2.4.1. The Relevant Product Market

Although the concepts of demand-side substitutability and supply-side substitutability are widely recognised by the Commission (see the Commission’s Notice) and the Community Courts, they have not been applied consistently in the past. The test of demand-side substitutability employed by the Commission seeks to determine which products are sufficiently similar to be regarded by users as reasonable substitutes for one other. In the case of AÉROSPATIALE-ALENIA /DE HAVILLAND, the Commission stated that a relevant product market comprises, in particular, all those products that are regarded as interchangeable or substitutable by the consumer with regard to their characteristics, their prices and their intended use.

This Commission’s practice concerning the recognition of supply side effects has been subject to harsh criticism. Neven et al. (1993, 77) wrote: ”The procedures used for market definition are frequently inconsistent; in particular supply substitution is sometimes taken into account at the market definition stage and sometimes at the stage of assessing dominance, with usually no clear rationale for the difference in treatment. Supply substitution is also sometimes counted twice. The failure to take supply substitution into account will probably tend on average to result in excessively narrow market definitions. Although it is not possible to point with confidence to particular cases in which this bias has made a difference to the market definition adopted, we indicate one or two instances where it could have been significant.” This evaluation was written some two years after the European Merger Regulation had been passed. Here, we set out to ask whether the Commission’s practice has been modified during the last ten years.

At times, the Commission points to supply-side substitutability. If this is the case, it is often, however, not incorporated into the relevant market. An example for this practice is VOLVO/SCANIA.

While the concept of supply-side substitution is recognised in practice, experience suggests that the Commission’s delineation of the relevant market focuses principally on demand-side considerations – supply-side substitution, if it is considered at all, tends to be more of an after-thought. Moreover, supply-side considerations play no role in the product market definition according to Form CO. The Commission’s view of whether two products or regions should be included in the same relevant market thus depends almost exclusively on their substitutability from the perspective of consumers.

A worrying aspect of the rare use of supply-side considerations by the Commission is that the Commission defines separate markets on supply-side where demandside considerations would suggest a single market. In these cases, the Commission used differences in conditions of competition and distribution to identify different product markets, although the products themselves were fully interchangeable or even identical. In the case of VARTA/BOSCH, e.g., the Commission introduced a distinction between starter batteries supplied to the original equipment market and those supplied to the replacement market. However, the Commission did not consider the relevant question whether suppliers of replacement batteries would switch to supplying original equipment batteries if prices of the latter were to rise. Another example is the case of SCA/METSÄ TISSUE: here, the Commission defines separate markets for branded and private label products of toilet tissues and kitchen towels, although the products were produced by the same producers using the same technology and identical machines.

This approach is problematic and cannot be easily reconciled with the concept of the relevant market. From the perspective of economic theory, a relevant market is defined as a set of products worth monopolising. This has entered into antitrust practice by way of the SSNIP test that was described above. As long as a market conjectured to be the relevant one is not sufficiently attractive for being monopolised, it can thus not be the relevant market and needs to be broadened.

When defining the relevant product market, the Commission has focused its analysis primarily on three factors:

(1) Physical characteristics of the product and service The European Commission has stated that if two products are physically very different to the extent that that they cannot in fact be used for the same end use, they will not be considered to be substitutes. In the cases of RENAULT/VOLVO and VOLVO/SCANIA, e.g., the Commission defined separate markets for trucks (below 5 tons, between 5 and 16 tons, and over 16 tons), on the grounds that they were technically very different and had different end uses. If consumers in the segment between 5 and 16 tons can, in fact, not easily use larger trucks for the same purposes as the smaller ones, then the delineation of two distinct product markets seems to be justified. However, it appears at least not unlikely that there are also consumers who would substitute between the two segments under some circumstances. For instance, some consumers, in response to an increase in the price of 17-ton trucks, might choose to buy a 15-ton truck instead. Conversely, if the price of a 15-ton truck were to increase, some customers might switch to buying a 16 or 17-ton truck instead. Therefore, the question is whether enough consumers would switch demand in order to make an attempt to increase prices in one category unprofitable.

In the cases of MERCEDES-BENZ/KÄSSBOHRER, VOLVO/SCANIA and MAN/AUWÄRTER, the Commission distinguished three individual segments of bus markets: the market for city buses, for inter-city buses and for touring coaches. The Commission defined the market narrowly, in part, because there were differences in the specific use of the bus. Hence, it was argued that there was no demandside substitutability between low-floor city buses and double-decker touring coaches with toilet, kitchen and video. The various types of buses can be produced using exactly the same machinery. Hence, the possibility of supply-side substitution imposes constraints on the pricing of the various types of buses. This suggests that the relevant product market did indeed encompass all three specific types of bus.

(2) Product prices
The Commission has often inferred that two products are not reasonably substitutable if they have substantially different prices. Examples for this line of reasoning can be found in NESTLÉ/PERRIER, PROCTER&GAMBLE/VP SCHICKEDANZ, AÉROSPATIALE-ALENIA/DE HAVILLAND; KIMBERLY CLARK/SCOTT PAPER. Price differences have therefore been used to distinguish between products that are perceived as different products by consumers but that may be functionally substitutable. If differences in price are caused by different characteristics of the good, they can, indeed, justify the argument for the existence of more than one relevant market. But at times, differences in price are completely unrelated to functional substitutability such as between branded labels and distributors’ own labels. Differences in price are thus not a sufficient condition for delineating two or even more different markets. Were the price of a premium product to raise substantially, consumers might be willing to change to another good, which is not in the premium segment of the market but which has been cheaper all along.

The argument can be highlighted by drawing on a real-world example: In the cases of KIMBERLY-CLARK/SCOTT and SCA/METSÄ TISSUE, branded products compete with unbranded, i.e., private label equivalents. A question central to the assessment of competition was whether private label toilet tissue was in the same relevant market as branded tissues. Often, branded tissue costs more than private label tissue, despite the fact that there is little to distinguish between them on a physical basis. Indeed, manufacturers of branded products produce many private label products. Using differences in absolute prices to delineate relevant markets would place the branded product and private label product in separate relevant markets. It is obviously the case that part of this price differential is due to perceived quality differences, which is valued by consumers. Still, one would have to ask whether the pricing of branded label products was constrained by the prices of private label products. This is clearly the case; hence, price differences are not sufficient for identifying two different markets. Rather, it is the interdependence of prices that matters.

(3) Consumer Preferences
The Commission also regards consumer preferences as relevant to delineating relevant markets. Despite the existence of substitutes at similar prices, the Commission may hold that consumer loyalty will limit substitution away from the product concerned following a price rise. Whilst question marks can be put against using consumer loyalty for the purpose of either market definition or barriers to entry, issue with this fundamental question is not taken here. If consumer loyalty is particularly high, then a price rise by a hypothetical monopolist may not induce substitution, suggesting separate markets. But in most decisions it is unclear how the Commission could even measure the extent of consumer loyalty. Therefore, without an empirical
test about the importance of brand or consumer loyalty, this argument introduces high subjectivity into the assessment of competition. In section 2.5 of this chapter, we develop a proposal how the relevance of customer loyalty for the delineation of the relevant product market can be assessed on a systematic basis.