This piece is the latest in our online symposium–spearheaded by Professor Laura Dickinson–focusing on the International Criminal Court’s (ICC) probe in Afghanistan and its implications for the United States.
As previous contributions to this Symposium have explained, the Pre-Trial Chamber of the International Criminal Court (ICC) will soon decide whether or not to authorize an investigation into war crimes and crimes against humanity allegedly committed in Afghanistan—including by U.S. military personnel and CIA officials—on or after May 1, 2003.
So, could U.S. personnel who “tortured some folks” in Afghanistan find themselves prosecuted before the International Criminal Court? What about Russian agents operating in Ukraine, Israeli soldiers operating in Palestine, or, perhaps, Myanmar security forces operating across the border with Bangladesh? Afghanistan, Ukraine, Palestine, and Bangladesh have either joined the ICC or accepted its jurisdiction. The United States, Russia, Israel, and Myanmar have not.
Ultimately, these questions will be answered in The Hague. But the answers may be preordained, or at least constrained, by a seemingly unrelated legal process currently underway in New York and Geneva. International law is fast approaching a crossroad. Let me explain.
Under customary international law, state agents enjoy functional immunity—immunity ratione materiae—from the exercise of national criminal jurisdiction by other states with respect to acts performed in an official capacity. According to many states, such functional immunity applies in respect of both domestic and international crimes—including war crimes, genocide, and crimes against humanity. Of course, an agent’s state may waive this immunity, but it is currently under no legal obligation to do so. Importantly, an agent’s state may assert this immunity without making any commitment to prosecute its agent for these crimes in its own courts. If that view seems grotesque, then you understand it.
Now, as David Stewart notes in his contribution, the United States may argue that its agents enjoy functional immunity from the exercise of the ICC’s jurisdiction in respect of international crimes committed in the territory of its member states. On this view, even if the ICC has jurisdiction in respect of international crimes committed in Afghanistan, it may not exercise that jurisdiction in respect of international crimes committed by U.S. agents. Importantly, functional immunity would bar ICC prosecution of U.S. agents independently of any special immunities arising from bilateral agreements between the U.S. and Afghanistan (expertly discussed by Beth Van Schaack here and here).
Indeed, the United States once seemed to take this position. As Ambassador David Scheffer put it before the United States Senate Foreign Relations Committee in 1998:
Our position is clear: Official actions of a non-party state should not be subject to the court’s jurisdiction if that country does not join the treaty, except by means of Security Council action under the U.N. Charter.
On this view, only the Security Council can override the functional immunity of non-party state agents with respect to their official actions, including international crimes committed in the territory of member states. Similarly, at the Rome Conference, the United States offered a proposal “to exempt from the ICC’s jurisdiction conduct that, in the absence of a Security Council referral, arises from the official actions of a non-party state acknowledged as such by that non-party.” The U.S. proposal was rejected, a point to which we will return.
So, the United States might argue:
(a) U.S. agents enjoy functional immunity from the exercise of jurisdiction by Afghanistan with respect to their official actions, including international crimes committed in Afghanistan,
therefore,
(b) U.S. agents enjoy functional immunity from the exercise of jurisdiction by the ICC with respect to such actions.
No doubt, some will argue that this argument is invalid, that (a) neither entails nor decisively supports (b). My own view is that this argument is unsound, because (a) is false. Still, I should say something about the validity of this argument, lest readers dismiss it (and this post) out of hand.
The argument begins with the idea that states may not evade the functional immunity of foreign state agents by delegating the exercise of their jurisdiction to the ICC. The argument proceeds in one of two ways. On one version, if Afghanistan delegates the exercise of its jurisdiction to the ICC, then the functional immunity of U.S. agents accompanies the delegation and directly binds the ICC. Accordingly, if the ICC exercises jurisdiction over U.S. agents, then the ICC directly violates their functional immunity. On a different version of the argument, if Afghanistan delegates the exercise of its jurisdiction to the ICC, and the ICC exercises jurisdiction over U.S. agents, then Afghanistan indirectly violates their functional immunity. Finally, the argument posits that the ICC should avoid interpretations of its Statute that would make it or its member states responsible for violations of international law.
Admittedly, article 27(2) of the Court’s Statute provides that:
Immunities or special procedural rules which may attach to the official capacity of a person, whether under national or international law, shall not bar the Court from exercising its jurisdiction over such a person.
Since the ICC has jurisdiction over persons who commit international crimes in the territory of its member states (arts 4 & 12), the ordinary meaning of article 27(2) is that immunities which may attach to the official capacity of such persons under international law shall not bar the Court from exercising such jurisdiction. However, on the argument we are considering, the Court should ‘read down’ article 27(2) to apply only to agents of member states, lest it or its members states incur responsibility for violating the functional immunity of non-member state agents.
For a more detailed treatment of the path from (a) to (b), I highly recommend this article by Roger O’Keefe. Before discussing (a) on its own terms, I should note that the inverse argument seems valid. In other words, if (a) is false then it is hard to see how (b) could be true. If U.S. agents do not enjoy functional immunity from Afghanistan’s exercise of its own jurisdiction in respect of international crimes committed on its territory, then it is hard to see how they could enjoy functional immunity from the ICC’s exercise of jurisdiction delegated to it by Afghanistan in respect of the very same crimes. All the more reason to inquire into the truth of (a).
Enter the International Law Commission
For the last decade, the expert members of the International Law Commission have studied the immunity of state officials from foreign criminal jurisdiction. Last year, the Commission provisionally adopted a Draft Article stating that functional immunity from foreign criminal jurisdiction shall not apply in respect of international crimes, including those crimes listed in the ICC Statute. Since the Draft Article rejects premise (a) of the argument sketched above, its reception by states, as well as its basis in legal principle, may well determine whether U.S. agents enjoy functional immunity before the ICC.
As I discuss in an earlier post, the Commission provisionally adopted the Draft Article by recorded vote, called by those opposed to it, rather than by consensus. 21 members voted in favor, 8 voted against, and 1 abstained.
Importantly, in the General Assembly, 24 states expressed support for Draft Article 7, while 18 expressed disagreement. The supportive states were Austria, Chile, Czech Republic, Denmark, El Salvador, Estonia, Finland, Greece, Hungary, Iceland, Italy, Mexico, Netherlands, New Zealand, Norway, Poland, Peru, Portugal, Romania, Slovakia, Slovenia, South Africa, Sweden, and Viet Nam. The opposed states were Australia, China, France, Germany, India, Iran, Ireland, Israel, Japan, Malaysia, Russia, Singapore, Spain, Sri Lanka, Switzerland, Thailand, the United Kingdom, and the United States.
For the most part, the states that supported the Draft Article argued deductively, claiming that functional immunity in respect of international crimes is unsupported by or incompatible with sound legal principles. In contrast, the states that opposed the Draft Article primarily argued inductively, claiming that an exception or limitation to functional immunity in respect of international crimes is unsupported by sufficient state practice.*
Interestingly, the statement by the United States was more moderate than one might have expected, and ambiguous in a number of ways. To be sure, the U.S. statement says that Draft Article 7 “cannot be said to represent customary international law or even the progressive development of existing law.” However, the statement ends by suggesting that the Commission should conclude that “we don’t know – the law is unsettled, state practice is sparse and uneven, and the issue is not capable of being properly resolved at this time.”
Importantly, the U.S. statement does not say that functional immunity applies to all acts performed in an official capacity:
This is not to say that all states have adopted an absolutist position regarding ratione materiae immunity; to the contrary, as noted above, there have indeed been prosecutions of foreign officials in some circumstances. Nor is it to say that there should not be any exceptions, even if immunity would ordinarily attach. However, in our view, the inconsistent nature of state practice means that premature attempts at codification can do more harm than good in this area.
Indeed, the statement notes that “there have, in fact, been prosecutions of foreign officials, including by the United States, for a range of conduct including corruption, violent crimes, and cyber crimes.”
In an interesting passage, the U.S. statement notes that:
there are basic methodological disagreements about how to identify customary international law, if any, in this area. In evaluating state practice, does one begin with a baseline of immunity, and then look for examples of exceptions? Or does one begin with a baseline of no immunity, and then look for examples of immunity?
Indeed, these are important methodological questions. In my view, the inherent right of states to exercise jurisdiction over persons and conduct in their own territory is the general rule, while the immunity of foreign state agents is the limited exception. Accordingly, the application of functional immunity to international crimes committed in the territory of the forum state must be demonstrated and should not be presumed.
Unfortunately, the U.S. statement seems to vacillate from one methodology to the other. When criticizing the ILC’s proposal, the U.S. seems to begin with a baseline of immunity, and then look for examples of exceptions, concluding that:
there is insufficient state practice to illustrate a “clear trend,” let alone the widespread and consistent state practice taken out of a sense of legal obligation required to create, or to demonstrate the existence of, sufficiently specific rules of customary international law to support the ILC’s proposal.
In contrast, when describing its own “prosecutions of foreign officials … for a range of conduct,” the U.S. seems to begin with a baseline of no immunity, and then look for (or perhaps ignore) examples of immunity. Certainly, the U.S. does not cite widespread and consistent state practice consistent with its own prosecutions, or even formulate the sufficiently specific rule of customary international law that supposedly permits them. Instead, the U.S. claims only that “there have indeed been prosecutions of foreign officials in some circumstances.” Apparently, that is enough to establish an exception or limitation to functional immunity under customary international law. Except, of course, when it isn’t.
Most interestingly of all, the U.S. statement says that:
We are also concerned by the cursory explanation in the Commentary about why Draft Article 7 does not include an exception for crimes by foreign officials in the territory of the forum state. This fundamental issue of territorial conduct and its effect on criminal jurisdiction warrants much more serious attention and analysis.
This passage certainly suggests that the United States believes that there is an exception for crimes by foreign officials in the territory of the forum state. In my view, there are principled grounds for such an exception. States may be sovereign equals, but their sovereign rights are not equal. If one state seeks to exercise jurisdiction over persons and conduct in its own territory while another state seeks immunity for its agents acting in the territory of another sovereign state, then the sovereign rights of the former should prevail. Yet if such a ‘territorial conduct’ exception exists, and applies in respect of “corruption, violent crimes, and cyber crimes,” then logically it must apply in respect of international crimes as well, most of which are themselves violent crimes. For example, a ‘territorial conduct’ exception to functional immunity that extends to corruption but not to genocide seems bizarre.
If the United States does, in fact, accept such a ‘territorial conduct’ exception to functional immunity from the exercise of national criminal jurisdiction, then it rejects the premise (a) of its only obvious argument that (b) its agents enjoy functional immunity from the exercise of the ICC’s jurisdiction in respect of international crimes committed in the territory of ICC member states. Put the other way around, if functional immunity does not bar states from exercising territorial jurisdiction over foreign state agents in respect of international crimes, then it is hard to see why functional immunity would bar the ICC from exercising the territorial jurisdiction of its member states on their behalf.
In these ways, the U.S. statement in the General Assembly seems to undermine the most plausible functional immunity-based argument it might make before the ICC. Conversely, if the U.S. asserts functional immunity from ICC jurisdiction for acts of torture committed by its personnel in the territory of Afghanistan, then it may implicitly concede the functional immunity of Russian, Chinese, or North Korean agents who commit crimes in the United States. Forced to choose, I would naturally encourage the United States to endorse the territorial conduct exception and renounce functional immunity from ICC jurisdiction.
I should note that, on one view, a ‘territorial conduct’ exception would not apply if the territorial state consents both to the presence of the foreign state agent and to the activity that leads to the crime. If that is the U.S. position, then I suppose it could argue that Afghanistan consented both to the presence of its personnel and to the detention activities that led to torture. Even if these facts could be established, this form of argument would not speak to the ICC’s exercise of jurisdiction over agents of non-member states who enter and commit international crimes within member states without their consent—as credibly alleged, for example, of Russian agents operating in Ukraine, which has accepted the ICC’s jurisdiction.
Conclusion
If the United States asserts that its agents enjoy functional immunity from the ICC’s exercise of jurisdiction, then it will have to explain whether foreign state agents enjoy functional immunity in respect of violent crimes—including international crimes—committed in the United States. If the U.S. answer is ‘yes’, then that would be news to our adversaries. If the answer is ‘no’, then that seems hard to reconcile with the assertion of immunity before the ICC.
In my view, the United States should accept that functional immunity does not apply in respect of international crimes committed in the territory of the state exercising jurisdiction or delegating the exercise of its jurisdiction to the ICC. The United States should not compromise its own inherent right to prosecute violent crimes committed in its own territory, or the ICC’s delegated right to prosecute international crimes committed in the territory of its member states.
One last observation. Earlier, I noted that some will argue that the inference from (a) to (b)—from functional immunity from the exercise of national jurisdiction to functional immunity from the exercise of ICC jurisdiction—is invalid. In a way, I hope this inference is valid because, in an indirect way, this would help show that it is unsound. Let me explain.
123 states have ratified the ICC Statute, 31 have signed but not ratified, and several have accepted the ICC’s jurisdiction with respect to particular situations. They have done so knowing full well that article 27(2), by its terms, provides that immunities under international law shall not bar the Court from exercising its jurisdiction. And they have done so in full knowledge of U.S. objections, including its failed proposal to exclude official actions of non-party states from the exercise of the ICC’s jurisdiction.
In my view, the ICC should presume that these 123-154 states accept the ordinary meaning of article 27(2) and do not think that either they or the ICC would violate international law if the ICC exercises its jurisdiction as its Statute provides. In other words, the ICC should presume that these states reject (b).
Importantly, if the inference from (a) to (b) is valid, then so is the contrapositive inference from the negation of (b) to the negation of (a):
-(b) state agents do not enjoy functional immunity from the exercise of jurisdiction by the ICC in respect of international crimes committed in the territory of a member state,
therefore,
-(a) state agents do not enjoy functional immunity from the exercise of jurisdiction by other states in respect of international crimes committed in their territory.
So, if we should presume that ICC member states reject (b), then we should presume that they also reject (a).
In my view, when states endorse a legal rule or principle, we should presume that they would also endorse its logical consequences. Of course, states are free to prove us wrong and adopt logically inconsistent positions. But until they do so, we should presume that they will not. Accordingly, we should presume that ICC member states that have not taken a public position on (a) would reject it, that is, they would deny that functional immunity applies in respect of international crimes committed in the territory of the state exercising its jurisdiction or delegating the exercise of its jurisdiction to the ICC. In this way, we might fill gaps in state practice and opinion with logical deductions from the very act of ratifying the ICC Statute itself.
What about those member states that publicly endorse (a)—including France, Germany, and the United Kingdom? Ideally, they would revise their view and reject functional immunity in respect of international crimes committed in the forum state. Alternatively, they might deny that (a) entails (b), potentially in direct opposition to the United States. Finally, they might embrace (b), declaring years after ratifying the ICC Statute that it does not mean what it says, or that they did not know what they were doing when they ratified it. In my view, the ICC should not do such distasteful work for them. The ICC should presume that its member states will resolve contradictions in their legal positions in the most reasonable way available to them. If member states wish to prove such charity improvidently granted, they are free to do so. In the end, international law is what states make it. More’s the pity.
*Astute readers will notice that France, the United Kingdom, and the United States opposed the ILC’s Draft Article. Supposedly, these states support accountability for international crimes, but they simply cannot find sufficient evidence that customary international law permits the national prosecution of foreign state agents who commit such crimes. Of course, last week, these three states used armed force against Syrian government targets, based on legal theories that do not reflect general state practice accepted as law. One might wonder what suddenly happened to their rigorous, inductive methodology. But, then, there is no more consistent and virtually uniform state practice than hypocrisy.