The Second Circuit’s Incorrect Interpretation of Customary International Law (UPDATED)

The Second Circuit’s Incorrect Interpretation of Customary International Law (UPDATED)

I was shocked — to say the least — when I read Roger’s post on the extremely important Khulumani case. With due respect to the Second Circuit, Judge Katzman’s analysis of the mens rea of aiding and abetting under customary international law is completely incorrect. Although I can see how a judge who is not familiar with international criminal law could be misled by the Rome Statute’s idiosyncratic language, there is absolutely no question that the customary mens rea is knowledge, not purpose or intent.

The best place to start is Judge Katzman’s statement that “those who assist in the commission of a crime with the purpose of facilitating that crime would be subject to aiding and abetting liability under the statutes governing the ICTY and ICTR.” That is correct, but incomplete — the ICTY has consistently held that those who assist in the commission of a crime with the knowledge that the principal intended to commit a crime are equally liable:

Aleksovski (Appeals Chamber), para. 162: “[I]t is not necessary to show that the aider and abettor shared the mens rea of the principal, but it must be shown that the aider and abettor was aware of the relevant mens rea on the part of the principal. It is clear that what must be shown is that the aider and abettor was aware of the essential elements of the crime which was ultimately committed by the principal.”

Furundzija (Trial Chamber), para. 245: “[I]t is not necessary for the accomplice to share the mens rea of the perpetrator, in the sense of positive intention to commit the crime.”

Vasiljevic (Trial Chamber), para. 71: “The aider and abettor must be aware of the essential elements of the crime committed by the principal offender, including the principal offender’s state of mind. However, the aider and abettor need not share the intent of the principal offender. The fact that the aider and abettor does not share the intent of the principal offender generally lessens his criminal culpability from that of an accused acting pursuant to a joint criminal enterprise who does share the intent of the principal offender.”

Outside of aiding and abetting genocide, the ICTR agrees:

Bagilishema (Trial Chamber), para. 32: “An accomplice must knowingly provide assistance to the perpetrator of the crime, that is, he or she must know that it will contribute to the criminal act of the principal.”

Judge Katzman’s reliance on the Ministries case, decided by a Nuremberg Military Tribunal (NMT), is also misleading. It is true that Ministries “declined to impose criminal liability on a bank officer who was alleged to have ‘made a loan, knowing or having good reason to believe that the borrower w[ould] use the funds in financing enterprises [conducted] in violation of either national or international law’, but was not proven to have made the loan with the purpose of facilitating the enterprises’ illegal activities.” But that is literally the only time that a Nuremberg Military Tribunal ever held that knowledge was not enough for aiding and abetting — a holding that reflects the Tribunals’ general reluctance to hold bankers liable for their involvement in the Nazis’ crimes. (The IMT’s indefensible acquittal of Hjalmar Schacht, the President of the Reichsbank, is another example.) In every other case, the Tribunals held that knowingly assisting the commission of a crime was sufficient.

Consider, for example, the conviction of Waldemar Klingelhoefer, an SS Major who was assigned as an interpreter to Einsatzgruppe B, which was responsible for killing more than 45,000 persons. Although the Einsatzgruppen Tribunal concluded that Klingelhoefer directly participated in executions, it made clear that it would have convicted him even if his assistance to Einsatzgruppe B had been limited to translation, because he knew that his interpreting services would facilitate the executions:

Klingelhoefer has stated that his function in the Einsatzgruppe operation was only that of interpreter. Even if this were true it would not exonerate him from guilt because in locating, evaluating and turning over lists of Communist party functionaries to the executive department of his organization he was aware that the people listed would be executed when found. In this function, therefore, he served as an accessory to the crime.

The Flick case — most notable as perhaps the most lenient of all the NMT decisions — reached a similar conclusion regarding Bernhard Weiss, a high-ranking corporate officer of the Flick Concern who had procured an increase in the number of freight cars available to the company and an increase in the number of Russian POWs manufacturing the freight cars. Although the Tribunal found no evidence that Weiss issued his request for the purpose of facilitating the use of slave labor, it concluded that he was aware that the use of slave labor would result from his request. It thus convicted him of aiding and abetting the Nazis’ slave-labor program.

These examples could be multiplied indefinitely. Moreover, a number of other World War II cases held that knowledge suffices for aiding and abetting, such as Schonfeld and Zyklon B, which were heard by British military courts.

In short, Judge Katzman’s claim that his position is “based on [a] review of international law’s treatment of aiding and abetting liability over the past sixty years” is simply incorrect. It’s also troubling, because nearly all of the cases that I have cited (and more) were mentioned in Furundzija, on which Judge Katzman relies — correctly — for the proposition that the actus reus of aiding and abetting requires “substantial assistance” to the perpetrator of the crime. Indeed, here are the paragraphs that immediately follow paragraph 235, which Judge Katzman cites in defense of the substantial-assistance requirement:

236. With regard to mens rea, the Trial Chamber must determine whether it is necessary for the accomplice to share the mens rea of the principal or whether mere knowledge that his actions assist the perpetrator in the commission of the crime is sufficient to constitute mens rea in aiding and abetting the crime. The case law indicates that the latter will suffice.

237. For example in the Einsatzgruppen case knowledge, rather than intent, was held to be the requisite mental element.

238. The same position was taken in Zyklon B where the prosecution did not attempt to prove that the accused acted with the intention of assisting the killing of the internees
. It was accepted that their purpose was to sell insecticide to the SS (for profit, that is a lawful goal pursued by lawful means). The charge as accepted by the court was that they knew what the buyer in fact intended to do with the product they were supplying.

239. Two of the not guilty verdicts in Schonfeld also provide an indication of the mens rea necessary to amount to being “concerned in the killing”. Both concerned drivers who claimed to have followed instructions without knowing the purpose of the mission, and were therefore found not guilty. Despite having made a physical contribution to the commission of the offence, they had no knowledge that they were doing so.


245. The above analysis leads the Trial Chamber to the conclusion that it is not necessary for the accomplice to share the mens rea of the perpetrator, in the sense of positive intention to commit the crime. Instead, the clear requirement in the vast majority of the cases is for the accomplice to have knowledge that his actions will assist the perpetrator in the commission of the crime. This is particularly apparent from all the cases in which persons were convicted for having driven victims and perpetrators to the site of an execution. In those cases the prosecution did not prove that the driver drove for the purpose of assisting in the killing, that is, with an intention to kill. It was the knowledge of the criminal purpose of the executioners that rendered the driver liable as an aider and abettor


249. In sum, the Trial Chamber holds the legal ingredients of aiding and abetting in international criminal law to be the following: the actus reus consists of practical assistance, encouragement, or moral support which has a substantial effect on the perpetration of the crime. The mens rea required is the knowledge that these acts assist the commission of the offence.

That leaves us with Judge Katzman’s reliance on Article 25(3)(c) of the Rome Statute, which he correctly notes “makes clear that, other than assistance rendered to the commission of a crime by a group of persons acting with a common purpose, a defendant is guilty of aiding and abetting the commission of a crime only if he does so “[f]or the purpose of facilitating the commission of such a crime.” Scholars disagree about the impact of the “for the purpose of facilitating” language: some believe that it imposes an intent requirement, while others believe that it leaves the traditional knowledge requirement intact. It doesn’t matter, however, which position is correct — as Robert Cryer has pointed out, Article 25(3)(c) neither reflects nor declares customary international law:

The article also introduces a purposive, motive requirement that is not required by custom (under which knowledge suffices). Thus the crime is not defined in accordance with customary international law, but in practice the addition of the purposive intent will render liability under the Rome Statute more narrowly than in custom.

By his own admission, Judge Katzman was trying to determine what definition of aiding and abetting was so “well-established [and] universally recognized to be considered customary international law for the purposes of the ATCA.” That definition is the one articulated in Furundzija and the other cases cited above, not the definition provided by Article 25(3)(c). The Rome Statute thus should not have played a role — at least not a determinative one — in Judge Katzman’s analysis.

The Second Circuit’s decision in Khulumani is obviously a significant win for the plaintiffs, allowing them to proceed on their aiding and abetting claims. Unfortunately, because of Judge Katzman’s erroneous analysis of customary international law, succeeding on those claims will prove to be far more difficult than customary international law demands.

UPDATE: I just came across Anthony Sebok’s helpful FindLaw column on Khulumani. As he notes, the majority’s adoption of the Rome Statute standard may well spell doom for the plaintiffs:

As Judge Korman slyly noted in his dissent, when Judge Sprizzo, on remand, applies the Rome Statute, it is very likely plaintiffs will not meet its comparatively demanding standard. Thus, the Second Circuit panel may have handed plaintiffs a paper victory, one that will not stand up on remand. Still, the remand does keep the case alive, providing a possible setting for a settlement.

None of the pleadings so far indicates that it was the purpose of Barclay’s Bank or Ford to facilitate the violation of human rights by promoting the apartheid system. Instead, the most the plaintiffs have so far alleged, is that the corporate defendants were substantially certain that their efforts to sell products to the South African government would, in fact, have the effect of causing apartheid to survive. But substantial certainty, while good enough for the Restatement test, is not good enough for the Rome Statute’s test.

The Restatement test applied by the judge in the minority — the Restatement (Second) of Torts §876(b) — provides that an individual is liable for aiding and abetting the tortious conduct of another “if he… knows that the other’s conduct constitutes a breach of duty and gives substantial assistance or encouragement to the other.” The irony is obvious: properly understood, the customary international law standard and the Restatement test are the same.

Print Friendly, PDF & Email
Notify of
Marko Milanovic
Marko Milanovic

Kevin, I think that your analysis is spot on. Regrettably, a very similar thing happened in the ICJ’s recent Bosnian Genocide judgment (note: all disclaimers apply). In para. 421 of the judgment, the Court refused to give an answer to whether complicity in genocide requires specific, genocidal intent, or merely knowledge of the specific intent of the principal (as the ICTY ruled in Krstic when it comes to aiding and abetting genocide). The Court could take this apparently agnostic approach because there was no conclusive evidence that any organs of Serbia had actual knowledge of the Bosnian Serbs’ intent to commit genocide in Srebrenica, and consequently even the lower standard could not be met. The most likely reason why the Court took this approach is because the preparatory work to the Genocide Convention states that complicity in genocide must be ‘deliberate.’ However, as I tried to explain elsewhere, that simply means that the drafters wanted to exclude any potential negligent provision of assistance which facilitates the commission of a crime, not that they wanted to adopt the same standard of mens rea for commission and complicity. If we take the facts of the Zyklon B case, for example, of a… Read more »

cruz del sur
cruz del sur

Some people are not safe anywhere