2.2. Consequences of Recent Theoretical Developments

A highly concentrated market structure was the central most important indicator for the presence of market power within the Harvard approach. The New Industrial Organisation on the one hand, and Transaction Cost Economics on the other, have pointed to some problems that might arise when structure is the central preoccupation of competition policy.

Game Theory stresses the possible divergence between individually rational behaviour and collectively beneficial results: in certain interaction situations, rational individual behaviour will lead to suboptimal results on the collective level, Adam Smith’s “invisible hand” will thus not work in some cases. The most famous of these interaction situations is, of course, the Prisoners’ Dilemma. Game Theory has now shown that even duopolists who could make themselves better off by explicit or tacit collusion will not necessarily be able to reach that result if the incentives not to co-operate with the other player are sufficiently large. Structure as such would thus be meaningless, other factors that increase the likelihood of successfully implementing joint maximisation strategies have to be checked for. Section 4 of this chapter contains an extended checklist that can be used to assess the likelihood of collusion taking insights from game theory explicitly into account.

Representatives of Transaction Cost Economics point to the fact that obsession with structure might lead one to overemphasise the dangers of strategies pursued in attempts to monopolise and to underestimate the benefits of strategies pursued in attempts to economise. Williamson has often referred to the Harvard approach as the “inhospitality tradition” (e.g., 1985, 369f.). As was shown in more detail in chapter II, Williamson (1968) has made it very clear that trade-offs are inevitable. Suppose two merged companies are able to pursue the strategy of a monopolist (i.e., increase price to marginal revenue and reduce output accordingly). Focusing exclusively on market shares, such a merger would have to be prohibited. Williamson now points out that this conclusion might be blatantly wrong from a welfare-economic point of view: Before drawing conclusions based on having had a look only on the cost side of the merger, one should take a look at possible benefits, namely possible cost savings based on economies of scale or of scope. If these outweigh the welfare losses, the proposed merger should be waved through. This is thus an argument in favour of an efficiency defence.

In a study commissioned by DG Comp, Röller et al. (2000) have dealt with efficiencies and proposed to take rationalisation, economies of scale, technological progress, purchasing economies and reduction of slack explicitly into account.

Before an explicit case in favour of the introduction of efficiency as a merger defence into merger policy can be made, one fundamental problem must be convincingly solved. If efficiencies are a merger defence, firms willing to merge have an incentive to grossly overstate the amount of efficiencies generated as a consequence of the merger. Similarly, directly affected third parties that expect synergies to be substantial and who are, correspondingly, afraid of tougher competition have an incentive to belittle their expected size. None of the concerned parties thus has an incentive to reveal information truthfully.15What is thus needed is a mechanism to reveal the information as truthfully as possible. Two general paths are possible: (1) improving the incentives of the directly involved parties to reveal their private information at least to a certain degree and (2) having the revealed information monitored by outside experts. These two ways can be combined and are thus complementary.

Williamson was not only the first to point to the nature between welfare-gains (efficiencies) and welfare-losses (market power) of mergers but also the one who stressed that the realisation of efficiencies was not necessarily confined to cost savings in production but could possibly be extended to cost savings in transaction costs. In order to take this possibility adequately into account, one should therefore analyse the three determinants that Williamson believes to be decisive for an optimal firm structure, namely the degree of asset specificity, the degree of uncertainty present in those activities that two firms aim to incorporate into one hierarchical structure, and the frequency with which these transactions will take place. Here, size might be the consequence of economising on transaction costs and not an indicator for market power or dominance.

In addition to some theoretical developments, progress in empirical techniques might also impact upon the necessity to delineate relevant markets. The disadvantages of market definition are widely acknowledged by now. If one could assess dominance without having to resort to market definition, this could be an improvement. It has been argued that the widespread use of scanner technology allows exact estimates of own-price and cross-price elasticities (the original contribution is Deaton and Muellbauer 1981).16 Based on these estimates, the extent to which a
planned merger would reduce competition could be calculated.