Pick and Choose at the ICJ

The International Court of Justice (ICJ or Court) has apparently begun to deviate from its own long-standing standards. Its Provisional Measures Orders (PMOs) in Nicaragua v. Germany, Equatorial Guinea v. France and Mexico v. Ecuador illustrate the emergence of a new approach to assessing “circumstances” required for the indication of provisional measures under Article 41 of the ICJ Statute. Traditionally, the Court uses a chronological five-prong test since Belgium v. Senegal.

Recently, however, the Court has adopted what may be called a “pick and choose” approach to that test. While giving the Court a certain degree of flexibility in assessing the circumstances, this approach creates risks of arbitrariness and unpredictability.

The Five-Prong Test for Indicating Provisional Measures

According to Article 41 of its Statute, the ICJ has the power to indicate, if it considers that circumstances so require, any provisional measures which ought to be taken to preserve the respective rights of either party. What those “circumstances” are, remains, however, unclear. As such, the Court consolidated the five-prong test in Belgium v. Senegal. The said test involves the assessment of five conditions, namely, (i) prima facie jurisdiction, (ii) plausibility of the alleged rights, (iii) link between the alleged rights and measures requested, (iv) risk of irreparable prejudice, and (v) urgency.

Since Belgium v. Senegal, the Court has cumulatively assessed these conditions by following a logical sequence, albeit without expressly spelling out the mandatory nature of this sequence. The Court had earlier introduced (and to some extent consolidated) “prima facie jurisdiction” in UK and Germany v. Iceland in 1972, “plausibility of the alleged rights” in Belgium v. Senegal in 2009, the “link between the alleged rights and measures requested” in Argentina v. Uruguay (Second PMO) in 2007, and the “risk of irreparable prejudice” and “urgency” in Finland v. Denmark in 1991. Since then, all PMOs consistently adopted the absolute approach to the five-prong test – until the recent three cases.

As we will see, the “pick and choose” approach was adopted in PMOs in which the Court declined to adopt provisional measures (negative PMOs). To understand the differences between the traditional approach and the “pick and choose” approach, it is thus useful to examine the negative PMOs prior to Nicaragua v. Germany.¹⁾

In Belgium v. Senegal, the Court, after analysing all five conditions, declined to adopt provisional measures due to the non-existence of risk of irreparable harm and urgency (paras. 72–73). Similarly, in Nicaragua v. Costa Rica, the Court found the first three conditions satisfied but not the last two (paras. 34–35). In Ukraine v. Russia, the Court established prima facie jurisdiction for the requests related to the International Convention for the Suppression of the Financing of Terrorism (ICSTF) but did not find those rights plausible, and therefore correctly declined to adopt ICSTF-related measures without assessing the remaining conditions.

These cases show that the ICJ used to assess the five conditions in chronological sequence and ceased further analysis once one condition was not met. This was justified by the cumulative nature of the test: if any one of the cumulative five elements fails, the Court lacks the power to indicate provisional measures. Another reason is the doctrine of judicial economy, which aims at evading unwarranted efforts or expenses of the courts.

The (novel) “pick and choose” approach

Nicaragua v. Germany marks the starting point of the ICJ’s “pick and choose” approach to the five-prong test. In this case, the Court deviated from its long-established approach by not assessing any of the five conditions and directly concluded that “the circumstances are not such as to require the exercise of its power under Article 41 of the Statute to indicate provisional measures” (para. 20). As indicated by Judges Iwasawa and Cleveland (along with Judge Sebutinde), the Court reached this conclusion due to the lack of “risk of irreparable prejudice” and “urgency”. The Court’s direct move to the last two conditions, bypassing the first three, was regarded even within the bench as exceptional. However, given the appalling situation in Gaza, positive assertions of the German government, and the Court’s choice of approach, Wentker and Stendel described it as a “pragmatic approach.” Becker termed this approach as “unusual”. Yet, what seemed exceptional soon became the norm.

In Mexico v. Ecuador, the ICJ again adopted the “pick and choose” approach, this time with some clarification. It explained that its decision to decline provisional measures requests was due to the non-existence of “risk of irreparable prejudice” and “urgency,” without assessing the first three conditions (paras. 28, 33–34) – just as in Nicaragua v. Germany. But this time, the Court appears to justify its approach by reference to Ecuador’s unilateral declarations, noting that such declarations are binding upon States. In particular, the ICJ assessed the unilateral assurance of the Ecuadorian government to provide full protection and security to the premise and persons of Mexican Embassy in Quito against the above-mentioned conditions and concluded their non-existence (paras. 31-32).

The ICJ applied the “pick and choose” approach again in Equatorial Guinea v France, though this time considered only the second condition, i.e., “plausibility of the alleged rights”. Unlike the earlier two cases, it did not involve any unilateral declarations or undertakings from the Respondent State. This was striking, given that both parties had made extensive submissions on prima facie jurisdiction (see Separate Opinion of Judge Yusuf, paras. 4-7).

Although none of the judges disclosed the precise jurisprudential origin of this approach, it can be inferred from the Separate Opinion of Judge Iwasawa (Nicaragua v. Germany, para. 5) and the Declaration of Judge Nolte (Mexico v. Ecuador, para. 8) that the Court may have drawn inspiration from Greece v. Turkey, where it reached (negative) conclusions without exploring prima facie jurisdiction. However, the position of Judge Tladi in Nicaragua v. Germany (Declaration, paras. 9, 12) and Equatorial Guinea v. France (Separation Opinion, paras. 1-5, 17) implies that this novel approach (or, in his words, “a flexible and fluid approach”) should be limited to situations involving unilateral declarations and conditions of irreparable prejudice and urgency.

Why not to “pick and choose”

The “pick and choose” approach has already attracted strong criticism from the Bench. For instance, Judge Sebutinde termed it a “strange approach” and Judge Yusuf an “unsound approach”. Judge ad hoc Al-Khasawneh, who shared the bench as a regular member in Belgium v. Senegal, described it as “a novel approach constituting a departure from its established jurisprudence”. Judge Nolte, referring to this approach as “shortcuts”, expressed his expectation for adherence to the traditional approach, even in negative PMOs. Judge Yusuf questioned whether such an approach was a judicial innovation or a temporary glitch, ultimately expressing his conviction that it was the latter.

The “pick and choose” approach indeed appears to be glitchy. It departs from the ICJ’s legal framework on provisional measures, at least in three respects.

First, PMOs do not have res judicata effects, and Disputant States may apply for the revocation or modification of earlier provisional measures, or even for the indication of new ones (Article 76, Rules of the Court). Generally, the traditional approach to the five-prong test relieves the Court from reassessing all five conditions afresh in subsequent PMOs. However, if the Court follows the “pick and choose” approach, it may later be compelled to reassess all the conditions anew in subsequent provisional measures requests. In that case, taking a shortcut in the first PMO may ironically lead to a longer route in the subsequent ones, culminating a situation of false judicial economy.

Second, the ICJ’s reliance on the existence of unilateral undertakings to adopt the “pick and choose” approach is inherently problematic. Since such undertakings in questions are generally linked to the subject matter of the proceedings, considering their binding nature directly when assessing irreparable prejudice and urgency – without first confirming the Court’s jurisdiction over those subject matters, their connection to the requested measures, and the plausibility of the alleged rights – risks undermining the very purpose of provisional measures.

Moreover, the Court should exercise caution in evaluating the adequacy and appropriateness of unilateral undertakings, as it notably did in Timor-Leste v. Australia. Furthermore, adherence to the absolute approach would relieve the Court from the need to reassess all the conditions in the event of a request for modification of provisional measures following a breach of unilateral declarations during the pendency of the case (for context, see New Zealand v. France).

Finally, the “pick and choose” approach may frustrate the legitimate expectations of the parties and create a legitimacy crisis for the Court. This concern was highlighted by Judge Tladi, who endorsed the approach in Nicaragua v. Germany but later criticized it in Equatorial Guinea v. France. In that case, he said that it is arbitrary to begin with a condition and end with it without justifying such a choice (para. 6).

Importantly, this issue is not only relevant for Applicant States but also equally significant for Respondent States. Although the ICJ is not required to forecast the outcome of a case at the stage of PMOs, the outcomes of PMOs nevertheless assist both parties in strategizing during all stages, as the reversal of outcomes of PMOs at the preliminary objections stage in Ukraine v. Russia illustrates. On the other hand, unusual treatment of the five conditions in PMOs, e.g., the direct assessment of plausibility in Equatorial Guinea v. France, may also give the impression of arbitrariness by the Court (Separate Opinion of Judge Tladi, para. 6), and even prejudge a particular case (Separate Opinion of Judge Yusuf, paras. 16-17).

Conclusion

As Judges Yusuf and Nolte have emphasised, the ICJ would do well to return to its traditional approach to the five-prong test. That test was originally developed in a negative PMO involving unilateral undertakings (Belgium v. Senegal) and later applied in a positive PMO (Timor-Leste v. Australia). From its inception, then, the traditional approach has proved well-equipped to deal with similar situations.

By contrast, the current “pick and choose” approach, though offering some degree of procedural flexibility to the Court, risks rendering future PMOs more unpredictable and arbitrary – although the ICJ developed that test, and its logical sequence, precisely to prevent such arbitrariness and ensure certainty. A return to the traditional approach is therefore not only doctrinally warranted but also crucial for the sound administration of justice and for reinforcing public trust in the ICJ – particularly at a moment of profound crisis for the international legal order.