欧洲改革合并审查：针对问题领域的政策结果-Reforming European Merger Review: Targeting Problem Areas in Policy Outcomes
04-23, 2014
欧洲改革合并审查：针对问题领域的政策结果-Reforming European Merger Review: Targeting Problem Areas in Policy Outcomes
Abstract 摘要
This paper is an attempt to derive priorities for the reform of European merger policy from observed problems with decision making in merger cases. We identify problems in the application of theory and empirical method as well as the impact of resource constraints and the potential for systematic decision bias. In contrast to the intense debate about the merits of a switch from a dominance criterion to a significant lessening of competition test, we find that such a switch would not effectively address any of the problems in Commission decision making on mergers. Similarly, the lack of an efficiency defense does not seem to explain any of the weaknesses of European merger policy. In contrast, we find that most of the problem areas identified in current merger control practice can be effectively addressed by reforming the merger procedure, the internal organization of merger control in the Commission, as well as addressing resource issues. Together with carefully crafted merger guidelines these should have strong priority for the reform of merger policy in Europe.
本文就是试图优先获得观察到的问题与决策对欧洲并购政策的改革。我们确定在理论上和实证方法的应用，以及资源限制和系统决策偏差的潜在影响的问题。相反，大约从一家独大的标准切换到竞争考验的显著减持优劣的激烈辩论中，我们发现，这种委员会的决策不能有效地解决任何问题。同样，缺乏效率的防守似乎并没有解释任何欧洲并购政策的弱点。相反，我们发现大多数的合并控制实践中发现问题的地方都可以通过改革合并，委员会合并控制的内部组织，以及资源问题得到有效解决。加上精心打造的并购指引这些有很强的优先级的在欧洲并购政策的改革。
1.Systematic concerns about EU merger policy 关于欧盟并购政策系统性的担忧
Businesses and practitioners often express their concerns about different issues in the merger process in terms of ``fairness'' of the process. But many times the word` `fair'' issued to capture a number of very different issues ranging from a perceived bias on the part of the decision maker to a perception of a lack of serious consideration of the evidence or lack of knowledge of the industry. In this section, I will develop the main concerns I perceive from an economist's perspective, focusing on how certain aspects of current policy may lead to economically unreasonable merger decisions.
企业和从业者往往表达他们关注并购过程中的不同问题的``''公平的条款。但很多时候这个词`发出来捕获一些非常不同的问题，从对决策者的一部分感知偏见的感知缺乏认真的考虑这个行业的知识或证据不足'公平'。在本节中，我将从一个经济学家的角度来看，着眼于现行政策的某些方面可能导致经济上不合理的并购决策。#p#分页标题#e#
Instead of a systematic review of cases as in Uneven et al.(1993),I will focus on the issues raised by some of the recent controversial cases because they highlight the issues of concern. For this discussion it is of little importance whether, in my judgment, these cases were decided correctly. What is essential is whether features of the decision making process could have led to erroneous(or random)outcomes. The purpose of this section is to identify the possible sources for such decision error. This forms the basis for a discussion in later sections about the potential changes in policy that could reduce the potential for such errors.
1.1.Does the commission use inadequate theoretical frameworks for its analysis
One difficulty in assessing the adequacy of the theoretical framework used in merger analysis is that any competition authority attempting a sound theoretical foundation for its policies has had to shoot at a moving target. This is particularly true for any attempts to include concerns about collusion(or parallel conduct)and foreclosure into preventive emerge policy. The fold of Industrial Organization has been moving fast in these are as over the last 20 years and the appropriate incorporation of these insights into policy have not really been settled. But even in more traditional areas like market definition the developments in economics have changed what we would consider an appropriate rigorous review of mergers.
1.2.Empirical substantiation of theories
From an economic point of view evidence for anti-competitive effects of mergers is nothing else but empirical data that allows one to discriminate between likely anti-competitive effects from a merger in the market and the absence of such effects. I have discussed in the previous section, that there are often significant problems in the theoretical frameworks the Commission uses and that the empirical implications that are attributed to the theory are often invalid. However, there are also, somewhat distinct, issues with the acquisition and use of evidence in itself. The observation made by Evened al.(1993,p.77)that the Commission's assessments`` rely unnecessarily on qualitative assertions and hunches even when more quantitative evidence cold have been made available, ''is still a valid concern. Furthermore, the presentation of evidence in published decisions at times appears selective and unsystematic, or strong conclusions are drawn from inconclusive evidence .In this section, I will brieˉy discuss some of these issues in the context of the three main concerns in merger assessment: unilateral effects, coordinated effects, and arguments about foreclosure and leverage. In the area of unilateral effects of mergers empirical analysis has made the greatest progress. Its power has been demonstrated impressively in the United States in Home Depot/Staples in which the argument for a narrow market of office supply super stores was rigorously supported by market data. Even where sophisticated econometric methods are not or cannot be used US antitrust authorities have systematically pushed the use of hard data to obtain reasonable assessments of the order of magnitude of claimed effects.#p#分页标题#e#
1.3.Constraints on resources
One of the most important sources of error in any administrative procedure is a lack of appropriate resources. There is no doubt that the Merger Task Force has been strained to the limits in its capacity to process mergers simply as a result of the increase in the case load. Some cases have been farmed out to other units of DG Competition. At the same time the Commission appears to be more ready to accept requests from national competition authorities to take over responsibility for a case. Constraints on the number of available man-hours constitute only one aspect of such resource constraints.
For the quality of decision making of a competition authority it is crucial what stock of expertise it has ready access to. Originally, Merger Task Force personnel have come from a background as career civil servants with a heavy emphasis on training in law. The average member on a case team will often lack intensive economic training or have little experience with looking at merger cases from an industry perspective(as consultants, competition lawyers, or investment bankers).I will discuss the problems generated by a generally limited range of experience in Section 1.4.Here I concentrate on the issue of capacity constraints on the side of highly qualified economists.
There is no doubt that the Commission has made an effort to hire economists and increase the MTF's capacity in economic expertise. Commissioner Monty has made the hiring of quailed economic staff a priority. But it is important to note two aspects of the problem. First, the necessary level of economic expertise for merger cases is fairly high.
Second, we are starting from a relatively low base.
Let merest consider the level of expertise needed. The economic questions discussed in recent merger cases, including empirical applications of market definition tests and unilateral effects analysis, are, as I have described, often at the forefront of industrial economics research. New economic theory was written for cases like GE/Honeywell or UPM developments in empirical Industrial Organization were at the basis of the empirical analysis of unilateral effects in theVolvo/Scania case. All of the cases that have led to considerable controversy have revolved around issues associated with fairly recent academic research, especially those on foreclosure and leverage This dramatic development in the degree of sophistication in the economic analysis of merger cases is reˉacted in the market place. Specialized consultancies in competition policy anti-trust today have a policy to hire economists with a Ph.D. from a leading university for any job above the level of basic research assistant positions.
2.Addressing problems through substantive changes in merger assessment 通过兼并评估实质性的改变解决问题
In its Green paper the European Commission has proposed two substantive changes to the test applied in merger control analysis. One is a switch away from the dominance concept to a standard of ``substantial lessening of competition’ as in the United States, Canada or even the United Kingdom. The second is the introduction of an explicit efficiency defense in merger proceedings.#p#分页标题#e#
Both switches tend to be appealing to economists. The ``substantial lessening of competition'' standard is phrased in a way that sounds more like the language an economist would use. The efficiency defense explicitly incorporates one side of the trade-off involved in mergers into the analysis. But the real question to ask is whether such changes have any possible real impact on merger decisions. Furthermore, it is usually not discussed whether such changes could at all address the decencies in merger policy that we have identified.
In this section, I will argue that the proposed changes are at best cosmetic. Given the context in which the change is proposed and given the nature of judicial review, there is area danger that the suggested switch could aggravate the problems in merger assessment that we have identified rather than reduce them. At the same time there are simple substantive rule changes that could eliminate many of controversies that we have seen in recent cases.
2.1.Efficiency defenses
Calling for efficiency defenses is another pet subject for economists. The essence of the subject of economics is to look for trade-offs and it makes economists nervous when policy apparently only addresses one side of the trade-off, namely the anti-competitive effects of mergers. Indeed, economists have made proposals to evaluate the welfare effects of mergers by looking directly at the tradeoff of price effects against cost changes in a welfare calculus.(See Farrell and Shapiro,1990 for a theoretical framework and Werden and Froeb,1994 and Roèller et al.,2000 for suggestions for practical implementation.)I will discuss in this section why efficiency defenses aimed ate valuating the market power/efficiency trade-off are highly problematic. I will also discuss an alternative view of efficiency arguments that are aimed not at such a trade-off. Instead I will argue that efficiency arguments can be used as evidence in establishing intent on the part of the parties. This approach is quite different from thirst in the intended use of efficiency arguments, as I will discuss below.
To understand the problem of efficiency defenses as commonly understood it is necessary to understand what kind of evidence can be used to evaluate the different effects of mergers. As we have discussed, unilateral effects can in principle be estimated from current data(assuming costs and other market circumstances)are unchanged after
the merger. Basically, pre merger variations in prices allow one to make inferences about demand elasticity’s, which in turn allow predictions about post-merger behavior. Things are much more difficult with coordinated effects of mergers and efficiencies. For these effects the post merger situation is counterfactual to the pre-merger market setting. There is no variation in pre merger data in the market that would allow one to make direct inferences about the changes in coordination or efficiencies post merger. In coordinated effects analysis we at least have a well-specie theory that can help in making qualitative predictions. But the prediction about behavior is relies much more strongly on assuming of a septic model of industry interaction than unilateral effects analysis does.#p#分页标题#e#
2.2.The potential dangers of a combined switch to SLC plus efficiency defenses
One important issue that is almost never discussed in the discussions on SLC and the efficiency defense is the way in which the two changes would interact. Given that most proponents of the SLC test appear to aim at a reduction in the critical benchmark fronting a presumption for significant anti-competitive effects of mergers we would clearly see a greater range of mergers which the Commission could prohibit without
running into problems of substantive review. As I have argued this would significantly shift the burden of proof to the merging parties. But this would also greatly increase the number of mergers that, according to the standard, the Commission should prohibit, but which it feels will not generate significant anti-competitive effects. The creation of an efficiency test would allow the Commission to get around this problem by accepting efficiency defenses. Since the order of magnitude of efficiencies cannot really be assessed with any precision, a wide range of assessments will be considered reasonable.
In effect, this would give the Commission wide discretion to use efficiency arguments to allow mergers above the intervention threshold to pass. The attractiveness of such a system with increased discretion for the Commission is obvious. Blocking mergers would require no more than the determination of market shares on the side of assessing market power and ease the blocking of mergers without detailed market analysis. The efficiency defense would allow wide discretion to permit those mergers that appear less anti-competitive to the Commission and avoid an impression of too interventionist a policy.
3.The role of procedural and administrative adjustments 程序及行政调整中的作用
We have seen that most of the issues concerning valid theoretical arguments and solid empirical evidence do not have a resolution through changing the substantive tests for merger analysis. Indeed, in the previous section I have argued that the proposed changes may in fact(depending on how they would be implemented)give the Commission even more discretion in decision-making. This would only reduce the pressure for more rigorous economic argument. I concluded, that the only substantial change in merger tests that could control some of the more problematic merger assessments would be to largely exclude the use of foreclosure and leverage arguments from merger assessment.
The only other way to improve the quality of decision-making is to create better incentives for the decision makers to correct the problems that are surfacing. These incentives are clearly not in place at the moment. Even where the Commission looses cases in court, as in the Air tours case, it tends to insist that there were never any problems in the theoretical substance of its case or in the empirical evidence provided. This is not necessarily a problem of attitude or lack of professionalism on the side of the staff at the Commission. Indeed, these arguments are clearly made by a committed, highly motivated staff that strongly believes in its mission. The problem is that there is no benefit in the system for Commission members to fundamentally question own practices. This means that the self-cornering bias existing for any individual or in any institution is not kept in check by an effective incentive system.#p#分页标题#e#
3.1.Creating checks and balances in through the merger proceedings
There is a general principle of judicial proceedings that investigative and prosecutorial functions are completely separated from the function of decision making. The Commission merger procedure has been widely criticized for putting the function of prosecutor and judge into the same hand: those who investigate and write a statement of objections also write the decision.
Why is such a procedure so problematic? In penal law this is easy to see for everybody. Police and prosecutors can get so worked up about a case and can at some point convince themselves so thoroughly of the guilt of their favorite suspect that they loose out of sight that they have not really collected sufficient evidence remedy discussions forward would further incentives a development in this direction,
which could ultimately benefit merging? Indeed, often the suspect is innocent. The separation of powers means that a different person, say the judge, takes close look at the case and all of the evidence and makes a decision on the basis of all the information provided by the prosecution and the defense. This has two big advantages:
First, the prosecutor has, by the time he has put his case together, a stake in the outcome. He wants to be right and will naturally overlookˉawes in his argument. Second, these parathion of powers creates a very nice incentive system for the prosecutor who has to generate the information. The measure of the prosecutor's success now is given by his performance in court. A prosecutor who brings too many cases that he does not win will loose in reputation and will be assigned less important cases. He is revealed not to be as good as his colleagues. The result will be a much greater effort in obtaining persuasive evidence and making a watertight case. Note that these incentives are created without themed of any formal incentive systems. The institutional structure naturally creates a situation in which a prosecutor will obviously be good when he prevails in front of a judge. All the non-monetary benefits in a job like prestige and respect are connected to such outcomes.
It is immediately apparent that exactly the same incentive problem is present in merger cases. The case teams investigating the merger and, possibly, putting a case against the merger together have exactly the same function as prosecutors. But currently we have a procedure in which these same people write the decision.
Currently we neither have the nice feature that someone with a fresh mind takes another thorough look at the case, nor do we Ben remedy discussions forward would further incentives a development in this direction,
which could ultimately benefit merging firms. from the incentive properties that a separation of investigation and decision-making has. This is different in other jurisdictions. In the United States the decision has to be made in a court, so that the relevant competition authority only has the investigative and prosecutorial function. Everything else is passed on to the court system.#p#分页标题#e#
In Canada, even investigation and prosecution are formally separated and the decision mistaken by the Competition Tribunal.
3.2.Creating internal checks and balances
A further way of creating checks and balances is to create organizational structures that systematically use the non-monetary incentives motivating the individuals and promote the presence of a variety of perspectives on a problem. This would affect both the internal organization of the Merger Task Force in the way that it uses existing resources as well as strategies that help in the recruitment of appropriate staff. In this section, I will
discuss improvements in the use of economic expertise win the Commission to serve these goals. I will then consider how some of these insights can be extended beyond the issue of economic expertise.
Currently, economists work in the MTF along side other staff as case team members in merger cases. This makes all of them fully integrated into the general staff and very much aligns their perspectives with those of the case team. This type of organization has several disadvantages and inefficiencies attached to it. First, it means that it is difficult to attract economists of a high enough qualification to join the MTF. Joining the MTF means to give up any academic ambitions for the future and also any hope to work with a strong team of other economists. This is different for economists joining consultancies. There they mostly work with other economists with strong academic training and also have greater exposure to working with academic economists in septic cases. A strong empirically oriented Ph.D. a consultancy environment highly rewarding but will not feel that his or her skills are really used in the environment of the MTF. For highly qualified Ph.D.'s a consulting job will therefore be of much greater interest than apposition in the MTF. More research-oriented academics can typically not be attracted.
Second, there is the problem that careful economic reasoning is a skill that is only maintained by regular practice. For many economic questions plausibility arguments are very seductive. They may seem very convincing attars sight and sound very clever to non-specialists. Unfortunately, more often than not, plausibility arguments turn outdo be incorrect when subjected to careful analysis. In a world in which there is a premium in providing a` `convincing argument’ but not much of an incentive to`` get it right,’ good economic reasoning easily falls by the wayside. Many excellent economists are known to have succumbed to the attractiveness of erroneous plausibility arguments that favored their clients whenever they felt that they were not under much scrutiny by their peers. In contrast, good economic analysis thrives in environments in which economists are encouraged to constantly challenge their own arguments because they have to prove themselves with respect to their peers. In other words, an organization should try to organize itself in such a way that professional pride can bemused for incentive purposes.#p#分页标题#e#
3.3.Checks and balances through outside review: Transparency
One of the features considered a strength of the European merger regulation by the Commission are its formal measures for transparency of Commission decisions. All decisions have to be published. Indeed, many economists would assess competition policy and the quality of the Commission arguments to a large degree from the published record.
Unfortunately, real transparency is by far not as high as suggested by such rules of publication. Someone who has participated in specific cases will quickly recognize that the published decision often differs greatly from what the parties thought the relevant issues in the case were. The case is always presented from the point of view of the Commission, and whole strands of argument the Commission did not consider relevant
are often simply suppressed. Furthermore, almost any decision is selective in the evidence it presents by suppressing data brought forward against the Commission's argument. But the greatest problems with transparency do not arise from the subjective decisions of the Commission about presenting the case. A crucial problem for effective transparency is
that most of the evidence never makes it into the public documentation of the case, because it will be purged from the public version of the documents due to confidentiality considerations. In most cases, the critical issues revolve around the determination of market boundaries and market shares and the assessment of evidence as to the degree of market power achieved through the merger. The relevant data necessary to evaluate thesis basically never a part of the public record. Hence, there exists no transparency such that outside observers could criticize the Commission in a rigorous way.
What this implies is that it is feasible for outside observers like academic economists to assess the theories on which the Commission decision is based. This is guaranteed by the publication of the Commission decision. However, it is impossible to assess from the outside whether the empirical methods used are valid and whether the evidence obtained by the Commission justices the conclusions in the decision.
4.Improving the remedies procedure 完善救济程序
Since the determination of appropriate remedies in merger cases is thorny one, it is unsurprising that a good procedure for discussing remedies is no less problematic. There’s little doubt that current practice is unsatisfactory both for the Commission and the merging parties. At the same time it is a very hard task tend a good way of dealing with remedies in the time frame allowed by the merger regulation. One of the greatest problems arises from the deadlines of the process. If the parties fully use the time allotted for developing suggestions for remedies, time becomes extremely short for the Commission to assess the adequacy of the suggested remedies. As I have discussed there are strong incentives for the parties not to propose remedies until late in the process because of their signaling function. In the Green Paper the Commission states that parties tend tend it problematic to defend themselves against the case and discuss remedies at the same time. This seems another way of stating the same problems. Parties understandably do not want decisions to be colored by remedy discussions.#p#分页标题#e#
The result is that in cases with serious and complicated remedy discussions the clock tends to run out on the Commission and the parties. With insufficient time to assess the remedies the Commission will have a tendency to block mergers that might have gone through with a more elaborate remedy discussion. The Commission has proposed a` `stop-the-clock'' provision. According to this provision the merging parties could stop the clock at the end of phase 2 in order to prolong the time for a satisfactory remedy discussion.(A similar procedure is considered for phase 1).The stop the clock provision in itself will not change the initial problem that there are strong incentives to delay the remedy discussion. Indeed, some of the delay problems should be expected to be increased. First, making proposals only close or at the deadline becomes more attractive because it is more likely that such a proposal will go through. Given that there is typically ongoing negotiation about the remedies(see the newspaper reports on GE/Honeywell cited above)such a stop-the-clock proposal would also give more incentives to the merging parties to make more incremental offers.