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

Cartels as Efficient Productive
Structures

Pascal Salin

artels are considered to be specific productive structures which

allow producers to exert a monopoly power. Evaluation of the

working of cartels is thus closely linked to the theory of com-
petition and monopoly which one adopts. This field of economics is
certainly one of those in which Murray Rothbard’ has made break-
through contributions. He has persuasively shown that there is no mo-
nopoly power as far as there are only voluntary arrangements. From
this point of view, freedom of entry in production is the only relevant
criterion to evaluate productive structures and one might dismiss as
irrelevant all the other traditional conditions of the pure and perfect
competition theory. In the present article we will not address this gen-
eral debate about competition and monopoly and we accept the views
of Murray Rothbard without any discussion. Our precise aim is rather
to look for the specific characteristics of cartels and to evaluate them un-
der the light of this approach. Cartels are generally considered negatively
as formal arrangements to restrict production. After having discussed
this approach we explain why cartels rather play a positive role in meet-
ing some specific demands of the market. As a consequence they mod-
ify the frontier between the firm and the market.

The Cartel as a Restrictive Structure of Production

The traditional theory of pure and perfect competition focuses on the
number of participants and, as such, introduces a strong opposition
between the competitive case in which demand is perfectly elastic
(from the point of view of the individual producer), and the monop-
oly case in which demand is inelastic, which makes possible for the
monopolist to extract a “super-profit.” In between is the case of the

*Pascal Salin is professor of economics at the Université Paris—Dauphiné.
Particularly in Man, Economy, and State, 3rd ed. (Auburn, Ala.: Ludwig von Mises
Institute, 1993). In the present paper we rely mainly on this book, from which we
extracted all quotations for which pages are given without any further indication.
The Review of Austrian Economics Vol. 9, No. 2 (1996): 29-42
ISSN 0889-3047

29


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

30 The Review of Austrian Economics Vol. 9, No. 2

oligopoly, namely the troublesome case in which there are a few pro-
ducers on a market.’

What is wrong in the traditional theory is not the formal analysis
of monopoly or oligopoly, but the confusion it has made between the
end result and the process. The number of firms on a specific market
at one given time is the provisional outcome of a specific process and
it has no meaning by itself. On the contrary, the fact that the process
is based on the use of public constraint or is purely spontaneous makes
the difference. Thus, it is not true that super-profit exists whenever
there is freedom of entry on a market, since one cannot explain, in such
a case, why there are no more producers trying to get such a super-
profit. But the traditional analysis of monopoly is clearly and directly
appropriate when applied to the case of a public privilege, namely the
existence of a public barrier to entry: In that latter case, the formal
analysis of monopoly is a correct description of reality. In this case only
does a super-profit exist.

Now, what about the case of a small number of producers, which is
the scope of analysis both of the theory of oligopoly and the theory of
cartels? Let us assume that for some technical or historical reasons,
only a limited number of firms exist at one given time (the possible
economies of scale not being large enough as to justify the existence of
only one producer). Each producer has to elaborate a strategy of his
own in an environment of imperfect information (since the decisions
of each producer affect the state of the market, but no one knows about
the intentions of others). Thus, authors such as Heinrich von Stackel-
berg have shown that the outcome of the process may be different ac-
cording to the strategies followed by different producers. There are, for
instance, cooperative and non-cooperative processes, equality between
the producers or a distinction between a leader and followers, etc.
Once more, these assumptions may have interesting applications.
They are debatable as far as they tend to introduce the idea of a mo-
nopoly situation in the absence of any public privilege.

A distinction seems to be frequently made between an oligopoly
and a cartel. A cartel could be defined as a specific oligopoly in which
the different producers, instead of just interacting, enter into a process
of explicit cooperation. Thus, a cartel is viewed as an agreement be-
tween different producers in order to follow common rules or behav-
lors, i.e., a system of mutual and freely accepted obligations.

2The definition of a market is arbitrary, since it implies the possibility to differen-
tiate perfectly specific goods. This would be possible without any ambiguity only if goods
were perfectly not substitutable one to the other. As far as some substitutability
exists—which is always the case—the definition of a good and of a market is arbitrary.


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

Salin: Cartels as Efficient Productive Structures 31

It is usually assumed that this agreement can be explained by a
common intention of participants, namely creating a monopoly posi-
tion. Therefore, participants who do not face a perfectly elastic de-
mand are able to extract a super-profit from purchasers, which they
would be unable to obtain without such an agreement. Thus, cartels
are mainly viewed as having a negative aspect, as is expressed in the
usual saying that cartels are collusive.

In the traditional theory the super-profit is shared between par-
ticipants according to a priori rules, for instance their market shares.
It is well-known that cartels are considered unstable structures, since
participants are tempted to become free riders: Although they have an
interest in others’ respecting the agreement, they can expect a larger
share of the common super-profit by introducing some degree of com-
petition with regard to their partners.

The cartel is sometimes considered as worse than the monopoly. In
fact, it is often assumed that a monopoly exists for purely technical
reasons, for instance because techniques are such that large econo-
mies of scale make it impossible for several firms to coexist in the most
profitable way (natural monopoly).’ In such cases no normative judg-
ment is issued against the monopoly, but it is argued only that the
state has to prevent the assumed exploitation of consumers by regu-
lating the monopoly or by nationalizing it. The value judgment con-
cerning cartels is more critical since it is assumed that there is no tech-
nical reason for any monopolist position, but that the cartel is created
ex nihilo by an explicit agreement between producers in order to ex-
ploit purchasers. From this point of view any cartel arrangement
ought to be prohibited, for the cartel exists in order to create restrictive
procedures.

In fact, as we shall see later, this negative approach to the cartel
is a direct consequence of the traditional theory of competition. Accord-
ing to it, pure or perfect competition can only exist when a great
number of (identical) producers produce a homogeneous good. In fact
this theory does not describe a real process of competition between real
entrepreneurs, but the technical organization of managers in a non-
innovative system, for instance the managers of plants in a Soviet-
style centralized system of production: It is assumed that there is one
single technique—an optimal technique from a purely technical point
of view—to produce a given good and all (numerous) managers have
to apply exactly the same technique to produce exactly the same good.

3Anyhow, those who support such views do not usually consider the possibility of
competition from producers located in other nations and are ready to support protec-
tionist measures.


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

32 The Review of Austrian Economics Vol. 9, No. 2

In that sense, it can be said that the theory of pure and perfect compe-
tition is in fact a theory of central planning.

The Austrian theory of competition leads to an opposite view. The
end result of a competitive process is not the production of identical
goods by numerous producers, but, on the contrary, the differentiation
of goods produced by entrepreneurs who are innovators and who are
induced by competition—namely the free entry of potential competi-
tors—to offer better goods at lower prices.

Therefore, as far as cartel agreements intend the homogenization
of goods (either of their specific characteristics or their price), one may
wonder for which category of theories it could be a problem. From the
point of view of a supporter of the traditional approach of competition,
the problem may not be that of homogenizing goods, but of creating
monopoly power. From the point of view of the supporter of the Aus-
trian theory of competition, there is no such thing as monopoly power,
as far as the process of monopolization by cartels is spontaneous. But
the reason for homogenization has to be clarified—which we will try
to do later—since the merit of competition relies on the fact that it in-
duces differentiation.

The approach used by Murray Rothbard to study the specific case
of cartels seems to be the following: He accepts the idea according to
which cartels are a system of collusion aiming at introducing some re-
strictive actions in a productive system (rarefaction of production).
However, he shows that, on the one hand, restrictive actions of produc-
ers are perfectly justified and that, on the other hand, cartels cannot
create monopoly power since such power cannot exist as long as there
is free entry on a market. Thus, in Murray Rothbard’s view, if produc-
ers organize themselves into a restrictive agreement-—a cartel—it is
to meet a demand by purchasers for a restrictive action. The argument
goes as follows. A monopoly position can be reached by a single pro-
ducer or a set of producers organized into a cartel only if demand is
inelastic, which makes restriction of production possible and profit-
able. But, it is up to demanders not to have an inelastic demand: If ever
they were unsatisfied with the cartel’s behavior, they could decide on
having a perfectly elastic demand, i.e., not accepting an increase in
prices through the restriction of production. As Murray Rothbard
wrote,

If the consumers were really opposed to the cartel action, and if
the resulting exchanges really hurt them, they would boycott the

we have developed this view in a short book, La concurrence, in the Que sais-je?
collection (Paris: Presses Universitaires de France, 1995).


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

Salin: Cartels as Efficient Productive Structures 33

“monopolistic” firm or firms, they would lower their purchasing so
that the demand curve became elastic, and the firm would be forced
to increase its production and reduce its price again. If the “monop-
olistic price” action had been taken by a cartel of firms, and the cartel
had no other advantages for rendering production more efficient, it
would then have to disband, because of the now demonstrated elas-
ticity of the demand schedule.’

By saying that demanders could react against the cartel by lowering
their purchases, he implicitly assumes that the good produced by the
cartel has close substitutes, so that the demanders can decline the re-
strictive action of producers (and the corresponding increase in prices)
by shifting to these substitutes, without any significant loss of util-
ity.

Thus, if a cartel burns coffee in order to increase its price, the
waste comes from the fact that there has been an excessive production
of coffee and not from the fact it is burnt (since coffee has no other
value than the one it gets from its relative scarcity in comparison with
other goods). In some sense the restrictive action is not specific to a
cartel, since on any free market one always restricts the production of
a good in order to produce another one. According to Murray Rothbard:
“The cartel’s action, in reducing the production of coffee and causing
an increase in the production of rubber, jungle guiding, etc., led to an
increase in the power of the productive resources to satisfy consumer
desires,” and “A cartel action, if it is a voluntary one, cannot injure
freedom of competition and, if it proves profitable, benefits rather than
injures the consumers.”” Thus, either the cartel results from the free
working of the market and its possible restrictive actions are wanted
by consumers, or it is the consequence of a barrier to entry and the
restrictive action is a pure monopoly—i.e., state created—exploitation
of purchasers.

However, Murray Rothbard seems to limit justification for cartels
to the case of voluntary and beneficial restrictions of production, al-
though he does not exclude the possibility of other cartel actions when
he writes:

To regard a cartel as immoral or as hampering some sort of con-
sumer’s sovereignty is therefore completely unwarranted. And this is
true even in the seemingly “worst” case of a cartel that we may
assume is founded solely for “restrictive” purposes, and where, as a

SMurray Rothbard, Man, Economy, and State, p. 565.
®Ibid., p. 569.
Ibid., p. 584.


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

34 The Review of Austrian Economics Vol. 9, No. 2

result of previous error and the perishability of product, actual
destruction will occur.®

However, the very notion of a “restrictive action” is by itself debat-
able, inasmuch as such an action can be labeled as restrictive only in
comparison with a norm which could undoubtedly be named non-re-
strictive. Let us take the example of coffee producers burning coffee in
order to increase its price and obtain a maximum profit. A cooperative
action is necessary in order to avoid free-riding. However, instead of
burning coffee after it has been produced, producers could have de-
cided on a cooperative action before producing it in order to avoid over-
production. They did not do it just because they lacked the necessary
information on the future state of the market. Therefore, burning cof-
fee is a restrictive action only in comparison with a specific state of
affairs which had been freely chosen by producers in a situation of im-
perfect information.’ Therefore, it would be preferable to avoid using
the normative word of “restriction” and to just assume that the cartel
aims at profit maximization.

Now, if there is free entry on the market, it is wrong to say that this
profit maximization corresponds to a super-profit. As is the notion of a
restrictive action, the notion of a super-profit is meaningless, since it
implies a comparison with a completely non-existing norm, namely the
rate of profit obtained in a situation of pure or perfect competition.

The mere definition of a cartel does not imply anything about the
degree of freedom of entry on the market. The theory of monopoly has
no scope as long as the reasons for the existence of one single producer
are not made precise (e.g., barriers to entry, economies of scale or in-
novation). Similarly, the theory of cartels is meaningless without a
clear understanding of the reasons for the existence of such a market
structure. As we have seen, the possibility to introduce a restrictive
action on a market is not an acceptable explanation and we have to
consider other justifications for cartels.

If there is not a single producer but a couple of producersin a given
market, it means for instance that there is no additional gain to be
obtained from shifting from several producers to one single producer
(the optimal size of firms has been obtained with several firms and
there is no marginal economy of scale). To most theorists, cartels appear
as unstable market structures, because they can find explanations for a
situation with one monopolist or for a situation with a great number

81bid., p. 570.

Moreover, in such a case, the cartel is only a transitory market structure to cope
with the unexpected consequences of non-cooperative behaviors. Now, in principle, the
market is a coordination process which makes cooperation unnecessary.


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

Salin: Cartels as Efficient Productive Structures 35

of producers, but not for any in-between situation. The cartel thus ap-
pears as a transitory solution to solve a specific problem. However, this
intermediate market structure can be perfectly stable and optimal.

The Cartel as a Value-Producing Organization

As we just stressed, contrary to the traditional theory, competition has the
merit of inducing producers to differentiate their production one from the
other and not to try to produce exactly the same good with the same tech-
nique. Now, if we define a cartel as a structure which allows different
producers to coordinate their production in order to suppress any differ-
entiation in their products, does it mean that cartels can be viewed as anti-
competitive organizations, or that some other reason might explain the
emergence of such a productive structure?

There is a very general and simple answer to this question: If compe-
tition prevails in the sense that there is free entry on a market and if a
cartel has existed for long, it necessarily means that this structure is the
best one to meet some specific demand of the market. In other words, a
cartel is not necessarily unstable—as is obvious from practical experi-
ences—and if it remains alive in spite of potential competition, it means
that it is a useful structure in this specific case. In other words, competi-
tion usually leads to differentiation of products, i.e., imperfect substitu-
tion between them, whereas a cartel tends to induce homogenization, i.e.,
substitutability. The benefits of differentiation are so obvious that there
must be serious reasons for limiting it and introducing a higher degree
of substitutability between products.

In fact there are many specific activities in which there is a de-
mand for homogeneity, especially in network activities, for instance
telecommunications, transportation, or money production (which are
frequently considered as public utilities and, even, natural monopo-
lies). Generally speaking such situations can be named situations of
“sub-additivity,”*’ which includes externalities, economies of scale and
economies of scope. In such cases one can obtain gains from coordinat-
ing several producers or from substituting one unique producer for a
number of them.

As an example, in the case of money production it can be consid-
ered that there are economies of scale and, therefore, decreasing mar-
ginal costs in the production of money, since, for instance, advertising
costs on the characteristics of a currency may be more or less fixed and
the centralization of reserves allows savings of resources; there are

10his term seems to have been used first by the economists of Bell Co. It is used,
for instance, in W. Baumol, J. Panzar, J. C. Willig, Contestable Markets and the Theory
of Industry Structure (New York: Harcourt, Brace, Jovanovich, 1982).


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

36 The Review of Austrian Economics Vol. 9, No. 2

economies of scope (since information obtained by a financial interme-
diary can be efficiently used to create money against credit); and, fi-
nally, there may be externalities, since, for instance, a currency is
more useful for one person the more widely it is used by others. Even
if one disagrees with the precise reasons for sub-additivity, anyone
may accept the idea that it would not be optimal to have a very large
number of different currencies.

Governments and all the experts who support them usually shift
from such observations to the conclusion that there are natural mo-
nopolies, so that public monopolies or regulations are necessary in or-
der to avoid the exploitation of demanders by producers. In fact, there
is only one conclusion to be drawn from such observations, namely that
there is a potential gain to be obtained from decreasing the degree of
differentiation in the production of such goods. There are potential
gains which can be obtained from substituting one or a limited number
of goods for a great diversity. However, it does not mean that the opti-
mal degree of diversification—or, conversely, the optimal degree of ho-
mogenization—can be decided a priori from a purely technical ap-
proach. Two remarks are important at this point:

(1) If ever gains can be obtained by diminishing the degree of differ-
entiation—for instance because of economies of scale—it does not
imply that it is also optimal to diminish the number of producers. As
an example, if it could be proven that it would be technically optimal
to have one single money or one single computer standard in the
world, it would not imply that there ought to be one single producer:
Either a monopoly or a cartel can do the job. In fact, a cartel is a
productive structure in which different producers produce the same
good, so that they can be as efficient as a monopoly in meeting
sub-additivity problems.

(2) One cannot know in advance and forever whether there is
sub-additivity in an activity. It has to be discovered. And, as is
well-known, competition—i.e., free entry—is the best way to dis-
cover to which extent sub-additivity does exist and to which extent
it may change over time. More specifically, it cannot be generally said
that sub-additivity exists in such and such activity, as it may exist at
one given time at one point of the production process, but not some
particular over its entire range. For instance in a telecommunication
or transportation network, it may not be efficient to have more than
one major highway in some part of the network, whereas in other
parts there would not be any economies of scale: the highway may be
operated by one monopolistic producer or by several producers coor-
dinated into a cartel, whereas the other parts could be managed by
different, not coordinated, producers as well as by a cartel (for


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

Salin: Cartels as Efficient Productive Structures 37

instance the one which is managing the highway). As new technolo-
gies are discovered, the place of the cartel may change over time.

Therefore, cartels exist not only, or even not mainly, in order to make
resources scarcer and to increase prices, but to increase the value of
production and improve the productive processes. The cartel is not cre-
ated to extract what the traditional theory calls a “super-profit,” but
to produce coordination gains.

Cartels, as well as monopolies, are the possible outcome of compe-
tition, i.e., free entry on a market: Competition makes it possible for
real entrepreneurs to innovate, so that they are the only suppliers of
the new product they have decided to introduce to the market. As com-
petition has the merit of inducing producers to be the first ones on a
market—i.e., to be what traditional theory calls monopolists—it is
meaningless to compare cartel situations to a situation with many pro-
ducers. But it makes sense to compare cartels to monopolies. In fact,
if ever there are gains to be obtained from homogenizing production,
because of sub-additivity phenomena, is it not more efficient to have
one single producer than several ones in a cartel?

In fact, in a cartel there are possible coordination costs which may
be opposite to the coordination gains obtained through the homogeni-
zation of production. However the cartel may be organized along two
different procedures, either spontaneous coordination or explicit coop-
eration. As an example of the first category, let us take the case of a
free banking system and let us assume that, in a given area, several
banks produce currencies which benefit from a convertibility guaran-
tee in terms of gold. However, each bank discovers that, in order to
make its own money more attractive, it has to increase its liquidity
and, therefore, it decides to accept the currencies issued by other
banks of the system against its own without limit, at the fixed price
given by their mutual gold prices. Such a system of spontaneous coor-
dination makes all currencies perfectly substitutable, which means
that the banks of the system have decided to eliminate any possible
differentiation of their products. However, one may imagine that some
degree of cooperation—i.e., explicit coordination of decisions—may
take place, for instance to decide a common name for the common cur-
rency, but it is not absolutely necessary, since each bank can decide
separately to use the name of a currency already issued by one of them.
Thus, one cannot define a cartel by the existence of cooperation—or
what is sometimes called collusion—but rather by the fact that there
is homogenization of goods produced by different producers, whether
it results from explicit, centralized decisions or from decentralized, in-
dividual decisions.


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

38 The Review of Austrian Economics Vol. 9, No. 2

Coordination costs are certainly higher the more cooperation exists
in the system, since it is necessary to monitor the explicit cartel agree-
ment. In such a case, as we have already recalled, each participant in
the cartel has an interest in others’ honoring the agreement, and in
trying to engage in free riding in order to get a larger share of the com-
mon market. But coordination costs are almost non-existent whenever
the cartel is the result of the spontaneous decisions of its members.

On one hand, a cartel may have some advantages in comparison
with the case of a single producer (monopolist). The main gain stems
from the fact that, although there is actual homogenization of produc-
tion, the possibility of future diversification remains. As we have al-
ready stressed, the existence of coordination gains may change over
time as new technologies are discovered. When a single good is pro-
duced by several producers in a cartel rather than by a single producer,
there may be more incentives to discover new techniques. The tradi-
tional assertion according to which a cartel is necessarily unstable is
not completely wrong, but it must not be considered as a negative aspect
of the cartel, but rather as a positive one. It means that the cartel subsists
as long as it is the most efficient productive organization. But that it may
burst whenever other productive structures appear as more efficient.
Moreover, the participants in the cartel are permanently induced to look
for the possibility of inventing more efficient productive organizations,
some of which includes the dissolving of the cartel or, maybe, the ar-
rangement of a new form of cartel.

Another possible gain brought by the cartel is a scale gain. Con-
trary to what is usually assumed economies of scale do not generally
exist. Moreover, whenever it can be assumed, that there are economies
of scale, they have to be viewed as technical and not as institutional
economies of scale.'’ But there are also diseconomies of scale, which
are mainly of an institutional nature. It is well-known that the inter-
nal organization of a firm does not rely on explicit exchange proce-
dures, so that the production of information may be more difficult. The
larger the firm the larger may be the organization cost. Thus, by coor-
dinating their production—or cooperating—in a cartel, different firms
are able to produce a good at the optimal scale, from the technical point
of view, and of minimizing institutional costs (diseconomies of scale).

The TATA (International Air Transport Association) gives an inter-
esting example of an efficient, rather stable, but changing cartel. Through

"As Murray Rothbard rightly wrote, “The critical problem is not the size of the
plant, but the size of the firm” (Man, Economy, and State, p. 577) and “Economics can
make few valid statements about the optimal size of a firm except that the free market
will come as close as possible to rendering maximum service to consumers, whether we
are considering the size of a firm or any other aspect of production” (Ibid., p. 578).


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

Salin: Cartels as Efficient Productive Structures 39

an agreement which is not very costly to manage, different firms are
able to give more value to their services, since the tickets issued by dif-
ferent firms are (nearly perfectly) substitutable one for another, at least
as regards regular tickets. However, airlines are allowed some degree of
differentiation—for instance in the quality of service—and, on the other
hand, they also produce services (e.g., special rates and charters) which
do not enter the cartel agreement. Thus, the airline industry is charac-
terized by an optimal degree of differentiation and homogenization from
both the points of view of travelers and airlines. To some travelers, the
substitutability between airline tickets is very valuable, for some other
ones the priority consists in getting the lowest possible prices. Thus, the
airline industry meets the different needs of customers.

The traditional theory of competition, as well as the traditional
theory of the natural monopoly, have a global view of what they call a
good. It is considered, for example, that there is something such as air
transportation or telecommunications. Moreover it is assumed that
sub-additivity exists in such network activities, so that there is the
possibility of a natural monopoly. In fact, as it has been made clear by
the new theory of consumption,'” people do not demand goods, but
characteristics and physically distinct goods are supplying bundles of
characteristics, in various proportions. The problem of production is
to adjust to the immense variety of characteristics desired by different
individuals. Now, homogeneity may be one valuable—and, therefore,
demanded-—characteristic, as is the case for money or air transpor-
tation. Therefore, to be efficient a productive structure has to pro-
duce baskets of characteristics which are viewed as optimal by de-
manders, which implies that some baskets may include homoge-
nized goods and some others differentiated ones. The cartel—which
is a mix of differentiation and homogenization—contributes to this
adjustment. From this point of view it is an essential feature of pro-
ductive structures. Contrary to the usual view according to which car-
tels are fundamentally unstable and, therefore, transitory, and con-
trary, even, to the view of Murray Rothbard who also considered car-
tels as somewhat transitory organizations, cartels have a durable, al-
though changing, role to play.

The Cartel as an Intermediate Productive Structure

Thus, the cartel plays an important role in allowing an optimal com-
bination of diversification and homogenization in production, according

12K T. Lancaster, “A New Approach to Consumer Theory,” Journal of Political
Economy (April 1966); reprinted in Modern Consumer Theory (Brookfield, Vt. and
Aldershot, England: Edward Elgar, 1991).


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

40 The Review of Austrian Economics Vol. 9, No. 2

to the needs of demanders; and in providing an optimal combination
of coordination and cooperation. From this latter point of view, it plays
an interesting and intermediate role in productive structures.

Murray Rothbard considers that'a cartel aims at organizing a “co-
operation to increase the incomes of the producers” and he adds: “For
what is the essence of a cartel action? Individual producers agree to
pool their assets into a common lot, this single central organization
agrees to make the decisions on production and price policies for all
the owners and then to allocate the monetary gains among them. But
is this process not the same as any sort of joint partnership or the for-
mation of a single corporation?”"® Comparing the creation of a cartel
to that of a centralized firm, he concludes, “there is therefore no essen-
tial difference between a cartel and an ordinary corporation or partner-
ship.” However, as we already stressed, if ever a cartel and a big firm
are exactly the same, why would cartels exist? Seeing no basic differ-
ence between them, Rothbard assumes that cartels are mainly transi-
tory structures, contrary to what can be shown by reasoning or expe-
rience." In fact, his conclusions are dependent on his definition of the
cartel. Viewing the cartel as an explicit cooperation of firms in order
" to increase the joint profit—according to the traditional definition of a
cartel —he assumes that the most efficient firms will be tempted to
break the cartel in order to increase their market shares. In fact, as
we have seen, a cartel cannot be defined by a strategy of market shar-
ing, although it does exist, but by a coordinated—not necessarily coop-
erative—effort to homogenize production (which may imply identical
prices). If this homogenization is desired by the market, the cartel is
efficient and it will last. If it does not meet any specific need of the
demanders, but is only the end result of an effort by producers to maxi-
mize joint profits, it may fail more or less rapidly.

As a consequence of his restrictive definition of a cartel, Murray
Rothbard thinks that either the cartel is efficient and a merger will
rapidly take place between its members, or it is not, and it will break
down. As he wrote, “if joint action is the most efficient and profitable
course for each member, a merger will soon take place.”’® In reality
efficient cartels can and must last, possibly by transforming their
structure and activities or the number of their participants. The best
example may be that of money production. Under free banking the pro-
duction of money by members of cartels was efficiently made without

3Rothbard, Man, Economy, and State, p. 572; emphasis in the original.

In many cases, a cartel can be considered as simply a tentative step in the
direction of permanent merger,” ibid., p. 573.

Ibid., p. 579.


===== PAGE 13 =====

Salin: Cartels as Efficient Productive Structures 41

any destruction of the cartel or any merger into one single big pro-
ducer. In fact, if there are no unlimited economies of scale, there is no
reason for a merger.

Mergers do not occur precisely because in a cartel firms are inde-
pendent profit centers, which makes economic calculation more effi-
cient. Instead of viewing a cartel as a set of firms which are about to
merge, it may be both more realistic and more efficient to consider it
as the ultimate stage of a process by which a big firm has been decen-
tralized into various coordinated decision centers and, ultimately, split
into independent profit centers with different owners.

As is well-known in his seminal work on the modern theory of the
firm, Ronald Coase'® gave an answer to the following question: If, as
it is rightly assumed, the market is an efficient way for individuals to
organize their mutual exchanges, why substitute other procedures, for
instance the cooperative and command procedures which are used in-
side a firm? The market makes it possible to coordinate relations be-
tween individuals through voluntary exchange, whereas individual
actions are made compatible inside a firm (or any other organization)
through cooperation, i.e., a complex mix of spontaneous and con-
strained (command) actions.

Therefore, it is now widely admitted that an optimal organization
of production stems from the juxtaposition of two non-excluding
schemes, cooperation and coordination. The market is a coordination
process between voluntary actors and the firm—as well as any other
organization——is a cooperative system in which the productive process
is based not on spontaneous interactions between individuals through
contracts but through more vague processes of cooperation (for in-
stance through command, although in any firm there is a mix of com-
mand, voluntary decisions and initiatives, coordination through infor-
mation processes, etc.). Now, there is a sort of frontier between the
market process and the organizational process of firms. It can be as-
sumed that the larger a firm is the more difficult is the internal organi-
zation process. However, the firm may gain from either the possibility
of developing economies of scale or—according to the traditional the-
ory—exploiting a monopoly position and a super-profit. From these
conflicting tendencies, an optimal size of the firm results under spe-
cific conditions.

Instead of this one-or-the-other approach, a cartel makes it possi-
ble to better combine these conflicting tendencies via a better use of
both cooperative solutions (internal organization) and coordination

16Ronald H. Coase, “The Nature of the Firm,” Economica (1932).


===== PAGE 14 =====

42 The Review of Austrian Economics Vol. 9, No. 2

processes (market). This is the reason why there is a great variety of car-
tels combining, in different ways, coordination and cooperation proce-
dures. This is also the reason why cartels are not necessarily unstable.
They are part of a firm's strategies: If ever a firm considers any change
in its production, it does not have to choose only between extending the
internal process of cooperation or entering into voluntary exchange on
the market. It may be preferable to combine cooperation and coordi-
nation under the form of a cartel.” The micro-computer industry
gives an interesting example of such strategies. According to the evo-
lution of markets, strategies, and technologies, producers decide their
standards independently or enter into cooperative agreements which
can be considered as cartel arrangements.

From this point of view it is preferable to abandon the definition
of a cartel as an agreement between firms which intends to exert a
restrictive action or any sort of specific action. The actual intention of
participants is not relevant. Any action results from an intention, but
the content of the intention does not matter from the market point of
view. It may be that an entrepreneur enters into an agreement with
some specific intention, but the outcome of the agreement is not the
one intended, but another one which appears as beneficial, so that the
agreement will be maintained. What is important in a cartel is that
some mix of coordination-cooperation efficiently blurs the frontier be-
tween organizational processes and market processes.

Franchising in retail trade is a well-known example of such strategies.