2.1.1. The relevant product market

Market power is said to be present if a producer can profitably raise prices by, e.g., five or ten per cent. Usually, increases in prices lead to a fall in demand. The price elasticity of a product is defined as the percentage change in demand in relation to a one per cent price change. If a five per cent price increase leads to a decrease in demand of ten per cent, the corresponding price elasticity would thus be –2. If the price elasticity of a product is very low in absolute terms (i.e., is close to zero), this is a sign that no close substitutes for this product are available, otherwise consumers would switch demand to a competing product. The lower the price elasticity of a product in absolute terms, the higher the market power of its producer.

Cross-price elasticities are defined as the percentage change in demand of a good y in relation to a one per cent price change of a good x. If price increases of good x do not lead customers to switch from x to y, then they will not be considered as substitutes. A low cross-price elasticity is thus an indicator that the two products do not belong to the same product market, whereas a high cross-price elasticity would be an indicator that they do belong to the same market.

Competition policy experts have developed a variety of procedures that take these basic relationships into account in order to delineate the relevant market. The first and probably most widely used one is based on a “reasonable consumer” and the question whether she would consider two products as substitutable. If she interprets them as substitutable, they belong to the same market. This concept is often referred to as demand substitutability.

There is one fundamental problem with this concept, namely its subjectivity. It is simply impossible to define a “reasonable consumer” by a number of abstract standards or criteria. If each of a dozen antitrust experts sits behind his/her desk and thinks very seriously about the products a reasonable consumer will consider as close or adequate substitutes for a particular product, each one of them will supposedly come up with a list for which a number of arguments can be named. But the chance that any of the twelve lists are identical to any other list should be pretty slim. In other words, as long as research concerning substitutability is done only by introspection, its results will remain very subjective.

Economists often pretend to know all the relevant preferences of actors without ever actually asking them. One could now be inclined to think that actually surveying consumers might lead to less subjective results. Surveys might indeed help to come up with some quantitative measure. Yet, the subjectivity of the “reasonable consumer” can never be completely eliminated. Take as an example the invited guest to a dinner party who wants to give a little present to his host. For him, a bottle of wine might be just as good as flowers, a compact disk or a new book. The list is surely incomplete. But how exactly it is extended depends on individual preferences. Explicitly recognising this is difficult even when actual consumers are asked: In order to take this into account, surveys would have to be constructed as open-ended surveys. Depending on time constraints and incentives of the interviewed, one can expect very long individual lists. Whether they contain substantial overlap with each other is a second question.

Until now, we have implicitly assumed that other suppliers of a product will not react if one firm changes its prices. This is, however, not necessarily true. A given price increase could make the production of close substitutes, and thus market entry, profitable. This is often referred to as supply substitutability.