. 9
( 10)


Suppose that a particular resource appears in the nature-endowed factor
supplies of resource owners sufficiently few in number for all of them to
enter into a cartel agreement. Under such an agreement the owners of the
resource attempt to earn greater revenue through the elimination of compe-
tition among themselves. Each seeks to offer the market less attractive
opportunities (that is, to obtain opportunities more advantageous to him-
self) through the assurance that no other owners of the resource will offer
opportunities more attractive to the market than his own.
Such a cartel, in theory, could operate in exactly the same way as the
single owner of a monopolized resource. If demand conditions are pro-
pitious, the cartel may be able to raise the price of the resource. This higher
price, however, will be maintained only if all cartel members refuse to sell
4 A special case may exist where in the absence of monopoly a resource would have
been a free good. Here a monopolist may be able to hold off sufficient quantities of the
resource to enable it to command a price. See p. 131, ftnt. 15.

for less than the agreed price. This will result in a smaller aggregate
quantity of resource sold, leaving some of it unsold in the hands of the
owners. A cartel agreement will have to provide for a definite method
whereby the sales revenue should be distributed among the cartel members.
(Or, to put the same thing the other way around, the agreement must
specify clearly the basis on which the loss of revenue attributable to the
unsold quantities of the resource is to be borne by the cartel members.)
If the cartel agreement is fulfilled, the group as a whole will gain addi-
tional revenue in exactly the same amount as would be gained by a single
monopolist-resource-owner. This gain will have been distributed among
the members through the arrangement mentioned in the preceding para-
graph. The cartel members as a group will have denied the market the
output obtainable from the unsold quantity of resource, just as the single
monopolist did. In both cases the loss suffered by the market as a whole
is the inevitable accompaniment to the additional revenue gained by the
monopolist or the cartel.
It is, however, precisely this additional revenue gained through the
strict fulfillment of the cartel agreement that makes such an agreement
appear exceedingly difficult to set up and maintain. There is a built-in
tendency for members of a cartel to break away from it. This can easily
be understood. Under the cartel agreement each member, in effect, gives
up his supply of the resource to the cartel as a whole; the cartel as a whole
holds back a quantity of the resource and is able to sell the remaining quan-
tity at a higher price; the revenue is then distributed among the members.
Each member receives in this way more than he would have received if there
was no cartel. But (and it is this that makes a cartel agreement precarious)
each individual resource owner can probably see that he could obtain even
more revenue if he remained outside the cartel arrangement and sold all
his supply of the resource to the market at the price achieved by the cartel.
Now it is true that where one or several resource owners refuse to join
a cartel, it may still be worthwhile for the remaining resource owners to
form a cartel. But these remaining resource owners would now possess,
as a group, only an incomplete monopoly over the resource. In making
their calculations as a group they must now realize that if they force up
the price by holding off some of their supply from the market, they must
share any resulting gain with outsiders who shoulder none of the neces-
sary cost; namely, the loss of revenue on the unsold portion of the resource
supply. With only an incomplete monopoly over the resource, a cartel or a
single resource owner holding a large portion of the resource supply must
lose all the revenue attributable to the resource supply held off the market
(since those outside the cartel are eager to sell all they can at the ruling
price). Any price increases can be maintained only if the cartel holds the
required quantity off the market. Thus, in calculating the wisdom of

pursuing a policy restricting supply, the cartel must offset, against only part
of the additional revenue gained through such a policy, the entire loss of
revenue on the unsold quantity.B
Nevertheless, when the number of resource owners is sufficiently small,
it may be possible to maintain a collusive price-fixing arrangement. In the
literature such cases are frequently called cases of collusive duopoly (where
there are two sellers) or oligopoly ("few" sellers).6

A special case of considerable interest may be considered as follows.
Consider a resource which is present in the original endowments of a num-
ber of resource owners, too large for a stable cartel to be successfully estab-
lished. Suppose, however, that through some special device (legal, institu-
tional, or other), a group of the resource owners are able to sell their
resources to the producers of a particular product (or group of products)
5 This may be illustrated by a diagram. Here DD' represents the market demand
curve for the resource. For a monopoly, this line then represents the monopolist's line
of average revenue, with MR the corresponding line of marginal revenue. The monop-
olist's best possibility, assuming he does not wish to use any of the resource for himself,
is then to sell the quantity OA at price AB (so that his marginal revenue is zero). His

Figure 12-1
total revenue is then OA — AB. If, however, a cartel has only partial monopoly over
the resource, things are different. The line SXCS'XC represents the aggregate supply
curve of the resource owners outside the cartel. Assuming the DD', SNCS'NC curves are
known, the cartel operators may calculate the demand curve that they face. At each
proposed price they can calculate the quantity that the cartel will be able to sell by
subtracting, from the aggregate quantity that the market will buy at the price, the ag-
gregate quantity that the non-cartel suppliers will supply at the price. (Thus, at price
OE, the cartel may expect to sell the quantity EF - GH = EH ” EG.) The line DCDC'
thus obtained is the demand curve facing the cartel; MRC then represents the cartel's
marginal revenue line. The best decision for the cartel is then to announce a price LM.
At this price they can sell the quantity of resource OL yielding the greatest possible
revenue OL — LM (marginal revenue being zero). This revenue is clearly much less than
that for the complete monopolist, and will be correspondingly lower as the SSCS'NC line
moves to the right.
6 A very large literature has emerged dealing in great detail with these cases.
Much of the analysis required for these cases depends on postulates that must be
imported from outside price theory proper. In this book we do not enter into
these problems. For one excellent review of such problems see \fachlup, F., The
Economics of Sellers' Competition, Johns Hopkins University Press, Baltimore, 1952,
Parts 5, 6.

without fear of competition from the other owners of the resource. In
other words the favored group of resource owners, although unable to
control the entire supply of the resource to the market generally, has gained
complete control over the supply of the resource available to all producers
of a particular product or group of products.7
In such a situation the favored group of resource owners may act in a
way that is in some respects similar to the actions of the resource cartel,
but that is in other respects significantly different. Since the owner group
faces no competition in its own "preserve," it may (like a cartel) ask a price
(within this protected area) without regard to the price being asked by
the other resource owners (outside the area). Moreover, although the
owner group realizes that the higher the price it asks (within the protected
area), the smaller the quantity of the resource that will be bought (in the
area); the group is still free to offer (if it wishes) the unsold quantity of re-
source to buyers outside the area (in competition with all the other resource
owners).8 Thus, it may appear extremely profitable for the group to force
up the price of its resource to an area where the group can restrict the supply,
well above the resource price elsewhere.
The consequences for the market are generally different from those
brought about by a cartel with complete or with only incomplete monopoly
of the supply of a resource. In the present case no quantity of the resource
is kept altogether off the market. What is not sold in the protected area at
the high price is sold elsewhere at the lower price. On the other hand,
the artificially high price in the protected area must necessarily generate
important consequences with respect to production plans and the allocation
of resources. Within the protected area the producers will seek to adjust
their production plans to the higher price. They will substitute other re-
sources for the "restricted" resource at the margin; they will alter the scale
of their output in the light of the new configuration of resource prices. In
the long run they may move into other branches of production, outside the
protected area.
On the other hand, producers outside the area will find that a larger
quantity of the restricted resource is being offered for sale to them at any
given price (this quantity including those resources barred from employ-
ment in the protected area by the artificially high price). This will result
generally in a somewhat lower price than would have prevailed in the ab-
7 Strictly speaking, this case is unlikely to be altogether compatible with the definition
of a free market system developed in Ch. 2.
8 If the group that has gained the favored control over the supply is not a group
of resource owners but a group of entrepreneurs (who admit resource owners as
partners in order to supply the "protected area"), then there will of course be no
problem of unsold resources. The group will merely admit to partnership only that
number of resource owners necessary to ensure supply of that quantity of resource that
maximizes the group's revenue.

sence of all supply restrictions outside the protected area. Producers out-
side the area will adjust their short-run and long-run plans to this situation.
In general, the result will be that the restricted resource is used in the pro-
tected area in such a limited degree that the efficiency of the resource at the
margin is high so that buyers in this area find it worthwhile to pay the
high price; while outside the area the resource is used so freely, in view of
its especially low price, that its efficiency at the margin is much lower. The
supply restriction, while not denying altogether to the market the output
of any quantity of the resource, has succeeded in forcing some quantity of
the resource to be used where its efficiency at the margin is lower than in
the protected area.
This will be eventually reflected, of course, in the pattern of product
prices and the quantities bought of these products. It is observed, once
again, that these consequences of the supply restriction result directly from
the gain received by the favored group of owners”this gain, in the present
instance, not being offset by any loss of revenue due to any unsold quantity
of the resource.9

In the short run it may be possible not only for sellers to combine, but
also for all the buyers of a particular resource to combine, in this case, for
the purpose of forcing down the price of the resource. (Alternatively, it is
possible that all the supply of the resource is bought by a single entrepre-
neur, and that in the short run he will be able to exploit this monopsony
position in order to force down the price.) In the long run, if the price
of a versatile resource is very low, there is no a priori reason why in the
absence of institutional barriers the superior advantages secured by pur-
chase”so cheaply”of the resource should not attract competition from
fresh entrepreneurs. (There is thus an important asymmetry in this respect
between the buyers' and sellers' sides of the market.) But in the short run
the entrepreneurs who buy the resource may feel reasonably secure against
outside competition and may seek additional advantage by eliminating com-
petition among themselves.
Such a combination of buyers will be able to offer a low price for the
resource without fear that anyone else will offer sellers of the resource a
more attractive price. The result will be a lower price for the resource,
and a consequently smaller quantity of resource supplied to the market.
Buyers of the resource, if they choose to force down the price in this way,
will have to adjust their production plans to the availability of only smaller
9 Some revenue loss may be suffered, of course, due to the lower price the re-
source must be sold at outside the protected area.

quantities of the resource.10 The lower resource price may yield short-run
advantage to the buyers in their capacity of producers. The other pro-
ducers of the products that these monopsonist-buyers produce, whose plants
and long-range production plans require no inputs of the monopsonized
resource, will find themselves at a cost disadvantage.
It is observed, however, that there is a fundamental difference between
the previously considered consequences wrought by the monopoly power of
a single resource owner, and the consequences of the buyers' combination
discussed here. In the monopoly instance, the control over supply (coupled
with the existing demand conditions) made it in the interest of the resource
owner to hold back from useful employment (in fact to destroy) an available
quantity of resource that consumers (through their "agents" the entrepre-
neurs) would have gladly used (and for which they were willing to pay a
price, which, in the absence of monopoly, would have brought all the re-
source quantity into employment). In the present case of a buyers' combina-
tion, on the other hand, the buyers have merely decided to offer, in concert,
a price so low that it is worthwhile for resource owners to yield only a smaller
quantity of the resource to the market. (Even a resource price established
in a competitive market, we observe, is probably able to attract resource
owners to yield only a smaller resource quantity than they would be pre-
pared to yield at a still higher price.) Resource owners are not hurt by
monopsonistic action on the part of the buyers of resources in the same
way the ultimate buyers of resources are hurt by monopolistic action on the
part of the sellers of resources.
10 This may be illustrated with the help of the diagram. The line SSr represents
the supply curve of the resource that faces the monopsonist group. Each point on the
curve reflects the quantity of resource that the resource owners in aggregate will sell
to the monopsonist group if they offer a particular price. The MC line then expresses

Figure 12-2
the marginal cost to the buyers' group of advancing purchases of the resource by suc-
cessive units. The line MP reflects the respective increments to revenue that the em-
ployment of successive units of the resource is able to afford to the buyers. (The;
downward slope of this line reflects, among other possible things, the laws of variable
proportions.) Clearly, the monopsonist group will do best by offering a price AB, so
that they will be able to obtain the quantity OA. (At higher prices they would be able
to secure greater quantities of the resource.) It should be observed that the selection
by a monopsonist of his preferred position does not differ essentially (either diagram-
matically or logically) from the selection made by a non-monopsonistic resource buyer.
The only difference is that for the latter the supply curve is likely to appear far more
elastic (in special cases, even perfectly elastic).

Nevertheless, in the short run, the combination of buyers will have its
effect on consumers. Since we are assuming that the entrepreneurs who
are members of the buyers' combination produce their products in competi-
tion with other producers, there will not result directly any contraction in
product output. Any reduction in output by the members of the buyers'
combination, due to the smaller quantity available of the resource that they
buy as a group, will be made up by other producers, possibly at somewhat
higher prices.
As we have seen, in the long run and, for many resources, even in the
short run, even these effects cannot last. Barring institutional restriction
upon entry into the ranks of the entrepreneur, the lower costs achieved by
the members of the buyers' group will attract competition. If the new
entrepreneurs are unable (even by joining the buyers' group) to secure the
quantities of the resource they would like (due to the small quantity supplied
at the low price even by the buyers' combination), they will offer resource
sellers somewhat higher prices in competition with the group. The compe-
tition of these new entrepreneurs, bringing about a tendency for the product
prices to fall, and for the resource price to rise, will eventually wipe out
any profits that the members of the buyers' group had gained.

An important monopoly case may arise when an entrepreneur pro-
ducing a particular product has monopoly control over a resource abso-
lutely essential to its production. We may for simplicity imagine a favored
resource owner, the only person in whose resource endowment any of this
resource is included, acting as entrepreneur-producer of a product that must
include a fixed quantity of the rare resource per unit of product. Since
there is no buying and no selling of the monopolized resource itself, the
monopoly power conferred upon the favored resource owner can be ex-
ploited only in the product market.
The consideration determining his production and pricing policy are
similar to those governing the decisions of the monopoly seller of a resource.
He is in a position to ask consumers any price he chooses for his product,
without fear that anyone else will offer the same product to the market for a
lower price. No one else, in fact, can produce the product at all, since no
one else is permitted to buy the monopolized factor indispensable for its
production.11 On the other hand, the monopolist-producer knows that he

Where it is possible for other resources to be employed as more or less imperfect
substitutes for the monopolized resource, certain modifications must be made in the
analysis in the text. To the extent that the monopolized resource is superior to the
substitute resources, the monopolist-producer may yet be able to exact from the market
a monopoly gain. See p. 285, ftnt. 19.

faces the very real competition of other products bidding for the consumers'
incomes, both the competition of the products that are substitutes in con-
sumption for his own product, and the competition of other products.12
(All consumer products, of course, are "substitutes" for one another in the
attainment of "satisfaction.") He knows, therefore, that the higher the
price he asks for his product, the smaller will be the quantity bought by con-
sumers. The keenness of the competition provided in general by all other
consumer products will express itself in the elasticity of market demand
for the monopolized product. (It is the market demand that is relevant to
the decisions of the monopolist-producer, since he has to deal with the entire
demand of the market for his product. The demand curve that he faces
is the demand curve of the entire market.)
A set of factors not considered by the monopolist-seller of a resource
complicates the decision of the monopolist-producer. These relate to the
costs of production of the monopolist-producer. Like the monopolist-seller
of a resource, the monopolist-producer must weigh, against the increased
revenue that can be obtained from what he is able to sell at a higher price,
the loss in revenue that he suffers due to what, precisely because of this
higher price, must remain unsold altogether. But in addition the monopo-
list-producer must consider the effects of asking a higher product price upon
his aggregate and per-unit costs of production. At the higher price he
will sell less of the product, will produce less of the product, and will in
consequence, in the short run certainly, have lower per-unit costs of produc-
tion. Thus, offsetting the loss in revenue on potential units of product that
will not be sold or produced due to the higher price, the monopolist can
weigh (besides the higher revenue on the products produced and sold) the
saving in costs of production both on the units of product not produced,
and also on the units that are produced (at costs that are lower due to the
smaller volume of output).
The deliberations of the monopolist-producer can be conveniently
schematized by means of diagrams. In Figures 12-3a, b, and c, on the follow-
ing pages we assume (heroically) that the producer knows (or believes that
he knows) both his cost curves and the market demand curve for his product.
The diagrams show the short-run average and marginal cost curves of the
monopolist-producer (these costs not including any cost attributable to the
use of the monopolized resource). Each of the diagrams reflects a particular
(different) demand situation shown by the relevant market demand curve
for the product (which is, therefore, also the monopolist-producer's average

12 As usual, the elasticity of the market demand curve for the monopolized product
reflects the degree to which it faces the competition of other products in general. The
chief purpose of the notion of cross elasticity of demand discussed on pp. 99 ff, is to
measure the degree of competition offered to the monopolized product by any one particu-
lar product.

revenue curve). For each average revenue curve the corresponding marginal
revenue curve has been drawn. In each diagram the line BE marks the
absolute upper limit to the volume of output of the product permitted in
each period by the available supply of the monopolized resource, no matter
how high the product price may be. This maximum output is, for each of
the diagrams, the quantity OB.


ß.C Ouon†i†y

Figure 12-3a

It is clear that in each of the cases shown, the monopolist-resource-
owner-producer will seek to produce that quantity at which marginal cost
of the other factors required for expanding output just balances the cor-
responding marginal revenue. This output decision on the part of the
resource-owner-producer is completely analogous to what we know to be
the optimum decision (mutatis mutandis) for an entrepreneur who is a
producer but not a resource owner. At any smaller output volume, it is
obvious, marginal revenue (derived from the use of the monopolized re-
source) is greater than marginal cost (of the other required factors). The
producer stands to gain, by a unit expansion of output, an addition to
revenue that is greater than the required addition to costs of production.
(No additional cost would be involved by the increased use of the monopo-
lized resource.) Thus, a smaller output volume would not exhaust all the
possibilities open to the monopolist-producer. On the other hand, a greater
volume of output (than that at which marginal revenue just balances the
marginal cost of expanding output) would also not be the best for the inter-
ests of the monopolist. At greater volumes of output, the marginal cost
curve is higher than the marginal revenue curve. A unit cutback in produc-
tion would save the monopolist, at the margin, an amount greater than the

corresponding loss in revenue. The best output from the point of view of
the monopolist-producer is thus shown on each of the diagrams by the
distance OC (to be sold at the price DC), corresponding to the intersection
(at F) of the marginal revenue and marginal cost curves.
1. In Figure 12-3a this output happens to coincide exactly with the
maximum output (OB) that the monopolist is able to produce with his
limited stock of the monopolized resource. In this diagram the demand
situation, therefore, is such that it does not pay the monopolist-producer to
restrict his employment of the resource that he monopolizes. If he holds
any quantity of the resource "off the market" (that is, if he refrains from
using the whole supply in production), he will be sacrificing, on the units
of revenue not produced, a potential revenue that (even after it is reduced
by the corresponding saving in costs of production) is not offset by the
resulting increased revenue obtained on the units of product that are pro-
duced. In this case, demand is sufficiently strong and sufficiently elastic to
force the monopolist-producer to use his monopolized resource just as fully
as it would have been used in the absence of monopoly. The upper limit
to output set by the quantity available of the monopolized factor is fully
achieved despite the ability of the monopolist to restrict production.
If, with the same cost and demand structure of Figure 12-3a, the resource
(now monopolized) would have been available in a competitive market (in
the same aggregate quantity OB), there would have been essentially similar
market results in equilibrium. The full quantity of the (now monopolized)
resource would have been bought by entrepreneurs at a price, for the fixed
quantity of the resource required per unit of product, somewhat less than
FD. At this price for the resource, it would just have paid the entrepreneurs
to produce the units of product requiring the final units of the (now monop-
olized) resource. To produce a unit of product they would at this point
have been paying the sum FC for the other complementary factors of pro-
duction (as does now the monopolist-producer also), together with the sum
FD (or somewhat less) for the required additional quantity of the (now
monopolized) resource. The competitive market would have been in equi-
librium. It would just have paid the entrepreneurs to produce an aggre-
gate output of OB: the marginal cost of production (FC + FD) being exactly
covered by the marginal revenue (DC) (which is the price that consumers as
a whole are willing to pay for the supply OB, as seen from the demand
curve). Any smaller aggregate volume of output would have sold at a price
high enough to leave a profit. Competition would wipe out this profit
margin through output expansion up to OB.
With the resource monopolized, on the other hand, and with the monop-
olist-resource-owner himself the producer, the demand pattern in Figure
12-3a, brings the same results. The monopolist produces output OB, and
sells it at price DC per unit, paying the sum FC for the other factors re-

quired for the marginal unit of output, with the difference being the net
proceeds that he receives (as resource owner) from the employment of the
marginal units of the monopolized resource.
2. In Figure 12-3b the cost curves, and also the limit-to-output line EB,
are all exactly similar to those in the previous diagram. The demand
situation, however, is different. In the present case the market demand
curve for the product (the monopolist-producer's average revenue line AR)




C HB Quantity

Figure 12-3b

intersects the line denoting the marginal cost of the other factors to the left
of the line EB. This means that if the entire supply of the monopolized
resource were to be employed in production, the resulting output volume,
in contrast to the preceding case, would be so large that it could be sold
only at a price per unit insufficient to cover even the costs of the other
factors required (for the production of the last possible unit of output).
If there were no monopoly of the resource, it is clear that some quantity
would remain unused. Competition among sellers of the (now monop-
olized) resource would force down its price to zero,13 and entrepreneurs
would employ it only up to the point where the additional revenue gained
by employing the marginal unit is just greater than the additional costs in-
curred by the employment of the other factors of production complementary
to it. This would result in an aggregate output OH (assuming that the
13 Compare, on this point, the discussions on p. 131, ftnt. 15; p. 229, ftnt. 12; and
p. 268, ftnt. 4. An example of such a case is where a single producer has sole possession of a
piece of technological information that he is able to keep secret. Under competition
such information, vital though it might be to a certain branch of production, could
command no price. Knowledge of the technological secret could produce, with freely
available complementary resources, any desired quantity of product; the distance OB
would be infinitely great. Monopoly over the secret (conferred institutionally, for ex-
ample, by patent) would result in the consequences discussed in the text.

costs for the monopolist are no different than they would be for a competi-
tive industry as a whole) and a competitive price GH.
In such a situation it is clearly in the interests of the monopolist-
resource-owner-producer to restrict the employment of the monopolized
resource so that the volume of output is cut to OC. For this output his
marginal revenue line intersects his marginal cost line at F, with the prod-
uct selling at a price CD per unit. The configuration of demand is such
that the interests of the monopolist-producer run counter to those of the
consumers generally. Although a sufficient further quantity of the monop-
olized resource is available to produce the additional quantity of product
CH, which consumers value more highly than the bundle of other com-
plementary factors required, the monopoly position of the resource owner-
producer leads him to withhold the required units of the monopolized


***`¯ ˜/¯"**^”*- i D

i< `^AR

. ¦
I 1

CB Quantity

Figure 12-3c

3. In Figure 12-3c we have still another possibility. Here again we
have the same cost curves and upper-output-limit line EB as before. The
market demand curve for the product intersects the line of marginal costs
of the other complementary factors to the right of the EB line (exactly as it
did in the case of Figure 12-3a). If the now monopolized resource would
not have been monopolized, it would have been fully employed and would
have brought a price in the market (again, exactly as in the case of Figure
12-3a). Competitive output would have been OB, selling at a competitive
price BG. However, in the present case (unlike the case of Figure 12-3a
but like the case of Figure 12-3b), it would be in the monopolist's interest
to restrict employment of the resource that he monopolizes, and conse-
quently, of course, also the volume of output, below the corresponding levels
in a competitive industry.

This is so because at the "competitive" level of output OB, the marginal
revenue (associated with the monopolist's average revenue line) is less than
the marginal costs (incurred by the employment of the other resources
necessary for the production of the last unit of output). Thus, it would
pay the monopolist to restrict output to OC, corresponding to the point of
intersection of the marginal revenue and marginal cost lines.
The difference between Figures 12-3a and 12-3c thus depends on the
relation of marginal revenue to marginal cost, for a volume of output that
would exhaust the monopolized resource. If the marginal revenue is not
below the marginal cost (of the other required factors), the monopoly posi-
tion of the resource owner will be innocuous, with no divergency from the
price-output pattern that would prevail in a corresponding competitive in-
dustry. If marginal revenue falls short of marginal costs at this maximum
possible output volume, on the other hand, the monopolist's interest will
result in an output restricted below the potential competitive level, with
price correspondingly higher.
Figures 12-3a and 12-3c differ from Figure 12-3b in that in the latter
case the monopolized resource would be in a competitive world, a free good.
This was expressed in the diagram, we have seen, by the intersection of the
market demand curve and the line of marginal costs, to the left of the EB
line. In this case, as we have seen, it would always be in the interest of a
monopolist-owner of the resource to restrict its employment. Where the
demand for the product is sufficiently strong for the (now monopolized) re-
source not to be a free good (even in a competitive market), then, as we
have seen in Figure 12-3a and 12-3c, it may be in the interest of the monop-
olist-producer to restrict output below the level of a corresponding competi-
tive market. In such cases, with a given price the maximum possible output
can be sold at (the distance BG in Figures 12-3a and 12-3c), it would be the
elasticity of demand (at the relevant point G on the market-demand curve)
that will determine whether or not the monopolist-producer will attempt to
force up the price. As in Chapter 6 14, the marginal revenue corresponding
to any point on a demand curve (such as G) is given by the formula MR =
P ¯¯ Pl*> w n e r e P is t n e height of the point above the quantity axis, (such
as BG), and e is the elasticity of the demand curve at the point. Thus,
with a given distance BG for the average revenue obtainable by the sale of
output volume OB, the corresponding marginal revenue will depend purely
on the elasticity of demand (the required marginal revenue being less than
BG by the quantity ”BG/&). The more inelastic the demand curve is at
the point B (reflecting the weakness of the competition of other products),
the greater will be the value of ”BG/s, and, therefore, the lower will be
the relevant marginal revenue. For sufficiently low elasticity, marginal
14 See p. 98, ftnt. 7. (It will be recalled that for downward-sloping curves, the
elasticity is negative.)

revenue will fall short of the relevant marginal costs, and, as we have seen,
make it in the monopolist's interest to exploit his monopoly position through
output restriction.15

If conditions are favorable, we have seen, it may be possible for a
market participant, who is the sole owner of a particular resource, to monop-
olize the output of a particular product and bring about a price-output pat-
tern for the product different from what would prevail in a competitive
situation. In the absence of the particular required constellation of de-
mand and costs, we have seen, the mere fact that the sole control over an
essential ingredient in a product gives a particular producer the monopoly
of the product's output will not lead to any deviation from what would
prevail in the absence of monopoly. The phenomena prevailing in a gen-
eral market, therefore, where a host of products are produced by the coopera-
tion of a host of different productive factors will not necessarily be distorted
merely because of monopoly control over some of the resources, even if this
results in monopoly control over the output of particular products.
Where conditions do favor monopolistic output restriction, the conse-
quences are not difficult to understand. The monopolized resource is
employed, and the product produced, in smaller volume than under compe-
tition. Complementary factors of production that, in the absence of monop-
oly, would have been employed in the monopolized industries will seek
employment elsewhere. In these other industries their productivity will
be lower, and consequently the price that these complementary factors will
bring will be correspondingly lower. On the other hand, the output vol-
umes of other products will be increased somewhat due to the transfer of
these other productive factors. The owner of the monopolized resource,
even after market forces have eliminated all entrepreneurial profits, will
still finish with a more desirable income than he would have been able to
secure without exploiting his monopoly power. The owners of the other
factors will be somewhat worse off, both as a result of the possibly lower
prices they may be receiving for their resources, and as a result of the shift
of production from the more desirable (monopolized) product to other,
somewhat less urgently desired products. These consequences will be
affected by the revisions in consumer income allocations induced by these
income and price changes, and also by the consequent ripples of change
affecting the organization of production.

is A monopolist, like any producer, may select one price-output decision as the best
that he can achieve with a given plant, but may select quite a different plan when he is
free to construct an entirely new plant. In the long run a monopolist's cost curves are
(like those of all producers) different from those relevant to short-run decisions.

The greater the number of resources that are monopolized by the same
single resource owner, the more powerfully he will be able to distort market
activity. Monopoly over many resources, making possible monopoly in
the production of many products, will mean correspondingly weaker compe-
tition from non-monopolized products. This will provide the monopolist-
producer with exceptionally attractive opportunities to gain by raising the
prices of his products.

We have already seen that where a particular entrepreneur or group of
entrepreneurs is the only buyer of a particular resource, he or the group may
be able to obtain a short-run advantage over competitors (who only use other
resources) by forcing down the price through restricting their purchases of
the resource.16 We saw that this possibility is by no means completely
analogous, however, to the case where a monopolist-owner of a resource is
able to force up its price by holding some of it off the market. The analysis,
in the preceding sections, of the effects on the market of monopoly in the
production of a particular product (arising from a monopolized resource)
makes possible the exploration of a further case involving monopoly on the
part of a resource buyer.
Suppose that a producer monopolizes the production of a particular
product by virtue of his sole ownership (in his capacity of resource owner) of
a resource (say, resource A) essential to its production. Suppose further
that the production of the monopolized product calls for the employment
of (among other productive factors) a resource (say, resource B) specific to
the production of this product. Then it is clear that the monopolist-pro-
ducer can enjoy complete freedom from competition in buying this specific
resource B. No other producer will ever desire to buy this resource, so
long as the production of the only product it can be used for is monopolized
by the monopolist-owner of resource A. The monopolist-producer will
adjust his purchases of the specific resource B, as we have seen in a previous
section, so that the marginal revenue that he can derive from the last unit
purchased of it is just higher than the increase in costs necessitated by its
purchase. In the present case the producer will be able to rely on the low
price that he thus secures, not only for the short run, but also for the long
run. So long as he monopolizes production, he will be the only buyer of
resource B who purchases it at a lower price (but for this reason being able
to buy only a smaller quantity of the resource) than would prevail in a
competitive market.
Under certain conditions the position of the monopolist-producer as

16 See pp. 272-274.

sole buyer of the specific resource B may bring about results that seem
analogous to what the monopolist-seller of a resource is able to do. In the
diagram, OBA represents the upper limit to the volume of output obtainable
from the supply of the monopolized resource A (obtainable; that is, if all
other inputs, including the specific resource B} were plentiful). On the
other hand, OBB represents the limit to volume of output obtainable with



Figure 12-4

the actual supply of the specific resource J3. The market demand curve
for the product (the monopolist-producer's average revenue curve) and the
corresponding marginal revenue line are also shown on the diagram. The
cost line shows, for each successive unit of output, the increment in costs
of production attributable to all the quantities of resources required for
its production except resource A and resource B.
If the resource A were not monopolized, the situation would then be
as follows (assuming other things to be unchanged). Output would be
produced by competing entrepreneurs in the aggregate volume OBB> this
quantity being sold at the price BBD. Since this aggregate output requires
all the available supply of resource B, but not all the available supply of
resource A, the latter resource (if specific to the production of this product)
would be a free good. Competition between sellers of resource A would
force down the price to zero. Resource B would command the price DH
in the resource market. Since, however, resource A is monopolized by the
producer, it can be in his interest to restrict output to the quantity OC
(corresponding to the intersection at F of the marginal cost line and the
marginal revenue lines).
Such a restriction of output means that the producer will be employ-
ing less of the specific resource B than is in fact available. Competition
between the various owners of this resource will therefore force down its
price close to zero (assuming the owners of resource B do not form a cartel).
The monopoly position of the producer, gained from his control of resource

A, has thus made it possible for him to cut back output, and hence the
employment of resource B to a point where the latter resource has a zero
price. His monopoly position, as in cases considered earlier, has made it
in his interest to deny consumers the output obtainable from quantities of
the resource which he monopolizes (even though consumers value additional
units of product more highly than the cost of other required productive
factors); but in addition this same interest of the producer has implicitly
required that he deny consumers the output obtainable from a quantity of
resource B which he does not monopolize. Since both A and B are specific
and essential to the product, any restriction of the supply of A allocated to
production, implies also a corresponding "waste" of some of B. The
monopoly position of the owner of resource A, coupled with the specificity
of resource B, together have robbed owners of B of any income they might
have been able to obtain in the market through the sale of their endowments
of resource B, and also robbed consumers in general of the quantity CBB
of output, for whose production they would have been prepared to pay.17

Until now we have considered the possibility of the monopolization of
production, and consequent restriction of output, only as the result of the
sole control by a producer of a resource essential to the production of a
particular product. All the consequences for market phenomena that we
were able to deduce as resulting from such a monopoly of production sprang
thus from the favored position of a producer, not in his capacity of entre-
preneur, but as an owner of resources. The monopoly gain obtained by a
monopolist-producer who has successfully exploited his position is thus not
a kind of entrepreneurial profit, but a kind of gain that may be extracted
from the market by a monopolist-seller of a resource. Where no monopoly
of any single resource exists, there is, in the absence of institutional barriers,
no a priori reason why any one producer-entrepreneur should find himself
in a favored position concerning any particular branch of production.
There may be cases of monopoly in production where the existence of
a monopolized resource may not be immediately perceived. It may happen
that a number of producers are competing with each other in the output
and sale of a particular product, and yet each of the producers feels that
his product commands the loyalty of at least some of his customers. Each
producer feels confident that even if he were openly to raise the price of his
product a little higher than the prices charged by his competitors, not all
of his customers would switch to the products of his competitors. Clearly
such a situation must mean simply that each of these producers is producing
17 Compare with Mises, L.v., Human Action, Yale University Press, New Haven, Con-
necticut, 1949, pp. 380-381.

a product that is not exactly the same as the products of the other producers
in the group”at any rate from the point of view of consumers (which is all
that matters). There may be numerous factors capable of differentiating
the product of one producer from closely similar products produced by
other producers, in the eyes of consumers. The packaging, the color, the
location of production, the name a product is marketed under”these and
similar factors may make two products "different" from one another to
consumers, even though an outsider might pronounce them "the same."
While each of the producers may be producing a product that, in this
sense, is "unique," they are, of course, still competitors with each other.
We have seen that a producer of any product experiences the competition
of producers of other products; certainly a producer will feel the competi-
tion of producers whose products differ only slightly from his own. On the
other hand, in the strict sense of the term, the sole producer of a product,
no matter how slightly it is different from others, may still be called a
monopolist of his product if he has sole control over a resource that is
responsible for the uniqueness of his product. (If he does not possess sole
control over any such resource, then there is no reason why any uniqueness
that he imparts to his product should not be achieved by other producers,
too, if this proves profitable.)18 But there may well be monopolized re-
sources that make possible the product differentiation, and these may not
always be immediately perceived, and may sometimes be the result of institu-
tional barriers.
A catchy trade name, for example, may be such a resource that could
be monopolized as a result of appropriate laws. A special location of
production, "superior" in the eyes of some customers to alternative loca-
tions, may be another such monopolized resource.19 The good name ac-
quired by a particular producer through past activities may be yet another
such resource (one which, in the nature of things, is monopolized, at any
rate in the short run). These may not be immediately recognized as being
resources, so that the source of a monopoly in the production of a differen-
tiated product may not be immediately perceived as resulting from a re-
source monopoly. But from the point of view of pure theory, it is clear,

18 In the price theory literature these cases have acquired the name monopolistic
competition. In this (very voluminous) literature the existence of a resource monopoly
(as foundation for the restriction of output) has not been emphasized. Within the
framework of discussion adopted for the present chapter the cases labeled "monopolistic
competition" differ or do not differ from bona fide monopoly cases insofar as they do or
do not involve resource monopoly.
19 These resources, it is noticed, confer an advantage over the similar, but "inferior"
resources, used by the other producers. The monopolized resources, in these cases, are
"indispensable" only with respect to the advantage which they confer. For an excellent
discussion of this point see Bain, J. S., Pricing, Distribution and Employment (rev. ed.),
Holt Inc., New York, 1953, p. 195.

anything that contributes toward making a product superior in any way,
from the point of view of consumers, is a factor of production.
As far as the impact upon the market exerted by the uniqueness of
such a differentiated product is concerned, the relevant analysis is no dif-
ferent from the analysis of the activities of all monopolist-producers. We
have seen that the monopoly of production, which is the result of sole con-
trol over an essential resource, may or may not lead to monopolistic restric-
tion of output below the competitive level. Where there are a number of
producers each producing a product that he is able to differentiate from
the others by virtue of a resource that he monopolizes, each of them will
certainly produce an output where marginal revenue is just balanced by
the marginal cost of all resources except the monopolized one. This may or
may not call for monopolistic restriction of output. The less important
the differences between products, the less likely it will be, other things being
the same, that a producer will stand to gain by monopolistic restriction of
output. Even where the difference between two products is considerable,
the higher price obtained for the superior product of course simply may
reflect its relative superiority in the eyes of the public, rather than be the
result of monopolistic restriction of its supply.
So far from resulting in monopolistic exploitation of the market, the
various methods whereby producers differentiate their products, moreover,
sometimes may be simply the very means with which they compete with
one another. We know that the essence of the competitive market process
is that each participant seeks to obtain more desirable opportunities for
himself through offering the market opportunities superior to those avail-
able elsewhere. Entrepreneurs with superior knowledge of the availability
of resources and of the demand for various products can earn profits by
offering consumers better and cheaper products than other entrepreneurs
less well-informed about market conditions. The attempt to offer to sell
a given product at a lower price is only one of the dimensions competitive
market activity may proceed along (although it is, admittedly, the dimension
analyzed most thoroughly in the literature).20 Entrepreneurs will compete
with each other, in addition, as we have seen, in the selection of which prod-
uct to produce”and this includes of course the selection of quality (or
qualities), which packaging to use, which location to produce at, which name
to assign the product, and so on. Thus, if (without any monopoly of a
resource) an entrepreneur is the only one among a group of producers of
a product who chooses to package the product in a particular way, this
simply means that other entrepreneurs believe they can compete more
effectively by other means. Just as we know that, until equilibrium has
20 The traditional emphasis on price competition seems partly due to the fact that in
the analysis of the "very short run" (the market where no further production is possible),
it is through price competition that the market does, in fact, achieve results.

been attained, different entrepreneurs may be asking different prices in
the market for the same product, so also may they be offering different
varieties of the product to the market in their attempts to most successfully
cater to the wishes of consumers.

The remarks in the preceding sections should help, in addition, in
explaining the case where a particular product happens to be produced by
only one producer who does not control the supply of any of the resources
required (either by technology or by institutional conditions) for his prod-
uct. Such a producer, it is clear, may be the only producer in his "industry,"
but certainly does not monopolize production. His situation is usually
described as one in which he faces potential competition. The situation
might be one, for example, where all other entrepreneurs happened to
believe that this particular product could be produced only at a loss, so that
only one entrepreneur undertook the risk of building and equipping a
plant for the production of this product. The single producer may know
that it is perhaps within his power in the short run to restrict output, and
to raise the price that he asks for his product, without fear that his customers
will turn to another source of supply for this same product at a lower price.
On the other hand, he also knows that there is nothing to stop the eventual
emergence of competing producers of this product, and that a restriction
of his own long-run capacity in order to secure higher prices will certainly
invite the competition of other producers eager to sell the additional units
of output for whose production consumers are prepared to pay. If the
single producer is intent on avoiding long-run losses as well as on securing
short-run supernormal gain, he will avoid a restrictive price-output policy.
A special case of considerable theoretical (and practical) interest arises
where a particular product happens to be produced by only one producer
as a result of the economies of large-scale production. If the long-run
average cost curve for a particular product is declining throughout its rele-
vant extension, the competition of entrepreneurs will eventually bring about
the emergence of bigger and bigger producers. The industry will not be
in equilibrium with a large number of small producers. Whatever the
price of the product may be, a firm that has been satisfied to produce with
a plant designed for a small output volume will realize that it could do
even better with a bigger scale of plant. In the long run, therefore, competi-
tion between producers will force them to seek a bigger output volume.
The bigger the scale of plant, the lower the price a producer can afford to
sell the product at. Producers will therefore seek to offer consumers lower
prices than others are offering through continual increases in the scale of
their operations. On the other hand, of course, if bigger producers are to

do well enough in the industry to wish to stay there, the aggregate output
must not be larger than that which can be sold at a price high enough to
cover costs of production.21 Thus, in the long run the competition among
producers will force out of the industry a sufficient number of producers
so that those remaining can cover their costs. Eventually, it is conceivable
that a single producer may be able to produce the entire supply of the
product at so low a cost and therefore at so low a price, that it pays no one
else to remain in the industry.
A tendency toward the emergence of big-scale production will certainly
evolve in such an industry. So long as this is the result of competition, it is
clear this tendency operates consistently with the tenor of the competitive
market process, in general, to force entrepreneurs to organize production
as efficiently as possible. On the other hand, it is also clear that where only
one or only a very small number of larger producers are left as a result of
this competitive process, a cutback in production may be tempted (if demand
conditions are propitious) in order to achieve greater gains. As we have
seen, such a single producer may be in a position to do this during the short
run. A producer with a specialized large-scale plant, which would require
much capital and time to duplicate, does in fact monopolize a resource
essential to the production of his product. However, it is important to
recognize that he monopolizes this resource only from a short-run viewpoint.
In the long run, anyone who believes he can do better in this industry than
anywhere else can raise all the necessary capital and buy all the productive
factors required to erect another plant large enough to secure all the
economies of big-scale production. (If the first single producer has been
using a scale of plant that has not yet exhausted all possible economies of
scale, then in the long run it will certainly pay other entrepreneurs to con-
tinue the competitive process whereby ever bigger and bigger plants emerge.
Moreover, if we momentarily relax our habitual ceteris paribus assumptions
just sufficiently to consider the impact of a progressing technology, it is clear
that in the long run competing entrepreneurs will be able to set up newer,
more efficient plants than those of the existing "short-run monopolist.")
Thus, while a single large producer might be tempted to underutilize
his plant (in other words to deny consumers the output obtainable from a
resource that he monopolizes in the short run”even though consumers are
willing to pay the additional costs of the other required factors), he would
know that in the long run this would only attract other entrepreneurs into
the industry who will be able to produce as least as cheaply as he himself

21 This assumes that the market demand curve for the product at least for large out-
puts, does slope downward. Considering the analysis of Ch. 5, this assumption is em-
inently reasonable. Economies of scale will boost industry output to the point where
the demand curve, in fact, does slope downward.

can. Potential competition may thus effectively bar even short-run restric-
tion of output by the single producer.

Thus far in this book very little explicit mention has been made of a
model very much used by writers on price theory; namely, the model of a
"perfectly" (or "purely") competitive economy. In this model it is assumed,
in addition to the general assumptions that set up a market system, that
there are so many buyers and sellers of each resource and product that no
one buyer or seller is able by himself to influence market prices, and also
that there is nothing preventing any market participant from entering into
the production of any product he chooses. (Many writers also include the
further condition of perfect knowledge, especially where they refer to the
perfect competition model.) Although models based on these assumptions
have played a very important part in the development of price theory in
this century, and despite the considerable pedagogical usefulness of such
models, they do not contribute significantly to an understanding of the
market process. Analysis of perfectly competitive models is usually con-
fined almost exclusively to the state of competitive equilibrium. (In fact it
has frequently been pointed out that rather serious logical problems arise
when an attempt is made to find out how a purely competitive industry
can ever conceivably attain a state of equilibrium from any other initial
position.) 22
One implication of perfectly competitive models is of particular im-
portance in connection with the discussions of the preceding sections.
Implied in the definition of a perfectly competitive industry is the condition
that each seller of a resource or of a product faces a perfectly elastic demand
curve for what he sells, and also that each buyer of a resource or of a
product faces a perfectly elastic supply curve of what he buys. (Sometimes
perfectly competitive conditions are defined in these terms.) These con-
ditions reflect the assumptions that no seller can raise the price (even
slightly) no matter how he may restrict the quantity that he offers the
market, and also that he will not lower the price no matter how much he
offers to sell to the market; and that no buyer can lower price no matter
how little he buys, and also that he will not raise price no matter how
much he seeks to buy. It follows from these perfect-elasticity assumptions
that to any seller under perfect competition, marginal revenue is equal,

22 In addition, it has frequently been complained, the term pure (or perfect) "com-
petition" is a misnomer, since it requires conditions that prevent individual market par-
ticipants from engaging in any of those activities usually understood by the verb "to

for all possible sales quantities, to his average revenue (which is of course
the market price of what he sells).23 Similarly, to any buyer under perfect
competition, marginal cost is for all possible quantities purchased equal
to average cost (which is simply the market price of what he buys). Now,
since every seller of a product will always seek (with or without perfect
competition) to sell a quantity for which his marginal revenue just balances
his marginal cost of production, it follows that in perfect competition,
equilibrium requires that for all producers output be such that marginal
cost of production just balance product price. And similarly since every
buyer of a resource seeks to employ just enough for the increment in rev-
enue obtainable through the employment of a marginal unit of it to be
just balanced by its marginal cost to him, it follows that in perfect com-
petition, equilibrium requires that for all producers output, and the
proportions of inputs, be such that for each resource the additional revenue
obtained from the marginal unit be just balanced by the resource price.
As a result of the attention paid to the model of perfect competition,
a special significance has frequently thus come to be attached to the equality
for a producer both (a) of marginal cost of production and product price,
and (b) of additional revenue derived from the marginal unit of each re-
source and resource price. Any excess in the price of a product over
its marginal cost of production (or any excess in the revenue obtained
from the marginal unit of a resource, over the price of the resource) being
a departure from perfectly competitive conditions, is immediately associ-
ated with monopolistic or monopsonistic control. Thus the possibility
of a producer being faced with demand curves and supply curves of less
than perfect elasticity (and thus leading to a volume of output where
product price is greater than its marginal cost of production, and the price
of a resource less than the additional revenue obtained through the em-
ployment of a marginal unit of it) is described as monopolistic deviation
from the standards of a perfectly competitive market.
It should be emphasized that such conditions (while certainly inconsist-
ent with the assumptions of a perfectly competitive economy) need not be
accompanied by the monopoly of any one resource and are consequently
different from conditions involving deliberate restriction of output through
denying to the market the use of an available resource. The monopo-
listic restriction of output that we found to be a possible consequence of
monopoly control of a resource should therefore not be considered as the
case symmetrically opposite to the perfect competitive model.24 Rather,

23 See p . 98.
In the context of the "perfect-competition" models, and hence also of the monop-
olistic-competition literature, the polar opposite to perfect competition is provided by
the case of the single producer in an industry that (a) does not permit entry of new
producers and (b) is not faced with the competition of close substitutes.

monopolistic restriction of output resulting from sole control over a resource
should be seen as analytically counterposed to the situation in a "competi-
tive" market25 where competition means simply the freedom for a person
to produce anything that he chooses (without the assumption that when
any one product is produced, it is in fact produced by a very large number
of "atomistic" producers).
When attention is focused exclusively on the state of equilibrium,
a significant difference may appear between the performance of a market
model where each participant faces only perfectly elastic supply and de-
mand curves, and the performance of market models where these curves
("monopolistically") have some slope. But when, as in this book, the
focus of interest is in the market process (leading to equilibrium, possibly),
then the significant distinction is the one emphasized in this chapter;
namely, whether or not market conditions make it worthwhile for the
monopolist-resource-owner-producer to deny to consumers a quantity of
output (one of the resources for which the producer himself has available,
and the remaining resources for which consumers are willing to pay for).
Certainly the idea should be avoided that the assumptions that character-
ize the perfect competition market are in any sense "normal" or "standard"
for a market economy.

Finally, we consider the possibility that the existence of monopoly
control over supply may lead to the emergence of more than one price for
a particular good. Under competition, we have seen, such a state of affairs
must be intrinsically unstable. Should two competing sellers charge differ-
ent prices for the same good, buyers will cease buying (as soon as they
discover the true state of affairs) from the higher priced seller. Of course
where a seller is able to sell at prices considerably below those of his com-
petitors, he may be in a position to demand different prices for his product
from different buyers. But, with no monopoly over required resources,
competition between sellers will eventually enable them all to sell for the
same low prices. For this reason the analysis of price discrimination”
the sale of the same product by a seller to different buyers at different prices
”is usually confined to monopoly situations.
Under certain conditions it may be feasible for, and in the interest
of, a monopolist-seller (either of a resource or of a product) to sell to

25 The term free competition sometimes has been used to denote closely similar mod-
els (but also has been used to cover other cases). See Scitovsky, T., Welfare and Com-
petition, George Allen & Unwin, Ltd., London, 1952, Ch. 15; and also Machlup, F., The
Economics of Sellers' Competition, Johns Hopkins University Press, Baltimore, 1952,
p. 104.

some buyers at prices lower than those that he obtains from other sellers.
For this to be possible, the seller must feel sure that the buyers from
whom he demands the higher prices are not able to buy the good from the
other buyers to whom he is selling for lower prices. Clearly if this con-
dition is not fulfilled, it will pay the latter group of buyers to buy at the
low prices and then resell to the first group of buyers at prices below those
demanded by the monopolist-seller. For price discrimination to be worth-
while an additional condition is that net proceeds with discrimination be
higher than without. This condition, it will be seen shortly, depends on
the respective conditions of demand within each of the groups of buyers it
is possible to discriminate among.
Suppose that a monopolist-seller knows that those who buy from him
(or who might buy at low enough prices) fall naturally into two separate
groups between which no resale of the good (which he sells) is technically
feasible.26 Suppose further that he has available a given quantity (q) for
sale, and, pondering on how to secure the greatest possible revenue from
its sale, is considering asking a price that is the highest price the entire
quantity can be sold at (without discrimination and without holding any
units entirely off the market). At this price, the seller knows, the first
group of buyers (group A) will buy altogether a quantity qa, and the second
group (group B) will buy quantity qh, (qa + qb = q). Now, the respective
demand conditions in group A and group B may be such that the marginal
revenue derived from the last unit sold to group A is less than the marginal
revenue that would be obtained through the sale of an additional unit to
group B. In this case it is in the seller's interest to sell (at a higher price)
a quantity (qa ” 1) units to group A (rather than qa) and a quantity (qb + 1)
units at a lower price to group B (rather than qb), since he would gain a
greater increment in revenue from the latter than he would have to sac-
rifice in group A. The demand situation within each of the two groups,
A and B, is such that the (qb + l) st unit is valued more highly by group
B (as measured by the sums that the group as a whole is prepared to pay
respectively for qb units and for (qb + 1) units) than the (qa)th unit is
valued by group A (as measured by the sums that the group as a whole is
prepared to pay respectively for (qa ” 1) units and for qa units). So long,
then, as a given aggregate sales volume is distributed among the two groups
in such a way that a significant discrepancy exists between the marginal
revenues associated with the last units sold in each group, an opportunity
exists for profitable price discrimination. By exploiting the division be-
tween the two buyer groups, the seller may take advantage of the greater
eagerness of some of the buyers to buy in the one group at the same time
as he taps the revenue obtainable from the large number of potential buy-

26 A standard textbook example is provided by the market for electric power.

ers in the second group who are prepared to buy only at low prices. The
seller will have exhausted all opportunities for further profitable price dis-
crimination when he has adjusted prices in the two groups so that marginal
revenues are the same for both groups.27
Where discrimination in the price of a product is possible in this way,
the monopolist-producer will determine his optimum output accordingly.
As always, he will seek to adjust his output to the point where his marginal
revenue is just balanced by his marginal cost. The marginal revenue rele-
vant to the case where discrimination is possible is of course the additional
revenue obtained through a unit of expansion of total output when output
(both before and after the proposed expansion) is distributed between
the groups by means of the different prices asked so that the marginal
revenues of the quantities sold in each of the groups are equal to one an-
By discriminating in this way between the two groups, the monopolist-
producer may be able to profitably employ all of the resource that he mo-
nopolizes, even though, without price discrimination, it might have been
in his interest to raise the over-all product price through monopolistic
output restriction. Price discrimination enables the monopolist-producer
to gain at least some of the additional revenue resulting from a higher
price, without having to sacrifice all the revenue that he would have to
lose (without discrimination) on the units that cannot be sold at the high
price. The price-discriminating seller is able to sell the units that cannot
be sold at the higher price to a group in which they can be sold at a
lower price.
Where price discrimination is practised, those charged the higher price
are being deprived of part of the consumers' surplus ^8 that they might have
enjoyed in a market without discrimination. Without discrimination
those buyers most eager to buy would not have had to pay a price any
higher than the price paid by the least eager buyer. Now the division of
the market into buying groups forces the buyers in each group to pay a
price no lower than that paid by the least eager buyer within the group.
The segregation of the more eager buyers into one group thus forces them
all to pay prices higher than would have been paid when less eager buyers
were in their market as well. Of course, each of the buyers, even those
paying the highest prices, consider themselves better off by buying than by
refraining from buying (or else they would not be buying); nevertheless

27 T h e point made in the text is frequently expressed alternatively by saying that
discrimination will be worthwhile where the aggregate demand curves of the two (01
more) sectors of the market have respectively different elasticities at a given price. Since
MR = p + p/e, it follows that where the sector demand curves have different elasticities
for a given value of p, the respective marginal revenues will not be the same.
28 See p . 110.

the division of the market has enabled the monopolist-seller to prevent
them from gaining an even greater advantage from their purchases.29
A special case where this can be achieved almost completely is some-
times termed perfect price discrimination. Perfect price discrimination
is possible where the seller divides buyers from each other so completely
that each of the buyer's "groups" consists of only one buyer.30 By dealing
with each buyer individually, a seller conceivably might charge (assuming
he possesses complete knowledge of each buyer's eagerness to buy) each
buyer a price so high that all consumers' surplus is wiped out for all buyers.
Where a number of buyers are included in a group of buyers, even where
they are all very eager buyers, the most eager still gain some consumer
surplus, since they pay a price no higher than is sufficiently low to induce
the least eager in the group to buy. When the size of a "group" dwindles
to one buyer, however, it may be possible for the seller to extract from
each buyer the highest price that he is prepared ever to pay for each
unit bought. (This implies, of course, that a different price will be ex-
tracted from a buyer for each of the units that he buys.) The seller can
achieve this by offering a given quantity to a buyer and demanding a price
for the whole quantity (the price being what the seller believes will just
leave no consumer surplus), with no option to the buyer of purchasing
any smaller quantity at a proportional price. (The seller will determine
the sizes of the lots he will offer to the various buyers, in this all-or-nothing
fashion, in the way that will maximize his own net proceeds.)
Analysis analogous to that presented in this section can be developed
to deal with the conditions price discrimination might be practised under
by a buyer.*1 In the absence of institutional divisions between different
groups of sellers, however, it is doubtful whether monopsonistic price dis-
crimination could be maintained for any length of time.

This chapter has examined the modifications in the general market
process that are introduced as a result of the concentration of the supply
of particular resources (or the production of particular products) in the
hands of single market participants.
When the entire natural endowment of a particular resource is con-
centrated in the hands of one owner, it may or may not be profitable for
29 For a situation where each of the buyers is better off with price discrimination
than without it, see Mises, L.v., Human Action, Yale University Press, New Haven,
Connecticut, 1949, p. 387.
30 The standard textbook example of this possibility is a physician selling medical
services to his patients.
31 See Robinson, J., The Economics of Imperfect Competition, The Macmillan Com-
pany, London, 1933, pp. 224-228.

him to force up the price by restricting supply (depending on the elasticity
of demand for the resource). Where the monopolist finds it worthwhile
to hold some of the resource off the market, corresponding changes are
brought about in the methods and volume of production affecting the
availability of goods to consumers. Where a resource is exclusively owned
by a group of owners able to act in concert, they too may conspire to force
up the resource price by restricting supply. (Several variants of resource
cartel possibilities can be analyzed.) However, such cartels may face seri-
ous problems of enforcing the respective cartel agreements. Where all the
buyers of a resource combine, they may be able to exert short-run effects
on prices and production.
Where the sole owner of a resource chooses not to sell any of it to
other producers, but establishes himself as the sole producer of a product
for whose production the resource is essential, he can employ his monopoly
power in the product market. Detailed analysis shows the conditions under
which he will be able to profit by using his monopoly power to raise the
product price through output restrictions. Further analysis explains how
his favored position may also have an impact upon the markets for the
other resources required for the production of the exclusively produced
In a market where there are numerous, slightly differentiated compet-
ing products, it may not always be immediately apparent whether or not
some of the resources are monopolized.
The analysis also clarifies the existence and impact of the sole producer
in situations where he does not have monopoly power, as defined in this
chapter. In such cases market activity is carried on under the influence
of potential competition. A special case typical of this kind of situation
is where the economies of large-scale production result in only one producer
or a very small number of producers.
The absence of monopoly power, as defined in this chapter, does
not imply that each buyer of any good or service faces a perfectly elastic
supply curve, nor that each seller face a perfectly elastic demand curve.
These latter conditions are usually required for much discussed models
of "perfect competition." The idea should be avoided that such models
are in any sense "normal."
One particular possible result of monopoly control is that more than
one price may emerge for a particular good, even in equilibrium. Such
possibilities are investigated by the techniques of the theory of monopo-
listic price discrimination.

Suggested Readings
Mises, L. v., Human Action, Yale University Press, New Haven, Connecticut, pp.
Hayek, F. A., "The Meaning of Competition" in Individualism and Economic
Order, Routledge and Kegan Paul Ltd., London, 1949.
Machlup, F., The Economics of Sellers' Competition, Johns Hopkins University
Press, Baltimore, 1952, Ch. 4.

The Price System and the
Allocation of Resources

this book has been con-
cerned, "positively," with the operation and mechanics of a free enterprise
system and the market process. We have discussed the process by which
the market determines (a) the prices and quantities produced of each possi-
ble product, (b) the prices and quantities employed of each of the available
resources, (c) the particular group of resources used for the production of
each of the products produced, and (d) the income secured in the market
by each of the consumers and the particular group of products each con-
sumer spends his income on. In Chapter 3, as part of our over-all pre-
liminary survey of a market economy, it was noted that such a system can
(like so many other things) be viewed not only positively but also norma-
tively. That is, a market system can be examined not only in order to
discover chains of cause and effect, which may exist under such a system,
but also in order to judge the degree of success with which the system
achieves specified goals. In the present chapter we return to such an ap-
We have seen that each market participant takes part in the market
process only because he believes that he can in this way achieve his own
goals more fully than by acting completely on his own. In Chapter 3
we saw further that each of the participants is concerned that the system
coordinate the activities of all the participants. A participant will special-
ize in repairing other people's automobile engines only if he can rely on
the market system to ensure that other people will bake his bread, build
his home, and produce his clothes. The more efficiently such coordination
is achieved, the more fully each of the participants will be able to fulfill his
own goals through the market. Coordination, we found, must involve

(a) the priority system according to which the wishes of consumers are suc-
cessively satisfied, (b) the method of production employed for the produc-
tion of each of the products produced, and (c) the means by which the
several contributions of different individuals, who have cooperated jointly
in a single productive process, can be separated for the purpose of assigning
incomes corresponding in some way to individual productive contribution.
In the market system, we found, it is through the assignment of market
prices to resources and products that these coordinating functions are ful-
filled. In the present chapter, within the framework of such a price-co-
ordinating system, we appraise the market system as a means to achieve
the appropriate allocation of the available resources, as judged from the
point of view of the market participants. Market participants, in general,
will wish to know how faithfully the market process impresses upon the
organization of production the pattern that "efficiency" requires, as meas-
ured with reference to the very price system upon whose coordinating prop-
erties the market participants are relying.

Inquiries into the allocative efficiency of an economic system usually
are termed "welfare economics." (This term goes back to a time when
economists uncritically believed it possible to talk meaningfully about the
"total welfare" of a group of individuals. Since then it has come to be
used to cover discussions of the efficiency of a social apparatus in which
"efficiency" is far more carefully defined.) It should be stressed that in-
quiries into the allocative efficiency of a market system can be attempted at
tïuo levels, and that it is only one of these that primarily concerns us here.
The first kind of welfare inquiry assumes all the relevant data are
known, in principle, to the inquiring economist as well as to the market
participants. The initial problem for the economist is to devise "op-
timum" patterns of productive utilization of the known quantities of all
resources, and of distribution of the resulting products among participants
with known tastes. A market system will then be appraised as to whether
its freedom from ignorance enables it to attain such an optimum-allocation
pattern of activities. With full knowledge of all relevant data assumed,
the market position that is set up for appraisal on this level of inquiry is
the position of full equilibrium. The conditions that spell out an equi-
librium position for a market economy (endowed with a given initial set
of factor endowments and with participants of given tastes) are appraised
and compared for their consistency with the conditions for optimality.
We do not consider this kind of welfare inquiry in this chapter.1
1 This first kind of welfare inquiry, presents an essentially mathematical problem. The
general results of this kind of welfare inquiry usually lead to the conclusion that the so-

The second kind of welfare inquiry we are concerned with proceeds
from the assumption that each of the participants is to a large extent
ignorant of the body of information that includes all the "data" of the
market. The initial position assumed for the market is thus a state of
disequilibrium. Initially, the market is understood to be making numer-
ous "errors"; the initial decisions of the various participants are to a large
extent z¿rccoordinated with one another. The market process brings al-
terations in these decisions. The process may be appraised as to the
efficiency with which, employing the limited scraps of information scattered
among the participants, it discovers and corrects the initial errors and fail-
ures in coordination. In this second kind of appraisal, it is the market
process that is being judged rather than the state of equilibrium the proc-
ess leads toward. In many respects this second kind of inquiry is the one
that market participants may be expected to be the most interested in.
After all, in a changing world, a state of market equilibrium, as we have
seen, is hardly an attainable goal. The precise degree in which the state
of market equilibrium deviates from the conditions of optimality is there-
fore likely to appear a distinctly academic question. On the other hand,
participants will be most interested in knowing the direction the market
process moves in; they are vitally concerned with the efficiency whereby
existing m^allocations are discovered and removed, and with the faith-
fulness and speed whereby the market process tends to adjust market ac-
tivities to changes in the basic data. (Of course, participants would hardly
be concerned with the efficiency of a market process unless they also knew
that the final state of equilibrium the process tended toward was also at
least reasonably efficient from the point of view of the first of the two kinds
of welfare inquiry mentioned in this section.) It is the normative exam-
ination of the market process that concerns us in this chapter.

First of all, we should fix in our minds precisely what is implied in the
statement that a resource has been misallocated in a market system. A
unit of a particular resource, let us say, has been employed together with
quantities of other productive factors in the production of a particular
product. The employment of this unit of factor in this way has deprived
consumers of the productive contributions that it might have rendered
in an alternative employment. On the other hand, consumers under the
existing arrangement, are enabled to enjoy the productive contribution
that the unit of factor is making in its present employment. In a market
called "welfare conditions" for optimality, with some reservations, are fulfilled by the
equilibrium conditions for an economy where "perfect competition" prevails in all

system there is a market value placed upon each of the various foregone
productive contributions that might have been rendered elsewhere by the
factor, and there is also a market value placed upon the productive con-
tribution that the factor actually does render. If the market value of
any one of the foregone productive contributions is greater than the value
of the actual contribution of the unit of factor, then we say that this unit
is being employed in the "wrong" use. Measuring "usefulness" by market
value of productive contribution (since we are conducting our examination
of the market system in terms of its own "guide lines"), it is evident that
the unit of factor is being employed less usefully than is possible.
The market value of a productive contribution is an objective mag-
nitude determined jointly (a) by the physical increment of product at-
tributable to the employment of the unit of factor, and (b) by the market
value of a unit of the product. The physical increment of product at-
tributable to the factor depends upon the technological laws of production
and upon the quantities of other factors the unit of the first factor is to
cooperate with in production. The price of a product depends, as we
know, on the willingness of buyers to buy, and of producers to produce and
sell, the particular product. The difference between the market values
of the different possible productive contributions that a unit of factor may
be able to make may thus be due to the different degrees of physical pro-
ductivity of the factor in the various proposed processes of production and/
or to the different conditions of market supply and demand for the relevant
products. Misallocation of a resource may thus be due to its employment
in a productive process where its potential physical productivity is not
being exploited to the full, and/or to its employment in the production
of a product that the market pronounces less important ("importance"
being measured, once again, by market price) than another potential prod-
Our statement of the meaning of the term "misallocated resource"
refers to any given state of affairs (insofar as concerns other market phenom-
ena). We do not here speak primarily of a resource that is not being
employed as it would be under conditions of equilibrium. A resource is
misallocated if it is in the "wrong" place in terms of actual market prices
and with respect to a state of the economy as it is. Our task is to examine
the effectiveness of the market process in detecting and eliminating this
kind of "waste." This is waste (a normative word) because, under the
current conditions of the market, a resource is being used in an employ-
ment that the market declares to be less important than an alternative
available employment.

The discussions in Chapters 7, 10, and 11 concerning the market proc-
ess commencing from a state of disequilibrium clarified the reasons for any
resource being misallocated in a competitive market economy. A resource
may be misallocated only as a direct result of the imperfection of the knowl-
edge of market participants. If knowledge of all relevant data were
possessed by all participants, no perverse discrepancy could exist between
the market value of the productive contribution of a factor in its actual
employment and the value of its potential contribution elsewhere. With
perfect knowledge the price of the unit of factor would be the same in
all areas of the market; differences in the technological efficiency of the
factor in different uses, and in the desirability to consumers of the different
products, would be fully reflected in the prices and output volumes of
the various different products. No room would be left for a perverse
difference between the market values of actual and potential productive
But we proceed here from a position where all the available informa-
tion is initially widely scattered in the form of scraps of knowledge pos-
sessed by individual participants. Resources will be misallocated as a
result of this incomplete knowledge. A resource may be employed in a
less important manner because the entrepreneur is unaware of the more
important employments possible, or because those who are aware of the
more important possible employments do not know of the availability of
the resource. In the first ease, the entrepreneur using the resource in
the less important employment may be unaware of the greater technological
productivity of the resource in other branches of production, and/or of
the higher prices obtainable in the market for the other products. In
the second case, the entrepreneurs who are unaware of the more important
productive contribution that such a resource can make elsewhere may
mistakenly believe that the price of the resource is too high to make its use
worthwhile in these more important employments.
In general, then, the misallocation of a resource can be equated with
widespread (if uneven) ignorance of the gaps in pertinent information.
Some market participants may know all about one piece of information
(for example, the availability of the resource); others may know all about
a second piece of information (for example, the value of the contribution
that the resource could render). But because nobody simultaneously
knows both these pieces of information, nobody is aware of any possibility
of improving the existing allocation of resources. An appraisal of the
efficiency of the market process therefore involves an appraisal of the way
the market process disseminates these missing links of information neces-

sary for the discovery of superior opportunities for the allocation of re-
sources. In the case of the changing economy, the basic data (concerning
resource availability and productivity, and consumer tastes) are free to
change. The efficiency of the market process in this case is again a question
of its ability to transmit to the relevant decision makers those pieces of new
information necessary for the "correct" allocation of resources in terms
of the new conditions.
It should be apparent by now that the answer to an inquiry into the
efficiency of the market process is embedded in the very description and


. 9
( 10)