Is Economic Value Still a Problem?

Author(s): ADOLPH LOWE

Source: Social Research , WINTER 1981, Vol. 48, No. 4, Politics: The Work of Hans
Morgenthau (WINTER 1981), pp. 786-815
Published by: The Johns Hopkins University Press

Stable URL: https://www.jstor.org/stable/40970847

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/40970847?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and
extend access to Social Research

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 Is Economic Value /
Still a /
Problem?* / ADOLPH LOWE

A he formulation of the theme of this paper is inten

ambiguous. It may indicate that the question it rais
become redundant because there exists a generally a
solution. This was certainly so during the era of class
nomics, and again from the Marginal Revolution until t
of the Second World War. Since then, however, the rev
interest in Ricardian and Marxian economics and, more sub-
tly, in the microfoundations of macroeconomics have dis-
turbed this consensus - so far without any traces of reconcilia-
tion among the contestants.
But my question may be interpreted quite differently. Con-
sider the statement with which Paul Samuelson concluded his
survey of utility analysis. In trying to show that the subjective
theory of value "is not in a technical sense meaningless," he
established two conditions - the nature of which need not con-
cern us here - which would raise the theory above a mere
tautology. Still, at the end he wondered "how much economic
theory would be changed if either of the two conditions . . .
were found to be empirically untrue. I suspect, very little."1 In
other words, solution or no solution, the entire problem is
seen there as being of little relevance for the major issues in
economics.

During my own academic training early in the century, I

was fed ad nauseam with the controversies between classical

1 Paul A. Samuelson, Foundations of Economic Analysis (Cambridge: Harvard Univer-

sity Press, 1955), p. 92, 117.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 787

and neoclassical value theory. But

shared Samuelson's suspicion. Ju
Walras, and Marshall assembled in
fronted with some issue of econom
of a general rise in wages, the intr
imposition of a sales tax. I am co
most instances arrive at the same
might disagree - for example, a
interest or of flexible factor pr
other than differences in value t
In contrast with this skeptical
cently taken a radically different
burning issues of our day - inflat
tion explosion, destruction of th
tion a few - can be truly understo
Marxian theory of value and capita
this exaltation of Marxian theory
relative inputs of labor quantities
rectly, the ratio of the prices of p
late of Marx's value theory.
I do not in principle reject the
to Marxian theory, for which he w
standing" in place of "explaining
called by Schumpeter the "pre-anal
phenomenon - as a well-ordered
as a time-independent or an evolu
of shared benefits or of exploita
that a theory of value can be press
vision.

However, my interest in raising the question "whether eco-

nomic value is still a problem" is much more limited. It is the
same that dominated the minds of the classical and neoclas-

sical writers when they inquired whether a theory of value can

2 Edward J. Nell, "Value and Capital in Marxian Economics," Public Interest, Special
Issue (1980): 174-200, at pp. 194ff.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 788 SOCIAL RESEARCH

serve as the basis for the explana

cal prices. In other words, I sh
with the inner consistency and th
contestants - the labor theory an
I do so for two reasons. On the one hand, if one or even
both these theories should prove defective under the aspect of
consistency and/or explanatory power, my purpose would be
served without entering into the much more intricate problem
of the adequacy of the underlying "vision." On the other
hand, I believe that the gradual purging of utility theory of
unessential elements and the reinterpretation of the Marxian
enterprise - for example, in the work of Morishima3 - open
the way for a rapprochement of what have hitherto been
regarded as irreconcilable positions.

Defining the Problem

In starting but on our undertaking, the first difficulty arises

from the danger of a confusion of terms. To give a represen-
tative example, let us consider the following statements in
Schumpeter's History of Economic Analysis. The headline with
which he introduces marginal utility reads as follows: "The
Revolution in the Theory of Value and Distribution." To this a
footnote is attached stating: "Instead of using this traditional
phrase, I might have used exchange ratios or relative prices."4
Now if value and relative prices are one and the same, the
former can hardly explain the latter. But two pages later
Schumpeter correctly states that all of the early marginalists

aimed at the same goal, which was to prove that the principle of
marginal utility suffices to deduce [my italics] the exchange ratios

3 "I believe that Marx would have accepted the marginal utility theory of con-
sumer's demand if it had become known to him" (Michio Morishima, Marx's Economics
[Cambridge: Cambridge University Press, 1973], p. 40).
4 Joseph A. Schumpeter, History of Economic Analysis (New York: Oxford University
Press, 1954), p. 909.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 789

between commodities that will estab

tive markets. ... In other words, the
Ricardo and Marx had believed to
exchange value can be explained in

In the same context Schumpeter refers also to the rival

"theorem that 'prices' of commodities tend to be proportional
to the labor quantities embodied in them," indicating that for
both schools exchange value results from a more basic
determinant - use value in the form of marginal utility, or
labor quantity.
It is on this basis that we shall approach the problem by
asking whether there is indeed a factor that can serve as an
explanation of the ratios of empirical prices (and possibly of
the quantities transacted). Or, as the same idea has been for-
mulated more pointedly, suppose that a universal amnesia were to
wipe out the knowledge of all present prices, would there be a rule for
reestablishing them? In other words, what cannot be accepted as
valid is an "historical" solution of the price problem, that is, an
explicit or implicit derivation of present prices from past ones.
We shall see that this exclusion is of great significance in our
further deliberations.

It is trivial to restate that the two conflicting "causes" of

relative prices - labor input and utility - refer to experiences
occurring in production or consumption, or that they stand
for a technological or a behavioral factor, respectively.5
Perhaps it is less trivial to stress from the outset that the factor
which dominates in classical economics is a measure of the
resistance with which those who furnish the means of provision
have to contend, whereas the neoclassical factor is supposed to
measure or otherwise to express the welfare that accrues to the
beneficiaries of provision. An analogy would be the difference
between ohm and volt in electrophysics - an indication that the

5 Applying the term "cause" to utility is, of course, correct only for Jevons or
Menger and their followers, in contrast to the "functional" school inaugurated by
Walras. For our purposes this difference will prove irrelevant.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
790 SOCIAL RESEARCH

two approaches may deal with qu

that reason not be a priori incom

Ina systematic treatment we wou

history of the problem from it
going back to Aristotle and the
limited purpose we can confine o
few salient aspects of the class
proaches. We shall begin with the
that a return to Ricardo or Ma
modern theory cannot offer an
nately, we can dispense with a det
utility theory itself. The featur
common both to the early "psyc
among the marginalists and to the
or Debreu. We shall also skirt suc
to whether the utility experienc
rectly ascertained or must be inf
whether utility is measurable in
terms, and other such issues. Wh
beyond controversy within the ran
the very foundation on which th

The Utility Theory of Value

Some roots of the utility theory

certain writings of the classical age
and Cournot. Still, the theory aros
the classical labor theory of valu
range of explanatory power. If a
theory seemed pertinent only to

6 For a brief sketch of this doctrinal developm

Economic Value," in Human Values and Econom
York: New York University Press, 1967), pp

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 791

period equilibrium - prices, and ev

unrealistic assumption of equal ca
the system. Fluctuating market
ruling under natural and contract
to a law of supply and demand w
were left largely unexplained.
Thus it appeared as a big step t
simplification of analysis when t
able to explain all the variants of
principle - the universal aim of con
Whether understood as a motivat
tern, this maximization was tak
choice among the possible objects
about only by the operation of
referring to an inverse relationship
quantities at hand, and to the equa
when several goods were to be a
It has never been quite clear whet
to be facts of experience or post
What is clear is that the claim to u
can be maintained only if the mar
ers determine not only the relati
but also those of the factors eng
show this is not difficult for the s
themselves. A worker's idea of wha
can be related to the utility he e
consumer goods he is subsequentl
he experiences in expending effort
is to be consulted by the prospect
the hourly wage he is to offer?
This is the famous problem of
the members of the Austrian sch
headache. The problem will occup
marginal productivity theory. In t
portant to realize that, in all con
doctrine, the utility principle is

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 792 SOCIAL RESEARCH

behavior of householders, where

profit (not utility) maximization
action directive of firms when se
productive services.7 Accepting t
principle at least suffice to determ
ior of householders?

Limiting Conditions

One crucial test is the derivation of an unambiguous law of

demand. Let us consider the response of the buyer of a con-
sumer good to a fall in price. If his wish to maximize utility is all
we know about his aim, we can say no more than that he will
raise the quantity demanded or reduce it or leave it as it is. If
he raises it, he will indeed respond in accord with the con-
ventional law of demand, according to which a change in price
evokes an inverse change in the quantity demanded. But there
is no doubt that the other modes of behavior, though con-
flicting with the conventional law, can be interpreted equally
well as "utility-maximizing." If demand falls as a response to a
price fall, we may deal with the Veblenesque case of ostenta-
tious expenditure, in which expensive goods are valued more
highly than cheaper ones. Or the good in question may be
regarded by the buyer as what Hicks calls an "inferior" good
on which he spends less when its price falls, to devote the gain
in purchasing power to increasing purchases of more highly
valued goods. Finally, he may read into the price fall the
beginning of a trend indicating a further price fall to follow,
so that postponement of purchases to a future date promises
still greater satisfaction. And we can interpret such abstention

7 It seems unnecessary to quote examples for this proposition because maximization

of utility, together with maximization of profits, is the general standard today for
market behavior in neoclassical theory, although I am not aware of any explicit
justification why half of the bargainers should be excluded from the utility rule.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 793

from still other considerations

for the good concerned or, mo
in the past may yield him max
consumer's homeostasis.

In the face of such plain objections, if the protagonists of

utility theory still maintain their claim of being able to predict
consumer behavior in accord with the conventional law, one
suspects that there must be somewhere a limiting assumption
in what looks on the surface like a mere tautology. And in-
deed, closer examination reveals no less than three special
conditions that are implied in the conventional proposition.
The first concerns the nature of "things," the utility of which
is to be maximized, namely, priced goods and services, rather than
intangibles such as Veblenesque "ostentation." Second, and
more important, any consumer who tries to maximize utility is
supposed to be in a state of under satiation - a state that implies
not only the painful experience of a discrepancy between
means and ends, but also the desire to attain fuller satisfaction.
Third, the time span over which the consumer is supposed to
maximize utility is limited to the immediate future, thus
eliminating the disturbing effect of other than static expecta-
tions. In other words, the utility principle is an operational
concept only in a world of under satiation* and instantaneous pro-
vision.

Those limiting conditions indeed eliminate the exceptions

from the conventional law stated above, exceptions which can
all be interpreted as specific states of satiation. But, we must
ask, do those exclusions significantly reduce the empirical rel-
evance of the utility principle as an explanation of the behav-
ior of householders? Is not undersatiation - the psychological
reflection of scarcity - man's existential fate? Even if for a
fleeting moment he can fully satisfy one or another want, is
not the totality of his wants without limit, so that any stock of

8 The undersatiation condition has been formalized in the convexity postulate of

indifference analysis.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 794 SOCIAL RESEARCH

resources must always appear t

conventional law of demand the iron rule for his behavior?
Though it is a cornerstone of modern analysis, the assump-
tion of a transhistorical absolute scarcity of resources is a
fiction. Such scarcity prevails only in two historical situations.
One is a state of destitution in which the available stock of
resources does not assure physical survival. At the other end
of the scale is a state in which even the largest conceivable
stock of resources cannot satisfy all wants because consumers
are insatiable. However, far from reflecting unalterable natural
constraints or intrinsic human propensities, destitution and
unlimited appetition are historical phenomena, related to par-
ticular stages in the technological and cultural evolution of
mankind. Stated differently, above the level of destitution
scarcity of resources becomes a relative notion - relative to the
scale of wants that the ruling cultural system sanctions.
Very likely, the natural and social pressures of early
capitalism - mass poverty, unbridled competition, and Puritan
work ethic - created a climate in which adherence to the con-
ventional law of demand was an imperative for survival.9 But
the affluent marketers of present-day organized capitalism
present a different picture. They seem to move somewhere
between the two extremes of absolute scarcity, exhibiting
growing tendencies toward homeostasis rather than maximi-
zation. Maintenance of asset values or of market shares are
examples on the supply side; insensitivity to price fluctuations
over time or to price differentials for physically homogeneous
goods in favor of routinized purchases of branded com-
modities are examples on the demand side.10

9 It is those conditions of the environment which also explain profit maximization as

the main action directive of firms. Moreover, even for householders they resolve the
apparent clash between maximization of utility and the classical action directive of
maximization of receipts and minimization of expenditure. The former - reflecting the
"dynamics of undersatiation" - is a psychological substructure of the latter.
10 The first to draw attention to the significance of homeostatic tendencies m
modern markets was Kenneth E. Boulding in A Reconstruction of Economics (New York:
Wiley, 1950), ch. 2.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 795

Thus at least one of the excep

of demand - marketers' homeostasis - seems to be far from
"rare and unimportant," as Hicks rightly considers the case of
inferior goods.11 Even more damaging is the assumption of
"timelessness." It confines the validity of the utility approach
to the Marshallian very-short period, restricting it to a static
framework. It thus excludes all dynamic phenomena, contrary
to the claim on which the marginalists base their superiority
over their classical predecessors.

Deriving Prices from Utility

So far we have still been dwelling in the forecourt of the

problem, studying the effects of given prices and price changes
on consumers' behavior. But the central problem is, of course,
the derivation from utility considerations of the prices them-
selves. For this we shall take our bearings from the most
sophisticated construct - Walras's theory of the equilibrium of
production, refined by Hicks's application of indifference
analysis. We can do so without biasing our verdict because the
two objections we are going to raise apply to all neoclassical
constructs, causal and functional, that try to derive equilib-
rium price ratios from utility arguments.
To put it in the simplest terms, the theory asserts that, given
the marginal rates of substitution of all householders for all
commodities, also the quantities of such commodities pos-
sessed at the outset by each - what is now often called their
initial "endowment" - together with the production functions
and factor supply schedules, we can derive prices and quan-
tities of commodities and factors representing the final market
equilibrium.
We shall not raise such much-debated questions as whether
the final equilibrium thus attained is "unique," or whether

11 John R. Hicks, Value and Capital (Oxford: Clarendon Press, 1939), ch. 1, sec. 6.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
796 SOCIAL RESEARCH

equality of unknowns and of eq

lationship between knowns and u
tence" of such an equilibrium. Ou
assumed independence of certain
knowledge of certain prices and (
implied behavioral rule.
As far as the first point is conc
are the initial endowments of each
on the one hand, on the marginal r
the other hand, on the factor su
role of those endowments has be
but the crucial objection has been m
defendants, as shown by the follo
of these initial endowments as ex
and much misunderstood, in partic
instance, it is often argued that th
income and wealth exogenous."12
analysis the objection is the very co
those endowments are not exoge
serve as data or independent vari
ments and their apportioning amon
result of the process of distribut
The misunderstanding seems to
tinguishing between personal an
Personal distribution, the manner
come and wealth - labor and proper
the householders, is mainly the r
and legal order. As such, it is in
and thus a legitimate datum of p
functional distribution, that is,
components of the householders'
profit, interest, and rent. This w
relative factor prices of the preceding

12 Frank Hahn, "General Equilibrium Theo

(1980): 123-138, at p. 124, n. 1; my italics.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 797

the then ruling commodity prices

regress. Since the marginal rates
schedules of factor supply, vary w
they partake in this "historical
themselves. And in order to esta
indeed have to "start with Neande
the era when a general price sys
Moreover, there is another "historical" limitation of the
Walrasian model of quite a different nature. It concerns its
general frame of reference - competitive equilibrium. This
confines what validity it possesses to the competitive era of
capitalism, excluding from its explanatory range the "fixprice"
structure of contemporary organized capitalism. The writings
of Kalecki, Hicks, and Morishima have explored the "abnor-
mal" consequences of fixprices for the operation of market
systems. But I know of no attempt to relate fixprices and, in
particular, the markups in question to any value theory.
We turn now to our second objection, namely, to ascribing
to Walras unreservedly the conventional interpretation of the
rule of market behavior. The problem concerns the path over
which final equilibrium is to be established or reestablished
when it has been disturbed. Again, we disregard the well-
known objections to Walras's notion of "crying" prices or to
Edgeworth's idea of "recontracting." What we are concerned
with are the "tâtonnements" which are supposed to reduce any
gap between potential purchases and sales to the point where
the market is cleared.
From our earlier discussion we know that this path is fully
determined and leads in the right direction only under the
restrictive assumption of timeless undersatiation. For this rea-

13 Ibid., pp. 124, 127. The gist of the above argument was presented half a century
ago in Hans Mayer, "Der Erkenntniswert der Funktionellen Preistheorien," in Die
Wirtschaftstheorie der Gegenwart, 4 vols (Wien: J. Springer, 1927-32), 2: 147-239, esp p.
196. However, Mayer overlooked the fact that his objection to an indefinite regress
was equally valid for the "causal" theories which Mayer favored over and above the
functional ones.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
798 SOCIAL RESEARCH

son it is significant that Walra

passages the "positive" validity
formalization of actual behavior, offers in addition another
and quite startling interpretation. In putting forth what he
calls the "law for the establishment of equilibrium prices," he
has this to say:

Given several commodities, which are exchanged for one an-

other through the medium of a numéraire, for the market to be
in a state of equilibrium or for the price of each and every
commodity in terms of the numéraire to be stationary, it is
necessary and sufficient that at these prices the effective demand
for each commodity equal its effective offer. When this equality
is absent, the attainment of equilibrium prices requires a rise in
the prices of those commodities the effective demand for which
is greater than the effective offer, and a fall in the prices of
those commodities the effective offer of which is greater than
the effective demand.14

The importance of this passage lies in the fact that here

Walras no longer claims that the forces of competition and the
utility considerations that impel them will, in fact, cause the
system to approach general equilibrium. Rather, he establishes
certain behavioral requirements - requirements that are seen as
prerequisites for maintaining equilibrium once it exists, and
for establishing it when it is absent. In other words, a norm of
behavior is postulated as a means for a specified end -
macroequilibrium. Why such an equilibrium should be chosen
as the end is not explicitly stated. In now examing more
closely its nature, we shall bring to light a possible reason for
such choice.

The Theorem of Marginal Productivity

An easy access is through the theorem of marginal produc-

tivity. As was indicated earlier, the theorem was originally
14 Léon Walras, Elements of Pure Economics, translated by William Jaffé (New York:
A. M. Kelley, 1954), p. 172; my italics. An equivalent statement for the case of
production appears on pp. 253-254.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 799

devised as a tool of imputation in

is, as a way of splitting the atom
the component elements of factor
not seem to be of much interest i
is to be decomposed - consumer
explained only under the extrem
set out above. However, once we draw the ultimate conse-
quences from the theorem of marginal productivity -
consequences that are usually not drawn - we shall understand
those extreme conditions for what they really are.
What is at stake can be most easily explained by using J. B.
Clark's analytical tool chest. According to him, a productive
factor will be employed to the point at which the addition to
output of the last unit equals the cost of that unit. It follows
that under perfect competition each factor receives as its re-
muneration the equivalent of its marginal product. In con-
centrating on the two basic factors, labor and real capital,
Clark establishes a stationary state with positive returns to both
factors by placing production in the range of decreasing
physical returns.
There is nothing wrong with the logic of this argument if
one implicit assumption is accepted, namely, that returns of
capital are spent on consumer goods. But is this a realistic
assumption in a competitive market? Are we not told that,
whatever rules consumers' behavior, firms aim at the maximi-
zation of profits - in the given case, of returns of capital? So
we must assume that the returns of capital received in what
now appears only as a pseudostationary state will be saved and
invested. We are reciting here what nowadays is scoffed at as a
neoclassical "parable." It asserts that, under the stated condi-
tions, a process of accumulation will be initiated that leads to
steadily rising capital-labor ratios associated with rising per
capita output, rising real wages, and falling returns to capital.
Where this process will end has been aptly described by Joan
Robinson, as follows:

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 800 SOCIAL RESEARCH

In the special case where the total labo

... if accumulation is going on, a scarc
later emerge. A rise in the real wage r
lation draws labor out of the investment sector. The rate of
accumulation falls, but if it is still positive the scarcity of labor
will sooner or later emerge again; a further rise in wages wil
further reduce the rate of accumulation, and so on, until re-
placement absorbs the whole of gross investment, and the stock
of capital ceases to increase. All labor is then employed on
producing consumer goods and maintaining capital, wages
absorb the whole net product of industry, and the rate of profit
is zero. This is ... properly described ... as the state oí economic
bliss, since consumption is now at the maximum level which can
be permanently maintained in the given technical conditions.15

In passing, we must remember that the above proposition

must defend itself against objections that have recently been
raised under the heading of "switches of technique" and "cap-
ital reversing."16 However, within the limits of a constant sup-
ply of labor - one of the assumptions underlying both Joan
Robinson's analysis and Clark's pseudostationary state - the
steady movement toward economic bliss may be temporarily
interrupted but not finally deflected from the state of bliss, so
long as accumulation of the returns to capital continues. Tem-
porary blocks to progressive "mechanization" may arise
whenever a more labor-intensive technique promises higher
returns. But any such change in technique is bound to in-
crease competition for the constant stock of labor and thus to
raise wages, ultimately to the point at which more capital-
intensive techniques again yield higher profits.
A more serious objection to the realism of the construction
concerns the doubt that, considering the steady fall of returns

15 Joan Robinson, The Accumulation of Capital (London: Macmillan, 1956), pp.

81-83. The same reasoning underlies Schumpeter's notion of "circular flow equilib-
rium" and Keynes's concept of a "quasi-stationary community." See Joseph A.
Schumpeter, The Theory of Economic Development (Cambridge: Harvard University
Press, 1934), chs. 4-5; John Maynard Keynes, The General Theory of Employment, Interest
and Money (London: Macmillan, 1936), ch. 16.
16 See G. C. Harcourt, Some Cambridge Controversies in the Theory of Capital (Cam-
bridge: Harvard University Press, 1972).

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 801

to capital, accumulation will continue beyond a certain

point - Marx's notion of the "blunting of profits" as a block to
further accumulation. In acknowledging this objection, we
cannot claim for the state of bliss a positive role as describing
an empirical phenomenon - it is a norm postulating the condi-
tions for maximizing real net output.
In retrospect, the Walrasian model too can be fitted into this
interpretation. We remember that its a- theoretical, "historical"
nature is due to the effect which functional distribution has on
the relative size of the individual endowments. If, as in the
foregoing analysis, the number of priced factors is reduced to
one, functional distribution will be eliminated and the model
will be logically self-contained - though in a form that has no
realistic counterpart and, like the Clarkian model, can serve
only as a norm.

Quantities and Qualities

We have arrived at an altogether disappointing conclusion.

Even within the limited range of competitive capitalism, em-
pirically valid price ratios cannot be derived from utility
theory through procedures of self-contained pure theory. We
are forced to accept the order of relative prices of the period
preceding the one under examination as a datum of
analysis - we must implant utility theory in an historical con-
text to achieve testable results. And even within an historical

context the conclusions of the theory apply only to a world of

undersatiation and even this only over the very short
period - thus excluding the empirically important cases of
homeostasis and of nonstatic expectations.
Fortunately, this is not the end of the story. In his famous
polemic with Jevons, Marshall opposed to the "catena" estab-
lished by Jevons, namely, cost of production determines
supply - supply determines final degree of utility - final de-
gree of utility determines value, a catena of his own. It reads:

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
802 SOCIAL RESEARCH

utility determines the amount

supplied - the amount that has
cost of production - cost of pr
The argument is set forth in d
hardly any role in Marshall's s
theory. But it is a useful remind
ing of the market mechanism,
what price" but also "what and ho
It is as an answer to that questio
its versions plays its part. Of cou
modities demanded depends on
lished factor prices enter here t
difference maps quite explicitly
tities demanded at given prices
been reduced to a secondary role
the price problem can be solved i
role that has never been claimed
of value. At the same time it i
solution of the price problem an
full understanding of market st

Labor Input and Price

If the foregoing argument is acc

approach does not offer a satisf
distinct from historical, answer
mately determines relative pric
examination of the classical app
confronted by a long history ext
Mill and culminating in Marx.

17 See Alfred Marshall, Principles of Econo

pp. 817, 819.
18 An all-inclusive study of the problem
contributions of Sraffa and von Neumann. B
to the extent to which it touches on the iss

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 803

writers was on "costs of product

several factors of production, at l
a commodity," in Mill's words, "can be indefinitely in-
creased."19 Still, though labor input always played an impor-
tant role, it was raised to the dominant "cause" of pricing only
by Ricardo and Marx. Both reduced the operative factors of
production to two - labor and real capital - and tried to show
that the effect of the latter on pricing can be expressed in
terms of the former. It was Marx who, by proclaiming "socially
necessary labor time" as the ultimate determinant, seemed to
give the clearest expression to a price-independent variable as
the source and measure of exchange value.
In what follows we shall concentrate on the Marxian version
of the theory. The reason is that the aims which such a theory
pursues, as well as the difficulties it encounters, can be exam-
ined there most thoroughly. Perhaps it is in order to stress at
this point that, whatever else Marx intended to demonstrate in
discussing value, no careful reader of Capital, vol. 3, ch. 9,
can doubt that one of his major concerns was the relationship
of the "prices of production" - his term for the relative prices
prevailing in equilibrium - to the original quantities of labor
input.
Before saying more about this, it should be emphasized that
a price theory based on the input of socially necessary labor
time may, as we shall see, run into serious difficulties. But at
first sight it seems to escape the vicious circle of indefinite
regress that bedevils all neoclassical price theories when they
refer to the real world. Therefore, Samuelson's advice to "rub
out ... as an unnecessary detour . . . Volume I's analysis of
values" may annihilate more than he envisages.20 Rather,

19 Where natural or artificial monoplies prevail, price was supposed to be deter-

mined by the interplay of "effectual" demand (Mill) and supply. Effectual demand
was seen to depend on "the wish to possess," as well as on "the power of purchasing."
Thus the argument is involved in the same historical chain that creates the indefinite
regress in utility theory.
£V Paul A. Samuelson, "Understanding the Marxian Notion of Exploitation," Journal
of Economic Literature 9 (June 1971): 399^431, at p. 421.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
804 SOCIAL RESEARCH

Marx's analysis, resting on an "

cal, and thus price-independent
whose status as a datum no log

The Transformation Problem

Now to begin with, there is no

exceptional and unrealistic assu
ratios throughout the system wi
each enterprise coincide with t
ratios of labor inputs be identica
of production. Strangely eno
Boehm-Bawerk, critics treated
refutation of the theory. This w
connection could be established between' those deviations and

the quantities of embodied labor - the essence of the so-called

transformation problem.
In a brilliant paper, Meek has established such a connection
by relating Marx's procedure to Sraffa's notion of a "standar
industry."

Sraffa is postulating precisely the same relation between the

average rate of profit and the conditions of production in his "stan-
dard" industry as Marx was postulating between the average
rate of profit and the conditions of production in his industry of
"average organic composition of capital. "

With one important correction introduced by Sraffa allowing

for the effect that a change in wages has on the prices of the
means of production, both Marx and Sraffa conclude that

the average rate of profit, and therefore the deviations of price

ratios from embodied labor ratios, are governed by the ratio of
direct to indirect labor in the industry whose conditions of
production represent a sort of "average" of those prevailing
over the economy as a whole.

Thus "the very deviations of equilibrium price ratios from

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 805

embodied labor ratios are them

tities of embodied labor.'"21*22
But we must now ask: What is the role, if any, oí utility or use
value in the classical conception of price determination? Friend
and foe seem to agree that Ricardo expressed the classical
consensus when he stated on the very first page of the Princi-
ples: "Utility ... is not the measure of exchangeable value,
although it is absolutely essential to it. If a commodity were in
no way useful ... it would be destitute of exchangeable value,
however scarce it migh be, or whatever quantity of labor might
be necessary to procure it."23
Such exclusion of utility and thus also of demand from price
determination is easily defended when we deal with industries
operating at constant costs. It is no longer tenable when we are
confronted with rising-cost industries. Even then it can be said
that their price ratios coincide with the ratios of the labor
input required at the margin of output. Therefore labor input
can still serve as a standard of measurement of prices. But it is the
quantities and qualities demanded that determine where on the

21 R. L. Meek, "Mr. Sraffa's Rehabilitation of Classical Economics," in Economics and

Ideology (London: Chapman & Hall, 1967), pp. 161-178, esp. pp. 175-177; his italics.
22 It may be worth noting that even Schumpeter, in the face of other grave objec-
tions to Marx's approach, thought it "possible to hold that the relative prices of
commodities, as deduced in the third volume, follow from the labor-quantity theory
of the first volume" (Joseph A. Schumpeter, Capitalism, Socialism, and Democracy [New
York: Harper, 1947], p. 29, n. 9).
Very different is the case of von Neumann. First of all, his "Model of General
Economic Equilibrium" (Review of Economic Studies 13 [1945-46]: 1-9) is concerned
with dynamic equilibrium, more precisely, with balanced growth. It is to this frame-
work that his major innovations apply - a new treatment of fixed capital and its
depreciation, and his criterion for the choice of different techniques - not to the
stationary framework in which Marx discusses the transformation problem. In the
latter context, the assumption of a stationary age structure of fixed capital - in any
period the quantity of fixed capital produced equals the quantity worn out and thus to
be replaced - is not only permissible but indispensable if equilibrium is to persist.
When transferred to such a stationary framework, von Neumann's "inequalities"
change back into Marxian equalities.
23 For Marx's formulation of the same idea, see Capital, vol. 1, ch.l, sec. 1. Consider-
ing the many editions that are in use, my quotes refer to chapters and sections rather
than to the pages of a particular edition of Capital.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
806 SOCIAL RESEARCH

production scale the margin of output lies. They thus

codetermine the quantity of marginal labor input, denying the
latter the role of the sole cause. In view of the strategic role of
natural resources in the process of production, most of which
are subject to increasing costs, this modification must not be
minimized.24
On the other hand, the labor theory of value is not limited
by the "undersatiation" postulate as is the nontautological util-
ity theory. Even in a homeostatic system, all priced
commodities - that is, all commodities the provision of which
depends on labor input - exchange in equilibrium at ratios
determined by the latter.

The Nonhomogeneity of Labor

So far the labor theory of value seems to come out with

flying colors as a perfect tool of measurement of stationary
price ratios and as greatly superior to utility theory in the
determination of empirical price ratios - at least to the extent
to which equilibrium can be regarded as an approximation to
the real world.

Alas, this favorable result will be challenged when we now

turn to an examination of what we mean by labor input in
view of the great disparity of actual performances. In other
words, before resting content we must deal with what critics
have defined as the "nonhomogeneity" of labor. In fact, we
are confronted with no less than three different types, which
we shall label types of "form," of "grade," and of "skill." Only
if it is possible to reduce all three types to a common denomi-
nator does the labor theory of value become operational.

24 It is noteworthy that, contrary to the categorical denial of any influence of use

value and thus of demand on price, as expressed in Capital, vol. 1, Marx accepts this
modification in vol. 3, ch. 10, when he writes: "If demand is so strong that it does not
contract when price is regulated by the value of the commodities produced under the
most unfavorable conditions, it is these conditions that determine the market value."

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 807

This is not difficult in cases w

this we refer to labor input of ot
in different occupations, such a sh
we can accept Ricardo's answer th
forms will be achieved through
market, or Marx's equivalent re
that goes on behind the backs o
However, as has often been po
not permissible in dealing with
Speaking first of grades, by whic
handy as compared to average ca
comparable to differences in fert
of natural resources. And the solution lies where Ricardo

found it in the latter instance. Because they are able to

duce more output per unit of labor time, the higher gr
hold a monopoly position that will yield them a "rent" o
and above the wages of the marginal workers. However,
the Ricardian case, this rent determines neither the costs of
production nor consequently the prices of output. As in the
case of increasing costs discussed earlier, those prices and thus
the rents of skill are determined by the input of marginal - in
a preferable formulation, of average grade - labor.
Turning now to different skills, that is, acquired capabilities,
Marx himself has given what looks like the beginning of an
answer:

In order to modify the human organism, so it m

and handiness in a given branch of industry,
labor-power of a special kind, a special educatio
requisite, and this, on its part, costs an equiv
modities of a greater or less amount. . . . The e
education . . . enter pro tanto into the real val
production.26

Alas, closer examination of this procedure

difficulties. Reducing differences in skill to d
25 Marx, Capital, vol. 1, ch. 2.
¿* Marx, Capital, vol. 1, ch. 6.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 808 SOCIAL RESEARCH

labor input in education and tr

relative inputs can themselves be
training time required. This, i
ferent educational labor times ca
a common denominator. In other
express one hour of skilled lab
labor performed with no skill.
In reality all skills are created b
educators and trainers, an inpu
the skilled labor of other educato
an indefinite regress. To reach
labor as the ultimate creator of skills, one must take into
account the entire evolution of civilization from its most

primitive stages to the present. Obviously there must h

been a gradual evolution of higher from lower skills. Th
however, would be another historical approach, formally
sembling what was said above in the context of a critique
utility theory.
Within the framework of a given market order there is
common denominator to which the hierarchy of labor sk
could be reduced. Rather, the wide dispersion of such sk
must be accepted as an ultimate datum. There may be no
objection to this as a practical expedient, but it prevents
from treating the relative prices of commodities produced b
different skills as if they represented multiples of some
mental unit. In fact, we have here a perfect analogy to an is
that has come up in the recent debate on capital theory -
irreducibility of different types of real capital to a comm
denominator.27

27 The same conclusion has been drawn by Edward J. Nell in "Understanding the
Marxian Notion of Exploitation," in George R. Feiwel, ed., Samuelson and Classical
Economics (Boston: Kluwer-Nijhoff, 1981). The problem has a long history dating back
more than 90 years to von Boehm-Bawerk's Karl Marx and the Close of His System,
where he sharply criticized the idea that skilled labor could be reduced to simple labor
without circular reasoning. On the other hand, an interesting literature in favor of
"reducibility" has recently sprung up in neo-Marxist circles, of which Yilmaz Akyuez,
in "Heterogeneous Labour and Labour Theory of Value," Memorandum from the
Institute of Economics, University of Oslo, 1976, offers a good survey. Morishima, in

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 809

However, there is another limita

of the labor theory of value, t
which extend far beyond the d
reducibility of labor skills. It is t
reduce the applicability of utility
ally consistent. Like the latter,
competitive markets and thus c
tions of organized capitalism. Nev
some meaning to it, one that pert
we must again interpret it as a n
performance of real markets c
This has been the very conclus
analyzing the utility approach. I
even further. Remembering our
ductivity, we saw that in that con
system coincides with the "idea
capital fall to zero because the
saturation" where each worker
stock of real capital applicable w
logical horizon.
Now Marx's model of "simple r
analysis of prices of production
"pseudoequilibrium," as we labeled the intermediate state
which conventional analysis of marginal productivity treats as
a state of rest. Though the source of the returns to capital
differ in Marx's scheme - being sociopolitical rather than
technological - these returns too are supposed to be consumed.
If we now postulate, as we did for the marginal productivity
model, that those returns should be saved and invested - a

Marx's Economics, pp. 190-193, takes an intermediate position. Though presenting a

"generalization of the labor theory of value" that allows for different kinds of labor,
he finds it in conflict with Marx's theory of exploitation.
I am grateful to my colleagues Robert L. Heilbroner and Anwar Shaikh for helping
me to understand the intricacies of these attempts. They all have one element in
common: they attribute from the outset to simple labor a significant role in creating
skills - a proposition which seems to me unrealistic at the level of practice and
question-begging at the level of theory.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 810 SOCIAL RESEARCH

postulate that is in full accord wit

pattern Marx assumed with regard
same state of bliss. That this mo
empirically applicable result was
proach to bliss being gradually s
profits." But this does not preve
process from serving, as in the f

Conclusion

We have reached the point at which our results can be

summarized in a few general propositions:
1. Our first and primary problem has been the derivation of
empirical price ratios. It has turned out that neither the utility
theory nor the labor theory of value succeeds in presenting a
self-contained analytical solution. Both are inseparable from
two kinds of historical context. One concerns specific features of
the respective theory - the initial endowments in utility theory,
the role of different skills in labor theory. The other concerns
the evolution of the modern market system, for whatever
empirical validity either theory may possess refers only to the
competitive stage of capitalism. Recalling our criterion for a
"nonprice" determinant of actual price ratios, we must admit
that neither approach would permit us to restore the present
price cosmos if amnesia were to wipe out our recollection of
them, other than by a complex process of trial and error.29

28 Marx's notion of "simple commodity production" (Capital, vol. 3, ch. 10) resem-
bles the structural features of the state of bliss. It even refers to empirical, though
precapitalist states in which farmers and craftsmen own their means of
production - an indication that exchange value is not conditional on the presence of
"alienated" labor.
29 The ingenious attempt of Anwar Shaikh, made in an unpublished study, to
reduce empirical prices directly to labor values derived from input-output tables, does
not refute what is said in the text. Even if his statistical technique goes unchallenged,
his original data necessarily reflect what was called above the "dispersion of skills."
What his study does achieve is to raise doubts about the empirical significance of the
transformation problem.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 811

These defects common to both approaches must not

obscure important differences between them. Within the lim-
itation stated, the labor theory is more comprehensive because
it can account for price ratios in a homeostatic as well as a
maximizing system. On the other hand, once prices are de-
termined, utility theory can establish the qualities and quan-
tities of the commodities transacted - an issue that the labor
theory has never raised.
All in all, the historical nature of both solutions does not
come as a surprise to those who have always conceived of
economic processes as embedded in a wider sociohistorical
context. Lacking the elegance of self-contained pseudo-
mechanical models, the historical frame of reference in
which both theories must be understood points to the con-
tinuity as well as to the transformation that characterize actual
market systems.
2. The situation changes drastically once we transform the
exchange values arising in equilibrium into a norm. It is true
that, to arrive at an unambiguous standard in this way, we
must impose restrictive qualifications that go far beyond the
conventional conditions for equilibrium. The normative state
then appears as one in which, within a given technological
horizon, the maximum physically possible level of output is
attained, with zero profits and the rewards of labor as the only
net income. In this ideal state, the two approaches converge.
3. And yet this is not the last word. We have not yet spelled
out the ultimate criteria that are to raise this state of bliss and
its price order to the rank of a norm.
There is, first, a technological criterion. By assuring the
maximum net output, such a system operates with maximum
technical efficiency. But such a technological maximum satisfies
also the welfare criterion of maximum provision - a criterion
valid so long as the system is exposed to scarcity in the sense of
objective destitution, as well as of the subjective experience of
undersatiation. Finally, we must accept as a third criterion
equality of provision in accord with the input of labor skills. In a

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 812 SOCIAL RESEARCH

word, our norm refers to an egalita

vision under conditions of scarcity
that none of these criteria has a
economic aspect, the "relativity"
welfare criterion is especially relev
in homeostatic systems where som
achieve full satiation because the w
identical with the technological ma
lier that, above the level of destititution, a homeostatic
economy is today by no means a Utopian data. Moreover, not
only a goodly number of primitive societies but the highly
civilized communities of the medieval town economies resem-
bled such an order as well. Therefore, it is not without interest
to ponder the views of a late classical economist as to the
prospects of homeostasis in a fully industrialized society.
In his well-known discussion of the stationary state, Mill
waxes quite enthusiastic when imagining a society with these
features:

A well-paid and affluent body of laborers ... a much larger

body of persons than at present, not only exempt from the
coarser toils, but with sufficient leisure, both physical and men-
tal, from mechanical details to cultivate the graces of
life. . . . There would be as much scope as ever for all kinds of
mental culture and moral and social progress; as much room for
improving the Art of Living, and much more likelihood of its
being improved, when minds ceased to be engrossed by the art
of getting.30

What Mill depicts as an ideal state is a society in which all

strata of the population shift an ever larger part of their
energy and time from labor-requiring priced goods to the
immaterial "utilities" of mental culture.
An even more radical transformation, in which labor input
loses its power of value creation altogether, was anticipated by
Marx himself:

... to the degree that large industry develops, the creation of

real wealth comes to depend less on labor time and on the
;J0John Stuart Mill, Principles of Political Economy, bk. 4, ch. 6.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 813

amount of labor employed than o

in motion during labor time, w
itself in turn out of proportion to
their production, but depends r
science and the progress of tech
ifests itself ... in the monstrous d
time applied, and its product, as w
ance between labor, reduced to
power of the production proce
longer appears so much to be in
process; rather, the human bein
and regulator of the productio
labor in the direct form has cease
wealth, labor time ceases and must c
exchange value [must cease to be
The free development of individ
tion of the necessary labor to a m
to the artistic, scientific, etc. dev
the time set free. . . . The measure of
any way, labor time, but rather disp

What was at the time when M

vision more and more resemb
capitalism with its progressiv
lated by what has been named
tion."
4. This raises a final question, on the surface merely a
terminological one, but really a paradox for which I have not
found a solution. Why has the standard of economic mea-
surement that is supposed to underlie price ratios been associ-
ated since the days of Aristotle with the notion of "value"?
Our attempt at moving the concept of exchange value from
the empirical to the normative level might be taken as an ex
post justification of that terminology. However, when we con-
sider the meaning that attaches to the term value in the fields
of ethics and esthetics, not to speak of human relations gener-
ally, we should expect optimum welfare - the normative
goal - to coincide with the maximization of exchange values.

31 Karl Marx, the Grundrisse (New York: Random House, 1973), pp. 704-708; my
italics.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 814 SOCIAL RESEARCH

This is indeed so under the mi

individual's command over prov
exchange value of what he has
forget that someone else's provis
But what about the macroecon
assess optimum aggregate welf
acter of exchange value as a meas
comes into the open. To see th
attention from labor input rat
ratios to the absolute inputs of
achieve a stipulated level of agg
ous that the welfare attached to
related to the absolute labor re
timum being the Garden of Ed
and thus the related value is ze
In this we only reformulate the
stated in ch. 20 of his Principles,
"riches." He correctly identifies
tity of goods obtainable in excha
zation of which indeed coincides
fare, always assuming scarcity. B
not on abundance, but on the dif
tion." If this is so, what sort of
of which would be the conditio
true?

Now is economic value still a problem? When first raising

this question we found that it could be answered from two
quite different viewpoints: whether a theoretical solution can
be found acceptable to both parties or, solution or no solution,
whether the debate is irrelevant at the level of economic prac-
tice. Taking the latter viewpoint first, we cannot help admit-
ting that, even if a perfect solution were presented, the discus-
sion of the problem moves at much too high a level of abstrac-
tion to matter when we try to explain or predict a concrete
event, much less when we try to frame economic policy. For

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC VALUE 815

this reason, in practical enonom

from a given historical situatio
order as a datum. We then arg
forward, as the analytical task r
tler issues that relate to the "basement" on which the theoreti-
cal edifice rests.
The case is different for the pure theorist. If for no other
than methodological reasons, he is interested in what goes on
in that "basement." For him our results are gratifying and
disappointing at the same time. They are disappointing be-
cause there is no ultimate determinant of empirical prices
without abandoning the realm of pure analysis for that of
history. It is gratifying because what there is - a normative
principle by which we can adjudicate the performance of ac-
tual competitive markets - can be derived from either of the
two rival approaches. In part their answers are complemen-
tary. Thus the psychological underpinning of maximization or
the determination of the qualities and quantities transacted
are achievements of the utility principle. Mastery of homeo-
stasis and the stress on the "resistance" character of exchange
value go to the credit of the labor principle. For the rest, quite
surprisingly their answers turn out to be identical.

* I wish to express my sincere thanks to my colleagues Ron Blackwell, Robert L.

Heilbroner, Edward J. Nell, and Anwar Shaikh for their criticai comments on the first
draft of this paper, from which I have greatly benefited.

This content downloaded from

134.193.238.8 on Wed, 17 Nov 2021 01:13:42 UTC
All use subject to https://about.jstor.org/terms