Printer Friendly

The political economy of international antitrust harmonization.


International harmonization of competition laws is in the air. A large number of academics have called for harmonization of the substantive content of antitrust laws. (1) They have noted the potentially large costs of divergent national antitrust laws applied globally. These costs include the transactions costs for companies complying with multiple regimes (2) and the costs of being governed by the most restrictive antitrust regime, even if that regime is suboptimal. (3) More importantly, as a practical matter, the recent Doha meeting of the World Trade Organization (WTO) has called for the next round of world trade talks to take steps toward harmonization, including harmonization of certain core substantive standards. (4) Thus, the intellectual impetus for harmonization now may enjoy a plausible forum for its realization.

In my view, substantive harmonization, even if limited to core competition standards, would be a major mistake. It is undoubtedly true that multiple regimes impose some costs, but substantive harmonization--by which I mean a single international regime binding on all nation states in at least some areas of antitrust--also has potential costs. An international lawmaking regime creates high agency costs because it is less subject to democratic control than national regimes. (5) It also imposes costs by discouraging beneficial change, as the regime once in place will be difficult to alter. (6) Moreover, the appropriate scope of antitrust law in different nations may differ, depending on such factors as the size of their markets, their openness to trade, and their administrative competence in enforcing regulatory laws. (7) Thus, an international regime might well lead to an overall worse world competition policy.

The long-run costs of a substantive antitrust regime are particularly problematic in a world that is not static. As information costs, transportation costs, and trade restrictions decline, it may well be that the appropriate scope of optimal antitrust rules will tend to narrow as market processes become better correctives to market imperfections than government intervention. The lock-in costs of an international regime thus are particularly high in a world in which the pace of change is ever increasing.

In contrast to substantive harmonization, I offer an argument for a limited and modest antidiscrimination international antitrust regime located within the WTO. The rationale for this regime, however, comes principally from international trade law rather than antitrust law. Foreign bias in competition laws is likely to become a greater problem as the WTO eliminates tariff and other barriers to trade in goods and services. The WTO should block substitution of discriminatory antitrust law for barriers that it has removed in order to sustain progress in world trade. This effort would be a modest extension of its existing mission: it already attempts to prohibit many other forums of regulatory discrimination that interfere with exporters' market access.

The antidiscrimination model also has advantages over substantive harmonization, because formulating and applying antidiscrimination rules have fewer agency costs than formulating and applying substantive rules. (8) Moreover, the antidiscrimination model permits continued innovation and change in substantive rules, thus facilitating continued debate regarding the optimal content of regulation. (9)

Part I of this Article will critique the arguments for substantive harmonization of antitrust laws. It will suggest that these arguments are unpersuasive because they fall to show that the costs of our decentralized system of competition law are greater than the agency costs and associated pathologies of more centralized rule making and enforcement. In particular, arguments for substantive harmonization fail to recognize that a decentralized system has a certain dynamism over the long-run: the conflicts between different systems may become a focus of public attention and lead to better laws. (10) In contrast, a harmonized international regime faces a greater danger of stasis. (11)

Part II of this Article will address a new argument that international cooperation on substantive antitrust is necessary because domestic antitrust regimes neglect foreign consumer and producer interests, thus departing from an optimal antitrust standard. First, this Part will use the U.S. antitrust regime as an example to suggest that some institutional structures, such as an independent judiciary, tend to constrain this bias in some circumstances. (12) In other cases, this bias, even if unrestrained, may compensate for public choice flaws in a domestic antitrust system and actually may result in a more efficient competition policy. (13)

Part III of this Article will provide a sketch for a limited antidiscrimination model of international competition law. It will first show that the neglected rationale for some policing of domestic antitrust law is that antidiscrimination requirements are necessary to protect the efficacy of both the WTO's tariff reductions and its elimination of other trade restrictions.

I then discuss briefly the limited and incremental nature of antidiscrimination principles that the WTO should embrace. WTO enforcement of antidiscrimination rules should be limited to discrimination that creates nontariff barrier substitutes for tariff barriers such as antitrust rules that discriminate in market access. Only that kind of discrimination undermines the WTO regime. Until more evidence develops that foreign bias is a serious problem in the administration Of antitrust laws, I might well apply the antidiscrimination regime only to expressly discriminatory laws. Third, even if the WTO were to broaden its antidiscrimination regime to include discriminatory application of antitrust laws, it should consider ways of assuring nondiscrimination through the use of domestic institutions rather than increasing the power of WTO tribunals.

I also describe the concrete advantages of an antidiscrimination regime over substantive harmonization. First, it will preserve substantive diversity and therefore experimentation, preventing lock-in problems. (14) Second, it will avoid many public choice problems that would make suboptimal substantive harmonization likely. (15) Third, an antidiscrimination model will comport with the rest of WTO jurisprudence and, if the model is located there, will benefit from the WTO's already established institutional structure, which includes a sophisticated jurisprudence for ferreting out discriminatory laws. (16) Unlike substantive rules, an antidiscrimination model will not require substantial transformation of WTO institutions, change that would be counterproductive to the rest of the WTO regime.


A variety of arguments have been made to support the international harmonization of substantive competition law. Two of the oldest are also the simplest. First, in a world of multinational corporations, multiple antitrust regimes raise transactions costs. (17) Second, in a world with extraterritorial application of antitrust laws, the most restrictive antitrust laws will always govern business behavior even if those antitrust laws prove suboptimal. (18)

These traditional arguments do not seem very persuasive by themselves because they do not attempt to compare the substantial costs of diverse antitrust laws with the costs of harmonization. These omissions place us in danger of committing the internationalist version of the nirvana fallacy: even if uncoordinated national regimes are inefficient, it does not follow that an international regime will be more efficient. (19) The transactions costs of complying with different laws can be substantial in terms of the costs of professional services, (20) but these costs seem to be dwarfed by the question of whether the international rule will prove substantively better. Moreover, some of the transactions costs of compliance with multiple laws can be reduced by nonsubstantive harmonization--i.e., harmonization of the forms of requests for information. (21) Steps for such procedural harmonization are already being taken bilaterally and unilaterally. (22)

The debate over substantive harmonization, thus, should center on whether the decisions under a harmonized legal system are likely to be better than the sum of the legal applications of diverse systems. This comparison requires an evaluation of both the benefits and costs of an international harmonized regime as well as the benefits and costs of diversified national regimes. As the standard of comparison, this Article will first utilize the consumer welfare model of the Chicago School as an economic efficiency standard. (23)

A. The Political Economy of Domestic Antitrust

We can best understand whether an international standard would tend to depart from a consumer welfare model of antitrust if we understand why national standards, at least within developed nations, tend to depart from this model in the first place. In developed nations, several factors tend to make their competition law depart from the consumer welfare ideal. (24)

The first distorting factor is the interest government antitrust officials have in more than optimally interventionist antitrust rules. This distortion reflects two kinds of bias. First, such an interventionist set of rules allows antitrust officials to gain more rents. (25) In return for enforcing the antitrust law against a company, officials can gain rents from competitors. (26) It is true that in refusing to exercise official discretion to enforce laws against a company, antitrust officials can obtain rents from that company; however, to make this threat of enforcement credible, they occasionally have to enjoy the discretion to intervene and occasionally exercise it. Second, the Antitrust officials charged with intervention have an interest in the larger budgets and staff that a more interventionist policy brings. Such a policy permits more prestige to agencies. (27) It also brings more postbureaucratic employment opportunities, as an interventionist policy requires many consultants and lawyers to handle the effects of such policies on behalf of corporations. Broadly speaking, the first kind of bias--campaign contributions and other favors from excessive intervention--will tend to affect elected executive officials and the second kind of bias--prestige and employment opportunities from the intervention--will tend to affect the permanent bureaucracy. (28)

The judiciary or legislature is also unlikely to limit the pathologies of antitrust enforcement. For example, in the United States antitrust enforcement offers legislators rent-seeking opportunities similar to those of antitrust officials: they can extract rents by establishing a system that permits excessive intervention because they can use legislative hearings and the budget process to influence the cases executive agency regulators, i.e., the Department of Justice's Antitrust Division, bring. (29) Moreover, even if legislators were motivated entirely by the public interest, antitrust statutes cannot be written like a code, because no code can capture all anticompetitive business practices. (30) Thus, executive regulators inevitably enjoy substantial discretion to shape the interpretation of antitrust statutes in a manner enabling them to pursue their own interests.

The judiciary's ability to restrain regulators can also be limited. (31) First, regulators can act in a way that makes judicial review of their actions ineffective. For instance, particularly in Europe, regulators can act to block a merger and by the time effective judicial review can reverse that decision, business considerations may force termination of the merger. (32) Second, even in cases where judicial review is effective, the judiciary has often deferred to the regulators because there are costs to the judiciary in aggressively standing up to the democratically elected branches, particularly when the law, such as antitrust law, does not provide bright-line rules. (33) Finally, the tendency toward self-aggrandizement that drives bureaucrats is also known to the judiciary. For instance, the prospect of transforming and supervising an entire industry, as Harold Greene did with the phone industry, cannot be discounted as a possible distorting motive. (34) Thus even if institutionally constrained, the basic hum an tendency toward self-aggrandizement contributes to rules that will permit excessive intervention.

The public, furthermore, in the form of voters, cannot be expected to exercise much control over the form of competition law. Because of rational ignorance, the public confronts large agency costs in monitoring the behavior of legislation and executive action in any area. (35) Competition law tends to be quite technical, which tends to exacerbate the difficulties of democratic control over bureaucratic elites. (36)

Finally, a heuristic bias contributes to over enforcement. One must always remember that the central question in antitrust is whether government intervention will serve to correct anticompetitive practices better than the market itself. (37) Only when antitrust intervention will better correct anticompetitive practices is government intervention justified. The answer to this central question depends largely on whether one is considering correction in the short run or long run. In the long run the market erodes monopoly power, decreasing cartels and monopolies. (38) Humans, however, have a heuristic bias that causes them to discount these long-run facts. Because the future is uncertain, people use a "representativeness" heuristic, by which they assume that future patterns of events will resemble the past patterns with which they are familiar. (39) Political actors, both legislative and executive, will thus tend to underestimate the likelihood of the destabilizing change that the market brings to cartels and monopolies in the long run.

B. The Political Economy of International Antitrust

After considering the reasons why national antitrust laws tend to be suboptimal, we are in a better position to assess whether internationally harrmonized antitrust will prove more effective than the current decentralized system. In my view, the international regime has distinctive costs in the form of higher agency costs and deprives us of distinctive benefits of the diversified regime in the form of potentially useful innovations in competition law.

1. Higher Monitoring Costs

International harmonization has a nice ring to it, conjuring up "an image of citizens of many nations happily singing in harmony." (40) In the context of polities, as opposed to music, however, where individuals may use centralization to gain resources for themselves, international harmonization can become the song of the oligarchs. (41) Agency costs are likely to be higher at an international level and thus bureaucrats will have more ability to fashion rules in their own interest. First, average citizens will find the international process even more opaque than the domestic process and thus have more difficulty monitoring those charged with carrying out antitrust policy. (42) For the American citizen, Geneva is more distant than Washington. Second, more rents are available on a global scale. If business groups can gain international intervention in their favor, they can disable a whole world's worth of competitors. (43) Finally, international antitrust regulators, like other regulators outside the control of government, become a distinct class with a distinctive interest--growing the international antitrust apparatus. This interest, however, is not likely to mirror the interests of their government leaders, let alone the interests of the average citizen in the world. (44) Because of these costs, rules fashioned at the international level may be less in the public interest than the rules in many individual nations.

2. Higher Enforcement Costs

We also must consider the costs of enforcing an international harmonized regime. The kind of enforcement costs differ depending on whether an international agency enforces international competition law directly or individual nations enforce the international standards themselves. This section will consider each set of costs in turn. If a substantive regime is to be enforced directly as a matter of domestic law by international bureaucrats, it will be a regime without precedent. Its novelty alone may entail substantial costs. First, any new regime has start-up costs and is more likely to make errors because of a lack of appropriate traditions, rules, and institutions. International decision makers will create high agency costs, for the reasons discussed above. (45) Finally, because such an international corps of decision makers will depend ultimately on the coalition of support from leaders of nation states around the world, these decison makers will need to take the national identity of actors into account when making their decisions, thus distorting the agreed upon competition principles. Even if they did take substantial account of national government interests, it remains an open question whether international regulators would have the political legitimacy to make their decisions stick if a nation state vehemently disagreed.

Perhaps most costly of all, such a regime would necessitate changes in the constitutional structures of nation states. For instance, in the United States, the Appointments Clause (46) and the Article III judiciary requirement (47) represent a serious constitutional bar to giving international decision makers authority to render decisions that would directly affect the decisions of the United States. (48) The costs of changing such constitutional provisions are not limited to the costs of the constitutional amendment process. They also include the cost of destabilizing the constitutional norms that serve to Ameliorate public choice problems of government by assuring An accountable executive and an independent judiciary. Even if these norms are worth relaxing for the cause of international antitrust harmonization, relaxing them may create costs in other areas where their loss would not be worth the cost. It is no answer to say that we can make a case-by-case decision as to whether the costs are worth paying. Our decision to constitutionalize accountability requirements, such as the Appointments Clause and the Article III judiciary requirement, represents a judgment that day-to-day political assessments regarding these matters will be systematically incorrect.

Another mechanism for antitrust harmonization would be to set international standards but allow individual nations to enforce these standards themselves. In other words, the rules would be centralized, but enforcement would be decentralized. Such a structure, however, allows for nations to diverge dramatically in their application of the standards over time. Even within a single nation, the judiciary has essentially interpreted the same antitrust law with spectacularly different results in different eras. (49) The results of a structure with centralized standards and decentralized enforcement could vary widely between nations because a nation's enforcement of the international standards will be influenced by that nation's preexisting substratum of competition law. (50)

3. The Loss of Innovation and Competition Through Substantive Harmonization

It is also necessary to evaluate whether the international regime would lose any benefits distinctive to the diversified regime through harmonization. One could argue that antitrust law is unlikely to be worse under an international system because under a decentralized system in a globalized world, the most restrictive system will govern. Even apart from the problem discussed above that agency costs may make international standards more excessively interventionist than those of any major nation, this argument may not be correct. (51) First, the most restrictive system in a decentralized, global antitrust system will only govern if that system has jurisdiction with respect to a given antitrust enforcement matter. Even in a globalized world, some nations will not always have jurisdiction to apply their antitrust law. For instance, a merger may occur that has implications for the United States and Asia but which does not effect Europe. Assuming that Europe is the most restrictive antitrust regime, its antitrust law would not apply to the merger when the United States and Asian nations are the only nations to have jurisdiction. Thus we would have to compare the sum of the antitrust applications in the decentralized system to that of the harmonized system to determine the harmonized system's comparative benefit.

Second, and most importantly, we should investigate the distinctive benefits of a diversified regime in a world that is not static. In the 1980s, U.S. antitrust law and policy was transformed to embrace a much more optimal economic regime with a better focus on consumer welfare. (52) One question is whether there will be some process by which excessively restrictive regimes will undergo similar transformations. In my view, the different rules operating within a diversified regime may move to a more optimal level by virtue of their very diversity because diversity creates opportunity for legal innovation and change.

Commentators have rightly noted that competition law applied to international transactions will not necessarily be optimized by jurisdictional competition. Because of its extraterritorial reach, a nation can apply its competition law to firms that do not choose to locate there (53) and firms thus cannot exit to put pressure on the antitrust law to become more optimal. Conversely, if firms could choose their antitrust law without regard to their domicile, some might choose the least restrictive regime which may well prove suboptimal. (54) Nevertheless even with its imperfections one should not dismiss the possibility that jurisdictional competition will do some good. Antitrust law, like business and tax law, affects business conditions, and nations will prosper insofar as they have good antitrust rules, at least as applied to purely domestic transactions. While they may then choose to discriminate and apply worse law to antitrust practices in order to help domestic producers at the expense of foreign producers or consumers, such a strategy can be policed by an antidiscrimination rather than substantive antitrust regime.

Moreover, even in the absence of the formal conditions for effective jurisdictional competition through movement of firms, a decentralized regime will foster innovation and debate in antitrust law. (55) Certainly the European Union's (EU) merger laws have been heavily criticized when they have prohibited mergers--such as the GE-Honeywell merger--that the United States has allowed. (56) Moreover, this criticism appears to have prompted changes in the EU regime. (57) Conflicting decisions and the resultant tensions can bring beneficial change by focusing the attention of the press and other elites outside government circles on the wisdom of government antitrust intervention. Such focal points of criticism would be lost with harmonization. (58)

It might be argued that centralized institutions could presently take advantage of the growing knowledge of the Chicago School antitrust benefits disseminated by graduate students worldwide. Public choice analysis, however, suggests that knowledge is not a sufficient condition for good antitrust decision making. (59) The decision makers need to have the correct motivation. Conflict is more likely than increased knowledge to produce the correct motivation, because conflict will give these academic views salience Among the press and other elites.

A thought experiment is helpful in illustrating the way in which a decentralized regime may be more open to beneficial innovation than a centralized regime. Assume that antitrust law today was that of the Warren Court era (a very sorry state from a consumer welfare perspective). (60) Under those circumstances, it would be very likely, even without the additional agency costs of formulating international rules, that the harmonized regime of that era would have been suboptimal from the consumer welfare perspective. Assume further that, as in the 1980s, a laissez-faire revolution sweeps the United States but does not reach Europe. Changing the international regime would be more difficult than changing the United States' regime alone because the international regime would be much more impervious to being destabilized by a revolution limited in geographic scope. If the United States' regime were not transformed, it could not provide an implicit critique of other national regimes.

Despite the public choice tendencies in the developed world for excessive Antitrust law intervention, revolutions in social thought and changing political institutions can sometimes upset these interventionist policies. These revolutions are much more likely to succeed bit-by-bit in a decentralized regime as they take hold in one nation and spread by example to others. Innovations emanating from the United States, in particular, seem to have been responsible for many beneficial transformations in international competition law, as well as many other areas of law and culture. (61) In my view, it would be a mistake to set up international structures that inhibit the dynamism of the United States from continuing to exert maximum influence by example.

The need for continually innovative perspectives on antitrust law is especially important because the comparative advantage of government intervention depends on how swiftly the market will correct for supercompetitive prices in the absence of such intervention. (62) This market process in turn depends on transportation, information costs, and trade restrictions, as a world in which supplies can be redirected easily from one country to another is a world in which the market can correct supercompetitive prices more quickly. (63) The point here is that the appropriate set of rules authorizing government antitrust intervention may differ depending on the effectiveness of general market discipline. Intervention in markets through any bureaucratic rules, no matter how sensible in the abstract, has potential costs because the government may make mistakes in rule application. Thus, the optimal shape of antitrust rules, like market prices themselves, is always changing. International antitrust enforcers, therefore, benefit from continuing experimentation to gauge the best contours of competition law.

In contrast, an international regime faces a higher degree of stasis in its rules than national regimes because the transactions costs of getting agreement from nearly all nations to change the rules are higher than the transactions costs of getting change from a single nation. (64) If, on the other hand, one were to delegate substantial discretion to international bureaucrats, their power would exacerbate the problem of agency costs. (65)

Of course, this Analysis assumes that economic efficiency should be the goal of antitrust law. Other objectives of antitrust law, however, would be subject to a similar public choice inquiry. First, why do national antitrust laws fall short of their identified objectives? Second, why do these same difficulties not prevent international harmonized antitrust from better realizing these objectives? (66) Thus, whatever the appropriate objectives of antitrust, their optimal realization depends on considerations of political economy.

It may be thought that international antitrust norms have some compensating advantage. I have suggested that in the developed world there is a tendency for antitrust law to be excessively interventionist. (67) In the developing world, however, the absence of antitrust law often leads to underenforcement of competition norms. (68) The developing world lacks government infrastructures to effectively enforce competition law. (69) Moreover, in the developing world, government officials or their families are often partners in major industries and thus, in contrast to the officials in the developed world, they can collect rents directly by sharing in the monopoly profits from collusion and other anticompetitive practices in which their companies engage. (70)

One might argue that an agreement forged between the developed and developing world may reach a happy medium, but I am skeptical. As discussed above, antitrust law must be written with bread concepts and those charged with enforcing them will be largely Western bureaucrats who are motivated by the prospect of the greater prestige and rents that rules permitting excessive intervention promote. Without a world government or global demos, the actions of these international enforcers will be difficult to monitor.


A. Introduction

A new and interesting argument for greater cooperation on substantive antitrust principles is that nations will systematically fail to take account of the interests of foreign producers and consumers in their antitrust law and policies. (71) Consideration of foreign producers and consumers, however, is necessary to maximize global consumer and producer welfare. Thus, application of antitrust law under national systems will be suboptimal, because it will not even attempt to maximize consumer and producer welfare generally. (72)

The argument that decentralized systems result in suboptimal antitrust enforcement represents the best and most comprehensive attempt to show that a structure of international competition law--or least substantive international principles--will represent an improvement over the present decentralized regime. This argument is better than traditional arguments (73) for four reasons. First, the objective sought--that antitrust laws should be applied without respect to nationality--is modest and may well command consensus. Second, the claim that domestic antitrust laws neglect foreign interests has the ring of truth. Third, as globalism makes trade more important, it is at least superficially plausible that a country's failure to consider foreign interests has large costs because corporate actions--potentially regulated by antitrust--affect more and more foreign producers and foreign consumers. Finally, the problem of foreign bias has a clear nexus with cooperation on substantive principles. If the world were to move toward a single antitrust law enforced by an international organization, parochial bias might be reduced substantially, if not eliminated, just as the federal enforcement of antitrust laws in the United States has reduced concern about state bias. (74)

Professor Andrew Guzman, the leading proponent of this argument for substantive coordination, employs a relatively simple model to show that countries will take account of only their own citizens' interests. (75) Nations simply neglect the interests of foreign companies and foreign consumers in their calculations of welfare-maximizing competition policies. (76) For instance, a country with only exporters of a good and no consumers of that good will turn a blind eye to anticompetitive practices that detract from consumer welfare. (77) Conversely, nations with only consumers and no producers will not take the increased foreign-producer profits from transactions such as mergers into account in assessing whether these transactions should be permitted. (78)

This tendency to neglect foreign interests can theoretically result in two kinds of divergences from the competition law that a nation would have chosen in the absence of trade with other nations. First, it may simply result in either de jure or de facto exceptions to the antitrust laws. For example, a nation that generally prosecutes cartels could simply exempt domestic cartels that affect only foreign consumers from the Ambit of its laws. A nation could also decide to prevent vertical integration of foreign manufacturers that want to sell in the nation's markets while permitting vertical integration of the nation's domestic companies. We will call these divergences "exception divergences."

Second, if such outright exceptions were not possible, a nation might change its antitrust law from what it would optimally deploy in a closed economy--i.e., an economy where only domestic consumers and producers were at issue. (79) If a nation consumed less than it produced and did not distinguish between antitrust laws affecting consumers and those affecting producers, the nation would tend to have weaker antitrust laws than it would in a closed economy, because antitrust laws protect consumers while depriving producers of monopoly profits. (80) The tendency toward weaker antitrust laws would be particularly pronounced if the nation exported imperfectly competitive products, such as pharmaceuticals and other intellectual property-rich products, because monopoly profits from exports would redound to the benefit of the nation. (81) Conversely, if a nation consumed more than it produced, then it might tend to have stronger laws than it would in a closed economy. (82) This tendency would be particularly pronounced if the nation exported perfectly competitive products because antitrust laws would not affect these products anyway. (83) I will cell this kind of divergence "distortion divergence" and address these two kinds of divergence separately.

Professor Guzman suggests that cooperation on substantive international antitrust provides a potential solution to the problem of divergences, because international rules and enforcement would have no reason to discriminate on a parochial basis. Thus, at least potentially, antitrust enforcement would move to a more optimal level.

The attractiveness of this solution depends on how problematic parochial bias is for antitrust and whether there are features of international harmonization that actually make it likely that international antitrust principles will move away from the optimal level of antitrust enforcement. First, I believe that bias may be less of a systemic problem than Professor Gunman's formal model suggests. Second, and more fundamentally, bias might compensate for other public choice problems of antitrust, paradoxically leading to more optimal competition laws. Third, even assuming that local enforcement is suboptimal, coordinating substantive international antitrust rules may sacrifice so much in experimentation and potential for beneficial transformation that the costs may outweigh the benefits.

B. Institutions and Localism

Professor Guzman is certainly correct in his claim that antitrust will sometimes be deployed systematically to consider only local interests and will be biased against foreign interests in some circumstances. (84) In these cases, exception divergences will result. For example, Congress has exempted export cartels from the reach of antitrust laws, (85) and may be influenced by the predominance of exporter interests in exempting certain sectors from antitrust altogether. Parochial bias strongly influences Congress and other democratic legislatures around the world. (86)

There remains the question of whether other nations can attack such cartels under their own laws, thus making it unlikely that domestic producers can take advantage of the foreign bias in their domestic laws. For instance, even if the United States authorizes a cartel, Europe can bring an antitrust complaint under its law if the cartelized products are shipped to Europe. Now it is true that some nations may not be able to gain jurisdiction over U.S. companies or are generally less sophisticated, but these would tend to be relatively small nations with consequently small losses to allocative efficiency. Others more familiar than I with jurisdictional and procedural issues in international antitrust have argued that there is no need for international agreement on substantive antitrust principles in this area because individual nations can challenge cartels. (87)

In contrast to such express discrimination, many antitrust laws are neutral on their face, particularly as between foreign and domestic producers. (88) When the laws themselves are neutral it is not clear that foreign interests will be neglected because the legal institutions charged with carrying them out may not be as subject to parochial bias as legislators. Let me illustrate this latter point with some examples from U.S. antitrust enforcement. First, assume that the United States has only consumers of a particular product. Will our antitrust system fail to take account of producer surplus of the foreign companies in deciding whether to challenge a merger? First, classifying the nationality of multinational corporations is often difficult. They often have shareholders and employees around the world and thus government decision makers of many countries often take substantial account of their interests. Second, advanced industrial democracies like the United States have developed formal rules and legal cultures that require decision makers to consider only those criteria that the law makes relevant. Thus, the judiciary, and to some extent the bureaucracy, will be constrained to take foreign interests into account so long as the law does not formally eliminate them from consideration.

In fact, the International Operations Guidelines issued by the U.S. Department of Justice and Federal Trade Commission make clear that unless the law otherwise requires, "[t]he agencies do not discriminate in the enforcement of the antitrust laws on the basis of the nationality of the parties." (89) Of course, it is always possible that decision makers will slight these interests, but it is not entirely clear what mechanism will guide them to do this. They have craft interests that militate against such blatant favoritism.

If the antitrust division disapproves a merger, the company has judicial recourse. (90) Judges, moreover, are even less likely to be swayed by such local bias. (91) To be sure, sometimes, because of timing issues, regulators can stop mergers before judges can review their decisions, but judicial review nevertheless offers some protection in many cases. Thus, there are a variety of institutional screens that in the context of the United States make bias less likely in cases assessing the anticompetitive conduct of foreign companies.

Now assume that the United States has only producers of a product and no consumers. Will U.S. antitrust enforcers fail to intervene to protect foreign consumer interests? This scenario may present more reason to worry. A recent act has suggested that the interest of foreign consumers need not be taken into account in most circumstances in initiating antitrust actions. (92) The Antitrust Enforcement Guidelines for International Operations confirm this Stance. (88) Even in the absence of such a direction against foreign bias, in initiating a prosecution, enforcement officials must consider budgetary constraints. In the context of case selection, foreign bias may be more likely to creep in under the guise of budgetary concerns where it is less obvious. Moreover, because there is no judicial review of an agency's failure to take action, the judiciary will not be a bulwark against bias.

The general lesson of this review is the importance of making our models institutionally rich. Only then can we assess the degree of damage that a diverse antitrust regime will inflict through the tendency toward bias against foreign interests.

C. Public Choice Considerations Countervailing to Foreign Bias

If one believes that public officials do not necessarily enact and enforce competition laws in the public interest, bias against foreign interests will not necessarily make antitrust laws depart from an economically optimal regime. (94) Indeed, such foreign bias may counteract the public choice-driven biases against wealth maximizing laws and thus move competition law toward a more optimal state.

For example, let us assume that, contrary to their claim that they treat producers of all nationalities equally, prosecutors in the United States give greater weight to producer interests when they are primarily domestic, and less weight to consumer interests when they are foreign than is justified by antitrust laws. In many instances, this reweighting may move us toward optimal antitrust enforcement because it will make prosecutors less likely to go after monopolies and other commercial practices in the marginal cases where the public choice factors described above may lead to overenforcement. (95)

One might suggest that bias against foreign interests would have the opposite effect--creating less optimal antitrust enforcement-when the United States has almost all the consumers and few of the producers of a given product. For example, in deciding whether to allow a merger initiated by a foreign corporation, administrators' tendency to disallow the merger of a foreign corporation because of bias toward overenforcement may be exacerbated by a bias against foreign producer interests. I have already suggested, however, that because of institutional constraints like an independent judiciary, foreign bias may have less effect in the merger context.

In any event, without a careful investigation of public choice and institutional factors, one cannot really measure the extent to which exception divergence moves us away from the optimal enforcement of neutral antitrust laws.

D. Distortion Divergence of Neutral Laws Because of Foreign Bias

In addition to specific exemptions from antitrust, some suggest that foreign bias may lead to distortion divergence leading a country to alter its laws from the unbiased laws which would have been in place under a closed economy. (96) While distortion divergence is a possibility based on an economic model, it is supported by less empirical evidence than exception divergence. In fact, there are some reasons to doubt that distortion divergence is more than a theoretical possibility. First, a nation's balance between production and consumption changes, as does the kind of goods it exports. It seems unlikely that a nation changes its antitrust law to follow changing patterns of consumption and production--the transactions costs are too high. (97) Second, antitrust law does not appear to track closely the predictions of the model. Developed nations that export imperfectly competitive goods tend to have more interventionist laws than developing nations, (98) although the theory suggests that the tendency would be the opposite. (99)

In any event, even if such tendencies exist, it is not clear whether they will make antitrust law better or worse. The optimality of Antitrust law depends on what other forces shape antitrust law. If, as I have suggested, developed nations tend to deploy excessively interventionist competition laws, and developing nations tend to deploy overly lenient competition laws, (100) trade is likely to make the laws better than they would be under a closed economy. For instance, because of their export of imperfectly competitive goods, the developed nations will adopt laws that are less suboptimally stringent than they otherwise would in a closed economy.

Thus, even before we look at the costs of choosing a set of international antitrust principles, it is not clear that the costs of parochialism are as high as might be assumed initially. The size of these costs is crucial because they must be weighed against the cost of coordination of substantive antitrust principles, which I have suggested may be quite high. First, the high agency costs may lead to less optimal enforcement than at the local level. (101) Second, the loss of experimentation and conflict may decrease the opportunities for beneficial transformation of competition laws. (102)


A. Introduction

Quite apart from consideration of antitrust law, the world trade regime contains a rationale and indeed already a structure for opposing discrimination in antitrust law. (103) In the absence of a prohibition, nations under the pressure of special interests are likely to increase the use of discriminatory competition laws as nontariff barriers, particularly as WTO agreements progressively reduce tariff barriers. (104) The prospect that discriminatory competition laws will be substituted for other trade barriers eliminated by the WTO, although never discussed in the debates over antitrust harmonization, is in my view the best reason for an antidiscrimination antitrust regime. (105) Such substitution furnishes a concrete reason to believe the impulse to foreign bias will become stronger as the WTO clears away other trade restrictions. The WTO already attempts to prohibit discriminatory regulations that interfere with the market access of exporters, providing the new regime with precedent for preventing similar discrimination in competition regulation.

I stress that my support for an antidiscrimination antitrust trust regime based on a trade rationale is nevertheless limited in scope, incremental in application, and sensitive to issues of institutional design. WTO enforcement of such rules should, at least at first, be limited to discrimination that creates nontariff barrier substitutes for tariff barriers such as antitrust rules that discriminate in market access. Second, until evidence develops that antitrust laws are applied in a discriminatory manner, the antidiscrimination rules should apply only to expressly discriminatory laws. Third, if the WTO thickens its antidiscrimination regime to include discriminatory application of antitrust laws, it should consider ways of assuring nondiscrimination through domestic institutions rather than increasing the power of WTO tribunals. For instance, the antidiscrimination regime might provide a safe harbor for nations that have administrative directives not to discriminate on the basis of nationality and an independent judiciary.

Finally, there appears to be evidence that antitrust issues were some of the so-called Singapore issues that contributed to the failure of the recent trade talks in Cancun. (l06) Although the modest antidiscrimination provisions that I would favor adding were almost certainly not at the center of the breakdown, I believe that these provisions are not nearly important enough to warrant endangering the talks which have much more important issues, like reducing agricultural subsidies.

In any event, an antidiscrimination approach has several advantages over substantive harmonization. It would preserve a diversity of antitrust approaches. Conflict on the substance of antitrust law could still occur, generating publicity and cutting through agency costs. Better norms could thus still evolve from the debate that diversity may engender. Another advantage of this system would be that neither its formulation nor enforcement would be subject to the peculiarly high agency costs and other public choice problems involved in choosing substantive rules.

I caution that I am sketching a regulatory ideal of an antidiscrimination antitrust regime. Many of the best features of such an ideal, such as excising antidumping laws applied to foreign producers when they are inconsistent with the antitrust laws applied to domestic producers, are unlikely to be realized in the short teem because of political obstacles in the Doha Round. (107) It is useful, however, to understand our baselines before the political compromises begin.

B. The Shape of a WTO Antidiscrimination Regime in Competition Law

With respect to the antidiscrimination model, the WTO offers precedent for the content of the model's rules because the WTO, with few exceptions, forbids its members to discriminate against other members in their regulatory regimes rather than imposing substantive restrictions on the content of regulation. Under a limited antidiscrimination regime, the WTO thus affords a better institutional match than United Nations Committee on Trade and Development, the Organization for International Cooperation or Development, or other forums that have been suggested for the development of an international competition structure. (108) Moreover, An antidiscrimination international antitrust regime will not threaten to transform the WTO into a regime with very high agency costs, because it will not set a precedent for substantive regulatory rulemaking.

To understand the way the WTO naturally generates an antidiscrimination model, it is useful to explain the basic political economy of the WTO regime. The WTO creates a structure moving toward freer world trade by creating a reciprocal relation between domestic tariff reductions and those in other countries, thereby changing the constellation of political interests at play in debates over tariff reductions. (109) The benefits that exporter groups gain from low tariffs abroad cause them to enter the political struggle against the protectionist interest groups at home, thus virtually representing the public's interest in wealth-enhancing free trade. (110) The WTO thus blunts the ability of protectionist interest groups to obstruct lower tariffs that would benefit the majority of citizens in their countries. (111) Under this view, the WTO is a regime that facilitates democratic choice within individual states even as it increases wealth by decreasing tariff barriers. (112)

The reduction of tariffs, however, does not make interest groups disappear. Rather, it merely causes them to adopt more subtle strategies of protectionism. They have a tendency to substitute nontariff barriers for the reduced tariffs so they can continue to earn rents. (113) An important kind of nontariff barrier is discriminatory regulation. Discriminatory health and safety regulation is well known, and much of the current WTO dispute settlement regime is Rimed at providing avenues for government to challenge such discriminatory regulation as part of the dispute settlement system. (114)

Quite apart from any interest in harmonization for the sake of the efficiency of the antitrust law, the WTO regime should begin to prohibit discriminatory antitrust laws that interfere with market access. Otherwise, protectionist interest groups will have a tendency to substitute such discriminatory rules and frustrate the effect of tariff reductions. This development can detract from the entire WTO regime because exporter groups will not have ex ante incentives to fight for lower tariffs in their home country if they recognize that their reciprocal gains abroad can be reduced by such a strategy. One example of discriminatory antitrust regulations that detract from market access is discriminatory restrictions on importing companies that prevent these companies from integrating vertically with domestic companies, if such mergers would facilitate competition against domestic companies. Other discriminatory practices might include refusing to apply for the benefit of foreign companies' rules against boycotts or a domestic requirement of providing competitors an "essential facility," (115) even when domestic companies receive the advantage of such rules.

As best construed, Article III of the General Agreement on Tariffs and Trade (GATT), (116) which embodies GATT's national treatment requirement, already provides such protections in some instances. But it would be immediately helpful to clarify that Article III applies to competition law. Moreover, it is clear from the case law that GAfFs Article III as written would not reach all discriminatory antitrust legislation that could interfere with market access. (117) In the long run the WTO could establish a separate agreement to make national treatment principles more comprehensive and concrete in the area of competition law, as previous GATT rounds have provided separate agreements that make national treatment principles more comprehensive and concrete in such contexts as the regulation of organisms and their products (118) and in technical regulation of products. (119)

Article III, paragraph 4, in particular, requires only that a member nation refrain from providing less favorable treatment to foreign "like products." (120) Given the "like products" element of the requirement, a nation conceivably could apply its competition law differently depending on the product at issue, so long as it applied the competition law equally to the same product. (121) If the new low cost provider of a product, for instance, were a foreign importer, a nation could permit boycotts and other anticompetitive processes to frustrate all new entrants, both domestic and foreign, on the grounds that the actual incidence of the anticompetitive process would fall mostly on the foreign imports. A WTO panel actually gave this limited but defensible reading to Article III, paragraph 4 when it held that Japan had not discriminated against U.S. photographic film producers because its competition law treated Japanese film producers in the same way. (122) Left unexplored in the opinion was whether Japan was applying its competition law in the photographic film industry as it does in other industries. (123) The problem of divergent application of antitrust laws driven by foreign bias is likely to become a bigger problem as the General Agreement on Trade in Services reduces other kinds of barriers in an increasing number of service sectors. (124)

If evidence develops that discriminatory application of antitrust laws does become a substantial problem, the WTO can require that antitrust laws be applied consistently at least within the same sectors of the economy where there is little justification for applying different kinds of antitrust laws. (125) There is precedent in the WTO for a broader requirement of legal consistency. For instance, the Sanitary and Phytosanitary Standards (SPS) Agreement--the GATT agreement that regulates food regulation--requires members to avoid "arbitrary or unjustifiable discrimination between members where the same conditions prevail or a disguised restriction on international trade." (126) The contemplated WTO competition agreement could deploy a similar consistency requirement, requiring nations not to vary their competition law arbitrarily depending on the products and services at issue at least when such products or services are within the same sector. As discussed below, such a requirement should be even easier to apply than the SPS consistency requirement. (127) To avoid requiring WTO dispute resolution process to engage in unnecessary scrutiny an agreement could create a strong presumption that a law is not discriminatory when nations have enforcement guidelines that do not distinguish among nationalities of corporations and when they have an independent judiciary to scrutinize the application of antitrust laws.

Of course, even a requirement of legal consistency will not assure market access if a nation like Japan refuses to punish anti-competitive behavior even when directed against its own citizens. (128) In that case, however, the nation must resist the application of an Antitrust regime that would benefit not only their consumers generally, but domestic producers entering the full range of their product markets. Thus, an invigorated consistency rule would make it harder for a nation to engage in covert discrimination in market access.

It is important to note one potential conflict between an antidiscrimination antitrust model and the WTO regime. (129) Currently, the WTO regime permits antidumping laws under certain conditions. (130) Antidumping laws generally permit tariffs to be raised on foreign imports if they are sold at a price lower than the price in their home country or lower than the cost of production. (131) In contrast, domestic antitrust often (as in the United States, for instance) permits suits against domestic companies only if their cost is lower than their cost of production and even then only under very defined circumstances. (132) More generally, a nation's antidumping regime is at war with the consumer welfare model of its antitrust regime, because antidumping remedies are intended to protect competitors rather than competition. (133) Thus, any regime that has consumer welfare as its goal in antitrust, but deploys antidumping laws, is potentially subject to attack under a serious antidiscrimination antitrust regime that seeks to guarantee market access. For political reasons, it may not be possible to subject antidumping laws to the discipline of an antitrust anti-discrimination model. Indeed, the difference between our antitrust regime and the antidumping regime offers a practical measure of the entrenched power of interest groups and xenophobia to inhibit wealth-creating free trade. Nevertheless, there would be no difficulty from the perspective of ideal free trade if an antitrust antidiscrimination regime eventually undermines GATT's permission for antidumping laws.

Finally, the WTO regime also is best interpreted to permit nations to apply their antitrust laws to activities in other jurisdictions so long as those activities affect that jurisdiction. As I have discussed elsewhere, permitting nations to regulate activities that affect their citizens, wherever they occur, is consistent with long-established legal norms and promotes more regulations that will be more responsive to the interests of citizens than those generated by a distant international bureaucracy. (134) In contrast, for some of the same reasons, nations generally should not be permitted to apply their antitrust law to practices whose effects are limited to foreign nations. Domestic authorities are not well positioned to assess the needs or views of foreign citizens, nor are they easily accountable for the effects of their decisions on foreigners. (135) Thus, the WTO rationale and practices would suggest that nations have the authority to apply their competition laws extraterritorially so long as their regulation meets an effects test.

C. The Constraints of Antidiscrimination Rules and the Minimization of Agency Costs

The public choice problems described in Part I will affect antidiscrimination rules, even Ambitious rules beyond the modest ones I recommend currently, far less than substantive rules because antidiscrimination rules remove most policy discretion from international decision makers. The WTO, for instance, already enforces antidiscrimination principles with respect to other regulations with a variety of procedure-oriented rules that eschew the kind of policy decisions that can be distorted by public choice factors, including bureaucratic interests. The WTO's antidiscrimination jurisprudence considers whether a nation's rules are transparent, are consistently applied with respect to foreign and domestic producers, And avoid regulatory processes that are unnecessary to advance the regulations' objective. (136)

For instance, the transparency requirement does not require substantive judgments and yet helps assure that both foreign and domestic producers face similar compliance costs for antitrust regulations. (137) Foreign firms may have greater difficulty complying with generally applicable regulations because domestic producers may better understand the opaque requirements of their own bureaucracy. (138) Transparency helps exporters compete on an equal basis by making regulatory compliance easier. (139)

Even transparent regulations, however, can have an unequal incidence on foreign and domestic producers or foreign and domestic consumers if they are applied unequally. (140) Thus, if evidence developed inconsistent application were a serious problem, the antidiscrimination regime would need to make sure the regulations are applied consistently. Evaluating consistency would not require a review of the substance of a nation's competition regulations, just a comparison of its treatment of foreign producers or consumers and its treatment of domestic producers or consumers.

This comparison in fact should likely be easier than similar comparisons undertaken by the WTO in evaluating health and safety regulations because a consistency evaluation in that context requires identifying the health and safety objectives served by a myriad of health and safety regulations. For example, in order to decide that an Australian regulation prohibiting the importation of salmon was discriminatory, the WTO compared this regulation with weaker regulations relating to domestic herring which it concluded posed higher risks. (141)

In the competition context, in contrast, nations tend to have somewhat more unified rules focusing on improving competition. A WTO consistency requirement would not address the difficult question of which regulation affecting domestic producers provides an appropriate comparison to the allegedly inconsistent regulation affecting foreign producers, but rather would address the simpler question of whether the nation's competition law was applied consistently across different products. The WTO Competition Code should also adopt a stance of substantial deference to enforcement decisions if those decisions were reviewed domestically by a judiciary genuinely independent of the executive and structured in such a way that it substantially eliminates the possibility of foreign bias. (142)

One final danger would be that nations might add procedural or substantive requirements that do not advance their legitimate antitrust objectives simply because those regulations impose greater costs on foreign producers. If evidence of this kind of problem developed the antidiscrimination regime could use precedent in the WTO to apply the least restrictive alternative requirement to eliminate provisions extraneous to competition objectives. The review would not be deeply substantive and thus subject to public choice pathologies, because it would not need to identify the competition objectives that the law served, but rather would only make sure that the regulations served those objectives and were not the cause of discrimination. (143)

D. Enforcement of WTO Antidiscrimination Competition Law Decisions

If antitrust harmonization took place within the WTO, another advantage of a nondiscrimination regime would be that such harmonization would not require construction of a new and untested institution, because tribunals are already familiar with applying a nondiscrimination jurisprudence. The dispute resolution system within the WTO also minimizes the enforcement problems discussed earlier. (144) It requires international review of domestic decisions, avoiding the parochial interpretation of its principles that will, over time, lead to divergence. (145) Yet WTO rulings are binding only as a matter of international law. (146) They have no direct effect, leaving the final decision to implement the ruling to member states. (147) They thus permit domestic systems of constitutional accountability to work and avoid the substantial costs of direct implementation of international rulings. (148)

Even without direct effect, the WTO nevertheless deploys a fairly effective method of gaining compliance with its rulings. If a nation does not comply with a ruling of the WTO, the offended nation generally can seek to withdraw concessions from the offending nation in value equal to the harm that the violation of WTO rules causes. (149) For instance, if a nation were found to have violated the Competition Code by failing to afford a foreign market entrant the protection that it gave to domestic market entrants, the offended foreign nation would be authorized to raise its tariffs on a product of the offending nation in the Amount of the value of the violation. The theory is that such sanctions will energize the exporters adversely affected by the sanctions to lobby their governments to comply with the WTO ruling. (150)

Given the institutional precedent of the WTO, dispute resolution under the WTO (or at least withdrawal of cencessions authorized by dispute resolution) should be available only for violations of the WTO Competition Code provisions that impair market access. Thus, for instance, decisions by nations designed to immunize export cartels from antitrust scrutiny would not be subject to dispute resolution. The reason is that extending dispute resolution to competition law that does not affect the market access of exporters would impose costs on exporters with no corresponding benefits.

As discussed above, the WTO structure relies on exporters to lobby for lower tariffs in their own countries in return for concessions abroad. (151) The prospect of future sanctions will reduce ex ante the value of these concessions and cause exporters to lobby less for the lowering of tariffs in their own countries. This cost is worth paying if, as I have argued above, enforcing a prohibition on discriminatory regulation is necessary to prevent nations from substituting discriminatory regulations--including discriminatory antitrust regulations that affect market access---for tariff barriers. (152) If the discriminatory antitrust regulations, however, do not affect market access, as when a nation permits export cartels at the expense of foreign consumers, the use of WTO sanctions to enforce a prohibition against such regulations will impose costs on a class of companies that does not receive benefits from the enforcement. As a result, expanding the withdrawal of concessions beyond market access issues would tend to reduce the expected value of the WTO regime to exporters and thus diminish their activism in support of the rounds of reciprocal tariff reductions that are at its heart. (153)

Even if the WTO should not provide full dispute resolution for violations of antidiscrimination norms that are unrelated to market access, other vehicles may be available for the WTO to monitor compliance with such norms. A competition committee could issue periodic reports And such reports could create pressure for nations to change their practices. (154) Because of the WTO's experience in applying antidiscrimination rules, as discussed above, it may be a useful forum for such reports. (155)

E. Objections

I address two possible objections to limiting international antitrust within the WTO to a nondiscrimination regime rather than adding the requirement, as the Doha Declaration appears to contemplate, that all WTO members also be held to certain "core principles of competition." (156) The first is that nondiscrimination principles are insufficient to force nations to treat foreign interests equally. Some countries, for example, may choose to have no Antitrust law because they want to help their producers at the expense of foreign consumers. Moreover, it may not be rational for a small country to have an antitrust law, because it will not possess the clout to apply its competition law extraterritorially to protect its own consumers. (157)

An international agreement could address this second point by requiring other nations to permit the extraterritorial application of another nation's laws, at least on the same antitrust theories deployed by the nation whose producers are the target of the antitrust enforcement. Thus, a nation would no longer have un international process obstacle to the enforcement of antitrust laws.

The WTO regime should not, however, require nations to choose some particular form of antitrust law. Small, undeveloped nations rationally may choose a low level of antitrust enforcement for reasons wholly related to their domestic situation rather than to any foreign bias. (158) First, given vigorous international trade, small, developing nations may best conclude that monopoly prices of both domestic and foreign products can be best constrained by foreign competition rather than domestic antitrust intervention. (159) The smaller the size of a nation's market, the easier it is for a foreign company to redirect part of its supply to discipline any super-competitive prices in that nation. (160) Thus, a small nation may rely on free trade rather than competition laws to promote consumer welfare. Second, if, as I have argued, the central question in antitrust is the comparative advantage of corrective market processes and government intervention,1el such nations may rationally take account of their relatively weak institutional competence in assessing the likelihood that government intervention will lead to better results than market processes. (162)

Finally, there are nonfrivolous arguments that, in a variety of situations, no antitrust enforcement will be superior to the antitrust enforcement of a particular government. (163) An advantage of a diversified system is that some countries can test this regime while others study its fruits. It is no answer to say that cartels are always bad, because distinguishing a horizontal agreement that constitutes an anticompetitive cartel from one that may have market benefits is a matter of debate even under U.S. antitrust law. (164) Thus, the antidiscrimination model of antitrust harmonization should oppose substantive harmonization even if it is limited to "core principles" or some other formulation that would impose a standard scope for some kinds of antitrust and inhibit innovative rules, including the absence of government intervention, in the area of horizontal agreements. (165)

Next, some may suggest that each WTO member should be required to enforce core antitrust principles on the analogy that the WTO Agreement on Trade Related Aspects of Intellectual Property Rights (TRIPs) requires WTO members to enforce core intellectual property principles. (166) The theoretical and practical reasons that call for enforcing substantive intellectual property principles through the WTO do not apply to substantive antitrust rules. First, certain nations have no incentives to provide any intellectual property laws at all. (167) For instance, consider a small developing nation with few, if any, inventors. (168) That nation has no incentives to provide patent protection, because it would conclude rationally that it is better to free ride on the inventions of others. (169) Indeed, even if a small developing nation had some inventors, it would have no incentives to provide patent protection because creators in their nations would reap sufficient rewards by obtaining patent protection abroad. (170)

The case concerning antitrust is different. All nations have consumers and thus all nations have incentives to have antitrust laws. These incentives may not be perfect and may be affected by the mix of imported and exported goods, but that is true of all regulations. Nevertheless, the WTO does not and should not require nations to have a floor of health, safety, labor or other regulations because of the public choice defects involved in international regulatory regimes. (171) The WTO requires only that these laws be nondiscriminatory--this is the proper analogy for a competition regime. This Part has shown how these antidiscrimination principles may be deepened in the context of competition law.

Second, TRIPs, unlike international antitrust standards, may have been so important to further tariff reductions that they needed to be included within the WTO. The intensely pragmatic reasons for the inclusion of TRIPs into the WTO go to the heart of the dynamic that fuels the WTO's reciprocal tariff reductions. The developing world most wanted tariff reductions in textiles and agriculture where it often holds a comparative advantage. The developing countries, however, could only succeed in obtaining these reductions with the help of the most powerful exporters in the developed nations. These exporters, such as film producers and pharmaceutical companies, had exports that made use of intellectual property. The exporters would not have been enthusiastic, however, about the prospect of increasing their exports to the developing world through reciprocal tariff reductions, if the developing world continued to expropriate what was, in their view, their intellectual property. Accordingly, the grand bargain at the heart of the Uruguay Round required nations, particularly developing nations, to adopt minimum standards for intellectual property in return for the agreement of developed nations to lower tariffs on textiles and agriculture. (172) I know of no evidence that key export sectors are demanding substantive harmonization of competition law as a price for making fundamental tariff reductions in the coming rounds of WTO talks.


Substantive antitrust harmonization is inadvisable because it has high agency costs and will reduce beneficial, long-run experimentation and innovation in antitrust. In contrast, an antidiscrimination antitrust model has fewer agency costs, particularly if it is limited in scope, because the antidiscrimination precedent of the WTO will provide a stable guide to the development of antidiscrimination rules. The antidiscrimination regime also will allow continued beneficial experimentation and innovation in competition law. Moreover, insofar as this regime helps police nations' denial of market access to importers, it will advance the goals of free trade by inhibiting nations from substituting discriminatory competition laws for tariff and other barriers that the WTO has already eliminated.

(1.) See, e.g., Sharon E. Foster, While America Slept: The Harmonization of Competition Laws Based Upon the European Union Model, 15 EMORY INT'L L. REV. 467 (2001); Eleanor M. Fox, Toward World Antitrust and Market Access, 91 AM. J. INT'L L. 1 (1997). For early works in this canon, see, for example, F. M. SCHERER, COMPETITION POLICIES FOR AN INTEGRATED WORLD ECONOMY (1994); Eleanor M. Fox, The End of Antitrust Isolationism: The Vision of One World, 1992 U. CHI. LEGAL F. 221.

(2.) See Spencer Weber Waller, The Internationalization of Antitrust Enforcement, 77 B.U.L. REV. 343, 385-86 (1997).


(4.) Paragraph 25 of the Doha Declaration addresses harmonization issues, and calls for considering the inclusion within the WTO of "core principles, including transparency, nondiscrimination and procedural fairness, and provisions ... [prohibiting] hardcore cartels; modalities for voluntary cooperation; and support for progressive reinforcement of competition institutions in developing countries through capacity building." See World Trade Organization, Ministerial Declaration, WT/MIN(01)/DEC/1 (Nov. 20, 2001), 41 I.L.M. 746 (2001) [hereinafter Doha Declaration].

(5.) See discussion infra Part I.B.2.

(6.) See discussion infra Part I.B.3.

(7.) See infra text accompanying notes 157-64.

(8.) See discussion infra Part III.C.

(9.) See infra Part III.B.

(10.) See infra notes 55-62 and accompanying text.

(11.) See infra note 65 and accompanying text.

(12.) See infra Part II.B.

(13.) See infra Part II.C.

(14.) See infra Part III.A.

(15.) See infra Part III.C.

(16.) See infra Part III.C.

(17.) See Waller, supra note 2, at 385-89 (noting burdens of multiple antitrust reviews on business).

(18.) See FINAL REPORT, supra note 3, at 52 (noting that the most restrictive nation prevails in merger review).

(19.) Cf. Harold Demsetz, Information and Efficiency: Another Viewpoint, 12 J.L.& ECON. 1 (1969) (discussing tendency of theorists to propose a new policy on the basis of a critique of the old policy without considering the defects in the innovative structures). Demote described the "nirvana approach" to public policy as considering "between an ideal norm and an existing 'imperfect' institutional arrangement." See id. at 1.

(20.) See Edward T. Swaine, The Local Law of Global Antitrust, 43 WM. & MARY L. REV. 627, 629 (2001).

(21.) See Daniel Steiner, The International Convergence of Competition Laws, 24 MANITOBA L.J. 577, 580 (1997) (explaining procedural harmonization).

(22.) See id. at 579-80(displaying in Table 1 examples of recent procedural harmonization).

(23.) For a brief summary of the Chicago School, Welfare Maximization Model, see Jessica L. Goldstein, Note, Single Firm Predatory Pricing in Antitrust Law: The Rose Acre Recoupment Test and the Search for an Appropriate Judicial Standard, 91 COLUM. L. REV. 1757, 1776 n.136 (1991).

(24.) Full illustration of the several public choice pathologies of antitrust are beyond the scope of this Article. Accordingly, I will discuss only a few of the more relevant pathologies.

(25.) "Rents" in this context refers to the antitrust officials' political influence. Even besides the influence of rent seeking, the mismatch between the short-run horizons of politics and the long-run horizons of markets may lead to excessive antitrust intervention. See Donald I. Baker, Government Enforcement of Section Two, 61 NOTRE DAME L. REV. 898, 926 (1986).

(26.) See Fred S. McChesney, Economics Versus Politics in Antitrust, 23 HARV. J.L. & PUB. POL'Y 133, 133-34 (1999) (giving examples of political influence on antitrust enforcement).

27. See id. at 140 (suggesting that increased antitrust enforcement brings favorable (publicity to antitrust agencies).

(28.) In the United States private actors can also enforce much of antitrust law through private causes of action. See 15 U.S.C. [section] 15 (2000). Nevertheless, private enforcement is unlikely to correct for public choice bias in favor of excessive antitrust intervention. First, executive regulators tend to shape the law with their decisions on antitrust, both because judges are more likely to defer to regulators' judgments and because much private litigation simply piggybacks on their decisions. See Spencer Weber Waller, Prosecution by Regulation: The Changing Nature of Antitrust Enforcement, 77 OR. L. REV. 1383 (1998) (explaining that the U.S. antitrust policy largely is defined by the enforcement agencies). Second, the structure of private litigation has its own structural difficulties. Since private litigation can be filed by competitors as well as consumers, much private litigation will attempt to advance theories that protect competitors rather than competition, and as such be excessively interventionist. See 15 U.S.C. [section] 15 (2000) ("[A]ny person who shall be injured in his business or property by a reason of anything forbidden in the antitrust laws ... shall recover threefold the damages by him sustained, and the cost of the suit, including a reasonable attorney's fee." (emphasis added)).

(29.) See Michael P. Kenny & William H. Jordan, United States v. Microsoft: Into the Antitrust Regulatory Vacuum Missteps the Department of Justice, 47 EMORYL.J. 1351, 1376-77 (1998) (detailing how Senator Orrin Hatch, who was the state Senator for many of Microsoft's competitors, tried to influence the Department of Justice to bring the Microsoft case).

(30.) It is for this reason that United States antitrust law employs a common law rather than code approach. See Standard Oil Co. v. United States, 221 U.S. 1, 59-60 (1911); see also Alan J. Meese, Price Theory, Competition, and the Rule of Reason, 2003 Ill. L. Rev. 77, 89-92 (explaining that Standard Oil in particular and Rule of Reason in general contemplate application of evolving economic theory in a manner that could produce changing treatment of various business practices). Christoper M. Barbuto, Note, Toward Convergence of Antitrust and Trade Law: An International Trade Analogue to Robinson-Patman, 62 FORDHAM L. REV. 2047, 2055 (1994) (describing the common law character of the Sherman Act, one of the primary U.S. antitrust laws).

(31.) See Waller, supra note 28, at 1394-95 (suggesting that antitrust regulators often set rules with minimum judicial involvement).

(32.) See Eric S. Hochstadt, Note, The Brown Shoe of European Union Competition Law, 24 CARDOZO L. REV. 287, 300-01 (2002) (suggesting that European regulators have substantial discretion because of lengthy appellate procedures for judicial review of their decisions).

(33.) See Waller, supra note 28, at 1421-23 (explaining why the judiciary has been reluctant to play a significant role in the shaping of antitrust law).

(34.) Harold Greene was even called the czar of telecommunications during the pendency of his oversight regarding the 1984 AT&T/Baby Bell restructuring. See Leslie Cauley, Telecom Czar Frets Over New Industry Rules, WALL ST. J., Feb. 12, 1996, at B1.

(35.) Rational ignorance describes the systematic tendency for citizens to pay little attention to political information. Anthony Downs developed the first theory of voters' "rational ignorance." See ANTHONY DOWNS, AN ECONOMIC THEORY OF DEMOCRACY (1957). The phenomenon occurs because acquiring information is both costly and unproductive. See DENNIS C. MUELLER, PUBLIC CHOICE II 205-06 (1989). It is costly to acquire such information because individuals must invest time that they could be using in other more productive enterprises. See id. It is unproductive because, although the principal instrumental use of such information is to guide voting, the vote of any one individual has little influence on the outcome of an election, See id.

(36.) For a discussion of why cognitive ignorance exists on complex matters and how cognitive ignorance increases agency costs, see John O. McGinnis, The Partial Republican, 35 WM. & MARY L. REV. 1751, 1792 (1994) (book review).

(37.) See Frank H. Easterbrook, The Limits of Antitrust, 63 TEX. L. REV. 1, 2-3 (1984).

(38.) Id.

(39.) The "representativeness" heuristic tends to make people overconfidently extrapolate predicted characteristics of a class from a small sample size of which they happen to be aware. See Amos Tversky & Daniel Kahneman, Belief in the Law of Small Numbers, in JUDGMENT UNDER UNCERTAINTY: HEURISTICS AND BIASES 23, 24 (Daniel Kahneman et al. eds., 1982). If the sample consists of events rather than objects, the heuristic should tend to make people extrapolate uncertain future events from events of which they are aware in a similarly irrational manner. See id. Thus, individuals will tend to think that future events will resemble past events more than probability warrants. See Kenneth J. Arrow, Risk Perception in Psychology and Economics, 20 ECON. INQUIRY 1, 5 (1982). For an important present day application, see ROBERT J. SHILLER, IRRATIONAL EXUBERANCE 137 (2000) (using work on "representativeness" heuristic to suggest that people will think that today's stock market patterns will predict the stock market patterns of tomorrow).

(40.) John O. McGinnis, The Political Economy of Global Multilateralism, 1 CHI. J. INT'L L. 381, 393 (2000).

(41.) See id.

(42.) See Paul B. Stephan, Accountability and International Lawmaking: Rules, Rents and Legitimacy, 17 NW. J. INT'L L. & BUS. 681,699 (1997) (noting that "international lawmaking tends to involve somewhat greater secrecy than does domestic lawmaking").

(43.) Cf. Frank H. Easterbrook, The State of Madison's Vision of the State: A Public Choice Perspective, 107 HARV. L. REV. 1328, 1337-38 (1994) (noting that interest groups can obtain greater gains by influencing national, as opposed to local, legislation).

(44.) For discussion of the rise of a network of international regulators to enforce such matters, see Anne-Marie Slaughter, The Real New World Order, FOREIGN AFF., Sept.-Oct. 1997, at 183, 189-90. I believe that Professor Slaughter underestimates the public choice costs of such networks, because those regulators will be interested in their own mission and status and will thus not take account of the public interest.

(45.) See supra notes 42-44 and accompanying text.

(46.) U.S. CONST. art. II, [section] 2.

(47.) U.S. CONST. art. III.

(48.) See Jim C. Chen, Appointments With Disaster: The Unconstitutionality of Binational Arbitral Review Under the United States-Canada Free Trade Agreement, 49 WASH. & LEE L. REV. 1455, 1456-57 (1992) (explaining that direct enforcement of the United States-Canada Free Trade Agreement by international arbitrators violates Article II's Appointments Clause because of the procedures used to select international arbitrators and also violates Article III because federal judges would be replaced by international arbitrators).

(49.) See Hochstadt, supra note 32, at 322.

(50.) See Spencer Weber Waller, Neo-Realism and the International Harmonization of Law: Lessons from Antitrust, 42 U. KAN. L. REV. 557, 562-69 (1994) (discussing general theories on the transferability of laws between diverse nations and implying that harmonized rules enforced nationally rather than internationally will tend to diverge because of different conceptions and values in different nations).

(51.) See supra note 19 and accompanying text.

(52.) See ROBERT H. BORK, THE ANTITRUST PARADOX 427 (1993) (suggesting that recent American courts adopted economic efficiency as the touchstone of antitrust).

(53.) Andrew T. Guzman, Essay, Antitrust and International Regulatory Federalism, 76 N.Y.U.L. REV. 1142, 1149 (2001) ("Regardless of the location of a firm and its activities, many jurisdictions--most notably the United States and the European Union--apply their own laws to actions that have a local affect.").

(54.) Id. at 1148.

(55.) See John F. Duffy, Harmony and Diversity in Global Patent Law, 17 BERKELEY TECH. L.J. 685, 707-09 (2002) (positing that global harmonization may restrain innovation). These information flows conduce to a kind of jurisdictional competition, as Professors Kerber and Budzinski show. See Wolfgang Kerber & Oliver Budzinski, Competition of Competition Laws: Mission Impossible?, in THE NEW ANTITRUST PARADOX: POLICY PROLIFERATION IN THE GLOBAL ECONOMY (M.S. Greve & R.A. Epstein eds., forthcoming 2003).

(56.) See Hochstadt, supra note 32, at 327-76 (critiquing the EU decision).

(57.) See, e.g., Press Release, EU Institutions, Speech by Mr. Marie Monti, European Competition Commissioner Review of the EC Merger Regulation--Roadmap for the reform Project, Conference on Reform of European Merger Control, British Chamber of Commerce--Brussels, June 4, 2002, available at (acknowledging criticisms of European merger review process in setting forth reform proposals).

(58.) One might argue that European law can be criticized even in the absence of conflict. Of course, it is true that academics and others can critique antitrust law at any point, but the question is whether anyone will listen. Political theorists since Machiavelli have noted that conflicts between officials on fundamental points can engage the public's attention. See NICCOLO MACHIAVELLI, DISCOURSES ON THE FIRST TEN BOOKS OF TITUS LIVIUS, reprinted in 2 THE HISTORICAL, POLITICAL, AND DIPLOMATIC WRITINGS OF NICCOLO MACHIAVELLI 98-102 (Christian E. Detmold trans., University Microfilms Int'l 1978).

(59.) Professor Alan Msese has noted the many instances in which U.S. regulators have failed to follow the best economic theory. See Meese, supra note 30, at 130-131 (recounting the United States' position in TOPCO, including assertion that government control of entry would be superior); Alan J. Meese, Farewell to the Quick Look: Reconstructing the Scope and Content of the Rule of Reason, 68 ANTITRUST L.J. 461, 486 (2000) (explaining that government simply ignored defendants' explanation of the virtues of the restraints involved in TOPCO); see also Alan J. Meese, Rasing Rivals Costs: Can the Agencies Do More Harm than Good?, 11 GEO. MASON L. REV. (forthcoming 2003) (detailing government theories on tying and monopolization that are contrary to sound economic understanding); Alan J. Meese, Don't Disintegrate Microsoft (Yet), 9 GEO. MASON L. REV. 761, 786-790 (2001) (arguing that standards sought by the United States were outmoded in light of advances in economic theory from transaction cost economics); Timothy J. Muris, The FTC and the Law of Monopolization, 67 ANTITRUST L.J. 693 (2000) (arguing that the FTC's approach to monopolization law is biased unduly toward the government).

(60.) See Christopher R. Leslie, Achieving Efficiency Through Collusion: A Market Failure Defense to Horizontal Price-Fixing, 81 CAL. L. REV. 243, 251-52 (1993) (explaining that "the Warren Court placed emphasis on social and political factors, rather than efficiency").

(61.) Cf. GEOFFREY BLAINEY, A SHORT HISTORY OF THE WORLD 400-01 (2002) (suggesting that the United States is an empire by virtue of the power of its ideas rather than its military conquests).

(62.) See Easterbrook, supra note 37, at 2 (suggesting that the scope of antitrust should be limited to instances in which it has a comparative advantage over market-correcting forces).

(63.) See A.E. Rodriguez & Mark D. Williams, The Effectiveness of Proposed Antitrust Programs for Developing Countries, 19 N.C.J. INT'L L. & COM. REG. 209, 218 (1994) (noting that competition from imports "is a countervailing force against whatever power domestic firms may have to raise prices above the competitive and socially efficient level").

(64.) Cf. William J. Alceves, Institutionalist Theory and International Legal Scholarship, 12 AM. U. J. INT'L. & POL'Y 227, 244 (1997) (detailing transactions costs of international agreements).

(65.) See supra Part I.B.1.

(66.) Even those who understand the political economy of antitrust differently from the view advocated here must still identify the assumptions underlying their predictions of government actors' behavior and show why international harmonization would be superior to national regimes under these assumptions.

(67.) Cf. William H. Page, Antitrust Review of Mergers in Transition Economics: A Comment, with Some Lessons from Brazil, 66 U. CIN. L. REV. 1113, 1115 (1998).

(68.) See Julian Epstein, The Other Side of Harmony: Can Trade and Competition Law Work Together in the International Marketplace?, 17 AM. U. INT'L L. REV. 343, 360-61 (2002) (discussing the lack of antitrust law in developing nations).

(69.) See William E. Kovacic, Merger Enforcement in Transition: Antitrust Controls on Acquisitions in Emerging Economies, 66 U. CIN. L. REV. 1075, 1092 (1998).

(70.) See id. at 1093-94.

(71.) See Andrew T. Guzman, Is International Antitrust Possible?, 73 N.Y.U. L. REV. 1501, 1512-21 (1998).

(72.) See id.

(73.) The traditional arguments in favor of a centralized regime are: (1) decentralized regimes raise transactions costs; and (2) because in decentralized regimes the most restrictive regime governs, decentralized regimes result in suboptimal antitrust enforcement. See supra Part I.

(74.) Federal antitrust enforcement has not eliminated this concern, as shown by the suggestion that Senator Hatch, who came from the state of a number of Microsoft's competitors, influenced the initiation of the Microsoft litigation. See supra note 29 and accompanying text.

(75.) See Guzman, supra note 71, at 1512-21.

(76.) See id.

(77.) See id. at 1512-15.

(78.) Id. Professor Guzman's general result is: Assuming that nations apply their antitrust laws extraterritorially, the general result is that
   [a] country whose firms are responsible for x% of global production
   will take into account x% of the change in global producer surplus
   generated by a particular activity. A country whose consumers
   account for y% of global consumption will take into account y% of
   the total change in global consumer surplus generated by the

Id. at 1520.

(79.) See Andrew T. Guzman, International Antitrust and the WTO: The Lesson from Intellectual Property, 43 VA. J. INT'L L. 933, 936 (2003).

(80.) Id. at 943.

(81.) Id. at 947.

(82.) Id. at 943.

(83.) See id. at 947.

(84.) See supra notes 75-83 and accompanying text.

(85.) See, e.g., Webb-Pomerene Act, 15 U.S.C. [sub section] 16-65 (2000).

(86.) See Robert Sheppard, Towards a UN World Parliament: UN Reform for the Progressive Evolution of an Elective and Accountable Democratic Parliamentary Process in UN Governance in the New Millenium, 1 ASIAN-PAC. L. & POL'Y J. 4 (2000). That foreign bias can affect antitrust laws should come as no surprise to public choice analysis. Legislators do not gain by promoting the interests of foreigners, because foreigners do not vote and foreigners are often prohibited from providing material aid to politicians, such as through campaign contributions. See, e.g., 2 U.S.C. [section] 441(e) (2000). For similar reasons, constitutions that attempt to discipline legislators through rules encouraging neutrality among their own regions are likely to be indifferent, even antagonistic, to foreign interests.

(87.) See Harry First, The Vitamins Case: Cartel Prosecutions and the Coming of International Competition Law, 68 ANTITRUST L.J. 711 (2001). First notes, for instance, that in 1999, the United States alone had fifty active grand jury investigations against cartels by foreign producers. See id. at 711. He also points out that the United States prosecuted all sorts of foreign cartels during the 1940s. Id. at 720-32. He concludes powerfully:
   Consider these rhetorical questions: Is there any careful antitrust
   lawyer today who would advise a client that there is no risk in
   forming a cartel to restrict exports to a "foreign" jurisdiction
   that has an antitrust regime? Is there any such lawyer who would
   advise a client that even if one jurisdiction grants a statutory
   immunity for anticompetitive export behavior, that the immunity-will
   be a defense if the "inbound" jurisdiction decides to prosecute? Is
   there any such lawyer who would advise the officers or directors of
   a client corporation not to worry personally about participating in
   a cartel that affects prices in a foreign country?.

Id. at 726-27. I am indebted to Alan Meese for discussions on this point.

(88.) See, e.g., Sherman Act, 15 U.S.C. [subsection] 1-7 (2000); Clayton Act, 15 U.S.C. [subsection] 13-15 (2000). It is interesting to speculate why Congress does not make such parochial distinctions between foreign and domestic companies in merger laws. If Congress did shape our laws to be so blatantly discriminatory, the laws would invite retaliation against U.S. companies by foreign interests. U.S. companies thus might lobby against such discrimination. In contrast, U.S. consumers are a diffuse group and are not in a position to prevent discrimination against foreign consumers even if it were to encourage discrimination against U.S. consumers in foreign competition laws. U.S. law, however, does seem to discriminate against foreign consumer interests. See supra notes 85-86 and accompanying text.



(91.) Cf. Kevin Claremont & Theodore Eisenberg, Xenophilia in American Courts, 109 HARV. L. REV. 1120, 1132 (1996) (suggesting on basis of study that American courts did not harbor bias against foreigners).

(92.) See Foreign Trade Antitrust Improvements Act, 15 U.S.C. [section] 6(a) (2000).

(93.) See GUIDELINES FOR INTERNATIONAL OPERATIONS, supra note 89 (requiring an effect on U.S. commerce or U.S. producers to assert U.S. antitrust jurisdiction).

(94.) See Guzman, supra note 71, at 1531 (acknowledging possible interaction of distortions driven by foreign bias and those driven by public choice considerations).

(95.) See supra notes 27-36 and accompanying text.

(96.) See supra notes 84-88 and accompanying text.

(97.) Cf. Charles R.B. Macedo, Note, First-to-File: Is American Adoption of the International Standard in Patent Law Worth the Price?, 1988 COLUM. BUS. L. REV. 543, 586 (noting the high transactions costs of changing U.S. patent law).

(98.) See supra note 68 and accompanying text.

(99.) See supra notes 80-81 and accompanying text.

(100.) See supra notes 67-70 and accompanying text.

(101.) See supra Part I.B.2.

(102.) See supra Part I.B.3.

(103.) The proposal in this Article is very similar to one contained in a paper presented at the same American Enterprise Institute Conference where I first presented this paper. See Michael Trebilcock & Edward Iacobucci, National Treatment and Extraterritoriality: Defining the Domains of Trade and Antitrust Policy (forthcoming 2003) (manuscript at 2), available at The main difference is my conclusion that only discriminatory rules that interfere with market access should be subject to dispute resolution. See infra note 152 and accompanying text.

(104.) See John O. McGinnis & Mark L. Movsesian, The World Trade Constitution, 114 HARV. L. REV. 511, 549-52 (2000).

(105.) This new rationale for a limited antidiscrimination model in antitrust may respond to some of the objections thoughtful scholars have been lodging against even this modest form of international antitrust. For instance, as this Article was in press, Professor Edward Swaine noted that an antidiscrimination regime would still have to permit nations to enter into bilateral deals for information exchange and other cooperation, thus weakening the force of the nondiscrimination principle. See Edward A. Swaine, Against Principled Antitrust, 43 VA. J. INT'L L. 959, 970 (2003). These deals would not appear to detract, however, from the overriding purpose of an antidiscrimination model offered here because they would be unlikely to raise substantial substitute barriers against goods from any nation.

(106.) See Jeffrey Schott, Unlocking the Benefits of World Trade, THE ECONOMIST, Nov. 1, 2003, at 65.

(107.) For a discussion of the various political obstacles facing an international antitrust regime in the Doha Round, see Kal Raustliala, The Architecture of International Cooperation: Transgovernmental Networks and the Future of International Law, 43 VA. J. INT'L. 1, 37 (2002).

(108.) See Consumer Policy and Multilateral Competition Frameworks Discussion Paper (copy on file with author).

(109.) See JOHN H. JACKSON, THE WORLD TRADING SYSTEM: LAW AND POLICY OF INTERNATIONAL ECONOMIC RELATIONS 157-74 (2d ed. 2002) (discussing the reciprocal nature of the Most Favored Nation obligation of the WTO).

(110.) See McGinnis & Movsesian, supra note 104, at 545 (noting that "producere that enjoy a comparative advantage gain new markets when foreign countries reduce tariffs ... [which] creates incentives for such producers to lobby for lower tariffs in their own countries").

(111.) Id. at 546.

(112.) See id.

(113.) For a general discussion of the manner in which concentrated interest groups will attempt to substitute one kind of rent seeking for another kind that is blocked by the political system, see John O. McGinnis & Michael B. Rappaport, Supermajority Rules as a Constitutional Solution, 40 WM. & MARY L. REV. 365, 428-29 (1999).

(114.) See McGinnis & Movsesian, supra note 104, at 566-72 (explaining the antidiscrimination model of the WTO).

(115.) For a discussion of "essential facilities" doctrine, see Epstein, supra note 68, at 367 (advocating the WTO's adoption of the doctrine).

(116.) General Agreement on Tarriffs and Trade art. III, Oct. 30, 1947, 61 Stat. A-ll, 55 U.N.T.S. 187 [hereinafter GATT.

(117.) See infra notes 120-24 and accompanying text.

(118.) Agreement on the Application of Sanitary and Phytosanitary Measures, Apr. 15,

(119.) Marrakesh Agreement Establishing the World Trade Organization [hereinafter WTO Agreement], Annex 1A, LEGAL INSTRUMENTS--RESULTS OF THE URUGUAY ROUND vol. 27 (1994), available at http'J/ [hereinafter SPS Agreement].

(119.) Agreement on Technical Barriers to Trade, Apr. 15, 1994, WTO Agreement, Annex 1A, LEGAL INSTRUMENTS---RESULTS OF THX URUGUAY ROUND vol. 27 (1994), available at

(120.) GATT art. III. Article III, paragraph 4 of GATT provides:
   The products of the territory of any contracting party imported into
   the territory of any other contracting party shall be accorded
   treatment no less favourable than that accorded to like products of
   national origin in respect of all laws, regulations and requirements
   affecting their internal sale, offering for sale, purchase,
   transportation, distribution or use. The provisions of this
   paragraph shall not prevent the application of differential internal
   transportation charges which are based exclusively on the economic
   operation of the means of transport and not on the nationality of
   the product.


(121.) See id.

(122.) See Japan-Measures Affecting Consumer Photographic Film and Paper: Report of the Panel, WT/DS44/R, [paragraph][paragraph] 10.378-10.382 (Mar. 31, 1998) [hereinafter Japanese Film Report].

(123.) See id.

(124.) See generally General Agreement on Trade in Services, Apr. 15, 1994, WTO Agreement, Annex 1B, LEGAL INSTRUMENTS--RESULTS OF THE URUGUAY ROUND vol. 28-30, 33 I.L.M. 1167 (1994).

(125.) The Doha proposals would permit nations to have different antitrust rules for different sectors. See, e.g., Working Group on the Interaction Between Trade and Competition policy, Including Transparency, Non-Discrimination and Procedural Fairness: Background Note by the Secretariat, WT/WBTCP/W/209, at 35 (Sept. 19, 2002), available at

(126.) SPS Agreement, supra note 118.

(127.) See infra Part III.C.

(128.) See supra note 125 and accompanying text.

(129.) See generally Wesley A. Cann, Jr., Internationalizing Our Views Toward Recoupment and Market Power: Attacking the Antidumping/Antitrust Dichotomy Through WTO-Consistent Global Welfare Theory, 17 U. PA. J. INT'L ECON. L. 69 (1996) (describing tension between antitrust and antidumping law).

(130.) See GATT art. VI (permitting GATT members to raise tariffs in response to dumping by importing nations).

(131.) See, e.g., 19 U.S.C. [section] 1673 (2000).

(132.) See Lan Cao, Toward a New Sensibility for International Economic Development, 32 TEX. INT'L L.J. 209, 219-21 (1997).

(133.) For discussion of this tension, see generally Richard D. Boltuck & Seth Kaplan, Conflicting Entitlements: Can Antidumping and Antitrust Regulation Be Reconciled?, 61 U. CIN. L. REV. 903 (1993).

(134.) See McGinnis & Movsesian, supra note 104, at 574 (noting that transparency requirements in international law facilitate representative democracy).

(135.) See id.

(136.) For a fuller discussion of this jurisprudence, see id, at 572-83.

(137.) Id. at 574.

(138.) Id.

(139.) Id.; see also Edward A. Laing, Equal Access / Non-Discrimination and Legitimate Discrimination in International Economic Law, 14 WIS. INT'L. L.J. 246, 296-97 (1996).

(140.) See McGinnis & Movsesian, supra note 104, at 575 (discussing unequal application as an "opportunity for disguised protectionism").

(141.) Australia-Measures Affecting Importation of Salmon: Report to the Appellate Body, WT/DS18/AB/R, [paragraphs] 154-58 (Oct. 20, 1998), available at http://www.wto/org/english/tratop_e /dispu_subjects_index_e.htm.

(142.) See supra notes 90-91 and accompanying text (noting that, in the United States, review by an independent judiciary mitigates concerns regarding antitrust enforcement bias).

(143.) See Alan O. Sykes, Regulatory Protectionism and the Law of International Trade, 66 U. CHI. L. REV. 1, 25 (1999) (noting that the least restrictive means test allows governments to pursue their chosen objectives).

(144.) See supra Part I.B.2.

(145.) For problems of national interpretation of international principles, see generally supra note 50 and accompanying text.


(147.) See id.

(148.) For problems involved in loss of constitutional accountability, see supra notes 46-48 and accompanying text.

(149.) Agreement on Subsidies and Countervailing Measures, WTO Agreement art. 4.10, 7.8, LEGAL INSTRUMENTS--RESULTS OF THE URUGUAY ROUND vol. 27 (1994), available at; see also Understanding on Rules and Procedures Governing Settlement of Disputes (1994), WTO Agreement, Annex 2, art. 22.4, LEGAL INSTRUMENTS--RESULTS OF THE URUGUAY ROUND vol. 31, 33 I.L.M. 1226 (1994), available at (providing that the WTO can authorize an offended member to either receive compensation or retaliate by withdrawing concessions and that the retaliation must be "equivalent" to the loss the offended member has sustained).

(150.) See Mark L. Movsesian, Enforcement of WTO Ruling: An Interest Group Analysis, 32 HOFSTRA L. REV. (forthcoming 2003).

(151.) See supra notes 110-12 and accompanying text.

(152.) See supra Part III.A.

(153.) It might be argued that some exporters already expect that they will not be net beneficiaries from the dispute resolution process, because they will know from their position that they will be a likely target of sanctions. I agree that because exporters do not operate under a perfect veil of ignorance with respect to the likelihood of sanctions, some might, on net, prefer no dispute resolution regime. That is no reason, however, to add previsions to the dispute resolution regime that will not help any exporters. That burden is likely to weaken support among all exporters who are keys to the dynamic that powers the GATT.

Others suggest that some other provisions of the GATT, such as antidumping previsions, do not help exporters at all. It is true that exporters would be better off generally without any antidumping provisions. The dispute resolution system, however, enforces GATT's provisions that limit the scope of antidumping provisions and these limitations do serve the interests of exporters.

(154.) The WTO has created more informal structures, such as the Trade Policy Review Mechanism, which encourages nations to improve their laws from a trade perspective. Trade Policy Review Mechanism, WTO Agreement, Annex 3, LEGAL INSTRUMENTS--RESULTS OF THE URUGUAY ROUND vol. 31 (1994), available at 29-tprm_e.htm (last visited Aug. 28, 2003).

(155.) See supra Part III.C.

(156.) For a description of the Doha Declaration, see supra note 4 and accompanying text.

(157.) See Guzman, supra note 71, at 1539-40.

(158.) See id.

(159.) See id.

(160.) See Rodriguez & Williams, supra note 63, at 218.

(161.) See supra note 37 and accompanying text.

(162.) Conversely, antitrust scrutiny might be more stringent in countries that have government-created barriers to entry than in countries that facilitate entry. Antitrust also might be more stringent in countries with poorly developed capital markets, where starting a new, large business is difficult or in countries with poorly developed corporate law, since the absence of such law might make it more difficult for new firms to enter. Mark Roe similarly suggests that the content of a country's corporate law and nature of relationships between shareholders and managers can depend upon the nature of other institutions. MARK J, ROE, STRONG MANAGERS, WEAK OWNERS: THE POLITICAL ROOTS OF AMERICAN CORPORATE FINANCE 223 (1994). I am grateful to Alan Meese for drawing my attention to the analogy to Roe.

(163.) See generally Lino A. Graglia, Is Antitrust Obsolete?, 23 HARV. J.L. & PUB. POL'Y 11, 21 (1999) (arguing that in many instances antitrust law "does more harm than good").

(164.) For instance, in BMI v. CBS, 441 U.S. I (1979), the Supreme Court held that price fixing ancillary to the creation of a blanket license of copyrighted works should be under the Rule of Reason, even though it was literally price fixing. See also United States v. Brown Univ., 5 F.3d 658 (3d Cir. 1993) (price fixing by Ivy league schools analyzed under the Rule of Reason); Rothery Shortage v. Atlas Van Lines, Inc., 792 F.2d 210 (D.C. Cir. 1986) (price fixing ancillary to a joint venture analyzed under the Rule of Reason).

(165.) See supra note 4 and accompanying text (discussing Doha agenda that contemplates enforcement of core antitrust principles).

(166.) For discussion of the enforcement provisions of TRIPs, see WATAL, supra note 146, at 49.

(167.) See Duffy, supra note 55, at 698-99 (explaining the international externalities of intellectual property systems that justify TRIPs).

(168.) This example was developed originally by John F. Duffy. See id.

(169.) See id. at 698.

(170.) Id.

(171.) See McGinnis & Movsesian, supra note 104, at 552-53.

(172.) See, e.g., WATAL, supra note 146, at 9-47 (discussing this bargain).

JOHN O. MCGINNIS, Professor, Northwestern University Law School. I want to thank Richard Epstein, Eric Hochstadt, Nelson Lund, Fred McChesney, Alan Meese, Mark Movsesian, and Spencer Waller, as well as participants in workshops at DePaul and the University of Alabama Law Schools and at the American Enterprise Institute Conference on International Antitrust. An abbreviated and thus somewhat less nuanced version of this piece will appear in THE NEW ANTITRUST PARADOX (Michael Greve & Richard Epstein eds., 2003). The support of the American Enterprise Institute made this project possible.
COPYRIGHT 2003 College of William and Mary, Marshall Wythe School of Law
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2003 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:McGinnis, John O.
Publication:William and Mary Law Review
Date:Dec 1, 2003
Previous Article:Patent first, ask questions later: morality and biotechnology in patent law.
Next Article:Voter knowledge and constitutional change: assessing the New Deal experience.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters