تبعیض قیمت مکانی و تناقض ادغام
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|17956||2000||16 صفحه PDF||سفارش دهید||5963 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Regional Science and Urban Economics, Volume 30, Issue 5, September 2000, Pages 491–506
A familiar result in the literature on mergers is that the principal beneficiaries from such activity are the firms which are excluded from participation. The possible existence of this ‘merger paradox’ contrasts strongly with the frequently expressed view that merger is anti-competitive. This paper examines the question within the context of a model of spatial competition in which firms choose their locations in anticipation of forming a merger, and practise price-discrimination. We allow for differences in firms’ shares in the benefits of merger, and for the possibility that the firms will attach probabilities to merger formation.
One of the robust insights afforded by traditional models of oligopolistic behaviour is that the principal beneficiaries from a merger are the non-participants. In the standard Cournot model,1 for example, the parties to a merger exploit their increased market power by restricting output and raising price. But the consequence is that excluded firms increase their output to raise their own profits and, as Salant et al. (1983) have observed, this increase is generally sufficient to offset any profit gain by the merged entity. 2 The literature which has grown since the work of Salant et al. broadly reinforces their original conclusion (see Pepall et al., 1999, for an extensive treatment). The theoretical findings which have emerged stand in marked contrast to the widely held perception that, in practice, mergers are anti-competitive (see Boyer, 1992, for a discussion). Indeed, as White (1988) has observed, the parties most likely to file suits intended to overturn mergers are those firms which are excluded from the process. Excluded rivals may legitimately fear that the merged firm will acquire the larger financial resources necessary for successful predation, in the form of ‘dumping’ or other types of underpricing, or for encroaching on established product lines. There may arise also the perception that competitors which are enlarged by merger may be in a position, through the use of potentially resource intensive pre-emptive tactics, to block the expansion of smaller rivals into other markets. In any case, given the conflict between theory and experience, the natural question is whether there can be found circumstances in which horizontal mergers can be simultaneously profitable for the participants (a requirement which must be satisfied if the merger is to take place), anti-competitive (the basis for the filing of a suit) and socially inefficient (a justification for government intervention). This paper addresses that question. The approach we take is based upon a model of spatial competition in which firms practise price discrimination. The ‘spatial’ approach to the question of collusion and merger is not in itself new. Writers such as Jehiel (1992) and Friedman and Thisse (1993) have investigated differing degrees of collusion between spatially competitive firms, but on the assumption that customers pay transport costs (so-called f.o.b pricing). The work of Levy and Reitzes (1992) is similarly focused. Our justification for considering price discrimination is that, as several writers have noted (see, for instance, Thisse and Vives, 1988; Norman and Thisse, 1996), such pricing may be preferred by both customers and firms over f.o.b pricing. This difference between our approach and that of Jehiel and Friedman–Thisse, aside, the importance of their work is that it shows the extent to which location choices which are made in anticipation of the possibility of merger may be different from those which are not.3 This idea is an entirely reasonable one, and we accord it a central position in the present paper. Previous research allowing the possibility of price discrimination, for example Reitzes and Levy (1995),4 does not consider the impact of anticipated merger on location choice. As a consequence, merger raises the prices set by the merged firms while leaving unaffected the optimal discriminatory prices of excluded rivals. Moreover, since merger simply results in a transfer from customers to the merged entity, there are no welfare effects. The present paper considers within a framework of spatial price discrimination the possibility that (a) in making their location choices, the parties to a prospective merger, as well as those which will be excluded, anticipate that such merger will take place, and (b) the profits from such merger may not be equally shared amongst the participants.5 In these two respects the paper is an extension, to a three-firm environment, of work by Gupta et al. (1997). Their paper involves a duopoly, but no ‘outside’ firms which might be excluded from a merger and, therefore, cannot examine the merger paradox. Moreover, in Gupta et al., the maximum price which firms can set post-merger is a constant. By contrast, in the present analysis this maximum depends, in a sense to be made precise below, upon the existence and location of the outside firm. Many of the practical examples of the theoretical problem we consider here follow from interpreting the spatial dimension as an ordered product characteristic (see Schmalensee and Thisse, 1988; Tirole, 1988). For example, newspapers or magazines differ from one another in their editorial policies. For each publication, editorial policy could, in principle, be represented as a ‘location’ decision on a political spectrum running from ‘left’ to ‘right’. The worldwide wave of mergers and consolidations which has taken place among newspapers and magazines should then help to determine the editorial policy (location) to be adopted by a new entrant. In turn, the editorial policy of the entrant will influence the profits associated with a potential merger amongst existing publications. A second example involves departure ‘slots’ at airports. In this case, the spatial dimension can be thought of as representing departure times over an interval running from early morning to late evening. Here, a new entrant’s choice of departure time (location on the interval) is influenced by the departure times offered by existing airlines. The profitability (and proximity to one another) of the departure times offered by the latter is determined in part by opportunities for merger, as well as by the departure slot chosen by the entrant. Recent work in other branches of industrial organization theory also touches upon some of the issues to be considered here. For example, Eaton and Schmitt (1994) address the question of flexible manufacturing in a spatial framework. In their discussion the delivered price schedules of the model of spatial price discrimination are replaced by marginal cost schedules which reflect the (increasing) costs of producing progressively larger variations on a basic product. The two schedules are in this sense analogous. The authors investigate the relationship between, on the one hand, the propensity to merge and, on the other, the profitability of market entry by firms which remain outside of the merger. They argue, inter alia, that, given the products produced by the established firms, the profitability of entry is independent of whether or not merger occurs. It will become clear that the approach which we adopt here bears also upon this question. The paper proceeds as follows. In the next section we set out the model in detail; Section 3 analyses the implications of merger activity involving two firms and one excluded rival; Section 4 offers some concluding comments.
نتیجه گیری انگلیسی
The ‘merger paradox’ arises when the principal beneficiaries to a merger are those firms which are excluded from the process. As noted in the Introduction, the theoretical possibility that this phenomenon might exist sits somewhat uncomfortably alongside the popular perception that, in general, merger harms outsiders. The purpose of this paper has been to attempt a reconciliation of theory and practical experience. Our interest has therefore focused on the circumstances in which merger is simultaneously (a) individually rational, (b) anti-competitive and (c) welfare reducing. Our results show that, first, on one criterion of welfare (viz transport costs), all merger, irrespective of the distribution of benefits between the firms involved, reduces social welfare. We have also been able to show that the magnitude of the effect on welfare does depend upon the precise distribution of these benefits. Second, our results show that when merger is individually rational, the gains to the participants exceed that of the excluded firm. Third, with or without merger, the ‘middle’ firm (L2) obtains lower profits than does the potentially excluded rival. 13 Moreover, as the share of L1 in the gain from merger increases, all firms move rightwards, and consequently the profits of both L2 and the excluded rival fall. Fourth, there exists a wide range of α for which merger is individually rational (that is, [0.34,0.89], approximately). Yet, only a small part of this range ([0.87,0.89], approximately) is associated with harm to the excluded rival. Thus, individually rational participation is generally not anti-competitive. With the exception of the narrow range identified above, anti-competitive mergers occur only when L2 is ‘pressured’ into unprofitable participation. The most damaging instance of the phenomenon occurs when α=1, when the negative deviation (for both L2 and L3) from the no-merger level of profits is maximised. Fifth, when merger does not occur with certainty (that is, when 1>ρ≥0) the profits of the three firms can be shown to reflect the impact of changes in α, given ρ, and changes in ρ, given α. In all of these cases, it is clear that, irrespective of ρ, the party with the larger expected share in incremental profits gains at the expense of the other firms. Conversely, given any expected allocation of incremental profits between the parties to the merger, and subject to both the individual rationality requirement and the effect of large α on the excluded rival, the profits of all firms will in general be lower as the probability of merger is smaller.