2 - Common Law History

We agree with the court below … that “since it is the trade, and not the mark, that is to be protected, a trademark acknowledges no territorial boundaries of municipalities or states or nations, but extends to every market where the trader’s goods have become known and identified by his use of the mark. But the mark, of itself, cannot travel to markets where there is no article to wear the badge and no trader to offer the article.”

Hanover Star Milling Co. v. Metcalf, 240 U.S. 403, 415–416 (1916)

Introduction

Comparing the development of German and European law with American doctrine reveals a number of critical structural differences.¹ Unlike German doctrine, which has always been founded on formalist privilege theory, US law is distinctively nonformal. The concept of goodwill has governed both trademark protection and unfair competition prevention since the 1800s. While substantive trademark law has been wrought with debate on the extension of goodwill protection ever since, neither the realist attack of the 1900s nor the enactment of federal trademark law in 1946 nor the law and economics movement of the 1980s led to a jettisoning of goodwill as the central concept; not surprisingly, trademark-as-property protection remains the order of the day (see infra p. 77 et seq.). Nonetheless, what has remained widely unexplored to date is the relevance of the goodwill concept for trademark and unfair competition conflicts law. A historical perspective reveals several stages of development, including the establishment of equity jurisdiction over cases of trespass on trademark property, a model of virtually unlimited common law rights, and the Supreme Court’s Tea Rose/Rectanus doctrine. In the course of this evolution, trademark and unfair competition law transformed from a domain of absolute and universal rights into a system of market-related goodwill protection. This also laid the foundation for the extension of international goodwill. Another facet unexplored to date is the US federal legal system and its contribution to the “unboundedness” of market rights. While a matter of course for US theorists, the intricacies of Swift and Erie are a maze to civil lawyers. Here, the understanding of “federal common law” under Swift has been particularly important. An inherent tendency to disregard interstate variations of the common law under the pre-Erie system contributed to a general neglect of state sovereignty concerning issues of trademark rights extension. Quite surprisingly, the federalization of US trademark law under the 1946 Lanham Act and preceding statutory trademark laws also failed to substitute the common law foundation of rights acquisition and extension (see infra p. 127 et seq.). Hence, today, it is still Tea Rose/Rectanus that provides for a genuinely market-oriented theory of rights and a general disregard for political boundaries. This is lucidly revealed by a look at the Supreme Court’s 1952 decision in Steele v. Bulova Watch Co., the court’s only precedent on the issue. As revealed by a critical historical analysis of the Steele reasoning and a closer look at its progeny, the tendency of US trademark and unfair competition conflicts law to overextend the protection of domestic rights and competitors is due to its common law foundations and its borrowing of “effects on US commerce” testing from international antitrust doctrine (see infra p. 151 et seq.).

Section 1 Substantive Trademark and Unfair Competition Law

Several aspects are important for this chapter’s historical account. First, the roots of US trademark propertization must be traced to their beginnings—found in eighteenth-century England—in order to understand how substantive law came to be what it is. Second, within the paradigm of trademark-as-property protection, “goodwill” has become the most determinative element. At the same time, trade diversion has been the mirror image of goodwill protection. Indeed, US law has always been a system of trade-diversion prevention. Over time, the system of goodwill-as-property protection that had developed throughout the nineteenth century came into conflict with the structures of a modern society and economy. With the burgeoning of transcontinental trade and business activities at the beginning of the twentieth century, the idea of trademark property became less suitable. As a result, the property paradigm of US trademark and unfair competition law was modernized toward a market-oriented perspective.

I The Early Straightjacket: Equity, Passing Off, and Universality

The historical development of American society and economy differed from that of Europe in a number of respects. Yet, as in Europe, the United States witnessed a dramatic shift in its socioeconomic conditions in the nineteenth century. After the Civil War, a delocalization of trade and an extension of business activity commenced throughout the country. Prior to the 1860s, production and trade had been local, and the need for and use of identifying symbols in trade had been small. The subsequent expansion in territory, population, wealth, and income, however, soon led to a drastic proliferation and extension of marketplaces.² Production and distribution became more sophisticated due to technological innovation and enhanced infrastructural conditions.³ With the concomitant increase in consumers’ per capita income, the diversification of products and a multiplication of intranational and international trade ensued. Intensified competition was the result. Both marketplace expansion and intensified competition, in turn, led to the emergence of new advertising methods—notably, brand-name marketing. In essence, the replacement of direct transactions between producers and consumers by anonymous sales through retailers and middlemen made trademarks an important marketing tool.⁴ The legal arena reflected this development: while disputes over trademark and unfair competition conflicts rarely occupied US courtrooms during the first half of the century,⁵ they assumed a more conspicuous presence after the Civil War.⁶ With this rise in case numbers, US law embarked on an adventurous journey toward a modern trademark and unfair competition regime.

A Trademark Protection in the Distorting Mirror of Law and Equity

The first obstacle in the way of a modern law was a remnant of medieval times. The demarcation between law and equity proved particularly burdensome with respect to a growing demand for judge-made redress among traders and merchants. Of course, after the merger of law and equity in the nineteenth century, common law courts no longer inquired about a special jurisdictional basis for ordering injunctive relief when a trademark infringement was at stake.⁷ But the road that had brought doctrine this far was a winding one. A right owner’s position had been significantly different in the eighteenth and early nineteenth centuries. At that time, trademark protection was based on the concept of fraud. No property right in the trademark was recognized.⁸ It was thus questionable whether a court of equity would grant injunctive relief; after all, this always required the infringement of a subjective right, and not just fraudulent activity by the defendant. Accordingly, alleged infringers would often successfully object to bills in equity and assert that the suit should be brought in a court of law.⁹ There, proof of the defendant’s fraudulent intent was required—and this was not always easy to establish.

A prominent example of the courts’ hesitation to enforce individual trademark rights is the 1742 English case Blanchard v. Hill,¹⁰ in which the court denied relief against the defendant’s use of the plaintiff’s stamp on playing cards. The court was eager to explain that the royal charter entitling the plaintiff to the exclusive use of certain stamps on playing cards amounted to a “plain monopoly” and was therefore “illegal.” Indeed, the anticompetitive nature of the charter as such appears to have been the main reason for the court’s refusal to grant trademark protection.¹¹ But the overall climate at the time was not beneficial for an extension of subjective rights to trade names and marks, either. In particular, the general condemnation of trademark rights as anticompetitive disfavored protection. Upholding a strict requirement of fraudulent intent was one way to keep perceived detriments within narrow confines.

Yet, over time, cases of successful trademark infringement suits became more common. This was often due to a more generous handling of the fraud requirement. A famous example where the plaintiff managed to overcome the obstacles of contemporary law and equity doctrine is the 1824 case Sykes v. Sykes.¹² The defendant had marketed shot-belts and powder-flasks with imitations of the plaintiff’s mark. The court found an infringement, noting that the plaintiff’s sales had decreased after the defendant had started marketing identically labeled goods. What still seemed to be important for the court, however, was that the defendant had marked his wares “in order to denote that they were of the genuine manufacture of the plaintiff.”¹³

Soon after, the courts’ rejection of a subjective rights theory in trademark protection started to falter. Indeed, the 1838 case Millington v. Fox seems to mark the first time that a court recognized a right to the exclusive use of marks.¹⁴ This reflected a dramatic change of direction, particularly since it did not require fraud on the side of the defendant. The case, which appeared before an English court of equity, centered on an allegation that the defendants had marked steel with the plaintiffs’ names and symbols. Lord Chancellor Cottenham, while not using express property terminology, declared that equity could be invoked even absent evidence of fraudulent intent on the side of the defendant:

I see no reason to believe that there has, in this case, been a fraudulent use of the Plaintiffs’ marks. … That circumstance, however, does not deprive the Plaintiffs of their right to the exclusive use of those names; and, therefore, I stated that the case is so made out as to entitle the Plaintiffs to have the injunction made perpetual.¹⁵

By 1863, the courts’ adoption of property terminology had become evident. In Edelsten v. Edelsten, Lord Chancellor Westbury stated:

At law the proper remedy is by an action on the case for deceit: and proof of fraud on the part of the defendant is of the essence of the action: but this Court will act on the principle of protecting property alone, and it is not necessary for the injunction to prove fraud in the Defendant, or that the credit of the Plaintiff is injured by the sale of an inferior article. The injury done to the Plaintiff in his trade by loss of custom is sufficient to support his title to relief.¹⁶

In the same year, Westbury further explained in Leather Cloth Co. v. American Leather Cloth Co.:

It is correct to say that there is no exclusive ownership of the symbols which constitute a trade mark apart from the use or application of them; but the word “trade mark” is the designation of these marks or symbols as and when applied to a vendible commodity, and the exclusive right to make such use[] or application is rightly called property. The true principle therefore would seem to be, that the jurisdiction of the Court in the protection given to trade marks rests upon property, and that the Court interferes by injunction, because that is the only mode by which property of this description can be effectually protected.¹⁷

Ultimately, trademark infringement had evolved from fraudulent passing off to trespass on property.¹⁸ In prominently cited terms, the Supreme Court’s 1879 Trade-Mark Cases illustrate what has been regarded by later courts and legal scholars as the final stage of the development of a “whole system of trademark property”:

The right to adopt and use a symbol or a device to distinguish the goods or property made or sold by the person whose mark it is, to the exclusion of use by all other persons, has been long recognized by the common law and the chancery courts of England and of this country, and by the statutes of some of the States. It is a property right for the violation of which damages may be recovered in an action at law, and the continued violation of it will be enjoined by a court of equity, with compensation for past infringement.¹⁹

When other courts added that trademark property conferred “an exclusive right good ‘as against all the world,’”²⁰ the concept of trademark-as-property protection seemed to have gained universal hold.

One caveat is worth mentioning, though. Mark McKenna has recently raised doubts as to whether the distinction between actions at law and actions in equity is as clear-cut as it appears.²¹ Nineteenth-century courts often used concepts of law and equity interchangeably, discussed the same precedents for different concepts, and spoke in the same terms regardless of the form of action. A distinction was—and is—therefore difficult to draw.²² McKenna is right, and there is additional indicia suggesting that the terminology of “trademark property” was not as widely established throughout legal practice as has sometimes been posited. In 1857, for instance, the court in Collins Co. v. Brown insisted that it was “now settled law that there is no property whatever in a trade-mark.”²³ Similar doubts can be found in other decisions.²⁴ Adoption of the property paradigm was often more a result of common sense and concrete case facts than of doctrinal necessity and reason.

Nevertheless, one thing remains for us to conclude. We can state indisputably that what had started as legal action on the basis of fraud gradually grew into a system of subjective rights protection. At the end of the nineteenth century, trademark law was on its way toward recognizing the individual rights character of trade names and marks.

B Passing Off: “The Whole Law and the Prophets on the Subject”

At first glance, the areas of trademark protection and unfair competition prevention—like the domains of law and equity—seem to have been clearly separated. However, the dichotomy between the protection of trademark “property” and the prevention of unfair competition “conduct” was superficial. Unlike German law, US doctrine was never strictly divided into two distinct sectors. Goodwill protection was and remains the common denominator.

As in European doctrine, the earlier development of trademark protection in the United States had led to an initial dichotomy within the field.²⁵ Formally, the distinction between technical trademarks and trade names (or “rights analogous to trademarks”) was what drew the line. There was a general agreement in early doctrine that some indicia would always be considered common property. In the 1883 case Avery & Sons v. Meikle & Co., the court expressed this understanding:

The alphabet, English vocabulary, and Arabic numerals, are to man, in conveying his thoughts, feelings, and the truth, what air, light, and water are to him in the enjoyment of his physical being. Neither can be taken from him. They are the common property of mankind, in which all have an equal share and character of interest. From these fountains whosoever will may drink, but an exclusive right to do so cannot be acquired by any.²⁶

Accordingly, while everyday words and symbols were considered off-limits for private appropriation, words and symbols that were of a new and unknown structure or usage were not. This category of technical trademarks—or trademarks proper, as it evolved during the nineteenth century—was capable of private appropriation.²⁷ Under today’s trademark doctrine, this category comprises arbitrary, fanciful, invented, distinctive, and nondescriptive trademarks. Their illegitimate appropriation was a tort, and injunctive relief was available upon showing that the defendant had made use of an identical or similar trademark for the same product.²⁸ Quite differently, the protection of designations other than technical trademarks—namely, trade names; personal, corporate, and firm names; and geographical and descriptive terms—was not founded on a theory of formal property rights. These designations were deemed nonprotectable within the category of technical trademarks.²⁹ Yet protection was possible under a doctrine of unfair competition prevention, notably as “cases analogous to trademarks.”³⁰ Over time, state and federal courts extended this doctrine of unfair competition to comprise ever more instances of unfairness. Ultimately, a wide range of unfair competitive conduct was covered.³¹

Even though, at that time, it seemed as if a line had been drawn between property and fairness protection, we must question whether this dichotomy ever actually existed. Despite the lack of formal property in unfair competition doctrine, protectable rights could be acquired by showing that the plaintiff had established secondary meaning.³² In this regard, although property doctrine had not absorbed nontechnical trademarks, the general distinction between technical trademark property and unfair competition prevention was not well defined—and, in fact, was widely ineffective. Some courts were even willing to also find property rights in trade names and other nontechnical trademarks. One example is the 1904 case Sartor v. Schaden, in which the Supreme Court of Iowa started with a general recognition that “[t]here is a well-marked distinction between what is known as the ‘infringement of a trade-mark’ and ‘unfair competition.’ ” The court explained that a trademark would be the “exclusive right of its proprietor.” With regard to nontechnical trademarks, it stated:

[A]side from the law of trade-marks, courts will protect trade-names or reputations, although not registered or properly selected as trademarks, on the broad ground of enforcing justice and protecting one in the fruits of his toil. This is all bottomed on the principle of common business integrity, and proceeds on the theory that, while the primary and common use of a word or phrase may not be exclusively appropriated, there may be a secondary meaning or construction which will belong to the person who has developed it. In this secondary meaning there may be a property right.³³

The last part of this illustration, a concept of secondary-meaning-as-property protection, would later return in other court decisions and scholarly commentary.³⁴ Without belaboring the point, a basic fact is eye-catching: both sectors were founded on the principle that no competitor had a right to pass off her goods as those of another.³⁵ The prevention of passing off was designed to protect against the improper invasion of goodwill.³⁶ And impropriety was found in consumer confusion. James Love Hopkins described this in 1905:

Unfair competition consists in passing off one’s goods as the goods of another, or in otherwise securing patronage that should go to another, by false representations that lead the patron to believe that he is patronizing the other person.³⁷

As he went on, “The principles involved in trademark cases and tradename cases have been substantially identical.”³⁸ Even though the facts that a plaintiff had to prove may have been different, the common foundation of all cases was the diversion of trade by misinformation. This has remained the touchstone of both fields in the United States ever since.³⁹ As Judge Learned Hand famously stated in his 1928 Yale Elec. Corp. v. Robertson opinion:

The law of unfair trade comes down very nearly to this—as judges have repeated again and again—that one merchant shall not divert customers from another by representing what he sells as emanating from the second. This has been, and perhaps even more now is, the whole Law and the Prophets on the subject, though it assumes many guises.⁴⁰

This common foundation of trademark and unfair competition law also surfaces with regard to the debate on the fields’ interrelation. For quite some time, it was unclear whether trademark law was part of the domain of unfair competition prevention, or vice versa. One reason the issue was so vexing was that, on the basis of the fields’ common principle, either trademark or unfair competition law could be duly characterized as the fundament.⁴¹ And even though the question was formally answered by the Supreme Court in Hanover Star Milling Co. v. Metcalf in 1916, the homogeneity of policies has remained a critical point until today. As the Supreme Court majority explained, “the common law of trademarks is but a part of the broader law of unfair competition.”⁴² Repeating what had been established under nineteenth-century English precedent, the court emphasized that “[the] essential element is the same in trademark cases as in cases of unfair competition.” In particular, the court observed:

Courts afford redress of relief upon the ground that a party has a valuable interest in the good will of his trade or business, and in the trademarks adopted to maintain and extend it. The essence of the wrong consists in the sale of the goods of one manufacturer or vendor for those of another.⁴³

C Kidd/Derringer: Trademark Universality “US Style”

As illustrated in chapter 1, German law in the nineteenth century widely adhered to the idea of international trademark universality.⁴⁴ A look at what the US courts did at that time—notably how they interpreted the geographical scope of trademark rights protection and what they understood as rights universality—sheds a very different light on the issue. Curtis A. Bradley has argued that the universality theory was never “embraced wholesale” by US courts. Since the Supreme Court, under its Tea Rose/Rectanus doctrine,⁴⁵ had early on limited a trademark’s scope of protection to the territory of its use, European-style universality never came into existence.⁴⁶ However, case law prior to Tea Rose/Rectanus suggests a different picture—one of virtually unlimited rights extension and trademark universality. Here, as in Germany, the boundlessness of nineteenth-century property doctrine actually did account for an interim peak in trademark extension.

Essentially, nineteenth-century trademark protection is part of contemporary legal doctrine on the creation of rights in nonphysical values.⁴⁷ As with other kinds of intangible value protection under the guise of formal “property” rights, trademark policy was designed to accommodate pressing socioeconomic interests in a preindustrialized country. Both scholarship and practice agreed that valuable interests had to be protected, regardless of whether the form of wealth was tangible or intangible.⁴⁸ Political consensus was that the protection of investment had priority within a society and economy faced with the challenges of industrializing a scarcely populated continent. In this regard, it was contended, legal certainty and predictability were necessary to encourage economic activity.⁴⁹ In many cases, such an extension of investment protection could be achieved only by jettisoning the Blackstonian conception of property as overly physicalist. If no physical or material thing to be protected existed, the interest or value at issue would have to be fictionalized as a position of “property.” Such an extended conception of intangible values, of course, confronted the most basic problem of property theory: the unrestricted protection of an individual’s property was impossible without a correspondingly absolute limitation of other individuals’ freedom of activity.⁵⁰ This absolute doctrine was impractical at best—and detrimental and immoral at worst. Over time, therefore, any and all positions of property had to be limited. For fictionalized matter, the restrictions were “invisible,” and, hence, there was endless matter for dispute. Accordingly, legal practice was often based on a trial-and-error approach rather than a structured and consistent system of property rights and limitations.

With respect to trademarks, legal practice reflects the judiciary’s struggle in a number of different ways. One example is the dichotomy between technical trademarks and nontechnical rights. What had begun as a quasi absolute concept of trademark-as-property protection was gradually downsized on a sliding scale of protection. In the end, as we have seen,⁵¹ courts distinguished between a highly competition-sensitive area of nontechnical trademarks (e.g., descriptive or geographic indications), where market competition depended on maximum availability, and the area of technical trademarks, where the risk of monopolization if a trademark was appropriated was not deemed too pressing.⁵² With respect to the interstate economy and its marketplaces, another modification was required regarding the geographic extension of rights. Here, too, an initially absolute dominion of rights protection had to be broken down over time. The California Supreme Court’s 1865 Derringer v. Plate⁵³ case and the US Supreme Court’s 1879 Kidd v. Johnson⁵⁴ decision illustrate the difficult correlation between absolute rights and an unrestricted geographical protection.

Kidd centered on a trademark for whiskey. The dispute arose over concurrent trademark use in the owner’s initial place of business in Cincinnati (by his distillery’s purchasers) and in New York (by his removed business). The Supreme Court’s characterization of trademark rights, though deftly short, expresses the contemporary concept of absolute and exclusive trademark rights:

The right to use the trade-mark is not limited to any place, city, or State, and, therefore, must be deemed to extend everywhere. Such is the uniform construction of licenses to use patented inventions. If the owner imposes no limitation of place or time, the right to use is deemed coextensive with the whole country, and perpetual.⁵⁵

The Derringer⁵⁶ decision of the California Supreme Court was similarly unrestricted in its approach to the geographical scope of protection. The plaintiff, a resident of Philadelphia, sold pistols under his trademark, “Derringer, Philadel.” The defendant manufactured similar pistols in San Francisco, and he employed the plaintiff’s trademark. Under the heading “Right to a trade mark at common law,” the California Supreme Court explained:

[The] right to the trade mark accrues to [the trademark owner] from its adoption and use for the purpose of designating the particular goods he manufactures or sells, and although it has no value except when so employed, and indeed has no separate abstract existence, but is appurtenant to the goods designated, yet the trade mark is property, and the owner’s right of property in it is as complete as that which he possesses in the goods to which he attaches it …. [D]octrine has been uniform for many years, that the manufacturer or merchant does possess an exclusive property in the trade mark adopted and used by him. … [L]ike the title to the good will of a trade, which it in some respects resembles, the right of property in a trade mark accrues without the aid of the statute. The right is not limited in its enjoyment by territorial bounds, but subject only to such statutory regulations as may be properly made concerning the use and enjoyment of other property, or the evidences of title to the same; the proprietor may assert and maintain his property right wherever the common law affords remedies for wrongs. The manufacturer at Philadelphia who has adopted and uses a trade mark, has the same right of property in it at New York or San Francisco that he has at his place of manufacture.⁵⁷

The last part of the court’s argument in particular provided room for divergent interpretation. While it was widely acknowledged that common law trademark protection extended beyond areas of trading activity, it was not clear how far such protection would reach. A broad interpretation projected trademark rights beyond state and even national borders. As long as the jurisdiction at issue granted trademark protection under a common law system, trademark rights detached from their origin jurisdiction could be protected.⁵⁸

What ultimately has proven more important, however, is something else. The concept of unlimited trademark rights was difficult to uphold in a world of expanding marketplaces. As had become increasingly evident, the overextension of property rights affected the public good. With the advent of transcontinental trade and commerce, the issue of protecting good-faith market investment progressively acted as a counterbalance to formal trademark property. Consequently, the principle of strict priority combined with potentially unlimited trademark protection was no longer adequate.⁵⁹

II The Right/Markets Connex: Materialization, Goodwill, and Trade Diversion

At first glance, it may appear that German and US trademark and unfair competition doctrine underwent similar processes of de-ideologization. Indeed, Josef Kohler, in a comparative account of US and European law, actually described the United States’ property paradigm as equivalent to his theory of personality rights protection. With only a trace of arrogance, he explained:

In France, England, and America [reference to Derringer case], protection of the individual right of product designations is considered an emanation of general principles; and the merit of this perspective is not lessened by the fact that these regimes often operate with the category of property rather than with the category of individual right, for construction—as is well-known—is not the most valuable asset of these regimes. As with Roman law, their major aplomb is the secure manner in which their jurisprudence finds its way through all troubles, regardless of the momentary system and the possibilities of rational-juridical reason—and a good jurisprudence with wrong arguments is still ten times better than a bad jurisprudence with good arguments.⁶⁰

In this light, one might have expected the Kidd/Derringer doctrine to be jettisoned in the same way that personality rights universality was rejected in Germany. After all, in both countries, unlimited geographical trademark protection had become increasingly inapt at regulating expanding marketplaces. But US law took a different turn. Unlike German doctrine, American legal thought did not shrink rights geographically to the owner’s place of business.⁶¹ Instead, the subject matter of protection was transformed. Goodwill remained the foundational concept, and the diversion of trade became its practical metric. The Supreme Court’s Tea Rose/Rectanus doctrine established protection against goodwill invasion in accordance with the parties’ marketplace activities and investment. Trademark and unfair competition doctrine thereby first became detached from the competitor’s place of business or residence, and then from the state’s territory.

A The Materialization of Trademark Rights

The detachment of trademark rights from both their owner’s personality and from the place of business is characteristic of US law. While in Germany a trademark remained connected to its owner’s business and state territory, US doctrine established a model of market-related rights; neither personality nor business place determined a right’s location. This attachment of trademark goodwill to the marketplace has proven significant for conflicts law.

As Kidd and Derringer illustrate, nineteenth-century doctrine conceived of trademark rights as providing protection against the entire world.⁶² Over time, the subject matter of protection was shrunk. The emphasis shifted to actual commercial activity. This development, however—from personal rights and rights attached to a place of business into a scheme of marketplace rights—did not occur instantaneously. Early definitions of goodwill in scholarship still focused on a localization of values in at least some tangible element of the business.⁶³ This corresponded to an environment of local communities and local trade where goodwill was attached to individuals or small businesses.⁶⁴ Joseph Story’s mid-nineteenth-century definition of goodwill (frequently referred to in later trademark treatises and commentaries) specified the establishment of a business as a determinative factor. He defined goodwill as

the advantage or benefit, which is acquired by an establishment, beyond the mere value of the capital, stock, funds, or property employed therein, in consequence of the general public patronage and encouragement, which it receives from constant or habitual customers, on account of its local position, or common celebrity, or reputation for skill or affluence, or punctuality, or from other accidental circumstances or necessities or even from ancient partialities, or prejudices.⁶⁵

Later scholars, building on this definition, referred to business-owner personality as the foundation of goodwill value. A. S. Biddle, for instance, posited in 1875 that goodwill was “a species of incorporeal personalty, … subject with but few exceptions to the general laws which regulate that kind of property.”⁶⁶ In this regard, scholarly opinion in the United States still resembled the contemporary German doctrine of personality rights protection. Yet the foundation on personality rights never completely took hold in the United States, to the contrary. By 1883, for instance, Adelbert Hamilton had explained the concept of goodwill as being founded on the business as such: “Good-will denotes a relation existing between a man or firm and the public with reference to a particular business. It is the good-will of the public to the man or firm.”⁶⁷

This early separation of goodwill value from an owner’s personality and a business’s physical existence was implemented in practice as well. Over time, courts shifted the focus of protectable subject matter to all instances where a plaintiff’s investment in general was at issue. As a result, the need for a tangible thing to support or to connect to the intangible interest or value was gradually abandoned.⁶⁸ At the beginning, English courts still interpreted goodwill as being founded on incidents of real property. One example is Lord Eldon’s definition of “goodwill” in the 1810 case Cruttwell v. Lye, where he explained that “good-will … is nothing more than the probability, that the old customers will resort to the old place.”⁶⁹ Indeed, courts in both England and the United States went on for some time to describe goodwill as an appendage of real property, particularly the place of business.⁷⁰ This tangibility, however, faded toward the end of the century. The Supreme Court’s 1893 decision in Metropolitan Bank v. St. Louis Dispatch Co. illustrates the shift. Starting with the general position that goodwill “is tangible only as an incident, as connected with a going concern or business having locality or name,” the court went on to describe the goodwill of a newspaper company:

As applied to a newspaper, the good will usually at[t]aches to its name, rather than to the place of publication. The probability of the title continuing to attract custom in the way of circulation and advertising patronage gives a value which may be protected and disposed of, and constitutes property.⁷¹

The US Court of Appeals for the Second Circuit added in 1897:

Nor is [goodwill] indissolubly connected with any particular locality, or any specific tangible property. … If good will be a “parasite,” it is a “parasite” of the business from which it sprung, not of the mere machinery by which that business was conducted.⁷²

These and similar cases⁷³ marked the end of a line of decisions that led trademark and unfair competition doctrine to radically detach value protection from both tangible business assets and personality. In this regard, the US conception of business goodwill (unlike the static understanding in contemporary German doctrine) evidenced a genuinely economic analysis. Goodwill was, as J. Roberton Christie explained in 1896, “the aggregate advantages arising from the business connection, reputation, and favourable situation of an established trading concern.”⁷⁴ Customer relations and the public’s favorable regard became the central aspect.⁷⁵ More concretely, it was the likelihood that customers would repeatedly return to a business or product that was seen as determinative.⁷⁶ English doctrine later came to characterize this phenomenon as “dog” goodwill, because dogs (unlike cats) are loyal to their owners.⁷⁷ In the United States, the same was expressed by reference to a “probable expectancy” of attracting the consuming public.⁷⁸ Ultimately, it was the information capital accumulated by performance and advertising investment in the marketplace that accounted for the scope of goodwill.⁷⁹

We can thus conclude that, over time, the American conception of trademark goodwill grew less attached to productive resources and more attached to the marketplace. The customer became the ultimate reference point. Quite differently, German legal doctrine at the time still adhered to a static concept of owner-centered rights protection. There, neither trademark nor unfair competition law were ever fully emancipated from personality rights protection.⁸⁰ Part of this distinction between German and US trademark doctrine has endured until today. As we will see in the following, it was the peculiar transformation of goodwill into a subject matter of market relations that particularly influenced the formation of US conflicts law.⁸¹

B The Reverse Picture: Trade-Diversion Prevention

While, formally speaking, the trademark right was always at the center of the plaintiff’s claim, the real object of protection was the business’s goodwill against invasion. The trademark as such was rarely characterized as the property right itself; indeed, in 1879, the Supreme Court clarified that words or symbols could not be the object of protection.⁸² As Edward S. Rogers explained in 1909, “Each [tort] is a trespass upon business goodwill,”⁸³ and “every trader has a property in the good will of his business, that he has the right to the exclusive benefit of this good will.”⁸⁴ At stake in both trademark and unfair competition disputes, therefore, was an injury to the plaintiff’s business relations. In practice, actionable goodwill invasion was most conveniently found in cases of stealing customers, attracting patronage, or diverting trade. Indeed, court rulings regularly indicated that even the potential to divert trade was sufficient. For instance, in the 1845 case Coats v. Holbrook, Nelson & Co., the New York Court of Chancery enjoined product imitation by a competitor, providing the following explanation:

A man … has no right, and he will not be allowed, to use the names, letters, marks, or other symbols, by which he may palm off upon buyers as the manufactures of another the articles he is selling, and thereby attract to himself the patronage that without such deception, use of such names, &c., would have enured to the benefit of that other person who first got up, or was alone accustomed to use such names, marks, letters or symbols.⁸⁵

Around the same time, in the 1849 case Amoskeag Manufacturing Co. v. Spear, another New York court said:

He who affixes to his own goods an imitation of an original trade-mark, by which those of another are distinguished and known, seeks, by deceiving the public, to divert and appropriate to his own use the profits to which the superior skill and enterprise of the other had given him a prior and exclusive title. … [T]he owner is robbed of the fruits of the reputation that he had successfully labored to earn.⁸⁶

Numerous examples can be found in subsequent case law.⁸⁷ In addition, scholarly commentaries identified trade diversion as an indicator of illegitimately caused injury or harm. A particularly instructive explanation can be found in Hopkins’s 1905 edition of The Law of Trademarks, Tradenames, and Unfair Competition:

Unfair competition consists in passing off one’s goods as the goods of another, or in otherwise securing patronage that should go to another, by false representations that lead the patron to believe that he is patronizing the other person. … It is apparent that the simplest means of depriving another of the trade he has built up is to copy the marks he places on his merchandise. This is the easiest method of stealing his trade, and most universal because of the general use of marks or brands upon personal property. The use of such marks runs far back into the shadows of history …. It is only natural that these marks used in trade, or trademarks, should have first become the subjects of judicial consideration, and that the law concerning them should have reached a state of comparatively complete development before infringers began to employ other and more obscure means to divert trade.⁸⁸

Among the most prominent twentieth-century decisions concerning the question of whether early trademark doctrine sought to protect consumers against fraud and deception is the Seventh Circuit’s 1912 case Borden Ice Cream Co. v. Borden’s Condensed Milk Co. As is commonly known, the court rejected a theory of consumer protection. Its reasoning, however, also illustrates the dominant perception of trade diversion at the time:

It has been said that the universal test question in cases of this class is whether the public is likely to be deceived as to the maker or seller of the goods. This, in our opinion, is not the fundamental question. The deception of the public naturally tends to injure the proprietor of a business by diverting his customers and depriving him of sales which otherwise he might have made. This, rather than the protection of the public against imposition, is the sound and true basis for the private remedy.⁸⁹

Frank I. Schechter summarized the relevance of trade diversion in his 1927 analysis of English and US unfair competition and trademark law: “‘The diversion of custom’ is the gravamen of the action in either ‘passing off’ or ‘unfair competition.’”⁹⁰

As this summary reveals, trade diversion constituted an essential element of common law doctrine—and it became particularly determinative with regard to the localization of infringements in conflicts law. This is another striking difference from German doctrine, in which the place of conduct or the victim-competitor’s place of business determined the applicable law.⁹¹ Seen in this light, it becomes evident that German doctrine prior to the 1960s was virtually devoid of the considerations that American courts and scholars had undertaken much earlier. Localization of the customer base and the place where “lost transactions” would occur were of secondary concern at best. In the United States, by contrast, the marketplace became the governing paradigm in 1916, with the Supreme Court’s introduction of a new doctrine on the geographical scope of trademark rights.

C Tea Rose/Rectanus: The Doctrine of Market-Based Rights

Indeed, the Supreme Court’s Tea Rose/Rectanus doctrine marked the turning point for common law trademark rights’ geographical protection. With a doctrinal shift, the court (in two decisions of 1916 and 1918⁹²) ultimately curbed the extension of trademark rights, which had for a long time been interpreted as virtually unlimited. Tea Rose/Rectanus, though not inventing a completely new rule, provided the foundation for the modern concept of immediately market-based rights. The Supreme Court’s majority opinion pointed out the “fundamental error of supposing that a trade-mark right is a right in gross or at large” and stated that “[t]here is no such thing as property in a trade-mark except as a right appurtenant to an established business or trade in connection with which the mark is employed.”⁹³ The following discussion illustrates how the court deconstructed contemporary substantive law. The international effects of Tea Rose/Rectanus will be addressed later.⁹⁴

Hanover Star centered on a dispute over the “Tea Rose” trademark. This trademark had been used by three parties, each of which claimed exclusive rights. The dispute resulted in two lawsuits, one in Alabama and one in Illinois.⁹⁵ The facts of the case are complex, but a short illustration suffices here. Essential to note is the fact that the parties’ areas of trademark use never overlapped geographically. In addition, the second-comer’s use of the trademark was coincidental, not in bad faith.⁹⁶ The first party, Allen & Wheeler Co., had started manufacturing flour under the “Tea Rose” trademark in Ohio in 1872. The company was able to demonstrate significant sales under this trademark only north of the Ohio River, not in the southern states of Georgia, Florida, Alabama, or Mississippi. The second party, Hanover Star Milling, had adopted a similar trademark—“Tea Rose”—in good faith in 1885 and was extensively advertising and marketing its flour under this trademark in Alabama and other southern states, particularly Florida and Georgia. The third party, Metcalf, was a retail seller of flour in Alabama that was produced by another party, yet also marketed under an identical “Tea Rose” trademark. Allen & Wheeler alleged trademark infringement against Hanover. The latter sued Metcalf for trademark infringement and unfair competition. Metcalf, inter alia, contested Hanover’s allegedly exclusive rights by reference to a prior use by Allen & Wheeler. The Supreme Court granted certiorari, and Hanover prevailed in both disputes.

The majority opinion, which began with the finding that neither party had a registered trademark, started its analysis on the basis of general common law: “Nor does it appear that in any of the states in question there exists any peculiar local rule, arising from statute or decision. Hence, the cases must be decided according to common law principles of general application.”⁹⁷ Under the principles of federal common law, the court repeated its prior characterization of trademarks as property rights.⁹⁸ At the same time, it limited the scope of protection by reference to the trade and market relevance of trademark functions:

[I]t is plain that in denying the right of property in a trademark it was intended only to deny such property right except as appurtenant to an established business or trade in connection with which the mark is used. … In short, the trademark is treated as merely a protection for the good will, and not the subject of property except in connection with an existing business. …⁹⁹

That property in a trademark is not limited in its enjoyment by territorial bounds, but may be asserted and protected wherever the law affords a remedy for wrongs, is true in a limited sense. Into whatever markets the use of a trademark has extended, or its meaning has become known, there will the manufacturer or trader whose trade is pirated by an infringing use be entitled to protection and redress. But this is not to say that the proprietor of a trademark, good in the markets where it has been employed, can monopolize markets that his trade has never reached, and where the mark signifies not his goods, but those of another. We agree with the court below … that “since it is the trade, and not the mark, that is to be protected, a trademark acknowledges no territorial boundaries of municipalities or states or nations, but extends to every market where the trader’s goods have become known and identified by his use of the mark. But the mark, of itself, cannot travel to markets where there is no article to wear the badge and no trader to offer the article.”¹⁰⁰

In the end, the majority rejected the interpretation that territorially unlimited trademark protection had been established under Kidd v. Johnson and Derringer v. Plate.¹⁰¹ The geographical area of a trademark’s protection could never exceed the reach of the trade in which the mark was used.¹⁰² At this point, I will not address the question of whether the majority denied the relevance of political borders.¹⁰³ Important here is that the new doctrine was tangibly market oriented. Trade and commerce were to determine the scope of the market—goodwill would be deemed to extend only so far.

In United Drug Co. v. Theodore Rectanus Co.,¹⁰⁴ the other half of the Tea Rose/Rectanus doctrine, the trademark “Rex” was used by Ellen Regis, a Massachusetts resident, for medicine starting in 1877. The business was continued locally as a partnership with her son, and the trademark was recorded. In 1911, United Drug purchased the company with all trademark rights. Meanwhile, around 1883, Kentucky druggist Theodore Rectanus started using, in good faith, the same trademark for medicinal preparations. His use was limited to the city of Louisville and its vicinity; the same was true for the respondent purchaser who later acquired Rectanus’s business in 1906. United Drug did not sell the first “Rex” products in Louisville until 1912. In its decision, the Supreme Court rejected a theory of trademark infringement, particularly the contention that a business owner having started trademark use in one place would be protected against second-comers if she subsequently decided to extend her trade:

The asserted doctrine is based upon the fundamental error of supposing that a trade-mark right is a right in gross or at large, like a statutory copyright or a patent for an invention, to either of which, in truth, it has little or no analogy. … There is no such thing as property in a trade-mark except as a right appurtenant to an established business or trade in connection with which the mark is employed. The law of trade-marks is but a part of the broader law of unfair competition; the right to a particular mark grows out of its use, not its mere adoption; its function is simply to designate the goods as the product of a particular trader and to protect his good will against the sale of another’s product as his; and it is not the subject of property except in connection with an existing business.¹⁰⁵

The court—once again—rejected the idea of trademark rights’ extension beyond the actual marketplace:

It results that the adoption of a trade-mark does not, at least in the absence of some valid legislation enacted for the purpose, project the right of protection in advance of the extension of the trade, or operate as a claim of territorial rights over areas into which it thereafter may be deemed desirable to extend the trade. And the expression, sometimes met with, that a trade-mark right is not limited in its enjoyment by territorial bounds, is true only in the sense that wherever the trade goes, attended by the use of the mark, the right of the trader to be protected against the sale by others of their wares in the place of his wares will be sustained.¹⁰⁶

Of course, the Supreme Court did not completely jettison the paradigm of absolute property rights in trademarks. As Kenneth J. Vandevelde has pointed out, the Hanover Star majority still preserved a potentially absolute and unlimited concept through a flexible application of the estoppel doctrine: a first user could not claim trademark rights in a geographical area where she had failed to extend her commercial activity; the formal ground for rights limitation here was abandonment.¹⁰⁷ This juggling with formal doctrine, however, did not mean that the conception of rights’ extension and scope of protection had remained unaltered. Even though the Supreme Court literally upheld the idea of trademark property, the paradigm had gained a qualitatively different foundation. The court rejected the absolute extension of trademark rights as an end in itself. Protectable subject matter was limited to what could be found within the marketplace; protection was “coterminous with the market actually covered.”¹⁰⁸ In the wake of the Tea Rose and Rectanus judgments, courts no longer adjudicated on conflicts between abstract rights but instead began to separate different markets.¹⁰⁹

The paradigm of a market/rights correlation is part of a bigger picture. A similar trend has actually been identified regarding the contemporary extension of rights into markets for unrelated goods. Shortly after the turn of the century, courts had also begun to extend trademark protection to separate product markets under a theory that would become known as the Aunt Jemima doctrine.¹¹⁰ In short, this doctrine provided that if there was a likelihood that consumers might be confused about the source of a product, a trademark owner could protect even unexplored markets. As Steven Wilf has pointed out, both Aunt Jemima and Tea Rose/Rectanus reflect the conquest for consumers’ minds.¹¹¹ Under both doctrines, the consumer is the cynosure of market allocation and the delimitation of rights.

Under this perspective, it also becomes clear that Tea Rose/Rectanus reflects a natural law approach. Earlier common law trademark doctrine had regularly made reference to a theory of natural rights protection, similar to the creation of copyrighted works. As Blackstone’s Commentaries stated in 1884, the “right to the exclusive use of distinctive trade marks” was “somewhat analogous to literary copyright” for one reason: similar to literary property, the right to the exclusive use of a trademark was deemed to flow from a natural right to appropriate the fruits of one’s own labor.¹¹² But it was not an act of creation per se that would promulgate the protectable res. Copyrights and patents were (and still are) protected as products of the mind. Common law trademarks, by contrast, were and are protected for their distinguishing function alone, a value that must flow and result from their actual use in the marketplace.¹¹³ Therefore, the creation of trademark rights in a Lockean sense was not a singular act but rather the constant flow of marketing activities. Accordingly, the relevant “fruit of labor” in trademark terms was market investment (under the shorthand of “goodwill”).¹¹⁴ The trademark had become an instrument for securing its owner the benefit of her efforts within the marketplace.¹¹⁵ Finally, under this perspective, it is also clear that the Supreme Court’s doctrine of marketplace/rights correlation reflects the Lockean no-harm principle. By definition, the principle contradicted unlimited rights extension. When an individual had appropriated an object from the public domain through labor, it was clear that she was not to be deprived of it.¹¹⁶ Anyone could acquire property through labor, but acquisition was limited by the public’s claims in the commons—in other words, property could be acquired only as long as there was “enough, and as good left in common for others.”¹¹⁷ In this regard, the universal acquisition of trademark rights by simple use in one part of the state or national territory was questionable. Tea Rose/Rectanus implemented this concept of not taking more than necessary. According to Locke, “God [had given] the world … to the use of the industrious and rational …; not to the fancy or covetousness of the quarrelsome and contentious.”¹¹⁸ But even the industrious and rational had to respect some limitations:

For as a man had a right to all he could employ his labour upon, so he had no temptation to labour for more than he could make use of. … What portion a man carved to himself was easily seen; and it was useless, as well as dishonest, to carve himself too much, or take more than he needed.¹¹⁹

This last point is also important with regard to the question of how goodwill could be created and accumulated. By the late nineteenth century, advertising had become a progressively important marketing tool.¹²⁰ Theory and practice subsequently came to recognize the connection between goodwill and advertising efforts. Courts and scholars alike agreed that advertising investment, in addition to actual trading in the marketplace, generated goodwill.¹²¹ The New York circuit court’s 1897 case Hilson Co. v. Foster illustrated this point:

Where the goods of a manufacturer have become popular not only because of their intrinsic worth, but also by reason of the ingenious, attractive and persistent manner in which they have been advertised, the good will thus created is entitled to protection. The money invested in advertising is as much a part of the business as if invested in buildings, or machinery, and a rival in business has no more right to use the one than the other.¹²²

Mere advertising of a brand—without actual sales—might not have been enough to generate goodwill. But the use requirement for rights acquisition was still low. The branded product only had to be offered with an intention of continued marketing.¹²³ Hence, advertising could span geographical areas in advance of actual commerce.¹²⁴ Even though this model came under pressure with the advent of radio and television advertising, it would coin the doctrine of common law trademark acquisition for decades to come.¹²⁵ In particular, conflicts resolution was to be significantly influenced by the idea that goodwill transcended national borders upon advertising.¹²⁶

III The Realist Attack: Much Ado about … Quite Little

The end of the nineteenth century was an era of formalism, a time of “mechanical” jurisprudence. In academic fora and courtrooms alike, legal reasoning and decision making were often reduced to a process of deducing mechanical rules from broader principles. The language of the law was paramount, and cases were decided by a rigid adherence to existing precedents—often, as perceived by critics, in the interest of business corporations in their struggle with workers, consumers, or other opposing parties in the market.¹²⁷ The theory and practice of trademark-as-property protection provides one example of such formalism. The doctrinal countermovement to formalism was so-called legal realism—or, more figuratively, the “realist attack.”¹²⁸ Notwithstanding its overall groundbreaking impact on modern legal thought, the realist attack’s practical consequences on trademark doctrine were humble. Property terminology may have been rethought. Yet neither the concept of goodwill nor the idea of private rights protection was replaced or reconceptualized. In fact, one might even conclude that some proponents of realism set the stage for a modern repropertization of trademarks.

A The Turn-of-the-Century Crisis

An oft-enunciated example of the formalist/realist debate was the 1918 Supreme Court case International News Service v. Associated Press.¹²⁹ Daniel M. McClure has aptly characterized the majority’s opinion as a “high water mark of formalist conceptualism in trademark-unfair competition law.”¹³⁰ The court granted the plaintiff a quasi property right to news stories that the plaintiff had written and published and that had been, according to the majority, misappropriated by the defendant, who had rewritten and published the news stories as its own. Particularly famous is their characterization of the defendant’s activities as a “reap[ing] where it has not sown, and … appropriating to itself the harvest of those who have sown.”¹³¹ Justices Holmes and Brandeis each wrote dissenting opinions. Holmes’s critique has been recited ever since:

Property, a creation of law, does not arise from value, although exchangeable—a matter of fact. Many exchangeable values may be destroyed intentionally without compensation. Property depends upon exclusion by law from interference.¹³²

The open conflict between formalist and realist ideas of “property” in International News Service laid the foundation for a modern critique of the classic unfair competition doctrine.¹³³ It is actually not surprising that the realist attack, among other things, targeted the concept of trademark-as-property protection and the various ideas of what competitor goodwill protection should include. Not only had legal scholars been unable to agree on a uniform, consistent, and comprehensive definition of “trademark property” and “goodwill,”¹³⁴ but courts had also been unable to provide for workable standards—and they openly expressed their discontent with the void of theoretical insight and instruction. Indeed, legal thought had not managed to provide a theoretical structure or a practically workable model. An explanation for why certain conduct should be deemed admissible while other instances of business activity should be enjoined was amiss. Not surprisingly, practical outcomes were often unsatisfactorily diverse and imbalanced—while the idea of property rights seemed to overextend protection in some instances, it prevented adequate relief in other cases, even where commercial dishonesty was evident.¹³⁵ One example of the courts’ self-acknowledged desperation is the 1935 case Premier-Pabst Corp. v. Elm City Brewing Co.:

[S]ome have vaguely suggested that a right to a name may be a part of one’s “good will” which is a subject-matter of property from which all others may be excluded. But such an assertion gets us nowhere. For “good will” itself is too loose and uncertain a quantity for aid in definition. As commonly conceived, it is a compound of many factors, and those factors chiefly associated with the concept seem to have little association with rights in a name. Thus value “as a going concern” is frequently considered as a part of “good will.” But such value seems quite distinct from the value attributable to the right to a name. And again, good will is somewhat vaguely considered as the favorable regard of the purchasing public for a particular person, or for goods or services known to the public to emanate from a particular source; a regard founded (usually) on past dealings or reputation and of value in so far as it may be expected to produce further dealings. But good will so construed certainly is not property in any technical sense; for no man can have, either by prescription or contract, such a proprietary right to the favorable regard of the public that he may exclude others therefrom.¹³⁶

Apart from the critique that contemporary formalism was biased toward protecting the corporate haves and disfavoring have-not newcomer and weaker parties, the debate was also seen as illustrating the disciplinary limitations of jurisprudence. With respect to goodwill in particular, the problem was how to “translate” genuinely economic concepts into policies of trademark and unfair competition law. In light of the complexity of real-world market transactions and inter-competitor and consumer-competitor relations, however, the goodwill paradigm was increasingly unmasked as being too unstructured and indeterminate. In other words, the spheres of marketplace economics and of legal doctrine were too far apart to allow for a smooth osmosis of ideas and concepts. As Christie’s 1896 critique of “goodwill” highlighted, “The term was originally one of the market-place rather than of the law courts.”¹³⁷ Accordingly, it seemed that the concept’s time had expired. Indeed, during the first decades of the twentieth century, the classic doctrine of unfair competition had evolved into one of the realist’s favorite bête noires.

B Courts’ Adherence to “Transcendental Nonsense”

Arguably, the most prominent critique of formalism came from Felix S. Cohen. His iconic 1935 Columbia Law Review article, “Transcendental Nonsense and the Functional Approach,”¹³⁸ became world-famous for its cynical account of legal doctrine at the time. Mocking legal formalism and categorization as transcendental nonsense, Cohen described such legal reasoning—especially that regarding trade-name protection—as devoid of a true policy foundation.

His critique started with an explanation of common myths and metaphors employed in traditional jurisprudence in order to masquerade the social forces that were actually molding the law and shaping the outcome of interest conflicts.¹³⁹ As he pointed out, contemporary doctrine’s foundation on self-contained definitions and rules created a system of adjudication isolated from social reality. Per se, the justification and critique of legal rules in purely legal terms meant arguing in a vicious circle.¹⁴⁰ As he further illustrated, again referring to the historical development of trademark-as-property protection, courts and scholarship focused on the protection of intangible values, which ultimately resulted in a “divorce of legal reasoning from questions of social fact and ethical value.”¹⁴¹ In particular, he attacked the shift from passing-off theory and from the concept of preventing deception to a system of property protection. In this regard, he specified the “thingification” of property as the primary evil that circular reasoning had created.¹⁴² In the end, the propertization of consumer responsiveness had become an instrument for creating and distributing a “new source of economic wealth or power.” In short, property had become a perverted function of inequality.¹⁴³

Cohen’s critique not only alluded to the general “monopoly phobia” of the 1930s¹⁴⁴ but also demanded a new understanding of policies concerning marketplace and competition regulation. Clearly, a simple balancing of interests was inadequate. Cases of trade diversion by confusion and the large array of other scenarios of improper misappropriation would no longer fit under the same umbrella of property protection, prevention of unfairness, and trespass. The maze of “economic prejudice masquerading in the cloak of legal logic,” as Cohen suggested, could be lifted only by a clear analysis of socioeconomic factors. The long-perceived homogeneity of policies—traditionally pushed into the catch-all concept of goodwill—was gone:

The prejudice that identifies the interests of the plaintiff in unfair competition cases with the interests of business and identifies the interests of business with the interests of society, will not be critically examined by courts and legal scholars until it is recognized and formulated.¹⁴⁵

Yet, as seen above, despite the fact that this realist critique was compelling and pointed, it did not lead to a sustainable modification of trademark doctrine. By contrast, it appears as if realism ultimately contributed to an ever-deeper implementation of propertization tendencies. A look at case law from the beginning of the nineteenth century sheds some light on this development.

Notwithstanding the growing distrust of formalities, and regardless of the realist acid, courts and scholars continued to rely on the concept of goodwill and property rights protection. The eradication of meaningless concepts that Cohen strived for was never achieved.¹⁴⁶ Thus, even today, the concept of goodwill protection remains central to trademark and unfair competition doctrine, and a clear definition of confusion prevention is rarely sought after. Even though critical scholarship identified policies that courts should take into account, legal practice continued to adjudicate on the basis of traditional structures. A look at part of what became known as the so-called Holmes/Hand doctrine of the qualified nature of trademarks illustrates the meandering between modern policy analysis and traditional goodwill protection.¹⁴⁷

Until the 1930s, decisions authored by Justice Holmes and Learned Hand served as major precedents for federal and state courts throughout the United States.¹⁴⁸ I have already alluded to Learned Hand’s famous allegory of “the whole Law and the Prophets on the subject” in Yale Elec. Corp. v. Robertson.¹⁴⁹ As he explained, “The law of unfair trade comes down … to … that one merchant shall not divert customers from another by representing what he sells as emanating from the second.”¹⁵⁰ Hand never accorded significant weight to an understanding of trademark or unfair competition law in terms of property protection. On the contrary, in subsequent cases, he stated that a trademark “never really gives any property in the words themselves” and that “[a] trade-mark is not property in the ordinary sense but only a word or symbol indicating the origin of a commercial product.”¹⁵¹ Nonetheless, his conception of the field still situated individual goodwill at the center of infringement analysis. Revealingly, he amended his explanation in Yale Elec. Corp. by an open individual rights focus—one not too different from the Kohlerian conception of personality rights protection in nineteenth-century German doctrine:¹⁵²

However, it has of recent years been recognized that a merchant may have a sufficient economic interest in the use of his mark outside the field of his own exploitation to justify interposition by a court. His mark is his authentic seal; by it he vouches for the goods which bear it; it carries his name for good or ill. If another uses it, he borrows the owner’s reputation, whose quality no longer lies within his own control. This is an injury, even though the borrower does not tarnish it, or divert any sales by its use; for a reputation, like a face, is the symbol of its possessor and creator, and another can use it only as a mask. And so it has come to be recognized that, unless the borrower’s use is so foreign to the owner’s as to insure against any identification of the two, it is unlawful.¹⁵³

A similar focus on right owners’ concerns coined Justice Holmes’s trademark jurisprudence. Since Holmes was much more of a realist, one could have expected him to be clearer about the fact that trademark protection was an issue of public policy, not of private property. However, his use of terminology also illustrates the entrapment in traditional goodwill terminology and doctrine.¹⁵⁴ Indeed, a look at some of his decisions reveals an inconclusiveness that Frank I. Schechter characterized as an “indication of the shifts and shadings of judicial thought” on the issue of trademark property.¹⁵⁵ Early, when Holmes was on the bench of the Massachusetts Supreme Judicial Court, he explained in Chadwick v. Covell:

When the common law developed the doctrine of trade-marks and trade-names, it was not creating a property in advertisements more absolute than it would have allowed the author of Paradise Lost, but the meaning was to prevent one man from palming off his goods as another’s, from getting another’s business or injuring his reputation by unfair means, and, perhaps, from defrauding the public.¹⁵⁶

Here, though Holmes did not completely reject a property right, the prevention of palming off (in the sense of injury to reputation and business) was the policy behind trademark protection. This understanding also looms in Holmes’s famous good-faith analogy in the Supreme Court’s 1917 decision in E.I. Du Pont De Nemours Powder Co. v. Masland:

The word “property” as applied to trademarks and trade secrets is an unanalyzed expression of certain secondary consequences of the primary fact that the law makes some rudimentary requirements of good faith.¹⁵⁷

But this critical stance seems to have taken a back seat some years later in the 1927 case Beech-Nut Packing Co. v. P. Lorillard Co., when Holmes again made use of “qualified” trademark-as-property and goodwill protection language:

A trade-mark is not only the symbol of an existing good will although it commonly is thought of only as that. Primarily it is a distinguishable token devised or picked out with the intent to appropriate it to a particular class of goods and with the hope that it will come to symbolize good will. Apart from nice and exceptional cases and within the limits of our jurisdiction a trade-mark and a business may start together, and in a qualified sense the mark is property, protected and alienable, although as with other property its outline is shown only by the law of torts, of which the right is a prophetic summary.¹⁵⁸

While it might be overly critical to imply that Holmes’s arguments were vague or meandering with regard to the property concept of trademarks,¹⁵⁹ one thing is evident: his use of terminology never said farewell to the notion of trademark “property.” Most notably, however, the individualistic concept of goodwill protection—which served the interests of right owners above all—was evident in both Learned Hand’s and Holmes’s understandings of trademark protection and unfair competition prevention. This concept dominated trademark doctrine at the time and continues to do so today. Hence, given that even the most prominent critics of legal formalism adhered to traditional terminology, it is not surprising that the realist attack was no true purgatory for trademark and unfair competition doctrine.

C Frank I. Schechter: The Victory of Goodwill

Trademark-as-property and goodwill-as-property terminology was not the only thing that survived. Another aspect is still characteristic of modern law. In fact, the foundation for a shift toward even further goodwill extension was laid by Frank I. Schechter in 1927. Schechter’s article “The Rational Basis of Trademark Protection”¹⁶⁰ is one of the twentieth century’s most influential contributions to trademark doctrine.¹⁶¹ Read together with his doctoral thesis at Columbia Law School, The Historical Foundations of the Law Related to Trade-Marks,¹⁶² published two years earlier, this article laid the foundation for modern antidilution doctrine. Generally, Schechter rejected the concepts of trademark property and goodwill protection. With regard to the protection of trademark property, his 1925 critique stated authoritatively, “To say that a trade-mark is property and therefore should be protected clarifies the situation no more than to say that a trade-mark is protected and is therefore property.”¹⁶³ Similarly, he deconstructed contemporary understanding of trademark goodwill. In “Rational Basis,” he explained:

The true functions of the trademark are, then, to identify a product as satisfactory and thereby to stimulate further purchases by the consuming public. The fact that through his trademark the manufacturer or importer may “reach over the shoulder of the retailer” and across the latter’s counter straight to the consumer cannot be over-emphasized, for therein lies the key to any effective scheme of trademark protection. To describe a trademark merely as a symbol of good will, without recognizing in it an agency for the actual creation and perpetuation of good will, ignores the most potent aspect of the nature of a trademark and that phase most in need of protection. To say that a trademark “is merely the visible manifestation of the more important business goodwill, which is the ‘property’ to be protected against invasion” or that “the good will is the substance, the trademark merely the shadow,” does not accurately state the function of a trademark today and obscures the problem of its adequate protection. … [T]oday the trademark is not merely the symbol of good will but often the most effective agent for the creation of good will, imprinting upon the public mind an anonymous and impersonal guaranty of satisfaction, creating a desire for further satisfactions. The mark actually sells the goods. And, self-evidently, the more distinctive the mark, the more effective is its selling power.¹⁶⁴

Schechter called for a functional understanding of trademark protection. It was the mark’s selling power, which he later also described as “drawing power” or “magnetism,”¹⁶⁵ that formed the subject matter of protection.¹⁶⁶ In his appeal for a new and unmasked look at trademark functions, Schechter found legal practice to be on the right path in extending doctrine beyond the traditional confines of unfair competition. Yet, as he posited, “the process ha[d] been one of making exceptions rather than of frank recognition of the true basis of trademark protection.”¹⁶⁷ It was no longer trade diversion founded on misleading or deceptive conduct that accounted for a doctrine of trademark infringement. He defined a new metric for assessing remediable damage to trademark owners:

The real injury in all such cases can only be gauged in the light of what has been said concerning the function of a trademark. It is the gradual whittling away or dispersion of the identity and hold upon the public mind of the mark or name by its use upon non-competing goods.¹⁶⁸

Against this backdrop, Schechter’s theory of dilution may be duly characterized as an example of contemporary legal realism. His arguments in “Rational Basis” display, as Robert Bone has explained, “all the elements of a typical realist project.”¹⁶⁹ Schechter’s attack on the concept of trademark protection, his critique of property formalism, and his ultimate suggestion for a reconstruction of trademark law expressed a “common impatience with old theories,” mirroring a similar pattern of realist critique in other sectors of the law.¹⁷⁰ His rejection of formal property rights in particular suggests that he was influenced by the realist critique.¹⁷¹

Nevertheless, Schechter did not fully demolish the cathedral of trademark protection—in fact, the opposite is true. Even though Schechter’s invention of trademark uniqueness had made symbols part of the branded “‘goods’ themselves,”¹⁷² he had not broken with traditional goodwill protection doctrine as fundamentally as it appeared.¹⁷³ Indeed, he added yet another facet of goodwill value to the trademark. Under his guidance, trademark law evolved from a tort model to a proprietary model of protection.¹⁷⁴ In the end, his ideas laid the ground for an even wider extension of goodwill protection. Regarding the practical implementation of his ideas, by 1932, New York courts had already begun to refer to Schechter’s ideas,¹⁷⁵ thus providing relief to trademark owners. Furthermore, starting in the 1940s, state legislators had begun enacting antidilution statutes, and by 1995, a federal law on dilution prevention had been enacted.¹⁷⁶

Here, it is not necessary to explore the practical impact of antidilution doctrine on US (or other jurisdictions’) trademark law.¹⁷⁷ For the purpose of this inquiry, however, one aspect is particularly important: Schechter’s theoretical achievement not only helped lift doctrine to a higher level of sophistication¹⁷⁸ but also led to a significant extension of goodwill protection. He extended the value basis of trademark protection. While his approach might appear to have been influenced by the realist critique, his advocation of broad property rights contradicts an interpretation as purely realist. This is the reason why he, even though having pointed out the circularity of contemporary property doctrine, ultimately became a target of Cohen’s critique some years later:

In practice, injunctive relief is being extended today to realms where no actual danger of confusion to the consumer is present, and this extension has been vigorously supported and encouraged by leading writers in the field.¹⁷⁹

Schechter was one of these “leading writers,”¹⁸⁰ along with Harry D. Nims, Milton Handler, and Charles Pickett. Even though he had actually started on the realists’ plane of restricting trademark-as-property theory, his foundation for modern trademark law unhooked protection from the requirement of actual or potential trade diversion by consumer confusion. In the end, his theory is emblematic of the realists’ failure. The relevant conclusion here is as simple as it is sobering: while a formal concept of property protection had already become obsolete by the pre-realist era, the concept of goodwill has resisted all attempts at politicization and functionalization. It is still central to modern doctrine.

IV Modern Theory and Practice: Economic Analysis and Repropertization

Post-realist reconceptualizations could not change the picture, either. After the Second World War, US trademark law underwent a significant transformation. Both court practice and scholarly analyses have become increasingly “economized.” Yet even though these developments actually shed more light on the underlying policies and thus should have restricted the trademark-as-property and goodwill overgrowth, the opposite is true. US trademark law in the twenty-first century has actually attained a level of almost unlimited private property/goodwill dominance.

A The 1946 Lanham Act: Monopoly Phobia Well Cured

Of course, the realist attack was not limited to the language of the law. It also provided the groundwork for a more wide-reaching and fundamental interdisciplinary critique. In trademark law, it was economic theory that seemed to challenge the age-old concepts of trademark-as-property and goodwill protection. Indeed, early on, economists had uttered doubts with respect to trademark protection as such. By 1933, Edward Chamberlin’s Theory of Monopolistic Competition had already formulated a powerful argument against product differentiation through brand loyalty. As he argued, trademarks could be used not only to insulate market shares from price competition but also to create undue barriers to entry for other branded products. Since the trademark owner could differentiate products from competitor products by mere advertising, consumer loyalty would, over time, lead to an isolation from competition.¹⁸¹ As Ralph S. Brown, Jr., explained, advertising would do more than simply inform the consumer—it would “persuade and influence,” creating fake perceptions of product differences and ultimately a “waste of resources.”¹⁸² Consumers would no longer base their decisions on quality and price but on a misguided perception of the brand as distinct from alternative products. In essence, Brown, Chamberlin, and contemporary critics found trademark and unfair competition protection to create monopoly rights, leading to higher prices to the detriment of the consumer.¹⁸³ They advocated for a reduction of protection in order to eliminate monopolistic elements.¹⁸⁴ This approach, adopted sporadically in practice,¹⁸⁵ never fully took hold.¹⁸⁶ Courts acknowledged the general problems of monopolistic trademark rights but did not implement the theory beyond incidents of an occasionally narrower interpretation of the scope of trademark protection in single cases.¹⁸⁷

A deepening of the debate on monopoly phobia¹⁸⁸ is not necessary here. It suffices to state that the Lanham Act’s drafting (even though debated during an era of antitrust critique) was not noticeably influenced by fears of monopoly enlargement. On the contrary, as the Senate Committee Report remarked, “Trade-marks, indeed, are the essence of competition, because they make possible a choice between competing articles by enabling the buyer to distinguish one from the other.”¹⁸⁹ In addition, the concept of investment protection was considered a part of the modern act’s purpose:

To protect trade-marks … is to protect the public from deceit, to foster fair competition, and to secure to the business community the advantages of reputation and good will by preventing their diversion from those who have created them to those who have not.¹⁹⁰

In other words, the Lanham Act did not alter the existing design of common law rights protection at the interstate level. On the contrary, the goodwill paradigm even became ennobled by its inclusion into lawmakers’ statutory policies.¹⁹¹

B The Economization of US Trademark Law

Notwithstanding legislators’ optimism, from the beginning, courts and academics struggled to reconcile the Lanham Act’s rationale with the field’s common law foundations. This was especially due to the divergence between traditional protection patterns and modern concepts of information economics. In fact, Brown’s aforementioned 1948 critique of trademark rights extension was based on an early economic analysis.¹⁹² And it was not long until law and economics theory took over completely. Toward the end of the twentieth century, a wide array of scholarship became dominated by the Chicago school of economics. For modern trademark law, there is little doubt that an economic rationale, most prominently explained by William M. Landes and Richard A. Posner,¹⁹³ has become the most influential theory. One can agree with Barton Beebe that in the United States “[n]o alternative account of trademark doctrine currently exists.”¹⁹⁴

Under the economists’ credo, the function of trademark law is to reduce consumer search costs. Trademark and unfair competition doctrine is part of the law of torts, whose overall purpose is to promote economic efficiency.¹⁹⁵ Each trademark communicates a particular set of information that the consumer does not need to gather herself every time she considers a purchase.¹⁹⁶ By preventing the improper use of trademarks by nonproprietors, the system ensures that consumer reliance on a product’s source is correct. The law thereby fosters the flow of true information in the marketplace. As Stacey L. Dogan and Mark A. Lemley posit, trademarks have “the potential to lead to better-informed customers and more competitive markets.”¹⁹⁷ The reverse side of this protection of information infrastructure is the creation of incentives for producers to maintain or improve quality.¹⁹⁸ In this regard, trademark protection confers a benefit that all property rights provide: a right owner will invest in the creation or improvement of a resource only if she is certain that no one else can appropriate the fruit of her efforts.¹⁹⁹ Suppose that a trademark owner could not be assured that her use of the trademark is exclusive. In this case, she would have to expect dishonest competitors to apply her trademark to lower-quality products. These competitors might charge less and divert patronage from the trademark owner. Such a system would arguably produce no incentives for trademark owners to invest in the quality of their products.²⁰⁰ To avoid misunderstanding, this incentive must be distinguished from the incentive referred to in the field of copyrights and patents. Trademark law provides a strictly limited—one could say conditional—incentive only. Without ongoing investment and marketplace activity, trademark protection will cease to exist. There is no value in the creation of the trademark as such. It is only consumers’ expected behavior within a functioning system of use and protection that opens an opportunity for trademark owners to reap the benefits of investing in quality and reputation. The premium that a right owner can charge for her products is thus not the result of the initial creation or invention of a trademark; it flows from the constant upholding of a certain quality standard and its communication to the public.²⁰¹ By and large, therefore, trademarks are instruments of market information. They are a part of the information infrastructure that connects producer and purchaser and channels the flow of communication in the marketplace.²⁰² Ideally, the protection of transmission structures for correct and true market information is to be understood as the dominant policy of any trademark protection system.²⁰³

C Modern Propertization and Repropertization

Notwithstanding its recent economization, trademark doctrine still contains numerous policies that defy a justification under economic theory. This is due largely to an adherence to traditional patterns of the common law—notably the unimpaired implementation of goodwill protection structures in both common law and modern statutory trademark law. In fact, the Lanham Act drafters did not intend to change the principal doctrinal foundation of use-based rights as developed under the reign of common law. As a result, federal law is still based on common law principles.²⁰⁴ It is thus not much of a surprise that late twentieth-century law brought forward a number of peculiarities that go beyond concerns for consumer protection and information economization, and that these peculiarities found shelter in the paradigm of goodwill protection. In the end, this traditionalist character of federal statutory law may have been a determinative factor for the reinvigoration of property-based trademark doctrine.²⁰⁵

Scholarly commentary has particularly criticized the shift toward an extension of protection beyond the core of immediate trade-diversion-by-consumer-confusion. Under this extended doctrine, for noncompetitive or not directly competitive uses, protecting goodwill no longer needs to be connected to an attempted or actual diversion of trade. In essence, the actionable invasion of trademark rights has become an issue of goodwill misappropriation rather than of the prevention of confusion-caused trade diversion.²⁰⁶ One example of the extension is so-called initial interest confusion. It applies when a second-comer uses a competitor’s trademark to attract the attention of consumers who would not have purchased her product otherwise. Accordingly, the policy of prevention aims at consumer protection in a presale setting—notably at saving the costs of having to search again for the product the consumer had been seeking prior to coming across the confusing information.²⁰⁷ The issue of “confusion” has, however, been detached from the point of sale or transaction and thus also from the consumer’s actual decision making. Therefore, in general, the subject matter of protection is goodwill beyond the search-cost rationale of the economic trademark protection model.²⁰⁸ Similarly, the doctrine of so-called postsale confusion has projected traditional infringement theory away from the point of sale. In postsale confusion cases, the defendant’s product creates a risk of confusion only after the point of sale or transaction. The actionable wrong appears to lie in the confusion of consumers regarding their postsale interaction with a competitor. In these cases as well, goodwill misappropriation theory, not genuine confusion prevention, governs.²⁰⁹ Finally, the protection of merchandising rights and modern antidilution doctrine are further examples where protection has been extended beyond the core of efficiency-based confusion prevention policies.²¹⁰

As all these examples illustrate, twentieth-century trademark law has extended the right owner’s exclusive domain into numerous dimensions far beyond former protection levels. It is no longer market information transmission prior to the consumer’s decision making that determines whether an infringement exists. Traditional confusion theory has lost its once governing status as basic trademark doctrine. Even economic theory, as the dominant approach in modern US law, has ultimately failed to delimit the scope of private rights protection. As it appears, the paradigm of trademark goodwill has reconquered the field and arrived at a stage of almost maximum propertization again.

Section 2 Interstate Trademark and Unfair Competition Law

Early twentieth-century trademark law did more than change the Kidd/Derringer paradigm of rights protection from universality to market-based rights. As a closer look at interstate trademark and unfair competition law of that time illustrates, the Supreme Court’s Tea Rose/Rectanus doctrine reflects a second characteristic that would prove determinative for the development of conflicts law: under Tea Rose/Rectanus, trademark rights were distinctively apolitical. Like mushrooms growing in a forest, common law rights would cross state borders following any market extension initiated by their owner. The perceived interstate universality and homogeneity of the states’ common law provided the ground for the non-territoriality of rights. Initially, this non-territoriality was further solidified under the Swift lens of a federal common law. Even the Erie shift in directions did not do away with a general common law of trademark protection and unfair competition prevention. Ultimately, the conception of virtually borderless, organic market rights had become so deeply implemented in the fundamentals of American trademark and unfair competition law that it would not change colors even by statutory federalization of this law under the 1946 Lanham Act.

I The “Market Universality” of Trademark Rights

As we have seen, similar to Germany’s embrace of trademark and personality rights universality, the United States witnessed an era of absolute protection for trademarks. Unlike German doctrine, however, US trademark law remained an issue of local law for a long time. It took more than seventy years before a uniform federal statute came into place. Yet, even today, one could still claim that US trademark law is a domain of common law rights. This localization of rights has influenced the concept of rights territoriality in particular and, thus, necessarily also trademark conflicts law.

A A. Bourjois & Co. v. Katzel: The One-Way Street of Trademark Extension

Prima facie, the rejection of universal trademark validity was implemented by the Supreme Court’s 1923 decision A. Bourjois & Co. v. Katzel,²¹¹ a case concerning parallel imports. The issue at hand was the reach of foreign trademark rights into the United States, not the extension of domestic trademark rights to foreign territories. The plaintiff held domestic trademark rights, and the defendant had imported branded goods from France. In France, these goods were legitimately sold under the French trademark. Nonetheless, the Supreme Court found the defendant liable for trademark infringement. As the court explained, trademarks were of an explicitly territorial character. A domestic trademark right would reach only as far as the national boundaries, never beyond.²¹²

Courts and legal scholars agree that Katzel marks the end of universality theory in US trademark law and that this rejection established acceptance of the territoriality principle.²¹³ However, the picture is more complex for two reasons. First, strictly speaking, the case concerned only the issue of parallel importation and the validity of foreign trademarks in the United States; it did not consider whether an extraterritorial extension of domestic rights was possible. The situation thus differed significantly from the facts that the Supreme Court had to decide on thirty years later in Steele v. Bulova Watch Co.²¹⁴ At best, therefore, Katzel marks an end to foreign rights’ extraterritorial extension into the United States. It had no impact, however, on the issue of domestic rights’ extraterritorial extension. The restriction of universality was unidimensional. Second, a closer look at the development of trademark conflicts at the interstate level reveals that even though US courts generally adhere to the principle of territoriality in intellectual property conflicts,²¹⁵ what they say and do differs in international trademark law. Here, as we will see, the doctrine of goodwill extension in Tea Rose/Rectanus is critical for international trademark conflicts.

B Tea Rose/Rectanus: The Doctrine of Nonterritorial Rights

As this inquiry has revealed, Tea Rose/Rectanus established a concept of rights acquisition through market occupation and investment.²¹⁶ But this is just one aspect of the doctrine. A second characteristic can be explained as a peculiarity of interstate trademark adjudication and is particularly important for the genesis of conflicts law with respect to international trademark disputes.

Let us recapitulate the Supreme Court’s approach. Concerning the emphasis on market activities, the majority explained that a trademark “extends to every market where the trader’s goods have become known and identified by his use of the mark. But the mark, of itself, cannot travel to markets where there is no article to wear the badge and no trader to offer the article.”²¹⁷ The area of protection could never exceed the reach of the trade.²¹⁸ As we have seen, the understanding of Kidd and Derringer, as it had developed in the meantime, was practically invalidated.²¹⁹ There was no longer any immediate or automatic extension of use-based trademark rights under common law doctrine, and the possibility of universality and extraterritoriality appeared to be spellbound. Yet Hanover Star still offered a new and different version of extraterritoriality. As the majority, agreeing with the court below, explained:

[S]ince it is the trade, and not the mark, that is to be protected, a trademark acknowledges no territorial boundaries of municipalities or states or nations, but extends to every market where the trader’s goods have become known and identified by his use of the mark.²²⁰

This “no territorial boundaries” formulation by Justice Pitney, pitting market foundation against territoriality, can be characterized as Hanover Star’s most significant influence on conflicts doctrine. From this moment on, the concept of market rights would prevail over the idea of territorially limited entitlements. Ultimately, the factual dissolution of political boundaries would prove to have far more drastic consequences in the inter-nation context than at the interstate level.

Hanover Star’s nonterritorial obliviousness is further reflected in the conception of different geographical zones of trademark protection that have been distinguished ever since by reference to the decision.²²¹ In the first zone, the “zone of actual market penetration,” a trademark user has sold goods or rendered services with such intensity that a second-comer’s use of the same mark would create a likelihood of confusion. This zone comprises all areas inhabited by consumers who customarily purchase the goods or services.²²² In the second zone, the “zone of reputation,” a trademark may be so well known among consumers that the use of the mark by more than one party would also create a likelihood of confusion.²²³ Finally, the “zone of natural expansion”²²⁴ covers areas into which the trademark owner has the potential to expand.²²⁵ The zone of actual market penetration and the zone of reputation are based on the concept of preventing consumer confusion.²²⁶ By contrast, the zone of expansion is not so evidently justified by reference to information-economization concerns. Rights of this kind are not genuinely use based. Instead, scholarly analyses have referred to more individualistic and property-based ideas of “room to grow”²²⁷ and “breathing space” for right owners.²²⁸ The last zone in particular reflects the organic nature of goodwill and the inherent unboundedness of rights extension. Not only will local confusion prevention policies prevail beyond political boundaries, but, as the third zone implies, there also is a quasi natural right of goodwill growth and projection.

Hence, under Tea Rose/Rectanus, the legitimacy of rights acquisition and protection is an issue of market dynamics, not of interstate or international politics and sovereignty. For almost a century, infringement analysis in US trademark law—at the local, interstate, and, ultimately, international level—has been an issue of market invasion.

C Holmes Concurring: A “Passive Figurehead” of State Sovereignty

Since legal analysis of this kind requires an economic rather than a political perspective, it is evident that a divergence of markets and political boundaries will rarely be a problem. In particular, such a divergence will not stand in the way of an extension of rights. Indeed, in an interstate setting, the unitary concept of goodwill under Tea Rose/Rectanus may be adequate, particularly if conflicts occur between common law jurisdictions where substantive law is nonstatutory. However, such an understanding of organic goodwill growth had a fundamental flaw from the beginning—a flaw that was foreshadowed by Justice Holmes in his concurring opinion in Hanover Star.²²⁹

Holmes agreed with the majority that trademark rights might extend within a zone of probable expansion. In addition, he further pushed the geographical scope of protection to state boundaries. Yet Holmes’s concept of territorial rights also gave regard to state sovereignty, which had been neglected by the majority. As he explained:

The question before us … is a question of state law, since the rights that we are considering are conferred by the sovereignty of the state in which they are acquired. This seems to be too obvious to need the citation of authority, but it is a necessary corollary of the Trade-Mark Cases …. Those cases decided that Congress cannot deal with trademarks as used in commerce wholly between citizens of the same state. It follows that the states can deal with them, as in fact they sometimes do by statute …, and when not by statute by their common law.

As the common law of the several states has the same origin for the most part, and as their law concerning trademarks and unfair competition is the same in its general features, it is natural and very generally correct to say that trademarks acknowledge no territorial limits. But it never should be forgotten, and in this case it is important to remember, that when a trademark started in one state is recognized in another it is by the authority of a new sovereignty that gives its sanction to the right. The new sovereignty is not a passive figurehead. It creates the right within its jurisdiction, and what it creates it may condition, as by requiring the mark to be recorded, or it may deny.²³⁰

Holmes’s theory of territoriality never gained a foothold in case law or commentary. Courts notably rejected it on the grounds that since markets are not necessarily circumscribed by state boundaries, the extension of rights cannot follow political limitations.²³¹ This critique was no doubt justified with regard to the lack of a solid policy foundation. The mere extension of rights within a granting sovereign’s boundaries without concurrent use of the trademark within the whole territory, as we have seen, defies the market information rationale underlying modern trademark policy.²³² Holmes’s theory was thus subject to the same critique that was to be launched much later against the Lanham Act.²³³

A close reading of the concurrence, however, reveals that his theory is more than a “good in one part of the state, good in all” concept. Holmes’s argument also is one of political segmentation for trademark extension. Whenever the trademark owner’s business crosses state lines, protection will be granted under a different legal regime. Accordingly, the owner’s goodwill and its corresponding trademark protection consist of a bundle of different states’ common law or statutory trademark rights.²³⁴ Goodwill, as Holmes correctly understood, is not a uniform or homogeneous subject matter—it is a patchwork of different goodwill segments. In 1927, Holmes extended this conception of political rights to international trademark law. In Ingenohl v. Walter E. Olsen & Co.,²³⁵ he explained—by reference to Tea Rose/Rectanus—that “[a] trade-mark started elsewhere would depend for its protection in Hongkong upon the law prevailing in Hongkong and would confer no rights except by the consent of that law.”²³⁶

In the interstate context, two years later, Justice Pitney casually put forth an apparently similar understanding. In United Drug, he stated that “[p]roperty in trade-marks and the right to their exclusive use rest upon the laws of the several states, and depend upon them for security and protection.”²³⁷ But this apparent wisdom on the divergence of markets and political territories was never implemented in practice. As Holmes had pointed out, there was no practical necessity to give regard to sovereignty in the interstate context since “[i]n most cases the change of jurisdiction will not be important because the new law will take up and apply the same principles as the old.”²³⁸ Accordingly, the actual consequences of interstate political segmentation of trademark rights and goodwill portions were never drawn. Yet the fact that political boundaries were irrelevant under common law doctrine and at the interstate level did not make it a negligible factor for the international arena. We will see in the following how the concept of common law uniformity contributed to modern international trademark extraterritoriality—notably how the Supreme Court neglected Justice Holmes’s early wisdom on political rights.²³⁹ But first I must illustrate another prominent characteristic of the US state/federal system that has proven critical for trademark conflicts doctrine.

II The Federal Common Law of Trademarks and the Erie Doctrine

As the debate on Holmes’s concurrence unveils, the development of Tea Rose/Rectanus, particularly its virtually apolitical extension of goodwill and trademark rights, has its roots in a distinctive feature of the common law. Just as the distinction between law and equity led to an early propertization of trademark protection, the federal system of the common law under Swift v. Tyson accounts for the development of a widely homogeneous body of trademark cases and a corresponding disregard for states’ substantive law policies.

A The Traditional Hodgepodge of State and Federal Common Law

Prior to the Lanham Act’s enactment in 1946, US trademark and unfair competition law was a conglomerate of federal and state rules.²⁴⁰ Under the doctrine of Swift v. Tyson,²⁴¹ each court had to apply either the law of the respective state (if a state court) or federal law (if a federal court). Accordingly, two separate bodies of case law evolved. While state courts promulgated principles of state common law, federal courts adjudicated on the basis of substantive federal common law.²⁴² Even though federal courts formally acknowledged that substantive rights in trademarks rested on the laws of the states,²⁴³ federal common law was applied in infringement disputes before federal courts.²⁴⁴ Not surprisingly, a consistent and uniform treatment of trademark and unfair competition cases was far from guaranteed. In light of the clutter of state and federal precedents, the resolution of a conflict depended on the forum in which the case landed.²⁴⁵ In addition, early federal statutory trademark law was limited to procedural rules. The 1905 act,²⁴⁶ for instance, provided for federal courts’ jurisdiction in cases involving registered trademarks but left the nature and scope of trademark rights under the domain of the common law.²⁴⁷ As commerce expanded across state lines, diversity-of-citizenship jurisdiction brought more and more cases involving disputes over unregistered trademarks and unfair competition into federal courts. Consequently, cases were increasingly decided by federal courts under rules of federal common law and without regard to state precedents.²⁴⁸ As a result, in the cross-border regime of trademark and unfair competition law, state sovereignty was a matter of negligible concern.

But adjudication in state courts also neglected choice-of-law issues. In particular, local rules on pleading, proof, and presumptions concerning the content of “foreign” laws (i.e., the legal regimes of other states) contributed to this development. While state courts always took judicial notice of forum law, the laws of other states were regarded as fact—these laws had to be pleaded.²⁴⁹ Accordingly, unless established by a legal presumption, the content of foreign laws had to be proven.²⁵⁰ In addition, any presumptions on foreign law were founded on an assumption of common law homogeneity. Hence, it was presumed that a foreign regime would accommodate the general principles of common law. Furthermore, if the forum’s common law differed from other states’ common law rules on a specific issue, and if the foreign common law was also unclear concerning its content, the “general rule [was] that that view of the common law taken by the courts of the forum will prevail in the absence of evidence of contrary rulings by the courts of the foreign State whose law [was] in question.”²⁵¹ In general, however, it was assumed that the common law was about the same everywhere.²⁵² At the turn of the century, Raleigh C. Minor expressed this in his treatise on Conflict of Laws with a simple but universal (and still modern) rationale:

The true basis of this presumption … is to be found in the unwillingness of the courts to deny relief to litigants coming before them, merely for want of a law to administer. Certainly the great weight of authority is in favor of the rule. Nor is it in most instances apt to work any material injustice, since a failure of both parties to present to the court any evidence of the proper foreign law may reasonably justify the court in presuming that neither party finds anything there which would place him in a position more advantageous than he occupies under the lex fori, or which would place his adversary in a less advantageous position. It is not unfair to presume therefore, whatever the real differences may be between the “proper law” and the lex fori, that for the purposes of the case in hand neither party can be injured by the presumption that the two laws are similar.²⁵³

As a consequence, trademark and unfair competition law at that time was governed by a hodgepodge of state and federal common law rules. There was no clear distinction between different states’ laws. Consequently, courts rarely gave regard to questions of choice of law or to the fact that regulatory norms of different sovereigns might diverge. Necessarily, there was also no awareness of trademark territoriality.

B The Erie Impact: The “Passive Figurehead” of State Sovereignty Reloaded

This situation would change fundamentally after the Supreme Court’s 1938 decision in Erie Railroad Co. v. Tompkins.²⁵⁴ As Justice Brandeis famously explained:

Except in matters governed by the Federal Constitution or by acts of Congress, the law to be applied in any case is the law of the state. And whether the law of the state shall be declared by its Legislature in a statute or by its highest court in a decision is not a matter of federal concern. There is no federal general common law. Congress has no power to declare substantive rules of common law applicable in a state whether they be local in their nature or “general,” be they commercial law or a part of the law of torts. And no clause in the Constitution purports to confer such a power upon the federal courts.²⁵⁵

Since the Erie doctrine applied to equitable suits and remedies based on legal rights,²⁵⁶ all trademark and unfair competition cases fell within its scope.²⁵⁷ Prima facie, therefore, Erie appeared to put an end to the existence of parallel state and federal regimes on trademark and unfair competition regulation. Some even predicted that the federal common law on trademarks, as a sophisticated body of case law, would disappear, leaving in its wake an underdeveloped common law of the states.²⁵⁸ Indeed, the invalidation of existing federal common law was seen as a significant hindrance to the development of a comprehensive and sophisticated doctrine in unfair competition law.²⁵⁹ This concern, however, was unjustified.

Shortly after the Erie decision, its application to trademark and unfair competition cases appeared mandatory and comprehensive.²⁶⁰ In the end, however, all attempts to establish a principle of state common law prevalence proved unsuccessful. The Supreme Court’s first trademark case considered after Erie was decided in the same year. In Kellogg Co. v. National Biscuit Co., Justice Brandeis included a footnote justifying the court’s application of federal precedents:

The federal jurisdiction rests on diversity of citizenship … . Most of the issues in the case involve questions of common law and hence are within the scope of Erie …. But no claim has been made that the local law is any different from the general law on the subject, and both parties have relied almost entirely on federal precedents.²⁶¹

In other decisions, even lip service to Erie was amiss. One example is the US Supreme Court’s 1938 case Armstrong Paint & Varnish Works v. Nu-Enamel Corporation,²⁶² in which the court did not refer to state law at all.²⁶³ Similarly, circuit courts were ambiguous about applying Erie to trademark and unfair competition disputes. While, for example, in the 1939 case Sinko v. Snow-Craggs Corp.²⁶⁴ the Court of Appeals for the Seventh Circuit founded its application of equitable principles on both pre-Erie federal court decisions and a Massachusetts state court decision, it adhered strictly to the Erie distinction in Addressograph-Multigraph Corp. v. American Expansion Bolt & Mfg. Co. two years later.²⁶⁵

In addition, courts were insecure in their application of federal and state law to different issues of trademark and unfair competition infringement. Whenever a federally registered trademark was involved, federal law governed procedure and remedies.²⁶⁶ In terms of parties’ “substantive rights,” however, the question was not clear. This issue was contested if, inter alia, the case concerned only intrastate commerce or if both the plaintiff’s and the defendant’s trademarks were unregistered under federal law.²⁶⁷ As it seemed, the applicable law depended much more on the allegedly infringing activities than on the trademark rights at issue.²⁶⁸

Apart from insecurity concerning the reach of Erie, other factors contributed to the factual survival of federal common law. Many states’ case law in the field of trademark and unfair competition law was far less developed than the federal law. The scarcity of state precedents provided federal courts with the discretion to continue adjudicating on the basis of old precedents and to further develop the body of federal common law that had technically been abolished. Furthermore, within the states, pre-Erie case law had often relied on federal precedents and doctrines. In this regard, the federal common law survived under the guise of “state precedents.” Not surprisingly, many federal courts, searching for applicable state law, justified recourse to federal precedents by reference to an alleged identity of rules under both regimes. One example is the Seventh Circuit’s decision in American Photographic Pub. Co. v. Ziff-Davis Pub. Co.:

Although local law applies to unfair competition and common law trade-mark infringement where federal jurisdiction is based on diversity of citizenship, … the applicable local law does not differ from the general common law of trade-marks. Accordingly, decisions of federal courts and other jurisdictions are in point as illustrations of the common law.²⁶⁹

In sum, fears that the United States would become a legal checkerboard of dozens of state regimes on unfair competition repression proved unwarranted.²⁷⁰ Although Erie may have altered the concept of federal trademark law and ultimately spurred the promulgation of federal statutory trademark law,²⁷¹ the existing body of federal trademark case law was never truly invalidated.

Most importantly for this inquiry, with regard to common law uniformity, Erie did not significantly affect the universality of interstate trademark protection and unfair competition prevention. Hence, the inherent extraterritoriality of trademark rights survived.

III The 1946 Lanham Act: An Innovation of Almost Territorial Rights

Even though the 1946 Lanham Act stems from lawmakers’ intent to give registered trademark rights a maximum extension throughout the territory of the United States, nationwide protection is still subject to a number of exceptions that can be traced back to the common law foundations of US trademark doctrine.²⁷² Modern domestic trademark doctrine is thus a system of “almost” territorial rights.

A The Common Law Foundation of Federal Statutory Rights

As described above, for a long time, the only source of rules for trademark protection had been judge-made common law. Beginning in the mid-nineteenth century, case law was gradually amended by states’ statutory rules. Congress enacted the first trademark statute in 1870. Interestingly, the statute was described as part of “[a]n Act to revise, consolidate, and amend the statutes relating to patents and copyrights.” Whether Congress was oblivious to the differences among intellectual property rights is not clear.²⁷³ In 1879, the Supreme Court declared this first statute unconstitutional, thereby clarifying the difference between copyrights and patents on the one hand and trademarks on the other. The US Constitution,²⁷⁴ the court argued, did not give Congress the authority to legislate in the area of trademark law; rather, the field was reserved for the states.²⁷⁵ Correspondingly, the next attempt at federal legislation, in 1881, strictly adhered to the confines of Congress’s authority granted under the Constitution’s trade clause, concerning only the registration of marks that were “used in commerce with foreign nations, or with the Indian tribes.”²⁷⁶

The first broadening of federal trademark protection prior to the Lanham Act occurred in 1905.²⁷⁷ Notwithstanding a new option to federally register, the 1905 act did not alter the existing concept of use-based rights. But registration of a trademark under the act provided standing to sue in federal courts. In addition, the act enabled the plaintiff to enforce an injunction in any US court.²⁷⁸ What was unclear under the 1905 act was whether federal registration would grant preemptive trademark protection beyond the actual area of use. Tea Rose/Rectanus had established a narrow rule of use-based trademark acquisition and protection. Against this backdrop, the reach of federal authority was critically important. Since federal power was limited to the regulation of interstate commerce, it was questionable whether a federal registration could protect against intrastate infringements that were remote from the area of actual use by the owner.

In 1929, the Supreme Court decided on this issue in U.S. Printing & Lithograph Co. v. Griggs, Cooper & Co. In this case, the plaintiff had used the trademark “Home Brand” for food in several states and had registered the mark federally. The defendant had used the word “Home” on similar products and in combination with other words in states where the plaintiff had not done business before.²⁷⁹ The Supreme Court of Ohio had decided that Tea Rose/Rectanus did not apply due to the plaintiff’s federal registration and that the plaintiff’s trademark rights would therefore be “project[ed] … into all the states even in advance of the establishment of trade therein, and … afford full protection to such registrant and owner.”²⁸⁰ The Supreme Court, however, did not see such a preemptive extension of rights beyond the common law basis. Justice Holmes declared:

[N]either authority nor the plain words of the [1905] Act allow a remedy upon it for infringing a trade-mark registered under it, within the limits of a State and not affecting the commerce named. More obviously still it does not enlarge common-law rights within a State where the mark has not been used.²⁸¹

But the Home Brand holding was of limited value for a comprehensive resolution. It concerned only intrastate competition.²⁸² For interstate competition, there was no Supreme Court precedent. Such competition had been an issue a few years earlier in the Second Circuit’s 1916 Bismarck case. The plaintiff owned a federal registration, “Bismarck,” that the defendant had allegedly infringed on by making use of the trademark in several states.²⁸³ As the Court of Appeals for the Second Circuit explained, “The rights which a person obtains by registration of a trade-mark under those statutes are coterminous with the territory of the United States.”²⁸⁴ The Supreme Court had granted certiorari,²⁸⁵ but the case was withdrawn before a decision could be rendered.²⁸⁶ In the 1930s and after, the Bismarck holding was harshly contested in scholarly commentary, mostly by reference to the Supreme Court’s rejection of federal rights extension in the Home Brand case.²⁸⁷ Even though the two cases were not on all fours, dominant opinion at the time seemed to agree that there was no extension of trademark rights ab initio. Hence, registration did not create new rights; instead, it merely recognized preexisting common law entitlements.²⁸⁸ In sum, federal rights protection in the first half of the twentieth century was holey at best. Registration would not grant advance protection against infringements in intrastate commerce. With regard to interstate commerce, the situation was unclear but strongly tended toward the same result.²⁸⁹

It was therefore the Lanham Act of 1946 that created the first comprehensive system of nationwide registration and protection.²⁹⁰ The act formally eliminated the effects of Erie and expanded the scope of trademark protection beyond the zones of protection that had been acknowledged under the common law rules.²⁹¹ Under the act, trademark rights could be established throughout the entire national territory by simple registration, regardless of the registrant’s zone of actual use.²⁹² In this regard, the provision on “constructive notice” in section 22 has been characterized as potentially the greatest advantage of registration.²⁹³ By establishing constructive notice of the registrant’s prior use, the Tea Rose/Rectanus doctrine was cut back. Once the mark was registered, a second-comer’s use could no longer be excused by reference to her good faith and lack of knowledge about the senior trademark.²⁹⁴ By the Trademark Law Revision Act of 1988, these effects were finally extended from registration to mere application.²⁹⁵ For the first time, then, actual use was no longer a prerequisite for rights acquisition.²⁹⁶

Notwithstanding these modernizing amendments, the goodwill paradigm has remained the foundation of federal trademark protection.²⁹⁷ And even though the Lanham Act has been characterized as placing federal trademark law “upon a new footing,”²⁹⁸ trademarks under the act are not aliud to rights acquired under common law. The act has not changed the system’s doctrinal foundation on use-based rights. Federal law as well is based on common law principles.²⁹⁹ In fact, it has even been contended that the act changed nothing at all.³⁰⁰ A look at some characteristic features of contemporary federal law can clarify this point.

As just mentioned, federal application and registration affords nationwide constructive notice of use or constructive use.³⁰¹ This largely prevents trademark right duplication within the United States. In most cases, therefore, a federal trademark owner is protected against other users’ adoption of identical or similar marks in remote areas. Yet common-law-based exceptions still exist. First of all, neither application nor registration of a federal trademark can wipe out another’s common law right acquired by use prior to the date of application.³⁰² In addition, even after a federal trademark application has been filed, a junior user may acquire common law trademark rights by use. In this case, the federal applicant may not be protected from a subsequent user’s adoption of an identical or similar trademark prior to actual registration.³⁰³ By this means, the statutory system restricts the effects of granting nationwide rights. If an independent common law right has been acquired prior to application or even prior to registration, the federal statutory right is ineffective throughout the local area of the preexisting use-based right. And the common law basis of the exception is also reflected in its inherent limitation: upon registration, the federal statutory right “freezes” the locally preexisting common law right in its current territorial expansion.³⁰⁴

The most intriguing example of common law pedigree is the so-called Dawn Donut doctrine. In Dawn Donut Co. v. Hart’s Food Stores, Inc.,³⁰⁵ the Court of Appeals for the Second Circuit established that a right owner is entitled to injunctive relief only if her mark has significance in the market—and such significance can exist only where the right owner actually serves her customers.³⁰⁶ In Dawn Donut, the senior user had federally registered trademarks (“Dawn” and “Dawn Donut”). The junior user, a retail seller of donuts and baked goods, started to use the senior trademark in good faith, serving a geographic market different from that of the senior user. As the court concluded, even though a valid registration existed, there was no automatic protection; actual competition was required for relief. Even for registered rights, therefore, the marketplace focus has survived:

We hold that because no likelihood of public confusion arises from the concurrent use of the mark in connection with retail sales of doughnuts and other baked goods in separate trading areas, and because there is no present likelihood that plaintiff will expand its retail use of the mark into defendant’s market area, plaintiff is not now entitled to any relief under the Lanham Act.³⁰⁷

B Scholarly Distortions: A Mirage of “Territorial Extraterritoriality”

Comparing the Lanham Act’s system of rights acquisition and extension with pre-1947 law, Roger E. Schechter has posited that trademark law was originally grounded on an “explicitly territorial foundation”³⁰⁸ but that a different system has since been established due to the Lanham Act’s peculiar features of registration, priority, and constructive notice. According to Schechter, the Lanham Act has created a situation of trademark rights’ “domestic extraterritoriality.”³⁰⁹ The once territorial foundation has been enlarged beyond its initial scope to an area of nationwide protection. Because Lanham Act registration grants the owner a right to control others’ uses outside her actual trading area, each area of non-use-based protection, Schechter argues, must be defined as “extraterritorial.”³¹⁰ I will address his approach in more detail in the next chapter.³¹¹ At this point, it suffices to take a closer look at Schechter’s understanding of “territoriality” and “extraterritoriality.” Even though his model may not be representative of scholarly commentary and practice, it is typical of the understanding of trademark rights extension in US doctrine. One aspect in particular is eye-catching: characterizing rights extension by the Lanham Act as extraterritorial illustrates the inseparability of rights and geography. Here as well, the common law model of use-based rights dominates legal thinking. Tea Rose/Rectanus made trademark acquisition inseparable from the geographic area of use. The implementation of a federal registration system with an option of immediate trademark acquisition upon application or registration has not altered this structure. An extension of rights beyond the area of actual use is therefore deemed extraordinary—in other words, extraterritorial.

A similar characteristic of legal doctrine is reflected in the hesitation to implement a federal law of unfair competition prevention. Suggestions were submitted even before Erie. None of these suggestions was implemented in practice, though: no uniform federal statute was enacted, and no common law solution was applied. Most prominently, based on its section 44(i), Edward S. Rogers suggested construing the Lanham Act as having laid out a federal action against unfair competition, covering all conduct that was condemned by either the revised 1883 Paris Convention or the 1929 Inter-American Convention.³¹² The Ninth Circuit developed this idea into the Stauffer doctrine, named after its 1950 case Stauffer v. Exley.³¹³ Under this approach, any US citizen would receive the same protective benefits that foreigners were entitled to under the United States’ international obligations. A federal action against unfair competition would have eliminated the oft-enunciated detrimental effects of Erie.³¹⁴ Other circuits, however, did not follow Stauffer.³¹⁵ Nor did Congress adopt the suggestion of creating a federal cause of action. This cause of action could have been based on section 5 of the Federal Trade Commission Act.³¹⁶ In paragraph 1, this provision declares unlawful “unfair methods of competition in or affecting commerce, and unfair or deceptive acts or practices in or affecting commerce.” In addition, it authorizes the Federal Trade Commission to prosecute unfair practices. With regard to the creation of a federal cause of action, it was suggested that the provision also be considered as a basis for private litigation among competitors.³¹⁷ But this was unsuccessful. In the end, unfair competition prevention protection remained largely an issue of common law and state law.³¹⁸ Until today, the concept of territorially uniform rules of unfair competition prevention has been only rudimentarily developed under the Lanham Act’s provisions against unfair competition (e.g., section 43).³¹⁹

IV Summary: Nonformalism and the Nonterritoriality of Trademarks

Political borders within the United States were practically irrelevant as far as the acquisition and protection of common law trademark rights was concerned. Early nineteenth-century trademark protection gave scant regard to state or even national boundaries. And even though the Supreme Court subsequently reduced the initial excess extension, the territoriality of rights remained a nonissue. Based on Tea Rose/Rectanus, courts found trademark and unfair competition law to constitute a uniform and comprehensive system of goodwill protection. As a result, the understanding of trademark protection as an instrument of market segmentation and allocation of market shares became the most fundamental characteristic of US trademark doctrine. In a sense, Tea Rose/Rectanus established a common law trademark model of immediate market/right correlations. And this distinctly apolitical common law pedigree of US trademark law also survived federalization beginning in the twentieth century. While Erie could have been deemed to have put an end to the casual cross-border adjudication in trademark and unfair competition cases, its impact was far less effective than expected. Courts rarely deal explicitly with rights territoriality or issues of state sovereignty. Finally, the Lanham Act, although implementing the first federal regime of acquisition and protection, did not jettison use-based rights as the doctrinal paradigm.

A final conclusion can be drawn in light of this genuine American trademark doctrine. It reflects a significant counterposition to the formalism of German and European law.³²⁰ The acquisition and protection of rights, until today, has scarcely depended on formalities. In 1947, Nims summarized the history of US doctrine by comparing it with British trademark law (which had implemented a statutory registration system in the nineteenth century). He explained that in the United States, “trade-mark statutes, state as well as federal, play a less important part.”³²¹ In the same year, Daphne Robert built on this understanding by arguing that “[a] trade-mark or service mark is not a Government grant.”³²² Nims’s and Robert’s characterizations are representative of the understanding of trademark rights being founded, in large part, on use within the marketplace, not on state-granted privileges. In fact, state and federal registration were sometimes even explicitly deemed irrelevant and ineffective. A 1935 bulletin of the New York Bar Association put it clearly:

Registration in the U.S. Patent Office is not at all essential for the protection of vested trade-mark rights. … Vested trade-mark rights are recognized and protected by the courts in all the states irrespective of state or federal registration. … State registration is helpful only in exceptional cases and the trade-mark owner should not be burdened with the large expense involved in securing such registration except in unusual cases.³²³

The association’s praise for common law “vested rights” was issued in order to warn the public about alleged “Trade-Mark Specialists” trying to solicit business by overemphasizing the relevance of registration for rights acquisition, maintenance, and protection.³²⁴ In explaining the current state of the law, the association unmistakably clarified the reluctance of traditional and contemporary trademark doctrine to adopt the formalities of registration.³²⁵ In other words, a privilege theory never existed in American trademark law.

Section 3 International Trademark and Unfair Competition Law

The paradigm of market/rights correlation has not only survived the federal unification but also affected trademark and unfair competition conflicts law. As Dan Burk has suggested—aptly illustrating the common law approach—the grounding of likelihood-of-confusion testing on market analysis has made common law principles of trademark protection “quite capable of spanning national borders.”³²⁶ Accordingly, the question is not “whether an unauthorized use has occurred within a certain territory, but whether a particular use is likely to cause confusion.”³²⁷ In other words, it is an analysis not of territorial sovereignty but of market effects. Accordingly, the “Bulova test” established by the Supreme Court in 1952 for international trademark and unfair competition conflicts is founded on the “effects on United States commerce.” This focus on commerce may have a basis in constitutional law; in essence, however, Bulova testing reflects a conventional common law analysis. Before I begin a more specific discussion on this aspect, it is necessary to point out a general characteristic: even though details of foreign rights’ extension into US territory are still widely ambiguous, it is uncontested that market extensions do “carry” concurrent rights across national boundaries. National borders may be pierced by trademark rights—this happens not only from the outside into the United States but also the other way around.

I The Porosity of National Borders and International Goodwill Theory

Prima facie, questions of foreign rights’ extension into the United States are not central to the field of conflicts law, or choice of law. After all, within the confines set by international agreements, nation-states are generally free to regulate trademark use and competition on their own soil; the issue, thus, seems to be primarily governed by domestic law. A look at the “well-known marks” doctrine, however, reveals a number of problematic aspects that are also important for this inquiry.³²⁸ The doctrine implements US obligations under article 6^bis of the Paris Convention.³²⁹ Its aim is to avoid the registration and use of marks that might cause confusion with other marks that, albeit unregistered and unused, are already well known in the country of registration or use.³³⁰ Even though foreign rights’ extension into the United States presents the reverse scenario to domestic rights extraterritoriality, a look at how case law handles the protection of foreign trademarks is revealing for an understanding of Bulova.

A The Well-Known Marks Doctrine: Transnational Goodwill Misappropriation

Early illustrations of transnational goodwill protection can be found in the 1936 and 1959 New York Supreme Court cases Maison Prunier v. Prunier’s Rest. & Cafe³³¹ and Vaudable v. Montmartre, Inc., respectively.³³² In Maison Prunier, the plaintiff was operating a restaurant in France that had become famous under the name “Prunier” since the restaurant’s founding in 1872. The restaurant had developed international repute, and the owners had opened a branch restaurant in London in 1935; they were also interested in extending their business to New York. Yet, in 1935, the defendants began implementing a business scheme for operating under the plaintiff’s name in New York. In his decision, Justice Shientag began by lamenting the existing doctrine on rights acquisition and protection in geographically separated markets and the rules to be applied in zones of business expansion.³³³ Closely following Tea Rose/Rectanus, he emphasized that both the reputation of the senior user and the good or bad faith of the second-comer would determine the outcome. As he acknowledged, “The protection may be extended to the market in which the meaning of the original mark has become known.”³³⁴ Moreover, he added that “[t]he deliberate appropriation of the name ‘Prunier’ is some evidence at least of plaintiff’s wide repute.”³³⁵ And even though he refused to decide whether the defendants’ activities were to be seen as “indefensible from an ethical viewpoint and [as] amounting to an aggravated form of commercial piracy,”³³⁶ he enjoined them pendente lite from using the plaintiff’s name in New York City.

In the second case, a restaurant operator had adopted the name “Maxim’s” for his New York city dining place. The original world-famous “Maxim’s,” founded in Paris in 1893, had become famous for, among other things, having been a setting in Franz Lehár’s operetta “The Merry Widow.” The court found that there was “no doubt as to [the original restaurant’s] unique and eminent position as a restaurant of international fame and prestige.”³³⁷ Accordingly, Justice Greenberg enjoined the New York restaurateur from using the name, even though the name owners had expressed no intention to expand their business activity to New York.³³⁸ The court’s reasoning, short as it was, displayed a distinct aspect of universality in misappropriation prevention and property protection:

The trend of the law, both statutory and decisional, has been to extend the scope of the doctrine of unfair competition, whose basic principle is that commercial unfairness should be restrained whenever it appears that there has been a misappropriation, for the advantage of one person, of a property right belonging to another.³³⁹

Over time, however, this doctrine of an international zone of expansion has become increasingly problematic. Under modern socioeconomic circumstances, two bedrock principles of American trademark law have come into sharp conflict. One is the requirement of territorial trademark “use” as a precondition for rights acquisition. The other is an understanding that trademark protection must be in conformity with market penetration. In their struggle to reconcile the two principles, courts have increasingly reverted to the paradigm of goodwill and its detachment from national-political territories. Three recent cases are telling.

In the 2003 Int’l Bancorp LLC case, a majority of the Fourth Circuit found the Monte-Carlo Casino’s provision of services to American tourists in Monaco and its concurrent advertising for casino services in the United States sufficient to constitute “use in commerce” as a precondition for trademark protection under the Lanham Act.³⁴⁰ In essence, the requirement of actual use “in the United States” was deemed dispensable. Against a vigorous dissent by Judge Motz,³⁴¹ the majority found it sufficient that modern American consumers would travel abroad to the place where services were rendered. Territoriality of use was substituted by customer-base mobility.

The issue of consumer mobility became even more pressing the following year. In the Ninth Circuit’s 2004 Grupo Gigante SA De CV v. Dallo & Co., Inc. case, the dispute centered on a Mexican chain of grocery stores’ use of the mark “Gigante” and an American party’s use of the mark in Southern California. Even though the American party had priority of use in California, Judge Kleinfeld ruled in favor of the Mexican right owner:

We hold … that there is a famous mark exception to the territoriality principle. While the territoriality principle is a long-standing and important doctrine within trademark law, it cannot be absolute. An absolute territoriality rule without a famous-mark exception would promote consumer confusion and fraud. Commerce crosses borders. In this nation of immigrants, so do people. Trademark is, at its core, about protecting against consumer confusion and “palming off.” There can be no justification for using trademark law to fool immigrants into thinking that they are buying from the store they liked back home.³⁴²

In order for a mark to be characterized as “well known,” the court required more than the mere existence of secondary meaning in the relevant market. The mark had to be familiar or known to a “substantial percentage” of consumers in the relevant market sector.³⁴³

This approach was rejected in 2007 by the Second Circuit in ITC Ltd. v. Punchgini, Inc.³⁴⁴ The plaintiff, ITC, was an Indian corporation that owned and operated the world-famous restaurant “Bukhara” in New Delhi, India. In the 1980s, ITC had further licensed the name to numerous restaurants around the world, including in Chicago and New York. It had also acquired a US trademark registration for the name. Yet in the 1990s, ITC ceased its activities in the United States, and both restaurants were closed. The defendants opened their restaurant in 1999 in New York under the name “Bukhara Grill,” with similar décor. Contrary to the Ninth Circuit’s decision in Grupo Gigante, the Second Circuit denied an implementation of the well-known marks doctrine in federal trademark law.³⁴⁵ Instead, Judge Raggi referred the case to the New York State high court and, inter alia, certified the question of whether the state’s trademark and unfair competition law recognized such a doctrine. And even though the New York Court of Appeals responded that state law did not contain this doctrine, it acknowledged that unfair competition law provides for a claim against misappropriation in the tradition of Prunier and Vaudable. This is where the doctrine comes full circle:

Under New York law, “[a]n unfair competition claim involving misappropriation usually concerns the taking and use of the plaintiff’s property to compete against the plaintiff[’]s own use of the same property” …. The term “commercial advantage” has been used interchangeably with “property” within the meaning of the misappropriation theory …. What Prunier and Vaudable stand for, then, is the proposition that for certain kinds of businesses (particularly cachet goods/services with highly mobile clienteles), goodwill can, and does, cross state and national boundary lines.³⁴⁶

What all decisions make clear is that the existence and extension of goodwill—and, in its wake, trademark rights protection—are largely independent of political borders. Goodwill has a rather organic structure: it grows and extends with its owner’s marketing activities. Once goodwill has crossed a political boundary, the “new” market territory beyond this border becomes part of the uniform and homogeneous whole. This holistic understanding of goodwill has also influenced the reverse scenario: whenever owners of domestic trademarks seek protection against foreign-based conduct and invasion from abroad, the apolitical nature of goodwill tends toward an extension of domestic rights.

B Rudolf Callmann: A Theory of International Unitary Goodwill

While a porosity of national borders for goodwill and trademark rights seemed to be established from the beginning, the issue of where a particular business’s goodwill should be situated troubled courts and legal scholars for some time. One famous and illustrative scholarly endeavor was Rudolf Callmann’s suggestion that the situs of certain “worldmarks” be the place of manufacture and that there be no separate national goodwill or trademark rights in other jurisdictions.³⁴⁷ Callmann’s theory of unitary goodwill was based on two foundations. The first basis was the concept of trademark use in and across many different jurisdictions. According to Callmann, worldmarks identified a product that had been sold in so many countries and so successfully that the trademark had become known in a considerable part of the world—not only to the actual purchasers, but also to sectors of the public that would not consider a purchase. In the eyes of the public at large, he concluded, these trademarks enjoyed a worldwide status as the trademark of a certain business.³⁴⁸ The second foundation was the idea that “a trademark has only one goodwill.”³⁴⁹ As Callmann posited in language resembling nineteenth-century personality rights theory, a business’s goodwill could not be “divorced from the source that supplies the market any more than the reputation of a person can be separated from the person.”³⁵⁰ Since a trademark’s goodwill was held to be inseparable from the underlying business activity of the trademark owner, the business establishment and the trademark were interconnected with regard to both location and ownership.³⁵¹ As he concluded:

[T]he situs of a worldmark’s goodwill is the situs of the international business that produces the article, unless that business uses different national trademarks in the various countries where the article is made and/or sold. The public will, by and large, identify Ford and Coca-Cola with the United States, Coty, Chanel and Cointreau with France, Guiness [sic] and Jaguar with England, Fiat and Olivetti with Italy, 4711, Zeiss and Bayer with Germany, and Omega with Switzerland. In all these cases the situs of the goodwill of those marks is the situs of the main business in the United States, France, England, Italy, Germany and Switzerland, respectively. In the case of Unilever, however, its margarine is identified as “White Lune” in England, “Blauband” in Germany, “Start” in Holland, “Solo” in Belgium, “Astra” in France, and “Sava” in Turkey; the situs of the goodwill of each such mark would be in the country where the particular mark is used.³⁵²

This theory, which Callmann termed an “indivisible or unitary theory of goodwill,” was also asserted by a handful of other voices.³⁵³ One example was the decision by the US Commissioner of Patents in Ex parte E. Leitz, Inc.:

It is true that as a result of the sale of German Leitz products in the United States by its American distributor, New York Leitz, a considerable amount of trade mark goodwill was generated in the United States, but such goodwill was not separated, indeed, it was inseparable, from the mark itself. In other words, the goodwill in the United States which was symbolized by the trade mark “Leitz” had its situs in Wetzlar, Germany, where the manufacturer was located. The American distributor acquired no rights in the trade mark or in the goodwill symbolized by it merely as a result of importation and sale in this country of the products of German Leitz.³⁵⁴

Yet the majority of courts³⁵⁵ and legal scholars³⁵⁶ rejected the concept of unitary goodwill. The idea that goodwill needed a situs was not contested. Critics did argue, however, that goodwill’s extension was dependent on the scope of a business or trade. If the trade covered several national territories, each constituted a separate compartment of independent goodwill.³⁵⁷ Most simply, for instance, a national market might be separated from neighboring states’ markets by a modification of the product. This was Walter Derenberg’s point of criticism. He argued that an international product may have different characteristics in different countries, reflecting local preferences. Each national product, due to these differences, would then constitute a separate market. Accordingly, different goodwill “portions” were to be distinguished.³⁵⁸ Another reason for distinguishing markets was customer perception. On this basis, the District Court for the District of Columbia, in another Leitz case, rejected the concept of unitary goodwill: “if the public ever understood or now understands all products bearing the ‘Leitz’ mark as having originated with German Leitz, its understanding was and is erroneous.”³⁵⁹

Notwithstanding the majority’s rejection of goodwill transnationality and homogeneity, the core question remained unanswered: What would happen in cases where neither product differentiation nor customer perception provided a clear guideline for the geographical or territorial separation of markets? If goodwill really was a subject matter of organic growth, and if it was also apt to transcend national frontiers with the stream of commerce or even market communication, arguing in favor of a strictly political segmentation would be difficult. As we will see, the Supreme Court’s 1952 Steele decision and its progeny have implemented a doctrine of unitary goodwill in the interest of national right owners.

II Trademarks’ Extraterritorial Scope: Steele v. Bulova Watch Co. and Its Progeny

As seen earlier, the once-governing concept of trademark universality, a product of nineteenth-century property theory, was aptly illustrated by Derringer and Kidd. Tea Rose/Rectanus subsequently reduced these quasi unlimited rights.³⁶⁰ The factual universality of rights, however, was never fully abolished. This is due to the fact that, unlike in Germany, in the United States this universality was not superseded by a political theory of rights. The disregard for state sovereignty and boundaries would prove determinative. Indeed, in 1952, the Supreme Court implemented the paradigm of an apolitical market relatedness in international trademark conflicts.

A The Epicenter of Extraterritoriality: Steele v. Bulova Watch Co.

The Supreme Court’s 1952 decision in Steele v. Bulova Watch Co.³⁶¹ has been duly earmarked as the landmark or “seminal case” of US trademark and unfair competition conflicts law.³⁶² The majority’s opinion was groundbreaking not only because it represented—and continues to do so—the sole Supreme Court decision on the issue. Far more influential than many of the court’s precedents in other fields, the majority opinion in Steele linked different eras and sectors of US law. First, it connected the early common law precedents on unfair competition conflicts doctrine with a modern test for the then-new Lanham Act’s subject-matter jurisdiction. Furthermore, the newly established analysis under the so-called Bulova test implemented a number of different strands of conflicts doctrine. Not only does the test require considering concepts of public international law, but it also connects the fields of tort choice of law, trademark conflicts, and international antitrust. In its combination of common law precedents, public international law doctrine, and transnational regulatory litigation, Steele was as reactionary as it was innovative.³⁶³

“The issue,” as Justice Clark started the majority’s analysis in a rather circular fashion,

is whether a United States District Court has jurisdiction to award relief to an American corporation against acts of trade-mark infringement and unfair competition consummated in a foreign country by a citizen and resident of the United States.³⁶⁴

As would soon become clear, this formulation invited a maximum scope of application for domestic trademark and unfair competition law.³⁶⁵ As we have seen, traditional concepts of trademark territoriality dominating in contemporary German and European law would have rejected a similar formulation of the issue ab initio. Their starting point was strictly territorial: without inland conduct, no domestic rights could be found to be infringed on.³⁶⁶ While the Supreme Court’s dissent argued similarly, the majority disregarded old-fashioned territoriality.³⁶⁷

The case facts are as follows: Sidney Steele, the primary defendant in the case, was a US citizen residing in Texas. The plaintiff, Bulova Watch Co., was a watch manufacturer that had registered the “Bulova” trademark in the United States but not in Mexico. Upon learning about the company’s lack of formal rights, Steele registered the mark in Mexico, bought watch parts in Switzerland and in the United States, and then had the parts assembled into watches, stamped with the mark “Bulova,” and sold. All of this happened exclusively in Mexico. When Bulova learned of Steele’s activities, it initiated litigation in the Texas district court. Meanwhile, upon parallel litigation started by Bulova in Mexico, the Mexican registration “Bulova” was eventually canceled. In the United States, the district court dismissed Bulova’s complaint on the ground that the court lacked jurisdiction over the cause;³⁶⁸ there was no illegal act committed within US territory. The case was then brought to the Court of Appeals, which reversed the decision. The Supreme Court affirmed.

B The Steele Progeny: A Motley Crew of Circuit Court Tests

Subsequent case law and scholarship have interpreted the Supreme Court’s reasoning and holding as having established three test factors—known as the Bulova factors—for Lanham Act extraterritoriality: (1) “nationality or citizenship of defendant,” (2) “effects on United States commerce,” and (3) “conflicts or potential conflicts with foreign law.”³⁶⁹ Based on these factors, a variety of tests has developed among the circuits. Most prominent among these tests are the Second Circuit’s Vanity Fair test, the Fifth Circuit’s American Rice decision, and the Ninth Circuit’s Wells Fargo or Timberlane rule of reason. In addition, the First Circuit has recently established a new test in McBee v. Delica Co. All of these tests consider the three Bulova factors. And even though the Ninth Circuit is somewhat the outlier, balancing “effects on United States commerce” in a rule of reason derived from antitrust extraterritoriality, the special comity factors integrated into the rule-of-reason test also contain “nationality” and “conflicts with foreign law,” among others.

The Second Circuit’s 1956 Vanity Fair Mills, Inc. v. T. Eaton Co.³⁷⁰ decision marks the beginning of what has come to be called the Vanity Fair test, a modification of the Bulova test.³⁷¹ The plaintiff sued for trademark infringement stemming from the defendant’s allegedly unauthorized use of the “Vanity Fair” name. The plaintiff was a Pennsylvania corporation that had sold women’s underwear in the United States (since 1914) and Canada (since 1917). The defendant was a Canadian corporation that had been granted the Canadian trademark registration “Vanity Fair” for similar products, which it began selling in 1915. Due to the defendant’s prior registration, the plaintiff was denied a trademark in Canada. When the defendant started selling both the plaintiff’s “Vanity Fair” products and its own merchandise under the brand, the plaintiff sought an injunction against the defendant’s use in both Canada and the United States. The Second Circuit started by analyzing the Supreme Court’s Bulova decision and then explained its own version of the three factors: First, the defendant’s conduct had to have a “substantial” effect on US commerce. Second, the defendant had to be a US citizen. And finally, conflicts with foreign law were to be avoided. The Vanity Fair test was significantly relaxed in later decisions, due mainly to modifications in light of other circuits’ interpretations of the Bulova test.³⁷²

In 1977, the Ninth Circuit adopted its own test version. In Wells Fargo & Co. v. Wells Fargo Express Co.,³⁷³ it formulated a rule-of-reason approach for assessing the Lanham Act’s international reach. In this case, the plaintiff had used the registered trademark “Wells Fargo” throughout the United States. The defendant, a foreign corporation, was using the same trademark in the United States and Europe. After the district court had rejected subject-matter jurisdiction based on Vanity Fair, the circuit court vacated the verdict and developed a circuit-specific test based on the jurisdictional rules of reason established in the Ninth Circuit’s case law on antitrust extraterritoriality, particularly Timberlane Lumber Co. v. Bank of America.³⁷⁴ This balancing test—a “ ‘jurisdictional rule of reason’ of comity and fairness”— required only “some”³⁷⁵ effects on US commerce and an additional analysis of several comity factors, notably:

[1] the degree of conflict with foreign law or policy, [2] the nationality or allegiance of the parties and the locations or principal places of business of corporations, [3] the extent to which enforcement by either state can be expected to achieve compliance, [4] the relative significance of effects on the United States as compared with those elsewhere, [5] the extent to which there is explicit purpose to harm or affect American commerce, [6] the foreseeability of such effect, and [7] the relative importance to the violations charged of conduct within the United States as compared with conduct abroad.³⁷⁶

The Ninth Circuit’s then-new balancing effort has been interpreted as a stark contrast to the Supreme Court’s and the Second Circuit’s allegedly bright-line tests.³⁷⁷ While the Bulova and Vanity Fair tests required taking into account only three test factors, the Timberlane formula seemed to establish a more sophisticated—and more problematic—analysis, allowing the Ninth Circuit’s courts an allegedly wider range of interpretations.³⁷⁸

The last circuit to establish its own test was the First Circuit in McBee v. Delica Co.³⁷⁹ In this case, the plaintiff, an American jazz musician, sued a Japanese clothing retailer that had adopted the trademark “Cecil McBee” (identical to the plaintiff’s name) for its adolescent female clothing line. The defendant company held a Japanese trademark. Though it did not market its products outside of Japan, the company maintained a website where the trademark was displayed. After the plaintiff’s unsuccessful attempt to have the trademark invalidated in the Japanese trademark registry, he filed a complaint asserting trademark dilution and unfair competition. The district court applied the Vanity Fair test and denied subject-matter jurisdiction. The circuit court formulated a new test, albeit producing an identical result. Under McBee, an inquiry into the defendant’s nationality is the mandatory first step of any analysis. Only if the defendant is not a US national will “substantial effects” on US commerce become the determinative factor. As the court further explained, however, even if substantial effects on US commerce are found to exist, a separate comity analysis might still result in the nonapplication of US trademark law.³⁸⁰

This multitude of tests suggests that subject-matter jurisdiction under the Lanham Act is prone to various interpretations. The outcome depends on which version of the test is applied. Some courts (such as those in the Ninth Circuit) seem to be more deferential to international concerns and therefore tend to limit the reach of US law. The Second Circuit’s Vanity Fair test, by contrast, is considered a bulwark for trademark owners against foreign-based infringing activities.³⁸¹ The most significant overextension of trademark protection, however, has probably occurred in the Fifth Circuit. In the 1983 American Rice case, both parties were US farmers’ marketing cooperatives acting in the United States and abroad. The defendant was selling rice in Saudi Arabia under a trademark similar to the plaintiff’s US registration.³⁸² Even though sales under the allegedly infringing trademark occurred solely in Saudi Arabia and “none of [the] products found their way back into the United States,”³⁸³ the court saw an infringement of the plaintiff’s US trademark. Effects on US commerce were seen in Saudi Arabian sales, particularly on the basis that the processing, packaging, transportation, and distribution of US-produced rice constituted activities “within commerce.”³⁸⁴

III Doctrinal Analysis: Use-Based Rights and Commercial Effects

Before taking a closer look at the Steele progeny, I will examine the Supreme Court’s majority’s opinion from a historical-doctrinal perspective. While a number of attempts have been made to explain the reasoning and holding, there has yet to be a comprehensive analysis exploring how the majority connected pre-Lanham Act case law with the new act. Such an analysis reveals that the judges of the Steele majority extended common law tort and unfair competition conflicts law into their statutory interpretation of the Lanham Act’s jurisdictional reach. As a result, the paradigm of market-based, organic, and apolitical goodwill extension seeped into modern trademark conflicts doctrine. In addition, in its reference to international antitrust precedents, the majority laid the foundation for a modern reliance on the effects-on-commerce test factor.

A The Common Law Roots of Lanham Act Subject-Matter Jurisdiction

As the Steele majority explained, prior to the Lanham Act’s enactment, courts had already granted relief to US trademark owners “[u]nder similar factual circumstances.”³⁸⁵ Looking at these cases, they concluded that the act’s language of reaching “all commerce which may lawfully be regulated by Congress” could “not constrict prior law or deprive courts of jurisdiction previously exercised.”³⁸⁶ The Lanham Act’s commerce provision thus became a conduit for incorporating common law doctrine into statutory trademark law. In its reference to pre-Lanham Act case law, the majority cited decisions by the New York and New Jersey circuit courts and the Supreme Court of New York.³⁸⁷

One of these decisions was the 1907 case Vacuum Oil Co. v. Eagle Oil Co., in which the court had to decide on allegations of international trademark infringement. Both parties to the case were US companies engaged in oil exportation. The plaintiff was doing business in the United States and in Europe. The defendant, Eagle Oil, was purchasing barrels of oil in the United States and shipping them to Germany, among other places, for sale. Eagle Oil attached the plaintiff’s trademark to these barrels, but not before their arrival in Germany. In addition to using the plaintiff’s trademark, the defendant made false representations concerning the products’ origin and production process.³⁸⁸ The court held that Eagle Oil, its manager, and certain officers had committed fraud and unfair competition not only in Europe (notably Germany) but also in the United States:

[T]he scheme was designed and intended to injure the defendant’s business by the false and fraudulent use of its trade-names, while at the same time maintaining so far as possible an unassailable position. Sufficient evidence has been given to satisfy me that the scheme was conceived and partially, but to a material extent, carried out in this country. … It cannot be that the arm of the court is too short to reach and stop this fraudulent conduct, or so much of it, at least, as is carried on in this country. … The purchase and shipment of this oil for the purpose of selling it under false representations, and the sale of it under false representations and trade-names abroad in unfair competition with the complainant, was a single business, and each step in the transaction was part of a single fraudulent scheme …. This unfair competition has inflicted injury upon the complainant’s business in this country by diminishing, or tending to diminish, its foreign trade.³⁸⁹

In addition to the domestic-injury-by-foreign-trade-impact paradigm, the court embraced an idea of internationally uniform standards of honesty in commerce. This, as explained in chapter 1, was a common perception at that time.³⁹⁰ As the court wrote:

The presumption is that the law in the foreign countries where any part of the fraudulent business was carried on is the same as our own, and that fraudulent acts are unlawful there as here. … It is apparent that an act that violates the law of fair dealing and good conscience must be of universal recognition. To assume the contrary is to suppose the foreign countries in question to be in a state of barbarism, and that is to assume a state of affairs that justify this court in applying the law of the forum. … But while the action is founded upon fraud, it is also of a transitory character, and the fact that some of the fraudulent acts were committed outside the jurisdiction of this court and outside of the United States will not avail the defendants.³⁹¹

Another decision cited by the Steele majority was the 1916 New York Supreme Court’s judgment in Morris v. Altstedter.³⁹² Both parties to the case were New York residents. The defendant had attempted to purchase the plaintiff’s product for resale. Since an agreement with the plaintiff failed, however, he copied the plaintiff’s artistic plaques and mottoes (known as woodenettes), which he then produced and sold solely in Canada. The proceeds, however, went to the defendant’s US business. Holding for the plaintiff, and relying on Vacuum Oil, the court emphasized the fraudulent character of the defendant’s conduct. Since fraud was frowned upon everywhere and an action was considered transitory, no question of territorial limitations came to the fore:

It has been repeatedly held that an act that violates the law of fair dealing and good conscience must be of universal recognition. Unfair competition in trade is made cognizable by a court of equity, because of its essentially fraudulent character. … It has also been held that, while the action is founded upon fraud, it is also of a transitory character, and the fact that some of the fraudulent acts were committed outside of the jurisdiction of this state or the United States will not avail the defendant.³⁹³

The third decision referenced by the Steele majority was the 1944 Second Circuit decision George W. Luft Co. v. Zande Cosmetic Co. In this case, as in Vacuum Oil, the aspect of injury to the plaintiff’s domestic business resulting from losses in foreign trade was present. The plaintiff and defendant, both New York corporations, were manufacturing and selling cosmetics. In a prior proceeding, the district court had found an infringement of the plaintiff’s trademark. In its decree, the court had comprehensively enjoined the defendant from using the trademark. As the Luft court later explained, “Read literally this seem[ed] to preclude defendants from doing business under [their trademark] anywhere in the world.”³⁹⁴ The Luft court deemed this too sweeping, at least with regard to the foreign business activities of both parties. Hence, the court began its modification of the district court’s decree by classifying the parties’ business activities into three categories:

As to the foreign business, the evidence … bears upon a classification that we regard as relevant, as follows: (a) countries where both parties are doing business and the defendants have established their right by the local law to use the [trademark]; (b) countries where both parties are doing business and the defendants have not established such right; and (c) countries where the defendants are doing business and the plaintiff has not proved that it has ever done business or is likely to do it.³⁹⁵

Regarding the first category—countries where both parties had undertaken business activities—the court explained that the defendant’s use of the allegedly infringing trademark could not constitute unfair competition or trademark infringement due to the existence of superior foreign rights. Granting a decree under such circumstances, the court felt, would give US trademark rights an extraterritorial effect, unduly interfering with foreign sovereign policies. The court also denied equitable relief against the defendant’s activities within the United States that were exclusively concerned with the relevant foreign countries; there was no fraudulent scheme fulfilling the requirements of Vacuum Oil. Notwithstanding this liberal stance toward the defendant’s foreign activities (founded on superior foreign rights), the concept of foreign-market protection surfaces with particular clarity in the two remaining categories. This is where the Tea Rose/Rectanus paradigm of market-based goodwill surfaces. The third category—that is, cases where the alleged infringer was doing business in a country where the domestic trademark owner was not active—is most revealing:

There remains for consideration class (c) countries where the defendants are doing business but the plaintiff has not proved that it ever has done business or is likely to do it. The Trade-Mark Act creates no new substantive rights in those who register their marks. … And it is well established that the right to a particular mark grows out of its use, not its mere adoption, and is not the subject of property except in connection with an existing business [inter alia reference to Tea Rose/Rectanus]. Hence if the defendants are doing business in Turkey, for example, but the plaintiff has never extended its trade to that country and there is no evidence that it is likely to do so, the plaintiff has not been damaged by the defendants’ Turkish business and is not entitled to restrain its continuance or to an accounting for damages and profits with respect to sales made there.³⁹⁶

Even though the Luft decision was later praised for avoiding sovereignty conflicts,³⁹⁷ it actually displays a distinctly different stance. The court did not double-check for territorial limitations. On the contrary, in its transnational projection of common law zones of protection, the court neglected the fact that US common law trademark rights, by definition, could never extend into foreign territories lacking a common law regime of trademark protection and unfair competition prevention—notably a civil law system like that of Turkey. The court’s arguments reflect a focus on universality that also existed in the domestic context: protection follows the right owner’s trade and business activities, and political boundaries are, at best, of secondary concern.³⁹⁸

The Steele majority, in its jurisdictional analysis, considered all these elements. First, it referred to an unlawful scheme on the part of the defendant. Since Sidney Steele had apparently acted with an intent to hide his improper activities in Mexico, it was easy for the majority to conclude that “[i]n sum, we do not think that petitioner by so simple a device can evade the thrust of the laws of the United States in a privileged sanctuary beyond our borders.”³⁹⁹ The defendant’s evident bad faith made the analysis a simple task.⁴⁰⁰ Less evident is the way the majority extended common law trademark principles. Nevertheless, as we will see below, this aspect has since become an important element of US trademark and unfair competition conflicts law. And it can also be traced to the Steele reasons: as the majority explained, the counterfeit watches, filtering into the United States from Mexico, “could well reflect adversely on Bulova Watch Company’s trade reputation in markets cultivated by advertising here as well as abroad.”⁴⁰¹ Protection of the US trademark “Bulova” against frivolous activities abroad was thus not only founded on the defendant’s activities inside the United States. In addition, the cross-border osmosis of common law rights—always closely tied to the extension of marketplaces—provided the basis for enjoining injurious activities abroad. In this regard, the court’s summary of relevant facts from the record illustratively emphasized the plaintiff’s extensive marketing “in the United States and foreign countries [in particular by] advertising [that] has penetrated Mexico.”⁴⁰² Read in the context of the Vacuum Oil parameter of “diminishing … foreign trade” and of the Luft common law rights extension, the Supreme Court formulated a theory of transnational goodwill protection.

Most interesting in this regard, finally, is the fact that neither the Supreme Court majority nor the dissent even mentioned the court’s 1927 decision in Ingenohl v. Walter E. Olsen & Co., where Justice Holmes had emphasized the political character of trademark rights.⁴⁰³ While the concept that “a trade-mark started elsewhere depends for its protection in a foreign jurisdiction on the law prevailing therein, and confers no rights except by consent of that law,” had still been highlighted as the guiding principle in Judge Russell’s dissent in the Fifth Circuit a few months before,⁴⁰⁴ the Supreme Court seemingly let these warnings of political-territorialist thought go unnoticed. In this regard, the American conception of transnational goodwill protection differs distinctly from the English doctrine of passing off in international infringement cases. Even though early case law pointed into the same direction,⁴⁰⁵ modern doctrine contradicts an unrestrained extension of goodwill across national borders.⁴⁰⁶

In the end, the Steele majority established both domestic and foreign-based goodwill as protectable subject matter in multijurisdictional trademark and unfair competition cases. Its substance ever since has been characterized by organic growth across state and national borders—necessarily, therefore, marketplace extension trumps political territory. With Steele, an apolitical and nonterritorial interpretation of Tea Rose/Rectanus had gone global.

B An Element of Modernity: The Effects-on-Commerce Factor

Apart from the cross-border extension of use-based rights, Steele laid the foundation of effects testing in Lanham Act subject-matter-jurisdiction analysis. By this means, unlike civil law territoriality doctrine, Steele quite early incorporated a characteristic of economic regulation into US trademark conflicts law. Ever since, goodwill as a subject matter of protection and economic effects as an indicator for finding an infringement have been fused into a combined test for jurisdiction.

The majority’s focus on the results of the defendant’s activities starts with a reference to American Banana.⁴⁰⁷ Initially, the court distinguished Steele from American Banana on the ground that Sidney Steele’s alleged infringements were private and individual conduct, as opposed to the sovereign-state acts in American Banana. Nevertheless, international antitrust doctrine was found to be relevant in trademark law as well. As the majority explained, there was no

blanket immunity on trade practices which radiate unlawful consequences here, merely because they were initiated or consummated outside the territorial limits of the United States. Unlawful effects in this country, absent in the posture of the Banana case before us, are often decisive.⁴⁰⁸

The concept of effects was still rejected by a strong dissent.⁴⁰⁹ But this warning from old-school territorialists went unheard. On the contrary, over time it became the most influential factor of the Bulova test. One of the last circuit courts to explicitly decide on an issue of extraterritoriality was the First Circuit in 2005. In McBee v. Delica Co.,⁴¹⁰ the appellate court rejected the first instance’s application of the Vanity Fair test and developed a new analysis in which it stated that the “sole touchstone to determine jurisdiction” over foreign defendants was whether the defendant’s acts have a “substantial effect upon United States commerce.” Interestingly, the First Circuit also referred to antitrust precedent, particularly the doctrine of Hartford Fire.⁴¹¹ In this regard, McBee marks the end of a long series of decisions that have developed the analysis of Lanham Act subject-matter jurisdiction into a genuine effects test.

IV A Bird’s-Eye View: Taking Stock of Lanham Act Extraterritoriality

The number of cases featuring opinions sufficiently detailed to allow for a closer analysis of the Bulova test or the circuit court variants is relatively small. Between 1952 and 2014, US federal courts have issued more than 140 opinions on the issue of Lanham Act subject-matter jurisdiction.⁴¹² This relatively small number is quite surprising if we look at the prophesies anticipating an enormous rise in conflicts litigation in the field of international intellectual property.⁴¹³ On the other hand, under a more pragmatic perspective, this outcome is not too perplexing. If it is true that the king’s writ reaches only as far as his sword, many, if not most, conflicts will remain unlitigated.⁴¹⁴ In a world where most foreign-based infringements cannot be prosecuted in domestic fora for lack of personal jurisdiction over the alleged infringer, it would be naïve to expect a large body of case law to have evolved. Despite this relative scarcity of cases, however, one may still attempt to examine case numbers and opinion content in order to verify two characteristics of US conflicts doctrine that have been highlighted thus far. One is the assumption of effects dominance in conflicts testing—that is, the idea that the existence and intensity of effects determine whether domestic rights will be protected and whether national policy will be enforced extraterritorially. The second aspect is the common law pedigree of modern trademark conflicts doctrine, particularly with regard to the traditional conception of market-related, organic, and thus necessarily nonterritorial rights under Tea Rose/Rectanus. While I will try to verify these two aspects by looking at the totality of opinions, I am of course aware of the potential objections from an empirical standpoint.⁴¹⁵ This is why I will not call my inquiry an “empirical” study; instead, I will characterize it as a bird’s-eye view of Steele and its progeny between 1952 and 2014.

A The Antitrust Gene: A Dominance of Effects

Looking at all of the opinions’ test-factor analyses for the relevant period, we can divide the case population into two groups. The first group consists of opinions where the court’s analysis of the three factors⁴¹⁶ found them all to point in the same direction—either in favor of or against extraterritoriality. Opinions in the second group decided the issue of Lanham Act application based on different test outcomes for the individual factors. The first group contains 48 opinions in which all three factors were found to either favor or disfavor application of the Lanham Act. The majority (40) of these opinions found the Lanham Act to apply, while the minority (8) rejected Lanham Act subject-matter jurisdiction. Since in all these opinions all three factors were found to point in the same direction, however, this group does not provide immediate insight into the relationship between factors.⁴¹⁷

The picture is much more revealing for the second group, whose 92 opinions involved different factor results. Two subgroups can be distinguished within this second group. The first subgroup consists of 12 opinions that expressly and separately tested and decided on the result of the test for each of the three factors. The second subgroup consists of 80 opinions that—even though they also featured a discussion of at least one factor—applied the test only to the extent that the court found necessary. In other words, these opinions left at least one factor untested or undecided. Both subgroups provide information on the different factors’ weight for the outcome.

My analysis of factor dominance begins with the smaller subgroup (12 in total) in which the courts expressly found different outcomes for each of the three test factors. Of the 10 opinions that found the Lanham Act not to apply, 7 denied the existence of “effects on United States commerce.”⁴¹⁸ In 3 of these opinions, both test outcomes for defendant “nationality” and “conflicts with foreign law” would have favored application of the Lanham Act.⁴¹⁹ The other 3 opinions having found “effects” to exist but still not applying the Lanham Act found the defendant’s foreign “nationality” and “conflicts with foreign law” to outweigh existing effects on US commerce. The first is a 1983 decision of the Western District of New York that still closely adhered to the circuit’s 1956 precedent. The court found “substantial effects” but denied application of the Lanham Act in light of the Vanity Fair holding, the defendant’s Canadian citizenship, and potential conflicts between the court’s own ruling and a Canadian court ruling.⁴²⁰ Another case was decided under the Timberlane rule of reason with the finding that the conflicts factor “weighs strongly against extraterritorial application of the Lanham Act.”⁴²¹ Ultimately, the Court for the Central District of California balanced the comity subfactors with special regard to the fact that—although substantial ties to the United States (and, hence, sufficient “allegiance” under the Timberlane factor list) existed—some of the defendants were foreign citizens.⁴²² The third case, decided by the Easter District of New York under the Second Circuit’s more recent Sterling Drug precedent,⁴²³ featured a finding of both nationality and conflicts pointing toward nonapplication of the Lanham Act.⁴²⁴ Of the 2 opinions that decided in favor of the Lanham Act’s application, both featured a finding of sufficient “effects on United States commerce.” One court found the defendant’s US nationality—broadly understood under the Timberlane comity analysis—and effects on US commerce sufficient to outweigh conflicts with foreign law.⁴²⁵ The other considered the defendant’s foreign nationality negligible based on a simultaneous finding of effects and missing conflicts with foreign law.⁴²⁶ In sum, therefore, it may appear as if “effects” alone or “nationality” and “conflicts with foreign law” together can make a difference.

This tentative finding is further strengthened by the results for the subgroup of opinions (80 in total) sparing a comprehensive discussion and application of all factors. Indeed, as a closer look unveils, these opinions reflect a significant pattern: they can be divided into a subpopulation where the court found “nationality” and “conflicts with foreign law” pointing toward nonapplication of the Lanham Act—then overcoming a positive finding of “effects” or making an “effects” analysis dispensable. In addition, a different segment of the population features opinions where the finding of “effects” or “no effects” was so significant that the other two factors were considered largely irrelevant. Excluding a few opinions that have featured extraordinary scenarios of the subject-matter jurisdiction test,⁴²⁷ the group can be divided into 41 opinions applying the Lanham Act and 35 opinions denying its application.

A rough summary of factor testing and neglect shows that courts treated the three factors differently: The effects factor was disregarded, considered irrelevant, or considered neutral, or a decision on the effects test was left open in 20 of these 76 opinions (26.32%). The defendant’s nationality and potential conflicts with foreign law, by contrast, were ignored 49 times (64.47%) and 55 times (72.37%), respectively. In 33 of 76 opinions (43.42%), the courts decided the case solely upon finding effects to exist or be missing, and either did not address the nationality and conflicts factor at all or left a test decision open. Moreover, 31 of the 41 opinions (75.61%) finding the Lanham Act to apply also found sufficient effects on US commerce. In 17 of these opinions (54.84%), the courts left both “nationality” and “conflicts” undiscussed—in any case, undecided—and based their decisions solely on “effects.”⁴²⁸ In 6 opinions, the courts did not discuss nationality.⁴²⁹ And in another 7 opinions, they did not perform a conflicts-with-foreign-law test.⁴³⁰ Among the opinions that did not expressly find effects and nonetheless applied the Lanham Act are, notably, decisions from the Fifth Circuit’s cascade of American Rice opinions, where the subject-matter issue is handled particularly liberally with a virtually unrestricted scope of the Lanham Act.⁴³¹ Furthermore, in a number of opinions, the effects prong was subdued in the discussion, especially since the court applied an older, shortened, or exotic variant of subject-matter jurisdiction testing.⁴³² Disregarding this small group of outlier cases, it is evident that US law will rarely apply absent an express finding of “effects.” In other words, the existence of “effects on US commerce” is a practical conditio sine qua non for Lanham Act subject-matter jurisdiction.

The effects prevalence is similarly evident with respect to opinions that ultimately denied subject-matter jurisdiction. Among the 35 that did not apply the Lanham Act, 23 expressly found “no effects” (65.71%). Among these, 16 opinions (69.57%) left out both nationality and conflicts testing or rejected a definite decision on both factors and based their finding that subject-matter jurisdiction is amiss on the lack of effects alone.⁴³³ Yet as this bird’s-eye view further unveils, despite its general prevalence, the effects factor is not completely unchallenged. Altogether, 6 opinions (including the 2 opinions in the Vanity Fair dispute⁴³⁴)—even though effects had been alleged to exist—ultimately rejected jurisdiction on the basis of both nationality and conflicts, pointing against application of the Lanham Act. An ultimate and definite analysis of effects was then not undertaken.⁴³⁵ Moreover, the Ninth Circuit’s Court of Appeals decided that potential conflicts alone could hinder the Lanham Act’s application even though sufficient effects on commerce were found to exist.⁴³⁶ And district courts in the Second and the Ninth Circuit found, inter alia, the defendant’s “nationality” (without an analysis of “conflicts”) to be determinative—even though “effects” seemed to exist.⁴³⁷

This analysis helps us draw a conclusion: while the concept of effects was initially an element of antitrust doctrine, it has since become the most relevant test factor for Lanham Act application in cases of international trademark infringement and unfair competition violations. The “effects on US commerce” factor has been the one most often used by courts to decide on the issue of extraterritoriality, while the two other factors, at least taken individually, are less influential. The impact of the effects factor can go in either direction—either in favor of or against a finding of Lanham Act application. If effects are amiss, Lanham Act subject-matter jurisdiction is hard to establish; and if effects are found to exist, both nationality and the conflicts factors regularly lose impact—at least if not marching together in the opposite direction.

B Common Law Goodwill Protection: Tea Rose/Rectanus Goes Global

It was not just the antitrust pedigree of the Steele majority’s arguments but also their common law foundation that has dominated trademark conflicts law ever since. While “effects on US commerce” have ultimately become the most relevant test factor in practice, the subject matter being protected was (and continues to be) the trademark owner’s use-based goodwill. Lanham Act application in international conflicts—like the domestic doctrine of trademark protection—is based on the conception of preventing improper goodwill invasion, even on foreign territories if necessary.

Typically, when testing for “effects on US commerce,” a court will start by defining the Lanham Act’s substantive law policy, particularly regarding the prevention of consumer confusion and deception. In international trademark doctrine, an occurrence of consumer confusion usually equals the finding of effects.⁴³⁸ In addition to consumer confusion, injury to the trademark owner may also indicate a relevant effect. In this regard, courts after Steele have developed and made use of a number of effects subfactors. This array of subfactors reflects the historical multitude of policies, primarily with respect to the development of legal doctrine in domestic trademark and unfair competition law. As we have seen, concepts of goodwill protection, along with diversion-of-sales and misappropriation prevention, have dominated the debate from the fields’ nineteenth-century incipiencies forward.⁴³⁹

A bird’s-eye view reveals that a substantial share of all opinions decided between 1952 and 2014 extended the effects test using a number of subfactor analyses. Overall, 119 opinions (85%) made use of different subfactors. The subfactors considered were the following:

• “consumer confusion” (tested in 51 opinions (36.43%))

• “diversion/loss of sales” (tested in 48 opinions (34.29%))

• domestic activities that provided “material support for foreign trademark use/business,” that constituted “essential steps [within the United States] in the course of business consummated abroad,” or that constituted the “orchestration of foreign activities” (tested in 46 opinions (32.86%))

• “damage to (ability to do) business and/or income” that affected the “value of plaintiff’s holdings” or caused “monetary harm” or “losses to the right owner” in general (tested in 44 opinions (31.43%))

• “damage to/adverse reflection on reputation/goodwill” (tested in 43 opinions (30.71%))

• “sale/offering of goods abroad with subsequent entering into the U.S.” (tested in 31 opinions (22.14%))

• “misrepresentation” without further specification (tested in 10 opinions (7.14%))

• “using/putting goods into the stream of U.S. commerce,” making “physical use of the U.S. commerce stream,” using “instrumentalities of U.S. commerce,” or “availing oneself of business opportunities inside the U.S.” (tested in 9 opinions (6.43%))

• “misappropriation/tarnishing of trademark rights/goodwill” (tested in 7 opinions (5%))

• a few more uncommon factors, such as “loaning funds in/transacting bank business in the U.S.” (tested in 6 opinions (4.39%)); “financial gain of a U.S. entity [i.e., defendant] received from abroad” (tested in 5 opinions (3.57%)); and whether defendant had violated “fair competition rules” (tested in 1 opinion (0.71%)).

A number of these subfactors can be traced to the Steele majority’s finding of an “unlawful scheme.”⁴⁴⁰ They aim primarily at the prevention of “unfairness,” not at goodwill protection. They cover defendant activities that provide “material support for foreign trademark use/business,” that undertake “essential steps [within the United States] in the course of business consummated abroad,” or that involve the defendant’s “orchestration of foreign activities.” All of these subfactors are distinctly focused on the defendant’s territorial activities. This means that they generally require conduct within the United States to be fulfilled.

In addition, one category of subfactors is only indirectly connected to the issue of goodwill protection; there is no connection to the right owner’s market position. These subfactors reflect a concern for unfairness prevention and an aim to protect the right owner’s financial assets or her business in general. Among this category are the subfactors “using/putting goods into the stream of U.S. commerce,” making “physical use of the U.S. commerce stream,” making “use of instrumentalities of U.S. commerce,” and “availing oneself of business opportunities inside the U.S.,” as well as a test for “damage to (ability to do) business and/or income” or for effects on the “value of plaintiff’s holdings,” “monetary harm,” or “losses to the right owner” in general.⁴⁴¹

Moreover, the subfactor “sale/offering of goods abroad with subsequent entering into the U.S.” is a direct descendant of the Supreme Court’s postsale confusion argument that the fake watches sold in Mexico might filter into the United States and injure the plaintiff’s domestic goodwill.⁴⁴² Even though this subfactor is largely goodwill related, it concerns the plaintiff’s domestic market and, accordingly, her domestic rights and goodwill position.

This is different for the remainder of the list:

• “consumer confusion,”

• “diversion/loss of sales,”

• “damage to/adverse reflection on reputation/goodwill,”

• “misrepresentation,” and

• “misappropriation/tarnishing of trademark rights/goodwill.”

A remarkable number of opinions in the group that used these subfactors followed what we can characterize as a “transnational goodwill” approach. This approach can be directly traced to the common law pedigree of the Supreme Court majority’s decision in Steele. It represents the international projection of the traditional common law conception of trademark and goodwill acquisition and protection.⁴⁴³ More concretely, in these opinions, the courts justified application of the Lanham Act based on the exclusive occurrence of one or more subfactors abroad, or on a simultaneous occurrence of one or more subfactors both abroad and within the United States. They thereby allowed for a transnationalization of the analysis—in other words, they permitted both domestic goodwill and foreign-based goodwill to be considered the subject matter of protection. Accordingly, under all these subfactors, both territorial and foreign-based conduct were qualified as “infringing upon” a plaintiff’s trademark or goodwill.

Let us start the more detailed analysis with a look at some concrete examples of the courts’ subfactor analyses. These opinions not only asked for a domestic occurrence of subfactor phenomena but also found sufficient effects to exist if these phenomena occurred on foreign territory. The Steele majority set the stage for this approach:

In the light of the broad jurisdictional grant in the Lanham Act, we deem its scope to encompass petitioner’s activities here. His operations and their effects were not confined within the territorial limits of a foreign nation. He bought component parts of his wares in the United States, and spurious “Bulovas” filtered through the Mexican border into this country; his competing goods could well reflect adversely on Bulova Watch Company’s trade reputation in markets cultivated by advertising here as well as abroad.⁴⁴⁴

In other words, the majority found effects in potential damage to the plaintiff’s goodwill that extended across the United States and Mexico.⁴⁴⁵ By connecting the Lanham Act’s jurisdictional grant with effects on commerce, and then connecting effects on commerce with a concept of rights that covers all geographic areas where the owner’s goodwill exists, the majority established the basis on which later courts built domestic rights’ extraterritoriality. Indeed, courts after Steele significantly extended this idea of goodwill protection on foreign territory. Of course, some courts distinguished between the national and international contexts, but many did not. The Second Circuit’s Atlantic Richfield decision illustrates a cautious analysis:

At best, [the plaintiff] has shown that [the defendant] has a geographic presence in the United States and … that some decision-making regarding [the defendant’s] foreign activities has taken place on American soil. We do not think that such a presence suffices to trigger an extraterritorial application of the Lanham Act. The ultimate purpose of the Lanham Act pertinent to this appeal is to encourage domestic sellers to develop trademarks to assist domestic buyers in their purchasing decisions. … Where (i) an alleged infringer’s foreign use of a mark does not mislead American consumers in their purchases or cause them to look less favorably upon the mark; (ii) the alleged infringer does not physically use the stream of American commerce to compete with the trademark owner by, for example, manufacturing, processing, or transporting the competing product in United States commerce; and (iii) none of the alleged infringer’s American activities materially support the foreign use of the mark, the mere presence of the alleged infringer in the United States will not support extraterritorial application of the Lanham Act. The presence of a foreign infringer, without more, simply does not call into play any purpose of that Act.⁴⁴⁶

Yet the majority of courts have been less critical. In 2005, the First Circuit—visibly aware of the problem of policy and goodwill overextension—illustrated this approach in its creation of a new test variant in McBee:⁴⁴⁷

McBee’s second argument is that Delica’s sales have confused Japanese consumers, hindering McBee’s record sales and touring career in Japan. Evidence of economic harm to McBee in Japan due to confusion of Japanese consumers is less tightly tied to the interests that the Lanham Act intends to protect, since there is no United States interest in protecting Japanese consumers. American courts do, however, arguably have an interest in protecting American commerce by protecting McBee from lost income due to the tarnishing of his trademark in Japan. Courts have considered sales diverted from American companies in foreign countries in their analyses. … Assuming arguendo that evidence of harm to an American plaintiff’s economic interests abroad, due to the tarnishing of his reputation there, might sometimes meet the substantial effects test. …⁴⁴⁸

Roughly speaking, three different aspects of the Steele common law extensions and its impact can be distinguished. The first is a general extension of legal policies underlying domestic trademark and unfair competition law; in particular, this concerns the subfactors of “consumer confusion” and “misrepresentation.” As courts in the Ninth and the Fifth Circuit have assumed, the Lanham Act’s policies extend beyond the domestic domain into international transacting. One example is the Southern District of California’s 1989 Van Doren Rubber Co., Inc. v. Marnatech Enterprises, Inc.⁴⁴⁹ decision, in which the court explained:

The Lanham Act imposes upon this Court “the duty to protect the entire gamut of purchasers, including non-English-speaking purchasers, in various countries throughout the world to which the defendants intend to export their [counterfeits].” … Moreover, “Congress has the power to prevent unfair trade practices (even) in foreign commerce by citizens of the United States, although some of the acts are done outside the territorial limits of the United States.”⁴⁵⁰

But Van Doren Rubber Co. not only reveals a view under which the Lanham Act’s policies are considered universal. In addition, it illustrates a second instrument of overextension—the diversion-of-sales subfactor. As many courts have assumed, a diversion of sales, even on foreign territory, should be considered an invasion of foreign-based goodwill and should thus suffice to trigger application of the Lanham Act. In Van Doren Rubber Co., the court also found a diversion of the plaintiff’s foreign-based sales (in Mexico) and a resulting “decrease [in] the value of the American plaintiff’s consolidated holdings,” as well as direct damage to the “plaintiff’s goodwill not only in Mexico but in the United States.”⁴⁵¹ Further examples of the diversion-of-sales factor are the Ninth Circuit’s Reebok Intern. Ltd. v. Marnatech Enterprises, Inc. decision and the Second Circuit’s Totalplan Corp. of America v. Colborne ruling. In Reebok, the Ninth Circuit wrote:

The district court found that, at the very least, Betech organized and directed the manufacture of counterfeit REEBOK shoes from the United States and knew that their counterfeit shoes went back to the United States with regular frequency. The district court further found that Betech’s sales of counterfeit REEBOK shoes decreased the sale of genuine REEBOK shoes in Mexico and the United States and directly decreased the value of Reebok’s consolidated holdings. … A review of the record indicates that those findings are in no way clearly erroneous. … Betech’s activities thus affect American foreign commerce in a manner which causes an injury to Reebok cognizable under the Lanham Act.⁴⁵²

The Second Circuit used virtually the same language in Totalplan, explaining that

the district judge did not err in finding that Totalplan failed to demonstrate that Lure’s shipment of Love cameras abroad had a substantial effect on United States commerce. Unlike Bulova, there is no evidence that infringing goods have affected United States commerce by re-entering the country and causing confusion. Furthermore, although Totalplan relies on the Fifth Circuit’s decision in [1983 American Rice] for the proposition that the packaging and shipment of goods from the United States constitutes a “substantial effect” on United States commerce, American Rice merely established that such activities, when combined with diversion of foreign sales from a plaintiff, constitute “more than an insignificant effect on United States commerce.”⁴⁵³

More recently, finally, the Southern District of New York openly drew a direct line from the diversion-of-sales subfactor to the Steele conception of transnational goodwill:

U.S. consumer confusion or harm to the plaintiff’s goodwill in the U.S. certainly suffices. … Financial harm to an American trademark owner whether from the loss of foreign sales or the damage to the trademark owner’s reputation abroad is at the very least, relevant to determining whether foreign infringement has a substantial effect on U.S. commerce. See Bulova Watch, 344 U.S. at 287 … (citing fact that defendant’s “competing goods could well reflect adversely on Bulova Watch Company’s trade reputation in markets cultivated by advertising here as well as abroad” as a factor weighing in favor of extraterritorial application of Lanham Act).⁴⁵⁴

Finally, a third aspect of transnational rights extension can be found in the Luft pedigree of Steele,⁴⁵⁵ which is still alive. Until today, a number of courts have expressly described international trademark conflicts as an issue of foreign-market protection asking for the parties’ positions in light of the traditional goodwill paradigm. In particular, this surfaces in the subfactors “damage to reputation” and “misappropriation of goodwill.” One drastic example of this perspective is the Third Circuit’s decision in the multijurisdictional trademark conflict Three Degrees Enterprise, Inc. v. Three Degrees Worldwide, Inc.,⁴⁵⁶ which refers directly to the Tea Rose/Rectanus doctrine:

[Plaintiff] Enterprise is unable to rely upon a registered mark. Accordingly, it is entitled to protection only in geographic areas where it has established a market for its goods. … The Court [in Hanover Star (1916)], held that the trademark of a prior user should be protected from infringement by a subsequent user of the same mark only in areas where the prior user has established a market for its goods: Since it is the trade, and not the mark, that is to be protected, a trademark acknowledges no territorial boundaries of municipalities or states or nations, but extends to every market where the trader’s goods have become known and identified by his use of the mark. … Thus, the senior user of a common law mark may not be able to obtain relief against the junior user in an area where it has no established trade, and hence, no reputation and no good will. … It is in this context that the district court concluded that Enterprise had demonstrated “no presence” in locations other than the United States and Monte Carlo. The fact that Enterprise may have had isolated contracts in the past to perform using its service mark in England, Japan and Bahrain does not establish that it has accomplished the kind of market penetration that would warrant a worldwide injunction or even an injunction covering those countries.⁴⁵⁷

The paradigm of transnational goodwill protection in the sense of an organic and apolitical extension of rights across state and national borders can also be explained in numbers:

Altogether, 119 opinions (out of 140 (85%)) have made use of one or more subfactors. Among these, 59 opinions (49.58%) considered subfactors under the transnational-goodwill paradigm with respect to foreign-based scenarios. Among other things, they considered a loss of sales abroad or confusion of foreign-based consumers as potentially relevant to trigger a positive finding of the respective subfactor—and thus of a positive outcome for the “effects on US commerce” test. The extraterritoriality rate among these opinions (i.e., the percentage of opinions that actually applied the Lanham Act) is 72.88% (43 out of 59), compared to the overall rate of 59.29%.⁴⁵⁸

V Summary: An Era of International Trademark Propertization

Trademark and unfair competition conflicts law in the United States reflects a certain paradox. At the level of substantive trademark law, courts and scholars became increasingly cautious about extending protection into distant product markets starting in the 1930s. In addition, the issue of preventing anticompetitive extensions was hotly debated in domestic law, and concerns about trademark monopolies troubled decision makers until the 1960s.⁴⁵⁹ And today, as we have seen, the extension of rights is still seen under a critical lens.⁴⁶⁰ This consideration of the downsides of extensive protection, however, was never reflected in conflicts doctrine. On the contrary, as the Steele reasoning and progeny reveal, trademark conflicts law has been driven by an opposing trend, extending property rights and domestic interests further and further.

There are numerous possible reasons for this development. Trademark extraterritoriality may be due to an overly casual or unmindful application of precedent, or to the virtuous (albeit naïve) ambition to protect foreign consumers and foster the efficiency of foreign markets. Also, the assumption of the superiority of American trademark policies may have nourished (similarly naïve) ambitions to increase global welfare by extending domestic law. However, one aspect in particular stands out as influential: from the courts’ point of view regarding private-party disputes, owners of national trademarks—the majority of whom consist of national entities—seem to be best protected against international trademark piracy and unfair competition by an extension of the Lanham Act. The idea that extraterritoriality is beneficial for domestic concerns continues to dominate the debate.⁴⁶¹ This also explains the difference with regard to the debate concerning right extensions in the domestic arena: for domestic trademark conflicts, the benefits and detriments of the overextension of rights will have to be allocated and distributed within the national economy; a zero-sum game seems inescapable. The international extension of domestic rights, by contrast, appears to generate unidimensional rent transfers with domestic gains and foreign-based costs—in any event, it seems to prevent unjustified and illegitimate rent transfers from domestic right owners to foreign infringers. In this regard, the phenomenon of excess Lanham Act subject-matter jurisdiction is part of a bigger picture—notably a tendency in international economic law in which domestic concerns are prioritized over the interests of foreign constituencies.⁴⁶²

I will address these issues—particularly the question whether an approach based on the extraterritoriality of rights is effective or detrimental—in chapter 5. At this point, it suffices to conclude that Steele has not only perpetuated American courts’ nineteenth-century tendency to protect exclusive rights against competition in favor of the prevailing mercantile and entrepreneurial elites⁴⁶³ but also transnationalized the once domestic dogma of investment protection. With respect to the doctrinal and structural foundations, it was the common law goodwill paradigm, particularly the organic, market-based, and nonterritorial nature of trademark rights, that provided the foundation.

Conclusions

This comparative look at trademark and unfair competition conflicts doctrine has revealed a key divergence between the common law and the civil law. The central paradigm of civil law doctrine is the concept of state-granted privileges. The territoriality of trademark rights is one result of this German formalism. US common law, by contrast, has never been similarly attached to territoriality; protection has been and remains an issue of use-based rights. These rights are connected primarily to their owners’ market activities. Nicholas de Belleville Katzenbach has lucidly explained this critical divergence in general terms with regard to English and continental law. His description also holds true for the comparison between US common law and German or European civil law:

Although it contained an element rooted in jus gentium, English private law was by and large a matter of remedies; historically its whole development had been in terms of expanding writs not rights. Statutes were few, loosely worded and drafted largely in terms of command to judges, hence in procedural language. Rights did not have their theoretical origin in the positive law, but rather in custom and morality—principles not formally tied down to political boundaries. The civilian, on the other hand, was greatly concerned with the other side of the coin. He, too, universalized his rights, but the limits of judicial authority were defined in terms of these rights (not writs) with their source in the written provisions of codes, statutes and ordinances.⁴⁶⁴

Seen in this light, the tale of German and US trademark conflicts doctrine is quite representative of the history of common law and civil law in general. For US common law, the fundament of trademark and unfair competition regulation has always been an issue of goodwill protection—which, by definition, is detached from political boundaries. For German civil law, the reign of the legal regime has always determined the scope of rights. Territoriality has therefore always been inherent to the system.

In addition, other commonalities and many more differences between German, European, and US doctrine have become visible in this historical comparison. First, the concept of trademark-as-property protection continues to be implemented in German and European law. Even though it appeared to fade in early twentieth-century German doctrine, the formalist trademark-as-property perspective has largely returned, particularly in the supranational rules of European trademark law instruments. In addition, the distinction between rights protection and conduct prevention has effectuated an internal separation. Early trademark law sought to protect the public from fraud. Over time, this impetus was lost—today, property protection is the main emphasis, while consumer protection is of secondary concern. Unfair competition law, by contrast, started on a foundation in tort law. Individual rights were paramount, and there was little regard for consumer protection. Yet contrary to trademark law, unfair competition doctrine has increasingly incorporated public policy concerns. It has become “politicized” and “socialized.” In US law, of course, trademarks also epitomize private rights and individual property. But American nonformalism stands in stark contrast to European doctrine. In the United States, both trademark protection and unfair competition prevention are founded on a paradigm of goodwill protection, which is closely tied to the extension of marketplaces. This common foundation represents the homogeneity and uniformity of the two fields. In addition, consumer protection has remained a key concern throughout. But here as well, a trend toward propertization exists. The extension of trademark-as-goodwill protection into areas beyond consumer confusion prevention increasingly disconnects the two sectors.

These substantive law characteristics have also coined the evolution of conflicts doctrine. Modern German law still reflects a historical concept of state-granted privileges. Ever since it abandoned the nineteenth-century theory of personality rights protection, trademark conflicts law has adhered to a principle of strict territoriality. For a long time, therefore, a conduct-oriented rule of the lex loci protectionis has governed. Unfair competition choice of law, by contrast, is governed by a collision-of-interests, or marketplace, rule. Application of the law at the place where competition actually occurs is the new paradigm. By and large, this appears to be much better equipped to handle the challenges of globalized societies and economies. Not surprisingly, the lex loci protectionis rule in trademark conflicts law has recently been watered down in order to adapt to modern communication and marketplace conditions. The US system of international trademark and unfair competition law, by contrast, has always been based on an idea of goodwill protection and commercial effects. Prior to the Lanham Act, trademark conflicts were resolved under common law principles. Courts did not distinguish between intrastate or interstate conflicts. Domestically, this approach raised few concerns. But the general disregard for state sovereignty has also come to influence international trademark doctrine. The Supreme Court’s Steele majority made Tea Rose/Rectanus virtually borderless.

In this light, it is evident that several dichotomies must be overcome in order to reconceptualize conflicts doctrine for the twenty-first century. For European doctrine, the hiatus between trademark protection and unfair competition prevention has become particularly questionable. The theory of state-granted privileges and an increasing propertization of trademark rights, accompanied by the concurrent socialization of modern unfair competition doctrine, have created a distinct bifurcation of the field. This not only distorts practical utility but fundamentally disregards the fact that trademark and unfair competition law has always been an area of market communication regulation. Current conflicts doctrine in Europe is accordingly distorted. While strict territoriality under the lex loci protectionis rule in trademark conflicts invites underregulation, the marketplace principle in international unfair competition conflicts is far from clearly defined. Most problematic is the adherence to conduct as the most relevant connecting factor. Under modern trademark and unfair competition regulation, conduct has become obsolete. Therefore, much more emphasis should be put on information infrastructure and consumer decision making. In this regard, at least upon first sight, the American Bulova testing promises a more consistent approach. Yet the concrete implementation of this commercial-effects testing has not only resulted in an overextension of domestic trademark rights but also perpetuated another obsolete paradigm of trademark and unfair competition law’s past. The protection of goodwill under a virtually apolitical, and therefore nonterritorial, common law approach has contributed to an overly extensive Lanham Act extraterritoriality. Here, the challenge is to formulate a more qualitatively governed effects test.