Competition policy is a contentious issue among policy makers today. Policies differ considerably: some may take a structuralist approach, others a more "pragmatic" or "control" approach. In this paper I wish to firstly discuss both. Secondly, I wish to show how recent work has derived a full definition of the "relevant market" as a tool of policy. In this essay I shall draw extensively from the reports of Fishwick (1983)for the Commission of the European Communities.
The structuralist approach could be considered the pro-active approach to competition. Policy is vigorously directed towards the maintenance of a competitive market structure, in preference to a single market structure which is concentrated or under a dominant monopoly. Active measures are taken to disarm potential and actual monopolies, the logic being that monopolies are not in the public interest, as well as being sub-optimal in the economic sense. The policy of the Commission of the European Communities (now eu) has been largely of the structuralist type since 1963. There has been a gradual trend towards structures as a means of preserving competition in this country. Other countries such as France and, to a certain degree, the uk have also adopted the structuralist approach like the eu. Does this suggest that the policy of a fully integrated Union will be adverse to monopolies and concentrations? The case study evidence would suggest the affirmative. The Commission has been acting as a watchdog on anti-disbanding of cartel-like agreements among dominant firms as called for under Article 85 of the Treaty of Rome which states that there is "a prohibition on all agreements between undertakings, decisions by associations of which have as their objective the prevention, restriction or distortion of competition within the common market" and Article 86: "any abuse by one or more undertakings of a dominant position within the common market or in substantial part of thereof is prohibited as incompatible with the common market in so far as it may effect trade between member States" . Commission policy has, however, in the past conflicted with decisions of appeal by the European Court of Justice concerning cases of alleged abuses of dominance, and of agreements between ventures that allegedly distorted competition or trade in the relevant market.
In contrast to the structuralist approach, American and, to a certain extent, German policy has adopted a less adverse view to monopolies and concentrations. This control approach some would claim is more natural than the structuralist one, the control approach being seen as leaving the market to decide its own course. If this leads to a concentration in the relevant market then the only grounds for intervention are to control abuses of the concentration or monopoly that act against consumer welfare. Concentrations are seen as the zenith of an efficient competitive structure. Monopolies and concentrations can bring with them economies of scale, increased research and development, and so on . According to Fox (1983), us policy has been changed over the past 20 years from a structuralist approach to an approach which realises that "gains in efficiency from mergers or increased concentrations may outweigh the adverse effects on competitive structure" . We could thus conclude here that the "control" approach refers only to supervision of concentrations, monopolies and the like, in contrast to the structuralist approach which in effect frowns upon and avoids such development in the first place.
Having looked at the two main approaches to policy formation in competition policy, I will now examine some of the "technicalities" of enacting policy. When we are discussing the issue of competition in a certain market, we must ask what exactly the market in question is. In other words, what is the good or service in question (the relevant product)? What is the geographical boundary of the market (the relevant geographic market)? What is the relationship to the relevant product of other goods or services; are they substitutes, complements, interdependent goods or otherwise? All these variables are grouped under the term "the relevant market". The question remains: how do we define such a broad definition of variables under one parameter? Do we need to define the relevant market at all? Relevant market analysis can prove detailed and at times seemingly obvious. For example, in a particular case, United Brands Corporation held a dominant position in the market for bananas and was accused by the Commission of abusing this position. The subsequent investigation compared the functional and reactive substitutability of bananas to other fruit under the subjects of texture, shape, flavour and type of consumer! For the purpose of analysis, we can in most cases define the relevant market in its broadest sense.
This is the actual good or service itself, defined in the narrowest terms (x). However, we must broaden this to encompass products which are perfectly substitutable to x. Substitution can be based on the following three criteria :
(i) Functional Interchangeability: does good y, the substitute, have the same physical or technical properties which enable it to do the task of x?
(ii) Reactive Interchangeability: Does y hold the same functionability and utility as a substitute to x, in the mind of the consumer?
(iii) Barriers to Substitution: do barriers to the substitution of x exist (special distribution channels, sunk costs and complementary goods for example.
Indirect tests such as the similar advertising or marketing of both goods may also indicate substitutability. Although neither the Horowitz test nor the Stigler-Sherwin tests for indirect substitution, shall be addressed directly, both are applicable. Direct tests such as cross price elasticity and substitution elasticity are examples of suggested tests of direct substitution. To sum up, though, all substitutes for x must be included in the relevant market. Potential substitutes will be dealt with later.
In order to define the relevant market we must outline the actual physical market in terms of size and scope, and whether it is a national or international market. Two goods could be considered in the same geographical market if suppliers cannot discriminate between them, thus ensuring that prices are equal. We can test for an integrated geographical market by considering whether or not there exists barriers to the area in question, as regards the supply and demand transferability of the relevant product between two geographical markets, and the level of transport costs.
When we suggest potential supply/competition we are not advocating the view that any Tom, Dick or Harry is a potential supplier of the relevant product. We are specifically concerned with the maximum amount that could potentially be supplied where possible. This would include the excess capacity of actual suppliers now, the output of actual suppliers of the relevant product not already sold, and the transferability capacity of "like-minded" production suppliers, in other words the ability and ease of transfer of production from one good to that of the relevant product. If the relevant product is produced in tandem with another good, as occurs often in the chemical industry for instance, then an increase in the profitability of this second good may bring about an increase in the production of the relevant product. Backward/forward vertical integration may also prove to be a potential supplier. Short term barriers to entry to the market may induce the misperception that there is no potential supply outside the current suppliers. Therefore, these barriers to entry must be assessed on a broad external scale. Such barriers may include geographical barriers (transport costs), sunk costs, legal barriers, and so on. These may all change in the long run, thus increasing actual competition.
In order to achieve a fuller, broader picture of the relevant market we may also extend our view to encompass products which may affect the relevant product either directly or indirectly. Vertical interdependence, where the relevant product is an intermediary good is a good example. Here a good, a, may be dependent on another good, b, along the vertical chain of production: for example, a drug for cancer (a) may be dependent on a raw material necessary to produce the drug (b). In order to acknowledge dependence we must consider whether "the buyer" of the relevant product (the drug) is able to substitute away from "the supplier" to a different supplier of the raw materials. This has been a contentious issue in its own right among competition case study history. The question of whether or not the relevant product is also a complementary good also surfaces. If so then there will be a need to examine the cross elasticity of demand in order to gauge the strength of the interdependence.We should note here the concept of "Partenaire Obligatoire". Briefly this states that the above step-by-step analysis may not always be suitable if there exists dependence in the case of a vertical relationship. In the case of an alleged abuse of dominance, such as an infringement of Article 86, the definition of the relevant market may not be of use in uncovering the details of the abuse. "Partenaire Obligatoire" studies the dependence of an obligatory trading partner where no transferable substitution is possible.Interdependence at a horizontal stage would suggest a case where the supplier of the relevant product was in some way dependent on the buyer of the relevant product. This is simple monopsony or oligopsony. These are rare examples and are to an extent mirror images of an abuse by a seller who holds a dominant position. In these cases the definition of the relevant market is necessary, as is perhaps the use of "Partenaire Obligatoire". Test for the strength of such a relationship are somewhat more difficult to procure. However, test for substitution regarding substitue outlets for the supplier to supply to, such as those based on functional and reactive interchangeability are applicable here. The dependence of the independent supplier of door panels specifically for Ford motor cars, on the Ford motor company and subsequently on the sales of Ford cars is a good example.
Finally in order to measure the relative market power of an undertaking in the relevant market a measurement of the relevant market sales may be a good indicator. This, combined with the above analysis of the dominant or interdependent relationships that the undertaking may possess can give a good indication of the strength of the firm in the relevant market.
In this paper we have taken a look at the two main types of policy in contemporary competition policy. Under the guise of various national approaches. We then went on to discuss the rather thorny issue of how one goes about defining the relevant market in cases concerning the competition authorities and undertakings etc. A step-by-step approach was thus derived. The definition of such a broad topic is bound to lead to debate, as indeed it has. It is likely that such a debate will continue into the future, and that the definition of the relevant market will sway in line with those arguments regardless of how broad or narrow they may be. What is important to note is that care must be taken not to ignore factors that will have considerable bearing on the actual market. In this light the steps taken by Fishwick and others, notably the Commission, are valid.
Fishwick, F., (1986): Definition of the Relevant Market in Community Competition Policy, Commission of the European Communities.
Glais, M. & Laurent, P. (1983): Traite d'Economie et de Droit de la Concurrence, PUF, Paris.
Fox, E.M., (1983) : "Abuse of a Dominant Position Under the Treaty of Rome: A Comparison with US Law", Annual Proceedings of the Fordham Corporate Law Institute.
Horowitz, L, (1981) : "Market Definition in Antitrust Analysis: A Regression Based Approach", Southern Economic Journal, Vol. 46.
Stigler, G.J. & Sherwin, R.A., (1985) : "The Extent of the Market", Journal of Law and Economics, Vol. XXVIII.
Commercial Solvents Corporation, (EEC Case), 1972.
Hugin v Lipton, (EEC Case), Dec. 1977.
Replacement Body Panels for Ford Cars (UK Case), Feb. 1985.
United Brands (Bananas - EEC Case), 1985.