Full text of The Four Kinds of Economic Value

View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

THE FOUR KINDS OF
ECONOMIC VALUE

LONDON: HUMPHREY MILFORD
OXFORD UNIVERSITY PRESS

THE FOUR KINDS OF
ECONOMIC VALUE
CORREA MOYLAN WALSH

CAM BRIDGE
HARVARD UNIVERSITY PRESS
1926

PRINTED AT THE HARVARD UNIVERSITY PRESS
CAMBRIDGE, MASS., U. S. A.

CONTENTS
I. Introduction ................................................. 3
II. History of their developm ent...................... 4
III. The first recognition of them as coordinate
species................................................... 11
IV. D efinitions..................................................... 15
V. Their proper o rd er......................................... 20
VI. Their objectivity............................................. 22
VII. Their re la tiv ity ...................................... 25
VIII. Their comparableness ................................. 27
IX. Their variability...................................... 34
X. Our attitude toward their variations. . . .
41
XI. Their knowableness............................... 46
XII. The measurement of the first three........ 49
XIII. The measurement of exchange-value . . . .
56
XIV. The question of cau ses............................ 62
XV. The causes of the first th ree.................... 64
XVI. The cause of exchange-value ■— Ricardo’s
T heory.................................................... 71
XVII. Its inconsistency....................................... 82
XVIII. On labor-cost........................................... 96
XIX. General exchange-values according to com
parative social esteem -values................100
XX. Their divergence in the course of time . . . 107
XXI. Practical uses of the measurements..........115
XXII. In which kind ought money to be stable? . . 126
x .

THE FOUR KINDS OF
ECONOMIC VALUE

THE FOUR KINDS OF
ECONOMIC VALUE
I

INTRODUCTION
E v e r y science needs to classify its subject-matter.
Chemistry deals with the substances in the material
world, and it tries to analyze them into their elements.
Physics deals with the forces of nature, and discovers
and investigates many such. And when the species and
varieties are found, they need to be named in order to
distinguish them from one another and from the genus
to which they belong. The generic term itself is prop
erly used only when all its species and varieties are
included: when these are referred to specially, their
distinctive names should be used. In the natural history
sciences the subdivisions are innumerable, and many
technical terms have to be invented. In the intellectual
and moral sciences the subdivisions are comparatively
few. The purpose of this paper is to show that, just as
in ethics virtue is the main theme and there are several
kinds of virtue, and just as in metaphysics causation is
the main theme and there are several kinds of causes,
so in economics value is a main theme and there are
— and it should be recognized that there are — sev
eral kinds of value. Rather curiously, the ancient phil
osophers had four (cardinal) virtues and four causes,
and modern economists have analyzed value into four
kinds. These are use-value, esteem-value, cost-value,
and exchange-value.

II
HISTORY OF THEIR DEVELOPMENT
A l r ea d y in antiquity Aristotle distinguished between
the utility of an article when put to its intended use, as
of a shoe being worn, and its utility in procuring for its
owner something else in exchange for it. In modern
times Locke applied the same distinction to value, and
called the one “ intrinsic value” and the other “ market
able value.” Before him, Petty made another division,
to the effect that things should be valued by “ land and
labor,” land standing for its material produce, and labor
for cost. Later, Turgot made another division into
“ valeur estimative” and “ valeur échangeable.” Then
all the kinds were exhausted. Meanwhile, other
economists were using the term “ value” in one or an
other of these senses, or in more than one. Later comers,
confining themselves to dichotomy, could only repeat
what had already been done, at best with improved
terms. Thus Quesnay and Adam Smith followed Locke,
the former speaking of “ valeur usuelle” and “ valeur
vénale,” the latter making the division into “ value in
use” and “ value in exchange,” or, translating Turgot’s
term, “ exchangeable value.”
In his comprehensive survey of the subject Adam
Smith could not confine himself to these two kinds of
value. The other two kinds obtruded themselves. Un
fortunately, instead of assigning them coordinate posi
tions, he subsumed the one under exchange-value and
went outside the concept of value to get a term for the
other. He conceived of exchanges as taking place, not
only between goods and goods, but between goods and

FOUR KINDS OF ECONOMIC VALUE

labor. What the Roman law said of contracts,1 he said
of exchanges. When the owner of a commodity A ex
changes it for a commodity B, Adam Smith said that
to this man the “ exchangeable value,” or simply the
“ value,” of A is B, but the “ price” of B is A. Here, to
the other party, the “ value” of B is A, and the “ price”
of A is B, so that, in such an exchange, “ value” and
“ price” are practically the same thing viewed from
opposite sides. Not so, when a commodity is supposed
to be exchanged for labor. This alleged exchange is of
so peculiar a nature that something new is introduced.
When A is given for a day’s labor, and a day’s labor is
given for A, Adam Smith now said that the “ real ex
changeable value,” or simply the “ real value,” of A to
its owner is the day’s labor he obtains for it, while the
“ real price” of the day’s labor is A; but to the laborer
the “ real price” of A is the same day’s labor he has
given to obtain it, and the “ real value” of his day’s
labor — this is the logical sequence, but it was omitted
— would be A. “ Equal quantities of labor at all times
and places may be said to be of equal value to the
laborer.” This obviously is not exchange-value, since
no reference is made to any exchange. A laborer may
be said always to set the same estimation upon his labor:
this can be only esteem-value. When a laborer ex
changes his (by himself) equally prized labor, he may
at one time and place get more of A for it than at others,
and therefore, to him, according to a logical deduction
from Adam Smith’s use of his terms, while the “real
price ” of A has decreased, since the original quantity of
A is obtained with less labor, the “ real exchangeable
1 Do ut des, do ut facias, facio ut des. The fourth, facto ut facias, our “ log
rolling,” did not interest him.

FOUR KINDS OF ECONOMIC VALUE

value” of his labor would have increased. On the other
side, as the employer now gets less of the laborer’s labor
for the original quantity of A, A is conceived by him as
having fallen in “ exchangeable value.”
Adam Smith took the employer’s standpoint, and
used labor as his measure of “real value” ; which in
volves a strange inconsistency, since labor is invariable
only to the laborer. He did, however, try to take a
broad and philosophic view of the case. Labor, he said,
is “ the first price” — also “ the ultimate price” — that
is paid for everything. This idea he may have got
either from Hobbes, who had spoken of commodities
which God “ for labor selleth to mankind,” or, more
likely, from Turgot, who had alluded to a commerce
between man and Nature, who exacts from him labor
as the price he must pay for what she yields.2 Yet even
so, this price is paid only by laborers, and when those
who have engrossed the sources whence Nature yields
and, having acquired some of the produce, “ buy labor”
with it, they do not buy what is always of the same
value to themselves, but only what is of the same value
(esteem-value) to the laborers. The laborers’ standard
cannot be properly made the employers’ standard.
All this refers to value “ at all times and places.” As
for the comparative values of different commodities
“ at the same time and place,” these Adam Smith ex
plained by the labor-costs of producing them, mixed in
with rent and profit. In other words, their exchangevalues are according to the prices, paid in labor, at
which they are sold by Nature. These prices are now
2 Hobbes, Leviathan, chap. 24. Turgot, Valeurs et Monnaies, ed Daire,
pp. 82-83. The idea dates back to the comic poet Epicharmus, as quoted
in Xenophon’s Memorabilia, II. i. 20.

FOUR KINDS OF ECONOMIC VALUE

more properly called costs.3 Adam Smith showed not so
much interest in this as in the other question.
This idea of an exchange of commodities and labor, —
not to stop at that between laborers and Nature, —
which underlies Adam Smith’s conception of value, is a
tremendous blunder, and has been a far-reaching cause
of error ever since his day. It is utterly improper, being
only metaphorical, since no such exchange is made in
reality. An exchange is a mutual transfer of two objects.
Labor cannot be transferred, but only the result of
labor — the work done. When we work to produce
something, we do not give any labor, we merely do
something, and it is done. When we hire laborers to do
something for us, we do not receive their labor, and we
do not care a rap for their labor: what we want is the
work done, and that is all we get. Labor itself is not a
commodity: it is rather a discommodity. A very
mischievous misuse of words is it to speak of laborers
“ selling” their labor. It gives them the impression that
all they have to do is to perform for a certain time cer
tain operations called labor, and thereby they are ful
filling their side of the contract and need not concern
themselves about the thing they are really giving in
return for their pay — the thing produced. What la
borers really give is evident in piece-work. In time
contracts, it should require little analysis to see that
the essence of the exchange is the same. Manufacturers
use the latter method for convenience, and, if they
choose, may talk of “ paying for labor.” There is no
3 “ Price,” in the improper wide sense, is what in an exchange one gives
and another receives; “ cost” is what one gives up for a thing and nobody
else receives. This refers to the real thing involved in this operation. In socalled money-costs, of course, the money element is transferred.

FOUR KINDS OF ECONOMIC VALUE

special convenience in this phraseology for the econo
mist, and to him even more than to the workman it is
misleading, since it is not a true statement of the things
with which he deals.
Fortunately, however, Adam Smith’s other mistake
of disguising cost-value under the misused name of
“price” was not continued by many of his successors.
The next great English economist, Ricardo, confined
“ price” to its usual reference to money. But Ricardo
fell into a confusion parallel to Adam Smith’s confusion
of esteem-value with exchange-value, for he confounded
exchange-value with cost-value. He accepted the im
perfect division of value into use-value and exchangevalue, and, like Adam Smith, maintained about “ value”
or “ real value” a doctrine which is properly applicable
neither to use-value nor to exchange-value; only it was
a different doctrine from Adam Smith’s, and as Adam
Smith’s was applied to an unnamed esteem-value,
Ricardo’s was applied to an unnamed cost-value. For
Ricardo held, like Adam Smith, that the “ value,” even
the “ exchangeable value,” of a commodity is not ac
cording to “ the abundance of other commodities for
which it will exchange,” but is according to “ the sacri
fice of toil and labor required to produce it” ; and this
he held not only of the value of one thing compared
with the values of others at the same time and place, in
which he differed from Adam Smith only in some de
tails, but — and here is where he chiefly differed from
Adam Smith — he held it of the value of the same thing
at all times and places. The doctrine he advocates is
nothing but a doctrine of cost-value, while the doctrine
he rejects is the proper conception of exchange-value.
Yet, having only the one term “ exchangeable value,”

FOUR KINDS OF ECONOMIC VALUE

since use-value has been passed by, he thus rejects from
exchange-value what belongs to it and tries to fasten
upon it what belongs to another kind of value. But the
force of words is too strong, and, at least when speaking
of contemporary things, he several times defined the
“ value” of a commodity as the quantity of other com
modities for which it will exchange. For this kind of
value, really exchange-value, he now had no distinctive
term, because the term properly belonging to it he had
misapplied to something else.
Ricardo was opposed by Malthus, who returned to
Adam Smith’s use of “ value,” even of “ exchangeable
value,” for esteem-value, and was followed by McCul
loch, who continued the use of “ value,” even of “ ex
changeable value,” for cost-value. Thus Malthus
maintained that the variable quantity of goods paid for
the same quantity of labor, and McCulloch that the
variable quantity of goods produced by the same quan
tity of labor, has constant value. English academic
economists for long were divided between these two
opinions, and some have held both without being clearly
conscious of what they were doing; while many out
siders have tried to stick to the obvious meaning of
exchange-value, or did so merely because they did not
go deeply into the subject. Prominent among those
who took exchange-value in its literal sense were Joseph
Lowe and Poulet Scrope; but their views attracted
little notice at the time, and had no influence. On the
continent, in France, Say did not employ any word
corresponding to use-value, and used “ real price” for
cost-value, confining “ value” to exchange-value, but
at times extending it to cover also cost-value. Like him
French economists have mostly followed Ricardo in

POUR KINDS OP ECONOMIC VALUE

paying more attention to cost-value than to esteemvalue. In Germany the dichotomy into use-value and
exchange-value long held sway, but more attention was
paid to esteem-value. This, however, was treated dif
ferently from the way it was treated in England (except
by Mill); for it was generally subsumed under usevalue, which, therefore, by this attachment to esteemvalue, was brought more into the foreground.4
4 References to the opinions of these many writers may be found in my
The Fundamental Problem of Monetary Science, Part I, chaps. 3-7.

I ll

THE FIRST RECOGNITION OF THEM AS
COORDINATE SPECIES
A t last, in the third and fourth quarters of the nine
teenth century, Roscher and Jevons put the several
kinds of value on a coordinate footing, although each
made an incomplete division. Roscher divided economic
value into
use-value,
cost-value, exchange-value;
and Jevons divided it into
use-value, esteem-value,
exchange-value.
Jevons rejected cost-value altogether; but Roscher had
to think of esteem-value, and he did so by extending the
meaning of use-value to cover it .1 Still, from these two
divisions the complete four-fold division plainly emerges.
Unfortunately, contemporary and later economists
have shown little inclination to adopt this classifica
tion, much less to employ these terms. Walras, who
shares with Jevons and Menger the claim to be a founder
of present-day economics, used only the two terms
“ value” and “ exchange-value,” which he at times
assimilated and at times differentiated, while the subject
of his inquiry mostly was esteem-value, for which he
1 Jevons did not use the above terms. He divided “ value” into value in
use, esteem, and -purchasing power or ratio of exchange, in the 2d edition of his
Theory of Political Economy (1879), p. 87. Roscher, Die Grundlagen der Nationaloekonomie, §§ 4, 5, used Gebrauchswert, Kostenwert, and Tauschwert,
which are immediately rendered by use-value, cost-value, and exchange-value
(“ exchange value” having been introduced by DeQuincey, who also men
tions “ cost value” as being in use), upon whose model the term esteem-value
is readily constructed. Roscher followed Ricardo in maintaining that ex
change-value is regulated by cost-value.

FOUR KINDS OF ECONOMIC VALUE

had no name except “rarity,” strangely employed.
And the Austrian school, following Menger, have be
clouded the matter by introducing a variety of long
terms drawn from metaphysics and psychology. Take
Bohm-Bawerk’s, for example. His “ value in the ob
jective sense” is use-value strictly confined to intrinsic
utility; his “ value in the subjective sense” is esteemvalue; his “ objective exchange-value” is simply ex
change-value; while his “ subjective exchange-value” is
nothing but esteem-value over again, confined to things
like paper-money that are useful only in exchanges.
Twenty-five years ago I recommended the use of the
four simple terms. My advice was proffered in the belief
expressed by the astronomer-statistician Quetelet, that
“ it is conducive to the progress of enlightenment to
substitute precise ideas for vague notions.” 2 It was in
strict accordance with the precept of the logician Mill,
who wrote that “ whatever we have occasion to think
of often, and for scientific purposes, ought to have a
name appropriate to it.” 3 Shortly afterward similar
2 Lettres sur la Théorie des Probabilités, p. 61.
3 Logic, III, vi, § 1. In his Political Economy Mill did not observe his
own precept. He there said : “ The word Value, when used without adjunct,
always means, in political economy, value in exchange” ; III, i, § 2; and cf.
viii, § 2. This implies that the only proper meaning of “ value” in economics
is exchange-value. But Mill himself, in his Summary, vi, 1, 2, spoke of “ cost
value” and of “scarcity value” (= esteem-value), and in distinction used
the full term “ exchange value.” A follower of his, H. Passy, in the Dictionnaire
de l’Économie 'politique, art. “ valeur” (ii, 810), made another recommenda
tion, on a line with Mill’s intention; which was that we should use “ value”
only in the sense of exchange-value, and for the other meanings we should
use other term s— “ utility” for use-value, “ cost” for cost-value (and ac
cordingly “ esteem” for esteem-value, as Jevons afterward did). But then
another generic term to embrace them all would be needed, which is lacking.
Also it requires that precedence be given to one of the kinds, on which agree
ment does not exist. Some economists (e. g., Carreras, Loria) would confine

FOUR KINDS OF ECONOMIC VALUE

advice was given, at least indirectly, by the Danish
writer of French works, Christian Cornélissen ; 4 but
unfortunately he has only Roscher’s three terms —
use-value, labor- or production-value for cost-value,
and exchange-value. More recently, in England, Mr.
G. B. Dibblee, in The Psychological Theory of Value,
teaches what he calls “ the multiple explanation of
value,” but fails to employ or to recommend appropriate
terms. The values he deals with are likewise only three
— those of Jevons, with esteem-value in the place of
cost-value. Both these writers suffer for not making
the complete classification. The latter also reverts, as
the title of his work betrays, to the old fault of seeking
a single theory of value. These are the only writers
whom I know of, outside the German-speaking coun
tries, who attempt to distinguish between the several
kinds of value. Meanwhile, the harm done by the old
mistake of endeavoring to get on with only one concept
of value is well shown by the falling-out of two such
men as Professor Irving Fisher and Dr. B. M. Ander
son. The former, in The Purchasing Power of Money,
has no use for any kind of value but exchange-value.
The latter, in The Value of Money, has only one value,
“ value” to cost-value; others {e.g., Menger) would confine it to esteemvalue; still others (e. g., Gossen) would confine it even to use-value, all em
ploying “price ” for exchange-value. It is now too late to adopt this recom
mendation. “ Value,” not only in popular usage, but in that of economists
in general, has acquired these various meanings, which cannot be taken out
of it. The only course remaining is to distinguish them by specific terms,
under “ value” as the genus.
4 His words are: “ Je suis convaincu que . . . on aurait pu s’épargner
beaucoup de malentendus et d’erreurs, si l’on avait mieux distingué la nature
de la valeur sous ses diverses formes et évité de désigner par le même terme
général et par conséquent vague de ‘valeur’ des notions tout à fait diffé
rentes.” Théorie de la Valeur, (1902), p. 21; 2d ed. (1919), p. 35.

FOUR KINDS OF ECONOMIC VALUE

which he calls social, psychological, absolute, and
which appears to be esteem-value, though he rejects
the prevalent theory of this kind of value and does
not elaborate a theory of his own. Consequently he
has been unable to appreciate Professor Fisher’s posi
tion. Rather curiously, Dr. Anderson conceives of all
value as “ social value,” and this he divides into three
genera — moral value, legal value, and economic value.
Strange, then, that economic value has no subdivision!

IV
DEFINITIONS
“ Value ” is one of those familiar words which it is diffi
cult to define, for the very reason that they are so well
understood without definition. For the same reason,
its definition is not so important as are the definitions
of its species. Value in general has a very wide range of
meanings in ordinary speech, and it is used even in
several arts and sciences, as in morals, music, painting,
and mathematics. Economic value, 1 in its generic
sense, is a quality or power in, or somehow connected
with, appropriable things with reference to the well
being and activity of the persons who do, or who would
if they could, appropriate them. The nature of this
quality or power is stated in the specific definitions.
Use-value is a thing’s power to serve our ends. Esteemvalue is its power to make us desire to possess it. Costvalue is its power to impose upon us effort to acquire it.
Exchange-value is its power to procure other things in
its place. The first is a power to preserve us alive and
provide the basis of a “ good life” in the Greek sense.
The second is a power to cause us to strive for the acqui1 From a Latin noun and the past participle of the corresponding verb,
the Italian language has two words, valore and valuta, the former properly
referring to exchange-value and the latter to esteem-value. The French gets
its single valeur from the former, and the English its single value from the
latter; which single terms have to do double duty, and are perfectly equiva
lent. Even the Italian valore has been extended to the wider service, valuta
being relegated to a few technical uses. (In English, worth is hardly employed
in economics with any distinctive sense, any more than “ work” differs from
“ labor.” This word, like the German Werth, originally combined the ideas
of utility and honor, and was therefore applicable also to persons. It thus
covered the two ideas of use-value and esteem-value.)

FOUR KINDS OF ECONOMIC VALUE

sition of the desired thing by means of the labor which
goes to the making of the third kind, and to keep it when
we have it so as to avoid the trouble or sacrifice of re
placing it. The fourth is its power enabling us, by
relinquishing our possession of it, to obtain another or
other things for which at the time we have a greater
desire, or, in other words, upon which we set a higher
esteem-value. The last is especially the kind which
permits the increase of wealth. In economics, therefore,
it is the most important kind of value. It is the one
kind which no economist has ever ignored.
On the other hand, the least important kind of value
in economics is use-value, because of its slight differen
tiation from usefulness or utility, which is a physical
quality in things in conjunction with physiological
qualities in us. In this first sense, the most valuable
thing in the universe for us is air, as without air we can
live only a few minutes. Air, in a way, may be said to
be appropriated when we breathe it into our lungs, and
for this reason it seems allowable to speak of it as hav
ing use-value. But no one thinks of calling the sun,
moon, and stars “ valuable,” — although they cer
tainly are very useful to us, — since they can in no way
whatever be appropriated. As for air, it ordinarily has
none of the other kinds of value; for, having it always
around us, we do not feel any want of it, do not there
fore desire it, do not need to exert ourselves to acquire
it, cannot lose it, cannot exchange it. It may, however,
have them in special circumstances. Thus, when a per
son who cannot swim falls overboard at sea, at once air
becomes to him an object of the intensest esteem-value,
impelling him to put forth all his efforts to acquire it
alone as the one thing in the world he then wants. And

FOUR KINDS OF ECONOMIC VALUE

to the professional diver, as to the miner, air has costvalue. If someone like TEolus could put the winds into
bags, he might sell them at high prices. Luckily this
cannot be done. But well-aired rooms command a
higher rent than otherwise equally good but badly
ventilated rooms, and therefore the air in those rooms
may be said to have exchange-value. So light in open
places by day has no value; by night, and in dark places,
it becomes endowed with all the values. Heat also loses
and gains them.
The next most valuable thing, according to its usevalue, is water, without which we can live only a few
days. Not so abundant as air, nor so easily acquired,
it ordinarily has some esteem-value, though it may be
without any to a person lying on the brink of a brook.
It has esteem-value, because, not being always at hand,
we sometimes feel the want of it, desire it, and then
have to go after it, put forth some effort to get it, and
take some pains to keep it till we consume it. It then
has cost-value, as especially to people in a city, who have
to build and maintain reservoirs and aqueducts. In
some arid regions it has exchange-value, as is evidenced
by itinerant water-venders, and by the sale of water for
irrigation. But, ordinarily, though its use-value is great,
its other values are small. When we do not distinguish
between the different kinds of value, this is a paradox,
which has been descanted on by poets and philosophers
from Pindar and Plato down. When we recognize the
different kinds, the paradox ceases.
After these two things, with the exception of land,
which properly has no cost-value, and possibly of a few
minor articles, such as ambergris, which is found only
accidentally, everything useful to mankind has all the

FOUR KINDS OF ECONOMIC VALUE

four kinds of value. It is not correct to except things
that are not ordinarily for sale, such as public buildings,
parks, and monuments, Roman and Greek Catholic
churches, entailed estates, and so forth, and to say that
these have no exchange-value. They lack exchangevalue only while held, and differ from other capital
goods only in being held longer.
As for money, its use-value resides in its serving as a
medium of exchange and as a measure of value not only
at the same time and place but at different times and
places. It has esteem-value, as is obvious, otherwise
there could be no misers. It has cost-value, though in
the case of paper money this is small. It has exchangevalue as its very essence. The use of a thing for obtain
ing other things we have seen Aristotle treat as a second
ary use. This is true of commodities, but not of money.
Money’s use to procure other things, either at once or
at a later time, meanwhile serving, whether hoarded
or loaned, as a store of value, is its only, hence its pri
mary use.2 This is its use to individuals. To the public
at large, in political economy, money is one of the most
useful of things. Perhaps, besides air and water and
the land on which we live, it is the only one kind of ma
terial things that is indispensable for the development
2 Because money is not used as consumable goods are used, Rau said that
money has no use-value, Volkswirthschaftslehre, §§ 64, 265. Similarly BohmBawerk, Positive Theorie des Kapitales, 4th ed., i, 205. This comes merely
from their arbitrarily confining “ use-value” to the first of Aristotle’s kinds
of use. In a way money is consumed, as its individual pieces wear out and
have to be replaced. So our individual houses wear out and have-to be re
placed, but houses as a whole class will probably last as long as mankind, and
money as a whole class also. The only difference is that it is the whole people
who consume and through their government replace their money, while it is
individual persons who consume and replace their own houses.

FOUR KINDS OF ECONOMIC VALUE

of civilization. Of course, food is indispensable for our
very existence, as we can live without food only a few
weeks; also shelter, without which, at least in tem
perate zones, we could live only a few months. But no
one kind of food or of shelter is indispensable.

THEIR PROPER ORDER

order in which the kinds of value have been named
is the right one. Say, apparently, would have put costvalue even before use-value; for he affirmed that it is
labor that gives utility to a thing, instancing that coal
under ground is useless and becomes useful only when
labor has brought it to a furnace. It would seem plain,
however, that all that labor does with regard to utility
is to bring a useful thing from where it cannot to where
it can be used. The same labor, if it brought sand to a
furnace, could not put utility into it. Use-value exists
first, and next comes, not yet cost-value, but esteemvalue. This is different from use-value in that esteem
in us for a thing, of which the esteem-value in, or of, the
thing is a reflection, is produced by something else in
addition to the thing’s utility, which other cause is its
condition of being scarce, or in not so great abundance
as to satisfy our need of it entirely. “ The rare,” said
Plato, “ is esteemed” or “ valuable” — it has esteemvalue.1 The utility of a thing remaining the same, we
desire the thing more if we have less of it, and less if we
have more of it. Obviously esteem-value is very dif
ferent from use-value, though it could not exist without
it. In fact, all the other kinds of value are based upon,
and presuppose, use-value, that is, utility and appro
priableness.
Having esteem-value, a thing may impel us to exert
ourselves to acquire it, and thereby it becomes endowed

T he

1 Euthydemus, 304B. The Greek language had two words for “ value” :
Ti/xi) for esteem-value, and
for exchange-value.

FOUR KINDS OF ECONOMIC VALUE

with cost-value. Cost-value thus comes after, and pre
supposes, esteem-value. Another way of acquiring a
thing is by giving in exchange for it something we
already possess. This involves exchange-value in the
two things. Thus exchange-value also follows, and pre
supposes, esteem-value. But exchange-value does not
necessarily presuppose cost-value, nor does cost-value
necessarily presuppose exchange-value. Exchange-value
is here placed last because of its greater importance.
What we need to bear in mind is that both are de
pendent on esteem-value. “A thing,” said Condillac,
“ has not value because it costs, but it costs because it
has value.” 2 This reminds one of Socrates’s question
whether certain things are holy because they are be
loved of the gods, or are beloved of the gods because
they are holy. It is of the utmost importance to get the
right sequence of cause and effect between many such
couples that are closely bound together. A fuller ex
amination of the causes of the various kinds of value
must be postponed.
2 Le Commerce el le Gouvernement, I, chap. 1. If for “ value ” in both parts
of this sentence be substituted the specific values, the same in both, the prop
osition is true except obviously in the case of cost-value, though it has been
questioned also in the case of exchange-value. The sentence is a good illus
tration of the need of using the specific terms, as different meanings may be
slipped, almost unconsciously, into the same word in the two parts.

VI
THEIR OBJECTIVITY

E x c h a n g e - v a l u e has several peculiarities differentiat
ing it from the other kinds, which we shall come upon
as we proceed. Here already appears an important one.
Things may have the first three kinds of value to a man
alone in the world, but they can have exchange-value
only to men in society. The things a man has with him
when he alights in a wilderness there lose all exchangevalue, while the other values of some increase and of
others decrease.
Because of this fact, it has been maintained that exchange-value cannot be a quality of things, for the rea
son that, if it were, it would exist in them always and
everywhere. The same denial of objectivity has been
extended even to all values in general, because the other
kinds also come and go in different circumstances —
because, in short, of their variability, which will be
examined presently. Value, it has been asserted, not
only is subjective, but it is only subjective; and a
couple of writers have declared that it signifies merely
“ a state of mind.” This opinion is not borne out by
facts in the case of other unimpeached and unimpeach
able qualities of things. The weight of one body re
quires the existence of other bodies, and in definite
relations. Take a stone from the surface of the earth
and place it at the center, or at a central point in
interstellar space, and it would lose all its weight. Heat
notoriously changes with position. Colors do not exist
without light. The taste of food exists only when one
puts it into his mouth. All qualities have a subjective
22

FOUR KINDS OF ECONOMIC VALUE

as well as an objective side. They exist first in us as
sensations, or as conceptions worked up from sensa
tions, of which we conceive causes to exist in the out
side things which occasion them; which causes we call
by the same names. There is no reason why value, or
in particular exchange-value, should be treated dif
ferently from the other qualities of things. Whether the
outside causes exist or not, is a question for meta
physicians. A German economist has said that value
is not a property of things in themselves.1 Well, Kant
said that weight, heat, color, and the like, even size,
mass, and motion, are not properties or states of things
in themselves. We may accept both statements, or
neither. Or we may cling to Descartes’s distinction
between primary and secondary properties. Then all
the kinds of value, not even excepting the first, would
have to be regarded as secondary properties. If prop
erties are not in the substances, we at all events, in our
conception of them, put them into phenomenal things,
and get along as well. Thus it does not matter how a
quality gets into things, when we are all agreed that it
is there. If many people living together put exchangevalue into certain things, any one person, living among
them, may find it there and take advantage of its pres
ence, and if he be an economist, he may study it as he
finds it .2 Psychological explanations are for psycholo1 A. Wagner, in his rewriting of Rail’s work, § 33.
2 Cf. Walras : “ Le fait de la valeur d’échange prend donc, une fois établi,
le caractère d’un fait naturel, naturel dans son origine, naturel dans sa mani
festation et sa manière d’être.” Eléments d’Economie politique pure, § 28.
Yet this did not preserve him from asserting that exchange-value is only sub
jective, only in us, and that the money-unit has no value, — § 144 (§ 146 of
the 4th ed.),— a position which, needless to say, he could not maintain.
Besides his defective nomenclature, this is the one flaw in Walras’s otherwise
almost perfect work — and it seems to spring from his defective nomencla
ture.

FOUR KINDS OF ECONOMIC VALUE

gists. Economists are concerned with economic facts,
and with the interplay of economic facts. Exchangevalue is now, in civilized peoples, an immense power.
By means of a little letter of credit, carried in his pocket,
a man can travel round the globe. The fact that if, by
mishap, he strands on a desert island, or one inhabited
only by savages, the exchange-value of this letter van
ishes, in no wise militates against the fact that it exists
when and where it does exist. Similarly in the case of
the other kinds of value. All values, like all qualities,
exist only under the conditions that make them exist;
and they exist as objective attributes, no matter how
subjective some of their causes may be.

VII
THEIR RELATIVITY

value is relative. The first three kinds are relative,
each to something outside its own kind. Use-value is
relative to the usefulness or utility of things; esteemvalue, to our desires for them; cost-value, to the costs
of producing them. But the exchange-value of one
thing is relative to the exchange-values of other things:
exchange-value may be said to be interrelative. Here
is another clean-cut distinction between exchangevalue and all the other kinds. In an unprecise way of
speaking, the others may be described as absolute, since
each of them in one thing is independent of, or out of
relation to, the similar values in other things, or nearly
so, not being quite so, as we shall see, in the case of
articles that have substitutes. The rise or fall of one
thing in any of these values does not involve any change,
much less any inverse change, in the corresponding
values of other things. It is possible to conceive of all
use-values rising or falling together, of all esteem-values
rising or falling together, of all cost-values rising or fall
ing together — and, in fact, the last two values are apt
to vary in greater or less agreement.
But all exchange-values cannot rise or fall together,
because the variation of one thing in exchange-value
means a variation of its exchange-value in some other
thing or things, and it involves an inverse variation of
the exchange-value of that or those things in it; and so
all along the line. At best, it is only the totality of all
exchange-values that is absolute, like the totality of all
energy, which in the physical world is the only kind of
A ll

FOUR KINDS OF ECONOMIC VALUE

force that is conserved.1 Strictly speaking, all we have
a right to say is that the first three values, taken singly,
are less relative than the fourth, none being really abso
lute. In this respect they still resemble all other quan
titative properties of things — size, weight, strength, and
the rest. We have no absolute length, no absolute
weight, no absolute force. The units of these quantities
are arbitrarily selected portions that seem to us to re
main constant relatively to all other similar quantities
taken together over against them.
1 See my The Measurement of General Exchange-Value, chaps. 3 and 13.

VIII
THEIR COMPARABLENESS
B e in g rela tive, all valu es in v o lv e com parisons.

First, they contain comparable elements. We desire
things because of their utility, and so there must be
some relation, permitting of comparison, between desire
and utility. We labor to produce things because of our
desire for them, and so there must be some comparable
ness between labor and desire. We exchange one thing
for another because we desire the latter more than the
former. Here the comparison is between desire and
desire.
Secondly, the different kinds of value are comparable
with one another. Esteem-value follows desire, and so,
like it, is comparable with utility, and through it with
use-value. Cost-value follows labor, and so, like it, is
comparable with desire, and through it with esteemvalue. Exchange-value, involving comparison of two
desires, involves comparison of two esteem-values, and
possibly of two cost-values, or of cost-value with esteemvalue. Cost-value and exchange-value have nothing to
do with use-value except through esteem-value.
In the relationship between desire and utility, or be
tween esteem-value and use-value, it may be said that
our desire for a thing and its esteem-value are, in a way,
according to its utility. But almost everything has
various utilities, some greater and some less, which may
be arranged in a descending or ascending scale; and we
always put a thing to its greatest use first, and as we
acquire more of it, we put it to its less use next, and so
on down the scale; or as we use it up, or lose some of it

FOUR KINDS OF ECONOMIC VALUE

(or cannot renew it), we cut off the least use first and
retrench backward up the scale. Now, desire for a
thing operates in two ways: to keep what we have, and
to acquire more. Our concern to keep a thing is to keep
the last and least use we put it to, since that is what we
should first give up if we lost any of it. Our desire to
get more of it reaches beyond the last and least use
to the next one below. These two halves of our desire,
therefore, go in opposite directions from a point which
is, so to speak, the end of the last and least use we make
of the thing. The amount of utility at this point, then,
is what determines or regulates the amount of our desire
for the thing, and according to this is the thing’s esteemvalue — to each of us, individually. But the thing’s
use-value is the sum of all its uses behind and above that
last and least use —■ the total of its uses to each of us as
we are able to use it; and to mankind at large use-value
is, of course, the sum-total of all its uses. Hence usevalue is always greater than esteem-value. This is why
a thing’s esteem-value is so strangely to a first view —
paradoxically, till it is explained — divergent from, and
always below, its use-value, the divergence being in some
proportion with the thing’s abundance.1
The relationship between labor and desire, or between
cost-value and esteem-value, is of an opposite nature.
1 Let us not hesitate to say this of precious things. A diamond of the
size of a drop of water is immensely more useful than that drop of water: its
use-value as well as its esteem-value is greater. "When a rich man buys a
diamond at a high price, he gets what is more useful to him than the sum of
money he gives for it. And all the diamonds in the world are immensely more
uesful to mankind than an equal quantity of water. It is only an enormously
greater quantity of water that is more useful than diamonds. As a fact, man
kind spends much more for water than for diamonds. But the esteem-value
of diamonds is probably very close to their use-value, while the esteem-value
of water is immeasurably far below its use-value.

FOUR KINDS OF ECONOMIC VALUE

The ultimate cost of things is the exertion of effort on
our part to produce them, and this effort is disagreeable.
Some economists speak of labor as a “ disutility.” We
desire to acquire things, and dislike the necessary effort.
We want more of the thing, and less of the labor. How
such dissimilar feelings as a like and a dislike can be
compared may be a problem for psychologists; but for
economists it is a matter of commonest experience.’
Every day people say, “ This is worth the trouble,”
“ That is not worth the trouble.” Now, what is worth the
trouble they will produce, and we interpret this as show
ing that their desire for it is greater than their dislike of
the labor, by a difference which may be more or less,
but which must exist, as otherwise there would be no
motive for the exertion. What is not worth the trouble
they will not produce; which similarly shows that their
dislike of the labor is greater than their desire for the
thing, also by a larger or smaller difference. If the two
opposites be exactly equal, one is not in the predicament
of Buridan’s ass, because no action is required here, and
no production is the result. Mathematical economists
easily dispose of the matter by giving utility a plus sign
and disutility a minus sign, and requiring the former
absolutely to exceed the latter, if production is to take
place.
Here we have considered the thing produced as a
single indivisible article. But economics deals prin
cipally with kinds or classes of things, which contain a
larger or smaller number of individuals and admit of
more or less. Now, we have seen that the more we pos
sess of a kind of thing, the less is our desire for it, or, we
may say, the less is our liking for it. Our liking for a
thing, in fact, decreases every time we use it, the longer

FOUR KINDS OF ECONOMIC VALUE

we use it, only to be reinvigorated after rest. But this
is another matter. The case of labor is precisely the
opposite. The longer we have to work at a stretch for
a thing, the more our dislike of the labor grows. Our
desire for a thing may sink to nothing, with satiety.
Our dislike of labor may rise till it becomes unbearable
and we must stop. So in the case of our daily toil. A
poor people, — poor because of barren soil from which
they must eke out their sustenance, or poor because of
inefficient labor, or the robbery of neighboring peoples,
or of a military caste amongst themselves, — because
of their greater need, must work more hours than a rich
people, who in their greater affluence have less desire
for more, and in whose case, therefore, the rising dislike
of labor meets the falling desire at a lower point.
In this sort of comparison, however, we are here not
so much interested as in another. The producers of
an article generally, because of different situations or
different skill, produce it at different costs of labor.
Just as the consumers put their acquisitions first to the
greatest use, so the producers first choose the places
where their labor will be the least. The same producer
may also produce different portions of his produce at
different labor-costs. For him the cost-value of his
whole product may be reckoned as the average cost.
But to the people at large, who get it from producers
other than themselves, the cost-value is according to
the cost of the portion ordinarily produced at the
greatest cost, because the portions produced with less
labor the producers are able to estimate, in their trades
with the consumers, at the cost-value of what is pro
duced and disposed of at the greatest cost; wherefore
the people at large must estimate cost-value at the

FOUR KINDS OF ECONOMIC VALUE

greatest cost at which they reward producers for pro
ducing the thing. Thus, just as esteem-value is de
termined by the last and least use, so cost-value is
determined by the last and greatest cost. If the esteemvalue, falling with increased quantity possessed, and
the cost-value, rising with increased quantity produced,
come close together, the process must stop; for costvalue to the community, just as to the individual, can
not be so great as esteem-value.
In the case of an exchange the relation stated is ob
vious; for unless each party gets something he desires
more than what he gives, he has no motive to make the
exchange. And accordingly in an exchange the esteemvalue of what each party gets is to him greater than the
esteem-value is to him of what he gives. But the two
articles exchanged have equal exchange-value; for this
is the essence of the idea of exchange-value. Unequal
esteem-values precede an exchange; equal exchangevalues are the consequence. Exchanges, however, of
similar articles of one kind for similar articles of an
other kind are of frequent occurrence, and so, within a
region of more or less extent called a market, and during
a period of more or less length, a prevalent rate of ex
change becomes established, and it is this rate which
determines equal exchange-values. When in an indi
vidual exchange the one party takes advantage of the
ignorance of the other party and gets more than the
prevalent rate warrants, he is said to “ get the better”
of the other, in “ an unfair bargain,” and the exchangevalues of the articles exchanged are said to be unequal,
because such an exchange does not fix the prevalent
rate. But when both parties, equally alert, depart from
it, this is a sign that it is no longer prevailing.

FOUR KINDS OF ECONOMIC VALUE

It is exchanges that bring cost-values into confronta
tion with esteem-values. For the producers of an article
generally produce much more of it than they themselves
consume — sometimes they consume none of it. To
such a one the esteem-value of his product may be said
to be rather the esteem-value of the things he believes
he can get for it, and, as before said, this esteem-value
is greater than its cost-value. To the consumer, on the
other hand, the esteem-value of what he gets from the
producer is always lower than what would be the costvalue to him (the potential cost-value, it may be called),
were he to produce the thing himself; which is the rea
son why he does not produce it himself, because he
receives more potential cost-value than is the actual
cost-value of what he gives, if he be the producer of
this. Here is merely another aspect of the fact that
wealth is increased by exchanges. When two producers
exchange their products directly, there is no reason why
they should always exchange them so that each gives
and receives what costs the same amount of labor. This
would mean that the exchange should be of equal actual
cost-values. But this would leave out of sight, not only
the potential cost-values, but also the esteem-values,
which may be different; and it is precisely for the sake
of the esteem-values, and of the use-values back of
them, that the exchange is made. If one of the parties
complains that to get the same esteem-value he has to
give more cost-value, he is at liberty to produce the
other thing himself — if he can. If his inability is
caused by the action of the other party, he has a right
to complain of that, and to join with others to correct it.
But if the inability is in himself, in his want of capacity,
he may complain of Nature or of Nature’s God; but

FOUR KINDS OF ECONOMIC VALUE

that will do him no good. That Adam Smith could have
taught such a doctrine with respect to primitive peoples
and that Ricardo could have extended it to civilized
peoples, must be attributed to their overlooking esteemvalue. They both knew that things are not exchanged
according to their general usefulness, and not knowing
the difference between total and final utility, between
use-value and esteem-value, the only thing left them
that could serve as a common measure was labor.

IX
THEIR VARIABILITY
B e in g relative, all value is variable. And, on compar
ing them, we find that the kinds vary differently.
Use-value is the most stable of the values, but it, too,
varies. It varies as the utility of the thing varies; and
the utility of a thing varies as its physical nature
changes — for instance, as an apple ripens and rots.
Apart from this, the physical nature being supposed
the same, utility may vary because persons change, as,
when a person is sick, or a people is afflicted with a
plague, the utilities of things vary from what they were
to them in a normal state of health. Or utility may vary
in different situations, especially in different climates,
as coal has more utility in a cold region than in a hot one,
more in winter than in summer. Again, all these things
being supposed the same, the utility, and consequently
the use-value, still may vary, but now more slowly. In
one sense, it might be said that the use-value of things
always is the same, since they admit always of being put
to the same uses. This means only that the potential
utility of things is constant. But their actual utility,
which is what their use-value follows, varies with the
actual uses to which they are put. A use must first be
known before advantage can be taken of it. For in
stance, the greatest utility of petroleum has been a
recent discovery. Also it may be dependent on the
existence of something else whose utility must likewise
first be known, as kerosene could not be used as an

FOUR KINDS OF ECONOMIC VALUE

illuminant without glass-chimneys. Thus use-value
increases with applied science.
In still another way actual utility, or use-value, may
vary, and this variation is very different from the varia
tion of the other kinds of value, use-value here standing
by itself. The actual uses to which things can be put
expand with the increased quantity of them in our pos
session, and contract with their decreased quantity.
This is the case at least with everything that has more
than one use, and all staple articles are such, only some
contrivances invented for a special purpose having a
single use, such as mouse-traps. Thus, in the case of all
the materials needed for the production of finished
articles, their use-value increases or decreases directly
with the increase or decrease of their quantity. Take,
for example, gold. We often hear it said that gold is not
a very useful metal because, besides its use as money,
it is used only in jewelry, in watch-cases, for gilding, and
for filling teeth. This is owing to the small quantity of
it in our possession. Were gold to become as plentiful
as iron, it could, and would, be used for the roofs of
houses, for the plating of the sides of wooden ships, for
ballast, for canning, and, with some alloys, for all the
household utensils and fixtures for which copper, brass,
tin, and nickel now are used. We should find it to be one
of the most useful metals; and so its use-value would be
much greater than it is now.
Not so its esteem-value. The more gold we should
possess, the less we should esteem a given portion of it,
the less we should value it, the less esteem-value it
would have, as would be shown by the lower uses to
which it would then be put. And similarly in the case
of all other things. Their esteem-value varies inversely

FOUR KINDS OF ECONOMIC VALUE

with their quantity. This is a simple fact of experience.1
And it varies differently with different things, and also
with different persons or peoples. The esteem-values
of necessaries vary most violently with inverse fluctua
tions in their quantities, as was better known in the
days when in regions of small extent those fluctuations
were much greater than they now are in our wide mar
kets; so that Gregory King was able to formulate an
empirical rule for the variations of the prices of wheat,
which prices, as we shall see, over short periods measure
variations of esteem-value. At the opposite extreme,
luxuries also vary much with slight variations of quan
tity, because of the comparatively small number of their
purchasers. Also, if with an increase of its quantity a
thing can be put to many uses, its esteem-value may
fall less than if it can be put to but few uses; for the ex
tension to the new uses, by diminishing the increase of
the quantity employed for the old uses, retards the fall
of its esteem-value. What is here said is important: it
shows that low uses are not the cause of the fall, but a
support of esteem-value, preventing its further fall till
continued increase of abundance overflows them. An
other reason for the differences is the presence or ab
sence of substitutes. If one kind of thing is easily re
placeable by others, its esteem-value may vary little
with its own quantity; but if it has few or no substitutes,
its esteem-value, depending more on its own quantity,
may vary much. A variation of a thing’s esteem-value
1 If the quantity of an ordinarily valuable thing becomes too great, so as
to interfere with the enjoyment of other things, the value of that thing for the
time being runs down below nothing into the opposite of value, or negative
value, becoming such that we desire to get rid of the superabundance of the
thing. This was found out in the case of gold by Midas. Many people have
experienced it of water, in times of floods.

FOUR KINDS OF ECONOMIC VALUE

may even be caused, without any change in its own
quantity, by changes in the quantities of its substitutes.
Moreover, in the case of articles that can be sub
stituted for one another, habit and fashion play a part.
Above, it has been said that our desire for things is ac
cording to their utilities. This is not entirely correct.
There is a still more subjective factor influencing our
desires. This is a preferential liking, hard to account
for, given to one thing rather than to another that may
perhaps serve the same purpose equally well. None but
scientists know whether wheat is more nutritious —
that is, more potentially useful — than rice; but in
occidental countries people prefer wheat, and in the
orient people prefer rice. This is an instance of habit,
which changes slowly. Fashion, which rests more on
the ornamental aspects of things, induces our liking for
different varieties of the same kind of things to change
quickly. Between the two, “ crazes” change more
quickly than the former, and less than the latter, such
as the “ bicycle craze” of several decades past and the
recent “ mahjong craze,” which cause the articles neces
sary for their exercise to be held in greater esteem, and
to receive greater esteem-value, while the craze lasts,
than at other times.2 It might be said that this subjec
tive element in esteem-value is nothing else than the
subjective reflex of utility, since our preference for one
2 Comparable to such a change of habit, fashion, or “ craze,” was the
demonetization of silver which took place so extensively about the year 1873.
This, by legal enactment putting into effect the desire of financiers, took the
use as money away from silver and put it all on gold. More recently carriages
have lost their use as vehicles, which has been put upon automobiles. Car
riages have hardly any other use, and so their esteem-value has dropped to
almost nothing, followed by their exchange-value. But silver has many other
uses, and therefore it has retained much of its esteem-value, and also much
of its exchange-value.

FOUR KINDS OF ECONOMIC VALUE

article leads us to esteem it more, so that while we do
so it has more actual utility. Still it would be our pref
erence that causes the greater actual utility, and not
the reverse. Now the more fully expressed law of the
esteem-value of things is that it varies directly with our
likes and inversely with their quantities. As for all
things taken collectively, as possessed and used by a
whole people, they of course have no substitutes; nor
can our liking for them, our desiring them, our need of
them, be said normally to vary: in them there is no
change of fashion. Hence, if our esteem for all things,
or the esteem-value of all things, does not vary, — and
certainly it does not increase to the extent of their in
crease, nor decrease to the extent of their decrease, •—
then the esteem-values of them individually must vary
in some inverse proportion to the variation of their total
quantity. What this proportion may be, will be ex
amined soon.
Cost-value has been seen to be in many respects the
opposite of esteem-value, along parallel lines. As a
thing is differently liked by different persons, so labor is
more disagreeable to some than to others. In fact, to
some persons labor is pleasant, at least in some occu
pations, and for a longer or shorter time after the com
mencement of the day’s work, only growing irksome
more and more during the remainder. Some persons
are industrious, and some indolent. There are habits
and fashions also with regard to labor, rendering some
kinds reputable and others disreputable, differently
among different peoples, and variably in the same.
Cost, we have seen, is what we give up for a thing. In
production what we give up is really something back of
labor — our ease, or the pleasures we might be enjoy
ing if we were not tied down to an arduous task. This

FOUR KINDS OF ECONOMIC VALUE

is what makes it difficult to wean a hunting or pastoral
people from their easy and enjoyable occupations to an
industrious agricultural or manufacturing life. Yet
when people are once accustomed to labor, it becomes
less tedious, and with the consciousness of acquired
skill they come even to find pleasure in it for a time,
while the more abundant products they have to con
sume heighten the enjoyment of their remaining leisure.
This, on the subjective side, is a reason why, as a people
becomes more industrious, the disutility of labor sinks
to them, the objective side being that their labor, be
come more efficient, produces more. There is a double
gain, more pleasure being acquired at a smaller cost of
displeasure. But the subjective aspect of labor must
generally be neglected in economics, except as it affects
the supply of labor in the so-called labor market. It
occasions no error, and it is sufficient, to speak simply
of labor as the cost of the things produced.
Most producers cooperate to produce one kind of
article under an employer and director. The employer
must consider not only his own labor, but that of his
employees, about which he knows nothing of their sub
jective elements, but only the objective elements of the
time they devote to the work and the efficiency of their
labor. A product of labor, therefore, has its cost-value
estimated by the time-quantity and the efficiency-qual
ity of the labor expended on it. Now, in many cases it
is found that, up to a certain limit, when more labor
is employed together, the quantity produced increases
faster than the quantity of labor, so that the cost-value
of a given amount is lowered. Thus, just as the increased
quantity of the thing produced has for its consequence
a fall of its esteem-value, the larger application of labor
to produce it may have for effect a fall of the individual

FOUR KINDS OF ECONOMIC VALUE

thing’s cost-value. The fall of esteem-value will, how
ever, overtake the other fall, and set a term to the
process. This is the case mostly with goods manufac
tured out of raw materials. The increased production
of bulky articles may lead to a rise of the cost-value
of their materials. For materials are mostly subject to
another rule, that of diminishing returns. They are
drawn from the earth, and the more that is drawn, gen
erally the more difficult becomes the drawing. Mines
begin on or near the surface, and have to be dug deeper
and deeper. Agriculture finds lands of different fer
tility, and applies itself first to those of greatest fer
tility. Later, therefore, the cost-values rise. A reverse
process may set in, when people find distant sources
more easily worked, and enough so as to surpass the
increased cost of transportation due to greater distance.
Decrease in the cost of transportation also helps here.
Machinery may be applied also to mines and to agri
culture. Growing population thus has opposite effects,
both to raise and to lower cost-values; and sometimes
the one tendency has prevailed, and sometimes the
other. For the last four centuries there has been cheap
ening. The coming centuries may witness a change.
Variations of exchange-value are of constant occur
rence. Their frequency is doubled by the fact that
every rise of one is a fall of another, and conversely. As
they cannot all move in the same direction together, as
can the preceding, our treatment of them must be very
different. Two great theories have been invented to
explain them, which would make them proportional to
variations of cost-value or to variations of es teem-value.
These theories presuppose measurement and considera
tions of causation, and so must be deferred.

X
OUR ATTITUDE TOWARD THEIR VARIATIONS
Otjr attitude toward the variations of the four kinds
of value is also very different in the case of all but two.
Toward use-value we have one attitude: we wish it to
be as great as possible, or in the course of time to in
crease. Toward esteem-value and cost-value we have
another attitude, common to both, and exactly the
opposite of the preceding: we wish them to be as low as
possible, or in the course of time to fall. Toward exchange-value we cannot rationally hold either of these
attitudes.
Value may increase in two ways: extensively or
laterally, and intensively or vertically. All the four
kinds may increase in the first way, as when we possess
more things, we may have more of all the values. It
sometimes happens, however, that after one has a cer
tain amount of a thing, as he acquires more of it, its
esteem-value falls faster than its lateral increase, so
that the total esteem-value of the thing may be to him
even less than when he had less of it. This can happen
even with the exchange-value of a particular article, if
we are all thus enriched by its unusual abundance and
are not at the same time equally blessed by a corre
sponding abundance of most other things. Farmers and
manufacturers sometimes find themselves impoverished
by over-production. Still, this is desirable for all of us
except the producers of the particular articles in the
infrequent cases when their exchange-values so behave.
At all events, we all wish for better things, that is, for

FOUR KINDS OF ECONOMIC VALUE

higher use-values. Also we all wish for more of most
things, that is, for lower esteem-values as the direct
consequence of their greater abundance. And we wish
to get them through less labor: we wish their costvalues to fall. Certainly the producers of every article
entertain this wish, and the rest of us wish the same
so far as they allow us to participate in their good for
tune.
The opposite attitude toward use-value and esteemvalue is another clear distinction between them, which
ought to keep all economists from confounding them.
Unfortunately, some economists, even since Jevons
expounded the distinction between total utility, which
applies to use-value, and final utility, which applies to
esteem-value, have thought of these two as utilityvalues, and so have renewed the old confusion. Even
the Austrian economists do not keep the two concep
tions as sharply distinct as they ought to be kept.1
1 The trouble occasioned by not clearly observing the distinction may be
illustrated by Bohm-Bawerk’s famous ‘ ‘casuistical ” hypothesis (or ‘ ‘thoughtexperiment”) of a Robinson Crusoe with five sacks of corn. Op. cit., book
in, chap. 3. Each sack is valued at 1 (the marginal utility of feeding par
rots), so that the total is 5. But if he had only three sacks, each would be
valued at 6 (the marginal utility of feeding chickens), so that together they
would be worth 18, and it would seem to be to his advantage to throw
away two sacks in order to increase the value of the corn in his possession.
This reductio ad absurdum has been urged against his theory, for instance, by
Cornflissen {op cit, 2d ed., p. 66), who does not distinguish between usevalue and esteem-value. Bohm-Bawerk himself, not clearly enunciating the
distinction, was conscious of the trouble, and tried to obviate it by denying
that the marginal values of the five sacks, each at 1, should be added. In
stead, he said that the sum should be made of the marginal value of the single
fifth sack at 1, plus the marginal value of the fourth sack at 4, and so on.
But just as the marginal value of the fifth sack is estimated at 1 on the
supposition that the five are possessed, so the marginal value of the fourth
is estimated at 4 on the supposition that only four are possessed, and so on.

FOUR KINDS OF ECONOMIC VALUE

As for exchange-value, our attitude toward this like
wise stands by itself. To be sure, the extension of the
exchange-values of all our possessions, the total valua
tion of our property, we all wish to be increased, except
in the way just described. But the heights of the ex
change-values of all things — the exchange-values of the
units in which their quantities are measured — must be
treated differently. We have seen that the use-values of
all things can rise together, and likewise the esteemvalues and the cost-values of all things can fall together,
and these are the variations which we desire. But we
have seen that the exchange-values of all things cannot
rise or fall together. Hence we cannot rationally desire
any change on the average, apart from individual ups
and downs, in the exchange-values, of all things to
gether.
What happens is this. Every article has its special
producers, and the producers of every article wish its
All these are incompatible suppositions, whose consequences hardly admit of
forming one sum. But with the distinction between use-value and esteemvalue there is no trouble. The use-value of the five sacks is greater than the
use-value of the three, each new use, though smaller, being properly added to
the preceding; and it is only the esteem-value of the three sacks, on the sup
position that he had lost two, that would have been greater than the esteemvalue of the five (and would have been shown to be so by his guarding the
remaining ones more carefully). But it is precisely use-value which this
Crusoe desires to be as great as possible, and it is precisely esteem-value
which he desires to be as low as possible. Hence it is in two ways to his ad
vantage to preserve all his five sacks. But if he were surrounded by other
men who produced many things but had no corn, while each of his sacks was
a warehouse full, then it might actually be to his interest to destroy two fifths
of his corn, as he might get in exchange for the remainder more of the other
things than he could get for the whole. Many monopolists have done this
very thing. The gain, however, would accrue to him alone; and to the com
munity at large it would be a loss. For wealth consists in use-values. Cour
not, in the first chapter of his Recherches mathématiques, made the mistake of
saying that wealth can be increased by such vandalism.

FOUR KINDS OF ECONOMIC VALUE

exchange-value to rise (whether by a rise of its own
esteem-value above those of other things, or by a fall of
the esteem-values of the others below its own); but all
the consumers of all articles wish their exchange-values
to fall relatively to what they have to give in exchange
for them. Thus especially persons with fixed moneyincomes wish the exchange-value of money to rise, but
each of the producers of what these persons consume
wishes the exchange-value of money with respect to his
own product to fall. And land-owners are happy if
people settle in their neighborhood and raise the ex
change-value of their land, while those who are about to
buy or to lease want its exchange-value to be low. All
these conflicting desires neutralize and destroy one an
other. An impartial onlooker — and such the econo
mist, such the statesman ought to be — cannot share in
all these special wishes, as he ought not to sympathize
with either of the opposite sides alone. His only wish
should be that every producer may produce equal quan
tities with less labor, and especially that he may pro
duce more with equal labor, and that every consumer
may have more to consume — that cost-values and
esteem-values of commodities may fall. Exchangevalues of commodities individually should not interest
him, since that would show favoritism. And exchangevalues as a whole take care of themselves. Only the ex
change-value of money needs looking after, as we shall
see later on.
Here, too, exchange-value differentiates itself from
esteem-value and from cost-value. There are some
points of contact between it and them, as we have seen
and shall see; and these explain, though they do not
justify, the persistent identification, or at least non

FOUR KINDS OF ECONOMIC VALUE

differentiation, of it with them even in the respects in
which they are different, on the part of so many econo
mists. But the distinction is perfectly clear, and is easy
to perceive when once pointed out. It too, therefore,
ought to be firmly held.

XI
THEIR KNOWABLENESS
B e in g variable, the kinds of value should, as far as
feasible, or as far as is needed, be measured. But before
taking up this subject, let us pause a moment to consider
one more characteristic wherein values differ, which is
their knowableness.
The uses of things it is the province of science to dis
cover, and no one is capable of knowing them all. But
as use-value follows only the known and actual uses of
things, use-values are completely knowable, and any
one can learn all he wishes about them on consulting
experts or searching through books on the arts and
applied sciences.
Personal esteem-values, or the esteem-values of
things to each individual person, are known to each one
within the limits he cares to become conscious of them.
The upper limits, however, are somewhat hazy; for
every one has but an ill-defined idea as to how highly
he prizes his various possessions or the things he wants.
It is the lower limits of esteem-values that become clear
as a person compares different objects with one another
by the lower uses to which he is able to put them. Yet
everyone in his bargains with others tries to keep secret
the height to which his desire for the article might rise
if he did not get it; and so the height of the esteemvalues of things to others is not completely knowable.
They are to a certain extent, however, ascertainable in
exchanges effected, especially as after an exchange
people are often willing to tell their friends how good a

FOUR KINDS OF ECONOMIC VALUE

bargain they have made. By means of exchanges, also,
to the people at large, the comparative esteem-values
that things have to them become apparent, as will be
shown more plainly when we examine the causes of
exchange-values. Thus, on the whole, esteem-values
are fairly well knowable.
The least knowable are cost-values. For cost-value
follows cost just as use-value follows utility and esteemvalue follows desire; but while everybody knows the
utility of everything he needs and his desire for it, only
the producers know the costs of the articles they pro
duce, and they are by no means willing always to dis
close them correctly. Thus the cost-value of each article
is knowable in the first place only to a few, and these
hide them. These few themselves hardly know how
easily they produce some portions of their product: it is
the highest cost of the whole product that interests
them. And the lowness of this highest cost, or the low
ness of cost-value, the producers try to conceal, or they
try to make it out higher than it is. They exaggerate it
for at least three reasons. The first is, in order to make
better bargains in their exchanges for other things. The
second is, to prevent others from learning their profits,
and thereby to keep competitors out of their line of
production. The third is, to get “ protection” from
their government by representing that their costs are
higher than the costs of the same articles in foreign
countries. It is, thus, the lowness of cost-values which
it is difficult for the people at large to know. Con
sumers, we have just seen, try to conceal the height of
their esteem-values from the producers; and now we
see that the producers try to conceal the lowness of their
cost-values from the consumers. But in this game the

FOUR KINDS OF ECONOMIC VALUE

producers have the immense advantage of being few in
the case of each article, and consequently of being able
to act in concert, while the consumers are many, are
scattered, and are little able to combine. Producers,
too, at least the managers, generally are consumers of
what they produce, and can easily estimate the esteemvalue in which others are likely to hold their products;
but consumers are not producers of most of the things
they want, and have no means of estimating correctly
their costs. Thus obscurity is one great peculiarity of
cost-value, separating it off from all the other kinds.
Exactly the opposite is the case with exchange-value:
its amount and its variations are the most apparent of
all. In fact, exchange-values exist only as they appear:
they are the most phenomenal of the kinds of value.
Moreover, their variations, of daily occurrence, are
reported from mouth to mouth, and quoted in the
newspapers. Everybody knows those in which he is
interested, though he may know little or nothing about
the cost-values of the articles, and may have only an
unprecise knowledge of their esteem-values even to him
self. Exchange-values are the most knowable of all the
kinds of value.

XII
THE MEASUREMENT OF THE FIRST THREE
We now come to the measurements of the kinds of
value. Being differently variable, they must be meas
ured differently.
Use-value, however, the lowest of the economic
values, hardly seems to be measurable with any pre
cision. Being so close to mere usefulness, its measure
ment, if possible in some cases, would belong rather to
physics and even to physiology. The various uses of
the same thing may be compared with one another
roughly; and also the total utility of one thing may be
compared with the total utilities of other things, and
their relative importance be estimated in round figures.
Nothing more than this seems to be needed. In the
case of money, its use-value resides in its services as
medium of exchange, as measure of comparative ex
change-values at the same time and place, and as store
of “ value” from one time or place to other times or
places (of which kind, must be examined later) ; and all
that we need say about it is that the more and the better
it serves these purposes, the greater is its use-value.
When the people of a country are passing from a régime
of barter to a régime of commercial exchanges on a
money basis, while this process is going on, money
flowing in, or being produced, and gradually filling, as is
said, the channels of trade, then its use-value is increas
ing. But when this process is completed, its use-value
may be said to be stationary. Again, however, when
substitutes are being introduced, such as bills of ex-

FOUR KINDS OF ECONOMIC VALUE

change, bank-notes, checks (transferring deposits), and
government notes promising to be received in place of
money and to be redeemed in money, then the original
money may be said to lose some of its use-value, which
is taken over by these substitutes; but if these all to
gether be regarded as money, — in Walker’s phrase,
that “ money is that money does,” — then the use-value
of money still is stationary (only its cost-value has
sunk). A smaller supply of money at higher “ value”
may, within certain limits, serve the same purpose and
have the same use-value as a larger supply at lower
“ value.” At the same “ value,” obviously more money
is needed if the country is growing in population and
prosperity. But here we are running ahead of the right
order of development of our theme.
Next to use-value comes esteem-value. This, we
have seen, varies directly with our likes or preferences,
and in some inverse proportion to the thing’s quantity,
differently of different things and for different persons.
Theoretically, the variations may be plotted out diagrammatically. But practically little more is known
than that the lines are curves sloping in certain direc
tions: they cannot be calculated individually. Luckily
we have little need of precision in individual cases.
What we most want to know is the variation of the
esteem-value of things in general, not only to individual
persons, but — and especially — to a people at large.
Here the problem is simplified by the fact that an in
dividual’s and a people’s liking or desire for all things in
general is not affected by modes and crazes; 1 wherefore
1 Except possibly religious crazes that induce some to despise property
and to seek poverty. Prospective wealth in heaven then takes the place of
wealth on earth. But this is outside of economics.

FOUR KINDS OF ECONOMIC VALUE

this factor in determining esteem-value may be elim
inated, quantity alone remaining.
In the case of individuals, a theory has long been
entertained that the whole of everybody’s property is
always of the same “ value,” -— or as Franklin put it,
a poor man’s all is as dear to him as is the all of a rich
man to him, —■ which can refer only to esteem-value.2
If by property be meant capital, it should obviously in
clude the capitalized value of one’s earnings, or rather
of one’s earning power. The theory, however, is more
simply applied to a person’s consumption of goods
during a given period, say a year, including the services
of the permanent goods which he uses as well as those
things which he uses up or destroys in using them. An
emendation was suggested by Daniel Bernoulli to the
effect that a certain minimum amount, just sufficient
to keep one alive without supplying pleasures, should
first be subtracted, and only the surplus be considered.
This would make the rise or fall of esteem-value slightly
greater than the decrease or increase of the whole for
tune. But in wealthy countries, where fortunes are far
above this minimum, the difference is small enough to
be neglected. Now, the variation of a person’s consump
tion of goods from one period to another can be best
measured by calculating his money-expenditures during
each of the periods compared, on the assumption that
the exchange-value or purchasing power of money has
been measured and found stable, or, if varying, that its
variation is allowed for. As for the variation of the
2 This theory is not invalidated by the fact that, temperaments being
different, some are content with little and others, possessing much, yearn and
strive for still more. The loss of half their property would be felt more keenly
by the latter than by the former. Still, the proportion to the whole would be
the same in both cases. The relativity of all value allows for these differences.

FOUR KINDS OF ECONOMIC VALUE

esteem-value of a person’s money itself, money may be
said to represent all the things it is exchanged for; and
like them, money is always in fashion, so that its esteemvalue is affected only by the quantity of it one has to
spend. It is important, then, to measure the variation
of the esteem-value of money as well as its exchangevalue.
And this is important especially in the case of the
esteem-value that is put upon money by a people at
large. For the theory just propounded applies to a
whole people as well as to individuals. A whole people
always sets the same esteem-value upon its whole an
nual consumption of goods, and consequently upon its
whole annual money-expenditures. As for a people’s real
annual consumption, which corresponds fairly well with
its annual production (plus the algebraic balance of im
ports and exports), this can be estimated by averaging
the ascertained variations of the amounts of all the
principal goods produced and consumed during the
periods compared. Professor Day and Professor Per
sons have recently been showing how this can be done.
In the case of money, it may be better to substitute
money-income for money-expenditures. The moneyincome of an individual can differ from his expenditures
during any period, on the one hand by his trenching
upon his capital or by borrowing, and on the other by
his paying off debts or by saving and adding to his
capital. But by the law of large numbers these differ
ences may be considered to be constant figures for all
the individuals in a country taken collectively; and
money-incomes are more easily estimated than moneyexpenditures. Now then, to measure the variation of
the esteem-value of money, all that needs to be done is

FOUR KINDS OF ECONOMIC VALUE

to calculate during different periods the average moneyincomes of a people. The average money-income, be it
said, because after the whole is calculated, allowance
must be made for increase or decrease of population,
and the making of this allowance is done by averaging.
In this case, contrary to the usual practice, it would
seem to be advisable to reckon only the adult popula
tion, as money has no esteem-value to babes, and to
other young children only as it is given to them. At
most, only self-supporting minors should be counted.
This question need not be decided here, the principal
purpose of this paper being to be suggestive. The prob
lem of calculating average incomes is one for statis
ticians. They have thus far found many obstacles in
their way, but none that are insuperable.3 This subject
is now attracting more attention, and the National
Bureau of Economic Research has already made some
progress in this direction. When both this measure
ment of the variation of the esteem-value of money and
the measurement of the variation of the exchange-value
of money shall have been carried out, the combination
of these two, as in the case of individuals above noticed,
will yield the variation of the people’s real income and
consequently the variation of the esteem-value of any
given amount of average goods consumed. More will
be said about this later.
As for cost-value, it is obviously determined by cost,
and the measurement of cost is a business affair. The
managers of every productive concern must needs try
3 Nothing would help them better than a strictly economic income tax,
levied on total net incomes (above a small minimum), from whatever sources
derived. Unfortunately our federal income tax, with its many exemptions
(even excepting the salaries of government officials!), does not give this full
aid.

FOUR KINDS OF ECONOMIC VALUE

to measure the cost, and the variations in the cost, of
producing their products, and they can generally do so
fairly accurately. They can calculate both moneycosts and labor-costs, their interest being chiefly in the
former, while economists are more interested in the lat
ter. Statisticians, by taking the results of such accounts
as far as they can get them, — and get them truthfully
reported, — can average their variations and thereby
make an approach to a more or less veritable measure
ment of the average cost of things, and so of their gen
eral cost-value. For the latter purpose, it may be ques
tioned whether the costs in the factories and on the lands
that are kept working at the margin of profit and with
out yielding rent, are the only ones to be reckoned.
Ricardo, to whom “ value” meant principally costvalue, denied that there could be any measure of this
value. He meant merely that no commodity could be
expected to be found that would always be invariable
in the cost of producing it. But this is not necessary for
the purpose. If the variation of the cost of producing
the usual unit (whether of weight or of volume) of any
commodity were ascertained at different periods, then
an inverse portion or excess of it would be known to
have stable cost-value over those periods. The calcula
tion of labor-costs and the averaging of them have
generally attracted more attention during eras of falling
prices, such as 1810-50 and 1873-96. The first of these
saw in England the labors of Tooke, which were con
tinued by Newmarch even beyond it; the second, the
efforts of Wells and Atkinson in America. Many prac
tical difficulties have been found to stand in the way,
which were pointed out in Carroll D. Wright’s Sixth
Annual Report of the Commissioner of Labor, 1891.

FOUR KINDS OF ECONOMIC VALUE

But again these might not be insuperable if there were
inducement enough to overcome them, although this is
doomed to be the least accurate of the measurements.
The examination into costs is now undertaken mostly
with a view to tariff regulations, because the applicants
for protection generally make out their costs to be
higher here than abroad, and their statements need to
be checked. In the special case of money, it may be
added that, where money is identified with gold, its
cost-value is of interest really only to the producers of
gold. Where money consists of paper notes, its costvalue is trifling. This is said from the standpoint of the
governments producing the money. To the people at
large it is really the esteem-value and the exchangevalue of money that are important.
What anyone at the present day has to say about the
above measurements must be uttered with diffidence,
as these subjects have not yet been fully discussed and
examined. The measurement of the variations of ex
change-value, to which we now turn, has been more
fully investigated, and we can speak about it with more
assurance.

XIII
THE MEASUREMENT OF EXCHANGE-VALUE
T h e measurement of exchange-value is very different
from the measurements of the other kinds of value,
owing to the peculiarity of exchange-value in being
interrelative. An article has exchange-value in another
article according to the greater or less amount of the
other article given in exchange for it in the open mar
kets of the country, whether the exchange be immediate
or through the intermediation of money; and the varia
tion of its exchange-value is measured directly by the
amounts of the other article exchanged for it. That
other article has exchange-value in the first article, and
the variation of the second article’s exchange-value in
the first article is simply the reciprocal of the variation
of the first article’s exchange-value in it. The first
article, moreover, has exchange-value in still another
article, and in still another, and so through the list of
all articles, and in every case its exchange-value is ac
cording to the amounts of the things exchangeable for
it. Some of the old economists called these the article’s
“ nominal values,” because each of the other things is
named. But they may be better called its particular
exchange-values. Now, the different particular exchangevalues of any one article may vary differently. If an
article should vary in exchange-value in every other
thing equally, its variation in exchange-value in all
other things —•or the variation of its general exchange-

FOUR KINDS OF ECONOMIC VALUE

value (the term is Mill’s) 1 — would be easily measured,
as its general exchange-value would vary simply as all
its particular exchange-values have varied. This may
happen at any time for at least a short while; for if the
price of an article be marked up at some moment, we
may consider it probable that for some few moments no
other change of price takes place, and therefore during
this interval the article has fallen in exchange-value,
not only in money, but equally in all other things. But
if a thing varies in exchange-value in one other thing
alone, its other particular exchange-values remaining
the same, which happens during that short interval
when only one article other than it has changed in price,
then its general exchange-value evidently is only slightly
affected and, though it has varied in the same direction
somewhat, it has varied to a much less extent. If,
again, it rises in exchange-value in one thing and falls
in exchange-value in another thing, all its other particu
lar exchange-values remaining the same, — for ex
ample, while its pi’ice is steady and only the prices of
two articles have been changed oppositely, — then
evidently its general exchange-value is affected op
positely, and these opposite variations of its two particu
lar exchange-values may be equal and neutralize each
other, in which case its general exchange-value remains
unchanged; or these opposite variations, being unequal,
may not neutralize each other, and then its general ex
change-value is affected more by the preponderating
exchange-value, and it inclines to that side. As a rule,
1 Pol. Econ., Ill, i, § 3. The idea being new and ill-defined to him, Mill
spoke of its “ necessary indefiniteness” (xv, § 1). If by this he meant that it
necessarily is and always will be indefinite, he spoke hastily and with un
necessary pessimism.

FOUR KINDS OF ECONOMIC VALUE

between two periods every one thing’s many particular
exchange-values have varied variously, and the varia
tion, or constancy, of its general exchange-value is a
resultant of these many particular variations.
Some economists have maintained that an equality
of a thing’s exchange-value, by which they mean its
general exchange-value, at different times is inconceiv
able except only in the case that it always exchanges for
the same amount of every other thing; that is, that all
its particular exchange-values remain unchanged; which
would require that all the particular and general exchange-values of all the other things should remain
unchanged too, in what one of them has called “ a petri
fied economic world,” such as does not exist. This
opinion is easily refuted. These economists admit that
the general exchange-value of a thing may at a later
time be considerably higher or lower than it was before,
and they do not require for this that all its particular
exchange-values must rise or fall alike, but admit that
its particular exchange-values may vary differently,
demanding only that the general trend should ob
viously preponderate on the one side or the other. They
must admit, therefore, that the thing’s general exchangevalue may vary from a higher to a lower, or reversely,
without requiring that all its particular exchangevalues vary alike. But by the principle of continuity,
when anything varies from more to less, or reversely,
it must pass through — or be able to pass through
without being required to jump over — the point of
equality. Hence it must be possible for a thing’s gen
eral exchange-value at a later period to be equal to
what it was at an earlier, without the requirement that
all its particular exchange-values must be the same at

FOUR KINDS OF ECONOMIC VALUE

both. Equality at a later period must be conceivable
if variation is conceivable, and of the same kind. And
if conceivable, it must be admitted to be possible. And
if possible, it can be measured.
To measure the variation, or absence of variation,
that is, constancy or stableness, of an article’s general
exchange-value, it is evident that an average must be
drawn of the variations of all its particular exchangevalues. In this averaging, of course, the relative quan
tities of the articles that appear in the market must be
taken into account; for if twice as much of, say, wheat
enters into traffic as of barley (reckoned by their total
valuations), the wheat should be considered as forming
two classes of similarly named and equivalent articles,
compared with barley as one class. This allowance for
relative quantities is called “ weighting” ; and for brev
ity of speech it is said that all articles should be weighted
“ according to their importance.” 2 In every measure
ment two different periods (or different countries) are
compared, and therefore the relative quantities, or rela
tive importance, of the articles at both periods (or in
both countries) should be taken into account, that is,
the weighting of both should be used, as neither period
(or country) has better claim to consideration than the
other.
Now, we have no direct information about every
article’s variations in exchange-value in every other
article, for the simple reason that in our markets com2 There always must be weighting, right or wrong. It is common to speak
of averaging as “ unweighted,” or “ without weighting,” if all the articles are
counted alike. It should be obvious, however, that this is merely even or
equal weighting. The difficulty of getting correct weighting is not avoided by
this ostrich-like subterfuge of trying to make out that no weighting is being
employed.

FOUR KINDS OF ECONOMIC VALUE

modities are not exchanged directly for one another.
But we have such information about the variations of
the particular exchange-values of money in all other
things, which information is given in the prices of all the
articles listed in the country’s markets; for the varia
tions of prices are the reciprocals of the variations of the
particular exchange-values of money. Thus in the case
of money we have at hand the data necessary for the
measurement of the variations of its general exchangevalue.3 The problem is now posed, and requires for its
solution an inquiry into the proper kind of average to
use and the proper kind of weighting. Here we need
not undertake this inquiry. Mathematical economists
have already worked over this subject sufficiently to
have reached an adequately accurate method of meas
uring the variation or constancy of the general exchange-value of money on the supposition that prices
and quantities are given. These it is the office of sta
tisticians to collect. The accuracy of the result is deter
mined partly by the accuracy of the data collected
(especially of prices), and partly by the extent of the
collection; for it is beyond human capacity either to be
accurate in all details or to make the collection com
plete. The articles included in the research can only
serve as samples. When we speak of a variation of
money’s general exchange-value in all things (that is,
in all things besides money), we can strictly mean by
“ all” only all the things included in the collection; but
3 General exchange-value, as was pointed out in my The Measurement of
General Exchange-Value, pp. 13, 39-41, may be further distinguished into
general exchange-value in all other things and general exchange-value in all
things, including the article itself. When we measure the variation of money’s
general exchange-value by means of the opposite variations of prices, we are
measuring the variation of its general exchange-value in all other things.

FOUR KINDS OF ECONOMIC VALUE

our meaning may be inductively extended to refer to
really all things with an approximation that is the
greater, the greater the accuracy of the data and the
greater their number; wherefore as many prices should
be included as can possibly be collected with satisfactory
accuracy.4 When this has been done, the mathematical
theory of probabilities may be invoked to calculate
what is the probability of error in the result; for it is
well known that no observation, no measurement, made
by man, ever is absolutely correct. Yet, notwithstand
ing its deficiencies, the measurement of exchange-value
has been brought to a closer approximation to truth
than the measurement of any of the other kinds of
value.
4 Wages, however, which with a view to greater completeness have been
included by some economists, ought to be excluded, as constituting an en
tirely different category of economic objects and forming an entirely different
line of research. Wages may be averaged by themselves, and the result gives
the variation of the esteem-value of money to the wage-earning class of the
community. They belong more properly, along with salaries, profits, and
even unearned incomes, as data to be averaged all together in the calculation
of the variation of a whole people’s money-income, for the purpose, as we
have seen, of measuring the variation of the esteem-value of money to the
whole people. This is an entirely different affair. To take wages out of this
combination and to put them with the prices of commodities for measuring
the variation of the general exchange-value of money (which money laborers,
like everybody else, use to purchase commodities with), is to make a hodge
podge without meaning. It is a survival of Adam Smith’s treatment of politi
cal economy from the standpoint of the employer.

XIV
THE QUESTION OF CAUSES
H a v in g examined how the measurements of the four
kinds of value are to be made, the next subject to which
we should turn would properly be an inquiry as to what
uses can be made of the measurements. There are
several practical uses to which some of them, at least,
can be put. But first there is a theoretical question to
be disposed of, which, now that economists have made
some progress in perfecting the measurements, we are
in a position to undertake; so that, in a way, this is the
first use we are to make of these measurements. This is
the inquiry into the causes of the variations of the kinds
of value, or simply into the causes of the kinds of value.
It is, according to Whewell, “ an important maxim of
inductive science, that we must first attain the meas
ure and ascertain the laws of phenomena, before we
endeavor to discover their causes.” 1 Economists, un
fortunately, began by searching for the causes of val
ues, confounded simply as “ value,” before they had
correctly distinguished and measured them. In this
respect, however, they proceeded as all scientists have
done; for physics and the other natural sciences have
run through similar courses. In antiquity the Greek
physicists inquired into the causes of things first
of all, before they measured and learned accurately
1 Philosophy of the Inductive Sciences (1847), ii, 241. So, of a particular
case, Newton had written: “ We must learn, from the phenomena of Nature,
what bodies attract one another and what are the laws and properties of at
traction, before we can inquire the cause by which the attraction is per
formed.” Optics, query 31, in Newton’s Opera omnia (Horsley’s ed.), ii, 243.

FOUR KINDS OF ECONOMIC VALUE

what it was whose causes they were seeking. They
stopped at metaphysics, or mental analysis, instead of
using objective research. The Greek scientists at best
made attempts at measurement only in astronomy and
acoustics, which were the only sciences (apart from
mathematics, wherein there are no causes) that they
really developed, except for a small advance in me
chanics. Physics in general did not become an indepen
dent science until accurate measurements of forces
were made in the seventeenth century of our era, and
chemistry not till accurate measurements of the com
binations of substances were made in the eighteenth
and nineteenth. Even in physics only recently have
such measurements been completed in the departments
of electricity and magnetism, and only now have these
departments been established on a scientific footing.
In the future, when the measurements of values shall
have been thoroughly carried out in economics, econo
mists will probably look back at the pre-measurement
economists very much as physicists nowadays look
upon the ancient physicists, regarding them as having
merely historical interest.

XV
THE CAUSES OF THE FIRST THREE
W e have, then, to inquire into the causes of the different
kinds of value — of all but use-value; for use-value re
quires no investigation, as its cause obviously is the
utility, or rather the utilities, the total utility, of the
thing in question. As for the causes of utility, they are
physical and physiological, and therefore lie outside of
economics. With regard to the other kinds of value, I
cannot go into so much detail as the subject properly
calls for, since my intention is not to develop a theory of
my own, but to point out the simplification effected by
recognizing the distinction between the different kinds.
Here, for an historical reason, the order of treatment
may be changed, and cost-value be examined before
esteem-value.
Ricardo and his followers — they were the first in the
field — assert that the comparative values of commodi
ties are determined, in the case of those freely producible
each at about the same cost on the part of all its pro
ducers, by their comparative costs of production (or of
reproduction); and in the case of those which are pro
duced at various costs owing to natural limitation of the
most fertile and most easily worked sources whence
they are obtained, by the comparative highest costs
which it is necessary to employ in order to produce all
that is needed of each. If they assert this of compara
tive cost-values, they have a perfect right to do so; for
we know that a thing will not in the long run be pro
duced unless it “ pays” to produce it — unless its cost,

FOUR KINDS OF ECONOMIC VALUE

or the disadvantage of laboring to produce it, does
not exceed the utility or advantage we derive from
consuming it. Hence “ value,” which always presup
poses utility or advantage, must, in one sense at least,
be as great as is the cost or disadvantage of producing
the thing which bears it; and where there are several
different costs, although the producers at the lowest
costs get the surplus benefit, the people at large do not
obtain the thing from these favored producers for less
than they have to give for it to the producers at the
greatest cost from whom they are willing to purchase.
Thus, general usefulness being presupposed, as is cost,
so is this kind of value: cost, or labor-cost, or simply
labor, is a cause of cost-value. This statement is in
controvertible, being nothing more than an analytical,
explicative, or definitional proposition, as logicians say,
the element in value which is so caused being picked out
and named.
Jevons, Menger, Walras,1 and their followers assert
that the comparative values of things would be deter1 Gossen, who is often said to have preceded these, is not to be mentioned
here because he held with Ricardo that exchange-value is determined by la
bor-cost. Gossen preceded only in recognizing the decreasing enjoyment we
get from using things and the increasing pain we suffer from laboring to pro
duce them. He used “ value” only in the sense of use-value. He knew “ the
value (= utility) of the last atom,” as he expressed the last increment; but
he did not recognize this as esteem-value, to be extended over the whole class
of the thing, and instead assigned a different “ value” to each portion of a
class according to the use it is put to. Hence he asserted that an exchange
creates “ value,” and denied that things having equal “ price” (= exchangevalue) are of equal “ value.” Accordingly he maintained that to multiply the
total quantity of a thing by the prices of the portions exchanged does not
give the true estimate of the thing’s total “ value” ; which of course is true
of its total use-value, else water would have little use-value. In an exchange,
he held, the only things equal are the pains of the last moments of the labor
that produced the quantities exchanged.

66
FOUR KINDS OF ECONOMIC VALUE
mined by their comparative utilities, if anything had
but one degree of utility; but as hardly anything of this
sort exists, they are determined by the comparative
final or marginal utility of the last and least use to
which people are enabled, by the abundance of the
things, to put them, and down to the last and least
degree of enjoyment obtainable from that use. If they
assert this of comparative esteem-values, they have a
perfect right to do so; for we know that, when we pos
sess a small quantity of some kind of thing, we apply it
only to its greatest uses, and get enjoyment to the ex
tent that we can apply it to them, since our enjoyment
decreases with the extension of our using it, until it
sinks to the level of another use, to which, if its abun
dance suffices, we then put it, and so on with increas
ing abundance, until the least use is reached, and down
to its last degree of enjoyment, when satiety ensues,
and if the abundance, as the thing is consumed, is main
tained by Nature’s bounty, we desire no more. Only
air, and in some places water, reach this ultimate degree
to all people; all other good things, which we appro
priate because of their lesser abundance, stop at some
higher degree of the last use, and generally at some one
higher than the absolutely least. And when we possess
things sufficiently to put them to several uses, we do not
distinguish the portions which we put to the greater or
lesser uses, or from which we get greater or lesser utility,
but we esteem them all according to the least utility of
the least use to which we are able to put the last portion
which we habitually possess. Hence value in one sense
must be as low as the least possible utility of a thing
according to the extent of our possessing it, and it falls
or rises with the extension or contraction of our pos

FOUR KINDS OF ECONOMIC VALUE

sessions. Therefore we say that the esteem-value of a
thing is determined and caused by the last and least
utility we are able to get from it. This statement, if
confined to esteem-value, is likewise a merely analytical,
explicative, or definitional proposition, necessarily
true, as it expresses what we mean by the term. These
economists generally go one step further and say that
as this final utility itself is directly according to the gen
eral utility or usefulness of the thing and inversely ac
cording to its abundance, or, in other words, directly
according to its scarcity; and as effects increase or de
crease with the increase or decrease of their causes,
therefore esteem-value is caused, not only by the general
usefulness of a thing, but by its scarcity. This further
statement is not an analytical or definitional, but a
synthetic proposition; nevertheless the proof of it is so
apparent in the commonest experience, that it is almost
equally incontrovertible. Thus, general usefulness being
presupposed, as is scarcity, so is esteem-value: scarcity
is a cause of esteem-value.
Thus have been established two great truths: that,
apart from general usefulness, which is presupposed in
both cases, cost-value is caused by labor, and esteemvalue by scarcity. There is an important relation be
tween these two propositions. As more labor is devoted
to produce a thing, its abundance increases and its
esteem-value falls. This may be only a horizontal ex
tension of labor, the labor-cost of producing each unit
of the thing remaining the same. Then, though costvalue is unchanged, esteem-value falls; but the fall of
esteem-value has a limit, a thing’s esteem-value not
being able to fall below its cost-value, except acci
dentally, as the thing will then cease to be produced

68
FOUR KINDS OF ECONOMIC VALUE
till its esteem-value rises again. There may be another
cause of the fall of esteem-value. Improved labor may
in an equal time produce more units, and so reduce the
labor-cost of the thing: then the cost-value falls, and
esteem-value may also come down to a lower limit.
Here we have spoken of production first, as if it leads
the way. But production is only for consumption:
consumption, as the purpose of production, really pre
cedes. And it precedes in time also, both to the race
and to the individual. When men were beasts, or lived
in the garden of Eden, they consumed what they found.
Children also consume what is given them. And chil
dren discover that their marbles become less attractive
as they acquire more of them, and they increase the
esteem-value of their possessions by swapping their
superfluous marbles for other things, and so get an idea
also of exchange-value before they know anything about
costs. These two kinds of value are thus learned before
people apply themselves to produce things and thereby
learn further their cost-value.2 Now, consumption, by
diminishing the abundance of the things possessed,
tends to raise their esteem-value; and it is precisely the
function of labor, by restoring the abundance, to keep
esteem-value from rising, and even, if possible, by in
creasing the abundance, to bring esteem-value still
lower.
Really, then, esteem-value precedes cost-value, exist
ing first, independently of cost-value. It is because we
set esteem-value upon a thing that we desire to renew
its supply; or in the case of a new thing, it is because
2 Therefore, in general works on economies, consumption, and also ex
change, should logically be treated of before production, contrary to the
usual practice, which was introduced by Say.

69
we believe in advance that it will have esteem-value,
that we wish to inaugurate a supply. And continuing
to produce the thing, and applying more and more
labor, and more intensive labor if necessary, we give it
more cost-value up, possibly, till it meets esteem-value,
which meanwhile, by our producing more of the thing,
has been coming down. When the two meet, they stop.
It is, however, important to note that there is no neces
sity that the two should meet, since Nature may not
yield enough of it to our labor to bring esteem-value
down sufficiently, — also the producers themselves may
not desire it and may be in a position to prevent it.
Esteem-value may, in some cases at least, in some places,
and at some times, — and for long times, — remain
higher than cost-value. Cost-value, on the other hand,
can only for a short while go higher than esteem-value,
doing so only by accident, the producers quickly rectify
ing their mistake.
Now, as it is a function of labor to lower esteemvalue, labor cannot be a cause of esteem-value. Rather
it might be said that esteem-value is a cause of labor.
We have seen that the esteem-value economists went
one step further than the cost-value economists. They
went from the analytical proposition about esteemvalue being caused by final utility, to the synthetic
proposition about its being caused by scarcity. The
others did not go beyond their analytical proposition,
did not inquire into the cause of labor. Now, as the
esteem-value of a thing is a cause of our laboring to pro
duce it, it might be claimed that esteem-value is a cause
of cost-value. But this is a final cause. Confining our
stand to the level of efficient causes, we must admit
that labor is the first and last cause of cost-value with
FOUR KINDS OF ECONOMIC VALUE

r
70

FO U R K IN D S O F E C O N O M IC VALUE

which economists need be concerned, just as scarcity is
the one cause of esteem-value with which they need
be concerned — apart from general usefulness, be it
always understood, which is a concurrent cause in both
cases.

XVI
THE CAUSE OF EXCHANGE-VALUE — RICARDO’S
THEORY
T h e position has, then, been won that labor is not a
cause of esteem-value, and scarcity is not a cause — not
an efficient cause — of cost-value. It is evident that
neither is a cause of use-value. But how about ex
change-value? Both the cost-value economists and the
esteem-value economists extend their doctrine to ex
change-value. The former do not say that the exchangevalue of no product can be above its cost-value; they
say, in order to allow for profits, that if the exchangevalue of any one commodity is above its cost, then the
exchange-values of all commodities are above their
costs in an equal proportion. Exchange-values, for
them, are regulated by comparative costs, and so agree
with cost-values. The latter do not allow that anything
can have exchange-value below its lowest esteem-value;
for if anything, through the bounty of Nature, has no
exchange-value, its esteem-value also is nil, though its
use-value may be great. They assert that exchangevalues, being essentially relative, are, in a way to be
explained presently, regulated by comparative esteemvalues. These economists make their doctrine abso
lutely general, because everything that has exchangevalue has utility, and therefore a final utility, and there
fore again esteem-value. But the former have to admit
exceptions to their doctrine. By its very wording their
doctrine is confined to the products of labor. Hence

FOUR KINDS OF ECONOMIC VALUE

lands,1 mines, waterfalls, and other such limited and
appropriable gifts of Nature, which have no cost-value,
need to have their exchange-values explained in some
other way, by bringing them into connection with the
exchange-values of the things by labor produced from
them or with their aid. Also the unique products of
past labor, which cannot be reproduced, such as an
tiques, and even the art works of living masters, which
cannot be reproduced by others, must, it is admitted,
have their exchange-values explained in still another
way.
Thus the only dispute is about the cause of exchangevalue. Which of these doctrines is correct? It would
seem that the mere exposition of them, when carried out
clearly with the aid of sufficient terminology, should
easily decide between the two. The need of admitting
exceptions at the very outset casts doubt upon the first
theory. Consider the opening words of Ricardo on the
subject, at the beginning of his Principles. “ Possessing
utility, commodities derive their exchangeable value
from two sources: from their scarcity, and from the
quantity of labor required to produce them.” He there
upon divides commodities into two classes, which have
their exchange-values determined, the ones in the first
way, and the others in the second. Of the first, compris
ing antiques and the like (even including “ wines of a
1 In the plural, because land is not a single kind of thing with uniform
value, but is composed of plots with different soils and in different situations,
having different values to be accounted for. Locke and some other econo
mists, among them Carey, tried to explain the exchange-value of agricultural
lands by the labor bestowed upon them that has brought them into a suitable
state. This is equally contrary to fact. Still more obviously labor cannot
account for the exchange-value of city lots, where previous labor often stands
in the way of their present value.

FOUR KINDS OF ECONOMIC VALUE

peculiar quality”), he says “ no labor can increase the
quantity of such goods, and therefore their value can
not be lowered by an increased supply.” This means
that with regard to the commodities in the second class
labor can increase their quantity and lower their value.
Certainly what reduces is not the cause (or the source)
of what it reduces, since it presupposes the existence of
what it reduces. Thus the cause (or the source) of ex
change-value in the second class also cannot be labor;
and as he has given a perfectly satisfactory explanation
of the exchange-value of the commodities of the first
class (which, in germ, is that of the esteem-value theo
rists), and has no other, this must cover the second class
as well. Still, the things in the second class, which he
considers by far the most important (though the first
by no means deserve to be neglected), may be exchanged
according to the comparative quantities of them pro
duced by the same quantity of labor. These things
Ricardo treats of under two subdivisions — freely
manufactured articles, and raw materials obtained from
limited sources of different fertility. We may review
the second first, because it is the more comprehensive,
and the first differs from it, as we shall see, only in
degree.
Ricardo supposes a people occupying a country whose
land has several degree of fertility, that of the first
quality, or No. 1, yielding 100 quarters of corn, No. 2
yielding 90, No. 3 yielding 80, and so on, the same acre
age being intended, and “ with an equal employment of
capital and labor.” At first the people will cultivate
only the best land; but as the population increases, this
increase will eventually, he says, “ make it necessary
to cultivate No. 2.” Then, as the same labor produces

FOUR KINDS OF ECONOMIC VALUE

less quantity, equal quantity is produced by more labor;
hence he says, “ the exchangeable value of [the] raw
produce will rise, because more labor is required to pro
duce it.” And similarly, when No. 3 has to be resorted
to, and so on. Here, and in most of his expositions of
his theory (but not in all, as we shall see later on), the
implication is conveyed that it is the cultivators of the
inferior soil who first raise the price (or exchange-value)
of their produce because of their higher labor-cost, and
then the cultivators of the better soil follow suit. At
all events, this implication is necessary for the theory
that exchange-value follows, or is regulated by, cost of
production. Ricardo, however, nowhere makes this
implication explicit, and we need not attribute to him
cleverness in hiding it: he was himself deceived into
entertaining it. But the real state of the case evidently
is that, as the population grows in numbers and riches,
and needs more corn, com having become scarcer to
them individually, they set a higher esteem-value upon
it, and are willing to give, and do give, more of other
things for it, so that its exchange-value rises, and rises
before the next inferior soil is cultivated, its rise there
fore being due to some other cause than the greater
difficulty of production, which other cause is the one
first assigned by Ricardo himself, namely scarcity; and
it is only this prior rise, as well of exchange-value as of
esteem-value, that permits the inferior soil to be culti
vated and thereby gives opportunity for the rise of
cost-value.
Ricardo was here explaining the origin of rent (in
which he had been preceded by James Anderson); for
the land that will just yield the lowest acceptable profit
to an agriculturist’s capital and labor cannot pay rent;

FOUR KINDS OF ECONOMIC VALUE

and that which yields more can pay rent because its
owner, instead of cultivating it himself, may lease it to
another, who will put up with the least acceptable profit
and hand over to him the surplus, which is rent. This
theory of rent is likewise held by the esteem-value theo
rists, with the difference that they give the right se
quence of cause and effect, which Ricardo inverted.
When the exchange-value of corn rises through scarcity,
if there is other poorer land to cultivate, the cultivation
of this, by increasing the supply, keeps the exchangevalue of corn from rising as high as it would have done
without this further cultivation. Thus, though greater
labor is employed to produce the same amount of corn
from the next poorer soils, it, instead of being a cause of
the rise of the exchange-value of corn, is a cause retard
ing the rise. But it retards it only for a time; for, as the
population continues to grow, the cultivation of the
second quality of land may not be sufficient to prevent
a further rise, which will lead to the employment of
more capital and labor on the third quality, and so on,
the rise of the exchange-value of corn always preceding
the rise of its cost of production; until at last, con
ceivably, there is no more land left for cultivation, and
the rise of the exchange-value of corn may then con
tinue unabated (and all land will pay rent), corn now
being like “ the wine of superior quality” which Ri
cardo placed in the first class, because no more labor
can be applied to its production, increase its quantity,
and keep its exchange-value from rising further. The
only absolute limit will be the stoppage of the growth of
the population by an approach toward starvation.
What Ricardo said of lands and of corn, he said also
of mines and of minerals, since mines pay rent, and dif

FOUR KINDS OF ECONOMIC VALUE

ferent rent, and there are some mines that cannot pay
rent. According to him, it is “ the return for capital
from the poorest mine paying no rent” that regulates
“ the rent of all the other more productive mines,” and
it is the cost of extracting and bringing to market (“pro
ducing” covers this, too) the minerals from the no
rent-paying mine — the least prolific mine that it pays
to work — that determines, regulates, or governs the
value (the exchange-value) of the mineral. The true
relation, however, is that it is the existing value (ex
change-value as well as esteem-value) of the minerals,
caused by their existing scarcity (in addition to their
utility), that permits the working of the least produc
tive mines, one or many, that are just workable, and
does not permit the working of still less productive
mines.2
In all this, the influence of the extension of labor to
less prolific sources, in consequence of diminishing sup
ply of the article, is parallel, in the opposite direction,
with the influence of the extension of the employment
of the article to lesser uses in consequence of an increas
ing supply. The latter process, as we have seen, tends
2 One of Ricardo’s followers, A. Houston, in his Principles of Value in Ex
change (London, n. d. [about 1865]), pp. 92, 93, drew a distinction between
mines and agricultural lands, in that the latter are nearly continuous in their
progressive fertility, while the former are much more discrete. Agricultural
produce, therefore, almost always finds some land where cost-value is equal
to exchange-value; but in the case of mines there may be a great gap be
tween the least productive mine now worked and the next one discoverable,
so that considerable latitude is left for the exchange-value to rise above costvalue till the next high-cost mine comes into operation and lifts cost-value to
the level of exchange-value. This seems like a clear recognition that cost is
not a cause of exchange-value. Ricardo himself, of course, did not make this
distinction, since he treated, in the above example, also lands as widely
discrete.

FOUR KINDS OF ECONOMIC VALUE

to keep the thing’s esteem-value from falling, and so is
not a cause of that fall; and the former, as we now see,
tends to keep its esteem-value from rising, and so is not
a cause of that rise. They have these contrary effects
also upon the exchange-value of this thing compared
with the exchange-values of other things. It would be
as great a blunder to say the lower uses are a cause of
the fall of exchange-value, as to say the greater costs
are a cause of the rise of exchange-value. Yet the lower
uses, acting to keep up the exchange-value, are much
more causes of exchange-value than are the greater
costs, the only action of which is to keep exchangevalue down. For what causes something to be as great as
it is, is much more properly described as its cause than
is something which only keeps it from being greater.
As to the things in Ricardo’s first class, which com
prises manufactured goods, their raw materials being
assumed to be given, these in a regime of free competi
tion are supposed to be produced each at one uniform
cost and at one uniform rate of profit. As a matter of
fact, different factories produce the same article at dif
ferent costs, owing partly to differently advantageous
situations, partly to different equipment, methods, and
processes, themselves due to different skill on the part
of the managers. Those run at less cost yield a surplus
which is exactly comparable with the rent of the better
lands, the only difference being that it is not generally
so great or so easily perceived and recognized. With
regard to different kinds of products, the theory requires
that all the profits gained in their manufacture be at a
uniform rate, as otherwise the relative exchange-values
would not be according to the costs. This also is con
trary to fact, since entrepreneurs and workmen have

FOUR KINDS OF ECONOMIC VALUE

different aptitudes and preferences for different kinds of
work. There may also be different combinations of pro
ducers of the various articles, and different legal aids
and restrictions, and customary hindrances affecting
them differently.
Still, there may be a pure theory of production under
free competition, on the assumption of perfect fluidity.
On this assumption, a commodity can be produced with
equal cheapness by an indefinite number of producers,
and as consumers will always purchase at the lowest
price, all who undertake to produce the thing will have
to sell it at the lowest price yielding an acceptable profit,
which will be the same to all. “ There cannot be two
rates of profit in the same employment,” said Ricardo.3
Nor can profits be higher in some employments than in
others, because new-comers will rush into those with
greater profits, and by increasing the quantity produced
will bring down the prices obtainable and consequently
the profits. “ In one and the same country,” said Ri
cardo, “ profits are, generally speaking, always on the
same level.” 4 Wages, at least for the same kind of
labor, will be uniform for a similar reason. For con
tinued production, prices must always be above costs,
and what keeps prices to a uniform rate above costs is
the perfect fluidity supposed.
On the other hand, there may also be a pure theory of
monopoly, where fluidity is replaced by solidity — by
a barrier somehow raised, which keeps others out.
3 Page 112; similarly, p. 37. Paginal references are to McCulloch’s edition
of Ricardo’s Works, 1886.
4 Page 76. He added, however, “ They differ only as the employment of
capital may be more or less secure and agreeable.” It is not so between dif
ferent countries because of obstructions preventing the perfect fluidity sup
posed. The pure theory does not abstract these.

FOUR KINDS OF ECONOMIC VALUE

This theory cannot be conceived by the cost-value
theorists without leaving the ground they stand on: it
can, logically, be developed only by the esteem-value
theorists. Ricardo, though occasionally referring to
monopoly prices, knew nothing about the theory of
them. The theory, developed since his day, is that, of a
monopolized commodity, there is, at any one time and
place, a certain price which yields a maximum profit;
and this price is higher, and the quantity produced and
sold is less, than what the price and the quantity are
in the régime of perfectly free and fluid competition as
determined by the other pure theory, which is held in
common by the cost-value and the esteem-value theo
rists. In practice, as the managers of a monopoly are
not omniscient, it cannot be supposed that they hit
upon the exactly right price that would give them the
highest profit; but it is supposed that they approach
more or less closely to it by trial, with a tendency per
haps to keep prices too high. For of course the favored
producer, or combination of producers, who have a field
of production to themselves, aim at the highest profit
for their own benefit, as otherwise there would be no
motive for their acquiring the monopoly.
Now, in the business world of to-day, and probably
of all time, conditions fluctuate between those assumed
in these two theories. There is no perfect fluidity as is
supposed in the first, and there is no perfect solidity as
is supposed in the second. The consequence is that not
only profits range between a minimum acceptable to the
majority up to a maximum striven after by a few, but
also prices range between the lower limits determined
by comparative cost-values not exceeding the corre
sponding esteem-values, and the upper limits deter

FOUR KINDS OF ECONOMIC VALUE

mined by comparative esteem-values alone. All pro
ducers strive for the highest profits, and to make them
they have the two opposite means, of lowering their
costs as much as possible, and of raising their prices as
much as possible; which last they can do by keeping
others out — hence the secrecy about their costs and
profits, hence the pouncing upon new competitors, hence
appeals to the legislature, hence unfair methods in
general.5 All consumers of what they do not themselves
produce are interested in helping the producers lower
their costs and in preventing them from raising their
prices, or from keeping them up when their costs are
falling. Everyone wishes every other occupation to be
ruled by the law of free competition, his own to be sub
ject to monopoly. Between these opposing tendencies
there is a perpetual seesaw. At a given time and place,
some manufacturers may be nearer the condition of free
competition, and the prices of their products are regu
lated more closely by their labor-costs, and their exchange-values are nearer to their cost-values. At best,
therefore, the cost-value theory of exchange-value is
only partially carried out. And, in general, all that can
be rightly said is that cost-values regulate the lowest
5 There is a peculiar way by which profits are made to appear equal. In
a market at the same time the rate of interest (risk abstracted) must be fairly
uniform, because interest is paid on money loaned, and there is no difference
between money units. But every business enterprise pays different dividends
on the original capital. If the surplus, like rent, in those with high dividends
be taken by the sellers of shares, the profits to the purchasers are equalized.
This, however, is a different matter, and more like interest than profits
proper. Still the higher prices paid for the shares reveal the higher profits on
the original capital. When through partial monopoly these become enor
mous, they are generally lowered in appearance by the injection of “ water,”
in order to head off legislative interference.

FOUR KINDS OF ECONOMIC VALUE

limits of exchange-values.6 But exchange-values do not
merely gravitate toward their under limits: they also
levitate toward their upper limits, and in fact float
variously between the two. There is therefore no pro
priety in picking out cost-value, the determinant of the
lower limit, as alone the regulator of exchange-value.
And the impropriety is all the greater in that, even when
the exchange-value of some commodity is in agreement
with its cost-value, the cause of its being as high as it is
is not that cost-value, and the law regulating it is the
same as when it is elevated above that lower limit.
6 At this point Jevons made an unfortunate slip. His famous sorites:
“ Cost of production determines supply; supply determines final degree of
utility; final degree of utility determines value,” Theory, 2d ed., p. 179,
merely played into the hands of Marshall, who, in his Principles of Eco
nomics, 3d ed., p. 562, drew from it the natural conclusion that cost of
production determines value. The mistake is in the first premiss. Cost
of production alone does not determine supply.

XVII
THE INCONSISTENCY OF RICARDO’S THEORY
I t may be objected, however, that Ricardo himself has
forestalled, met, and overcome this criticism. The
above exposition of his doctrine is based on his own ac
count of it in the long first chapter of his Principles, on
value, and in the two following chapters, on rent; and
the criticism has been directed against the doctrine
there expounded. But immediately after that exposi
tion Ricardo inserted a fourth chapter, “ on natural and
market price,” in which, besides using “price” in a wide
sense, contrary to his first definition of it, treating it
now as synonymous with value,1— and he later speaks
of “ market” or “ natural value” in the same sense, -—
he revises his doctrine and turns it into something quite
different from what he first advanced. As first ad
vanced, his theory was that cost, or last and greatest
cost, regulates or causes 12 value, by “ value” meaning
exchange-value. “ If the quantity of labor realized in
commodities regulates their exchangeable value,” he
there wrote, “every increase of the quantity of labor
must augment the value of that commodity on which it
is exercised, as every diminution must lower it” (p. 11).
Again: “ If fewer fish could be obtained with the same
capital and labor, fish would rise in comparative value.
Fish, then, would rise or fall in exchangeable value,
only because more or less labor was required to obtain a
1 He'could do so because he assumed, hypothetically, that money is
invariable in value, whereupon prices become true indicators of values.
2 This word is employed on pp. 13, 20, 24, also later, p. 44.

FOUR KINDS OF ECONOMIC VALUE

given quantity; and it never could rise or fall beyond the
proportion of the increased or diminished quantity of
labor required.” And of commodities in general: “ Un
less more labor were required for their production, they
could not rise in any degree whatever.” “ If more or less
labor were required in the production” of any commod
ity, “ this will immediately occasion an alteration in its
relative value.” 3 All this would be perfectly true if
stated of cost-value; and the criticism has been that it
is, as a matter of fact, not true of “ value” in the sense
of exchange-value, in which sense Ricardo here used it
and in which he always professed to use it, except when
he specifically speaks of “ value in use.”
But now in the fourth chapter we are suddenly in
formed that by value being regulated by cost, or last
cost, Ricardo meant only that (borrowing a term from
Petty) it is the “ natural price” or “ natural value” of
things that is thus regulated, whereas what he calls their
“ market price” or “ market value” may depart from it
in “ accidental and temporary deviations.” Market value
is the “ actual,” though only “ temporary,” value; natu
ral value is the “ primary,” “ necessary,” and “ per
manent” value.4 And later also he frequently speaks of
price or value, though varying “ always in the market,
and in the first instance, through the comparative state
of supply and demand” (p. 66), yet as being “ in the
market ultimately regulated by the comparative facility
or difficulty of producing” the thing, or “ by its natural
value” (pp. 114, 115). It is “ the natural price of com
3 Page 20; cf. later: “In all cases, commodities rise because more labor
is expended on them,” p. 65; similarly, p. 165. The italics are here inserted.
Kicardo never used them.
4 For “ natural value” he also uses “ real value,” pp. 171, 220, etc.

FOUR KINDS OF ECONOMIC VALUE

modities,” depending “ on the facility of production,”
“which always ultimately governs their market price”
(p. 130), or “ultimately regulates the prices at which
they shall be sold” (p. 227). “ The real and ultimate
regulator of the relative value of any two commodities,
is the cost of their production, and not the respective
quantities which may be produced, nor the competition
amongst the purchasers” (p. 208). “It is the cost of
production which must ultimately regulate the price of
commodities, and not, as has been often said, the pro
portion between the supply and demand: the propor
tion between supply and demand may, indeed, for a
time, affect the market value of a commodity, until it is
supplied in greater or less abundance, according as the
demand may have increased or diminished; but this
effect will be only of temporary duration” (p. 232).5 In
other words, it is only “ natural value,” which we recog
nize to be what is better called cost-value, that is im
mediately affected by any change of costs, and can never
diverge from costs, being exactly regulated by costs;
but “ market value,” which we recognize to be nothing
else than exchange-value, can diverge from costs, doing6
6 Yet this revised version he had inserted in his very first statement of his
theory in a pamphlet published in 1815. “ The exchangeable value of all
commodities rises as the difficulties of their production increase. . . . Whereever competition can have its full effect, and the production of the commodity
be not limited by nature, as is the case with some wines, the difficulty or
facility of their production will ultimately regulate their exchangeable value.”
And in a footnote: “ Though the price of all commodities is ultimately regu
lated by, and is always tending to, the cost of their production, including the
general profits of stock, they are all subject, and perhaps corn more than most
others, to an accidental price proceeding from temporary causes” (p. 337).
Remember we have seen that prices tend also toward a higher limit, and so
it simply is not true to say that they are “ always” tending toward the
lower limit.

FOUR KINDS OF ECONOMIC VALUE

so in the first instance under another law which regu
lates it and is only ultimately, in some cases, affected by
costs, after some interval, the length of which Ricardo
and his followers never attempt to define. So, after all,
the law of costs is admitted by Ricardo himself to have
only a more or less distant influence upon exchangevalue, which, whenever and while it deviates from costvalue, is regulated by the law of supply and demand.
Meanwhile the cost-value itself may be varying, so that,
if it was accidental for exchange-value to deviate from
cost-value, it may be equally accidental for it ever to
get back into agreement with it. But why is it acci
dental in either case? 6 By “ accidental” we mean that
we do not know the cause of what we call such. Ricardo
called the deviations “ accidental” because he did not
know, or at least did not have a satisfactory theory of,
their causes. But that is no reason why later economists
should not work out a theory satisfactorily explaining
market or exchange-value as well when it deviates from
as when it agrees with cost-value, the same theory ex
plaining both. If the law of supply and demand (which
is this theory, displeasing to Ricardo only because of its
vagueness as it was entertained in his day) can regulate
exchange-value while deviating from cost-value, how
does this law suddenly cease to operate when exchangevalue comes into agreement with cost-value? It would
6 Ricardo seems to have had in mind the accidental variations in the price
of corn arising from good and bad seasons: see another footnote on p. 877.
But these are causes, through their effect on the quantity produced, of varia
tions both in the cost-value (or “ natural value”) of corn and in its esteemvalue; and the fluctuations of the “ market value,” or exchange-value, of
corn result from the efforts of the producers not only to recoup their costs,
but to dispose of the whole crop at its proper esteem-value. After a harvest,
the quantity in the hands of the producers is a monopoly, to be treated with
out reference to its cost, which is a thing of the past and irretrievable.

FOUR KINDS OF ECONOMIC VALUE

seem to operate in all cases, and to be the only one of
the two laws that does operate in all cases, whereas the
law of costs, which always determines cost-value, de
termines exchange-value, concurrently with the law of
supply and demand, only when and where there is no
natural or artificial limitation to production.7
Similarly, Ricardo later on revised his theory of rent
and the value of raw produce. A greater demand for
corn, he later incidentally tell us, raises its price (its mar
ket price); this rise of market price or exchange-value
encourages greater production (pp. 252, 253), and then,
if the further supply “ be obtained from land of the same
fertility as that which was last in cultivation, and with
no greater cost of labor, the price will fall to its former
state; if from poorer land [at greater cost], it will con
tinue permanently higher” (p. 95; similarly, p. 66).
Certainly there is nothing accidental in all this. And
now, further, Ricardo once admits that, if land culti
vable even at greater cost is lacking, then “ every portion
of land in cultivation . . . will yield rent,” corn then
selling “ at a monopoly price,” and doing so “ perma
7 Shortly after the publication of Jevons’s work, in a defence of it from
Cairnes’s criticism from the Ricardian viewpoint, George H. Darwin, in an
article on “ The Theory of Exchange-Value ” in The Fortnightly Review, Feb.
1, 1875, well showed that, on the supposition of free competition and of un
limited sources, both the cost-of-production theory and the utility theory
are true, but that, as the cost of production and the final utility of the last
portion produced and consumed are equal, and as “ the utility theory of ex
change shows that commodities exchange against one another inversely pro
portionally to their final utilities; hence it appears that they also exchange
inversely as their cost of production. Thus Mr. Cairnes’s datum is deducible
from Mr. Jevons’s principles. I therefore hold that the utility theory is more
fundamental than the other” ; which other he also calls a “ secondary law,”
applicable only to special cases, while the utility theory is applicable to all
cases (pp. 250, 252).

FOUR KINDS OF ECONOMIC VALUE

nently,” since the “produce cannot be increased.” 8
Thus Ricardo himself corrected his theory very nearly
as we have done. But then this would be his theory,
and not the one which he expounded in the chapters
expressly devoted to it! But no, in other later passages,
considerably more numerous than these few, he re
lapsed into his original theory.9 Therefore really it is
that wrong and correction-needing theory which is
peculiarly his theory; and it is the cost-value theory
proper, for those who apply the cost-value theory to
exchange-value can apply it consistently only in this
way, and they can correct it only by invoking the es
teem-value theory of exchange-value. And in fact his
correction itself was not correctly made. Purely dog
matic was his assertion that the deviation of “ market
value,” or exchange-value proper, from “ natural
value,” or cost-value, is only accidental and temporary.
8 Pages 150, 151. In a later work he reverts to saying “ the only perma
nent cause of rise in the value of corn is an increased charge on its production,
caused by the necessity of cultivating poorer lands ” (p. 460). But above, for
the higher value, he has noticed a different permanent cause, so that this is
not the only one. In another passage he says the supply may not be deficient,
and yet the price will be permanently higher “ because there has been an in
creased cost in producing it ” (p. 96); for he conceived the people as needing
just so much corn and getting it at any cost. This is true, as a matter of fact,
only in the first stages, when the population is small and lavishly supplied
from the lands of greater fertility. In the later stages supposed by the pas
sage above in the text, the people simply do have to put up with a smaller
supply, and to make the best use of it.
9 Thus on p. 05 he speaks of certain commodities rising in price “in con
sequence of more labor being expended on them.” “ In another part of this
work [the first chapter], I have endeavored to show, that the real value of a
commodity is regulated, not by the accidental advantages which may be
enjoyed by some of its producers, but by the real difficulties encountered by
that producer who is least favored ” (p. 220). The original theory first ap
peared on p. 373, and is repeated on pp. 60, 70, 91, 104, 135, 150, 151, 171,
198, 200, 220, 221, 246, 251, 460, 475.

88
FOUR KINDS OF ECONOMIC VALUE
We have seen that it is just as natural to bull prices up
toward their upper limit as it is to bear them down to
ward their lower limit. If there are accidental and tem
porary causes affecting exchange-values, there also are
accidental and temporary causes affecting cost-values.
Moreover, Ricardo never corrected his erroneous state
ments about labor being the cause of exchange-value, as
far as labor went. He merely sought for a new cause
when exchange-value rose above what he could account
for by labor. Also his use of terms continues wrong;
for just as by “ market value” he meant “ market ex
changeable value,” so by “ natural value” he meant
“natural exchangeable value.” He tried to treat them
both as varieties of exchange-value; " whereas they are
coordinate species, and his “ natural value,” being costvalue, is not exchange-value at all, although it may
sometimes agree with it.
'
The later acknowledgment of error in the first ex
position is, after all, only a repetition of the admission
momentarily made at the outset. There one class of
objects were allowed not to have their exchange-value
regulated by cost, for the very good reasons that some
of them have no cost, others have no present cost, and
others again, by natural (or artificial) restriction, e. (/.,
the “ wines of peculiar quality,” have a monopoly price.
Of monopolized goods, he later tells us, “ the price has
no necessary connection with their natural value” (p.
234); wherefore, as he had no clear idea how it is regu
lated, he called it “ a fanciful price” (p. 150). Thus of
the things in the first class it was only the market value
that Ricardo momentarily considered, and then dis
missed on the ground that “ these commodities form a
very small part of the mass of commodities daily ex

FOUR KINDS OF ECONOMIC VALUE

changed in the market ” — in spite of the fact that the
paper money which, when he wrote, appeared in every
transaction in the market, belonged to that class. We
were there given to understand that nothing more was
to be said about these things with their unaccountable
values; yet now they, or others behaving like them,
constantly reappear, like Nature, pitched out with a
fork, always recurring, or, better, like the evil spirit
which, when driven forth, returned with seven others.
There we were told that he would pay attention only, to
the things in the second class — “ such commodities as
can be increased in quantity by the exertion of human
industry, and on the production of which competition
operates without restraint.” It was of these he there
maintained that their value (= exchange-value =
market value) always agrees with their natural value.
But in their case also he admitted, in the later revision,
that their market value, or exchange-value, can deviate,
though only more or less temporarily, from their natural
value, or cost-value. And we find that this deviation of
their exchange-value from their cost-value is due to the
same cause which regulates the exchange-value of the
things in the first class, namely, scarcity in addition to
utility, or the amount of existing goods in relation to
the people’s desire for them — in short, supply and de
mand. Thus Ricardo really entertained two theories of
exchange-value: the one, that it is due to supply and
demand (the esteem-value theory of exchange-value),
the other, that it is due to cost (the cost-value theory of
exchange-value). He used the first, really, of all com
modities, though of some only temporarily. He used
the second of as many commodities as he could, and as
often and as long as he could, and when he could not,

FOUR KINDS OF ECONOMIC VALUE

he fell back on the first. Evidently the first is the more
important; and yet he tried to make out that the second
is the only important one. The first he had abundant
facts to prove; the second he never even attempted to
prove. It was an a -priori theory. At the end of the
fourth chapter he says: “ In speaking, then, of the
exchangeable valu^ of commodities, or the power of
purchasing possessed by any one commodity, I mean
always that power which it would possess, if not dis,tur.bsd^b-y~*ny temporary or accidental cause, and
which is its natural price.” What does this mean but
that by exchange-value he meant the value which a
thing would have if it were determined by its cost-value,
— if, in fact, it behaved as his theory demanded? And
this after admitting that actually it does ROt so behave!
He simply did not mean always what he here tells us he
always meant. His theory is merely the analytical one
about cost-value, and it is true of exchange-value only
when exchange-value happens to agree with cost-value.
This absence of any attempt at proof of his theory
applied to exchange-value is serious. We have seen that
Ricardo denied the possibility of a measure of value,
that is, of cost-value (as well as of exchange-value), and
we have perceived the mistake in his denial (above, in
XII). If cdst-value exists, it should be measurable,
since it depends on costs, and costs are measurable.
This provides a means of testing the cost-value theory
of exchange-value. We have seen that if the variation
of the cost-value of the unit of something were deter
mined over several periods, the inverse portion or excess
of this unit would have a stable cost-value. Now, if
Ricardo’s theory that exchange-value is regulated by
cost-value were true, the variations of the units of all

FOUR KINDS OF ECONOMIC VALUE

other things could be found by comparing their prices
with the prices of these portions of the standard thing
that have stable cost-value. But the cost-values of
several other things over the same periods could equally
well be ascertained directly, in the same manner as were
the cost-values of that first thing, and these other things
could equally well serve as standards. Then it would be
seen whether they agreed or not. Yet if they did not
agree, Ricardo’s followers could say it was because the
prices used were only market prices. This merely shows
the hopelessness of the tangle in which they have wound
themselves. No practical use of the cost-value theory of
exchange-value can ever be made, because of the possi
bility, always existing, that, after all, exchange-value,
now disparagingly called mere “ market value,” may
deviate from it. The best his followers can do is to fall
in behind that most ardent admirer of his, DeQuincey,
who wrote: “ A measure of value which cannot be prac
tically applied is worthless; as a measure of value, there
fore, Mr. Ricardo’s law of value is worthless.” But no
blame was to be settled on Mr. Ricardo on that account,
“ for Mr. Ricardo never dreamed of offering a standard
or fixed measure of value.” 10 Ricardo, in fact, was a
broker, and knew nothing about the costs of producing
things, and apparently had no desire to know anything
about them.11 His followers may defend him by going
on with DeQuincey, who challenged “ any man to show
that the great interests of Political Economy have at all
suffered from want of such a measure, which at best
would end in answering a few questions of unprofitable
10 The Templars’ Dialogues, V.
11 DeQuincey somewhere tells of a Manchester merchant who dealt in
wheat for forty years, and never once saw a wheat field.

FOUR KINDS OF ECONOMIC VALUE

curiosity; whilst, on the other hand, without a know
ledge of the ground [ = source, cause] on which value
depends, or without some approximation to it, Political
Economy could not exist at all, except as a heap of base
less opinions.” 12 The uses of this knowledge are stated
to be its influence on all the derivative laws, and its
simplification of the science. DeQuincey could not fore
see that one of the uses of Ricardo’s theory would be to
establish Marxism and Bolshevism. Indeed, it may be
said that of all false theories Ricardo’s has proved itself
the most pernicious.
Ricardo himself made no such deduction. He used
his own theory when he could, and he used the other
when it suited him. His own theory is at the bottom of
his mistakes. The other is what led him to truth. One
of his mistakes was to maintain that wages and profits
can vary only inversely to each other, or at each other’s
expense. This is an obvious deduction from the prin
ciple that exchange-value always is exactly bound down
to, and is identical with, cost-value. But it is vitiated
by the fact that exchange-value, or market value, is not
so bound and is not identical with cost-value.13 The por12 Ibid. Cf. his later Logic of Political Economy, 1844, pp. 45, 49 (chap. 3,
section 3). There he admits one use of measurement*, to determine “ varia
tions in the stages of a paper money ” — not a matter of mere idle curiosity!
13 Even leaving the deviation of market value out of the question, this
proposition, though true in one sense, is false in another. Even if exchangevalue always were identical with cost-value at the same time and place, we
shall presently see that they are not identical in the course of time. If under
better management labor becomes more productive, there will be more to
distribute, so that the purchasing power (or exchange-value) of both wages
and profits may increase together. But Ricardo held that this greater prod
uct has only the same “ value” as before, because produced by the same
amount of capital and labor —■ it has the same cost-value, which he still
identifies with exchange-value. Now, of course, the invariable cost-value of
the variable product can be shared only by the laborer getting more as the

93
tion of Ricardo’s political economy that has been of
the greatest value is his discussion of monetary prob
lems, in which his profession made him well versed. Of
his treatment of money, nine tenths was an application
to it of the quantity theory (the esteem-value theory
of its exchange-value).14 And his one great error in the
theory of money was due to his cost-value theory of
exchange-value. He said the ratio of gold to silver, then
hovering between 15 and 16 to 1, was owing “ solely”
to the fact that “fifteen times the quantity of labor is
necessary to produce a given quantity” of gold than of
silver (p. 213). Thirty years later DeQuincey was able
to repeat the same assertion.15 That ratio had existed
already for many years, and it continued until 1873,
when it suddenly began to change and has been variable
ever since. It is the “ market value” ratio, and here
was a remarkable instance of the market values not de
viating from the alleged natural values for a very long
run. Ricardo and DeQuincey never examined whether
FOUR KINDS OF ECONOMIC VALUE

capitalist gets less, and conversely. This again is a worthless (except for being
a misleading) analytical proposition. Because of it, Ricardo maintained that
“ a real rise of wages is necessarily followed by a real fall of profits,” p. 534,
“real” here having reference to labor-cost, whereas it ought to refer to the
product, or to what the product can obtain in exchange — to the exchangevalue proper of the product, concerning which the statement is absolutely
false. It led him into the terrible blunder of saying “ there is no other way of
keeping profits up but by keeping wages down ” (p. 476).
14 He began with this in 1809 and 1811, before he distinctly developed his
cost-value theory; and he did not give it up afterwards. At the beginning
he said that, if a gold mine were discovered in some country, its currency
“ would be lowered in consequence of the increased quantity” (p. 264). Cf.
above, in n. 9 on p. 87.
16 In his alarmist essay on the gold discoveries in California. A few years
later Jevons, who was in Australia during the gold fever there, believed that
gold was being extracted at a cost greater than it was worth (Investigations,
pp. 70-71).

FOUR KINDS OF ECONOMIC VALUE

the mines were producing just fifteen times as much
silver as gold at the same cost: they never conceived
that to be necessary. And in fact it was not necessary;
for it is perfectly possible that some gold mine some
where and some silver mine somewhere else were yield
ing gold and silver at that ratio to the same amount of
labor, and no good would be accomplished by finding
just what mines were doing so. But to assert that these
two sets of mines caused gold to be so many times more
valuable than silver was to assign a most inadequate
cause, in view of the quantities of the two metals already
accumulated at all sorts of costs. With equal likelihood
there were other mines that could have been worked, but
were not because it did not pay to work them; and after
1873 some gold mines were opened which could not be
profitably worked before, and some silver mines were
closed because they could no longer be worked except at
a loss. Obviously it was the existing exchange-values
of gold and silver, due to the existing supplies and to
the monetary and other demands for them, that per
mitted some mines to be worked and others not. Ri
cardo reversed the order of cause and effect. And the
principal reason why gold and silver had the ratio they
had was before his eyes, staring him in the face; but
he could not see it, being blinded by his false theory.
Yet, with regard to paper money, Ricardo always said
that its value — its market value, its exchange-value —
depends entirely on its quantity (relatively to the de
mand for it), and that its exchange-value could be kept in
agreement with that of gold by regulating its quantity
(pp. 214, 265, 346, 398) — and not only temporarily,
but apparently forever. In fact, he considered paper
money, so regulated, the most perfect form of money

FOUR KINDS OF ECONOMIC VALUE

(pp. 218, 301, 397, 404^05). But the cost of producing
paper money has nothing to do with its exchange-value.
Then why is it that the cost of producing gold should
necessarily be the regulator, temporary or permanent,
of its exchange-value? Ricardo himself said of money
in general, “ The demand for money is regulated entirely
by its value, and its value by its quantity” (p. 114).16
Here, of gold money in particular, he added: “ From its
durable nature, and from the difficulty of reducing its
quantity, it does not readily bend to variations in its
market value.” This seems to be a misstatement for say
ing that, for these reasons, its market value, or exchangevalue, does not readily bend to its natural value, or
cost-value. This, however, would be inconsistent with
the notion that its market value and that of silver did
bend to their natural values for so long a time. Rather,
these reasons account for their natural values, or costvalues, not readily affecting their exchange-values.
That was why Ricardo was at liberty to apply the quan
tity theory to explain the variations of their exchangevalue. But in doing so he was not using his own costvalue theory.17
16 And (p.337 n.) lie spoke of “ an alteration in the value of the precious
metals, arising from their abundance or scarcity.”
17 Ricardo said that throughout his reasonings about the exchange-values
of commodities he would assume that the “ value” of gold, or of money, was
invariable (pp. 29, 60 n.). What could he mean by that? Suppose the costvalue of gold were invariable: still its market value could be variable. Perhaps,
then, he meant that the market value as well as the natural value of gold
should be invariable. But if the market value of gold were invariable, that
would be enough, and everybody (except gold miners)would be indifferent
to variations of its cost-value, about which they would know nothing. We
may take a hint from a passage already quoted. He meant that the exchangevalue of money should be what it would be if it remained in agreement with
an invariable cost-value of gold, whether gold happened to be invariable in
cost-value or not. This, of course, does not mean that the exchange-value of
money should be invariable.

XVIII
ON LABOR-COST

ic a r d o , however, was justified in objecting to Adam
Smith’s doctrine that rent is a component part of price.
But he should have objected likewise to Adam Smith’s
doctrine that wages and profit are component parts of
price. To the capitalist-employer producers two of
these, wages to all, and rent to those who pay rent, and
also interest (wrongly confounded with profit), are com
ponent parts of costs. Prices are what the producers
can get and what the consumers will give. They have
no component parts. They must, in the long run, be
above the costs, and the difference is profit. Profit is
not a component part, or cause, of price, but an effect
from the existent cost and price, which have different,
though interacting, causes.
Here we have unavoidably run over from exchangevalues to prices, and from labor to wages. Prices, at
any one time and place, correctly measure exchangevalues, and different wages indicate different kinds or
qualities of labor. To expound the labor-cost theory of
exchange-value entirely in terms of labor is an even
more impossible task, and involves a vicious circle.
The money-cost of the labor that has gone into the pro
duction of a commodity is composed of all the various
wages and salaries paid to the employees, including the
managers, all the different kinds of labor receiving dif
ferent pay, but all measurable in the money-unit.1 If

1 The subject of wages has at the beginning been excluded from the range
of this paper. Labor has no value, and there is no need of speaking of the
“ value of labor,” the term “ wages” being sufficient. Wages, however, have

FOUR KINDS OF ECONOMIC VALUE

the labor-cost simply — the mere expenditure of labor,
so many hours of so many laborers be considered, we
shall have to assign different “ values” to the different
labor of the different laborers without possessing any
labor-unit to measure them in. “ Value must here be
taken in an extra-economic and mathematical sense.
an hour of one man’s labor is to be considered as a
quantity of labor equivalent to two hours of another
man’s labor, when the former ordinarily, because of its
greater efficiency, at the same time and place receives
double the wages (or salary) of the latter. The costvalue theorists have generally slurred over this point.
They have generally taken for their unit the labor of
the worst paid large class of laborers they were familiar
with, and have then treated it as an easy matter to re
duce all the other superior labor to this unit. But the
only way they have of making this reduction is by means
of the wages or salaries the superior kinds of laborers
ordinarily command and receive. Thus if any one kind
of a commodity is produced by one hour of all the kinds
of labor required for its production, and an equally
valuable commodity of another kind is produced by two
hours of all the kinds of labor required for producing
it, the fact of the equal value of the two commodities is
taken as showing that really the same quantity of labor
has been employed to produce them, and therefore the
some analogy with value; which accounts for the practice of identifying
them. Similar theories have been invented to explain them: a cost-ofproduction theory, applied to the upkeep of the labor supply, and the scar
city theory, as the scarcity of competent laborers in different branches
affects their comparative remuneration. But the “ value” of the product
is always at the bottom of the question of wages. Hence the subject of
wages, which belongs under production, should be treated after the subject
of “ value,” which belongs under consumption. Cf., above, note on p, 68.

FOUR KINDS OF ECONOMIC VALUE

one hour of labor producing the former is the same
quantity of labor as the two hours of labor producing
the latter. Evidently the quantity of labor going into
the production of the two articles is now being deter
mined by the exchange-value of the articles, and not
the exchange-value of the articles by the quantity of
labor. There is no beginning or end to this method of
reasoning. At best, the cost-value theorists can hold
that labor is according to the laborer. This was main
tained by Aristotle, who said that as the agriculturist
is superior to the shoe-maker, so the product (and its
value) of the agriculturist’s labor is superior to that of
the shoe-maker’s. But this evidently inverts the se
quence of cause and effect. A great artist earns more
than, and is superior to, a shoe-maker, because his prod
ucts have greater “ value,” including greater exchangevalue.2
Thus, altogether, labor, though it creates wealth,
does not create value, except in the single sense of
“ value” meaning cost-value. Labor produces things
that have value, but it does not produce any value ex
cept cost-value. Neither use-value nor esteem-value
nor exchange-value is produced or caused by labor.
These values come to the products of labor, as to the
things not produced by labor, from elsewhere. The
three come to an object that has value from its utility,
and the later two both come to it from its scarcity.
2 A poor picture by a great painter may have greater exchange-value
than a better picture by an inferior artist, because, through association with
his other better pictures, the former acquires greater esteem-value than the
latter. The anomaly is to be explained under the theory of esteem-value.
Exchange-value here, as always (so the esteem-value theorists of exchangevalue maintain), is determined by esteem-value. But economists need not
pursue desire into all its vagaries.

FOUR KINDS OF ECONOMIC VALUE

People produce things only for the sake of these values
in the things. They expend labor upon them, if neces
sary, up to their value. Things have cost-value only
because they have the other values. Condillac was
right.

XIX
GENERAL EXCHANGE-VALUES ACCORDING TO
COMPARATIVE SOCIAL ESTEEM-VALUES
T h is leads us to the esteem-value theory of exchangevalue. The full explanation of exchange-value could not
be given till the theory of esteem-value was developed.
From of old people have held the rough-and-ready, ruleof-thumb theory of “ supply and demand,” as it is called,
which says that the “ value” of a thing varies directly
with the demand for it and inversely with the supply of
it. If by “ demand” be meant desire and by “ supply”
the quantity possessed, this is merely a statement of the
law of esteem-value. But “ demand” and “ supply”
mean more than that. “ Demand” for a certain kind of
things means the quantity of other things offered in ex
change for it by the persons desiring it, while the quan
tity of the thing offered by persons desiring the other
things is its “ supply.” All exchanges are between pos
sessors, however they came to possess the things they
exchange. They need not be the producers of them.
In regard to the things now possessed which were pro
duced by past labor and whose exact copies cannot be
reproduced, we have seen that the cost-value theorists
agree with the esteem-value theorists, having really aban
doned the field to them. We are concerned, therefore,
not only with exchanges between contemporary pro
ducers and consumers, but with all possible exchanges.
The same person qua consumer is very different from
himself qua producer. Of any given article the con
sumers are much more numerous than the producers:

FOUR KINDS OF ECONOMIC VALUE

101

of some articles the whole people are consumers. Con
sumers are concerned only with the esteem-value to
themselves of what they give and get in an exchange.
How this value is determined, we have already seen;
and the cost of production of articles had nothing to do
with it. Producers, however, generally produce their
own article in such excessive quantities that its esteemvalue to them as an article of consumption, after the
production, is practically nil; but as an article of ex
change, and useful to them for this purpose, its esteemvalue to them was, in anticipation, at least somewhat
greater than its cost-value, or else they would have had
no motive to produce it. In the open markets of a
country, people are able to compare the esteem-values
they set upon things by each one learning how much
others are desirous of getting and are willing to give;
whereas the cost-value of every article is known only to
its producers, and these generally conceal it, and even
try to deceive others about it. Moreover, even if it
could be found out, people are more interested with the
esteem-value to themselves of the things they exchange
than they are with the cost-value of every article to its
particular producers, about which they care so little
that they rarely make inquiries, and if told, pay little
attention. But as esteem-value is determined by quan
tity, which is variable, general utility being almost a
constant, everybody is interested in the quantities in
the market, whether recently produced and reproduc
ible, or produced only in the past. This is the reason
why the esteem-value theory accounts for monopolized
goods as well as for those freely produced. The proper
meaning of the terms “ production” and “ producer”
is the act, or agent, not only of making but also of bring

102

FOUR KINDS OF ECONOMIC VALUE

ing to market. In this broad sense there are producers
of antiques as well as of ordinary things.
Now, in every exchange each party receives an article
of higher esteem-value to himself than the article has
which he gives, because otherwise there would be no
motive for making the exchange. Each party wishes
the excess to be as great as possible. The consumer
wishes to purchase as far as possible below his esteemvalue, and the producer wishes to sell as far as possible
above his cost-value. When the exchange is of a single
commodity for money, the two parties set upon it dif
ferent estimates of its esteem-value to themselves, that
of the purchaser being the higher; and the price at
which the sale is effected will range between these esti
mates. In the open markets of a country, where many
producers come with various quantities of all commodi
ties, and many consumers (often the same persons)
come with variously intensive desires for various quan
tities of all the commodities, the estimates on both sides
in regard to every commodity will have a wide range,
crossing one another variously. Ordinarily, the higher
the price first set, the greater the offer and the less the
demand; the lower the price, the less the offer and the
greater the demand. So in the case of each commodity
individually,1each in turn when sold giving its producers
1 Of each article the supply and the demand may be represented diagrammatically by two curves, generally crossing each other. These curves
represent the conditions on one particular occasion, and they change per
haps every day. Sometimes, when the production of an article is exceeding
its consumption, prices fall, consumers (and the middlemen near to them)
stock up, and a day comes when “ the bottom falls out of prices” and there
is a “glut.” At other times, when consumption is exceeding production,
prices rising, there comes a day when producers (including the middlemen
nearest them) have no stock and are almost indifferent about sales. On the
former day the supply curve is so high as hardly to meet the demand curve.

FOUR KINDS OF ECONOMIC VALUE

103

opportunity to appear as purchasers of other things,
and all forming one great system of interactions. The
price at which the supply and the demand equal each
other is the price which will prevail. It may not be im
mediately hit, and meanwhile conditions may change:
the actual prices will trail the theoretical or normal
prices. The normal prices, on assumed conditions, can
be calculated mathematically. The best mathematical
treatment of the subject has been made by Walras;
the best in common language, by Bohm-Bawerk.2 Mill
On the latter, the demand curve is so high as hardly to meet the supply
curve. Between these there are many gradations. This remark is necessary
because these curves are often taken to represent the course, instead of the
state, of things.
3 Bohm-Bawerk, however, in one small detail has made a mathematical
mistake. He says, op. cit., pp. 267-270, that in an isolated bargain between
the seller and the purchaser of a single article, if the former estimates it at
100 florins and the latter at 300, and if they are both equally good bargainers
so that neither gets the better of the other, the price will be about 200, each
gaining a hundred. He here argues as Nozzolini did in a certain seventeenthcentury controversy reviewed by me in The Problem of Estimation, since he
lights immediately on the arithmetical mean as the proper one to draw be
tween the two estimates. But Galileo, as shown by his argumentation in that
controversy, would have argued that this was wrong, in that the seller would
gain the double of his estimate, but the purchaser only two thirds of his, or,
in other words, the seller has after the transaction 100 per cent more than
his estimate, but the purchaser has an article which he estimates only 50 per
cent more than he paid. Galileo would therefore have said that the equal
bargain, showing equal skill in both parties, would have been at a price
which is the geometric mean between 100 and 300, viz., a trifle over 173,
which would give the seller 73 per cent more than his estimate, and the pur
chaser an article which he estimates 73 per cent above what he paid. We
may even suppose that the estimates were actual exchange-values, — that
the seller had paid 100 for the article, and the purchaser later sells it for 300.
Or, with allowance for expenses, we may suppose the seller had bought the
article for 90 and after spending 10 on it estimates it at 100, and the buyer
after spending 10 on it sells it for 300. The equalizing price would then be
represented by x in the proportion 100 : x = (x + 10) : 300, which yields
x = 168j, giving to both the same profit of 68j per cent on each one’s outlay.

104

FOUR KINDS OF ECONOMIC VALUE

came very near to correctly stating the theory, in his
doctrine of the equation of supply and demand; but
missed it because of not clearly recognizing the distinc
tion between esteem-value and use-value.
The theory involves that every purchaser of a com
modity sets an esteem-value upon it at least as high as
the price, and has something to give which he does not
esteem so high, those to whom its esteem-value is not so
high being shut out, even though they have plenty of
other things to give for it. It involves also that every
producer who continues to produce the thing can pro
duce it at a cost at least as low as, or, rather, lower
than, the price. The highest cost required to produce
the amount demanded sets the lowest limit to the price
while the demand at that price continues; but, as we
have seen, nothing hinders the price, when the demand
increases, from rising above the highest cost, and above
the cost-value, if this be held fixed by natural or artificial
restriction. Not so with the relationship between the
price and esteem-value. The price cannot go above the
esteem-value which at least some one sets upon the
thing — that is, if it be an effective price, a sale taking
place; and it will not fall below the esteem-value repre
sented by the figure at which some one will purchase the
last increment offered. Exchange-value necessarily is
tied to some esteem-value; and it is determined by the
esteem-value which equalizes the demand to the supply
or the supply to the demand. The theory is, obviously,
the same for monopolized as for freely produced goods,
the difference coming in only on the side of produc
tion.
The prices of all articles, so determined, may be said
to indicate the comparative general esteem-values that

FOUR KINDS OF ECONOMIC VALUE

105

the people as a whole sets upon the articles in the mar
kets. These general esteem-values may be called the
social esteem-values of the articles, in distinction from
the various personal esteem-values in which they are
held by individuals. If the cost-value theory of ex
change-values were true, the same distinction would
have to be made in cost-values. There are, as we have
seen, various personal cost-values, every producer having
his own; but the idea of a social cost-value is a mere
fiction, since the people at large do not produce every
single thing, and the cost of a thing to its non-producing
consumer can only be transferred to it from the cost of
what he gives in exchange, and to some persons what
they give has cost them nothing. In exchange-value,
however, the same distinction reappears; for the prices
prevailing in the open markets indicate the social exchange-values of the articles, while particular exchanges
sometimes take place at other rates, which may be
called personal exchange-values, and, when differing
from the social, are not considered fair indications of
the thing’s true exchange-value.3 The term “ social”
applied to two kinds of value must not be taken to imply
that all value is social — far from it. Every purchaser
who keeps what he gets, does so because his personal
esteem-value set upon it is higher than its social esteemvalue; and any one whose esteem-value for a thing is
lower than its social esteem-value, must go without it,
or if he happens to have it, he is likely to sell it. Now,
3 General exchange-value is only social exchange-value, since personal
exchange-values, when different from the social, are only infrequent, not
repeated, and cannot be averaged. These are not a subject of interest. The
term “ social exchange-value” is not needed, since “ general exchange-value”
is better and is sufficient.

106

FOUR KINDS OF ECONOMIC VALUE

in the money régime, at any one time and place, prices
indicate exchange-values; for in every exchange in open
markets the articles exchanged have, by definition,
equal exchange-values. Hence exchange-values are ac
cording to comparative social esteem-values.

XX
THE DIVERGENCE OF EXCHANGE-VALUES FROM
ESTEEM-VALUES IN THE COURSE OF TIME
I t needs to be emphasized that exchange-values are
according to, or are determined by, comparative social
esteem-values; for they are not simply equal to, or
identical with, social esteem-values. Most of our treat
ment of the subject so far has been conducted on the
supposition that we are dealing with things in the same
place and at about the same time. It is necessary now to
deal with the values of things at different times,— dif
ferent places being mostly neglected for simplicity, —
that is, with things as they are valued by a people in the
same country through the course of time. Exchangevalues may, as time advances, cease to agree with social
esteem-values — or simply with esteem-values, as we
shall again generally omit the epithet, since our atten
tion is principally centered upon social values.1 For we
have seen that in the course of time all things can
rise or fall in esteem-value, whereas all things cannot
rise or fall in exchange-value. Exchange-values are in
terrelative, while esteem-values (along with cost-values)
have something absolute about them, since they vary
independently of one another; but every variation of
one exchange-value involves an opposite variation of
1 Naturally, in the course of time, personal esteem-values vary on the
whole with social esteem-values — in fact, leading the way. The personal
esteem-values of any one thing vary manifoldly, and the variation of its
social esteem-value, or simply of its esteem-value, is a middle term between
the extreme variations.

108

FOUR KINDS OF ECONOMIC VALUE

some other exchange-value.2 It is this companionship
at one time, or for a short time, and this eventual part
ing of company in the course of time, of esteem-value
and exchange-value, that has caused the confusion.
For while they are together, the same term is applicable
to both; but when they diverge in their variations, only
one term being still employed, inconsistency arises.
If all things fell exactly alike in esteem-value, their
exchange-values would remain entirely unchanged. Also
if all things fell exactly alike in cost-value, and if it fur
ther happened that their exchange-values were regulated
accordingly, and if the things that have no cost-value
behaved as if they had, then too their exchange-values
would remain unchanged. If things fell on the whole in
these values, but differently, there would be variations
of their exchange-values corresponding to the differences
in these falls, but exchange-values on the whole would
still be constant. In this respect esteem-value and
cost-value are alike, and form one category, over against
exchange-value as forming another. For generally
esteem-values and cost-values vary in the course of
time more or less closely together, since either the dis
covery or utilization of new sources that are worked
more cheaply, or the application of better machinery or
processes producing more than before by the same labor,
or both together, increases supplies and thereby per
mits a fall of esteem-values along with the fall of
cost-values; or conversely, with the increase of costs, the
scarcity which is either its cause or its effect, causes a
2 The exchange-values of all things can, of course, fall successively, just
as all the molecules of water in a pond can sink one after another, though
they cannot all do so at the same time. And the reason is the same in both
eases, in that the sinking of one thing shoves others up. It is a very simple
fallacy to think that all exchange-values can fall together.

FOUR KINDS OF ECONOMIC VALUE

109

rise of esteem-values along with the rise of cost-values.
The ulterior causes of these changes — whether they be
expansion and new inventions, or the impossibility of
further expansion, the exhaustion of sources, and de
generacy of the people — need not here concern us.
Mankind is now, and has been for several centuries,
and is likely to be for some time to come, in an era of
improvement and of expansion greater than the growth
of population, and both the cost-values and the esteemvalues of most, if not of all, things, have been, are, and
are likely to be, falling.
But the exchange-values of all things are not falling.
Only the prices of all things may fall — not necessarily
do fall. For all prices can vary together, like esteemvalues; but they do not necessarily do so. They may,
on the whole, remain stable, like exchange-values; they
may even, on the whole, rise, though esteem-values
(and cost-values) on the whole are falling. At a given
time and place prices correctly indicate exchangevalues, so that the terms may be interchanged. Con
sequently some economists have confounded them, and
have habitually used “ price” for “ exchange-value.”
This is unobjectionable for contemporary and coregional prices and exchange-values. It is, however, no
longer correct when prices and exchange-values are
compared at different times and places. This identifica
tion of prices with exchange-values (always so far mean
ing general exchange-values) and through them with
esteem-values (and less closely with cost-values) at the
same time and place, along with the possibility (though
not the necessity) of prices accompanying esteem-values
(and approximately cost-values) through the course of
time, is what has permitted economists to make the mis

110

FOUR KINDS OF ECONOMIC VALUE

take of thinking also of general exchange-values as
accompanying esteem-values (and cost-values) through
the course of time. But it is only the particular ex
change-values of things in money that are always iden
tical, in magnitude, with prices, by definition;3 and it is
only these particular exchange-values of things in
money that can accompany their esteem-values (and
cost-values) as they fall all together, or on the whole.
Then the particular exchange-values of money in the
things, the reciprocals of the prices, all rise together, or
on the average, which means that the general exchangevalue of money rises. In other words, the general ex
change-value of money rises if prices accompany the
falling esteem-values (or cost-values) of things. If
money is to remain stable in general exchange-value,
the average of prices must remain stable, some prices
rising and others falling as esteem-values vary among
themselves in their general fall. Then prices keep com
pany with the general exchange-values of things, varying
as they do; but part company with their esteem-values
(or cost-values), varying only as these do relatively to
one another, but not as they do absolutely.
Ricardo knew nothing about the compensatory action
of oppositely varying particular exchange-values and
the consequent possibility of measuring the variation of
the general exchange-value of a thing (especially of
money) by averaging the variations of its particular
exchange-values (or, inversely, of prices).4 He had
3 For the definition of price is the amount of money that can be obtained
in exchange for the thing, and this is also the definition of the particular ex
change-value of the thing in money, with the only difference that the latter
is, not the amount of money itself, but the power to obtain it.
4 “ Even this comparison [with the value of many other commodities]
does not afford a certain test [of the value of gold or money], because one

FOUR KINDS OF ECONOMIC VALUE

111

heard of the proposal to judge of the value of money
“ by its relation to the mass of commodities” (p. 400),
as it had been made in England in his day by G. Shuckburgh Evelyn, Arthur Young, and J. P. Smith, al
though the first serious effort to measure the general
exchange-value of money in this way and to reduce con
tracts to a stable value of this kind was made by Joseph
Lowe a year after the last edition in his lifetime of his
Principles. But because of his ignorance of the method
of executing this proposal, and because of his identifica
tion of exchange-value with cost-value at the same time
and place, Ricardo not unnaturally fell into the way of
identifying these two kinds of value at all times and
places. As all commodities can rise or fall in costvalue, “ why,” he asked, “ should all commodities to
gether be the standard, when such a standard is itself
subject to fluctuations in value?” (p. 166). This can
be true of “ value” only in the sense of cost-value (or
esteem-value), but Ricardo applied it to exchangevalue: “ the exchangeable value of all commodities
rises as the difficulties of their production increase” (p.
377). Still, he did dimly perceive that all commodities
cannot rise or fall together in exchange-value;5 which
precisely fits them for being a standard, wherefore if
any one thing, say money, preserved all its particular
exchange-values constant on the average, money would
half of the commodities to which they are compared may have varied in one
direction, while the other half may have varied in another: by which half
shall the variation of gold be tried? If by one, it appears to have risen; if
by the other, to have fallen” (p. 470). Similarly p. 228. The confusion is
caused by the idea of cost-value being implicitly present, as it is explicitly on
pp. 400-401.
6 “Corn rises because it is more difficult to produce, and its cost is in
creased; it would be no rise at all if all other things rose with it” (p. 463).

112

FOUR KINDS OF ECONOMIC VALUE

be stable in “ value” in the sense of general exchangevalue. Only this dim perception never rose to clear
conscious recognition, and therefore he had to deny that
all commodities together can serve for such a standard,6
and to fall back on his position that “ that commodity
is alone invariable [in value] which at all times requires
the same sacrifice of toil and trouble to produce it.” 7
Now, Ricardo had a right to the opinion, perfectly ob
vious and merely analytical, that only what is stable in
cost of production is stable in “ value,” provided he al
ways recognized, and proclaimed, that by “ value” he
meant only a kind of value, namely cost-value, which
by definition is tied to cost; but this opinion could then
exert no influence on the views about exchange-value
of those who clearly recognize the distinction, at least
through the course of time, if not at the same time, be
tween cost-value and exchange-value.
When the theory of esteem-value and its application
to exchange-value was worked out, its discoverers might
have made a similar mistake of inferring that because
6 “I cannot agree with M. Say in estimating the value of a commodity
by the abundance of other commodities for which it will exchange ” (p. 171).
Yet this contradicts the following: “ The only proof which we possess of the
relative value of money in two places, is by comparing it with commodities.
Commodities measure the value of money in the same manner as money
measures the value of commodities. If, then, commodities will purchase
more money [ = if money will purchase less commodities] in England than in
France, we may justly say that money is cheaper in England ” (p. 293; cf. p.
266). The solution of the contradiction is, of course, that the latter passage
refers to “ market value” and the former to “ natural value” ; yet the con
tradiction still exists when we remember that Ricardo subsumed both these
values under exchange-value.
7 Page 166. Because all commodities are produced by labor, “ labor is a
common measure by which their real as well as their relative value may be
estimated” (p. 171). Their “ real value” in the sense of cost-value, yes; but
their “ relative value” in the sense of exchange-value, no.

FOUR KINDS OF ECONOMIC VALUE

113

all exchange-values are in agreement with comparative
esteem-values at the same time and place, therefore
they all rise or fall through the course of time with the
rise or fall of all esteem-values. Jevons, who had already
himself devoted attention to the measurement of gen
eral exchange-value, and Walras, who afterwards did
the same, escaped making this mistake. Menger also
did not make it so baldly as Ricardo had done; but he
bordered upon it. Because of certain difficulties in
measuring variations of the general exchange-value of
anything, he concluded that this measurement is im
possible. And then, without showing how the measure
ment of variation of esteem-value is any more easily
and accurately attainable, he centered his attention
upon esteem-value. As Ricardo, merely because of his
theory explaining exchange-values by comparative costvalues, conceived only of money remaining stable in
cost-value as its one way of remaining stable in “ value,”
and desired it to remain stable in “ value” in this sense,
so Menger, simply because of his theory explaining
exchange-values by comparative esteem-values, like
wise desiring money to remain stable in “ value,” ap
proved of its remaining stable in esteem-value. Menger
would also have been perfectly justified in saying that
only in case money is stable in esteem-value is it stable
in “ value,” provided he clearly stated that by “ value”
he meant only esteem-value. Then this likewise merely
analytical proposition would have no influence on the
views about exchange-value of those who recognize the
distinction, through the course of time, between esteemvalue and exchange-value.
Similarly any one else who says that by “ value” he
means always exchange-value, would be equally justi

114

FOUR KINDS OF ECONOMIC VALUE

fied in maintaining that only in case money is stable in
exchange-value is it stable in “ value.” And this further
merely analytical proposition need have no more in
fluence on the views of those who distinguish cost-value
and esteem-value from exchange-value and who desire
money to be stable in the one or the other of these kinds
of value. The merely arbitrary restriction of the term
“ value” to one of its kinds cannot properly serve as the
basis of argument.
Some amplification of what is here advanced is needed.
This leads to two more subjects.

XXI
PRACTICAL USES OF THE MEASUREMENTS
F ir st , w e m u st exam in e som e o f th e p ractical uses to
w h ich th e m easurem ents o f th e kin d s o f v a lu e can be
pu t.

The practical use of a measurement of the use-value
of things, and in particular of money, would seem to be
small. In economics, beyond recognizing that use-value
increases in some proportion with quantity (except in
the case of money, when fully introduced, although in
a growing and progressing population more of it may be
needed for it to serve certain of its purposes well), we
have little need for further information on the subject.
The measurement of cost-value also seems of little im
portance, except to the producers, or from a technical
viewpoint. The variation of the average of all costs of
production may show the variation of the general costvalues of all commodities to their producers, but not
necessarily, as we have seen, to the people at large. The
people at large are affected only as lowered costs are
permitted, by the producers, to increase the quantities
of goods supplied to them — that is, to increase their
real income. When, or to the extent that, this is not
permitted, the measurement of the variation of costvalues is of no use at all in political economy; and when
it is permitted, then it agrees with, and is no better
than, the measurement of the variation of real income.
This latter measurement, which, inverted, is a measure
ment of the esteem-value of all things in general, may
not only satisfy our curiosity as to whether progress has

116

FOUR KINDS OF ECONOMIC VALUE

been made or not, and how great; but it may aid us to
experiment with different régimes — as, for instance,
whether more progress is made under free trade or un
der protection. Now, this information can be obtained
by a combination of two measurements: of the varia
tion of the social esteem-value of money, and of the
variation of the social or general exchange-value of
money, in the following way. It is especially facilitated
if the one or the other of these values of money be found
to have no variation, or to be constant or stable, al
though this is not necessary. These, then, are the two
measurements it is important for statisticians to de
velop.
Let us suppose that these two measurements have
been performed. We may consider two hypothetical
cases.
The first is that over a long tract of time the esteemvalue of money is found (or has been made) to be stable,
that is, the average of money-incomes has been constant.
If, then, the general exchange-value of money, which is
the more or less accurately weighted average of its
particular exchange-values, or inversely of prices, is
found to have risen or fallen to a certain extent, such a
rise (or fall of prices) indicates that the constant aver
age money-income purchases more commodities, on an
average, in the same proportion with the variation of the
general exchange-value of money (or with the inverse
variation of prices) ; and such a fall (or rise of prices)
indicates that the constant average money-income pur
chases less commodities proportionally with that fall
(or inversely with the rise of prices). The quantities of
the different kinds of commodities have had, of course,
to be collected as data for weighting the average of

FOUR KINDS OF ECONOMIC VALUE

117

prices. But only relative, not absolute, quantities need
to be used here, and a rough accuracy is sufficient, so
that this double method virtually accomplishes the
averaging of the variations of quantities with avoidance
of the new and difficult problem of properly weighting
them. Now then, a variation of the average real income
means an inverse variation of the esteem-value of things
in general; and the price of every particular thing at a
later year, compared with its price at an earlier, in
dicates how the social esteem-value of that thing has
varied, while the variation of its general exchangevalue is indicated by a comparison of its price-variation
with the average variation of all prices. All people
desire their real incomes to increase, and the esteemvalues of things to fall. Therefore, if their money-in
comes on the average are stable, which means that
money is stable in esteem-value, they must wish prices
to fall, and the general exchange-value of money to rise.
The second case is that over such a tract of time the
general exchange-value of money is found (or has been
made) to be stable, that is, the weighted average of the
particular exchange-values of money, as also of prices,
is constant. If, then, the esteem-value of money, which
varies inversely with the average variation of moneyincomes, is found to have risen or fallen to a certain
extent, such a rise (or fall of average money-income)
indicates a proportional fall of the average real income,
and such a fall (or rise of average money-income) indi
cates a rise of the average real income; for moneyincome and real income now vary together, since prices
have not changed on the average. The quantities of
goods purchased by the varying money-incomes, and
which constitute the real incomes, have had, as before,

118

FOUR KINDS OF ECONOMIC VALUE

to be collected more or less roughly at all the periods
compared; but the advantages of this double method
are the same as in the preceding case. Thus, in the
present case, while a variation of average money-in
come indicates the variation of average real income, the
general exchange-values of commodities, taken all
together, have necessarily remaine'd stable, and in
dividually, in agreement with their particular exchangevalues in money, have varied inversely with their
prices, some rising and others falling, and their average
esteem-value on the whole has varied inversely with the
variation of the average money- or real income. We
may get the variation of the social esteem-value of any
one thing between two periods by multiplying the
variation of its price (the ratio of its later to its earlier
price) by the variation of the esteem-value of money
(the ratio of its later to its earlier esteem-value). Our
desire with regard to real incomes and to the esteemvalues of things is the same as in the preceding case.
Therefore, if money is stable in exchange-value, people
must wish their average money-income to increase, and
the esteem-value of money to fall.
In these two cases it is not necessary that the general
esteem-value or the general exchange-value of money
shall be stable, since an allowance may be made for any
variation in the one which is taken for the standard.
Here for simplicity we may deal principally with moneyincomes and prices. Numerical examples will also be
more convenient.
In the first case let us suppose that between one year
and another the average of money-incomes is found to
have risen 10 per cent, and the average of prices to have
fallen 5 per cent. Now, if the money-unit were reduced

FOUR KINDS OF ECONOMIC VALUE

119

inversely, so that the new unit was equivalent to 1.10
of the old, the money-incomes would be constant in
this new unit. The later prices would then have to be
reduced at the same rate, by dividing them by 1.10;
and also the average of prices. This last operation is
0.95/1.10 — 0.8636, indicating a fall of average price in
the new unit of 13.64 per cent. Real incomes, therefore,
have risen to the reciprocal of this: 1.00/0.8636 =
1.1579, a rise of 15.79 per cent. Thus the reciprocal of
the reduced average price-variation indicates the varia
tion of the average real income; or, again, the reduced
average price-variation itself indicates the variation of
the social esteem-value of things in general. Of every
commodity also the price-variation being so reduced,
its reduced price compared with its original price will
show the variation of its social esteem-value. Such re
duced individual prices, compared with the prices at
the first year as unities, may be called esteem-value prices,
because they are reckoned in a money-unit correctly
serving as a measure of esteem-value.
In the second case, let us suppose the same actual
variations as before, the only difference being in our
use of them, our intention now being to reduce things
on the supposition that the average of prices had re
mained constant. This can be done by raising the
money-unit from 0.95 to 1.00, so that the new unit
equals 1.00/0.95 or 1.0526 of the old, in which new unit
the average of prices would be constant. Then all in
comes, and the average income, would have to be raised
in the same proportion, being multiplied by 1.0526.
The average money-income, so multiplied, becomes
1.1579, a rise of 15.79 per cent. This is the height to
which the average of money-incomes has risen in a

120

FOUR KINDS OF ECONOMIC VALUE

money-unit stable in exchange-value, and so it indi
cates the rise of the average of real incomes, giving of
course the same figure as before. When the later price
of any one commodity is altered in this way, the recipro
cal of its ratio to the price at the first period indicates
the variation of this commodity in general exchangevalue.1 Individual prices corrected in this way, com
pared with those of the first year, may be called exchange-value prices, because they are reckoned in a
money-unit correctly serving as a measure of exchangevalue.12
As a mere matter of calculation, therefore, it is in
different how money behaves, since the actual figures
may be corrected in either way we prefer to alter them.
But in practice there may be a great difference. In the
1 To be exact, the general exchange-value of a commodity so obtained is
its exchange-value in all things (including itself): see note on p. 60. It is
this general exchange-value of all things in all things that remains constant.
See chapter xiii of The Measurement of General Exchange-Value. The dif
ference between general exchange-value in all things and general exchangevalue in all other things is practically so small as to be negligible. It has only
theoretical interest.
2 In The Measurement of General Exchange-Value, pp. 466-468, 473—474,
these were called “ true prices,” on the assumption that the true function of
money as a measure of value is to serve as measure of exchange-value. There
—pp. 474-477—prices of this sort were worked out in the case of petroleum
for the years from 1863 to 1891; and it was noted that Laspeyres had worked
out some, and Sauerbeck and D ’Avenel had touched upon the subject. It
may be added that Jevons in his The Coal Question (1865, p. 62), remarked
that general prices between 1794 and 1860 had fallen from 100 to 62, and the
price of coal had risen from 100 to 120, so that, compared to general prices,
the price of coal had risen from 62 to 120, or had nearly doubled. This means
that coal, rising 20 per cent in actual price, had risen 93 per cent in general
exchange-value. More recently Professor Irving Fisher has worked out such
prices on several commodities, calling them prices “ in terms of general pur
chasing power” or, more simply, “in terms of commodities.” Stabilizing the
Dollar, pp. 41, 218.

FOUR KINDS OF ECONOMIC VALUE

121

above hypothetical cases we have altered only the static
relations of things after any dynamic influences affect
ing them have already had their effect, and we have
taken into consideration only the function of money as
a measure of value. But money, in its functions as a
medium of exchange and a store of value, has dynamic
influences upon the course of things. In the two later
modifications of our first suppositions, we might have
supposed that, instead of average money-incomes rising
10 per cent and average prices falling 5 per cent, money
had been made stable first in esteem-value, and secondly
in exchange-value. The first change could be brought
about by diminishing the volume of money (including
the quantity not only of legal-tender money, but of
credit substitutes, with allowance also for changes in
the rapidity of their circulation) so that average moneyincomes should be reduced by one eleventh; and in the
second, by increasing the volume of money so that
average prices should be raised by one nineteenth. But
if these changes in the money conditions had actually
been made, the resultant individual incomes and prices
would have been different, with an effect upon their
averages. A mere change in our idea of the money-unit
at the later period does not alter the relative values and
quantities of things, since we are merely correcting the
inferences we draw from the way money has functioned
as a measure of value. But actually to make the moneyunit stable in either of the two kinds of value, this
operation, by altering the volume of money, affects the
functioning of money as a medium of exchange and as a
store of value, and so has effect upon actual exchanges,
upon the distribution of products, and even upon the
quantities produced. In an era of industrial progress,

122

FOUR KINDS OF ECONOMIC VALUE

for money to remain constant in esteem-value or in
exchange-value, the volume of money must change
differently; and these different changes in the volume
of money have influence, and different influence, upon
the progress itself.
This difference of influence becomes apparent from
the following considerations. We may assume the con
tinuance of industrial progress, with falling costs and
increasing quantities of products per capita — in other
words, increasing real incomes and falling esteem-values
of commodities. This being presupposed, if money were
stable in esteem-value, money-incomes being constant
on the average and prices falling, this would benefit all
receivers of fixed money-incomes, including creditors,
bond-holders, holders of long leases, persons whose
salaries or fees are fixed by custom or by law, e. g., office
holders, who would resist reductions, and, to a less ex
tent, wage-earners, who would have the advantage of
position. Among these are idle rich men and women,
and the passive elements in production, who would be
favored, their real incomes automatically rising without
effort on their part; while the active factors in produc
tion, entrepreneurs and their partners (share-holders)
would be discriminated against, they having to give up
some of their gains, produced by their own efforts, to
the others. On the other hand, if money were stable in
exchange-value, average prices remaining constant and
money-incomes rising, all this would be reversed. Now
borrowers, share-holders, entrepreneurs would get the
increased real income, while the others would not have
their incomes increased, except as they might get a
slight rise of the interest rate, and might succeed in
getting some raise of salaries, wages,' and fees. These

FOUR KINDS OF ECONOMIC VALUE

123

latter, however, — certainly all creditors qua creditors,
— are not discriminated against. The active agents of
progress are merely getting the gains they have them
selves made, while the passive recipients are stationary,
except as they may turn to and become active agents,
whereupon they may make the gains that the others do.
Of course rising prices may intensify these effects, only
they now do discriminate in favor of the active elements
and against the others, with injurious effects; and fall
ing prices, as the fall approaches stability of money in
esteem-value, change the effects into the former, and if
the fall goes still further, intensify their injuriousness.
These contrary influences, arising from a difference in
the distribution of gains, must affect differently the
production itself, hindering or promoting its progress,
since the one favors idleness more than does the other.
Even wage-earners can only receive less in the former
case, since they get their wages from the entrepreneurs,
but if these have to give up more and more to their
creditors, there will be less for them to distribute in
wages. It is obvious that progress cannot be the same
under these two different behaviors of the monetary
system.
The question before us is somewhat like the question
between the Ptolemaic and the Copernican systems of
astronomy. Before Galileo and Newton discovered the
laws of motion and of force, it was, from the purely
geometric or static standpoint, a matter of indifference
whether the sun were viewed as revolving round the
earth, or the earth were viewed as revolving round the
sun. But since the discovery of those dynamic laws, it
has been necessary to adopt the latter view. Similarly,
although statically it is indifferent whether money be

124

FOUR KINDS OF ECONOMIC VALUE

stable in esteem-value or in exchange-value, it may be,
dynamically, very important that it should be stable in
the one way rather than in the other. In the Middle
Ages, when the supply of gold and silver was not suffi
cient to furnish all countries with a satisfactory cur
rency, and paper-money substitutes were impossible,
peoples struggled to get as much of the precious metals
as they could and set great importance upon them as
instruments of wealth. Later, when these metals be
came abundant, and especially when they were supple
mented by paper notes, ideas changed and a tendency
set in to disparage the importance of money. Money,
it was said, is a mere go-between, and economists tried
to leave it out of sight as much as possible. All that was
necessary, they said, was to keep the money-unit a
fixed quantity of metal. A large or a small quantity of
money was indifferent, as people could get along as well
on a high level of prices as on a low level, or conversely.
A change of level also seemed to many of them of little
importance. This latter view could be entertained only
because during the last two centuries, during which
economics was studied as a science, money was actually
varying not very much in its values, and mostly in the
less harmful of the two directions. Also there were
periods of a score or so of years when it gradually fell in
exchange-value, and then again rose in exchange-value,
so that those who were injuriously affected and who
argued for a change in any one of these periods, hardly
had time to get a hearing before conditions changed,
they were pleased, others were injured, and the argu
ments had to be done over from the opposite side. Dur
ing these times, most (though not all) economists held
the static view, and ignored the dynamic influences of

FOUR KINDS OF ECONOMIC VALUE

125

rising, falling, or stationary average prices upon the
different classes of society, or minimized it. The acci
dental fact that during these times these influences were
weak enough to permit their being overlooked, ought
not to have determined opinions. It is unworthy of
economists, if they pretend to be scientists, not to take
a broad and comprehensive view of all possibilities.
This brings us to our final subject.

XXII
IN WHICH KIND OF VALUE OUGHT MONEY
TO BE STABLE ?
A lmost all economists have maintained that money
ought to be stable in “ value.” But in which kind of
value ought money to be stable? Obviously this ques
tion could not come into consciousness until the distinc
tion between the different kinds was recognized. And
when this distinction is clearly recognized, it becomes
an important question, calling for decision. Between
the non-recognition and the clear recognition there are
many grades, and during this interval confusion of
thought was inevitable. At first a distinction was drawn
only between use-value and exchange-value. Then there
could be no doubt. It may be said that any currency
which fills the channels of trade is equally useful, or is
stable in use-value, however the “ value” of the moneyunit may otherwise vary. Or it may be said that that
is the most useful currency which behaves in the best
way, and therefore money must be variable in usevalue until it becomes stable in some other value. In
either case constancy in use-value is not enough. Then
it was necessary to say — and the economists who made
only this division, all said — that money ought to be
stable in exchange-value. But we have seen that, be
cause of the insufficiency of this dichotomy, the term
“ exchange-value” was made to cover also esteem-value,
or cost-value, or both. For a time cost-value was erected
into the principal claimant. But esteem-value was de
veloped out of the idea of use-value, and then became

FOUR KINDS OF ECONOMIC VALUE

127

a serious rival. Meanwhile exchange-value in its true
sense was always held by some economists to be the
right kind of value for money to be stable in.
There have even been other extraneous kinds of
“ value” set up as the proper kinds. In the old monetary
science a distinction was made between “ intrinsic” and
“ extrinsic” value of coin. The former was its metallic
content, the latter the estimation in the money of ac
count set upon it, both regulated by government. When
the metallic content was altered in the coin represent
ing the money-unit, or (which amounted to the same
thing) when the estimation set upon the existing coins
was altered, — to-day when gold has a variable agio in
the current paper-money, — should debts be paid in
the same “ intrinsic” or in the same “ extrinsic” value
— in the same amount of metal or in the same number
of money-units? The answer to this question is unde
terminable, since the question does not involve any of
the real economic kinds of value. Those who answer
that debts should be paid in the same “ intrinsic” value
are the holders of the bullionist theory of money, which
pretends that the money-unit is a certain weight of a
certain metal, to which it must be kept through thick
and thin, always. But we all know that if alchemy were
successful, these bullionists would be among the first to
clamor for a change. On the other hand, those who
would allow the government to alter the weight of the
metal in the coins at random, and maintain that always
a dollar is a dollar and so is the same thing, are equally
foolish. It may be added that there have been advo
cates even of the opinion that money should be stable
in “ value” as expressed “ in the street,” that is, it
should be supplied in such quantities as to keep constant

128

FOUR KINDS OF ECONOMIC VALUE

the rate of interest. This needs only to be mentioned,
to be rejected.
What is ultimately, fundamentally, though often con
fusedly, wanted by every sensible person is that money
should be stable in “ value” in one of the four kinds of
economic value. Use-value being set aside, there remain
three kinds. We have seen that in the course of time
cost-value and esteem-value keep pretty close com
pany, although they do part company when monopolies
become common. Hence these two have often been com
bined into one as a labor-standard, over against which
is set up a commodity standard, advocated by those
who wish money to be stable in exchange-value. Be
tween these two camps most clearly lies the bone of
contention. If the victory should be won by the laborstandard, it would then remain a fine question whether
money should be stable in cost-value or in esteem-value;
and the decision could hardly fail to be in favor of es
teem-value. If the commodity standard wins, there
will be no need of bothering about the distinction be
tween the other two.
One great error that has run through much of the
argumentation on this subject, due to indistinct and in
complete recognition of the different kinds of value, has
been to use the explanation of exchange-value as settling
the question. Those who hold with Ricardo that ex
change-value is regulated by cost-value, have generally
held that cost-value is “ real value,” as they express it,
and have immediately concluded that money ought to
be stable in cost-value. Ricardo himself did this because
he had only a vague conception of esteem-value, and
had no knowledge how exchange-value could be meas
ured. Some, but fewer, of the followers of Jevons, Wal

FOUR E3NDS OF ECONOMIC VALUE

129

ras, and Menger, led by Menger himself, have held that
because exchange-value is regulated by esteem-value,
therefore esteem-value is “ real value,” and money ought
to be stable in esteem-value. The latter have less justi
fication than the former for arguing in this way, as they
know the other kinds of value, and they know, or ought
to know, how exchange-value can be measured. But
the explanation about the cause of exchange-value, and
in particular about the exchange-value of money, is one
thing, and the opinion about the proper behavior of
money, or of prices, through the course of time, is an
other, and there is no proper connection between the
two. Exchange-value exists, however it be explained;
and money may be stable in exchange-value just as well
as it may be stable in cost-value or esteem-value. If all
commodities are falling in cost-value or esteem-value,
and money falls in these values equally with their aver
age fall, then money is stable in exchange-value. We
desire all commodities to fall in these values; then why
should we not desire money also to fall in these values,
and thereby to remain stable in exchange-value? On
the other hand, we might desire it to remain stable in
one of these values, and therefore to rise in exchangevalue. The question is: which of these desires is the
more reasonable one? Evidently the explanation why
money behaves in one or the other of these ways should
have nothing to do with settling this question. And to
call one of the kinds “ real value” is merely to express a
preference, and does not constitute an argument.
Another great cause of error of argumentation on this
subject has been the desire of the bullionists, especially
of the gold bullionists, to defend their standard by any
argument that comes readily to hand. Ricardo and his

130

FOUR E3NDS OF ECONOMIC VALUE

immediate adherents wrote in a period when prices
reckoned in gold were falling, and the idea that they
might be falling in agreement, more or less exact, with
the fall of costs, also known to be going on, no doubt
helped to make them satisfied with a money-unit at
tached to a weight-unit of gold. This satisfaction was
more consciously expressed in the later period of falling
prices from 1873 to 1896. Then the bimetallic contro
versy raged. The advocates of bimetallism wished, by
injecting silver into the circulation, to make money
more stable in “ value,” meaning exchange-value.
Their opponents maintained that money was stable in
“ value,” some of them meaning cost-value and some
of them esteem-value. Between these opposing parties
the terms “ appreciation” and “ depreciation” were
much bandied about. These terms, of course, contain
the same ambiguities as the term “ value.” They may
mean a rise and a fall respectively of esteem-value,
cost-value, or exchange-value.1 We might be captious
and claim that as the root-word in both the terms is
pretium or “ price,” they refer properly to rise and fall of
exchange-value. But nothing would be gained by this
preciosity. Now, many of the gold-advocates of that
period tried to make out that the average fall of the
cost of production of commodities had been about the
same as the average fall of prices, and then they said
1 There is still another common meaning of these terms. If a metallic
currency be recognized as the standard (without regard to the fluctuations of
its values), then a paper-money pretending to agree with it, if it does not,
but has a lower value relatively to that primary money, is “ depreciated.”
Such paper-money rarely is “ appreciated,” rising above the value of the
other, though this has happened. This is depreciation or appreciation within
money itself, between two kinds of money; and so is something entirely dif
ferent from the question before us.

FOUR KINDS OF ECONOMIC VALUE

131

that commodities had “ depreciated,” but gold had re
mained stable in “ value” ; and this they thought
enough. But if their contention was correct, it meant
only that commodities had fallen in cost-value, and that
money had remained, as they roughly estimated, stable
in cost-value. Others were content that wages had re
mained fairly stable, since, by Adam Smith’s and Malthus’s laborers’ income standard, they too thought
this indicated stability of money in “ value,” though it
was but a very imperfect measure of esteem-value.
Over against these many of the bimetallists merely
asserted that a fall of prices ipso facto was an “ apprecia
tion” of money. This is true of “ appreciation” in the
sense of a rise in exchange-value. The two sets of op
ponents, while seeming to be arguing with each other,
were merely describing the same facts in terms with
different meanings.
Their amplifications of these specious arguments
generally came but little closer to reality. The gold
advocates urged that since science and industry were
lowering costs, and because every article, as it was
cheapened, fell in price, therefore everything was as it
should be; or, in other words again, that the greater
abundance of goods, the greater supply, and the con
sequent lower prices of all things were a sign of advanc
ing civilization.2 The reply made by the bimetallists
was that if gold, too, had been cheapened by science
and industry, and become more abundant under the
impetus of advancing civilization, the prices of com
2 In the preceding period of rising prices before 1873, some of the gold
advocates had even held that it is the function of civilization to increase the
demand for goods, and therefore to raise prices! In our present period some
faint echoes of this have been heard.

132

FOUR KINDS OF ECONOMIC VALUE

modities on the average would have been restored, and
gold, being carried down the stream of progress on the
same level with the commodities, would have remained
stable in “ value,” that is, in exchange-value; and as
this was not the case, things were not as they should be.
To counter this, the other party asserted that gold
ought not to be cheapened, and its abundance ought
not to be increased, for the very reason that it was used
as the standard of “ value” and should remain stable in
“ value,” meaning cost-value or esteem-value. Still,
however, each side was only adducing its own prefer
ences for stability in the one or the other kind of value.
Nor could anything decisive result from each side trying
to cry down the other by asseverating that value is its
kind. Value is any of the four kinds, any one being as
good as any other. As a matter of fact, most of those
writers and speakers on the opposite sides did not use
these specific terms, and did not even recognize clearly
either their own or their opponents’ meanings. Their
arguments were mostly at cross purposes, and fell on
deaf ears. Nothing better than that futile debate shows
the need of the different terms for the different kinds of
value.
It surely is not sufficient to inquire whether during
any tract of time our money has been stable in one or
another of the kinds of value, and to pronounce it good
if only it has been stable in one of them, no matter
which one. Yet this is frequently done. At the present
day there can be little doubt that one reason why so
many economists are slow in supporting Professor Irv
ing Fisher’s project for stabilizing the dollar, is that
they have a notion at the back of their heads that when
prices begin to fall again, there are other kinds of value

FOUR KINDS OF ECONOMIC VALUE

133

in which money may be stable just as well as in ex
change-value. Then why bother specially about stabiliz
ing it in only one kind of value? It is this distraction
of ideas that needs to be overcome. The real question
calling for decision, we may repeat, is as to which kind
of value money ought to be stable in, — which kind
ought we to desire it, and, if possible, make it, stable in?
This involves the inquiry, in which kind of value is the
stability of money the best for all concerned, for the
whole people? This inquiry is partly one of morality,
and partly one of expediency. In the bimetallic contro
versy some of the disputants, especially on the propo
nent side, did raise their arguments to this higher plane.
But no conclusion was reached. Must inconclusiveness
reign for ever on this subject? Surely economists will
not be doing their duty if they allow it. Let us briefly
review some of the arguments.
Morality requires that parties to a contract carry it
out: the debtor must pay what he promises to pay. In
a contract to pay money what is it the debtor promises
to pay? On the face of it he generally promises to pay
so many money-units. Can he then honestly solve his
debt by paying the money-units that the government
happens to make current? No, say the bullionists: the
money-unit is so many grains of gold, and these he must
pay. But as a matter of fact, not one debtor in ten
thousand knows how many grains of gold he is contract
ing to pay. He regards money as a means to get other
things, and to everyone a dollar is a composite idea re
presenting so many things he can purchase with it: it
is a purchasing power, an exchange-value, or simply a
“ value” as it is usually most indefinitely called. A
debt, then, must be repaid in the same value that was

134

FOUR KINDS OF ECONOMIC VALUE

lent, plus interest.3 This is what morality calls for. Had
people always used the definite term “exchange-value”
there would probably have never been any question
but that money ought to be stable in exchange-value.
But as “ value” has other senses, the bullionists who
want the gold standard at all hazards have been able
to claim that a repayment in the given number of grains
of gold having an equal value of some other kind is
proper. Then, too, the idea that labor is a subject of
exchange has come in to aggravate the confusion.
Granted that money ought to be stable in exchangevalue or purchasing power, it is said that it ought to be
stable in its power to purchase labor, which means that
wages ought to be stable. Adam Smith and Malthus,
as we know, have asserted that this is sufficient: the
monetary system is behaving properly as long as wages
are stable, no matter how the prices of commodities
may be changing. This, of course, taking as it does only
the standpoint of the employer, is absurd. Less absurd,
and therefore more likely to be accepted, is the notion
that the purchasing power of money ought to be stable
over both commodities and labor. Suppose this were
so — suppose prices fell and wages rose in such a way
as to compensate each other. This could generally hap
pen only in case labor were becoming more efficient.
3 It not infrequently happens that creditors stipulate for repayment in a
unit of so many grains of gold. Here every idea of value seems to be ex
pressly excluded. But the creditors do this because they think gold more
likely to be stable in some kind of value than a possible paper substitute.
If a government really provided a paper-money stable in the proper kind of
value, it might be advisable to forbid such contracts in order to prevent
creditors, expecting gold to rise in this value, from taking advantage of the
need and ignorance of the debtor to impose a harder bargain than the debtor
himself thinks he is getting.

FOUR KINDS OF ECONOMIC VALUE

135

Then a debtor would repay in money able to purchase
more commodities but less labor of greater efficiency.
Certainly this money would have risen in purchasing
power, all told. The commodity element alone has to
do with exchange-value,4and labor is a factor in esteemvalue; and such a standard would be a mixture of two
kinds of value. As for esteem-value itself, is it expected
that debts should be repaid in equal esteem-value?
Take the case of individuals: a poor young man may
borrow, and may make repayment after growing rich,
his creditor, too, perhaps having advanced in the same
way. Nobody expects that he should augment the
quantity of money he repays so as to keep the sum of
the same esteem-value to himself and to his creditor.
But if this is not expected of individuals, how can it be
expected, and demanded, of all the individuals who
compose a people which is growing richer and whose
money, though stable in exchange-value, is falling in
esteem-value?
Nor is there any good reason for taking a half-way
stand between the two standards. Suppose a person
loaned money to another in 1800 and their descendants
settle the debt in 1900, and that between these dates
commodities have become twice as abundant on the
whole, per capita, so that their average esteem-value
(and probably their average cost-value) has fallen by
half, or by 50 per cent. If money-incomes had remained
stable and prices had fallen by half, the debt would be
paid in the same esteem-value, the creditor receiving
twice as great purchasing power as his ancestor loaned.
If money-incomes had doubled and prices remained
stable, the debt would be paid in the same exchange4 See the footnote on p. 61.

136

FOUR KINDS OF ECONOMIC VALUE

value, the debtor expending half the labor-cost of the
original loan. To let the two parties share the gain
equally, money-incomes should rise, not half way, but
to the geometric mean between 1 and 2, or from 1.00 to
about 1.41J, and prices should fall (on the average, of
course) from 2 to 1.41§, or from 1.00 to 0.70f. Then
both the average of prices and the esteem-value of
money would have fallen by about
per cent. (A
certain change in the volume of money would have both
these effects.) Then, as the debtor’s productive power
presumably has increased by 100 per cent, and as the
money-capital repaid purchases 41| per cent more than
it did, the real capital is produced by him at a laborcost equal to 70| per cent of the labor it cost the credi
tor’s ancestor to produce it. Thus the creditor gains as
from 1.00 to 1.41J, and the debtor gains as from 0.70f
to 100, which is the same as from 1.00 to 1.41 J, and the
gain of both is the same. But, as it was not the passive
side of the loan that created the improvements, why
should the creditor gain at all? Moreover, money in this
case would not be stable in any kind of value, and
economists would have to revise their almost unani
mous declaration that money ought to be stable in
value.
The arguments from expediency are either a priori or
a posteriori. The former go back to the considerations
reviewed near the end of the preceding article. In an
era of advancing prosperity money stable in exchangevalue does not discriminate in favor of debtors, but
money stable in esteem-value (or in cost-value), and
rising in exchange-value, does discriminate in favor of
creditors. Money falling in exchange-value, on the
other hand, discriminates in favor of debtors. Now, an

FOUR KINDS OF ECONOMIC VALUE

137

impartial observer should not desire to favor either
creditors or debtors, although, if the choice be neces
sary, it would be wiser to incline toward the latter. He
should not prefer, therefore, a monetary system which
takes the gains properly belonging to debtors and gives
them, either entirely or in part, to the creditors. This
is not a question between social classes, as debtors and
creditors do not fall into such, many persons being both
at once. But creditors qua creditors are passive and
merely wait to have their loans paid back in interest and
principal. Debtors qua debtors have to hustle to pro
duce the wherewithall to meet their obligations. Indus
trial debts are incurred to be used as productive capital,
and it is the borrowing capitalists that are interested to
make or adopt improvements, while lending capitalists
are not concerned how the money is obtained which re
pays their loans. Therefore a money stable in esteemvalue (or in cost-value), which gives the gains accruing
from improvements to idle lenders without effort on
their part, will induce many persons to retire fromactive
enterprise, who, under a money stable in exchangevalue, which leaves to the workers all the gains they
have produced, would be encouraged to become work
ers. Therefore the former system tends to retard, the
latter to further, industrial progress. Hence the opinion
has been more or less prevalent, ever since it was pro
pounded by Hume nearly two centuries ago, that rising
prices are better than equally falling prices. If that is
so, a fo rtio ri stable prices, which favor neither of the
parties to a loan, are better still, besides being more
just. And experience — this is the a posteriori argu
ment — seems to bear out this opinion. Unfoi’tunately
experiments cannot be made in this subject. History

138

FOUR KINDS OF ECONOMIC VALUE

alone can be appealed to. History offers few eras of
long continuance in which money has been stable in any
kind of value. But it offers many eras in which money
has fallen in exchange-value (and still more in the other
values), and others in which it has risen in exchangevalue (and been more nearly stable in esteem- or costvalue, as we have seen claimed by the gold-advocates).
And history shows that even the former eras have been
eras of greater prosperity and contentment than the
latter.
No more details need be gone into here. Only one
warning may be added. Malthus argued for his wagesstandard (or a partial esteem-value), that it should be
adopted because it was the only workable one. An
argument of the same sort has now been advanced by
some economists that the commodity-standard (of ex
change-value, by means of index-numbers) should be
adopted because it is the only workable one. This is not
right. If the income-standard (of esteem-value proper)
were the correct one, it should be adopted in spite of its
difficulty of operation, and the endeavor of economists
and statisticians should be turned to make it workable
by obtaining a good method of measuring esteem-value
and of collecting the necessary data. It is perhaps not
so much a piece of good luck that the measurement of
exchange-value is now the easiest, as it is a result of the
endeavors to make it so, themselves inspired by the
belief that this is the kind of value in which money ought
to be stable. At all events, the proper purpose of meas
uring the exchange-value of money is to prepare the
way for somehow rendering money stable in this value.
But first of all, this value must be cleared of confusion
with the other kinds of value.

Full text of The Four Kinds of Economic Value

FRASER