SHOULD AMERICAN COURTS PUNISH MULTINATIONAL COMPANIES FOR THEIR ACTIONS OVERSEAS?
More On Indirect Injuries And The Alien Tort Claims Act

This column is Part Two of a two-part series by Professor Sebok on the Alien Tort Claims Act, indirect injury, and joint action. Part One is available in the archive of Professor Sebok's columns on this site. - Ed.

The phenomenon of globalization has reshaped how we talk about the moral responsibility of consumers for what occurs in places far away from our own liberal democratic society. Even if one chooses, in the end, to buy sneakers from Nike, no one can pretend that what he or she buys in K-Mart does not affect the lives of workers in other countries.

But how seriously should we take the idea that "everything is connected"? It is one thing to make choices about what to buy--or even criticize one's friends for what they buy. But should American courts punish companies for the consequences - and even the indirect consequences - of their actions overseas?

That question is at the heart of the recent suit against Citibank, UBS, and Credit Suisse under the Alien Tort Claims Act ("ATCA") for their business activities in South Africa before the end of apartheid.

Why U.S. Tort Law Needs to Think Carefully About Indirect Injuries

As I described in my last column, the South Africa suit is unlike the Holocaust suit brought earlier (by some of the same lawyers) against Swiss, French, Italian and German corporations, in the following way:

Those suits alleged that the firms had directly injured Jews and other enslaved people by taking their money and labor without compensation. In contrast, the South Africa suit - like a handful of suits brought against multinationals operating in Myanmar (Burma), Nigeria, and Indonesia - allege indirect injuries. More specifically, the South Africa suit alleges the companies "did business" with governments that violated the human rights of the plaintiffs, and thereby indirectly caused the human rights violations.

In moral philosophy, there may be very little difference between direct injury and indirect injury. Sophisticated moral criticism can afford this luxury, since moral judgment is often fine-grained and measured. But tort law is not so subtle.

Thus, if the US is going to globalize its tort law with regard to human rights, it had better figure out what to do about indirect harm.

Being in A Foreign Country, and Observing Abuses, Is Not Enough.

As a starting point, it is well-known hornbook tort law that, absent a duty based on a special relationship, there is no duty to rescue an innocent party in peril. Whatever one may think about this rejection of Good Samaritanism, it is longstanding aspect of our law that is ingrained in our system.

This might seem to be an odd point to raise, since the multinationals that are being sued did something: They decided to establish a presence in the relevant countries. In contrast, someone who refuses to be a Good Samaritan refrains from doing something.

Yet the "no duty to rescue" point brings out an ugly truth - and a larger point - about tort law: In general, tort law does not impose duties on others to be moral. Rather, it imposes a range of duties that fall into a very narrow subset of the broader range of duties that are imposed by conventional morality.

Furthermore, simply being in a foreign country does not create duties under American tort law. For instance, a tourist in South Africa in 1975 would have had no tort duties to help anyone, even if she witnessed a police beating and could have intervened.

Anyone who thinks that the ATCA says otherwise has taken the "tort" out of the law and turned into something quite different than what Congress passed in 1798.

The Allegations Paint the Banks As More Than Passive Observers of Apartheid

Still, the Swiss banks are not accused of simply sitting and watching apartheid in action; their actions allegedly aided the violations of the plaintiffs' human rights. Specifically, the complaint alleges that from 1985 until 1993 the banks provided loans to a desperate South African government, which were "used to directly continue apartheid's reign of terror."

The real question, then, is when is economic activity in a nation that is engaged in human rights abuses actionable under tort law?

A Few Examples of Those Who Clearly Are And Are Not Liable Under the ATCA

Some of the limiting principles governing the circumstances under which economic activity should, and should not, lead to liability are clear.

For instance, consider the example of the banker Emil Puhl - raised by Anita Ramasastry in an earlier column for this site. Puhl is someone who would obviously be liable under the Alien Tort Claims Act.

Puhl was convicted by the United States Military Tribunal for crimes against humanity because of his role in managing the German Reich bank. He had not only knowingly received from the Germans military property taken from victims in concentration camps, he had actively organized the plundering in the camps. Moreover, the bank for which he worked was not a private bank--it was the state's central bank.

On the other hand, consider the example of an amoral tourist who visited South Africa during the apartheid era because it offered inexpensive vacations, and who spent $1500 while there. Despite the fact that he was knowledgeable about the human rights violations committed by the government, he did nothing to alleviate them. Instead, he propped up the regime by injecting his dollars into the South African economy.

The tourist, however poorly one might judge his conduct from a moral standpoint, is clearly not liable under the ATCA.

Searching For A Test For Indirect Liability Under the ATCA

So where do the banks fit in? Are they more like Puhl, or more like the tourist? Where are they on the spectrum between these two clear cases, and why?

For all but a handful of crimes against humanity, the ATCA has a "state action" requirement. Pursuant to that requirement, a plaintiff may sue only if the defendant was a state actor - for instance, a police officer or prison guard - or if he or she acted "under color of state law."

What does it mean to act "under color of state law" when one is not necessarily a state actor? The concept is taken from out own domestic civil rights regime - and specifically from the federal statute that allows individuals to sue for tort damages when their constitutional rights have been violated.

Borrowing from this statute (which is called "Section 1983") makes some sense. Like Section 1983, the ATCA allows individuals to sue for tort damages where certain human rights--those protected under international law--have been violated.

What are examples of conduct that might be done by someone who was not a government employee but who nevertheless acted under cover of state law? Think back to the South in the '50's when Jim Crow rules would be enforced by private citizens who were acting in close coordination with local police.

Or, think of an example given by the decision of the United States Court of Appeals in the Kadic v. Karazic case. There, the court held that a paramilitary commander was acting under color of the law of the former Yugoslavia when he committed mass atrocities. Though the paramilitary may not have been a state actor (after all, it was a paramilitary, not a military), it still acted "under color of state law."

"Under Color of Law" Tests That May Apply to the ATCA

Beyond these examples, what are the legal tests that determine when a private citizen is acting "under color of state law" for purposes of Section 1983 - and thus arguably for purposes of the Alien Tort Claims Act as well?

The case that best summarizes the tests - and that is most cited by courts borrowing the 1983 test and applying it in the ATCA context - is Gallagher v. "Neil Young Freedom Concert," which was decided in 1995. The facts are simple enough: the plaintiffs sued the private organizers of a concert at the University of Utah. They alleged that the promoters directed the security guards at the concert to subject them to "pat down" searches in violation of their constitutional rights. But the promoters argued that they could not be sued under Section 1983 because they were not state actors.

The U.S. Court of Appeals for the Tenth Circuit tested the promoter's argument by asking to what extent the promoter's relationship with the university (a state actor) meant the promoter had acted "under color of state law." It noted if the promoter and university were engaged in a "joint action" when the constitutional rights were violated, that might suffice to satisfy the "under color of state law" requirement.

Under Section 1983 - and thus arguably under the ATCA as well - "joint action" can occur in two different ways.

First, the private actor can conspire with the state actor to violate the plaintiff's constitutional rights. The private actor need not have done very much to interfere with the constitutional rights of the plaintiff. Instead, the plaintiff simply needs to prove that the state and private entities shared a specific goal to violate the plaintiff's rights.

Second, there can simply be a "substantial" degree of cooperative action between the state and private actor. The plaintiff need not prove that the private actor had a specific intent to violate constitutional rights. Instead, the plaintiff can simply show that the private actor engaged in actions that significantly promoted the violation of the rights.

The Problem With Using Section 1983 Tests in the ATCA Context

The problem with using the Section 1983 tests articulated in Gallagher is that they are designed for a different sort of rights violation than we see in the new ATCA cases.

In the Section 1983 context, the problem usually has been that the state hides behind private actors. The Gallagher case is a typical example: The plaintiffs were searched by a private security company under the control of a private concert promoter. The state had to be brought back into the case somehow if the plaintiffs were to prevail.

But in the ATCA context, the problem is often different. Think of the South Africa or Myanmar cases. Human rights violations were directly caused by state actors (soldiers, police, etc.). The allegation is that private actors are allegedly hiding behind the state.

Thus, the situation in the ATCA cases is the precise converse of the situation in Gallagher - where state actors allegedly hid behind private promoters. And the Section 1983 tests applied in Gallagher accordingly may not fit.

Shoplifting and Eviction Cases Provide Poor Comparisons for the ATCA

There are a handful of Section 1983 cases where plaintiffs have claimed "joint action" based on allegations that private actors acted under "color of state law" when they employed a state actor to violate the plaintiff's constitutional rights. Typically, a store uses city police to arrest shoplifters or to have customers removed by force from the premises, thus allegedly violating constitutional prohibitions against false arrest or racial discrimination in a public accommodation.

These cases usually fail. Conspiracy allegations are almost impossible to prove. Accordingly, the plaintiffs are left trying to show that the store substantially assisted the police - rather than just calling 911 as anyone can do. Naturally, this is also hard to prove, and so the cases are usually dismissed.

Were any of these cases winners? One example is Howerton v. Gabica, which was decided by the U.S. Court of Appeals for the Ninth Circuit in 1982. There, a private party was held to have been a joint actor with the police because he intervened at "every step" of an eviction process that violated the plaintiff's civil rights.

But that test may be too demanding if imported into the ATCA context: Must banks similarly intervene in every step of an apartheid abuse to be liable for it?

As these examples show, Section 1983 really hasn't been used to handle cases in which private actors hide behind public actors - and when it has been, has not been used very successfully. That suggests that maybe Section 1983 is not the right model.

Fortunately, there is another alternative - and it comes straight from American tort law. Like the purloined letter, this test has been staring courts in the face the whole time - yet many have overlooked it.

The Restatement On "Persons Acting in Concert": A Better Test?

Indeed, courts that have tried to determine "joint action" have openly admitted that the concept comes directly from common law tort principles - as described in the Restatement of Torts (Second), Section 876, which is entitled "Persons Acting in Concert."

The section addresses the question of when to attribute liability to a party who has assisted in the commission of a tortious act - but not performed the act himself, or performed it alone. There are two answers, depending on whether the assistance takes the form of conspiracy or of "aiding-abetting". As with the "conspiracy" and "concerted action" methods of proving liability in the Section 1983 context, discussed above in connection with Gallagher, each form has its own requirements.

"Conspiracy" requires proof that the party accused of indirect liability - for instance, the banks in the South Africa case - agreed in advance to a common scheme to participate in an unlawful act. Obviously, such a prior agreement probably cannot be proved in the South Africa Act; the allegation is not that the banks plotted abuses, but that they were indifferent to the knowledge that they were funding them by propping up an evil regime.

Aiding-abetting, however, requires no such prior agreement - and thus may be a better theory for the South Africa plaintiffs. It requires three other things, however.

First, aiding-abetting requires that "the party whom the defendant aids must perform a wrongful act that causes injury" - which may be easy to show in the South African case, since the plaintiffs need only prove the banks aided abuses by funding the apartheid government.

Second, aiding-abetting requires that the defendant must be generally aware of his role as part of an overall illegal or tortious activity at the time that he provides assistance. Again, this may be relatively easy to show in the South Africa case; the abuses of apartheid were well-known, and presumably known to the banks.

Third, aiding-abetting requires that the defendant must knowingly and substantially assist the principle violation. This, too, could conceivably be shown in the South Africa case, depending on the evidence as to whether the banks knew, for instance, that their funds were going towards abusive policing that included torture and murder.

This test is taken from Halberstam v. Welch, a 1982 decision written by Judge Pat Wald of the U.S. Court of Appeals for the D.C. Circuit. In that decision, Judge Wald also offers additional insight into the factors important to making out an aiding-abetting case.

In Halberstam, the estate of a man killed by a burglar sued the burglar's wife. The estate alleged that she was liable for the killing because she kept the books for her husband's business - which involved selling jewelry, antiques, furs, and melting down gold coins and rings - and knew of her husband's criminal activities. The wife argued that she could not be held liable for murder because regardless of whether or not she knew of her husband's criminal activities, she did not aid them.

Judge Wald set out five factors that matter to a determination of whether a particular defendant can be held liable for aiding-abetting. They are: the nature of the assistance, the amount of assistance, the relationship between the aider-abettor and the party assisted, the alleged aider-abettor's state of mind, and the duration of the assistance.

In Halberstam itself, Judge Wald held that these factors, taken together, added up to the conclusion that the burglar's wife had provided "substantial" assistance to her husband's crimes and thus could indeed be held liable as an aider-abettor of the crimes.

The definition for aiding-abetting is more precise than the definition of joint action. Nevertheless, they point to the same lesson.

For responsibility under American tort law, proof of awareness of the evils that will occur if one gives aid to a villain is not enough. There must also be proof that one has knowingly provided substantial aid to that villain's evil activity.

These principles should work in the ATCA context as well. There, we can ask: Did the banks knowingly provide substantial aid to the apartheid government's abuses?

More generally, it is worth noting that the tort law has already developed a considerable case law on the question of aiding-abetting. Some of the cases, such as Halberstam, directly raise the question of when financial assistance in exchange for financial gain crosses the line into joint liability.

These cases can provide rich, direct guidance in the ATCA context as well.

Thus, I hope the federal courts - including the court hearing the South Africa case - take advantage of the law that we already have before they choose to make new law when considering the problem of indirect liability under the ATCA.