rationally, then they will, as a general thing,
estimate the relative importance of different goods in accord-
ance with a scale of uses which exhibits such and such char-
acteristics,” the necessary implication being that it is possible
to determine independently of the specific acts predicted
whether or not they are acting rationally. The accuracy of
the prediction is then subject, in any case or group of cases
where it is applied, to the test of empirical verification.

The failure to distinguish between these two interpreta-
tions is, partly at least, attributable to the fact that the
definition is frequently stated in the same form as the asser-
tion of fact. That is, the sentence: ‘If a man acts rationally,
he will do so and so” may be interpreted in either way.3
Beneath the verbal ambiguity, however, is the more funda-
mental failure to draw a sharp distinction between purely
logical deductions, that is, the working out of implications

7. The specification of this internal structure of the system of ends
may be more or less complete according to the purpose the theorist
has in mind in constructing his concept. Cf. Kaufmann, Methoden-
lehre, p. 261.

8. At the risk of elaborating the obvious, this point may be restated
as follows: Is it possible for a rational being to act otherwise than in
accordance with the marginal utility principle? If the marginal utility
principle is viewed merely as a definition of rational conduct then it
clearly is not, since we are agreeing to call only that conduct rational
which exhibits the characteristics specified in the marginal utility
principle. If, however, what we mean to say in stating the marginal
utility principle is that we think rational people as a general thing do
and will act in the specified manner, then we of course have to admit
the possibility of exceptions and it is perfectly conceivable that a

rational being — however identified — would not act in accordance
with the marginal utility principle.


===== PAGE 6 =====

724 QUARTERLY JOURNAL OF ECONOMICS

already “contained in”’ the initial concept or system of con-
cepts, and empirical laws, which are predictions as to the
course of events in the real world.

The failure to observe this distinction, which has been
widespread among economists, comes out with particular
force in the methodological controversy between Menger and
the German Historical School, led by Schmoller. Menger
based his whole argument on what he conceived to be a
fundamental distinction between mere “empirical laws,” i.e.,
‘““observed regularities in the sequence and coexistence of
econcmic phenomena’? and exact theoretical laws, which,
arrived at by the process of abstraction, reduce the compli-
cated phenomena of reality to their “simplest, rigidly typical
elements.’

This notion that in economics there are two kinds of laws
— ‘“mere’’ empirical generalizations and “exact’’ theoretical
laws — both of which in some obscure way are applicable to
reality, has persisted down to the present time. Recent
examples may be found, for instance, in the writings of Pro-
fessor Robbins and Professor Hayek.?

In his recently published essay on ‘Collectivist Economic
Planning” Professor Hayek states categorically that since
experiment in the social sciences is “impossible” we can have
“no knowledge of definite regularities in the complex phe-
nomena in the same sense as we have in the natural sciences.”
There is, however, a different sense in which the social sci-
ences can know about the phenomena they are interested in.
Professor Hayek explains the difference as follows: “There

9. Irrtiimer des Historismus, p. 18.

1. Altho the Germans lacked the critical tools necessary to work out
their position in a thoroly satisfactory way, it seems to me that they
were guided by the sounder instinct of the two. Cf., for instance,
Schmoller’s warning that starting from hypotheses only hypothetical
assertions can be made, ‘to which one should not attempt, through
the use of the misleading word ‘exact,’ to give a false appearance of
scientific rigor (den Schein der strengsten Wissenschaftlichkeit).” Jahr-
buch fiir Gesetzgebung, 1883, p. 243.

2. Robbins, The Nature and Significance of Economic Science,

London, 1932. Hayek, Collectivist Economic Planning, London,
1935.


===== PAGE 7 =====

COLLECTED WORKS OF CARL MENGER 725

can be no doubt the social as well as natural sciences have to
employ deductive reasoning. The essential difference is that
in the natural sciences the process of deduction has to start
from some hypothesis which is the result of inductive generali-
zations, while in the social sciences it starts from known
empirical elements and uses them to find the regularities in
the complex phenomena which direct observation cannot
establish.” (Italics mine.) Itis too bad that Professor Hayek
does not enlighten us further as to the nature of the “known
empirical elements” which so conveniently absolve the social
sciences from the necessity of making ‘inductive generaliza-
tions.” In the absence of such enlightenment it is impossible
to determine more exactly the root of this peculiar concept
of law as applied to the social sciences.

A related misunderstanding, tho in its origin apparently
not identical, has been pointed out by Mr. T. W.Hutchison,?
in Professor Robbins’ Nature and Significance of Economic
Science: “In a recent pronouncement on these problems, the
following occurs: ‘It is this inevitability of economic analysis
which gives it its very considerable prognostic value. It has
been emphasized sufficiently already that Economic Science
knows no way of predicting what will be the given data at
any particular point of time. . . . But given the data in a par-
ticular situation it can draw inevitable conclusions as to their
implications.” A contradicticn appears at once, and a failure
to distinguish between implication and inference may be at
the bottom of it. This distinction may be summed up by
saying that implications take the form ‘if p, then q,” and
inferences the form ‘since p, therefore q.” The one shows
formal connections, the other states conclusions about the
real world. An inevitable implication is a tautology, and can
no more prognosticate anything than can the multiplication
table. An inference, on the other hand, requires some data
to be known in order to establish a ‘since.” But the foregoing
quotation says that ¢ Economic Science knows no way of pre-
dicting what will be the given data at any particular point of

3. “A Note on Tautologies,” The Review of Economic Studies,
February, 1935.


===== PAGE 8 =====

726 QUARTERLY JOURNAL OF ECONOMICS

time.” It follows, if this is to be taken au pied de la lettre, that
Economic Science can predict nothing.”

III

To this logical confusion must be added as a further source
of misunderstanding the elusive nature of the ‘“facts’”’ with
which utility theory deals. We have just seen that in its
attempt to explain exchange relationships utility theory goes
back to a scale of valuations which is supposed to exist in the
head of the consumer prior to his coming into the market.
On the basis of what is “known’’ about this scale of valua-
tions certain predictions are, in principle at least, possible as
to what the person in question will do. But the separation
of the known from the unknown in dealing with such psycho-
logical data as subjective scales of valuation is an extremely
slippery job. It is little wonder that most utility theorists
have fallen into the logical error of inferring the consumer’s
scale of valuations from his actual behavior in the market,
while at the same time pretending to explain the behavior on
the basis of the preéxisting valuations.

In the case of writers like Professor Mises,* who specifically
exclude any attempt to utilize psychological or biological
data, the circularity of reasoning involved may be so obvious
as to be harmless. But it is almost sure to be misleading in
the case of utility theorists who, like Menger himself, take
as their starting point a study of concrete needs and wants.5
The purpose of such a study is to discover in the relative
intensity of the different needs a clue as to the relative impor-
tance of the different uses for goods. Now the danger here
lies precisely in the fact that to a limited extent it is possible
to make significant statements about the order of intensity

4. Mises, Grundprobleme der Nationalékonomie, Vienna, 1933.
“People forgot that we can only know the need from the behavior, and
that consequently the concept of behavior which does not correspond
to the needs is nonsensical.” P. 140.

5. “The theory of needs is of fundamental importance for economics,
and at the same time the bridge which leads from the natural sciences,

especially biology, to the social sciences.” Grundsiitze der Volkswirt-
schaftslehre (2d ed.), p. 1.


===== PAGE 9 =====

COLLECTED WORKS OF CARL MENGER 727

of at least the simpler human needs and desires.®* Examples
of this can be found in the writings of all the earlier utility
theorists who were forever indulging in long discussions of the
relative intensity of the needs of a typical person for various
amounts of drink, food, clothing, shelter and the like.

Had they been willing to confine themselves to this limited
sphere no criticism could be made of their work. They were,
however, looking for completely “general” and absolutely
“exact” laws and as a consequence made the fatal mistake
of so broadening the concept of need as to take in every
motive or impulse to action of whatever sort. Any act of
choice on the consumer’s part came to be regarded as based on
an estimation of the relative intensity of competing needs,
the more intense need always winning out over the less
intense. The proposed explanation of choices on the basis of
a weighing of actually distinguishable needs — both the
needs and the weighing being regarded as independent of the
ensuing choices — thus ended up in a mere duplication of
terms.

This misconception has been particularly unfortunate since
it has greatly impeded progress in the realistic study of needs
and need-complexes. Such positive results as have been
achieved, as, for instance, in the analysis of complementarity,
periodicity, and divisibility,” have been incidental to the

6. Cf. Kaufmann, op. cit., pp. 265-266: “Of two needs, that one is
frequently characterized as the weaker or less important whose satis-
faction is actually given up. In this way it appears that the choice
acts are explained by psychological determinants, while in reality a
new label for the decision expressed in the choice act has simply been
introduced by way of definition. This misunderstanding is especially
dangerous here, since in certain subsections of the territory under
investigation such scales of intensity actually can be established. This
is particularly true of such ‘primitive’ needs as those for food, drink,
and clothing. Since, furthermore, the typical conditions of intensifica-
tion or weakening of these needs are known, intensity-patterns (Dring-

lichkeitstypen) can be constructed, which may be useful for the explana-
tion of economic behavior.”

7. In this respect Menger deserves high rank among utility theorists:
It is also worthy of note that most of the subsequent work of this sort
has been done by his successors in Vienna, notably Professor Hans
Mayer and Dr. P. N. Rosenstein-Rodan.


===== PAGE 10 =====

728 QUARTERLY JOURNAL OF ECONOMICS

illusory search for an absolutely general and universally valid
law of human bchavior.?

With the discussion of needs and their place in subjective
value theory we have covered the last three of the four major
assumptions involved in the subjective value approach.There
remain a few comments to be made on the first; that is, the
assumption that the explanation of prices should start from
the individual consumer.

The shift in emphasis from producer to consumer in the
writings of the earlier marginal utility theorists was all to the
good so far as it served merely to correct the classical econo-
mist’s one-sided neglect of demand and to provide economists
with a method of analyzing the factors that lie behind market
behavior on the consumers’ side. Other consequences of the
shift in point of view have not been so beneficial. For one
thing, it came, ironically enough, just at the turning of the
high tide of free competition and the beginning of the modern
development of monopolistic forms of market control. The
new approach was thus from its inception a generation or
more too late, the individual consumer obviously being a
more appropriate focal point for the study of a freely com-
petitive economy than for that of a partially trustified one.

Emphasis on the consumer also led easily, if not quite
legitimately, to the construction of a picture of the whole of
economic activity as directed toward the satisfaction of con-
sumers’ wants. Here again it is scarcely necessary to argue
that this picture has been steadily losing whatever justifica-
tion it may once have had. And finally the centering of

8. Offhand it seems somewhat paradoxical that the most serious
attempts to make concrete use of consumer needs in economic analysis
have come from writers like Loeb, Chart of Plenty, and Ezekiel,
$2,500 a Year, whose technocratic or institutional background would
lead one to expect a complete repudiation on their part of anything so
closely resembling a ‘““utility”’ approach. Most avowed disciples of the
marginal utility school, on the other hand, would roundly condemn the
procedure followed in the studies just mentioned. All of which leads
one to suspect that the bitter controversies between utility theorists and
institutionalists really had very little to do with the apparent question

at issue; namely, whether or not needs and desires might profitably be
utilized in the attempt to explain economic phenomena.


===== PAGE 11 =====

COLLECTED WORKS OF CARL MENGELR 729

attention on the individual consumer? has led to a serious
neglect of the various types of collective consumption, that
is, consumption in the planning of which the individual par-
ticipates as a “voter” and not directly as a buyer.!

IV

Menger’s contribution to the theory of value was unques-
tionably pioneer work of the highest order. With one brilliant
stroke he cleared away the perplexity and confusion that had
hitherto prevailed as to the role of subjective estimations of
utility in the determination of value. One has to read over
again some of the passages in which the older economists
struggled with the apparent paradox of utility to appreciate
fully the quality of Menger’s achievement in this respect.
The development of economic analysis was furthered in two
important ways. In the first place, many sterile controversies
and pseudo-problems were forever laid to rest, and in the
second, a new and potentially fruitful line of inquiry was
opened up.

But Menger and his disciples overreached themselves in
attempting to universalize the significance of their discovery.
Instead of being content to view the new subjective analysis
as a corrective, on the one hand, for the one-sidedness of the
older cost of production theories and as suggestive, on the
other, of the way consumers of various types would act under

9. It is obvious, of course, that individual purchase and sale is neither
the only, nor necessarily the best way — best, that is, from their own
individual points of view — consumers can seek to satisfy their wants.
Many highly important types of consumption are collective by their
very nature and require the exercise of a collective decision as to whether
they shall be enjoyed or not. Other types could be left to individual
decisions — as for instance, the choice of architectural designs for city
houses — but are of such nature that every one will be better pleased
with the final result if a measure of collective restraint is placed on the
individual decisions. The same applies to various forms of competitive
consumption: every one concerned would frequently be happier if a
collective agreement ‘to call it off’’ could be reached. These considera-
tions, while commonplace enough, have usually been overlooked in dis-
cussions of “maximum consumer satisfaction.”

1. Of course none of these defects is inherent in the marginal utility

principle itself. They are rather a consequence of the lack of discrimina-
tion with which it has been used.


===== PAGE 12 =====

730 QUARTERLY JOURNAL OF ECONOMICS

various circumstances, they professed to have discovered a
universally valid law of human behavior. This led, as already
pointed out, to a broadening of the concepts of need and want
until they lost all their original common-sense content. It
also led to Menger’s misguided excursion into the field of
scientific methodology: an excursion that was bound to end
disastrously because of the logically contradictory nature of
the goal it sought to reach.

In the recent revival of interest in Menger and the other
earlier Austrian economists, attention has been concentrated
on the supposed universal validity of the ‘laws’ of subjec-
tive value theory. The chief fruits have been more bad
methodology, combined with a neglect of much that was
sound and potentially useful in Menger’s treatment of the
subject. His fresh and vigorous, if somewhat naive, interest
in concrete needs? and their interrelations has been deemed
unworthy of the dignity and authority of economic science by
his modern disciples. For Menger himself, however, the
interest was a profound and continuing one. It is thus quite
legitimate to consider present and future experimenters with
consumption data — whether obtained from budget studies,
the psychological or physiological laboratory, or observation
of the experimenter’s self and his acquaintances — as carry-
ing on a line of inquiry in which Menger was one of the
earliest pioneers.

Aran R. SWEEzZY

HarvaArD UNIVERSITY

2. His effort to link utility theory to biology and physiology should
be recalled.