Erga Omnes Partes Standing and Procedural Issues in South Africa v. Israel

Written by

On 26 January, the International Court of Justice (ICJ) indicated provisional measures in a case brought under the Genocide Convention by South Africa against Israel concerning allegations of genocide related to the ongoing armed conflict in Gaza. South Africa relies on the theory of erga omnes partes standing which asserts that obligations under the Genocide Convention “are owed by any State party to all the other States parties to the relevant convention […] in the sense that each State party has an interest in compliance with them in any given case” (The Gambia v. Myanmar, para. 107).

While erga omnes partes standing has mostly been hailed as a long-awaited judicial tool to vindicate the rights of otherwise unrepresented victims of genocide or torture (here, here, here), this case demonstrates some of the procedural issues that erga omnes partes standing might cause in a Court whose procedure is strictly consent-based and tailored to the resolution of bilateral disputes between parties that are directly affected. To avoid these procedural shortfalls while still allowing erga omnes claims to be vindicated by non-injured parties, I argue that the ICJ should narrow its standing doctrine and declare applications by non-injured states inadmissible if a directly injured state was principally in a position to bring a case but did not do so.

Erga Omnes Obligations as a Foundation for Standing before the ICJ

While the notions of erga omnes- and erga omnes partes obligations have been around since the ICJ’s well-known obiter dictum in Barcelona Traction, it had been unclear whether they could serve as a basis of standing for non-injured applicants at the ICJ (here, 18-22 and here, 512). After the ICJ’s unambiguous confirmation of erga omnes partes standing in Obligation to Prosecute, it quickly became a popular basis for standing as it was invoked in recent applications by The Gambia against Myanmar, Canada and the Netherlands against Syria, and now by South Africa against Israel.

One common justification for erga omnes partes standing throughout the ICJ’s case law and academic literature is that it is necessary to vindicate legal claims of torture- and genocide victims who would otherwise have no representation (here, p. 20). This argument has so far been accurate, as most cases concern allegations of genocide or torture by a state against its own citizens or individuals within its territory. In such a setup, an actio popularis is the only conceivable way of allowing claims to be presented to the ICJ because a state cannot (and certainly would not) file against itself and diplomatic protection by another state is usually unavailable.

The structure of South Africa’s case against Israel, however, is fundamentally different in that it does not concern allegations of genocide by Israel against its own citizens or individuals on its territory, but against Palestinian nationals in Gaza—third-party nationals outside of Israel’s territory. Here, South Africa transformed a dispute between Israel and Palestine into a triangular lawsuit where the party on whose territory and against whose citizens genocidal acts are alleged, is absent. This novel case structure may lead to a number of procedural issues that are unresolved.

Absence of an Essential Third Party?

The first question relates to Palestine’s potential status as an indispensable party to the proceedings in whose absence the Court’s jurisdiction would be precluded. The ICJ held in the Monetary Gold case that even if it has jurisdiction over the parties to a proceeding, it is barred from entertaining their claims where the legal interests of a third party that has not consented to the proceedings “would form the very subject matter of the decision” (here, 32).

For South Africa’s claims against Israel (here, paras. 110-11), the Monetary Gold principle does not appear to be a substantial obstacle. South Africa’s claims are exclusively focused on Israel’s alleged violations of the Genocide Convention and it therefore seems feasible for the Court to write a narrow judgment that would not affect Palestine’s legal interests to a degree precluding jurisdiction.

One can easily imagine a scenario, however, where the adjudication of claims based on erga omnes obligations would require the determination of a third state’s legal interests. For instance, state A could bring a case under Article III (e) of the Genocide Convention against state B for ‘complicity in genocide’ in state C. Here, the determination of state A’s complicity would make the adjudication of state C’s commission of genocide the “very subject matter” of the dispute and the ICJ would be precluded from entertaining this claim. Thus, although the case between South Africa and Israel does not seem to be precluded under the Monetary Gold doctrine, future actio popularis cases may be barred.

Impossibility of Counter-Claims and Counter-Cases

A second concern relates to Israel’s inability of presenting meaningful counter-claims against South Africa or Palestine under the current structure of the case. Art. 80 of the ICJ’s Rules of the Court authorizes counter-claims “if [they come] within the jurisdiction of the Court and [are] directly connected with the subject-matter of the claim of the other party.” Thus, in the Bosnian Genocide case, where Bosnia accused Yugoslavia of various violations of the Genocide Convection, the Court permitted Yugoslavia, in turn, to introduce counter-claims that essentially alleged similar violations of the Genocide Convention by Bosnia (here, para. 5; Yugoslavia later withdrew its counter-claims).

In contrast, under the structure of South Africa v. Israel, the latter could only file counter-claims against South Africa since Palestine is not a party to the proceedings. Of course, there would be no point for Israel in bringing counter-claims against South Africa, unless Israel was to assert that South Africa is in some way complicit in violations of the Genocide Convention by Palestine vis-à-vis Israel. But, as explained above, such a claim would probably be precluded under the Monetary Gold doctrine.

The real issue here is, however, that Israel is barred from bringing a counter-lawsuit against Palestine even though Hamas’ actions might also fall within the scope of the Genocide Convention (here and here). This is because Palestine, as non-member observer state to the UN, has only recognized the ICJ’s non-member state jurisdiction pursuant to Security Council Resolution 9 for disputes arising under Article I of the Optional Protocol to the Vienna Convention on Diplomatic Relations. The result is a situation where Israel can be held accountable for alleged violations of the Genocide Convention on Palestinian territory, but cannot assert similar claims against Palestine as they would fall outside the Court’s jurisdiction. As Judge Xue already foresaw in her separate opinion in The Gambia v. Myanmar, similar concerns may arise when one of the parties to a conflict has entered a reservation with regards to the Court’s jurisdiction (here, paras. 28-29).

One might argue, as Judge Weeramantry did in his dissent in East Timor (172, also here, 427), that procedural fairness and equality of parties should yield when allegations of genocide are concerned. However, cases where both sides have legitimate claims to violations of erga omnes obligations, but where only one side can adjudicate them, may severely detract from the Court’s reputation for procedural fairness, equality of parties, and its overall legitimacy as an impartial judicial forum. Likewise, such a standing doctrine would invite ‘erga omnes partes standing shopping’ as states would be incentivized to present their claims through third states in order to avoid direct liability.

One-Sided Impact of Orders and Judgments

A third problem relates to the inter partes effect of the Court’s decisions. Article 59 of the ICJ Statute in connection with Article 94 of the UN Charter stipulates that the Court’s decisions only have binding force between the parties to a given case. Thus, third parties are generally not bound by the Court’s orders and judgments (here, para. 58).

In South Africa v. Israel this has the effect that both parties to the case are bound by the Court’s provisional measures and eventual judgment, but Palestine is—formally speaking—not. Practically, this is inconsequential when the Court limits its orders to the cessation of internationally wrongful acts, an obligation incumbent on all states in any case. In its provisional measures order, however, the Court ordered Israel to “prevent the destruction and ensure the preservation of evidence” and to report back after one month “on all measures taken” (here, para. 86). While these are reasonable provisional measures, it may again seem procedurally odd that the Court is unable to order similar ‘common sense’ measures vis-à-vis Palestine.

A related procedural issue concerns the principle of res judicata which mandates the finality and non-appealability of a judgment between the parties to a contentious case (here, para. 115). Generally, res judicata is understood as precluding future cases in which the parties, relief sought in the claim, and cause of action are identical (so-called ‘triple identity test’ here, para. 55). Applying this traditional standard to the new context of erga omnes partes cases may enable continuous re-litigation of similar issues by different non-injured states unless the Court applies its res judicata doctrine to cases where the claims are identical but the parties are not (here, 522).


Erga omnes partes standing has emerged as a powerful tool to bring claims on behalf of victims of the gravest international crimes. The dramatic case of South Africa v. Israel shows both the promise of this standing doctrine as a means for atrocity prevention, but also showcases a number of procedural issues the Court has yet to address. These issues are centrally connected to the triangular nature of this case in which South Africa, a state not specifically injured, represents legal claims of Palestine and its nationals even though Palestine would be able to bring these claims itself if it accepted the ICJ’s jurisdiction in this regard (as it did, for instance, to bring a case against the U.S., here paras. 25-35). The result is a situation where the Palestinian side can vindicate its erga omnes claims against Israel through South Africa as an interlocutor, while Israel is unable to make similar claims on behalf of its own citizens.

One way of resolving these issues would be by denying non-specifically-injured states standing as long as directly injured states are generally in a position to bring the case itself even if that would require them to accept the Court’s jurisdiction first. In the case at hand, the Court could deny South Africa’s standing by arguing that Palestine could have brought the case itself by simply agreeing to the Court’s jurisdiction over the Genocide Convention. This would, of course, make Palestine vulnerable to a counter-lawsuit by Israel, which, from a purely procedural perspective, appears consistent with the idea that both Palestinian and Israeli victims of this conflict should be equally able to bring erga omnes claims under the Genocide Convention.

Surprisingly, Israel did not challenge South Africa’s standing during the provisional measures hearings. It should do so when raising preliminary objections. Even though Myanmar’s somewhat similar objections regarding The Gambia’s standing were summarily dismissed (here, paras. 106-113), the facts of Israel’s case are sufficiently distinct for the Court to come to a different conclusion or at least clarify some of the questions raised in this post.

Print Friendly, PDF & Email

Leave a Comment

Comments for this post are closed


José Alves says

February 1, 2024

Great read, just two main points:

a) The ongoing (non-international armed) conflict is between Israel and Hamas. Whether the latter's acts could be attributable back to Palestine - whose statehood, by the way, remains contested by Israel and the West - remains open. I really doubt Israel would take any action that would (even if implicitly) recognize Palestine as a State.

b) The very definition of obligations "erga omnes partes" implies that the obligation is owed to all States parties. They are, therefore, entitled as an injured State to invoke the responsibility of another State who breached such obligations. Hence, the concept of "state not specifically injured" employed by the author doesn't make much sense to me, as it seems to go against the very notion of "erga omnes" in the first place.

Eugenio Carli says

February 2, 2024

Dear Dr Graf,

Thank you for this well-reasoned and thought-provoking post.

I agree that several procedural issues have already arisen and will likely arise in the South Africa v. Israel and in subsequent cases, were States’ trend to claim violations of erga omnes partes obligations before the ICJ be confirmed, as seems (fortunately) to be the case. Undoubtedly, the abovementioned dispute is quite exceptional, if only due to both its political and legal scope and implications.

I do share some of your concerns, but I would like to make a few comments.

As to the problem of an “essential third party”, rather curiously SA did actually invoke Israel’s responsibility for “complicity in genocide”, but did not substantiate that claim in its oral pleadings on PMs. Such a crime remains kind of obscure with respect to state responsibility. In the case at hand, it is unclear who or what the “principal offender” may be and in what way would have Israel contributed to the commission of genocide. One may even argue that the crimes of genocide and complicity in genocide are - to some extent - mutually exclusive (see the position of the ICJ in the Bosnia v Serbia 2007 judgment on this, para. 380). SA should have probably qualified its submission regarding “complicity in genocide” by Israel as alternative to the main one (as did Croatia in Croatia v Serbia).

As to the impossibility for Palestine to lodge a possible counterclaim, I agree that that can be problematic in “altruistic” procedures like in the case at hand, but I believe that the nature of the breaches claimed is such that the resort to some procedural mechanisms (such as counterclaims) is justifiably precluded or merely unlikely. In other words, the fact that a State sued by a non-injured State for a possible breach of an erga omnes partes obligation would reasonably not, or cannot, counterclaim against the Applicant for the same violations is, in my view, “part of the game” and it should be accepted as such. One of the effects of erga omnes partes obligations is indeed to allow all the States parties to a treaty to enforce them, and this feature has proved to be key in this case. Disputes regarding erga omnes partes obligations are by definition of a “collective” nature so that some strictly “bilateral” mechanisms – such as counterclaims – can be bypassed.
Israel has all the means to respond to the claims of SA; I do not see how its impossibility to make a counterclaim in this case would “severely detract from the Court’s reputation for procedural fairness, equality of parties, and its overall legitimacy as an impartial judicial forum”. This is true even considering that, under a practical perspective, the substantiation of a claim by Israel that Palestine, namely Hamas, has committed genocide in Israel – even where it was legally claimable – would be untenable.

Similar considerations apply, in my opinion, to the problem of the “one-sided impact of orders and judgments”. The exceptionality of the case and the importance of the “common interests” involved justify some possible procedural “oddity”.

Finally, the solution of “denying non-specifically-injured states standing as long as directly injured states are generally in a position to bring the case itself even if that would require them to accept the Court’s jurisdiction first”, that you have suggested, seems to go against the very notion of erga omnes (partes) obligations and the interpretation given by the ICJ to it so far. One of the reasons for the ICJ to recognize the ius standi of "non-directly injured States" lies in the will to ensure accountability in all those situations in which no "directly-injured State" can be identified, as is the case in most disputes of this kind (for example, no “directly-injured State” is identifiable in The Gambia v Myanmar and, as far as the Convention against torture is concerned, in Canada and the Netherlands v Syria). The ICJ wants to “equalize” the position of all the States concerned with respect to violations of erga omnes partes obligations and intentionally distances itself from the rigidity of the ILC articles on state responsibility under this profile, rightly so in my view.
Moreover, as I see it, the solution that you have proposed would create a situation in which the ICJ, on the basis of debatable (even arbitrary?) criteria, should deny the legal standing to a non-injured State in a particular case and recognize it in another, with the result of creating paradoxically more unfairness and inequality, and ultimately impairing accountability for these grave violations.

Again, the implementation of erga omnes (partes) obligations allows, in my opinion, some sort of procedural flexibility and adjustments.

As a final remark to emphasize the exceptionality of the case, please note the declaration of judge Xue, who, unlike in previous “actio popularis cases”, has not opposed the legal standing of the Applicant (though on the basis, in my view, of questionable grounds).

With kind regards,

Eugenio Carli