Ashley Deeks’ Problematic Defense of the “Unwilling or Unable” Test
Ashley Deeks, a fellow at Columbia and a former member of the Office of the Legal Adviser, has posted an essay on SSRN — forthcoming in the Virginia Journal of International Law — entitled “Unwilling or Unable: Toward an Normative Framework for Extra-Territorial Self-Defense.” Here is the abstract:
Non-state actors, including terrorist groups, regularly launch attacks against states, often from external bases. When a victim state seeks to respond with force to those attacks, it must decide whether to use force on the territory of another state with which it may not be in conflict. International law traditionally requires the victim state to assess whether the territorial state is “unwilling or unable” to suppress the threat itself. Only if the territorial state is unwilling or unable to do so may the victim state lawfully use force. Yet there has been virtually no discussion, either by states or scholars, of what that test requires. The test‟s lack of content undercuts its legitimacy and suggests that it is not currently imposing effective limits on the use of force by states at a time when trans-national armed violence is pervasive.
This Article provides the first sustained descriptive and normative analysis of the test. Descriptively, it explains how the “unwilling or unable” test arises in international law as part of a state‟s inquiry into whether it is necessary to use force in response to an armed attack. It identifies the test‟s deep roots in neutrality law, while simultaneously illustrating the lack of guidance about what inquiries a victim state must undertake when assessing whether another state is “unwilling or unable” to address a particular threat. Normatively, the Article plumbs two centuries of state practice to propose a core set of substantive and procedural factors that should inform the “unwilling or unable” inquiry. It then applies those factors to a real-world example – Colombia‟s use of force in Ecuador in 2008 against the Revolutionary Armed Forces of Colombia – to explore how the use of these factors would affect the involved states‟ decision-making and the evaluation by other states of the action‟s legality. The Article argues that the use of these factors would improve the quality of state decision-making surrounding the use of force in important substantive and procedural ways.
The essay is a very interesting read, and Deeks should be commended for trying to think systematically about what the “unwilling or unable” test would require in practice. There is, however, a fundamental problem with the essay: it completely fails to establish its thesis that “[i]nternational law traditionally requires the victim state to assess whether the territorial state is ‘unwilling or unable’ to suppress the threat itself.” The current state of the legal regime governing extraterritorial attacks against non-state actors is one of the most difficult and controversial areas of international law, requiring a careful analysis of state practice and opinio juris. Unfortunately, such an analysis is absent from Deeks’ essay. Instead, Deeks relies on a mistaken understanding of neutrality law, provides little more than a few isolated examples of extraterritorial attacks that have ostensibly been justified under the “unwilling or unable” rubric, and ignores all of the contrary examples. That is a methodologically unsound approach, and it significantly weakens what is otherwise a very good essay.
Deeks begins her discussion of the supposed “historical lineage” of the “unwilling or unable” test by turning to the law of neutrality, arguing (p. 19) that “neutrality law permits a belligerent to use force on a neutral state’s territory if the neutral state is unable or unwilling to prevent violations of its neutrality by another belligerent.” The law of neutrality, however, applies only in international armed conflicts between two legitimate belligerents; it says nothing about the use of extraterritorial force against NSAs — as Deeks herself recognizes (p. 16):
Although neutrality law does not directly govern uses of force between states and non-state actors, this section will show that the equities and concerns of the neutral state and an offended belligerent in the neutrality law context are analogous to those of the territorial state and the state seeking to use force in self-defense against a non-state actor on that territory.
This is a very significant admission. The law of neutrality may provide normative support for the “unwilling or unable” test in the context of attacks against NSAs, but it does not provide legal support for it.
Again, Deeks recognizes this — so she then cites (p. 22) UK and U.S. laws that effectively applied neutrality law to situations of internal armed conflict, arguing that the existence of such laws “explain[s] how neutrality rules developed to govern acts by states during international armed conflict expanded beyond that context to govern acts by non-state actors during peacetime (and in non-international armed conflicts).” But that is simply mistaken. It is true that the UK and the U.S. enacted such laws (state practice), but they never claimed that they did so out of a sense of legal obligation — the necessary condition of such laws counting as opinio juris in favor of the “unwilling or unable” test. On the contrary, both states recognized that the law of neutrality in no way obligated them to prohibit their nationals from (to quote Lauterpacht) “committing such acts as amount to making the national territory a base for military or naval operations against a friendly state.” As the U.S. Attorney General said in 1895 (emphasis added):
While called neutrality laws, because their main purpose is to carry out the obligations imposed upon the United States while occupying a position of neutrality toward belligerents, our laws were intended also to prevent offenses against friendly powers whether such powers should or should not be engaged in war or in attempting to suppress revolt.
To quote Tucker, we must always distinguish between “the operation of the law of neutrality as determined by international law and the operation of municipal neutrality laws. The latter may be applied to situations other than war in the sense of international law.” That distinction, unfortunately, is lost on Deeks. (For a longer discussion of the distinction, see my response to Karl Chang’s essay on neutrality here.)
There is, of course, another problem with using the law of neutrality to support the “unwilling or unable” test: that law predates the adoption of the UN Charter, which strictly regulates the use of interstate force. Deeks recognizes the problem, but barely addresses it — simply claiming (n. 33) that “[t]he better view is that neutrality law remains relevant and applicable, at least to international armed conflicts,” and that “[e]ven if neutrality law were defunct… the existence of the ‘unwilling or unable’ test in that law provides historical depth to today‘s rule.” Historical depth, maybe. But legal support? Definitely not — especially as Deeks admits that the current relevance of the law of neutrailty is limited to international armed conflicts.
The real question, then, is what the customary rule governing extraterritorial force against NSA might be in the post-Charter world. Deeks discusses three different positions on the relationship between Article 2(4) of the UN Charter’s prohibition on interstate force and Article 51′s exception to that prohibition for acts of self-defense in response to an armed attack: (1) that the armed attack giving rise to the right of self-defense must involve a state; (2) that the armed attack can involve a non-state actor (NSA), but the actions of the NSA must be attributable to a state; and (3) that the armed attack can involve a NSA and does not require any kind of attribution to a state. Deeks says that a “premise” of her article, supposedly based on “extensive state practice,” is that the third position is correct. I’ll return to that supposed state practice below, but it’s worth noting here that Deeks attributes the three positions on self-defense to “groups of scholars” — and then relegates to a footnote (n. 17) the rather important fact that the second position is the one that has been specifically endorsed, in multiple cases, by the International Court of Justice. ICJ decisions are not themselves primary sources of international law, but the failure to discuss those decisions is a serious problem with Deeks’ essay — especially as the essay does not even mention the Nicaragua case, in which the ICJ held the most clearly that state attribution is required.
So what is this “extensive” state practice that Deeks cites as evidence that the “unwilling or unable” test reflects customary international law? Actually, it’s not extensive at all. In the essay’s introduction, she mentions (pp. 4-5) Russia’s attacks on Chechen rebels in Georgia; Israel’s attacks on Hezbollah and the PLO in Lebanon; and Turkey’s attacks on the PKK in Iraq. Later on, she mentions the Soviet Union’s 1921 attack on White Guard bands in Outer Mongolia; U.S. attacks on Viet Cong soldiers in Cambodia during the Vietnam War; U.S. attacks on al-Qaeda in Afghanistan and the Sudan; and Colombia’s attacks on FARC in Ecuador. That’s it. Pretty weak tea indeed — especially when we factor in the international response to many of those uses of force against “unwilling or unable” states, such as the Organization of American States’ unequivocal condemnation of Colombia’s attacks on FARC as a violation of Ecuador’s sovereignty.
More importantly, Deeks simply ignores the numerous instances in which the Security Council and/or states have condemned extraterritorial uses of force against NSAs whose actions were not attributable to the state whose territory was attacked. Examples include Israel’s 1985 raid of a PLO headquarters in Tunis; Iran’s cross-border attacks throughout the 1980s on Kurdish fighters in Iraq (which were vociferously condemned by the U.S.); and Rwanda’s attacks in the late 1990s on Hutu rebels in the DRC.
To be sure, it appears that customary international law is slowly evolving away from the Nicaragua standard, especially in the wake of 9/11. But it is far from clear whether that standard has been replaced by the “unwilling or unable” test. Neither Christian Tams nor Tom Ruys, the two scholars who have examined state practice and opinio juris most closely, are willing to go that far. Tams concludes that state attribution is still required, but can be satisfied by something less than Nicaragua‘s “effective control” of the NSA. And Ruys concludes that “[d]e lege lata, the only thing that can be said about proportionate trans-border measures of self-defence against attacks by non-State actors in cases falling below the Nicaragua threshold is that they are ‘not unambiguously illegal’.”
Indeed, what is most surprising about Deeks’ essay is that Deeks herself admits that the “unwilling or unable” test cannot be considered customary international law. If you look at footnote 55 of her essay, you find this remarkable statement:
I have found no cases in which states clearly assert that they follow the test out of a sense of legal obligation (i.e., the opinio juris aspect of custom), nor have I located cases in which states have rejected the test. Even if one concludes that the rule does not rise to the level of custom, however, the rule makes frequent appearances in state practice and therefore is the appropriate starting point from which to determine how the norm should develop.
That is a remarkable admission — and one that directly contradicts Deeks’ thesis that “[i]nternational law traditionally requires the victim state to assess whether the territorial state is ‘unwilling or unable’ to suppress the threat itself.” If there is no opinio juris that supports the “unwilling or unable” test, it is difficult to argue that the test reflects customary international law — especially in light of the consistent and contrary pre-9/11 state practice and opinio juris that both Tams and Ruys discuss.
The bottom line: de lege ferenda, there is much to recommend Deeks’ essay. De lege lata, however, it completely fails to make its case.