George Mason Law University, George Mason Law Review's Antitrust Symposium
"The issue of extraterritoriality, you know, that's a little bit like modern art."(1)
Joe has argued that US assertions of extraterritorial jurisdiction are often problematic and that our courts tend to be cavalier about other nations' interests. I am going to suggest that what is an extraterritorial assertion of jurisdiction in today's globalized world is quite complex, often depends on where you sit, and need not be -- and generally in the merger context is not -- a provocative act. Using the Boeing-McDonnell Douglas case as a springboard, I am going to argue that we both have a cooperative framework between the EU and the US for reviewing mergers with transborder effects and need such a framework with other countries. But we first need to understand what set of merger cases each competition authority is and should be reviewing. Second, we need to determine on what subset of those cases competition authorities ought to be cooperating.
What was striking about Boeing-McDonnell Douglas was that the EC was looking at a merger between two non-EU firms, neither of which had any production assets in the EU. Usually the "foreignness" of a merger that a national authority is investigating is not quite so blatant. There is no question that as the economy globalizes more and more industries are operating in regional or worldwide markets. And antitrust enforcers encounter more and more transactions between firms of different nationalities or between firms of one nationality entering and setting up production facilities or establishing services in another country's markets. But most transborder mergers and joint ventures that we review involve a US and a non-US firm, or, if both firms are non-US firms, at least one has production facilities in the US. The same holds for the EU.
The civil aircraft industry, and consequently the Boeing case, is quite unique, largely due to the increasing returns to scale of plane production, the magnitude of R&D costs, and the long lead time for introducing new models. What these industry characteristics mean is that Boeing produces planes only in the US and Airbus produces planes only in Europe. Hardly any other regional or global industries exist in which firms have production facilities in only one, rather than many, countries.
But the EC's review of Boeing-McDonnell Douglas should not astonish anyone, particularly Americans. The US has occasionally reviewed and taken enforcement action against a merger of foreign firms, neither of which had production assets in the US. In 1994, the FTC reached a consent agreement with Oerlikon-Burle, a Swiss firm that proposed to acquire Leybold, a German company. We concluded that the merger would reduce competition in the markets for turbomolecular pumps used in manufacturing semiconductors and in compact disc metallizer machines. Both companies sold substantial amounts of their production into the US, even though both companies' production facilities were in Europe. The FTC required divestitures in both lines of business.(2)
Likewise, the EC has previously entered orders when only assets outside the EU were at issue. Last year, the EC blocked the merger of Gencor and Lonrho's platinum mines in South Africa. Gencor is a South African mining firm and Lonrho is a British company with interests in mining, refining, hotels, and agriculture. (Thus, in contrast to Boeing, at least one EU firm was involved.) Because platinum is a fungible asset, easily transported and readily traded without tariff barriers, the EC found the world to be the relevant geographic market. Moreover, Western Europe represented about 20% of the world's demand for platinum. The merger would have given the parties a market share of 35% based on production volumes, similar to that of another South African company.(3) However, about 90% of the world's platinum reserves are in South Africa, so allowing the merger would have left only two firms controlling 90% of the world's reserves.
There also have been several mergers reviewed by both the US and the EC where the two competition authorities have reached different enforcement results. So again Boeing was not unique, it just was one of the few instances where the differences raised eyebrows. For example, in the recent $60 billion merger of two Swiss firms, Ciba-Geigy and Sandoz, both the EC and the US had competition concerns. Both the FTC and the EC closely examined markets for flea control products and crop protection products. Both authorities agreed that there were separate, national geographic markets for each product group, largely because these products required national regulatory approval before marketing. But the EC found no Member State market in Europe sufficiently concentrated to require remedial enforcement action for either product. The FTC, in contrast, found a more anticompetitive situation in the US market for each product and required divestitures covering each product.
Likewise, both authorities examined the R&D market for gene therapies for the treatment of brain and other tumors. In that R&D market, the EC concluded that potential anticompetitive effects were too speculative to seek relief. The EC noted the questionable efficacy of gene therapy, the uncertainty surrounding the parties' patent claims, and the research that potentially competing firms were conducting. The FTC, in contrast, concluded that the merging firms were the leading commercial developers of gene therapies in a highly concentrated market where no new entry was likely to deter or counteract the anticompetitive effects of the merger. We therefore obligated the parties to license specified gene therapy technology and patent rights to Rhone-Poulenc Rorer, in order to preserve innovation competition in this critical R&D market.
The first question to be answered is should more than one country be reviewing these mergers? I think that the answer is an easy yes. In competition law, unlike trade law, the nationality of the firms involved is irrelevant, and the location of assets or production facilities is also unimportant, except insofar as the ability to reach them may be critical for effective relief. What does matter is whether anticompetitive activities -- wherever they occur and regardless of the nationality of the firms involved -- harm consumers or competition in one's market. I believe that both the US and the EC have a right to be interested in mergers occurring inside, outside and across their borders if those mergers are likely to affect their markets (or commerce) and lead to higher prices or lower quality goods for their consumers.
The next question is should different authorities reviewing the same merger be coming to the same result and does it matter if they don't? Sometimes two authorities looking at the same transaction should come to different results because the transaction will in fact have differing impacts on different markets. In the Ciba-Geigy/Sandoz case, the need for FDA or EPA regulatory approval to market the relevant products (flea control, crop protection and gene therapy products) in the US meant that the relevant US market was nationwide. In the EC, the regulatory approvals necessary entailed that Member State markets were the relevant ones. It is not surprising that the identity of firms competing in these different markets, and their market shares could differ, and that the merger would cause disparate effects in these separate markets. Nor is it unusual that the different markets may be served from different plants and that the remedies that may be necessary to solve the antitrust problem on one side of the ocean -- such as selling a plant or licensing technology -- can proceed without significantly affecting how the transaction is resolved on the other side.
Sometimes, however, the different standards in our laws will lead US and EC authorities to different results, even when looking at roughly the same facts. I do not believe this is bad because antitrust laws are not pure economics -- they reflect different nations' histories and cultures and values. We, for example, do not try to achieve any market integration with our competition law, as the EC does. Our Constitution's Commerce Clause, by prohibiting state actions that discriminate against interstate commerce, long ago laid the foundation for relatively unimpeded nationwide commerce. Perhaps it is worth commenting in this context that our Constitution's Commerce Clause functioned much like a trade law by eliminating governmental barriers between states. Consequently, when Congress passed our antitrust laws a century later, those laws simply had to insure that private barriers did not replace the prohibited state barriers. The EU's national boundaries and barriers have proven far more enduring and thus even competition law is enlisted to remove those barriers.
Moreover, I think it is fair to say that the EC focuses more on single firm dominance and the US focuses more on oligopoly coordination. I suspect that this is because the EC often encountered extremely large firms in certain markets, perhaps due to earlier state support or monopolies. Our antitrust history began with concern about industries like oil, steel and railroads, where a handful of large players coordinated their actions via trusts or holding companies. Moreover, we have always been more tolerant of single firm growth and even dominance on the theory that, if we are going to drive competitors to compete, we should not punish them when they have competed successfully, so long as they did not resort to exclusionary or anticompetitive tactics.
I do not expect EC competition law to reflect our values nor do I think it necessary that our law conform to the EU's values -- both systems reflect legitimate differences grounded in our pasts. Yet by and large, our two sets of competition laws and enforcement approaches and analyses are far closer to each other than are those of most other countries. In fact, even if countries were to agree on some perfect binational or international competition code -- which I regard as neither possible nor desirable -- we would still run the risk of looking at the same facts under the same standards and reaching different results. Antitrust merger analysis necessarily involves complex assessments about the future impact of a transaction and reasonable minds can differ on what the correct outcome should be. In the FTC's Boeing decision, it is thus not surprising that one Commissioner dissented from her three colleagues, even though she was reviewing the same facts and applying the same law.
So what can we do to minimize those differences in national merger laws that may impact adversely on other countries' important interests (as well as the firms involved) while tolerating those differences that are legitimate and nonthreatening? It was not unexpected to see the Financial Times quoting Sir Leon Brittan, in the aftermath of the Boeing matter, as advocating the need for an international agreement on competition rules and smoother co-operation between national competition jurisdictions. He urged the WTO to give high-level attention to how competition policy affects the world trading system, so as to avoid clashes when two or more competition authorities review the same case but apply different rules. What did surprise me was that some of the very businesses that have been extremely wary about the prospect of a uniform competition rule began suggesting that perhaps a single rule or a single reviewing authority made sense.
While Brittan is right to argue for more and improved cooperation among competition authorities, he and others are wrong to think that a single code is the answer. What would such a code look like? Interestingly, both the US antitrust statutes and Articles 85 and 86 of the Treaty of Rome are worded very broadly. This generality is highly desirable, since it allows national antitrust law to evolve and change as market dynamics change and the economics of industrial organization are better understood. A highly detailed code would tend to freeze antitrust thinking and ultimately impede globalization and innovation.
But a highly general code would be unproductive, if not hazardous, at the international level. Of course, a broadly worded code is all that could be hoped for among the hugely disparate WTO countries, many of which do not even have competition laws. It is noteworthy that currently even like-minded countries, such as OECD members, are having some difficulty defining what to jointly agree to prohibit as cartels -- the one area in which they all concur that their laws are most similar! Any agreement on prohibited anticompetitive conduct among WTO countries, which needs to either accommodate (or worse, ignore) the full spectrum of the world's varying competition laws, would by definition be so vague as to be meaningless. And such breadth and generality could lead to trouble if international panels of arbiters were allowed to interpret the code's vague provisions and second guess the complex and highly fact-intensive decisions that national courts make about the effects of a merger. In addition, extending the review of national merger decisions to allow for WTO scrutiny would effectively crater many merger deals. In today's fast-paced business world, where some high-tech products have life-cycles of three months, delay can be death.
Nor do I think that multi-country review of mergers, with the possibility that it poses for disparate results, can be solved by positive comity. As you may know, the US and EC have reached tentative agreement on an improved positive comity agreement, which interestingly covers all types of anticompetitive conduct other than mergers. What positive comity entails is that where the interests of two countries are adversely impacted by anticompetitive activities, the matter can be referred to the country within which the activities are principally occurring. This presumption that the country with the greater connection with the bad conduct will be the investigator and enforcer leads to a variety of efficiencies. The investigating country is better positioned to obtain evidence and fashion relief, dual prosecutions are eliminated, and the frictions caused by extraterritorial enforcement are minimized.
Yet mergers do not fall neatly within such a system. Both the US and the EC operate under very short deadlines when investigating mergers. In fact, the EC does not have discretion to allow us to investigate a merger on its behalf -- it must review all mergers that pass the EC's jurisdictional thresholds. But more fundamentally, neither country can risk asking the other to investigate a merger only to discover later that its consumers and important interests have not been adequately protected. At that point, it is too late to investigate on one's own, which is the option that always exists in other types of cases. Unfortunately, the deadlines that prevent firms from consummating the merger will have passed and the firms' assets will be scrambled.
Thus, we are left with the very practical, if perhaps unglamourous, reality of improved consultation, information sharing -- within the confines of confidentiality restrictions -- and cooperation. The amazing thing is that this works and could even work more effectively. Under our 1991 cooperation agreement, the EC notifies us whenever it reviews a merger involving US firms, assets or markets. We notify the EC whenever we begin investigating a merger involving an EU firm, whenever the EU's important interests are involved or whenever we seek documents or testimony from EU persons or entities. The constant contacts between officials and staff enable us to understand each other's analysis, lead to convergence in our approaches toward competition matters -- in some measure due to an increasingly common economic analysis -- and benefit parties insofar as we are often able to arrive at complementary remedies. Thus, even if the transaction needs to be addressed somewhat differently on both sides of the Atlantic because of differing market conditions and competitive realities, we reach solutions involving divestitures and licensing that neither conflict nor force firms to choose between complying with US or EC law.
I believe that the Boeing case in fact fits this pattern far more than it deviates from it.
Staff and officials on both sides consulted fairly regularly on developments in the case, sharing with each other how we were analyzing the various commercial aircraft and defense markets at issue and what factors were relevant under our respective laws. Yet there is no doubt that this is one case where modest differences in our competition laws made for somewhat striking differences in outcome.
Under US antitrust law, the FTC's decision not to challenge the transaction was straightforward. On first blush, the proposed merger appeared to raise serious antitrust concerns in the commercial aircraft market: Boeing accounts for roughly 60% of the sales of large commercial aircraft and McDonnell Douglas' market share was slightly below 5%. Airbus was the only other significant rival and barriers to entry were exceedingly high. But the most important focus for US competition law is what effect the merger is likely to have on prices, in particular, what airline customers likely would pay. Thus, the critical question became whether Douglas Aircraft, McDonnell's commercial arm, had prospects of playing a significant competitive role in the future in the commercial aircraft market. The virtually unanimous testimony of about 40 purchasers, both foreign and domestic, of aircraft was that Douglas' prospects for future sales were close to zero. Those purchasers did not view Douglas as a significant factor in future plane bids or competitions and would have said so in court. In addition, nothing indicated that McDonnell Douglas rationally could be expected to invest the vast amounts necessary to create even the possibility of turning itself around. Nor was any other market player ready to purchase all or part of Douglas and compete in the future. From a US perspective, accordingly, the merger did not reduce existing aircraft producers from three to two; rather, the market already consisted of only two significant players.
The EC's competition law, in contrast, focuses more on whether a merger increases the leverage that can be exercised by a dominant firm and the possible impact of the merger on competitors. The EC's merger law takes into account the merged firm's "economic and financial" power in a way that US law does not. Thus, it was also a straightforward application of the EC's law that led to its conclusion that the merger would increase Boeing's dominance to the detriment of Airbus and competition in the commercial aircraft market. In addition, the EC competition law is much tougher on exclusive agreements than US law. It finds exclusive arrangements to be anticompetitive when contracts foreclose a smaller share of the market and when the exclusive arrangements are of a shorter duration than would be needed to find a US antitrust violation. Moreover, the EC treats exclusive supply agreements as a relevant structural element in a market that is being analyzed for purposes of a merger review. So again, it is not surprising that the EC was concerned with Boeing's exclusive supply agreements with two, and later three, major airline customers. Indeed, the EC had previously required Fiat to give up an exclusive supply arrangement with its Italian distributors before its proposed merger with Ford's tractor and agricultural machinery businesses could be cleared. In contrast, the US in Boeing's case tended to regard the exclusives as business arrangements that antedated the merger and were not relevant to how the merger might change competition and prices.
Notwithstanding these substantive differences in our laws (as well as earlier cases, such as General Dynamics in the US and de Havilland in the EC, that foreshadowed, at least for antitrust initiates, how the case might play out on each side of the ocean), cooperation did work to a great extent. The EC acknowledged important US interests in declining to take any action with respect to defense markets, such as fighter aircraft. The US, in turn, acknowledged what were likely to be EC concerns, by announcing that Boeing's exclusive agreements with airlines were "troubling" and committing to monitor the anticompetitive effects of those and any future long term exclusive contracts. Finally, I believe that the compromise that Boeing and the Commission reached, against the backdrop of discussions between US and EC officials, in fact satisfied the principal competition concerns of the EC without undermining the competition interests of the US. Indeed, I do not believe that a WTO dispute settlement situation could have improved upon this resolution of the case, nor do I believe that either we or the EC would willingly give up our laws in favor of some uniform substantive multilateral code.
* The views expressed here are those of the author, and not necessarily of the Federal Trade Commission or any Commissioner
1. New York Times, A3 (Oct. 1, 1997) (quoting Felix Rohatyn, three weeks after presenting his credentials as the new American Ambassador to France, commenting on Total's $2 billion investment in Iran notwithstanding U.S. sanctions against foreign countries that invest more than $20 million in Iran). The opinions in this speech do not necessarily represent the views of the Commission or any individual Commissioner.
2. The FTC in 1990 also obtained a consent agreement in the Institut Merieux case, although neither party maintained production facilities in the relevant market -- rabies vaccine -- in the US.
3. Russia had the remaining production.