home biser trauma ozenama karadzic rusenje muskaracizena projekti novosti




Collective harms under the alien tort statute: a cautionary note on class actions

Catharine A. MacKinnon


A small but increasing number of class actions for mass human rights violations are being brought under the Alien Tort Statute, 28 U.S.C. § 1350.  Class actions are representative actions; the one stands for the many.  The mass accidents for which they are often used, such as large plane crashes, or slowly-unfolding corporate catastrophes, such as illnesses from asbestos exposure, involve discrete torts from a single physical cause in particular etiological scenarios. The injuries are not group-based in the human rights sense.  That is, many people are injured because of where they were (on a plane) or what they did (work with asbestos), but not because of who they are.  Human rights violations like genocide and crimes against humanity, by contrast, are not mass accidents.  They involve every imaginable tort to a human being and are done because of who the victims are, based on their race, ethnicity, religion, nationality, and sex.  People are also politically tortured on the basis of their politics and ethnicity, and war crimes are increasingly concerted acts against groups.  When war is an instrument of genocide, war crimes, too, can be group-based acts in the political sense. 

       The question here is whether United States class action instruments under Rule 23 are well suited to redressing international human rights injuries that take a collective form.  Focusing in particular on the Karadzic cases, and to some extent the Marcos cases and the more recent Holocaust Victim Assets cases, my concern here is with the fit between domestic class action techniques, particularly the “limited fund class action” device, under Fed. R. Civ. P. 23(b)(1)(B), and international human rights goals for group-based injuries to groups. 

       Class actions under Rule 23 are brought for injuries to large groups of people when common questions of fact and law are raised in situations where too many parties, plaintiffs or defendants, exist to make joining them all practical.  Some of the devices are mandatory; everyone who was hurt is deemed included.  Some permit people to opt out of the class to voluntarily exclude themselves.

       Class actions in the 60’s and 70’s, to speak very generally, could be called the darlings of the Left.  They made it possible to recognize and redress through law collective injuries, including to groups, on a scale something as vast as the injures were done.  Many of these class actions produced civil rights initiatives that resulted in social change, including prominently in education, employment, housing, and prisons.  The civil rights class action was assumed to be in the interest of the whole class it represented.  Without it, its members had nothing.  In this vision, dissenting class members were cast as greedy spoilers, self-seeking outliers, troublemakers.  Individuals who sought to opt out of these actions, actions assumed to be in the interest of every member of the group as a group including them, were considered self-seeking contrarians, free riders, denying and obstructing the group’s welfare.  They were also a huge pain in the neck for litigators, who imagined themselves on the way to the greatest good for the group, doing justice. 

       In the 80’s and 90’s, to again draw a bit of a caricature, class action devices became the darlings of the Right.  Corporations, even whole industries, found that class actions were made in heaven for controlling their legal exposure to victims of the widespread harms they did.  The utility of class actions emerged as limiting the liability of mass tortfeasors.  Some corporations, it was alleged, and some industry groups, or so plaintiffs asserted, went so far as to initiate their own collusive class actions against themselves, bringing together all the possible victims in one case that they in effect controlled, in order to settle low.  With mandatory classes, the result was that everyone’s liability was limited to whatever those who represented the plaintiffs—who could be real plaintiffs or not—settled for.  The class representatives were permitted to settle everyone’s claims in a way that bound class members whether they consented to be bound or not, then or later—with res judicata effect.  In massive and unpredictable cases like the asbestos litigation (in some jurisdictions, around a fifth of all civil cases were asbestos cases) mandatory class actions took place with the grateful acquiescence of the courts in their ever-persistent pursuit of docket control. 

       The tension between these two images of the class action came to a head of sorts last summer in the Ortiz case, producing a United States Supreme Court decision [3] that shows this shift from the earlier presumption that class plaintiffs had the interest of the class at heart, to the later suspicion that something else may be afoot.  More specifically, it revealed the change from assuming that it is best to get something as opposed to nothing for everyone to questioning whether it is valid to bind all to a group resolution whether they wanted it or not.  One view sees some relief as opposed to none for a group of disenfranchised people whom no one will otherwise represent.  The other sees an overly-hasty presumptive, even overreaching, potentially collusive (whether in fact or in intent) resolution of varied claims cutting off better possible relief later in a form that is susceptible to being run by agendas that are very far removed from full relief—far less justice—for the victims.  These agendas can include, to mention a few, politics, media attention and public speaking opportunities, career and turf-building, development of expertise or its appearance, credentialing and training, fundraising, and attorney’s fees.  When the interests of hurt people are not entirely driving the litigation, class actions can become more in the interest of perpetrators, and vehicles for the advancement of others, than engines of vindication and reparations for the survivors.  

       The class action chameleon that was the particular concern in Ortiz was the “limited fund class” arising under Federal Rules of Civil Procedure 23(b)(1)(B).  The classical limited fund class action arises, for instance, when an insured ship sinks.  The fund for everyone’s recovery is limited because the ship is only insured for so much.  A fixed number of people has a stake in the ship, and the policy limits total recovery, so the thought is that all claims should be litigated together.  Such classes are mandatory in the sense that no opting-out is permitted except rarely by judicial discretion.  No notice is required to 23(b)(1)(B) classes.  People can be bound by the adjudication without ever having heard that it happened. And the results bind all class members whether they took part in the litigation or not or even knew about it. 

       The Court in Ortiz, which concerned a settlement class in asbestos litigation, held that certification of a mandatory settlement class under b(1)(B) required a showing that the fund is limited independent of the agreement of the parties.  You can not just get the lawyers for the class together with the lawyers for the companies, agree to stipulate that “this is all there is,” and divide up the pie.  This invites abuse, such as exchanging avoidance of bankruptcy for the companies for large attorney’s fees to the class lawyers.  So, the Court held, the fund had to be shown to be limited in an external way, interclass conflicts had to be addressed, and class members had to be equitably treated.



       How do these concerns and safeguards map onto human rights concerns, particularly with large victim and survivor classes with collective injuries such as those increasingly occurring on the international stage?  Given a plaintiff class action, how do you know whether it is beneficial and progressive on the one hand, or complicitous and exploitive on the other? 

       Most Alien Tort cases have not been class cases.  They have been brought on behalf of harmed individuals whose human rights were violated, sometimes on or implicating group grounds, sometimes not.  In the spring of 1993, two actions were brought against Radovan Karadzic, the leader of a group of Bosnian Serb fascists who carried out a genocide through war to exterminate and eliminate non-Serbs in Bosnia-Herzegovina.  The two Karadzic cases were brought under the Alien Tort Act and the Torture Victim Protection Act for genocide, torture and war crimes in the Southern District of New York by Bosniac Muslim and Croat survivors, seeking relief for torts of ethnic cleansing committed against them.  One case, Kadic [4] emphasized claims for rape as genocide, rape as torture, and rape as a war crime.  We sought relief specifically for injuries of genocidal sexual atrocities perpetrated as a result of Karadzic’s policy of ethnic cleansing in collaboration with Slobodan Milosovic’s administration in Belgrade, Serbia.  Damages were sought for the named individuals and groups, with an injunction that Karadzic order the genocide to stop.  This is a representative action in the sense that the injuries had a group basis and the injunctive relief would have a group impact.  If you stop a genocide, you stop it for everyone—but the moving parties claimed to represent only those who brought the case.  The plaintiffs were one rather large survivor group, a smaller group, and the named individuals.  The second case, Doe, seeking damages, was brought on behalf of two unnamed young girls claiming to represent a class of “all people who suffered injury as a result of rape, genocide, summary execution, arbitrary detention, disappearance, torture or other cruel, inhuman or degrading treatment inflicted by Bosnian-Serb Forces under the command and control of defendant between April 1992 and the present.”[5]

       By court practice, these two cases proceeded in tandem under a single caption.  Jurisdiction was established over Karadzic by beating back varieties of immunity claims, some known, some previously unknown; a civil claim was permitted under the Alien Tort Act for rape as an act of genocide. [6] Then the Doe lawyers moved to certify the class, [7] which of course subsumed the Kadic plaintiffs’ case.  After some months, this motion was amended to seek, in the alternative, limited fund class certification because Karadzic’s assets were claimed to be limited.  This claim was based on a letter Karadzic had sent to the judge contending, inter alia, that he did “not have the financial resources to bring witnesses for my defense to the United States for either depositions, or trial.”[8] The Kadic plaintiffs sought to opt out of the class, and the class supported them.  The judge, however, certified the class on the limited fund theory and denied the Kadic motion to opt out [9]—over not only the support of the class but over the lack of opposition from the defendant as well.  The Kadic plaintiffs then moved to decertify the class.  The issues under domestic law will be resolved as procedural and due process matters under Ortiz and prior precedents. 

       Two other recent cases have raised similar issues—or potentially so.  Marcos was a limited fund case for torture that received a verdict of $2 billion at trial. [10]  The Holocaust Victim Assets cases were brought beginning during 1997 for claims under the Alien Tort Act for human rights violations, and violations of contract, conversion, breach of fiduciary duty, and other rights.  One claimed a class of all those persecuted and targeted for persecution by the Nazis, divided into three subclasses, those deprived of their assets by banks, subjected to slave labor, and forced to become refugees. [11]  The settlement proposal would permit opting out, even though a fixed amount of total recovery is agreed to between certain Swiss bank defendants and the plaintiffs. 



       This small cluster of critical cases raises two related issues for class actions: adequacy of relief and adequacy of representation.  The issue of adequacy of relief is illustrated in both the Marcos and Karadzic cases.  To us, limiting the relief of all the survivors of the Bosnian genocide because Karadzic says he cannot afford to come to New York seems both wrong and small.  In the Marcos case this last summer, a Philippine court disapproved a proposed settlement that would reduce the $2 billion verdict to $1.5 million based on a Marcos Swiss bank account, noting in particular that a quarter of this amount was slated to go to the lawyers. [12] That court also pointedly noted that it was principally in the interest of the Marcos estate, not the victims, to reduce the very large amount they had won to the much smaller amount of the settlement proposal “consequences which are extremely beneficial to the Marcoses and of minimal benefits for the human rights victims.”[13]

       The Philippine court, noting that Philippine procedural law is heavily based on United States law, citing United States authorities on class suits, observed that the settlement it disapproved would likely preclude future additional relief. 

Whether [the plaintiffs] will initiate a new action against new defendants over the same cause is open to question.  Whether they can even legally do so at this time is speculative . . . Whether human rights victims for the period 1972 to 1986 can still initiate separate suit against anyone else anywhere else is . . . doubtful (emphasis added). [14]

   The court also noted that the Hawaii decision in Marcos was “binding under res judicata principles upon all members of the class, whether or not they were before the court.”[15]  The res judicata effect of discrimination class actions have also precluded class members from suing subordinate tortfeasors for the discrimination. [16]  While this result should be resisted, it threatens to preclude future relief, for example, for individuals who run into their individual rapists on the streets of the United States, because Karadzic’s liability to the class in Doe is predicated on all the acts of all the people who carried out his orders and policies.  The complaint attributes all of it to him and the class definition seeks relief for all of it from him.  If relief from him is then limited by the limited fund, but his responsibility for the genocide is total, a vast amount of injury was just reduced, on the defendant’s “say-so”, to less than the price of a few tickets to New York.  And actual relief for the survivors’ injuries, in this or any other proceeding, is thus, if not undone, rendered speculative to nil. [17]

   The issue of adequacy of representation raises the dangers of asserting a class in human rights litigation when the members have, and can have, no real contact with their purported representatives.  The survivors are far away and speak another language; they may number in the thousands or, as in the Bosnian situation, in the hundreds of thousands.  In on-going policy development, one claims to represent huge numbers of people with whom one has no contact, speaking for them in public or policy settings, taking positions on issues that deeply and directly affect their lives, on which they have diverse and nuanced opinions.  The structure of the limited fund class claims in particular seems actually to discourage contact, even discourages telling members of the class that one is representing them.  Their involvement would make things cumbersome, complex, create cross-currents, become time-consuming, take resources.  Actually representing badly hurt people is a lot of work.  As some of the affiants in our motion supporting the de-certification motion noted, the Doe class usurps many of the functions of elected representatives, which is undemocratic [18]  It could even be termed colonizing. 

   Unsought and unwanted representation in a class raises the possibility that some of the intangible and expressive gains from human rights litigation, especially for group-based injuries like rape in genocide, may be undermined.  Human rights litigation offers people their humanity back. What is stolen from them when they are violated can be partially or potentially returned to them through a process that does not reduce them to the ciphers of group membership the way their perpetrators did.  It treats them as more than the sum of the injuries done to them.  It gives them back a voice in their fate, and the dignity of a place at the table.  For this to work, the process must be accountable, personal, and responsive.  Being forcibly lumped into a group-based class, thereby deprived of direct or actual representation, being represented in name (or no name) only, survivors of group-based atrocities can experience the process as furthering the deprivation of humanity that human rights law promises to restore. 



       On March 27, 2000, the Kadic plaintiffs won their motion to decertify the Doe class under Ortiz. Judge Peter K. Leisure cited, among other grounds, the Kadic plaintiff’s insistence that the Doe plaintiff’s “have been unresponsive to their attempts to secure adequate representation” and “perhaps their most serious accusation . . . the Doe plaintiffs’ willingness to accept defendant’s ‘profession of poverty’ in order to obtain mandatory class treatment [19]  On August 10, 2000, a New York jury awarded the Kadic plaintiffs a total of $745 million in compensatory and punitive damages and a permanent injunction.

[1] Panel on the Alien Tort Claims Act, International Law Workshop.  New York City, November 5, 1999.

[2]00000*000Elizabeth A. Long Professor of Law, University of Michigan Law School; long-term Visiting Professor, University of Chicago Law School; counsel for plaintiffs in Kadic v. Karadzic, 70 F.3d 232 (2d Cir. 1995), cert. denied, 518 U.S. 1005 (1996).  Particular thanks for our work together on these issues to Natalie Nenadic, Asja Armanda, Aida Daidzic, Liza Velazquez, and Maria T. Vullo. 

[3] Ortiz v. Fibreboard, 119 S. Ct. 2295 (1999). 

[4] Kadic v. Karadzic, 70 F. 3d 232 (2d Cir. 1995), cert. denied, 518 U.S. 1005 (1996).

[5] Doe v. Karadzic, 176 F.R.D. 458, 461 (S.D.N.Y. 1997).  Doe counsel later clarified the claim as covering persons injured from April 1992 to February, 1993.

[6] Kadic, n. 3 supra.

[7] All the other panelists worked on the Doe case, at one time or another. The class certification motion was filed when Beth Stephens, original lead counsel, was no longer actively associated with the case, and after Harold Hongju Koh, who contributed at a prior crucial period, had withdrawn to assume his position with the State Department. 

[8] Letter from Radovan Karadzic to the court 1 (1997), quoted in Doe v. Karadzk, 176 F.R.D. 458, 463 (S.D.N.Y. 1997).

[9] Doe, n. 4 supra.

[10] In re Estate of Ferdinand Marcos, 25 F.3d 1467 (9th Cir. 1994).

[11] In re Holocaust Victim Assets Litigation, No. 96 Civ. 4849 (E.D.N.Y. 2000).

[12] See, Philippines v. Marcos, Civil Case Nos. 0141 and 0185, slip op. at 17 (Republic of Philippines, Sandiganbayan July 27, 1999) (Consolidated Resolution, Garchitorena, PJ).  That court notes that one lawyer, Robert Swift, who also participated as counsel in making the limited fund claims in the Doe case, was claiming $34,585,000.00 of the $40 million sought by the lawyers in the Marcos settlement. 

[13] Id. at 20. 

[14] Id. at 15-16

[15] Id

[16] In adjudicating claim preclusion questions in claims brought by individual class members following even unsuccessful class actions, several circuits have found that a vicarious liability or principal/agent relationship provides enough privity to preclude their later claims.  See, e.g., Pelletier v. Zweifel, 921 F.2d 1465, 1502 (11th Cir. 1991), cert. denied, 502 U.S. 855 (1991); Lubrizol Corp. v. Exxon Corp., 871 F.2d 1279, 1288 (5th Cir. 1988); Cahill v. Arthur Andersen & Co., 822 F.2d 14 (2d Cir. 1987); Lambert v. Conrad, 536 F.2d 1183 (7th Cir. 1976). While a subsequent suit by a survivor absent class member against a lower-level perpetrator might not be precluded from seeking relief by a successful resolution of Doe, it might—a successful case backed up by the limited fund theory having a potentially more powerful preclusive effect.

[17] The proceedings in the Holocaust Victims Asset litigation are multiple and on-going.  Robert Swift is also involved in them. 

[18] See, e.g., Exhibit H. Decl. of Haris Silajdzic, Co-Chairman of the Council of Ministers of Bosnia-Herzegovina, Doe v. Karadzic, No. 93 Civ. 878 (S.D.N.Y. 1997); Exhibit E, Decl. of Mediha Filipovic, Parliament Member of Bosnia-Herzegovina, id.

[19] Doe I v. Karadzic, No. 93 Civ. 878 (PKL), 2000 WL 763851, at *8 n. 8 (S.D.N.Y. Mar. 27, 2000) (Memorandum Order).