By Michael S. Greve

May Texas courts decide antitrust cases over allegedly anticompetitive conduct that occurred wholly outside of Texas? The intuitive and correct answer is no. In the way, unfortunately, lay confounding factors: the liability explosion, needlessly vexing questions of antitrust jurisdiction, and above all the wretched condition of the constitutional and common law rules that govern state-to-state relations in our federal order. In an exceedingly important antitrust decision in October 2006, however, the Texas Supreme Court reached the correct decision. Along the way it articulated sound federalism principles that should resonate broadly beyond Texas and beyond antitrust. Although the court has been asked to reconsider its decision, it should stick to its guns.

Coca-Cola Company, et al. v. Harmar Bottling Company, et al.[1] has been rattling around Texas courts for more than thirteen years. At issue are so-called calendar marketing agreements (CMAs)–vertical agreements between soft drink bottlers and retailers–which govern the advertising, display, and sale of the products. The general structure of CMAs dictates that retailers obtain price discounts in return for compliance with the contractual marketing terms. For example, a CMA may specify that a retailer cannot conduct a marketing campaign for a rival product during a specified period. Economists believe CMAs may have potent pro-competitive effects, as they tend to prevent “free riding” by retailers and reduce prices to consumers. CMAs nearly identical to those at issue in Harmar are lawful everywhere in the country, including, after the Texas Supreme Court’s decision, in Texas.[2]

Naturally, bottlers often take a dim view of their competitors’ CMAs. (Producers often take a dim view of competition that impedes their success in any fashion.) In Harmar, some Royal Crown soda bottlers filed an antitrust suit against Coca-Cola and Pepsi bottlers in the so-called Ark-La-Tex market–in reality, not a single geographic market, but rather a collection of counties in Arkansas, Louisiana, Texas, and a stretch of Oklahoma. The suit was filed and litigated for many long years in Morris County, Texas, a place where, prior to the instigation of the case, some plaintiffs had never set foot. The plaintiffs sued in that county, abutting the Arkansas border, and obtained representation from Harold Nix, Morris County’s uncrowned king and, consequently, one of the most famous and richest trial lawyers in Texas. Nix owes his fame and fortune principally to a mass products liability case he filed in Morris County, which produced lucrative settlements in a case involving some “three thousand plaintiffs, five hundred defendants, three hundred lawyers, and no evidence.”[3]

In Harmar, a local jury returned a verdict on the antitrust counts. The trial court then rendered judgment against Coca-Cola (the Pepsi defendants having settled out of court) in the amount of $15.6 million, plus $500,000 in attorneys’ fees. The court also enjoined a wide variety of common business practices, ranging from exclusive advertising commitments to suggestions as to how retailers might organize their cold drink storage space. While technically confined to forty counties in Texas, Oklahoma, Arkansas, and Louisiana, the injunction effectively regulated competition much more widely, since supermarkets that operate in those and neighboring states are likely to want to operate under a single set of rules for the entire region. Moreover, Harmar was only the first of a raft of similar state court cases, all filed in Morris County or equally plaintiff-friendly Texas jurisdictions, involving competition from Kansas to Missouri to Louisiana.[4]

After an appellate court sustained the trial court’s ruling, the Texas Supreme Court granted certiorari. In October 2006, nearly two years after the case was argued, the court rendered its decision. In an opinion authored by Justice Nathan L. Hecht, it held that Texas courts may not apply Texas antitrust law to wholly out-of-state conduct. It further held that interstate comity barred Texas courts from entertaining the complaints under sister states’ antitrust laws. With respect to the alleged in-state violations, the court held that the plaintiffs had shown no facts upon which a reasonable jury could find defendants guilty of antitrust violations. The court was closely divided. A sternly worded dissent by Justice Scott A. Brister, joined by three other justices, rejected most of the majority’s conclusions.

A few weeks later, the plaintiffs and several amici asked for a rehearing and reconsideration. The Texas Supreme Court ordered briefing on that question and, as this goes to press, is considering the submissions.

Why It Matters

A $16 million verdict is hardly news. The injunctive relief at issue in Harmar is more consequential: it would block widely accepted, essential, and intensely competitive practices. But the true significance of Harmar lays elsewhere.

Federalism poses difficult issues of jurisdiction in both a vertical, state-to-nation dimension and in a horizontal, state-to-state dimension. Our country’s public and legal debates focus on the vertical aspect: should Washington or Sacramento regulate drug use in Santa Clara County? Horizontal federalism, in contrast, rests on an unarticulated and virtually uncontested premise: of course plaintiffs should be able to drag defendants into any forum they like. And of course that forum should then apply its own law if it sees fit to do so.

The trial bar as well as some judges and scholars defend this shopping spree as the very essence of federalism. Other judges and scholars–modern-day disciples of Felix Frankfurter–acknowledge horizontal federalism’s practical difficulties but deny that they have a constitutional dimension. A third group parks its discontents with our federalism’s unbearable results in odd places. In February 2007, for example, the U.S. Supreme Court determined that state juries may take defendants’ out-of-state conduct into consideration in determining punitive damages, but may not punish them explicitly for that conduct.[5] Behind that exquisite distinction lies a sound concern over extraterritoriality–that is, the danger that one state might punish defendants for conduct that may have been entirely lawful in other states. Because the constitutional norms and common law rules that would avert that danger are in disrepair, the Court rested its ruling on the all-purpose due process clause.

The Harmar majority, by virtue of its decision and opinion, has joined a fourth, still-tiny group. Members of this cadre of the wise insist that relations among states implicate serious constitutional concerns. They further contend that constitutionally acceptable relations among states cannot be maintained by force of the federal government– least of all if every jurisdiction insists on pushing its presumed authority to the outer limits or, more likely, accedes to being pushed to those limits by self-seeking private litigants. The system is in need of some restraint.

The Harmar majority showed sympathy and courtesy and taste in dealing with those problems in the distinctly difficult antitrust context. That is why its ruling has drawn considerable attention and organized opposition, and why it merits a defense.

Antitrust Jurisdiction: First Principles[6]

The Harmar Court started with the right premise: “No State can legislate except with reference to its own jurisdiction.”[7] American federalism rests on a principle of equal, territorial states.[8] That arrangement, in turn, requires a principle of nonaggression. One could grant states an equal right to tread on one another, but that would not be a net gain in states’ rights–just as my right to stick a fist in your face is not a net gain in individual rights.

Aggression, Alexander Hamilton explained in Federalist 7, encompasses not only physical assault, but also–and emphatically–the unwarranted extension and application of one state’s laws to another state’s citizens. The Constitution teems with provisions to guard against that menace: the Contracts Clause, for example, which sought to bar states from enacting debtor relief laws that would favor in-state debtors over out-of-state creditors; and the Privileges and Immunities Clause, which compels states to treat their own and other states’ citizens on equal terms.[9]

The principle, then, is clear enough, but it does not explain its own reach and application. What, precisely, does it entail and require in a world in which economic actors and transactions routinely cross state borders, and in which disputes over whose law governs are therefore bound to arise with great frequency? These questions are particularly difficult and pressing in the antitrust context for two reasons: the peculiar structure of our antitrust laws, and the presence of pervasive externalities.

Antitrust. The major federal antitrust statutes–the Sherman Act and the Clayton Act–were originally enacted long before the New Deal. They were intended to supplement, but not supplant, state jurisdiction: the federal statutes would operate on conspiracies against trade in interstate commerce, which states were then deemed unable to regulate. Conversely, the national government was thought to be constitutionally barred from reaching anticompetitive conspiracies contained within a state. The distinction between in-state and interstate commerce, however, vanished in the New Deal. In its wake, the notion that federal antitrust statutes “supplement” state law took on a very different meaning: state and federal laws henceforth governed private conduct alongside and on top of each other. The Sherman Act reaches purely local events with no cross-border effects–although, as Supreme Court Justice Antonin Scalia has observed, the statute by its terms applies only to interstate commerce.[10] At the same time, however, the Sherman Act does not preempt any state’s law.

To appreciate the salience of this point, consider this: in areas from airline regulation to telecommunications to financial services, pro-competitive federal statutes have freed industries and their consumers from hostile or parochial state and local regulations. Those salutary effects depend entirely on the preemptive force of federal legislation.[11] Our competition laws, incongruously, lack that force. One must look elsewhere for restraints.

Externalities. When more than one state’s law can apply to a given set of transactions, which state’s law should govern? In situations in which private contracts work well, the rules are conceptually straightforward: apply the forum and the law that the parties chose. If the parties made no explicit choice, go to the venue and the law that the parties most likely would have chosen. Those rules work well–or rather would work well, if we had them–for questions of jurisdiction (that is, questions of where a given defendant may be sued and a case may be heard) and for the choice of law.[12]

In the antitrust context, unfortunately, contractual rules fail for the obvious reason that the contracts themselves are suspect and, in some instances, criminal. For example, we cannot permit two Texas firms to fix prices in Arkansas and then defend the agreement on the grounds that it was made in Texas. No real significance attaches to the parties’ choice of law when the burdens fall on strangers to the transaction. The jurisdictional test has to go to the effects of anticompetitive practices, regardless of where they may originate. In that sense, state antitrust law has to be “extraterritorial.”

The corollary, however, is that states must not have jurisdiction over anticompetitive effects in other states–even if those effects should originate within the state’s own boundaries. The appellate court in Harmar mangled that issue and held that the defendants’ conduct outside Texas was subject to Texas law because some of the contracts had been drawn up in that state. But while that may be a reason to give the Texas courts personal jurisdiction over the defendants, it provides no reason for exercising prescriptive jurisdiction over events in other states.

The Texas Supreme Court got it right. “Why,” it asked, “should Texas law supplant Arkansas, Louisiana, or Oklahoma law about how best to protect consumers from anti-competitive conduct and injury in those states? . . . There is no good answer.”[13] There is a very good answer why Texas law should not have that effect. We do not like it one bit when the European Union “protects” us from Microsoft. Why, then, should consumers in Alabama be grateful for the Texas courts’ “protection” against lawful, procompetitive practices?

Admittedly, some messy antitrust cases threaten to confound the basic rules. Patents, for example, are good everywhere or nowhere. If one jurisdiction decides that a particular patent application violates competition law, the protection is lost around the world. There is no easy fix for that unfortunate result of an unadulterated effects test. Another problem arises when activities that are illegal in one state but are perhaps lawful in the foreign jurisdictions where they are conducted have price effects in the original state.[14] No such complications, however, were remotely present in Harmar. Some of the parties had not and could not enter Texas. (Bottling operations are exclusive territorial franchises.) At a bare minimum, those parties should have been dismissed at the outset of the litigation. The trial court itself managed to separate activities and effects on each side of the line, albeit belatedly: it awarded separate damages to the individual plaintiffs, including those operating wholly out-of-state.

The Principles Applied

The Harmar majority understood that the central issue in the case was jurisdiction. It started its analysis on that leg and it construed the Texas Free Enterprise and Antitrust Act (TFEAA)–the state’s basic antitrust statute–in that light. The ease and elegance of the court’s conclusions tend to cloud its considerable accomplishment: in an area in which issues of extraterritoriality, jurisdiction, and choice of law intersect, even sophisticated judges may lose track. Of that, the Harmar dissent provides an illustration.

Extraterritoriality. Remarkably, all nine members of the Texas Supreme Court agreed that Texas law could not apply extraterritorially. The justices differed, however, in their analyses.

The TFEAA provides that lawsuits shall not “be barred on the grounds that the activity or conduct complained of in any way affects or involves interstate or foreign commerce.”[15] The statute also says that its purpose is “to maintain and promote economic competition in trade and commerce occurring wholly or partly within the State of Texas and to provide the benefits of that competition to consumers in the state.”[16] But while the plaintiffs and the dissent made much of the “partly” language, it is best read as a sensible recognition of the already-observed fact that at some level, antitrust litigation must be “extraterritorial.” The mere fact that anticompetitive conduct has effects in other states cannot shield anticompetitive practices against Texas consumers. Conversely, however, the mere fact that such conduct extends into Texas cannot create a cause of action to demand relief on a global basis. As Justice Scalia put it in a strikingly similar (albeit international) context, the extraterritorial reach of the Sherman Act “has nothing to do with the jurisdiction of the [U.S.] courts.”[17] The same holds for the TFEAA.

Jurisdiction. The analysis becomes a lot more straightforward when one follows the Harmar majority in starting with the jurisdictional questions. The necessary corollary of federalism’s nonaggression principle is a principle of statutory interpretation: do not give a state statute extraterritorial effect unless the legislature clearly intended that result. On that basis, the Harmar majority recognized that the TFEAA’s language means to bar defenses that might block the full enforcement of the act within Texas. It does not invite a kind of antitrust imperialism, to be exercised even at some foreign plaintiff’s instigation, whenever the complained-of conduct touches Texas. As the court recognized, the TFEAA’s plain intent was to protect Texas consumers.

Read for all it is worth, the majority opinion says the legislative jurisdiction of the state of Texas, at least as exhibited in the TFEAA, extends no farther than the states’ boundaries. And since the Texas courts’ subject-matter jurisdiction extends no farther than the legislature’s, the courts lack such jurisdiction over events that have no anticompetitive effects in Texas or, for that matter, over the segregable out-of-state effects of conduct that affect some Texas consumers.

Choice of Law. Granted that Texas law does not apply to out-of-state conduct, should Texas courts have applied laws of other states to that conduct? The Harmar majority said no on the grounds that state courts should not readily volunteer to determine a sister state’s public policies.

The dissent observed that the defendants had never asked the trial court to apply foreign rather than Texas law–presumably, it surmised, because the antitrust laws of Oklahoma, Arkansas, and Louisiana look very similar to the Texas statute. That conclusion cannot be defended. State antitrust law, like federal law, is worded in broad generalities; almost all antitrust law is common law.[18] No state court can assume that sister-state policy is identical based solely on the statutory language; it would have to know in some detail what the sister states’ courts, and perhaps their enforcement agencies, think and have said about vertical restraints of the sort here at issue. If that policy is one of toleration, there may be no explicit policy or judicial precedent at all. In that event, asking the defendants to show that sister-state law is materially different is to confront them with an impossible task. That appeared to be the situation in Harmar. What evidence there was actually ran the other way: at least in Louisiana, CMAs appear entirely lawful.[19]

Surely, “hands off” is the proper rule. The plaintiffs’ rights remain unaffected: they can litigate their complaints and obtain their remedy in the appropriate state court, or else sue in federal court under federal antitrust law (and litigate the pendent state claims in that forum).

The Harmar majority characterized its holding not merely as a matter of comity but as abstention–probably, the dissent surmised, in analogy to the abstention doctrines that federal courts employ to avoid unnecessary interference with state courts and their proceedings. If this was the case, more power to the Texas Supreme Court: much can be said for horizontal abstention doctrines, and I have suggested elsewhere that such doctrines are required as a matter of federal constitutional law.[20] Our federal courts are light years from that recognition.[21] It is all the more encouraging that a state court seems to have adopted it unilaterally.

With Amici Like These, Who Needs Inimici?

The Harmar plaintiffs’ motion for a rehearing argues that the Texas Supreme Court failed to properly heed the jury’s findings and the voluminous record. That position is supported by a number of Texas law professors and civil procedure experts. The amici volunteer, however, that they are unfamiliar with both the record and with antitrust law. Why, then, should anyone listen? Antitrust experts disagree vociferously about the sorts of vertical restraints at issue in Harmar, but they all agree that it is extremely hard in this area to separate competitive from anticompetitive arrangements. They all agree, moreover, that the determination will at all events hang on a sophisticated assessment of complicated empirical facts and their fit with legal standards and economic theory. In that context, the amici’s prompting to trust the Morris County jurors’ intuitions is little more than a misplaced protest against the Lone Star State’s arduous tort reform efforts.

More incongruous still is the amicus support plaintiffs have procured from the attorneys general of two of the affected states, Arkansas and Oklahoma. One would think that an attorney general, charged with enforcing competition law within his state, would resist another state’s interference. One would further think that an attorney general, charged with protecting all citizens within his jurisdiction, would oppose any effort to have a home-state company conducting lawful business in the state dragooned into some faraway court and have its business practices declared unlawful in that forum.

One state, Alabama, articulated those positions in the Harmar litigation, in powerful amicus briefs to the Texas Supreme Court that lay out the relevant federalism considerations. (Among those considerations is the fact that Alabama, which does not apply its antitrust laws extraterritorially, should be entitled to equal respect from its sister states.[22])That intervention was initiated by former Alabama attorney general (and now Judge) Bill Pryor, who comprehends American federalism at an uncommonly deep level. Pryor’s exceptionally able staff has sustained Alabama’s Harmar intervention during the tenure of Alabama attorney general Troy King, who on this occasion and others has done his immediate predecessor proud.

The position of the not-so-exceptional officials in Arkansas and Oklahoma is that the Texas courts should determine competition law in those states. Comity concerns, they admit, are important, but they can be accommodated by certifying the Texas questions to the Oklahoma and Arkansas Supreme Courts, or perhaps by “domesticating” a Texas verdict in an Arkansas court.

Ignore the practical difficulties of that suggestion. (Harmar involves questions of law and fact. Are the courts in three or four different states supposed to review the full record and hear oral arguments? Suppose they then disagree: what happens next?) Ignore, too, that the attorneys general purport to speak for an independent branch of government. (Paraphrasing a famous quip of Judge Henry Friendly’s, the attorneys general are asking the Texas courts to act on what the attorneys general think about what Oklahoma and Arkansas courts might think about the referral of a question about which none of them have thought.[23]) Focus, for purposes at hand, on the intriguing question of why state attorneys general support the adjudication of their own laws in another state’s courts.

The Limits of Efficient Adjudication. It would be inefficient, Arkansas and Oklahoma say, to have the Harmar facts adjudicated yet again in an Oklahoma or Arkansas forum. That position, though, rests on a wholesale confusion between ex ante and ex post incentives. The non-Texas plaintiffs joined the Harmar parade in light of Morris County’s pro-plaintiff attractions and, moreover, to produce evidence by anecdote. (Take one CMA provision from Arkansas, another from a different time period from Louisiana, and a third from yet a different time in Texas, and in the words of Louis Armstrong and Bing Crosby, “Now You Has Jazz.”[24]) The attorneys general amici would countenance those maneuvers–and then plead for supposedly “efficient” litigation management. In that world, you wind up with Oklahoma’s wild proposition: we Sooners have an interest in having our own law determined in forums beyond our borders. But the problems would never arise if opportunistic forum shopping were blocked at the front end. The sane and efficient rule is that of the Harmar Court.

Cartel Federalism. For the entire half-generation that the Arkansas and Oklahoma bottlers’ complaints have been litigated in Texas, the attorneys general antitrust enforcers of those states have sat on the sidelines. Only now, after the complaints have been dismissed, has it occurred to them that the matter deserves a second look. It is hard to believe that their entreaties have to do with a desire for competition.

Long ago, Judge Frank Easterbrook suggested that it might be in every state government’s interest (though obviously not the citizens’ interest) to permit and indeed encourage export cartels–state-sponsored producer cartels that can extract surplus profits from the rest of the country.[25] The beneficiaries of those schemes are highly concentrated and in-state, whereas the losers are dispersed and mostly out-of-state. What public official would not ride that wagon? The odds of encountering opposition from sister-state officials are nil: they all have their own cartels to defend.

The state amici’s position in Harmar is consistent with that pattern. To be sure, the attorneys general do not explicitly endorse anticompetitive conduct or intergovernmental conspiracies. Still, when (as in Harmar) competitors rather than consumers or government enforcers complain about antitrust violations, the suspicion arises that they are attempting to serve their own interests rather than those of competition.[26] And when those competitors abscond to a foreign jurisdiction instead of filing a case in their own state or complaining to state officials, the suspicion hardens into a near certainty. The only conceivable point of the attorneys general amici’s submissions is to give the plaintiffs yet another bite at the apple. To what end?

Our Federalism

Under the Constitution, sound rules to govern state-to-state relations are every bit as crucial as state-to-federal relations. If anything, the Constitution is far clearer and more insistent on the former than the latter, and for good reason. Because the Founding Fathers expected the worst of the states, their Constitution economizes on state virtue–for example, by blocking egregious affronts to interstate comity in the Constitution itself.

As the Founders knew full well, however, that strategy has its limits. In the end, the system supposes some awareness within the states of unspecified but nonetheless vital federalism rules. The Harmar Court’s opinion is an all-too-rare signal of that awareness. Let it stand and let it thrive.

Michael S. Greve ( is AEI’s John G. Searle Scholar and director of its Federalism Project ( On account of his earlier work on federalism and antitrust issues (cited in this publication), the Coca-Cola Bottlers’ Association (CCBA), an amicus in the Harmar case, has retained Mr. Greve as an advisor on federalism issues of interest to the association. No fee or other compensation, however, was received by AEI or Mr. Greve in connection with this publication. The views expressed herein may or may not reflect the views of the CCBA or the Harmar litigants. They are, unmistakably, Mr. Greve’s. A version of this article also appeared on the Federalist Society’s website (

AEI research assistant Harriet McConnell and editorial associate Nicole Passan worked with Mr. Greve to edit and produce this Federalism Outlook.


Comments are closed.