TWENTY-SEVEN YEARS AGO, I TOOK MY law school antitrust course from a new assistant professor who had just left the Federal Trade Commission. My performance was adequate but not stellar.
In retrospect, I think my underperformance was because I was a soon-to-be economist as well as a budding lawyer. I approached antitrust with a presumption that the gears of industrial organization economics and antitrust law meshed more or less synchronously. But such a presumption was scarcely warranted; "competition" law was often sand, not grease, in the gears of competition. I went into the final exam conflicted, and evidently it showed.
Today, the conflicts are fewer, thanks to a new brand of antitrust thinking that has developed and a new breed of judges and antitrust enforcers who have arrived over the past generation. However, those conflicts could grow more numerous in the future as state attorneys general and European regulators and lawyers become increasingly active in the antitrust arena.
ANTITRUST LAW AND ANTITRUST ECONOMICS
The past generation of antitrust has witnessed much intellectual competition among economists and competition-minded jurists as to what that body of law is supposed to do. Section I of the Sherman Act outlaws "every contract, combination ... or conspiracy" that is "in restraint of trade," but it defines none of those terms. Likewise, Section 2 makes it illegal to "monopolize" (or attempt to "monopolize"), but does not define that term, either. The operative language of the other two important antitrust statutes, the Clayton Act and the Federal Trade Commission Act, is equally bare-bones.
Courts, lawyers, and economists were left to flesh out what would be deemed anticompetitive. At sea for the most part, judges initially used "per se" methodologies--declaring a practice illegal because it constituted a contract as described in the antitrust laws, regardless of whether the practice restrained or enhanced competition. As a result, courts declared per se illegal many contracts among horizontal competitors and many agreements among vertical contractors, particularly price agreements among vertically linked parties.
UNIFIED APPROACH Traditionally in antitrust, each sort of "contract, combination, or conspiracy" or allegedly "monopolizing" practice has been treated as requiring a separate mode of analysis. A separate body of case law specific to each contract or practice evolved, rather than a single system based on more fundamental notions of competition.
Increasingly, however, disparate strands of antitrust law have coalesced to ignore this needless taxonomy. Justice Sandra Day O'Connor fired an important shot across the bow with her 1984 concurrence in the tying case of Jefferson Parish Hospital District v. Hyde. Particularized rules for each sort of contract or practice coming under the antitrust lens made no sense, she wrote. Instead, she called for a unified approach to antitrust analysis based on a common rule-of-reason approach that compares a practice's economic benefits and costs. According to Justice O'Connor,
The time has therefore come to abandon the "per se" label and refocus the inquiry on the adverse economic effects, and the potential economic benefits, that the tie may have. The law of tie-ins will thus be brought into accord with the law applicable to all other allegedly anticompetitive economic arrangements.
Lower courts have leapt at the invitation to combine antitrust's disjointed jurisprudence into a single analytic model. For example, in its (hopefully) final opinion in United States v. Microsoft, the D.C. Circuit Court of Appeals interpreted the standards that should be applied under Sections 1 and 2 of the Sherman Act as a single test. The court said that regardless of statutory origin, an antitrust challenge should be evaluated by competitive costs and benefits of the challenged practice, noting that other circuits had concluded the same thing in other cases.
The growing unification of antitrust standards has resolved issues that have plagued antitrust for years. First, the law has finally ended any intellectual competition about what the goals of antitrust are. In particular, the debate about whether antitrust is to pursue economic or social goals is over--economics has won. "Anticompetitive" now is clearly defined as that which raises price, restricts quantity, or lowers quality. Social goals such as maintaining large numbers of smaller, less efficient firms in the market have largely been repudiated. The "Chicago school" focus on price, quantity, and quality is now accepted as the norm in antitrust.
A second development characterizing the past generation of antitrust is the demise of per se rules of illegality and the concomitant rise of the rule of reason as the dominant mode of antitrust analysis. The seemingly impregnable fortress of per se rules erected during the years of Justice William O. Douglas's pivotal antitrust role on the Court crumbled quickly after his departure. True, Justice Douglas's views are still represented on the Court, at least to some extent, by Justice John Paul Stevens. And not all areas of per se treatment have been eradicated. But lower courts have found ways to minimize the harm done by mistaken per se rules. Indeed, competition from lower courts has sped the demise of bad antitrust and the rise of better antitrust.
Starting in the mid-1970s, faced with per se Supreme Court pronouncements that made no sense from the standpoint of competition, American lower-court judges fought back in various ways. Some lower courts have respected the judicial hierarchy, applying misguided Supreme Court precedents but imploring the Court to reverse them. Perhaps the most famous example is Judge Richard A. Posner's 1996 opinion in Khan v. State Oil Co., concerning manufacturers' setting of retailers' maximum resale prices. Posner criticized as "unsound" the existing Supreme Court law making vertical maximum-price fixing per se illegal, noting that the prior rulings rested economically on "increasingly wobbly, moth-eaten foundations." But, he held, as an appellate judge he was bound by the Supreme Court's 1968 ruling in Albrecht v. Herald Co. holding maximum price-fixing to be illegal per se.
Yet Posner ultimately had his cake and ate it, too. The Supreme Court congratulated Judge Posner for applying stare decisis, despite his disagreement with the prior Court case: "It is this Court's prerogative alone to overrule one of its precedents." The Court then overruled Albrecht.
JUDICIAL NULLIFICATION At the other end of the spectrum, some lower courts have pretended that there is no Supreme Court decision of relevance in the first place. Judge Frank H. Easterbrook's 1985 opinion in Polk Brothers, Inc. v. Forest City Enterprises, Inc. is a noteworthy example. The two named parties in the case agreed not to sell certain products so that each could specialize in separate product lines. But the agreement applied only to their two respective stores in a single shopping mall; everywhere else in the relevant market, the companies competed in all the product lines concerned. Technically, by standard antitrust pigeonholing, the agreement between the two firms constituted a horizontal territorial allocation, even if that "territory" was just a shopping mall, and territorial allocations are per se illegal under the Supreme Court's 1972 ruling in United States v. Topco Associates, Inc. So, when Forest City informed Polk Brothers that it would no longer honor its agreement and Polk Brothers sought an injunction to compel performance, the district court held that the agreement was illegal per se.
On appeal, Judge Easterbrook explained why such an agreement was presumptively a good thing. Without the agreement, neither store would locate at the shopping mall in the first place, so consumers would have fewer, not more, choices. Under Topco, good things would be irrelevant; the contract would be per se illegal. Nonetheless, Judge Easterbrook did the economically honorable thing, reversing the district court without ever discussing, or even citing, Topco. Similar acts of unabashed judicial nullification in the face of robotic per se rulings from the Supreme Court have been routine in other areas of antitrust.
Other lower-court judges have worked around undesirable Supreme Court holdings by artfully distinguishing their cases from the Court's rulings. Two ploys have been used. In the first, the lower court notes that seemingly binding precedent comes from older cases--as all precedent must. The court then says that, if the current Supreme Court were deciding a similar case today, it would adopt a different rule. The lower court then resolves the current dispute "as if" the Supreme Court had abandoned its prior rule in favor of a more economically informed alternative.
So, for example, then-Judge Robert H. Bork discussed horizontal nonprice contracts (like those at issue in Topco) in his 1986 opinion in Rothery Storage & Van Co. v. Atlas Van Lines. Franchisee Rothery challenged various aspects of the way its franchisor, Atlas, had organized its operations, which required cooperation among competing movers using the Atlas name. "The business arrangement in Topco very closely resembles Atlas' policy," Judge Bork wrote for the D.C. Circuit Court panel. Thus, if Topco and other horizontal-restraint cases controlled, "the restraints imposed by Atlas would appear to be a per se violation of the Sherman Act." However, Judge Bork concluded, "examination of more recent Supreme Court decisions" indicated that those cases (including Topco) "must be regarded as effectively overruled." Treating the cases "as if' they had been overruled, Judge Bork held that the horizontal contracts at issue in Rothery did not violate the Sherman Act.
An alternative to arguing the law is arguing the facts. Lower courts sometimes simply distinguish the facts of their cases from the facts in the case on which the Court opined. For example, in the 1977 case Continental T.V., Inc. v. GTE Sylvania, Inc., a case involving exclusive territories and related restraints that a television manufacturer agreed to with its retailers, the Supreme Court noted much lower-court hostility to its per se ruling 10 years earlier in Arnold, Schwinn & Co. v. United States, a case involving a bicycle manufacturer's non-price vertical contracts with retailers. Opposition to Schwinn was clear from the appellate ruling in the Sylvania case. The Ninth Circuit had distinguished Schwinn on several grounds, all essentially factual--bicycles are not televisions--and thus held that the matter before it should be judged under the rule of reason rather than the per se standard dictated by Schwinn. In addition, as the Supreme Court noted, "the [Ninth Circuit] found support for its position in the ... decisions of other federal courts involving non-price vertical restrictions."
The Supreme Court disagreed with the various attempts to distinguish Schwinn, stating that those distinctions "have no basis" in that case. However, faced with large-scale judicial nullification of Schwinn in the lower courts, the Supreme Court noted that its prior per se ruling had little economic justification and concluded that the case "must be overruled." Leading from the rear, the Supreme Court validated what lower courts had already been doing with bad Court precedent.
TYPE I AND TYPE II ERRORS
The earlier dominance of per se reasoning necessarily implied that no error costs result from banning a questionable contract or practice. The only error would be in not prohibiting it. During the most recent generation of antitrust, however, the law has been increasingly influenced by courts' recognition of the harm that misguided antitrust can inflict.
More specifically, antitrust courts have recognized that two types of error should be considered, only one of which enters into per se reasoning. Type I error refers to a "false positive," analogous in the legal context to mistakenly imposing liability on an innocent defendant. Type II error is a "false negative," or failing to punish a guilty party. Per se reasoning entails little chance of Type II error, but great likelihood of Type I error.
Because judges are human, their decisions will sometimes be wrong. If a decision can never be correct with certainty, there is always some possibility of error in deciding one way or another. Each type of error has a cost associated with it. Optimally, decisions would be made so as to minimize the costs of being wrong. That decision standard is reflected in courts' choice of burdens of proof in different types of cases.
The trade-off between Type I and Type II error is common to all of law. But antitrust is different in one respect: Type II errors (failing to penalize anticompetitive contracts and practices) will be low as long as entry barriers into markets plagued by suspected anti-competition are also low. As prices rise because of anticompetitive contracts or practices, new entrants will emerge to alleviate, or even eradicate, the problem. Letting the guilty go free in antitrust is generally a self-correcting problem.
Type I error, however, is not subject to much self-correction. If liability is imposed on conduct that actually is beneficial (that is, innocents are punished), there is no market corrective for judicial mistake. Only judicial reversal of the case or legislative intervention to change the decision will undo the Type I error.
Antitrust developments of the past 25 years or so demonstrate growing awareness of the importance of the distinction between Type I and Type II error costs, with the balance shifting toward giving greater weight to the former. Traditionally, not only were Type I error costs treated as minimal (as reflected by the dominance of per se rules), but legislative correction was treated--or at least given lip service--as one reason not to be overly concerned about Type I error, even though legislative correction never occurred.
All that began to change a generation ago. First, the importance of Type II error itself starting getting attention, with the withering away of per se rules described above. Second, courts increasingly have looked for rules that are relatively foolproof in their application. Optimal avoidance of error requires not just rules that are substantively sound, but also ones relatively easy for courts to apply correctly. Using a mode of analysis that later courts will misapply is patently undesirable.
Sophisticated judges have recognized the need for simplicity. In his 1990 opinion in the case Town of Concord v. Boston Edison Co., then-Judge Stephen Breyer preceded his analysis with an encomium for simple rules:
We shall take account of the institutional fact that antitrust rules are court-administered rules. They must be clear enough for lawyers to explain them to clients. They must be administratively workable and therefore cannot always take account of every complex economic circumstance or qualification.... They must be designed with the knowledge that firms ultimately act, not in precise conformity with the literal language of complex rules, but in reaction to what they see as the likely outcome of court proceedings.
Concern about how judges will apply antitrust precedent extends as well to misgivings about antitrust remedies that will create de facto regulatory regimes, which in turn reflect concern for antitrust causes of action that would require courts to take continuing jurisdiction. Lessons from the mess created by judicial regulation of the telephone industry following the AT&T case have been learned.
HOW ANTITRUST ECONOMICS AND LAW STILL DIFFER
Concerns about Type I error plus the problems of judicial application of overly complex antitrust principles explain a major nondevelopment in the past generation of antitrust: the fact that "post-Chicago" economic approaches to antitrust have had no important impact in the courts. Post-Chicago economics, to borrow from Malcolm Coate and Jeffrey Fischer's 2001 Akron Law Review article, "relies on game-theoretic concepts, which emphasize strategic behavior among economic agents." Game-theoretic approaches typically model repeated interaction over time among competing firms and among firms and purchasers. Behavior that would make no sense economically if part of a one-time-only strategy, such as predatory pricing, can be shown more sensible as part of a repeated-strategy game--at least in theory.
The potential judicial effect of post-Chicago economics seemed especially high following the Supreme Court's 1992 opinion in Eastman Kodak v. Image Technical Services, which drew heavily on complicated game-theoretic models to provide a basis for plaintiff recovery. But Kodak has been narrowly interpreted by lower courts, which have again shown wariness of seemingly broad Supreme Court pronouncements.
More generally, the complexity of game-theoretic approaches has made them unattractive to modern antitrust judges eager for simple rules and worried about how complexity produces Type I error. In the one area in which economists' game-theoretic approaches seemingly offered new possibilities for plaintiff victories--predatory pricing--the post-Chicago approach has failed to deliver.
In an antitrust order increasingly respectful of economic learning, why has post-Chicago economics had so little impact? Given the increasing concerns about Type I error and concomitant desire for simplicity in liability rules, post-Chicago economics cannot be an easy sell. As William Kovacic and Carl Shapiro wrote in a 2000 Economic Perspectives article,
Game-theoretic methods dominated industrial organization theory in the 1970s and 1980s. The flexibility of game theory allowed economic theorists to generate equilibrium predictions in settings involving a wide range of conduct.... However, the same flexibility made general predictions hard to come by. Some types of conduct, such as long-term contracts with key customers or preemptive capacity expansion, could deter entry and entrench dominance, but they also could generate efficiencies. The only way to tell in a given case appeared to be for the antitrust agencies and the courts to conduct a full-scale rule of reason inquiry.
As Judge Easterbrook has written, it is hard enough for courts "to determine what is 'efficient' using a simple model" of competition. So, "how are courts going to decide cases based on complex, strategic models?" The answer, surely, is that they will do so badly, with considerable amounts of costly Type I error. The error costs must be higher when the game-theoretic economic models themselves do not generate consistent answers, which often happens. Judicial error costs are avoided by more simple rules, as courts themselves apparently have realized.
STATIC VS. DYNAMIC EFFICIENCY
The standard antitrust paradigm, even in the current era when price (or, reciprocally, quantity) is the principal focus, takes for granted that property rights are well-defined and enforced. When that assumption does not apply, the antitrust model has proven difficult to utilize, sometimes leading to perverse applications.
The tension between antitrust and property is well understood in the context of intellectual property. Legal protections afforded by patents, copyrights, and trademarks recognize that creation and enforcement of intellectual property entails a separate cost--the item must not only be produced, but first created--that does not apply to the standard widget already in existence. If so, prices above marginal production costs must be charged as an incentive to compensate for the fixed costs of creating the good in the first place. The higher prices necessarily result in lower quantities sold, compared to a price covering only production costs as in the standard economic model of competition.
This distinction between the static model with well-defined property rights and a more dynamic model that takes into account the need to create assets first would seem self-evident. But traditionally it has not been self-evident to antitrust enforcers. In the field of intellectual property, for example, the history of Department of Justice antitrust enforcement has been one of almost unbroken hostility toward patents.
RIGHTS CREATION Although the property-antitrust tension arises most frequently in the area of intellectual property such as patents, it is perhaps best illustrated in the context of more traditional property rights. Take the standard economic example, the fishery. Typically, fish are found in "open access," owned by no one until they are caught. Because access to a lake, stream, or ocean is open, over-fishing is a well recognized problem. The equally well-recognized solution to this so-called "tragedy of the commons" is some form of ownership, either communal or completely private. With private ownership, over-fishing ends.
But in an antitrust world in which low prices and high quantities are the goal, establishment of property rights seems at first an objectionable solution. Property rights mean excluding some fishers and ending exploitation of the open-access resource. As quantities taken diminish, prices naturally rise--a result striking at the core values of modern antitrust. To antitrusters, the result is particularly objectionable when, as is often the case, the solution to over-exploitation of resources available in open access requires a collective agreement among competing fishermen to reduce their catch. Then, it is a "contract, combination, or conspiracy" employed "in restraint of trade," with restricted quantities and higher prices. In the static antitrust world, Sherman Act liability would follow.
And so it has when private agreements have attempted to solve the tragedy of the commons. The Gulf Coast Shrimpers & Oystermans Association (GCSOA) was a private organization that regulated shrimp harvests along the Mississippi coast of the Gulf of Mexico. Its members agreed to sell only to certain packers, who in turn would pay GCSOA members a minimum price. The Justice Department ended the GCSOA's private definition of property rights in a 1956 criminal action brought under Section 1 of the Sherman Act. And so, an attempt to define private property, thus avoiding the economic waste created by open access, resulted in criminal conviction.
RIGHTS PROTECTION Just as it has been hostile to private creation of property--intellectual or marine--so has antitrust enforcement opposed private enforcement of property rights. To mention some of the better-known cases, the government has attacked manufacturers' collective attempts to safeguard their contract rights against fraud, protect their original fabric designs from being copied by pirates, and prevent reverse engineering of machinery protected by a web of patents and unpatented trade secrets.
Analyses that would reconcile property (including intellectual) law with antitrust, though voluminous, have not succeeded in resolving the essential puzzles. Although complex reasons are often offered for the incompatibility of the two systems, simple reasons suffice. Both intellectual property and antitrust law (as considered today) supposedly seek to maximize social welfare, net of costs. But one system (antitrust) maximizes welfare in a short-run, static sense. The other (property) recognizes that short-run losses from higher prices are necessary for the long-run existence of the good so that benefits will ultimately exceed costs. Thus, comparison of welfare benefits net of costs under the two models must by definition be an empirical exercise, comparing streams of benefits and costs over time, appropriately discounted for the time-value of money and for the risks of attaining the supposed net benefits. What is best in any particular situation requires empirical data that cannot be expected to emerge, at least not in the context of antitrust litigation.
Judge Easterbrook has proposed two basic tests for determining whether an antitrust case makes sense:
* Is there market power?
* Are consumers harmed?
But in situations where property rights are poorly defined or enforced, those tests are not helpful. The imposition of property rights in settings in which none exist will increase prices and "hurt" consumers in the short run.
The foregoing is not a criticism of Judge Easterbrook's filters. They have exerted an important influence in antitrust thinking since their appearance some 20 years ago, and deservedly so. The point, rather, is that they are effective in the standard antitrust paradigm in which property rights are already well defined and enforced. When antitrust cases arise outside that paradigm, standard antitrust thinking can diminish social welfare by applying the tools of maintaining competition in situations to which the standard assumptions do not apply.
MULTIPLE ENFORCEMENT OF ANTITRUST LAWS
The number of antitrust enforcers makes it unique in American law. Two federal agencies (the Department of Justice and the Federal Trade Commission) enforce the Sherman Act, Clayton Act, and the Federal Trade Commission Act. States enforce their own antitrust statutes (most of them based on federal law). And private parties also can bring antitrust suits under both federal and state law, usually with treble damages (plus costs and attorney fees) available for successful private plaintiffs. Needless to say, treble damages have created a powerful incentive for plaintiffs to file antitrust cases in the hope of at least being bought off by settlements.
An important improvement in antitrust law during the past generation has been the reduction of meritless private treble-damage actions in the overall scheme of antitrust enforcement. Private cases brought by one competitor against another (that is, competing sellers suing one another) complaining of actions that are actually pro-competitive now succeed less frequently. Also, courts are more alert to plaintiffs' understandable desire to characterize as treble-damage antitrust cases disputes that are really contract spats or business torts, and thus courts increasingly deny recovery on antitrust grounds. And, perhaps most important, the Supreme Court a generation ago began to limit private recovery for anticompetitive overcharges to plaintiffs who are direct purchasers from antitrust violators. No longer, then, can subsequent--so-called "indirect"--purchasers get access to the antitrust treble-damage trough. That space is limited to direct victims of illegal overcharges.
THE STATES But as private involvement in antitrust enforcement has diminished, public enforcement of antitrust has become more competitive. In particular, state attorneys general have increasingly insisted on mounting their own antitrust enforcement efforts. Most states had antitrust legislation when the Sherman Act was passed in 1890. Subsequently, the states legislated anew, adopting the language of the national antitrust statute.
Until a generation ago, however, state involvement in antitrust enforcement was negligible as compared to that of the federal government. States had only bit parts on the antitrust stage, in part because they could not bring actions patens patriae on behalf of their citizens. That changed in 1976 when the Hart-Scott-Rodino Antitrust Improvements Act, in addition to creating the regulatory regime that now applies to mergers, gave state attorneys general the ability to bring actions patens patriae for damages under the federal antitrust laws. (State legislatures have also passed statutes to allow indirect purchasers to recover under state antitrust statutes.)
State antitrust enforcers differ from federal enforcers in three important ways. First, given that most antitrust offenders will operate across state borders, any particular state by itself is at a disadvantage in pursuing antitrust offenses. Second, given that the state attorney general has responsibility for all of her state's legal work, there will he relatively few lawyers and staff devoted to antitrust work, and they--like the attorney general herself--will be relatively unspecialized. And finally, the state attorney general is an elected official, often one with aspirations for higher office. (The joke goes that "AG" refers to "aspiring governor.") And so, the antitrust agenda of the state attorneys general will be one driven more by political than truly economic concerns, as compared to that of the nonelected national enforcers.
Several predictions arise from those three differences. First, because a single state can do little by itself in ferreting out and pursuing multistate antitrust offenses, and given that few if any offenses are not multi-state, a state attorney general will only become interested in antitrust to the extent that her counterparts in other states are similarly willing to get involved. In fact, multistate antitrust enforcement is coordinated by the interstate National Association of Attorneys General. But second, to the extent that more and more states are willing to get involved in antitrust, the efforts of the state attorneys general merely duplicate what could be done by a national antitrust enforcer. If all 50 state attorneys general want to pursue a Fortune 100 company that operates in each state, they will have to coordinate their efforts and devise ad hoc ways to pursue the case. State enforcement thus will entail great fixed costs in organizing to undertake cases, and high marginal costs in actually pursuing them, as compared to how the national agencies operate. And finally, as politically motivated people, state attorneys general will look for high-profile but easily won cases--ones that will resonate with voters as the attorney general mounts the stump at the next election.
If so, further predictions follow as to the kind of cases states will pursue. Two types will dominate. First, states will seek cases in which the national enforcers are already involved or are likely to become involved. The states, as the more costly enforcers should they pursue their own cases, will be interested in "piling on" in cases where the national enforcers are already active. The Microsoft case exemplifies this first type of case, in which the states essentially took a free ride on the federal government. The states' theory of the case was no different from that of the federal government, Judge Thomas Penfield Jackson ruled early in the litigation. Years later, at the end of the case, state complaints about the allegedly lenient remedies the federal government had obtained against Microsoft were brushed aside by the federal district court.
LEAR, MEAN LAWSUIT For state attorneys general, free riding on the efforts of the national enforcers has its down side in that most of the glory will inure to the feds. Low costs are matched by low benefits. States will naturally be interested, therefore, in cases that the federal government declines to pursue. But what will those cases entail? By definition, conduct that the federal enforcers do not find worth pursuing. Moreover, the cases must deliver a victory at relatively little cost because the states will be bearing the financial burden themselves.
Consider a case like that against Salton, Inc., for resale price maintenance of its George Foreman grills, provisionally settled in September 2002. The case is one in which the federal antitrust authorities would have no interest. Resale price maintenance is now understood to be an intrabrand practice that enhances interbrand competition. Economists almost unanimously applaud resale price maintenance as a way to enhance distributor efforts to market the product vis-a-vis competing brands in ways that almost never have any anticompetitive aspects. Resale price maintenance simply has no place in the modern, economics-based enforcement agenda.
However, resale price maintenance cases like that against Salton are a natural for the state attorneys general. First, anomalously, resale price maintenance remains per se illegal under the Sherman Act and thus is illegal under states' antitrust acts. Therefore, victory is automatic--and cheap. All that need be shown is a contract to set resale prices, or something that a jury might so construe as such a contract. Victory is even easier when the states sue for hundreds of millions of dollars (as in the Salton case) and then offer a settlement for cents on the dollar ($8 million in the Salton case). No company, particularly one with public shareholders, could refuse an offer to settle for so little. To do so would invite a shareholder suit.
Salton's George Foreman grill is one of the great success stories in kitchen appliance sales. With unit sales in the millions, its high profile is guaranteed by George Foreman's name and ability to promote it. Hanging the scalp of a brand-name retailer and a phenomenally successful product on an attorney general's wall was not likely to discourage the two lead attorneys general in the Salton case, New York's Eliot Spitzer and Illinois's James Ryan. The former has shown himself not averse to publicity; the latter was running for governor at the time the suit's settlement was announced.
The suit certainly was valuable to the attorneys general. But what was in it for consumers, the supposed beneficiaries of antitrust? Nothing, apparently. Not only is resale price maintenance generally a beneficial practice socially, but the settlement amount was laughable in terms of redressing any supposed consumer injury. The settlement amounted to just pennies per grill sold. The attorneys general did not even try to get the money to the actual sufferers of any higher prices. Instead--attorneys general are politicians and 2002 was an election year--the money was destined elsewhere, as the attorneys general announced:
In view of the difficulty in identifying the millions of purchasers of the Salton grills covered by the settlement and relatively small alleged overcharge per grill purchased, the states propose to use the $8 million settlement in the following manner: Each state shall direct that its share of the $8 million be distributed to the state, its political subdivisions, municipalities, not-for-profit corporations, and/or charitable organizations for health or nutrition-related causes. In this manner, the purchasers covered by the lawsuits (persons who bought Salton George Foreman Grills) will benefit from the settlement.
This statement is commendably candid. Not only will supposedly wronged consumers not get any money, but the supposed overcharge was "relatively small" to begin with. If the overcharge was "relatively small," Salton could not have had much market power. Thus, the case flunks one of the principal filter tests that Judge Easterbrook rightly would impose to evaluate the worth of a standard antitrust case.
In the longer run, because of developments over the past generation, American enforcers will find themselves increasingly in competition with non-American antitrust regimes, most notably the European Union. The operative legislation under which EU competition law operates is not very different doctrinally from that of the United States. European law divides into three areas--collusion, dominance, and mergers--representing a more compact and intellectually more appealing taxonomy than that of American antitrust law. Several aspects of EU competition law are noteworthy, indeed praiseworthy. For instance, to revert to Justice O'Connor's lament in the Jefferson Parish case, there is no separate statutory "box" in EU competition law for tying akin to that under Section 3 of the Clayton Act.
European antitrust is a relatively recent phenomenon, just a generation old. The economics and politics by which it ultimately will work remain to be seen. But several lessons from the American antitrust experience are instructive.
First, it is not surprising that a large, trans-European antitrust regime has emerged via the EU. Individual European countries are at a disadvantage in antitrust enforcement similar to that of the individual American states. Most important, major enterprises operating in Europe do so transnationally, just as most large American firms operate across the several states. So any particular European country operates at a disadvantage in pursuing supposedly anticompetitive problems of any consequence. It makes sense to pursue antitrust offenses collectively. Hence the desirability, in principle, of European enforcement of otherwise-national antitrust law by cartelizing enforcement under an organization like the EU.
But the structure of European antitrust enforcement differs importantly from that characterizing the state-federal relationship in the United States. Most important, EU enforcement is not directed by an organization made up of representatives answering to national political authorities, the way that the National Association of Attorneys General is merely a coalition of representatives from state offices. Operating through the association, state politicians function ad hoc as a loose federation, focused on a particular case, working with personnel that are under their control (and thus whose careers depend on the respective attorney general's political fortune). However, EU antitrust enforcement includes a permanent bureaucracy, largely unreachable by individual national authority, applying and enforcing its own dictates.
And so, judging from the American experience, one would expect that individual European governments would take a back seat to the EU antitrust enforcement regime. In the United States, the states must gear up case-by-case to tackle each matter they decide to pursue, making them the higher-cost enforcer as compared to the federal government, with its two permanent bureaucracies (the DOJ and FTC). In Europe, however, it is the EU that has the permanent antitrust bureaucracy for pursuing trans-European antitrust enforcement. Thus, although the American government pursues the major antitrust cases (with the states either free-riding on those cases or instituting relatively unimportant actions of no interest to the federal government), the reverse would be true in Europe. The European Union permanent bureaucracy, not enforcers from a particular country, would undertake the principal enforcement actions in Europe.
INTERCONTINENTAL COMPETITION Predictably, given the institutional arrangements involving the EU and European national governments, how would EU antitrust enforcers treat contracts and practices that affect both the United States and Europe? The independent bureaucracy operating in Europe would respond to the incentives that it faces. The Europeans are antitrust newcomers, and so will justify their existence only by imposing more restrictive rules than the American federal enforcers impose. Looser rules would have the Europeans merely blessing actions already resolved in America, and such a "me too" attitude hardly justifies a separate European presence in global antitrust.
Modern standoffs between the United States and Europe on mergers, such as the proposed GE-Honeywell merger, illustrate Europeans' desires to be more restrictive, so as to override American antitrust. As Abbott Lipsky has written,
In GE/Honeywell, a foreign antitrust authority for the first time prohibited a merger that the United States had permitted. It was also the first time the world's major antitrust agencies confronted each other across a clear line of disagreement: the rules implicit in the European decision--vigorously defended by their authors--were declared by the U.S. "antithetical" to the purposes of competition law. The general public thereby became aware of conflicting approaches to business regulation in the world's two largest economies.
The concern engendered by the EU's GE/Honeywell decision has been considerable. But GE/Honeywell was not unprecedented; European objections to the Boeing-McDonnell Douglas merger, a largely vertical arrangement approved quickly in the United States, forced Boeing to abandon existing contracts that disfavored its European competitor, Airbus. The acrimonious disagreement over the merger came close to starting a trade war between the United States and Europe.
The need to avoid a "me too" stance by alleging offenses not pursued by Americans explains other European cases, such as the EU's case against Microsoft. Long after its American antitrust problems had been resolved with seemingly few important consequences, Microsoft faced continuing challenges from European antitrust enforcers. Predictably, the European case focused on peripheral areas such as servers and media players that did not figure in the American case, and demanded sanctions different from those sought by American antitrust enforcers. Just as predictably, Microsoft's competitors sought to take advantage of the separate European agenda to advance their own agendas. Having a European antitrust regime alongside that of the United States offers all the advantages of forum-shopping without the possibility that the cases might eventually be consolidated, and so be subject to a single standard.
Perhaps most egregious is the EU's pursuit of vertical nonprice restraints. Those agreements (involving things like manufacturers setting exclusive territories in their distribution chains) have practically no potential for impeding competition and so are almost always legal under American law. But they have frequently been declared illegal in Europe. Indeed, the hostility with which European antitrust enforcers pursue companies thought to be using vertical non-price contracts is sometimes remarkable.
There is only one body of economics that supposedly drives antitrust. But with the United States still the dominant antitrust enforcer and Europe striving to create a niche for itself, one body of antitrust law seems unlikely to emerge. In fact, there is demonstrated resistance in the EU against a single approach ("harmonization" is the term usually employed) to antitrust. Harmonization would mean that, once the American antitrust authorities had passed on a merger or a case with international significance such as Microsoft, there would be little for the Europeans to do.
The incentives to maintain a separate EU antitrust presence vis-a-vis the United States are the same as those that, in America, drive state attorneys general to maintain their own antitrust regimes. Just as state attorneys general find it useful for their own careers to bring high-profile antitrust cases, so it is widely suspected that EU antitrust actions are driven, at least in part, by what is perceived as politically and personally useful to the EU bureaucracy. That bureaucracy is headed by Mario Monti, the European commissioner for competition policy. Known as "Super Mario" for his ability to bring down companies such as General Electric and Microsoft, Montiis one of the most powerful men in Europe. Claims that EU actions are motivated by concerns about politics rather than economics are heard increasingly.
Indeed, one would predict further, the European permanent bureaucracy would sometimes advance its own ends at the expense of national enforcers, in ways that staff from the state attorneys generals' offices would never do as part of the National Association of Attorneys General. The more independent a bureaucracy, the more likely it will be "caught out" when political shifts in the underlying populace or national governments occur. That happened in the most recent wave of EU merger cases, in which several EU rulings objecting to mergers were overturned by reviewing courts in Europe.
The current antitrust generation has seen a pronounced shift in favor of the economic view of antitrust's role, to the diminution of any political or social objectives once thought to be important antitrust goals. Part and parcel of the economic approach to antitrust has been increasing judicial attention to Type I error in antitrust judgments, including the desirability of relatively simple antitrust rules. As long as entry barriers are not severe, any failure to implement an antitrust action that would have increased consumer welfare will correct itself. This concern over Type I errors and simple rules has made it difficult to implement post-Chicago economic approaches to antitrust.
Nonetheless, three developments over the past generation deleteriously separate antitrust law from economics. Antitrust suits sometimes interfere with private attempts to manage the commons. States' insistence on being involved in antitrust enforcement frequently results in harmful suits that make no economic sense. So does European enforcement of its own antitrust rules. The desires of both foreign enforcers and state attorneys general to assume larger roles on the global antitrust stage are particularly worth watching. How those issues are resolved in the next antitrust generation will be interesting to see.
* "The ABA Section of Antitrust Law and International Antitrust Convergence: Sketching the Grand Design," by Abbott Lipsy. Antitrust, Vol. 16 (2001).
* "Antitrust Policy: A Century of Economic and Legal Thinking," by William Kovacic and Carl Shapiro. Journal of Economic Perspectives, Vol. 14 (2000).
* "Can Post- Chicago Economics Survive Daubert?" by Malcolm Coate and Jeffrey Fischer. Akron Law Review, Vol. 34 (2001).
Fred S. McChesney is the Class of 1967 / James B. Haddad Professor of Law at Northwestern University and a professor in the Department of Management and Strategy in Northwestern's Kellogg School of Management. He may be contacted by e-mail at email@example.com.
This piece is adapted from a longer article that appeared in the Emory Law Journal in 2003.…