Take Efficiencies Explicitly into Consideration
As spelt out above, mergers may increase market power but still be welfareincreasing if they enable the merging companies to realise efficiency gains. In U.S. merger policy, the efficiency defence has been around for a while: Efficiencies are explicitly mentioned in the U.S. Merger Guidelines of 1992, but they had already been applied based on a publication of the Federal Trade Commission from 1984. The 1997 revision of the Guidelines meant a more restrictive application of efficiencies. The Guidelines now demand that the proof of cost advantages be clear and unequivocal; they specify more precisely what is meant by efficiencies. Some other legislatures have followed suit and have also incorporated an efficiency defence into their merger law. It can now be found in the U.K., but also in Australia, Canada, and New Zealand. The Guidelines on horizontal mergers that went into effect in May 2004 also contain an efficiency defence. Factually, efficiency considerations had, however, already played some role even before (see, e.g., NORDIC SATELLITE, AÉROSPATIALE/DE HAVILLAND, and HOLLAND MEDIA GROEP).
Efficiency Defence in the US
Drawing on the various versions of the US Merger Guidelines issued between 1968 and 1997, it is possible to describe the development of the relevance that efficiencies have played in US Merger Control Policy. The Guidelines that were issued in 1968 (U.S. Department of Justice 1968) are very restrictive concerning the possibility to take efficiency effects explicitly into account. Cost advantages as a reason to justify a merger were basically not accepted by the Department of Justice, except if exceptional circumstances were present. The realization of product and process innovations were the prime candidates for such exceptional circumstances. In its decisions, the Department of Justice consistently denied the presence of such exceptional circumstances. The main reasons offered by the Department of Justice for its restrictive stance on accepting efficiencies as a justification for mergers were that cost savings could also be realized through internal firm growth and that efficiency claims were notoriously difficult to verify.
This critical stance was retained in the 1982 version of the Guidelines (U.S. Department of Justice 1982). They reiterated that efficiencies could only play a role in extraordinary cases However, these extraordinary cases did not play a significant role in merger policy. In 1984, the Department of Justice explicitly introduced efficiencies into the Merger Guidelines (U.S. Department of Justice 1984). According to them, significant efficiencies could play a role if there was “clear and convincing evidence“ in their favour and it was impossible to realize them through alternative means. This was thus the first time that efficiencies appeared as a criterion of their own for justifying mergers. But simply pointing at expected efficiencies was never sufficient for getting a merger passed. They could only play a role in conjunction with other reasons.
The 1992 revision of the Merger Guidelines did not lead to substantial changes with regard to efficiencies (U.S. Department of Justice/FTC 1992). But the precondition of “clear and convincing evidence” was cancelled. This was not, however, accompanied by a reversion of the burden of proof. It is thus still with the merging firms. All in all, this made it easier to draw on the efficiency defence (Stockum 1993). Finally, the 1997 version of the Guidelines made the rather general formulations with regard to efficiencies more concrete (U.S. Department of Justice/FTC 1997).
The Revision of the Merger Guidelines of April 1997 served the purpose of making the recognition of efficiencies more concrete. Clearly specified criteria are to make it easier for firms to advance efficiency arguments and easier for courts to decide cases in which efficiencies play a role (Kinne 2000, 146). The Merger Guidelines (1997, 31) demand that “... merging firms must substantiate efficiency claims so that the Agency can verify by reasonable means the likelihood and magnitude of each asserted efficiency, how and when each would be achieved (and any costs of doing so), how each would enhance the merged firm’s ability and incentive to compete, and why each would be mergerspecific.” And further: “Efficiency claims will not be considered if they are vague or speculative or otherwise cannot be verified by reasonable means.” Relevant from a competition policy point of view are the so-called „cognisable efficiencies.“ These are savings in marginal costs, which can be realised in the short run. Cost savings in the areas of Research and Development, procurement or management are classified as not being sufficiently verifiable and in that sense not „cognisable.“
Efficiency aspects are estimated by following a three-step procedure:
(1) Determination of merger specific efficiencies Only cost savings caused by the merger itself are recognised. Efficiency advantages caused by increased market power must not be recognised.
(2) Analysis of the relevance of merger specific efficiencies The merging firms must document
– when, how, and to what degree and having to incur what costs efficiency gains
will be realised,
– why the claimed efficiencies can only be realised by a merger, and
– how these efficiencies will affect the competitiveness of the merging firms.
If the firm’s documentation is to the satisfaction of the competition authorities, the efficiencies will be evaluated as “cognisable.“
(3) Evaluation
After having estimated the size of the expected efficiencies they will be compared with the disadvantages that the consumers will most likely have to incur. The higher the expected disadvantages (measured by post-merger HHI, possible unilateral effects, and the relevance of potential competition), the more “... extraordinarily great cognisable efficiencies would be necessary to prevent the merger from being anticompetitive” (Merger Guidelines, 32). The creation of a monopoly as the consequence of a merger can never be compensated by efficiency arguments.
The 1997 version of the Guidelines clarified under what conditions efficiencies could be taken into account in notified mergers. Compared to the vague formulations of the 1992 Guidelines, this meant a substantial improvement. It is thus encouraging that the European Commission used the 1997 version of the U.S. Guidelines as a model rather than the earlier versions.
An increase in the predictability of European Merger Policy is possible if the basis on which efficiency considerations are taken into account are spelt out explicitly. Firms can then form expectations on whether the efficiencies that they expect to realise as a consequence of a merger will be taken into account by the Commission or not. But given that predictions concerning the size of realisable efficiencies rest on shaky grounds, there will be quarrels as to realistic levels. This will limit the gains in predictability.
The inclusion of efficiencies in the recently published Guidelines is thus welcomed. There, the Commission demands that the efficiencies have to “benefit consumers, be merger-specific, and be verifiable.” The Commission loosely follows the US-Merger Guidelines here. Unfortunately, the Guidelines contain a number of indeterminate legal terms, which need to be made more concrete. The terms just mentioned are described but concrete conditions that need to be met for efficiencies to be taken into account are not named. The US-Merger Guidelines are more concrete here and predictability could be further increased if the Commission were to follow the US Guidelines in this regard too.
The main issue to be decided for the inclusion of efficiencies are the criteria that need to be fulfilled. In the US, it is three criteria, namely (1) that gains in efficiency need to be passed on to consumers in the form of lower prices or increased quality, (2) that efficiencies must be merger-specific, and (3) that they must be verifiable. As just pointed out, the Commission is to use similar criteria in Europe. In order to ascertain possible consequences, these three criteria will be dealt with in a little more detail.
The probability that efficiencies will be passed on to consumers appears to be higher if they originate in lower marginal costs rather than in lower fixed costs. Expected cost savings in, e.g., the administrative overhead, do not have an impact on the marginal costs of a company. Reductions in its fixed costs will, however, only be passed on to consumers if the degree of competition is sufficiently high on a given market. But the reason for drawing on efficiencies as a justification for letting a merger pass is exactly that the creation of a market dominant position is suspected. It makes thus sense to weigh reductions in marginal costs higher than reductions in fixed costs. Reductions in marginal costs are to be expected if the merged entity is expected to be able to realise efficiencies on the supply side as well as during its production process, e.g., via economies of scale or scope. The more important the input factors on which the merged entity can save, the more important can the efficiencies expected to be. This is reflected in the Guidelines which attribute more weight to savings in marginal costs but do not exclude that savings in fixed costs might also lead to efficiencies that might benefit the consumer.
Additionally, efficiencies that can be realised in R & D are also highly welcome. Improved R & D will not only increase efficiency of the undertakings concerned, but will put other companies under pressure to improve their R & D. Overall welfare gains are thus to be expected.
The criterion that efficiencies need be merger-specific is more problematic than estimating whether efficiencies will be passed on to consumers. The basic notion is, of course, quite straightforward: efficiencies are only to be recognised as an offsetting argument if the merger is the only way to realise them. If there are other ways, such as internal growth, joint ventures, licensing agreements or other ways of cooperation between the firms, efficiency-gains are not to be used as an argument offsetting the creation of a market position which gives rise to serious concerns. From an economic point of view, the notion is not as straightforward as it first might seem.
In order to make the point, we draw on transaction cost economics, whose representatives are used to think in alternative institutional arrangements. Suppose that the management of an undertaking has incentives to choose that institutional arrangement that promises to maximise profits. It is then unclear why management should opt for a merger if an institutional alternative such as a joint venture or a license agreement could do the trick. A merger regularly involves high organisation costs. These costs will only be incurred if institutional alternatives are expected to be less beneficial. Moreover, it remains unclear why a joint venture is to be preferred over a merger. Joint ventures have often been the source of some cartel-like arrangements. Merger-specificity thus appears as a problematic criterion that should play no role in merger policy.
Critics might object that the criterion of merger specificity primarily serves to prevent the genesis of allocative inefficiencies that are due to market power resulting from a merger. But the efficiency defence will only trump a prohibition if the merger is expected to make the consumer better off, i.e. if it increases consumer surplus. But if post-merger prices need to be lower than pre-merger prices, the probability of allocative inefficiencies appears to be negligible. If, on the other hand, the criterion of merger specificity is applied rigidly, the could mean that mergers will be prohibited – or not even projected – although they would increase overall welfare. In order to protect consumers against profits due to market power, checking for the change in consumer surplus is sufficient.
Empirical Evidence on Efficiencies
The effects of mergers on the efficiency of merged firms have been subject to extensive empirical analysis. The main results can be summarized as follows (Gugler et al. 2003, Mueller 1997, Tichy 2001): Mergers often lead to higher profit levels while turnover decreases. This would mean that mergers increase market power but not efficiency. According to the available empirical studies, these effects appear independent of the sector or the country in which the merger occurs. The reduction in turnover is more pronounced in conglomerate mergers than in horizontal mergers (Caves/Barton 1990). On the other hand, and in contrast to the results just reported, one half of all mergers lead to a reduction in profits and in turnover which would mean that mergers lead to reduced efficiency (Gugler et al. 2003). If profits increase, these increases can be explained with increased market power as well as with increased efficiency. Market power effects are correlated with large mergers, efficiency effects with small mergers (Kleinert/Klodt 2001).
Concerning the relationship between mergers and technical progress, some studies did not find any significant correlation between the two (Healy/Palepu/ Ruback 1992). Some studies have found a negative correlation (Geroski 1990, Blundell, Griffith, Reened 1995), while others found a weakly positive correlation between mergers and technical progress (Hamill/Castledine 1986). The authors pointed at the possible relevance of the acquisition of patents as the consequence of mergers. The empirical results are thus not unequivocal and it is difficult to draw policy implications.
The few cases in which positive efficiencies as a consequence of mergers appear to be unequivocal seem to have occurred in mergers in which the merging parties were producing very similar products. Efficiencies are thus more likely in horizontal mergers with the merging firms producing close substitutes (Ravenscraft/Scherer 1989, Gugler et al. 2003). On the other hand, conglomerate mergers seem to be rather unlikely to lead to additional efficiencies. With regard to vertical mergers, no significant savings in transaction costs were found. Rather, the increased difficulty of entering onto a market and the extension of dominant positions in upstream or downstream markets seem to dominate (Gugler et al. 2003).
The empirical evidence with regard to the efficiency-enhancing effects of mergers is thus mixed at best. Yet, this mixed empirical evidence is not a sufficient reason for not relying on efficiencies in merger control. It is not entirely clear whether these studies do indeed measure what they pretend to measure. The most frequently relied upon indicator for the success – or failure – of a merger is the development of the profits, share prices or the return on turnover. The connection between these two indicators and the efficiency of a firm is, however, everything but clear-cut. Moreover, it is in the nature of these tests that the development of these indicators in the absence of a merger must remain systematically unknown.
This means that it would be premature to conclude from lower share prices or reduced profit margins that the merger must have been inefficient. The available studies provide some important insights into the conditions under which mergers are likely to be a success or a failure. But they appear to be far from conclusive. This is why they do not provide sufficient evidence against the incorporation of an efficiency defence into merger control.
The critical issue in the recognition of efficiencies certainly is the capacity to assess them. It is quite comprehensible that the burden of proof should be with the entities that want to merge. But that does not solve the problem of information asymmetries. It was already pointed out above that the main problem with the explicit consideration of efficiencies is that none of the concerned actors has any incentives to reveal them according to their true expectations. The pragmatic question thus is whether any second-best mechanisms can be thought of. Over the last number of years, a literature on the virtues of independent agencies has developed. It started out with the analysis of the effects of independent central banks but has been extended to a number of areas (Voigt and Salzberger 2002 is an overview). It is conceivable to delegate the task to evaluate the realisable efficiencies of a proposed merger to an independent “efficiency agency” that would specialise in such estimates. A similar suggestion has already been made by Neven et al. (1993). Unfortunately, the Guidelines on horizontal mergers did not realize any of these proposals. A number of indeterminate legal terms is used, but they are not made sufficiently concrete.
Assess Importance of Asset Specificity
When describing the insights of Transaction Cost Economics, it was pointed out that (1) asset specificity, (2) uncertainty, and (3) the frequency of interactions all played into the optimal governance structure. It was assumed that firms tried to economise on transaction costs and that unified governance – i.e., large firms – could be the result. This means that transactions cost arguments are basically efficiency arguments. They are dealt with separately here because they are intimately connected with one specific theoretical development, namely Transaction Cost Economics. Once invested, highly specific assets make the firm that has invested them subject to
opportunistic behaviour by its interaction partners. This might lead to investment rates below the optimum level. It can therefore be in the interest of both sides of an interaction to agree on specific governance structures in order to reduce the risk of being exposed to opportunism. This insight has potential consequences for merger policy: the higher the degree and amount of specific assets involved, the better the justification for a unified governance structure, in this case for a merger. In order to be able to take asset specificity explicitly into account, one eitherneeds to measure it or to use proxies for it. As described in the part on Transaction Cost Economics in chapter II, four kinds of asset specificity are usually distinguished, namely (1) site specificity (costs of geographical relocation are great), (2) physical asset specificity (relationship-specific equipment), (3) human asset specificity (learning-by-doing, especially in teams comprising various stages of the production process), and (4) dedicated assets (investments that are incurred due to one specific transaction with one specific customer). Physical proximity of contracting firms has been used as a proxy for site specificity (e.g., by Joskow 1985, 1987, 1990
and Spiller 1985) and R&D expenditure as a proxy for physical asset specificity. With regard to both human asset specificity and dedicated assets, survey data have been used.
It is thus possible to get to grips with asset specificity empirically. Since the merger rationale in cases of high asset specificity is quite straightforward, it shouldbe taken into account explicitly.
Assess Importance of Uncertainty
The theoretical argument concerning uncertainty has great similarities with the argument concerning asset specificity: if interactions could be made beneficial for all parties concerned, they might still not take place if too high a degree of uncertainty is involved. In cases like that, welfare could be increased if the interested parties are allowed to form a common governance structure in order to cope with uncertainty. With regard to merger policy, this means that in mergers in which uncertainty plays an important role, the evaluation should be somewhat less strict than in cases in which uncertainty is marginal.
Getting to grips with uncertainty empirically is no mean feat. In the literature, various proxies have been discussed; volatility in sales is one of them. Others (Walker and Weber 1984, 1987) have proposed to focus on one specific kind of uncertainty, namely “technological uncertainty”, measured as the frequency of changes in product specification and the probability of technological improvements. Given that technological uncertainty seems to have dramatically increased, it seems worthwhile to take it into account explicitly. The argument is that mergers are more likely in markets with high uncertainty as proxied by high volatility in sales or high technological uncertainty. These mergers are potentially welfare-increasing and should thus be passed.
Assess Importance of Frequency
Frequency is the last component to the Transaction Cost Economic Trias of asset specificity, uncertainty, and frequency. The argument is that the more frequent interactions between two specified parties take place, the higher the potential benefits from a unified governance structure. The implications for merger policy are obvious: assess the frequency with which parties willing to merge interact. The more frequent it is, the higher the chance that efficiencies could be realised, the more relaxed the competition policy stance should be.
Posts
