This document is available in two formats: this web page (for browsing content) and PDF (comparable to original document formatting). To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.

U.S. Department of Justice Seal   DEPARTMENT OF JUSTICE


Address by

William J. Kolasky
Deputy Assistant Attorney General
Antitrust Division
U.S. Department of Justice

Before the
George Mason University Symposium
Washington, DC

November 9, 2001

In the 1960s the United States experienced a wave of conglomerate mergers, driven in part by overly restrictive antitrust policies toward horizontal and vertical mergers. In response, the U.S. antitrust agencies and courts developed a number of theories of competitive harm with colorful names like deep pockets, reciprocal dealing, and entrenchment. As the Chicago School taught us the central importance of consumer welfare and efficiency in antitrust analysis,1 these theories faded away.2 This process was speeded by the failure of most large conglomerates to deliver shareholder value - in market after market, conglomerates underperformed their rivals, contrary to the fears of some theorists. After fifteen years of painful experience with these now long-abandoned theories, the U.S. antitrust agencies concluded that antitrust should rarely, if ever, interfere with any conglomerate merger.3 We simply could not identify any conditions under which a conglomerate merger, unlike a horizontal or vertical merger, would likely give the merged firm the ability and incentive to raise price and restrict output. We recognized, conversely, that conglomerate mergers have the potential as a class to generate significant efficiencies. These potential benefits include providing infusions of capital, improving management efficiency either through replacement of mediocre executives or reinforcement ofgood ones with superior financial control and management information systems, transfer of technical and marketing know-how and best practices across traditional industry lines; meshing of research and distribution; increasing ability to ride out economic fluctuations through diversification; and providing owners-managers a market for selling the enterprises they created, thus encouraging entrepreneurship and risk-taking.4

While I would not want to overstate the role antitrust policy played, the shift in antitrust policy that took place in the 1970s and 1980s (including the shift with respect to conglomerate mergers) was one of the factors (along with many other important changes in economic policy) that contributed to one of the most vibrant periods in our economic history. A more realistic antitrust policy toward mergers helped foster an active market for corporate control, which in turn forced managers to pay greater attention to delivering shareholder value. The result has been an enormous gain in efficiency and productivity as managerial slack has been eliminated.

It is particularly instructive to contrast our experience with that of Japan over the last decade. As an article in this month's Foreign Affairs reports, the Japanese economy is overburdened with a large number of small, inefficient firms that have been kept alive by loans from Japanese banks. The article concludes that by propping up these "zombie" companies Japan "has delayed the sort of structural adjustment that the country needs to eliminate its surplus capacity and allow the strongest firms to grow."5

Surprisingly, as we enter the 21st Century, we find ourselves replaying these old debates on a more global stage. The European Union ("EU") is raising increased concern about conglomerate mergers. This development became front page news this past summer with the EU's prohibition of the $42 billion GE/Honeywell merger - which had been billed as the largest industrial merger in history. That action triggered a firestorm of criticism, not just from the U.S. antitrust agencies and senior administration officials,6 but also from the business community generally and from leading economists,7 antitrust legal scholars,8 and editorial writers.9

To understand the reasons for this sharp divergence between the U.S. and EU, it is useful to compare the reasons we challenge horizontal and vertical mergers with the EU's reasons for its recent challenges of some conglomerate mergers. We challenge horizontal mergers because they eliminate a competitor and may thereby enable the merged firm to restrict output and raiseprice.10 Similarly, we challenge vertical mergers that eliminate a key supplier or customer where doing so may give the merged firm the ability and incentive to raise its rivals' costs, again allowing them to restrict output and raise price.11 By contrast, in its conglomerate merger decisions, the EU prohibits mergers because they will make the merged firm a stronger competitor that may ultimately be able to drive rivals from the market.12 Under this scenario the immediate effect of the merger is to reduce prices and increase output, with the anti-competitive effect occurring only if the other competitors cannot match the merged firm's offerings and exit the market.

As Charles James has said in a series of speeches,13 we view the EU approach to conglomerate mergers as inconsistent with the central tenet of U.S. antitrust policy -- that the antitrust laws "protect competition, not competitors."14 As the Supreme Court explained in Spectrum Sports, Inc. v. McQuillan:

The purpose of the [Sherman] Act is not to protect business from the working ofthe market; it is to protect the public from failure of the market. The law directs itself not against conduct which is competitive, even severely so, but against conduct which unfairly tends to destroy competition itself. . . . 15

Applying this principle to mergers, it is well established under U.S. law that the antitrust laws do not protect competitors from mergers that will make the merged firm more efficient, even if they fear they may as a result be forced from the market.16 This is because, as former Treasury Secretary Larry Summers reminded us at this year's ABA Antitrust Section Spring Meeting, competition is a means to an end, and not an end in itself: "The goal is efficiency, not competition. The ultimate goal is that there be efficiency."17 Production and transactional efficiencies benefit consumers by lowering the costs of goods and services or by increasing their value. Allocative efficiencies benefit consumers by moving the allocation of scarce resources toward the Pareto optimum; in lay terms, toward a situation where no rearrangement of those assets would enhance welfare.18 We value competition, not as an end in itself, but because itpromotes both types of efficiency.

Recognizing that efficiency is the ultimate goal should make us very cautious about adopting a merger policy that sacrifices short-term efficiencies in the name of maintaining competition. At a minimum, before applying such a policy, we should make certain we have a high degree of confidence that the trade-off we are making will ultimately benefit consumers. This would require quantifying the efficiencies and determining the likely duration of the competitive round that will occur before less efficient rivals are forced from the market. It would also require a high degree of confidence that the rivals will in fact be forced from the market - that they will not be able to develop counter-strategies that will enable them to become more efficient themselves in order to survive. Indeed, any business strategy that did not take into account competitive counter-strategies would fail the test for the Nash equilibrium -- a strategy that yields the best outcome assuming every other player plays her best strategy -- an insight for which John Nash won the Nobel Prize.19 It would also require us to estimate the size of the price increases likely to occur once the merged firm gains market power to determine whether, taking into account the efficiencies, future prices to consumers are likely to be higher or lower than they would be in a market populated by a several less efficient firms. Finally, we would have to determine the likely duration of the monopoly period - which would be dependent on entry conditions at the time the monopoly is finally achieved.

In the United States, we have very little confidence in our ability to make these judgments which would necessarily involve predictions far out into the future. We believe, in the immortal words of my favorite golfer, Tin Cup McAvoy, that we need to "be humble."20 We have more confidence in the self-correcting nature of markets.21 This confidence is especially strong when the markets are populated by strong rivals and strong buyers, who will usually find ways to protect themselves from an aspiring monopolist. Our strong belief in markets and our humility in our own predictive abilities lead us to be skeptical of claims by rivals that a merger will lead to their ultimate demise and to demand strong empirical proof before we will accept such claims.22


To illustrate these points, let me turn to GE/Honeywell.23 The Department has previously provided an explanation for our decision not to challenge that merger, except with respect to a couple of horizontal overlaps.24 Now that the EU's decision is public, we can examine its reasoning in order, as my good friend Bill Kovacic put it, to deconstruct the reasons for our divergent outcomes.

Before I turn to the substance, let me say a word about process. There have been some unfortunate suggestions in the newspapers that the Department turned a deaf ear to complainants and did not investigate the merger as throughly and seriously as the EU did. Those suggestions are wholly unfair to our very dedicated staff. We conducted a very thorough investigation, interviewing over 75 witnesses and reviewing hundreds of boxes of documents over a five-month period. We had a team with a substantial experience in the very markets under review: the same team that reviewed the AlliedSignal/Honeywell merger just one year earlier. Several of our witness interviews were done jointly with the EU team investigating the transaction. We met with every complainant who asked to meet with us, in some instances multiple times. We hiredan outside consulting economist who was generally hawkish on vertical foreclosure theories to assist us in evaluating the economic papers submitted by the complainants. So our difference of view did not stem from any lack of diligence or from any prejudgment of the issues.

Let me begin by briefly reviewing the facts. GE is a leading producer of jet engines for large commercial aircraft and large regional jets. Its two principal rivals are Pratt & Whitney (a subsidiary of United Technologies) and Rolls Royce. Honeywell is a leading producer of engines for small regional and corporate jets and of avionics and nonavionics systems, such as landing gear and auxiliary power units (APUs). It faces different rivals in each of its lines of business. For small jet engines, Honeywell's principal rivals are Pratt &Whitney and Rolls Royce. For avionics, its principal rivals are Rockwell Collins and Thales. For non-avionics, the principal rivals include United Technology (for APU's) and BF Goodrich and SNECMA for landing gear.

The EU had two theories of competitive harm. First, it found that the merger would strengthen GE's already dominant position in the market for large jet engines. Second, it found that the merger would enable Honeywell to gain a dominant position in the small engine, avionics and non-avionics markets in which it competes. It is worth examining each of these separately.

Large Jet Engines

The Department's disagreement with the EU begins with its finding that GE already has a dominant position in the market for large jet engines. We understand the EU's concept of dominance to be more or less synonymous with our concept of market power. Dominance is defined in EU case law as "a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market by affording it the power to behave independently of its competitors, its customers and ultimately of the consumers."25 Similarly, U.S. law defines market power law as the ability to raise price and exclude competitors.26 Both definitions are somewhat inexact from the perspective of a professional economist -- not even a monopolist can behave wholly independently of its rivals and customers -- but both seem designed to describe verbally the same basic economic position -- i.e., monopoly.

The EU's finding of dominance rested almost entirely on GE's large (65%) and growing share of outstanding orders for engines aircraft still in production, attributing to GE 100% of engine sales by GE's joint venture with SNECMA, CFM. We, by contrast, found these market shares only weakly indicative of competitive conditions in the market. GE's large share was almost entirely dependent on a single sole-source contract with Boeing for the 737, the most successful commercial aircraft in history. Excluding those sales would produce much more balanced market shares (even attributing 100% of CFM's remaining sales to GE): GE 44%, PW23%, Rolls Royce 27%.

Those shares comported much more closely to the actual competitive dynamics of the market, where the customers described the competition as "fierce," resulting (as even the EU acknowledged)27 in deeply discounted engine prices. Nor did we find any evidence supporting the EU's finding that Rolls Royce and Pratt & Whitney were no longer in a position to constrain GE's behavior. To the contrary, both companies had growing revenues and profits and were investing heavily in the development of their next generation engines. As of January 2001, both had actually outperformed GE in engine awards on the three major airframes offering a choice of engines. Rolls Royce, in particular, said publicly that it expected its installed base to practically double over the next five years as a result of market share gains.28 New engine awards to date this year seem to confirm our view that GE is nowhere near dominant: year-to-date GE has won 42% of contract awards, PW 32%, and Rolls Royce 27%.29

Obviously, this difference in view as to GE's position relative to its engine competitors colors the entire analysis of the merger. If, as we found, the engine market remains highly competitive, with GE losing rather than gaining ground, then what the EU calls "GE's toolkit for dominance"30 -- its immense size and financial resources and its aircraft leasing arm, GECAS --become much less threatening.

This factual issue aside, there is another, more fundamental problem with the EU's conclusion that the merger would strengthen GE's (non-existent) dominant position in engines. That finding necessarily rests entirely on "range effects" -- in this case, the theory that GE and Honeywell would engage in "mixed bundling" by offering a package of GE engines and Honeywell avionics and nonavionics systems at discounted prices because of the so-called Cournot effect.31 Honeywell brings nothing else to the party.

This theory was originally based largely on a model developed by Jay Pil Choi for one of the complainants, based on previous work by Barry Nalebuff. The parties retained Nalebuff to evaluate Choi's model. His conclusions, which he has now incorporated in an article, were that the model simply did not fit the competitive realities of aerospace markets, with highly differentiated products, powerful buyers, and individually negotiated transactions.32 Nalebuff further found that "[w]hen a model is built that takes into account the nature of competition in this industry, we do not find that product bundling creates an advantage."33

In its final decision, the EU eschews further reliance on Choi's model, but nevertheless inexplicably concludes (with no other support) that "bundling would lead to a re-allocation andtherefore to a shift of market share in favour of the merged entity"34 to such an extent that over the longer term GE's competitors would be unable to cover their fixed costs and would exit the market.35 Blocking a $42 billion merger on this basis, with neither theoretical or empirical support, is difficult to understand, to say the least.

In subsequent public statements, the EU has tried to minimize the role that the mixed bundling theory played in its decision. EU representatives have said that the Commission found the Cournot effect to be quite small36 and that GE/Honeywell was not primarily a portfolio effects case, but was based instead on the transfer of GE's financial strength to Honeywell.37 If that is so, it is hard to understand the finding that the merger would strengthen GE's -- not Honeywell's -- dominant position in large jet engines. GE already owns GE Capital and GECAS; Honeywell, a financially weak company with a falling stock price, adds nothing, except perhaps a drain on GE's resources.

Avionics and Non-Avionics

Let's now turn to the second part of the EU's case -- the argument that the merger would enable Honeywell to create a dominant position in the small corporate jet engine, avionics and nonavionics markets in which it competes. This is where the EU argues that GE's "toolkit for dominance" -- GE Capital and GECAS -- would come into play and tip the competitive balance decisively in Honeywell's favor. Neither claim can withstand scrutiny.

GE Capital

According to the EU, the first key factor in GE's toolkit is GE Capital.38 Noting that GE has world's largest market capitalization, the Commission argues that GE Capital offers GE businesses enormous financial means, enabling it to take more risk in product development than its rivals and to offer customers heavy discounts on the initial sale of engines, recouping those discount through sales of spares and repairs. This strategy, according to the EU, moves the break-even point of an engine project further into the future, forcing rivals to rely on external financing at a higher cost of capital than GE, with its triple AAA bond rating, enjoys. The merger extends this competitive advantage to Honeywell.

This is the first time I can remember seeing a triple AAA bond rating treated as an antitrust no-no, so let's "deconstruct" the Commission's reasoning.39 First, the size of GE'smarket capitalization is obviously competitively irrelevant. We have all seen how ephemeral market caps are - two years ago, at the height of the dot com bubble, Microsoft and Cisco both had market caps well in excess of GE's. More recently, we have seen even a company as large as Enron have its market capitalization shrink by 80% in a matter of months.

Second, more generally, an increase in aggregate firm resources does not necessarily mean that any one division of the enterprise will obtain capital more readily or more cheaply than its rivals. Capital markets generally work very efficiently and there is no obvious reason, absent some clearly defined market imperfection, why GE's cost of capital for a particular project should be any lower than that of its rivals. As a large diversified company, GE has many other uses for its capital and committing capital to one project entails opportunity costs because that capital is no longer available for other, perhaps more worthwhile projects. Taking these opportunity costs into account, GE's cost of capital with respect to any particular project should be equal to that of its competitors. This being the case, it is not surprising that, empirically, what we found when we examined the markets in which GE already competes was that GE's engine rivals are both investing just as heavily as GE in developing their next generation engines and have had no difficulty raising capital to finance that effort.

Third, to the extent GE does have access to cheaper capital, that is a source of efficiency just like any other valuable asset, be it machinery, knowledge, or management skills.40 Cheap capital serves to lower prices and promote innovation. Blocking a merger because of GE's lowercapital costs is therefore objectionable for the same reason as blocking a merger because of any other efficiency. Even if the number of sellers does decline, any tendency to higher prices would probably be at least partially offset by lowering of costs. Generally, the price level established by a small number of efficient firms will be lower than the price level in a market supplied by a large number of inefficient firms.41

Fourth, the EU's argument that because of the lower prices the merged firm will be able to offer Honeywell's rivals will no longer be able to cover the fixed costs required for the development of new products takes us all the way back to the concern over ruinous competition found in our very earliest cartel cases in the railroad industry at the turn of the last century.42 In the 1890s, economists had difficulty explaining how a competitive enterprise could ever recover its fixed costs. They feared that firms with significant fixed costs would be driven to "ruinous competition" resulting in bankruptcy and harmful destruction of assets. These arguments were quickly appropriated by defendants in the earliest Sherman Act railroad cartel cases, who argued that unregulated railroads would face ruinous competition unless allowed to fix their rates. Even Justice Peckham saw through that argument, leading the Supreme Court to hold that a "ruinous competition defense would force the court to decide what a reasonable rate of profit in a particular industry should be," and that the courts were not up to that task.43 How ironic that we should be hearing the same arguments on the other side of the Atlantic a hundred years later.

Fifth and finally, the EU's GE Capital theory is troubling because it lacks any limiting principle. Under this new variant of the old "deep pockets" theory, Honeywell couldn't merge with Citigroup or Microsoft and GE could not acquire any company with a leading position in any capital intensive business with high entry barriers. Surely that is not what the EU intended, but that is the logical extension of its decision.


The second key factor contributing to GE's dominance in engines, according to the EU, was its vertical integration into aircraft purchasing, financing and leasing through GECAS, which EU describes as the largest purchaser of aircraft in the world. With its GE-only procurement policy, the EU argued, GECAS has been able to influence the selection of engines by serving as a launch customer and by causing airlines to standardize fleets around GE-powered aircraft. The merger would enable GECAS to extend this influence to the markets in which Honeywell competes.

The problem with this story is that the facts didn't support it. GECAS's share of aircraft purchases is less than 10%, substantially less than what U.S. antitrust courts typically require to support a finding of potential foreclosure.44 Given its small share of aircraft purchases, we found no evidence that GECAS's policy of purchasing only aircraft equipped with GE engines had or could foreclose rivals from the market. Given GECAS's small share, unless its policy couldsomehow change equipment preferences of ultimate customers, the likely Nash equilibrium response by rival leasing companies should be to purchase proportionately more non-GE engines, in part to differentiate themselves from GE.45 We found several examples of rival leasing companies doing exactly that.

Our investigation also found that GECAS was not a significant launch customer and that the claims GECAS could "seed" airlines with GE engines because of the importance of commonality were seriously overstated; in fact, 90% of the world's aircraft are in mixed fleets. To the extent seeding was even a possibility, we could see no reason why rival engine manufacturers could not do the same simply by offering discounts off their engines. It makes no difference analytically whether GE is able to sell engines based on (1) a low lease rate from GECAS or (2) a low engine price. A low lease rate is simply another form of discount.46 To the extent airlines prefer discounts in this form, rival equipment suppliers could partner with leasing companies to offer favorable lease term for aircraft equipped with their engines. Once again the EU's theory fails the Nash equilibrium test.

We also examined the argument that GECAS could cause "share shifting" by inducing airframe manufacturers to sole source engines from GE. We examined each of the four transactions in which GECAS allegedly used its buying power to get GE engines sole source on new aircraft platforms. In each, we found that GECAS played no role in the customer's decisionto sole source and that in each case GE had won the competition on the merits, by offering the best engines for the customer needs at the best price.

Rival exit.

All the EU's theories of competitive harm are crucially dependent on its prediction that rivals would be forced to exit in the face of a strengthened Honeywell. These claims were, in our view, simply not credible. Honeywell's competitors include large, financially healthy companies like United Technologies, BF Goodrich, and Thales, each with important competitive advantages of its own. These companies are all more than holding their own currently. In addition, these aerospace markets are characterized by powerful buyers, Boeing and Airbus being the two best examples, with a strong incentive to maintain competition in the supply of avionics and nonavionics systems. These strong rivals, either independently or with the support of these powerful buyers, have a wide range of counter-strategies available, including mergers and teaming arrangements among themselves. The EU's dismissal of teaming arrangements as an effective counter-strategy is particularly ironic, given that GE's allegedly dominant position in engines results from just such a teaming arrangement with SNECMA and that the Commission found that teaming arrangements were an effective counter-strategy in this industry just one year earlier in its decision approving the Allied Signal /Honeywell merger.47 Indeed, immediately after the announcement of the merger, Rockwell announced that it was spinning off its avionics division, Collins, prompting speculation that it was positioning Collins to be a merger partner for another aerospace company. Again, the EU's theory fails the Nash equilibrium test.

A false efficiency debate?

In defending their decision, the EU has denied that it condemned the GE/Honeywell merger because of the efficiencies it would create, arguing that the parties in fact claimed no merger-specific cost savings of the type that might be a defense to a horizontal merger.48 In making this argument, the EU maintains that the price cuts that would flow from mixed bundling were not "real" efficiencies, but were simply "strategic pricing" of a kind that would not result in sustainable price reductions and that the parties did not claim any real efficiencies. There are several problems with this argument.

First, there is no sound economic basis for distinguishing between productive and allocative efficiencies. To the extent a merger of complements gives the merged firm the incentive to lower prices because it causes the firm to internalize the negative externalities associated with higher prices (the so-called Cournot effect), it moves prices in the right direction - toward marginal costs - enhancing allocative efficiency through the elimination of double marginalization and benefitting consumers with lower prices and increased output.49 Such price reductions are every bit as beneficial and sustainable as those resulting from cost savings. Applying pejorative labels such as "strategic" or "cross subsidization" to these price reductionsonly clouds the analysis. Strategic pricing and cross-subsidization across business lines do not violate the antitrust laws or cause consumer harm unless the resulting prices are below cost and therefore predatory.50

Second, to the extent the EU's decision to prohibit the merger rested on concern that Honeywell would gain a competitive advantage by having access to cheaper capital from GE, enabling it to invest more than its rivals in developing and discounting its products, that itself is an efficiency that would have benefitted consumers. Extending access to cheap capital to a broader range of business activities is just as much an efficiency as extending access to any other valuable asset that is currently under-utilized.51

Third, contrary to the EU's assertions, the parties did in fact claim the type of efficiencies most common to conglomerate mergers - namely, that the acquiring company would be able to manage the acquired business more efficiently than its incumbent managers were capable of doing. As Jack Welsh explains in its autobiography, Jack: Straight from the Gut: "I felt we could do so much more with Honeywell's assets by doing what we've done with GE: pushing more aggressively into services, and adding Six Sigma and e-business initiatives to Honeywell's operations. We figured on $1.5 billion in savings from these initiatives and other productivity measures."52 The EU offers no explanation to support its argument that these cost savings werenot merger-specific and would not have been passed on to consumers; indeed, its argument seems to contradict the central premise of the EU's decision -- namely, that the merger would enable Honeywell to lower prices to a point where its rivals would be driven from the market.

Why do we care?

That's enough about GE/Honeywell. That merger is history. Our reason for discussing it is based on concern for the future, not the past. From that perspective, the type of divergence between the two largest antitrust jurisdictions in the world we experienced in this case is very troubling for at least three reasons.

First, in cases involving mergers in global markets, there are serious externalities associated with one jurisdiction blocking a merger on the basis of theories that other jurisdictions believe risk sacrificing important efficiencies to prevent speculative future harm to competition. By so doing, that jurisdiction denies consumers around the world the benefits the merger might have delivered.

Second, divergent substantive standards between the U.S. and Europe are almost certain to increase the transactions costs associated with the merger clearance process. The result may well be to deter mergers that would have been pro-competitive and efficiency-enhancing.

Third, such a sharp divergence undermines the strong political consensus supporting vigorous antitrust enforcement, something none of us wants.

We recognize that the EU is entitled to make and interpret its own laws. We also recognize that we and the EU will not always agree and that our way is not always best. We have no power to change EU law, other than by persuasion. For this reason, we believe it is important that we discuss this issue in depth, both in private and in public. We also encourage the business community, the private bar, and the academy to participate in this debate.

As we discuss our divergent views, it is important to examine some possible explanations for our differences with respect to conglomerate mergers, in order to identify institutional changes that might promote greater convergence. The EU, to date, has suggested three possible explanations.

First, different legal standards. The EU Merger Regulation prohibits mergers that may create or strengthen a dominant position, whereas U.S. law prohibits mergers that may substantially lessen competition. Some have suggested that these different legal standards account for our differences in approach and have urged the EU to consider moving to a substantially lessening of competition standard.

One problem with this explanation is that the language of the EU Merger Regulation already incorporates a substantial lessening of competition requirement into the dominance test: it prohibits only those mergers that create or strengthen a dominant position "as a result of which effective competition would be significantly impeded in the common market or in a substantial part of it."53 On our side, we have defined a "substantial lessening of competition" in our Merger Guidelines as "creating or enhancing market power,"54 which sounds a lot like "creating or strengthening a dominant position." In addition, to the extent the dominance standard is believed to give less weight to efficiencies, the language of the Merger Regulation says otherwise. The Regulation expressly includes an efficiency defense, requiring the Commission to consider the effect of the merger on "the development of technical and economic progress provided that it is to consumers' advantage and does not form an obstacle to competition."55 This is not dissimilar to how efficiencies are treated under U.S. law. Nevertheless, there may be benefits to shifting from a "dominance" test to a "substantial lessening of competition" test if that serves to clear out the underbrush of old case law that has grown up over the years.

Second, less availability of ex post remedies. A second explanation offered by the EU is that merger review is their only shot because ex post remedies are less effective under EU law, inthe absence of treble damage exposure, than in the United States.56 Again, this explanation is not fully persuasive. The EU's conglomerate merger decisions rest primarily on concerns about types of bundling and discounting that would not necessarily be unlawful under the antitrust laws, so the availability of ex post remedies for anticompetitive conduct is largely irrelevant. To the extent the EU's conglomerate merger decisions rests on vague fears that the merged firm might engage in predatory pricing or unlawful tying, such conduct is best addressed when and if it occurs, rather than ex ante at the time of a merger when we have none of the facts we would need to determine whether the conduct, if it occurs, would, in fact, anticompetitive.

Third, short-term vs. long-term perspective. The third explanation offered by the EU is that they take a longer-term perspective than the U.S. Again, not surprisingly, we would disagree. The U.S. antitrust agencies are very much concerned about the long-term effects of mergers. That's one reason we will not challenge a merger unless we think its anticompetitive effects are likely to be durable, lasting at least two years. It's also why we challenge vertical mergers where the theory of anticompetitive injury may take several years to play out. The difference is not that we are short-sighted, but that we are simply more humble about our ability to predict the future.

In the interest of finding constructive solutions to this issue before it grows into a more serious problem, I would offer three alternative explanations. The first two lend themselves to concrete fixes that I believe we can and should take promptly. The third suggests the problem is more fundamental, in which case the solution is less obvious.

The first possible explanation may be institutional differences that are readily remediable. In the United States, we have a much larger professional staff, including more than 50 professionally trained Ph.D economists who are integrated into our case teams. We also have many more investigative tools - including the ability to compel production of documents and witnesses by both parties and nonparties. The EU's pending modernization proposals would improve its access to similar evidence and we understand the EU is seeking to hire more professional economists for the Merger Task Force. There may also be process and timing differences that contribute to different outcomes. We and the EU have agreed to discuss these differences and share best practices in our joint US/EU working group.

A second possible explanation is the difference in our judicial review mechanisms. If we decide in the U.S. to challenge a merger, we know we may have to go to court to convince a federal judge, by the proponderance of the evidence after an evidentiary hearing, that the merger may substantially lessen competition. This means that we know our witnesses will be exposed to the crucible of cross-examination before an independent fact-finder. By contrast, in Europe, the Commission is sometimes said to act as investigator, prosecutor, judge, and jury. Judicial review is slow and highly deferential to the Commission's factual determinations. After just six weeks at the agency, I cannot overstate how much knowing we may have to prove our case to an independent fact-finder disciplines our decisionmaking. Again, this is an institutional difference that we will be discussing in our joint US/EU Working Group.

A third, decidedly less optimistic, possible explanation is that we may have a fundamentally different view about the comparative ability of markets vs. government regulatorsto get it right. In the United States, we have much greater faith in markets than we do in regulators. Some commentators have suggested that by contrast the European Union comes from a more statist tradition that places greater confidence in the utility of governmental intervention in markets.57 Europeans may also simply be uncomfortable with our emphasis on efficiency and our unwillingness to cut competitors any slack. There is some support for this view, unfortunately, in the EU's recent criticisms of our approach as being too Darwinian.58 If this view is right, we may only be at the beginning of our disagreements. I hope it is not and that working together, with good will, we can repair the breach before it widens.

Thank you.


1 See Robert H. Bork, The Antitrust Paradox 89, 91 (1978) ("Whether one looks at the texts of the antitrust statutes, the legislative intent behind them, or the requirements of proper judicial behavior, . . . the case is overwhelming for judicial adherence to the single goal of consumer welfare in the interpretation of the antitrust laws.") ("The whole task of antitrust can be summed up as the effort to improve allocative efficiency without impairing productive efficiency so greatly as the produce either no gain or a net loss in consumer welfare.").

2 See generally United States Department of Justice Note for Discussion at Roundtable on Portfolio Effects in Conglomerate Mergers (Oct. 15, 2001) (on file with OECD and available upon request from Antitrust Division of United States Department of Justice).

3 The one exception is a merger of potential competitors which under the early cases was treated as a conglomerate merger. See, e.g., FTC v. Procter & Gamble Co., 386 U.S. 568 (1967). Today, however, we would classify a merger of potential competitors as a horizontal merger and would analyze it under our horizontal merger guidelines. See, e.g., United States v. SBC Communications, No. 1:99 CV00715 (D.D.C. filed Mar. 23, 1999 and settled by consent decree).

4 See Bork, supra note 1, at 248-49.

5 Leonard J. Schoppa, Japan, the Reluctant Reformer, Foreign Aff. Sept./Oct. 2001, at 76, 79.

6 John R. Wilke, U.S. Antitrust Chief Criticizes EU Decision to Reject Merger of GE and Honeywell, Wall St. J., July 5, 2001, at A3 (quoting Assistant Attorney General Charles James: "Clear and longstanding U.S. antitrust policy holds that the antitrust laws protect competition, not competitors . . . [The EU decision] reflects a significant point of divergence.")(also quoting somewhat more colorful language by Secretary of the Treasury Paul O'Neill).

7 See, e.g., Hal R. Varian, Economic Scene; In Europe, GE and Honeywell ran afoul of 19th-century thinking.," N.Y. Times, June 28, 2001 ("When evaluating a merger, United States antitrust officials tend to focus on the benefits to consumers, while European regulators give substantial weight to the impact on competitors, especially if they are `national champions'").

8 See, e.g., George L. Priest, The GE/Honeywell Precedent and Franco Romani, Wall St. J., June 20, 2001, at A1.

9 See, e.g., Editorial, Europe to GE: Go Home, Wall St. J., June 15, 2001, at A14 ("In the Honeywell case, novel antitrust theories have been dreamed up simply because it would be unthinkable to let a large U.S. company go about its business unmolested"); Editorial, Obstructionist Europe, Wash. Post, June 22, 2001, at A24 ("[F]or the first time ever, the Europeans seem poised to block a merger that has been approved by U.S. regulators; and . . . they are doing so on the basis of speculative reasoning about bundling that is controversial, to put it kindly.").

10 See, e.g., United States v. Franklin Elec. Co., 130 F. Supp.2d 1025 (W.D.Wis. 2000); United States v. Sunguard Data Systems, Inc., No. 012196 (D.D.C. filed Oct. 22, 2001).

11 See, e.g., Final Judgment filed in United States v. Premdor Inc., No. 1:01CV01696 (D.D.C. 2001). See generally Michael H. Riordan & Steven C. Salop, Evaluating Vertical Mergers: A Post-Chicago Approach, 63 Antitrust L.J. 513 (1995).

12 See, e.g., Boeing/McDonnell-Douglas, 1997 O.J. (L336) 16-47 (Case No. IV/M, July 30, 1997); Guinness/Grand Metropolitan, 1998 O.J. (L228) 24-54 (Case No. IV/M.938, October 15, 1997); Coca-Cola/Carlsberg, 1998 O.J. (L145) 41-62 (Case No. IV/M.833, September 11, 1997).

13 See, e.g., Charles A. James, Address Before the OECD Global Forum on Competition (October 17, 2001) (transcript available on U.S. Department of Justice Antitrust Division web site,

14 Brown Shoe Co. v. United States, 370 U.S. 294, 320, 344 (1962).

15 506 U.S. 447, 458 (1993). Accord, e.g., Eastman Kodak v. Image Technical Services, Inc., 504 U.S. 451, 482-83 (1992) (aggressive competition serves consumers and therefore is not condemned by the antitrust laws, even if it impairs rivals' opportunities, unless it is exclusionary, i.e., without valid business justification); see also, Muris, The FTC and the Law of Monopolization, 67 Antitrust L.J. 693 (2000).

16 See Monfort of Colorado, Inc. v. Cargill, Inc., 479 U.S. 104, 114-17 (1986) ("[C]ompetition for increased market share is not activity forbidden by the antitrust laws. It is simply vigorous competition. To hold that the antitrust laws protect competitors from the loss of profits due to such price competition would, in effect, render illegal any decision by a firm to cut prices in order to increase market share. The antitrust laws require no such perverse result, for "[i]t is in the interest of competition to permit dominant firms to engage in vigorous competition, including price competition.")

17 Lawrence H. Summers, Competition Policy in the New Economy, 69 Antitrust L.J. 353, 358 (2001).

18 Jean Tirole, The Theory of Industrial organization 6-7 (1988).

19 See Sylvia Nasar, A Beautiful Mind: A Biography of John Forbes Nash, Jr., Winner of the Nobel Prize in Economics in 1994 (1999). In a one-period prisoner's dilemma game, for example, the Nash equilibrium strategy would be to defect -- i.e., fink. In an infinitely repeated prisoner's dilemma game, the Nash equilibrium strategy would be to cooperate -- i.e., the code of silence. See William Poundstone, Prisoner's Dilemma 121-123 (1992).

20 See William Kolasky, Network Effects: A Contrarian View, 7 Geo. Mason L. Rev. 577, 584-85 (1999). See also Summers, supra note 17, at 358-59.

21 See generally Herbert Hovenkamp, Post-Chicago Antitrust: A Review and Critique, 2001 Colum. Bus. L. Rev. 257, 259; R.H. Coase, The Firm, the Market, and the Law (1988).

22 Our humility works in both directions. It also has led us to reject defenses in monopolization cases based on arguments that the markets involved are natural monopolies. Our view is that whether monopoly is the natural outcome is something to be decided by the market, not by courts, and that consumers have "an interest in competition even though that competition be an elimination bout." Union Leader Corp. v. Newspapers of New England, 284 F.2d 582, 584 (1st Cir. 1960). See generally Kolasky, supra note 20, at 596-98.

23 Commission Decision of 03/07/2001 declaring concentration to be incompatible with the common market and the EEA Agreement (General Electric/Honeywell, 3 July 2001, ¶¶353, 360, 376. I should disclose that before I joined the Department and while I was in private practice, my firm and I consulted with Honeywell with respect to the EU's review of the GE/Honeywell merger, after the Department has closed its investigation and the EU had issued its Statement of Objections. My discussion of the case in this paper is, therefore, based entirely on the Department's findings prior my joining it, as reported in our OECD white paper. When I use the pronoun "we," I am referring to the Department staff who reviewed the transaction and to the Department as an institution, not to my personal views.

24 United States Department of Justice Note for Discussion at Roundtable on Portfolio Effects in Conglomerate Mergers, supra note 2, at ¶¶ 53-60.

25 Hoffman-LaRoche v. Comm'n, Case 85/76, 1979 ECR 461 (CJ).

26 United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377, 391 (1956).

27 03/07/2001 Decision, ¶ 111.

28 Interavia (4/1/01).

29 This, of course, includes in GE's share sales of CFM engines. In attributing the sales to GE, the EU never explained why SNECMA, one of Honeywell's two principal competitors in the market for landing gear, would cooperate with GE in letting Honeywell dominant that market.

30 Götz Drauz, Unbundling GE/Honeywell: the Assessment of Conglomerate Mergers Under EC Competition Law, Speech Presented at the Fordham Corporate Law Institute, Oct. 2001, at 15.

31 Commission Decision of 03/07/2001 declaring concentration to be incompatible with the common market and the EEA Agreement (General Electric/Honeywell), 3 July 2001, ¶¶ 353, 360, 376.

32 See Barry Nalebuff & Shihua Lu, A Bundle of Trouble - Bundling and the GE-Honeywell Merger (October 2001) (not yet published but available from the authors).

33 Id. at 43.

34 Commission Decision of 03/07/2001, supra note 31, at ¶ 376.

35 Commission Decision of 03/07/2001, supra note 31, at ¶ 398.

36 Jaret Seiberg, Sticking to the recipe: Cross-Atlantic sniping continues over GE-Honeywell, The Deal, Oct. 24, 2001.

37 Drauz, supra note 30, at 14-15.

38 See id. at 16-38.

39 The theory is, however, reminiscent of the deep pockets theory in FTC v. Procter & Gamble Co., supra note 3, where the U.S. Supreme Court found ample evidentiary support for the FTC's findings that Procter & Gamble "with its huge assets" "might underprice Clorox in order to drive out competition, and subsidize the underpricing with revenue from other products." Id. at 575. The EU resistscharacterizing its theory as "deep pockets," but has not explained why.

40 See Bork, supra note 1, at 249.

41 Phillip E. Areeda & Donald F. Turner, Antitrust Law, ¶1135c (1980).

42 See Hovenkamp, supra note 21, at 257.

43 See id at 261, quoting United States v. Trans-Missouri Freight Ass'n, 166 U.S. 290, 372-73 (1897).

44 Under U.S. law, 35-40% share is typically the minimum threshold giving rise to foreclosure concerns. See, e.g., Jefferson Parish Hospital District No. 2 v. Hyde, 466 U.S. 2 (1984); United States v. Microsoft, 87 F. Supp.2d 30, 53 (D.D.C. 2000).

45 See The Economics of GE/Honeywell--Part 2: Foreclosure via "Share Shifting," Lexecon competition memo, 31 Oct. 2001 (available upon request from Lexecon).

46 See United States Steel Corp. v. Fortner Enterprises, 429 U.S. 610 (1977); see also Phillip E. Areeda & Donald F. Turner, Antitrust Law, ¶ 1105e at 19 (1980).

47 Allied Signal/Honeywell, (Case No. Comp/M.1601, January 2, 1999) ¶118.

48 See Drautz, supra note 30, at 25-28. It should not surprise us that the parties argued that the Cournot effects on prices would be minimal given that the Statement of Objections proposed condemning the merger because of those effects. In any event, the EU can not have it both ways. Either the Cournot effect was likely to be small and short-lived - in which case it is hard to understand how it could drive rivals from the market - or it was large and sustainable - in which case consumers would benefit.

49 See Jean Tirole, The Theory of Industrial Organization 173-77 (1988)("Consumers are better off under vertical integration because they face a lower price. Thus, welfare is unambiguously increased by the elimination of the double marginalization").

50 See, e.g., Atlantic Richfield Co. v. USA Petroleum Co., 495 U.S. 328 (1990)(holding that above-cost limit pricing does not violate the antitrust laws). See generally Areeda & Turner, ¶¶ 1136-39.

51 See Areeda & Turner, at ¶¶ 1105, 1135c.

52 Jack Welch, Jack: Straight from the Gut 362 (2001).

53 Council Regulation 4046/89, O.J. L 395/1 (Dec. 30, 1989), corrected version O.J. L 257/14 (Sept. 21, 1990), amended, Council Regulation 1310/97, O.J. L 180/1 (July 9, 1997, corrected version O.J. L 40/17 (Feb. 13, 1998), effective March 1, 1998, Article 2(3).

54 See U.S. Dep't of Justice & Federal Trade Comm'n Horizontal Merger Guidelines. 4 Trade Reg. Rep. (CCH) ¶ 13,104 (1992 and revised 1997).

55 Council Regulation 4046/89, supra note 49, at Article 2(1)(b).

56 See Seiberg, supra note 36.

57 Laura Tyson, The New Laws of Nations, N.Y. Times, July 14, 2001, at A29.

58 Drauz, supra note 30, at 24.