What entities can be liable under the alien tort statute? At a minimum, it has been established that natural persons and foreign sovereigns can be successfully sued under the ATS. Beyond that, there is, to put it mildly, quite a bit of disagreement, most of it centered around whether or not a juridical person can be a validly named defendant alongside humans and nations.

The legal historians’ amicus brief [PDF] in support of the cert petition in Kiobel addresses this question in the context of the common law’s historical approach to questions of liability, arguing that “the Second Circuit erred in concluding that ‘who is liable for what’ is a matter of customary international law,” and that, in the United States, entities are liable for their agents’ torts without regard to the source of the substantive norm of conduct, and this applies to the ATS no less than it would any other tort statute. Although I largely agree with the brief, I wanted to expand upon some of its arguments here, and, in particular, to challenge the claims that, under international law, a “corporation” is a distinct and insular category that can or ought to be afforded its own set of rules. Questions of corporate personality — and for that matter, natural personhood — were not cognizable on the international plane at the time of the ATS’s enactment, as distinctions between sub-state entities could only be made through a state’s domestic law. Accordingly, the question of whether a non-state actor has committed a violation of international law is entirely independent of how a state’s domestic law assigns legal personhood to its subjects.

Corporations & Humans: Pretty Much the Same, As Far as Customary International Law Is Concerned

Those who argue against corporate liability under the ATS, as does the majority in Kiobel, often attempt to portray the idea of corporate liability under international law as a revolutionary concept, as something qualitatively unique from the idea of individual human liability for violations of international law. But this argument places far too great of an emphasis upon the alleged distinction between “natural” and “juridical” persons — or rather, upon international law’s ability to distinguish between the two, in the context of civil liability.

Under the law of nations, there are only two real categories of beings — sovereigns and not-sovereigns. Sovereigns are states. Not-sovereigns are the subjects of states, and they take whatever form the state has established for them. As such, in the view of international law, corporations are not exotic creatures that exist in a legal realm distinct from humans; rather, the legal personhoods of corporations and humans alike are necessarily creations of domestic law, over which international law has traditionally had very little interest. There is no a priori distinction between natural personhood and juridical personhood; as far as international law is concerned, they are both just different kinds of non-sovereigns. Until the advent of the modern human rights regime, there was never any requirement that “legal personhood” include all humans, nor has it ever been required to exclude all non-humans — it was the prerogative of states to establish legal personhood for their subjects in whatever manner they saw fit.

Although non-sovereigns are capable of violating the law of nations, international law does not provide a civil remedy for not-sovereigns that are themselves victims of international law violations. Only domestic law does, unless and until states create through treaty an adjudicative body capable of hearing their grievances. As such, there simply cannot be any customary international law of non-sovereign remedies for civil violations of international law — it either exists by treaty or by domestic law, or not at all. Therefore, which forms of legal personhood are subject to domestic liability for violations of international law is a determination that is necessarily made by domestic law, or occasionally by treaty.

Thus, domestic remedies for violations of international law are matters left to individual states. International law did not force the United States to have an alien tort statute, nor did international law have any opinion on what sort of relief the ATS should afford — as far as international law is concerned, the United States could have enacted ATS jurisdiction solely for violations of international law that are committed on the last Tuesday of every month. The contours of the ATS were determined by the state, not international law; it is, above all, a domestic statute, and thus it is a step removed from pure incorporation of international law. Although agreement on this point is far from universal, the strongest argument is that the ATS has not provided a cause of action “directly” from international law– rather, in actual practice, the ATS has required the courts to create U.S. common law that is derived from international law. Thus, when our courts hear ATS cases, they are not technically deciding what international law is, but what federal common law’s conception of international law is; and, in effect, turning international law as a whole into persuasive rather than mandatory authority.

All of this is a round-about way of getting to the point of this post — that the idea that there is a binary distinction between “natural persons” and “juridical persons” is not supported by the history of the law of nations, nor is there any principled means for excluding incorporated entities from the kinds of non-sovereigns that are subject to liability under the ATS.

The State’s Prerogative Over the Legal Personhood of its Subjects

Because states are the only entities capable of establishing what entities possess legal personhood, only a state can establish which entities are capable of incurring liabilities under domestic law. States are thus the creators of all non-state liability units. For brevity’s sake, I am using here the term “liability unit” as shorthand to mean any legally defined entity that is capable of incurring private liabilities which will be enforced by a state’s civil processes. Although liability units can vary in scope — different rights and legal processes can be available to some liability units and not others — all liability units are based on the idea that a certain set of assets have been delineated and partitioned from the rest of the assets in the world, and given an accompanying assortment of possessory and contractual rights. And, where one unit seeks recovery against another unit, either in tort or contract, any recovery is limited to the assets associated with that liability unit. (For now, to simplify matters, I am going to ignore the fact that liability units are often rather fuzzy on borders — the partitions between them are porous, and dependent upon the particular contours of domestic law, but liability units do not need to be air-tight in order to have a meaningful existence).

In effect, a liability unit is an entity that has been incorporated by the state to possess an existence cognizable by its court systems. This form of incorporation can be attached to a single human being, or it can be attached to a pool of investors’ funds, or to a wide range of other possible groupings. It has two main components: partitioned assets over which ownership is either vested in the unit itself or vested in another unit(s) by contract, and a determination by the state as to who has decision-making authority over the use of those assets.

In most natural persons, there is a unity in ownership and control — the owner of the partitioned assets, i.e., that person’s property, is controlled by themselves. But the concept of separation of ownership and control is not unique to corporations. It can also apply to natural humans, as the state possesses the power to establish any sort of asset grouping as a liability unit — and to designate which human or group of humans it wishes to invest with the power to make decisions regarding use and disposition of those assets. For example, infants and wards of the state can still own assets and incur civil liabilities, but decisions regarding how to manage the property or what liabilities ought to be incurred are made by a guardian, not by the infant or the ward. A similar arrangement occurs upon a liability unit’s natural death; although a human being may die, the liability unit, which had been superimposed over the human as part of the state’s legal latticework, continues to exist in the form of his or her estate, which persists under control of a trustee until it is formally extinguished through the legal process. The human and the liability unit are related, in that the death of the human does have an effect on the corresponding liability unit, but their coexistence is mutable.

A corporation is, in effect, a liability unit that has partitioned assets in the form of its investors’ funds, over which decision-making control has been granted to a board of directors; it just happens to be a liability unit that is not directly mapped over an individual human being. Similarly, natural personhood under the law is a human being that has a corresponding liability unit, which provides them with legal recognition in the state’s court system and permits them to engage in private transactions that can be enforced through the use of state mechanisms. It is very much possible for an individual human being not to be vested with incorporation into legal personhood — and throughout much of human civilization, having the majority of individual humans be endowed with full legal personhood was the exception rather than the rule.

It is the state that has always established what human bodies are raised to the status of liability unit, and which human bodies are simply human bodies without a corresponding civil legal existence. Today, the overwhelming trend is for states to incorporate all human beings, so that all humans are established as entities that can sue and be sue, own property and incur debts, or enter into private contracts. (For humans below a certain age, or deemed by the state to be incompetent, the state generally continues to provide for full status as a liability unit, but separates control from ownership.) Liability units are certainly not limited to individual humans, but, as it is generally presumed today, every human is automatically its own liability unit.

This is an obviously rather modern state of affairs. For most of human history, fully independent liability units were bestowed only upon certain human bodies belonging to a particular class — it was not granted to all human beings as a matter of right. Nor is a liability unit necessarily limited to just a single, independent human being. A single liability unit can, as with the case of marriage, encompass two or more individual humans. Take, for instance, Blackstone’s description on the effect of the marital union:

“By marriage, the husband and wife are one person in law: that is, the very being or legal existence of the woman is suspended during the marriage, or at least is incorporated and consolidated into that of the husband: under whose wing, protection, and cover, she performs every thing[.]”

Traditionally, through marriage, a single liability unit consisting of two human beings was created, over which decision-making authority was vested with the husband — an arrangement which, under the common law, effectively prevented any plaintiff from bringing a tort claim against a spouse, as a single entity logically cannot sue itself. And, reaching even further back, under Roman law, the standard liability unit was a larger combination still: the family. The oldest living male and his descendants were, for the large part, treated as the standard unit of sub-state ordering, with the decision making authority concentrated in the oldest living male. (The pater familias could create a further sub-entity by spinning off some of the family’s assets into a peculium, which featured something of a one-way partition, and individuals could enter into partnership-like societas, but it was the family liability unit that was most closely analogous to the modern day individual human liability unit.)

Liability units can even consist of inanimate objects. As is well documented in the legal historians’ brief, there is a long history of individual ships being established as liability units under the domestic law of many jurisdictions, including the United States. Back in medieval Europe, the division of the liability unit could be smaller still — under the commenda, a single ship voyage had aspects of being itself a discrete liability unit.

And of course, for the first century of the United States’ existence, the legal personhood of corporations was more absolute than that of humans; prior to the 14th Amendment, there were corporations that could sue and be sued for tort, while there were human beings who could not. A slave was thus a human being that was not a liability unit. Although laws differed among individual U.S. states, in general, a slave could not bring suit in his own name, nor could suit be brought against him. In many cases, a slave’s owner was civilly liable for the torts of the slave, and would definitely be liable where the tort was committed in the course of the slave’s duties for his master.

Criminal law is different — an entity that is not a liability unit can be prosecuted for a criminal violation. A slave may not be sued for breach of contract, but he can be prosecuted for murder; a husband is held liable for a wife’s torts, but not her acts of treason. But civilly, there was no “customary international law of slave tort liability” or “customary international law of married woman tort liability” — some nations recognized such forms of liability, and some did not, but international law had nothing to say on the matter.

So what would have happened if, sometime in the early 19th century, a slave had, in the course of performing his duties for his master, carried out some blatant violation of international law? If the slave had (somehow) become a pirate, or failed to abide by a safe conduct issued to an alien, all for the financial benefit of his master? Would the alien that had been so aggrieved been left with no recourse under the ATS, because there was no customary international law requiring that all individual human beings be civilly liable?

Well, if a court in 1805 was somehow able to channel the decision in Kiobel, they might have held that the law of nations had no theory of liability for violations of the law of nations committed by slaves, let alone any theory of agency law under customary international law, and therefore find that they had no jurisdiction to adjudicate the claim. However, I think it much more likely that the court would have applied principles of agency law to find that the owner of the slave was liable for the tort, and that the question of whether or not a slave could be civilly liable for a violation of the laws of nations is answerable only by the domestic regime entertaining the cause of action.

Recognition of Foreign Liability Units

Because liability units are creatures of domestic law, their existence is only guaranteed within the territorial borders of the sovereign that granted them a legal form. As such, when liability units cross international borders, their continued existence as a liability unit is entirely at the mercy of the domestic law of the state where they now reside. Although in most cases, through comity, nations regularly recognize each others’ established liability units, unless it is otherwise regulated by treaty, this is merely a courtesy and not a guarantee.

This does not apply merely to juridical persons; it is equally applicable to natural persons. The question of whether an individual human is a liability unit is also a determination made by domestic law; although a person may be a liability unit in one nation, they are not necessarily so in another. As was noted in the Dred Scott decision, a slave going to England temporarily suspended his status of slave while there — for the “slave” was simply a human entitled to the same status as any other individual liability unit in England — but upon his return to the United States, “the slave resumed his original character of slave.”

Corporations, which are not bound up with the existence of any natural human being, pose an obviously more difficult situation when they cross international borders. When should one state recognize a corporate entity that has been granted status as a discrete liability unit by a foreign state?

To be continued in Part II. Because this post is already too long and rambling as it is.

-Susan