The appalling announcement of President Rodrigo Duterte of the Philippines that he would like “to slaughter” three million drug addicts in his country, much as “Hitler massacred three million Jews,” (never mind that it was in fact six million), requires an immediate response from the International Criminal Court (ICC). The time has come for the ICC to open a preliminary examination—the first step toward a full investigation—into allegations of the extrajudicial killings of as many as 3,000 suspected drug-dealers and users since Duterte became President last June.

The ICC has jurisdiction because the Philippines joined the Court in 2011. The killings could constitute crimes against humanity under the Rome Statute as murders “committed as part of a widespread or systematic attack directed against any civilian population.” Even before Duterte’s declaration, Senator Leila de Lima of the Philippines had warned that an investigation could be launched by the ICC:

There are some experts who are saying that…if this spate of killings go unabated and unchecked, it could reach that point that the ICC could send a prosecutor to our country and investigate all these for possible prosecution under the Rome Statute. That is not a joke. The last thing we need right now is for our dear President to be subjected to an investigation by an international tribunal like the International Criminal Court.

But the ICC generally moves slowly and cautiously, and here it may have hesitated to open an inquiry before now for two reasons.

First, the Rome Statute requires the prosecution to prove that the crimes against humanity were committed “pursuant to or in furtherance of a State or organizational policy to commit such an attack.” While the high number of killings in the Philippines already provided a strong indication of an underlying policy, Duterte’s statement erases any doubt. His words are compelling evidence that the killings to date have been sanctioned at the highest levels, as well as an invitation to government actors to continue targeting suspected drug dealers and users. Given that one of the goals of the ICC, as stated in the Rome Statute preamble, is “to contribute to the prevention of … crimes,” the ICC should act urgently and announce the launch of an inquiry.

Second, the Rome Statute only grants the ICC last-resort jurisdiction, meaning that the Court cannot intervene as long as national institutions are conducting their own criminal investigations in good faith. In this case, the ICC might have been inclined to hold back on opening a preliminary examination to see if the Philippines would conduct its own investigations into the killings. Again, however, Duterte’s declaration now indicates otherwise: if any investigations do occur in the current climate, they will surely not be conducted independently and honestly. There is no longer any reason for the ICC to wait.

Even after Duterte’s declaration, there may be lingering questions at the ICC about whether the targeting of suspected drug dealers and users constitutes “an attack on a civilian population.” In the Limaj case at the International Criminal Tribunal for the former Yugoslavia (full disclosure: I prosecuted the case as the Senior Trial Attorney), the Trial Chamber rejected the prosecution’s argument that an attack on Serbian and Kosovar Albanian civilians suspected of collaborating with Serbs was a crime against humanity in part because of the reason the civilians were targeted:

The means and methods used by the KLA in the period relevant to the indictment, in the abduction of Serbian and Kosovo Albanian civilians (whether considered together or separately) do not evince characteristics of an attack directed against a civilian population. At least in most cases of which there is evidence, the individuals who were abducted and then detained were singled out as individuals because of their suspected or known connection with, or acts of collaboration with, Serbian authorities — and not because they were members of a general population against which an attack was directed by the KLA.

In Bowoto v. Chevron Corp., the U.S. federal district court in the Northern District of California relied on Limaj in part to exclude crimes against humanity claims, reasoning that the persons subjected to “brutal attacks” in Nigeria were targeted because they were “oil protesters,” not because they were “civilians.” The notion that crimes against humanity must be aimed against civilians as such has even crept into ICC jurisprudence.  In its decision authorizing an investigation in Kenya, the Pre-Trial Chamber found that:

[T]he potential civilian victims of a crime under article 7 of the Statute are groups distinguished by nationality, ethnicity or other distinguishing features. The Prosecutor will need to demonstrate … that the attack was directed against the civilian population as a whole and not merely against randomly selected individuals.

This approach is a wrong interpretation of the law and could undermine crimes against humanity cases generally. A requirement that perpetrators intend to attack their victims because they are civilians has never been a feature of crimes against humanity, should not become one, and certainly should not be a basis to conclude that the extrajudicial killing of thousands of suspected drug dealers and users by state forces could not qualify as crimes against humanity.  To prove crimes against humanity, the prosecution must prove only that the underlying offenses at issue were committed with the knowledge that they were part of a widespread or systematic attack against a civilian population, or with the intent that they be part of such an attack. A specific intent, or motivation, to attack a group for a particular reason, such as nationality or ethnicity, is a feature of the crimes of genocide and persecutions, but not of crimes against humanity generally.  Importing this requirement into crimes against humanity would risk undermining nearly all such prosecutions:  perpetrators often attack their victims, even when they are members of an identifiable group, for some purported “reason,” however specious it may be, e.g. that the victim group poses a “threat” to the perpetrator group.  Moreover, as Professor Darryl Robinson has explained, such a requirement would also exclude cases that we would certainly want and expect to be included within crimes against humanity, such as if a state or organization sought to instill terror within a population by randomly killing civilians.

Part of the motivation for including an additional requirement that the victims be attacked as civilians appears to be to distinguish crimes against humanity from what we might describe as large-scale domestic crimes, or cases with a relatively small number of victims. As Robinson points out, however, much of these concerns are dealt with by the requirement that the attack on the civilian population be widespread or systematic, and by the additional requirement at the ICC that the crimes be committed “pursuant to or in furtherance of a State or organizational policy to commit such an attack.”

However, it is true that even with those limitations, one could argue that crimes that we might ordinarily consider to be domestic fit within the current definition of crimes against humanity. For example, if an organized crime syndicate had a policy of resorting to violence and systematically killed a significant number of people, the elements of crimes against humanity might be met. For some, such a result presents no difficulty, while for others it does. The fear of an overbroad definition of crimes against humanity motivated the late Judge Kaul at the ICC to dissent from the above-cited Kenya investigation authorization decision and argue for a more stringent and narrowing definition of the state or organizational policy requirement:

There is, in my view, a demarcation line between crimes against humanity pursuant to article 7 of the Statute and crimes under national law. There is, for example, such a demarcation line between murder as a crime against humanity … and murder under the national law of the Republic of Kenya. It is my considered view that the existing demarcation line between those crimes must not be marginalized or downgraded, even in an incremental way. I also opine that the distinction between these crimes must not be blurred. (Dissent, para. 9).

The error in Kaul’s thinking, however, is in imagining that it is possible to draw a bright line between crimes against humanity and domestic crimes (even assuming we could get agreement on these two categories). The definition of crimes against humanity will either be over-inclusive, capturing crimes that we might not expect to be called crimes against humanity, or under-inclusive, excluding those that we would want to be within the definition. A slightly over-inclusive definition is preferable given the record of under-enforcement of laws to counter atrocity. Moreover, the issue here is solely the classification of crime rather than the determination of criminality: regardless of what definition of crimes against humanity we settle on, all crimes against humanity will also be domestic crimes. And in practice, because of prosecutorial discretion in the selection of cases, Kaul’s fear is largely a phantom: nearly all of the crimes against humanity cases prosecuted by the international tribunals have been cases that we would all agree should be prosecuted as such.

In any case, the alleged extrajudicial killing of thousands of suspected drug dealers and users by state forces in the Philippines should pose none of these anxieties. By their scale and nature, these alleged crimes certainly rise to the level of crimes against humanity. And opening a preliminary examination now does not commit the ICC to any ultimate course of action: it could subsequently decide or not decide to begin a full investigation. Starting an inquiry now, however, would unquestionably send a strong signal to Duterte that the day may come when he will have to answer for any crimes he has committed, ordered, tolerated or encouraged in the Philippines.