What I have to say may, I think, best be developed by looking at a typical
instance that illustrates price determination through social control in a
particularly noticeable manner: the case of the settlement of wage disputes
by means of a strike.
According to the accepted formula of modern wage theory, based on the
marginal-utility theory, the amount of wages in case of free and perfect
competition would be determined by the "marginal productivity of
labor," i.e., by the value of the product that the last, most easily
dispensable laborer of a particular type produces for his employer. His wages
cannot go higher, for if they did, his employer would no longer gain any
advantage from employing this "last" laborer; he would lose, and
consequently would prefer to reduce the number of his workers by one; nor
could the wages be substantially lower, in the case of effective competition
on both sides, because the employment of the last worker would still produce
a substantial surplus gain. As long as this is true, there would be an
incentive to the further expansion of the enterprise, and to the employment
of still more workers. Under an effective competition among employers this
incentive would obviously be acted upon, and could not fail to eliminate the
existing margin between the value of the marginal product and the wages in
two ways: by the rise of wages, caused by the demand for more workers;
and by a slight diminution of the value of the additional produce, due to the
increased supply of goods. If these two factors are allowed to operate
without outside interference, they would not only delimit wages, but actually
fix them at a definite point, owing to the nearness of these limits, let us
say for instance at $5.50 for a day's labor.
But let us now assume competition to be not quite free on both sides, but
that it be restricted, or eliminated, on the side of the employers; either
because there exists only one enterprise of that particular branch of
industry over a large territory, thus giving it natural monopoly over the
workers seeking employment, or because there is a coalition of entrepreneurs
within that industry, who mutually agree not to pay their workers a wage
higher than, let us say, $4.50. In either case, this coming into play of
"control," a superior power of the employers, will certainly
suffice to lead the wages to be fixed at a point below $5.50, say at $4.50,
other conditions remaining equal.
How would this correspond with the standard explanation offered by the
marginal-value theory? The answer is not difficult. In fact, the solution has
been repeatedly stated in the fairly well developed theory of monopoly
prices. I shall merely try to restate the familiar arguments in a clear and
systematic manner.
We have before us a case of "buyers' monopoly." The widest
margin within which the monopoly price can be fixed is limited, from above,
by the value of the labor to be purchased by the entrepreneur exercising that
monopoly, and from below, by the value of unsold labor to the laborer
himself. The upper limit is determined by the value of the produce of the last
worker, for the reason that the entrepreneur will not assume any loss from
the last worker he employs and that the same amount of labor cannot be paid
for in unequal amounts. This upper limit of the possible wage would, in our
illustration, be $5.50.
More is to be said in regard to the lower limit. The very lowest limit is
determined by the utility that would be left to the worker if he were not to
sell his labor at all. It is thus, primarily, the use-value to the worker of
his own labor, provided he can make some use of his labor for himself alone.
In thinly populated new countries, with an abundance of unoccupied land,
where everybody may become a farmer at will, this labor-value might represent
quite a considerable amount. In the densely populated "old"
countries, however, this limit is extremely low, because most of the workers
lack capital, and can hardly ever profitably utilize their own labor as
independent producers.
A worker who has accumulated some savings may find some compensation for
not selling his labor in the escape from discomfort and hard work, or in the
enjoyment of rest and leisure. Those who have any such means of subsistence
will figure out just what minimum of wages would compensate them for the
effort of working. To those who have nothing to fall back on, the marginal
utility of a money income to be gained by working is so extremely high that
even a very low wage will be preferred over the enjoyment of leisure.
In order to illustrate this with actual sums of money, let us assume this
lowest limit, the use-value of labor and the enjoyment of leisure, to be very
low, say $1.50. This amount may be even far below the minimum of subsistence,
which, for well-known reasons, determines the lower limit of the possible permanent
wages without, of course, determining temporary wages or those of each
individual case.
But there may also arise other intermediate wage levels. In the foregoing
illustration we have excluded all competition among the employers in that one
particular branch of industry. If such competition were existing, it would
inevitably force up the wages to the upper limit of $5.50; but even in its
absence, there would still remain a certain amount of outside competition,
namely with employers in all the other branches of industry. This means that
the worker in our particular industry still has the alternative of escaping
the very low wage offered to him in his own line, by switching over into
other branches of production, although a number of circumstances may greatly
reduce the gains to be expected from this expedient. To change from one
occupation, for which one has been trained and adapted, into another, is
likely to result in less productivity, and the maximum wage level attainable
in the new occupation will be likely to remain far below $5.50.
The curtailment in wages will vary for each worker entering into a new
branch of production according to his adaptability, or his ability to perform
a different kind of skilled labor. The most painful cuts in wages will be
suffered by that probably largest portion of the workers, who are not
adequately trained to perform any other kind of skilled labor, and who will
have to switch over from "skilled" into "unskilled"
trades, and accept a poorer position in some type of common labor. Still
another slight lowering of the wage level may result from the fact that the
influx of new workers into that occupation may force down slightly the
marginal productivity of the last worker, and thus lower the wage level for
all.
Under the influence of all these circumstances we would now have to assume
that the various workers set for themselves a series of individual minimum
limits, below which no one would allow his wages to be reduced by the
monopolistic pressure of the entrepreneurs. To illustrate these various
gradations of minimum wages, let us assume the minimum of existence to be
$3.00, which, as has been said, would represent not the temporary, but the
permanently possible lowest wage level. The wages obtained by the most common
type of labor would thus be very near to $3, say $3.10. A smaller and smaller
number of workers could find employment in other occupations, as the wage
rate increased in the following ascending sequence: $3.50, $3.80, $4, $4.20,
$4.50, $4.80, $5. Note, however, that the upper limit of this wage scale
would still remain below the marginal product of the original occupation,
thus below $5.50.
What effects and limitations will result from this state of affairs in
regard to the monopolistic fixation of wages within the original widest zone
of $1.50 to $5.50?
Let us assume, to begin with, that the monopolistic entrepreneurs use
their power in an unrestricted, purely selfish policy, unaffected by any
considerations of altruism, or consideration of public opinion, uninfluenced
by any apprehension that the workers might fight back through means of a
labor union or strike, and convinced that they are absolutely assured of an
atomized, effective competition among the individual workers. Under such
premises, the rate of wages would be fixed according to the general formula
applying to a purely selfish monopoly, already mentioned before in another
connection: they would be fixed at that point which promises the largest
returns, after a careful consideration of all circumstances, and with due
regard to the inevitable fact that with changing prices, the amount of goods
to be disposed of profitably will change, only that in the case of a buyers'
monopoly the results are exactly opposite to that of a sellers' monopoly. Or
stated concretely: the lower is the wage rate fixed by the monopolist,
the smaller will be the number of workers available, and from a
correspondingly smaller number of workers will the entrepreneurs be able to
collect that increased return which might accrue from pushing the wage scale
down below the value of the product of the marginal laborer, i.e., below
$5.50; in fact, this value might even increase through a reduction in the
output, which would cause a rise in the price of the finished goods.
Of course, there may again enter certain counteracting tendencies, such as
increasing costs, with the restricted expansion of the enterprise, the growth
of overhead expenses, etc. With an increase in wages (which, however, we
always assume to remain below the marginal product of $5.50) the gain per
laborer would decrease; but, to offset this, the number of workers from which
that gain can be made will increase, or even be brought back to normal. From
these considerations, it would be most unlikely that the monopolists could
fix the wage rate at $1.80 or $2.00 or at any point below the minimum of
existence of $3, both because this rate would not be likely to remain in
force, and because it would be lower than the wage paid outside for common
labor, and therefore would at once cause the majority of the workers to
withdraw into those unskilled occupations which, in our illustration, receive
$3.10. This danger will diminish gradually with each increase in the wage
rate, and disappear almost entirely at some point, say at $4.50, at which
only a few exceptional workers might find it possible to obtain higher wages
in other skilled occupations, if such be open to them at all. Under the
assumed circumstances, the danger of men withdrawing would have almost
disappeared, and a successful attempt might be made by the monopolistic
employers to fix the rate of wages at this point, without running the risk of
any considerable restriction of output caused through a shortage of workers.
Two other considerations might influence an intelligent monopolist to
exercise his power "with restraint." First, a wage rate remaining
far below that of other skilled occupations may, if only in the long run,
lead to a shortage of workers, for while the laborers accustomed to their
occupation might hesitate to change their job owing to the difficulties of
transition, the new supply would fall off. Secondly, too high a rate of
profit per worker would exert too powerful a strain on the employers' union,
and is likely to lead to a dissolution of the coalition by those members
wishing to expand their business, or to the formation of new enterprises
outside of the coalition, thus creating new competition, likely to cut down
prices and to raise wages. Generally speaking, the fear of outside
competition forms perhaps the greatest safeguard against too unscrupulous a
use of monopolies preying on the general public.
I hardly need to re-emphasize the fact that if, under such conditions,
through the "control" of the monopolists the wage level were to be
reduced from $5.50 to $4.50, this would, from first to last, happen by virtue
of and in conformity with the elements of the price-law, as formulated by the
marginal-value theory. It is in consideration of these elements that both
contending parties would fix the price at that level, by
"delimiting" it from above and from below. By such action, no
"fixed" price would be determined, but merely a wider price-range,
as distinct from the case of perfect competition on both sides. The
monopolists might just as well decide upon $4.20 or $4.80 than upon $4.50.
This situation is explained by the fact that several factors entering into
the calculation, such as the number of workers likely to drop out at a
certain wage level, or the probability of outside competition, are not
definitely known, but only to be conjectured. The monopolists would naturally
try to select the most favorable point of the wage scale; but, owing to the
uncertainty of so many elements entering into the fixation of this optimum point,
there results a certain more or less elastic zone for its approximate
location, just as in ordinary market competition for prices, when
negotiations are carried on with covered cards, traders less experienced or
less shrewd commit errors in sizing up inside marked situations, so that
actual prices are caused to fluctuate over a wide range around the
"ideal" market price.
Let us now turn to the other case, equally interesting and complicated,
the influence of "control" exerted by labor unions, through the use
of their instrument of power, the strike. Let us retain all previous
assumptions with the same figures as above: $5.50 for the value of the
product of the "last" worker, $1.50 as the personal valuation to
the workingman of his unsold labor, $3 as the minimum of existence, etc., and
introduce into our assumed case only one novel element, namely that the
workers of the industry under discussion do not compete against each other,
but that they be unionized, and thus be in a position to enforce their joint
demand for higher wages by means of a strike.
Now I do not for a moment deny that this coming into play of
"power" on the part of the workers may profoundly influence the
price of labor. It might even raise it not only above the level of $4.50,
reached in the case of reduced competition among the monopolists, but even
beyond the level of $5.50, which would have been attainable under perfect
competition. This last fact is particularly noteworthy and striking, for
hitherto we had regarded the value of the marginal product of labor, precisely
that $5.50, as the upper limit of the economically possible wage, and at
first sight it might look as if "power" could actually accomplish
something in contradiction to the price formula of the marginal-value theory,
something that did not conform to this law, but disproved it.
Here now enters into our explanation the distinction between marginal
utility and total utility, i.e., the fact that the value of a total aggregate
of goods is higher than the marginal utility of each unit, multiplied by the
number of units contained in the total. The fundamental question in the
evaluation of a commodity or an aggregate of goods is always how much utility
may be derived from the command over the good to be valued. Under the assumption
of competition among all the workers, the thing to be evaluated by the
employer is always the labor-unit of each worker. If the employer had in his
employ, for instance, 100 workers, his negotiations with each one of the 100
workers over his wages would merely hinge upon the question of how much
additional profits the employers would make by employing that one additional
worker, or how much he would lose by not employing this one last worker. In
that case we were fully justified in arriving at the marginal utility of each
unit of labor, that is, the increase in output which the labor of the last
one of the 100 workers adds to the total output of the enterprise, or $5.50.
But now this is different: in the case of a joint strike of all the 100
workers, the point in question for the employer is no longer whether he is
going to run his enterprise with 100 or 99 workers, which to him would mean a
difference in the output of $5.50, but whether he is to keep his enterprise
going with 100 workers, or not at all. On this depends not 100 times $5.50,
but obviously much more than that, if for no other reason than that labor is
what is called a "complementary" good, a good which cannot be
utilized by itself alone, without the necessary other "complementary"
goods, such as raw materials, equipment, machinery, etc. If only one man out
of a hundred withdraws from the enterprise, the utilization of the
complementary factors will, as a rule, be little disturbed. One single
operation — the one which can be dispensed with most easily — will be
omitted, or replaced, as far as possible, through a slight change in the
division of labor, so that with the deduction of one man, not more is
lost than the marginal product of one day's labor, namely $5.50.
The withdrawal of ten men would cause a more serious disturbance. But a
changed disposition in the use of the remaining ninety workers would probably
make it possible to find some way for at least the most important functions
to continue unhampered, and the loss again to be shifted to that place where
it is least felt. A continued depletion of the complementary good,
"labor," would make itself felt more and more severely. While the
withdrawal of the first worker would have caused a decrease in the daily
production of only $5.50, that of the second might amount to a diminution of
the output by $5.55, that of the third by $5.60, and that of the tenth by as
much as $6. If, as would be the case in a strike, all the 100 men walked out,
there would be caused a loss, not only of the specific labor product of those
100 men, but additional productive goods would cease to be utilized. The
machinery would have to stand still, the raw materials would lie idle and
depreciate, etc. The loss in the value of the product would increase out of
all proportion, far beyond a hundred times the last laborer's marginal
product.
The loss, of course, would be subject to great modifications, according to
the actual conditions existing in each case. If the idle machinery and
capital do not suffer any other damage by being idle, the additional loss
would merely consist in a postponement of the completion of the respective
products from the capital goods, temporarily not utilized on account of the
lack of the complementary factor of labor. Their produce will be obtained in
an undiminished amount only at a later period, after the resumption of
production. This loss must at least equal the interest on the dead capital
for the period of idleness. It may amount to more, if the delay should
involve added losses, such as the inability to take advantage of favorable
business opportunities, whereby indirect depreciations may be incurred.
But the damage would be still further increased if the specific character
of the idle capital goods should not only cause a temporary delay, but a
definite curtailment in the profits, as for example in the case of perishable
raw materials, such as beets in an idle sugar refinery, or agricultural
products that cannot be harvested owing to the worker's strike, unused animal
power, such as horses, or the water power of an electric power plant. The
enforced shutdown may also threaten the fixed capital investments, as in
mines, where ventilation and water pumps must not stop, lest the entire plant
be destroyed.
How does all this affect the fixation of wages in the case of a strike?
Let us realize, first of all, that although the wage disputes are formally
concerned with the per capita wages for each individual worker,
to the manufacturer it is always a question of obtaining, or not obtaining,
the total labor of these 100 workers. He will either get all of the
workers, or none, according to whether the negotiations lead to an agreement,
or to a break. The decision as to how much wages he can pay at most will thus
hinge on the value that the hundred workers represent to him jointly. The per
capita wage is a secondary item, and is determined by dividing the total
value by the number of workers. To him, this quota represents only an
arithmetical concept, not a value; to him it does not represent the value of a
unit of labor.
But how high is the total value? This is explained by the theory of
imputation. The value of that aggregate of labor is derived from the value of
that amount of products which may be ascribed to the availability of that
particular total of labor, and this again is identical with the amount of the
product of labor.
Here comes into play a remarkable phase of the theory of imputation, which
I recently had to defend in detail against differing opinions.[1] For if the withdrawal of that amount of labor, whose
value we are trying to ascertain, not only prevented the use of that labor
itself, but also stopped the use of other, complementary goods, the utility
of these goods would have to be added to that of labor, regardless of the
fact that under certain circumstances the use of labor might have to be
imputed to its corresponding complementary good, without which the products
could not be obtained.
I shall merely recapitulate here without detailed discussion the various
steps of the argument leading to this conclusion. Fundamentally, the total
value of a whole group of complementary goods is dependent upon the amount of
the (marginal) utility which they possess jointly, and thus, in case of
complementary productive goods, upon the value of their common product.[2]
The distribution of this total value among the various units of the
complementary group may take different directions, according to the different
causation. If none of the units admits of any other use than joint use, and
if, at the same time, no one member contributing toward the joint use is
replaceable, then every single member has the full total value of the entire
group, while the other members are valueless. Each complementary unit is
equally capable of holding either one of the two valuations, and it is solely
the outside circumstances that determine which one of them shall be worth
"everything," by being absolutely essential in the ultimate
completion of the group, or which one is worth "nothing" through its
isolation.
In our case of an impending strike of all the hundred workers, the
employer is threatened with the total loss of the joint gain arising from the
use of the two complementary groups, labor and capital, to the extent stated
above, and this is why in that case he would have to attribute to labor that total
joint utility, including that part which under other conditions might have to
be attributed to the complementary capital goods. His subjective valuation of
labor must be based upon all these things.[3]
Consequently, the upper limit for the highest rate of wages will advance.
For all the hundred workers jointly it will rise beyond the hundred-fold
amount of the single value of each day's labor, that is, beyond 100 times
$5.50, at least by the amount of the interest of the capital left idle and
perhaps even above this, by the amount of the actual loss from perishing or
deteriorating complementary capital goods. Thus, for instance, in case there
be merely a postponement or loss of interest, it would rise above $550, up
to, say, $700 for each day; in case of a direct loss in the utilization of
the complementary goods, it would rise in proportion to the extent to which
an actual loss takes place, perhaps to $1000, perhaps even to $2000 per day. And
the maximum of the economically possible wage level for each individual
worker would thereby rise from $5.50 to $7 or even to $10 or $20. This means
that with any wage level remaining below this maximum, the entrepreneur
would, at least for the time being, fare better than if he were to cease
employing all the hundred men.
This "faring better" need, however, not imply actual profits to
the entrepreneur, but merely a smaller loss than he would incur in the other
alternative — the "lesser evil," which is, of course, to be
preferred to the greater one. This rise of the last possible per capita wage
to $7 or to $20, on the other hand, does not represent the subjective
valuation of one day's labor to the entrepreneur. This has already been
stated in the foregoing and it can hardly be sufficiently emphasized. The
employer would never pay that wage, if it were a question of employing one
laborer only. It represents the hundredth part of the total value of 100
laborers, which is a very different unit from the individual value of each
unit of labor.
In the wage negotiations between an employer and a labor union the range
would thus be limited by the value to the laborer of his unsold labor (i.e.,
the amount of $1.50 as his lowest limit), and by the per capita quota of the
total value of all 100 laborers at the rate of $10 as upper limit, to take
one of the three figures as an illustration.
In our imagined case, direct competition being absent on both sides,
entrepreneur and workers would meet each other within their limits on similar
grounds, just as the two parties of buyers and sellers meet in the case of
isolated exchange.[4]
In theory, it would not be unthinkable nor impossible for the rates to be
fixed at any single point within the wide range between $1.50 and $10. We
have, of course, come to know some circumstances that make it appear rather
unlikely, though not altogether economically impossible, that the wages be
fixed within the lowest section of the zone lying between the absolutely
lowest limit and the minimum of existence of unskilled labor; and for reasons
of similar nature, it is not very likely that the wage rate would be raised
up to a point near the upper limit of $10. That it could not be kept at such
a point for any length of time I shall try to demonstrate in a future
investigation which I consider of special theoretical import. But not even
temporarily will it readily be pushed so high. For any wage level
substantially exceeding the output of the "last worker" would meet
with a strong and increasing opposition on the part of the employers as
involving a loss to them. Before granting such a wage rate, they would
probably prefer to risk the decision of the supreme trial, consisting in
fighting matters out in a lockout or strike; although an intermediate wage,
approximating the actual service of the last worker, might conceivably be
granted by the employers, anxious to avoid the risk of the certain losses
involved in a strike, and the added uncertainty of its outcome. Nor would
workers find it to their advantage to push the wages up to level actually
causing losses to the entrepreneur, for this again might threaten them with a
restriction, or suspension, of work, and force them out of their jobs. Thus
there enters the question about the permanency of wages, which will be
investigated later.
On the other hand, the workers' difficulties will become all the greater by
the strike, the more excessive wage demands they make. The threat from strike
breakers or "scabs" from other branches of industry will increase
with the more favorable terms which the entrepreneur can still grant below
the refused rate of wages. If the striking workers should insist on a wage
rate of $9, a wage of $7 may perhaps already contain a very tempting premium
to scabs and substitutes, who in other occupations requiring similar
qualifications may obtain only a wage of $5.50, corresponding to the output
of the last worker. And once substitutes are employed, the cause of the
strike is usually lost, whereas, in the other alternative, the outcome is by
no means certain.
In a strike, that party wins, as a rule, which, popularly speaking, can
"hold its breath" for the longest time. To the worker, the strike
means unemployment. For the time being the worker may meet this loss by means
of savings accumulated for this purpose, by subventions from strike funds, by
consuming his property, by selling or pawning dispensable goods, or by
incurring debts as far as his credit will permit. With the longer duration of
the strike, these savings will become smaller and smaller until they are used
up. During the period of gradual diminution of savings, the marginal utility
of the rapidly decreasing means of subsistence goes up, more and more of
essential wants go unsatisfied, and more and more of the vital necessities
are neglected, with the increasing shortage of funds.
Finally the point is reached at which the very maintenance of life depends
on a renewal of income through work, if only at a modest wage: at this point
even the most obstinate resistance of the strikers is broken — provided, of
course, that the resistance of the opposite party, the employer, is not crushed
beforehand.
In the ranks of the employers there are the same phenomena. With the
increasing duration of the strike, the desire for a settlement becomes more
and more intense. The idle plant produces no income. Some of the costs of
production and at least the personal living expenses of the manufacturer
continue, and have to be met. If the entrepreneur has a large fortune, these
expenses may be covered from that. If not, then the pressure of the strike
will be felt much more rapidly and intensely. In any case, there are here two
very distinct phases of the effects of the strikes that should be
distinguished. The successive and increasing lack in the means of subsistence
may first threaten the business of the entrepreneur, and then, if there are
no funds left for the most urgent living expenses, his personal existence.
This latter, more intense effect of strikes, will normally arise only in
the most exceptional cases. Nor is it likely, for these and similar reasons
stated before, that in a strike wages will be fixed at the most extreme —
neither at the very lowest nor at the very highest — marginal regions of the
wide range "economically possible," at least for the time being. In
our illustration this zone was assumed to extend from $1.50 to $10, and a
wage rate below $3 would be just as unlikely as one above $8, although, as I
want specially to emphasize, such extreme wage rates are not unthinkable, nor
altogether economically out of question for a short period of time.
Most of what has been said so far is based on obvious and almost trivial
facts and observations which have become sufficiently familiar through common
experiences with strikes. I have merely restated these matters, so to speak,
in the terms of the marginal-utility theory, in order to make plain the
essential point of the theoretical principle under discussion, namely, that
the "influence of power" in the case of strikes, so familiar to all
engaged in industry, is not altogether distinct from, or opposed to, the
forces and laws of the marginal utility theory, but wholly in conformity
and in harmony with these, and that every deeper analysis of the
question, through what intermediate agencies and to what marginal points
"power" may control the course of events at all, must lead into the
more specific exposition of marginal utility, in the theory of imputation,
where the ultimate explanation is to be sought and found.[5]
There is another far more interesting question: When will the terms of
distribution, obtained through means of power, be of lasting effect?
This question is all the more interesting, in that it is by far the most
important one. Even the most ephemeral fixation of prices or wages may have
considerable importance to that group of individuals or for that short span
of time that happens to be affected by it. On the other hand, these temporary
fixations mean little or nothing for the permanent economic welfare of the
various social classes; just as the classical economists have held long-trend
prices to be far more important and challenging than momentary fluctuations;
thus Ricardo hardly touched upon the latter, and found it worthwhile only to
elaborate the theory of long-trend prices. Similarly, in the theory of
distribution, paramount importance is attached to the permanent trends
according to which the shares of the various factors of production tend to be
distributed as distinct from all ephemeral and temporary fluctuations. Even
the most ephemeral phenomena must also be understood and explained, if for no
other reason than that the laws controlling them are, in the last resort, not
different from those determining their permanent effects; but it goes without
saying, that that phase of our theory which covers those cases outlasting the
others in time and space will be far more important to us than the
explanation of rapidly passing exceptions.
But there is a second reason why it seems to me that the consideration of
the influences of "power" deserves greater attention from the
viewpoint of their permanency, for, as far as my knowledge of economic
literature goes, this most important phase of the subject has never been
investigated.
While the problem of the influence of power on prices as such has hitherto
been only scantily treated, and never in a systematic manner, in economic
theory, fundamental investigations into the permanent effects of such
influences of power seem to be totally lacking, so that here we enter, in a
certain sense, upon literary virgin land.
