Territorial Concessions to the Aggressor
(In-)Validity of a Prospective Peace Treaty Between Russia and Ukraine?
How can Russia’s war of aggression against Ukraine be brought to an end? With recent proposals by the Trump transition team, a possible peace treaty with Ukraine and Russia is gaining renewed attention. Gregory Fox predicts that “territorial transfers […] would likely be at the heart of any agreement”. Already in 2023, a NATO official considered territorial concessions as an element of a peace treaty, giving rise to political criticism and scholarly debate (e.g. here, here, and here). The pivotal question is whether international law allows aggressor states to walk away with territorial gains after the war, based on a peace treaty.
Fox argues that such a peace treaty might be invalid due to a conflict with jus cogens, as the treaty “would also ratify the unlawful force giving rise to the annexation”. The treaty could thereby come into conflict with the peremptory prohibition of the use of force, and the prohibition of annexation. However, this argument is based on an overly broad interpretation of what constitutes a “conflict” with a peremptory norm. Under a proper interpretation of “conflict”, jus cogens will not invalidate territorial concessions.
In contrast, the rule on coerced treaties (Art. 52 VCLT) poses a major legal obstacle to territorial concessions to an aggressor state. Under that rule, a treaty is void if its conclusion has been procured by an unlawful threat or use of force. That notwithstanding, a treaty invalid under Art. 52 VCLT can be validated by the UN Security Council (UNSC), a solution that is also compatible with jus cogens.
No case of jus cogens invalidity (Art. 53 VCLT)
A peace treaty whereby Ukraine transfers sovereign rights over parts of its territory to Russia would not be invalidated by jus cogens. The treaty-invalidating effect of peremptory norms follows from Art. 53 VCLT and customary international law (ILC conclusions on jus cogens, conclusion 10). The invalidating effect is preconditioned on the existence of a “conflict” between the treaty and a peremptory norm. A “conflict” arises, according to a classical test formulated by Vranes, “between norms, one of which may be permissive, if in obeying or applying one norm, the other norm is necessarily or potentially violated.”
Under that test, there is no conflict between a treaty making territorial concessions to an aggressor and the peremptory prohibition of the use of force. Regardless of the advantage the aggressor gains through this treaty, the aggressor’s use of force remains unlawful. The treaty confers no other verdict of un/lawfulness on that conduct. With the jus contra bellum prohibiting Russia’s use of force, and a possible peace treaty leaving this prohibition unaffected, no conflict arises.
Moreover, there is no conflict between such a treaty and the prohibition of annexations. Regardless of whether one considers this prohibition to be an integral part of the peremptory prohibition of the use of force, or assumes it to be a self-standing prohibition with peremptory character (see Brunk and Hakimi), that norm does not prohibit the transfer of territory through a treaty. What it prohibits is that states acquire the territory of another state through the threat or use of force. By conferring title over Ukrainian territory to Russia through treaty, such a peace deal would provide the basis for Russia’s territorial acquisition. While that treaty-based transfer of territory in a peace deal may be related to the preceding use of force, this relationship is addressed by Art. 52 VCLT, not by the prohibition of annexation. The peace treaty that makes territorial concessions does not aim to change the unlawfulness of annexations, such that no norm conflict arises.
As a consequence, a peace treaty with territorial concessions is fully compatible with the peremptory prohibition of force (and that of annexation). A finding of jus cogens invalidity could only be sustained on the basis of a broader interpretation of “conflict”. Several proposals for such a broader interpretation have been made (e.g. by Cannizzaro, Orakhelashvili). Under these proposals, any treaty that impairs the operation, purpose, or effectiveness of a peremptory norm should be considered to conflict with the latter. The argument can then be made that a peace treaty with territorial concessions rewarding the aggressor would counteract the deterrent effect of the prohibitions of the use of force and of annexation, thereby frustrating these peremptory norms.
However, three reasons speak against the expansive interpretations of “conflict”. Firstly, the expansive interpretation was implicitly rejected by the ICJ in its Jurisdictional Immunities of the State judgment (at paras 92-93). Responding to Italy’s argument that the rules of state immunity would impede the operation of the peremptory prohibitions violated by Germany, the ICJ held that the rules of immunity “do not bear upon the question whether or not the conduct [by Germany] was lawful or unlawful.” Thereby, the Court applied and affirmed the classical test of norm conflict. Secondly, the test of whether the purpose or operation of a peremptory norm is impaired is rather vague and difficult to apply. It may not always be clear what exactly the object and purpose of a peremptory norm is (for the prohibition of force, see the debate here). This can lead to competing predictions of whether the treaty may impair that purpose. Thirdly, expansive interpretations of “conflict” would entail a confusion with the rule on coerced treaties in Art. 52 VCLT. The VCLT clearly distinguishes between treaty invalidity on account of the content of the treaty (Art. 53), and the circumstances under which the treaty was concluded (Art. 52 VCLT). If treaties concluded in connection with an unlawful use of force were to fall under Art. 53, Art. 52 VCLT would be rendered obsolete. Therefore, the classical test of norm conflict should be retained, under which territorial concessions do not fall victim to jus cogens invalidity.
Invalidity of coerced treaties (Art. 52 VCLT)
The decisive rule for determining the (in-)validity of territorial concessions to an aggressor therefore is Art. 52 VCLT. This rule invalidates treaties procured by an unlawful threat or use of force. Emphasising the unlawful use of force seems to suggest the following distinction: Concessions to the aggressor state (who uses force unlawfully) entail the invalidity of the treaty, whereas a treaty benefitting the victim state is not procured by unlawful force (the victim presumably using force lawfully in self-defence), and therefore valid.
It is not quite that simple, though. Rather, the decisive question is whether a peace treaty has been procured by unlawful force. A treaty is only procured by force if the victim state “as the result of unlawful use of force, has been reduced to such a degree of impotence as to be unable to resist the pressure to become a party to a treaty” (Schmalenbach, Art. 52, at para. 25). Similarly, Vishchyk and Pizzi ask whether “the conclusion of a treaty will primarily be the result of the use of force” (emphasis added). These criteria entail that a victim state, which largely resists the aggressor, could still make concessions in a peace treaty, where these concessions are based on a voluntary choice to use this means for establishing a peaceful solution. Similarly, a comprehensive peace treaty negotiated only after Russian troops withdraw can arguably make concessions which would not be procured by Russia’s threat or use of force (see Rasmussen, at p. 51).
Assessing under what circumstances Ukraine might be considered able to resist the pressure to become a party to a specific treaty is beyond the scope of this contribution and is heavily dependent on circumstances constantly evolving on the ground. In any case, the invalidity of territorial concessions under Art. 52 VCLT should not simply be presumed. While one of the objectives of Art. 52 VCLT is to bar the aggressor from harvesting the fruits of its aggression (Schmalenbach, Art. 52, at para. 23), state practice suggests that certain concessions can still validly be made. First, the 1999 Lusaka Ceasefire Agreement between Uganda and the DRC was implicitly accepted as valid by the ICJ in Armed Activities on the Territory of the Congo (2005, at paras 104-105), even though the agreement authorised the presence of Ugandan troops on the territory of the DRC after Uganda’s unlawful use of force. Second, Ukraine accepted certain concessions in 2015 already in the Minsk II agreement, e.g. to modify its constitution. These cases suggest that peace treaties can validly make certain concessions to an aggressor.
Ex post-validation of an invalid peace treaty by the UNSC
Even if a peace treaty falls victim to Art. 52 VCLT-invalidity, a binding UNSC resolution can “cure” the treaty, or at least replicate its content. Fox argues that “a Chapter VII resolution obviously cannot cure a peace agreement that violates the jus cogens prohibition on annexation”. However, a UNSC resolution validating or effectuating territorial concessions is perfectly feasible from a jus cogens perspective.
The starting point is that the UNSC is bound by peremptory norms, including the prohibition of the use of force. Similar to the rule on treaty invalidity discussed above, a UNSC resolution is only invalid if it conflicts with a peremptory norm. This rule was controversially discussed by the ILC, but eventually included in its conclusions on jus cogens (conclusion 16). The decisive question is whether a UNSC resolution transferring sovereign rights over Ukrainian territory to Russia, or ex post-validating a peace treaty to that effect, would be in conflict with peremptory norms. For the same reasons set out above, that would not be the case. Neither would such a UNSC resolution seek to render Russia’s aggression lawful nor would it permit annexations. Therefore, there is no conflict between the hypothetical UNSC resolution and these peremptory norms – at least not under the classical test of norm conflict.
Another argument based on jus cogens could invoke the obligation of non-recognition. Under customary international law as reflected in Art. 41 para. 2 ARSIWA and Art. 42 para. 2 DARIO, states and international organisations may not recognise as lawful situations created through serious jus cogens breaches. Aggression always constitutes such a serious breach. Arguably, a UNSC resolution approving of, or effectuating, territorial concessions to Russia would imply recognising Russia’s unlawful control over these territories as lawful, thereby breaching the obligation of non-recognition. It is important to note, however, that the obligation of non-recognition is not itself a peremptory norm; there is hardly any evidence that the obligation of non-recognition would be accepted and recognised as a peremptory norm by the international community of states (opinio juris cogentis). Therefore, the resolution might only violate the customary obligation of non-recognition, from which the UNSC is permitted to deviate when acting under Chapter VII of the UN-Charter. This is based on a widespread interpretation of Art. 103 UN-Charter, according to which Charter obligations prevail over customary international law (see e.g. Leiss and Paulus, at paras 73-75).
Finally, the UNSC could be precluded from overriding the effect of Art. 52 VCLT – but again, that would only be the case if that rule itself was peremptory in character. For lack of opinio juris cogentis, that is not the case (Forlati at p. 331-332, but see Milano, at p. 1018).
Territorial concessions remain in the toolbox of the international community
None of this is to say that territorial concessions to Russia are politically wise or desirable. The argument can well be made that territorial gains would only encourage the aggressor to resort to an unlawful use of force again. A peace treaty with such an effect would do a disservice to regional stability and international law at large. This seems particularly pertinent in the case of Russia, which has continuously expanded its territorial control in a series of wars with its neighbours (Moldova, Georgia, Ukraine). In short, there are important reasons for Ukraine not to make territorial concessions.
Nevertheless, other cases may be fundamentally different, and even an aggressor state could have a legitimate claim to disputed territory. Therefore, territorial concessions by a peace treaty should be considered admissible under international law. To that end, the classical interpretation of “conflict” in the context of jus cogens invalidity should be retained, and the causality test inherent to the “procured by”-element in Art. 52 VCLT should be cautiously assessed. Otherwise, the aggressor would be capable of taking territorial solutions out of the legal toolbox of the international community for crafting a peaceful solution.
