(Reprinted from HKCER Letters, Vol.37, March 1996)


A New Approach to Competition,
Monopolies, and Cartels

Pascal Salin


There is currently a debate in Hong Kong about competition. Some people think it is not enough just to have competition. Competition must be fair so that firms will not exploit consumers. There is a temptation to introduce competition policy, and this is justified by what I would call the traditional theory of competition, which is the pure and perfect competition model.

The conditions under which competition prevails in this model are that there be a large number of producers supplying homogeneous goods, and that there be perfect information. Since information is perfect, all producers know and use the best technique to produce a given good. At the limit, firms make zero profit because they all use the optimal technique and transfer the benefits of this technique to consumers.

In reality, of course, it is impossible to have an infinite number of suppliers or perfect information. So while the theory appears quite scientific, the decision as to whether competition actually exists is somewhat arbitrary. Part of competition policy is to decide whether a certain activity is closer to the competition model or an oligopolistic model. And sometimes competition policy is used to change reality and make it closer to the pure and perfect competition model; for example, by breaking up a firm, regulating an industry, introducing anti-trust law, or nationalizing industries.

The traditional theory is mainly inspired by the desire to avoid market power and demonstrate that whenever an industry has a small number of firms, they will have some market power and there will be the possibility of exploiting consumers by making the goods scarcer and setting the price higher.

Also important to this theory is the idea that there exist natural monopolies, e.g., utilities, telecommunications, money, and sometimes transportation. The main reason for these so-called natural monopolies is economies of scale. Traditional theory has to solve the problem of making use of economies of scale by a small number of firms while avoiding exploitation of consumers. Competition policy is therefore designed to reconcile this by imposing price regulation, anti-trust policy, or production by the state.

Let me explain why I believe this traditional approach to be wrong. There are two types of entrepreneurs: those described in the traditional theory and those who make choices in a world of uncertainty. In the traditional theory, since everyone knows and employs the best technique, the producer optimizes the use of production factors in a way similar to other producers. However, in reality information is not given. An entrepreneur has to make decisions in an uncertain environment, and profits are the return to his decisions made in such environment. The traditional theory only compensates the entrepreneur for his work in allocating resources, not for his risk-taking. Thus, it is not the right model to describe a competitive economy, but again, it is this traditional theory that inspires competition policy.

I believe competition should be viewed as free entry to the market. Competition defined this way has a merit. It induces entrepreneurs to be unique so that they can get a monopoly position. So while the end result of the traditional theory of competition is a large number of firms producing the same product, the end result of competition viewed as free entry is to have one huge producer who is able to produce a better good with a better technique. If there is only one producer in a market, it is because he is more innovative than others, and he is rewarded for his risk-taking decisions.

It is a different situation if there is a single producer because he has obtained some privileges from the state. This is the result of control and not of innovation. I must say that consumer exploitation exists only if there are state-granted privileges, and it does not matter much whether the privileges are given to a public or a private firm. If some firms get privileges, that means there is no free entry and, what is worse, there is no possibility of innovation.

It is probably true that at one given moment many producers cannot coexist, especially in what can be called network activities such as telecommunications. From a technical point of view there may be one and only one optimal technique, i.e., there is only one way to organize the network in order to minimize costs. Given these technical considerations, the traditional theory tells us the optimal size for a firm. From an economic point of view, however, this is never true. Costs are known only ex post. A real entrepreneur looks at the market and tries to find a way to minimize costs. Quite often there are different ways to produce or organize a network, and the most appropriate way and the corresponding costs have to be discovered. The best way to make these discoveries is through competition, defined as free entry into the market.

It is often said that it would not be efficient to have several producers in telecommunications because there are economies of scale or economies of scope. Consequently, many countries have a regulatory agency distributing the frequencies. In the United States, the agency is the Federal Communications Commission. However, before this existed, it was possible to see the beginning of a free market in communications. It can be shown that there is always a possibility for a free market to work well if property rights are efficiently defined, such as the right of the first user. For example, if one firm is already producing radio services and another firm enters and disturbs the first firm's broadcasting, the first firm can go to the court to protect its property rights.

France also distributes air waves by a state commission. In the nineteenth century, the government decided it did not want a private firm to have the possibility of a natural monopoly, so it forbade anyone to organize any communications without authorization by the state. Then people came to believe there actually was a natural monopoly in broadcasting.

Electricity is often taken to be an industry of natural monopoly, but in fact, electricity is composed of many different parts. There is the actual production, there is transportation, there is distribution, and so on. It cannot be known in advance whether it is optimal to have one or many producers in each segment of this long chain of production. In countries where the market for electricity had been deregulated, it was discovered that some very small plants were highly competitive. The myth had grown that, in the production of electricity, plants had to be bigger and bigger to take advantage of economies of scale, but new techniques could be developed, and economies of scale may not exist forever. The future costs of new techniques can never be known today. The best way to discover new techniques is always the market. Whenever there is competition from free entry, potential producers will be induced to look for new techniques.

Suppose there is free entry to the market but that, at one time, gains can be obtained by some coordination process. In a competitive setting, producers will be able to discover that they can gain by producing at the larger scale, such as by forming a monopoly or by organizing a cartel. Cartels are often considered to be bad in traditional theory, but in my opinion cartels can be good if they are free cartels and not privileged cartels. A large scale of production does not mean that you need only one producer.

Recall that one merit of competition is that it gives rise to greater differentiation. In some cases, however, too much differentiation is not optimal. One obvious example is that of money. If each bank produces its own money, this would result in perhaps too much differentiation. A monetary system is an example of a cartel, but one which is regulated by the state and thus is not a good example of a spontaneous cartel. However, we know that there were private cartels producing gold currency in the eighteenth century. Banks themselves discovered that it was more efficient for each of them to anchor to some process of coordination to make a unique money produced by different producers.

A cartel may be superior to a monopoly because it has the same beneficial effect in terms of scale of production, and yet, may have more potential for innovation. In my view, the instability of cartels is actually an asset of cartels because it means that each member of the cartel is always considering the possibility of leaving the cartel and thus is induced to seek new techniques. Whenever the gains from cooperating with the other producers become lower than the gains from innovation, a member will leave the cartel. Such cartels are free in that they do not result from some privileges.

To end, I would like to address the current debate in Hong Kong on competition policy and its implementation. One possibility is to have an independent commission or judicial committee to be in charge of overall competition policy. However, we must recognize that bureaucrats have a vested interest in developing more and more regulations while justifying them on the grounds that they are necessary to protect consumers or to avoid the risk of monopoly power, etc.

Take the example of the Bell Company in the United States. The Bell Company got a monopoly privilege using the natural monopoly argument. In return for the benefits from this monopoly position, it agreed not to exploit all the possibilities for supraprofits. So there was a form of collusion between the regulators and the firm being regulated. Whenever a firm captures the market by a privilege granted by the state, there is a temptation to look for profits through protection instead of profits by innovation.

The traditional theory of competition defines competition from its results. It says that competition is lacking whenever a firm has too large a market share. In my opinion, competition cannot be defined from results but must be defined from the process, namely, that there is freedom to enter the market. Hayek has been quoted on the importance of processes versus results. A free economy is an economy in which you care only about processes, not about results. In a free society, the results will not be known and will not be modified, but they will have been obtained by free processes. So we cannot have a competition law which inquires into the results.

I also want to stress the risk of an independent commission on competition. First of all, the members would never truly be independent as they are ordinary people with their own ties. Secondly, I would even say it is dangerous that people be totally independent with no controls to ensure fairness. It is a merit of competition that it is a system of external mutual control. I find it ironical to pretend that a commission which benefits from a monopoly position can enforce competition.

What would be the best policy for Hong Kong? I would say that the best policy would be not to create any such organization in charge of competition. Learn from the errors of other countries. Also, Hong Kong should continue its process of deregulation. And finally, it must maintain free trade and free investment. Liberty makes prosperity possible. I truly hope that the so-called economic miracle of Hong Kong will be preserved.

Professor Pascal Salin is professor at the Universite de Paris-Dauphine, France. He is currently President of the Mont Pelerin Society and a member of the Board of Advisers, Hong Kong Centre for Economic Research. The above is an abridged version of a presentation he made during a recent visit to the Centre.


| Index | Research Projects | HKCER Letters |
| Speaker Program / Conference | Index of Economic Freedom |
The Hong Kong Centre for Economic Research
School of Economics and Finance
The University of Hong Kong
Phone: (852) 2547-8313 Fax: (852) 2548-6319
email: hkcer@econ.hku.hk