Compensation after the Climate Advisory: Reading the ICJ’s Remedies Jurisprudence

On 23 July 2025, the International Court of Justice delivered its Advisory Opinion on Obligations of States in respect of climate change. On remedies, the Court reaffirmed the familiar consequences of State responsibility—cessation and guarantees of non‑repetition, and “full reparation” through restitution, compensation and/or satisfaction (paras 446-455). The Court acknowledged that compensation remains anchored in Chorzów and ARSIWA Article 31’s full‑reparation principle (para 450). It also underlined three propositions central to climate litigation: (i) environmental damage “in and of itself” is compensable and may include indemnification for the impairment or loss of environmental goods and services (para 453); (ii) causation turns on whether a “sufficiently direct and certain causal nexus” exists, a flexible standard the Court considers suitable for climate claims (paras 436-437); and (iii) where the precise extent of damage cannot be established, the Court may, on an exceptional basis, award a “global sum” within the range indicated by the evidence, taking equitable considerations into account (para 454).

This post reads that framework against the Court’s recent compensation jurisprudence—Ahmadou Sadio Diallo (2012), Certain Activities (Costa Rica v Nicaragua) (2018), and Armed Activities (DRC v Uganda) (2022). The central claim is that Costa Rica v Nicaragua furnishes the most workable template for climate‑related compensation, whereas leaning on Armed Activities’ “global sum” risks amplifying the very uncertainties the Court aims to cabin.

Three building blocks of the Court’s recent compensation jurisprudence

1) Ahmadou Sadio Diallo (2012): Equity, but thin reasons

In Diallo, after finding breaches of human‑rights obligations, the Court awarded Guinea US$85,000 for non‑material injury and US$10,000 for certain material injury to personal property. The outcome is clear, but the reasoning for quantum is sparse; equity does most of the work without a transparent method (paras 13, 24-25, 36, 56). That sits uneasily with Chorzów’s demand that full reparation be determined through ascertainable legal standards, not numbers chosen without explanation. Members of the Court have long worried that equity, used this way, blurs the line between lawful approximation and arbitrariness  (Diallo, Declaration of Judge Greenwood, para 7; Diallo’s absence of any specific reason for awarding compensation was pointed out: Armed Activities, Separate Opinion of Judge Robinson, para 28).

2) Certain Activities (Costa Rica v Nicaragua) (2018): A replicable method

In Costa Rica v Nicaragua, the Court delivered its first judgment determining the amount of compensation for international environmental harm (para. 41). Its methodology rested on three key elements. First, it reaffirms that environmental damage “in and of itself” is compensable and may include indemnification for impairment or loss of environmental goods and services prior to recovery, as well as expenses reasonably incurred (paras 29-33, 41-43). Second, it sets out a stepwise assessment: identify each head of damage; determine whether a “sufficiently direct and certain” causal nexus exists between the wrongful act and that head of injury; and then quantify (paras 29-33, 72). The Court also reiterated evidentiary flexibility, allowing the burden to adapt where the respondent is better placed to produce relevant data (paras 33-35). Third, for quantification, the Court adopted an “overall assessment” of impairment or loss of environmental goods and services at the ecosystem level (paras 78-86). This approach accounts for interactions among different services (e.g., how tree removal cascades through raw materials, gas regulation and biodiversity), the protected‑area context, and natural regeneration capacity. On this basis the Court awarded compensation for impairment/loss “prior to recovery,” plus costs and expenses supported by evidence, while denying items not sufficiently linked to the harm (paras 84-87, 89, 92-101, 112-119, 140-155). The method is explicit, evidence-driven, replicable and reviewable.

3) Armed Activities (DRC v Uganda) (2022): The “global sum”

In 2022, in a reparations judgment spanning multiple heads of loss, the Court awarded a “global sum” of US$325 million (Ventura). The methodological hallmark—awarding a global sum on an exceptional basis—was justified by destroyed or inaccessible evidence and the long time since the conflict (para 106). The approach drew criticism for opacity and heavy reliance on equitable considerations absent a transparent computation (Armed Activities, Declaration of Judge Tomka, para 9; Separate Opinion of Judge Yusuf, para 23; Fikfak, p 224; Desierto). The Advisory Opinion nonetheless quotes the global‑sum device and signals its availability, but stresses its exceptional character (para 454).

What the Advisory Opinion actually says about compensation

The Opinion restates Chorzów’s principle — reparation must “wipe out all the consequences of the illegal act” — and situates compensation alongside restitution and satisfaction as alternative or combined forms of full reparation (para 450). Restitution will often be impracticable for environmental harm, though it may include reconstructing infrastructure or restoring ecosystems and biodiversity (para 451). On compensation, the Court does not quantify or apportion in the abstract, but it confirms that compensation could be owed for significant harm caused by climate change if a sufficiently direct and certain causal link is shown between the wrongful acts of one or more States and the harm (para 452). It also underscores that environmental damage itself and expenses incurred as a result of such damage are compensable (para 453). Exceptionally, a “global sum” may be awarded within evidential ranges and equity, while acknowledging that in climate cases quantification will often be difficult due to uncertainty about the exact extent of damage (para 454).

On causation, the Court affirms that the “sufficiently direct and certain causal nexus” standard—already used in its case law—is flexible enough for climate disputes (paras 434-437). The standard involves two inquiries: attribution of climatic events or trends to anthropogenic climate change (often addressed by science), and an in concreto link between a State’s wrongful conduct and the injury in the specific case (para 437).

How (and how not) to read paragraph 454 of the Advisory Opinion

Paragraph 454, already being flagged as a “[point] of interest,” reiterates that a “global sum,” within evidential ranges and with equitable considerations, may be awarded on an exceptional basis, and notes that climate compensation is often hard to calculate because there is usually uncertainty about the precise extent of damage. Read in context, four limits emerge.

First, equity is a guide when proof of extent is insufficient despite best efforts; it cannot supplant full‑reparation analysis. This is the restrained role equity plays in Costa Rica (para 35) and the best reading of Diallo in narrow compass. Second, the causation standard is neither relaxed to presumption nor tightened into impossibility. The “sufficiently direct and certain” test is said to be flexible enough for climate claims, with science addressing anthropogenic attribution and case‑specific evidence addressing the link to wrongful conduct (paras 435-437). That logic favours head‑by‑head assessment over single‑number awards wherever the record allows it.

Third, the Opinion’s nod to Armed Activities must be read alongside its embrace of Costa Rica. By recognising compensable environmental damage “in and of itself” and referencing goods and services prior to recovery (para 453), the Opinion imports Costa Rica’s categories, which are amenable to evidence‑based quantification even under uncertainty and do not require collapsing everything into one figure. Fourth, the Opinion emphasises case‑by‑case judgment and refrains from apportioning responsibility or quantifying harm in the abstract (paras 431-432, 452). That caution is acute in climate litigation, where multiple States contribute to a single harm. Responsibility may be invoked against each responsible State; apportionment and contribution are for in concreto assessment (paras 430-432).

In short, paragraph 454 is a safety valve: it prevents the absence of precise proof from defeating reparation altogether but does not relieve claimants of the need to structure and prove claims the Costa Rica way.

Practical implications for climate disputes

Head-by-head claims, evidence first. Climate cases should follow Costa Rica’s model: identify environmental goods and services harmed, cost out restoration, and document expenses. Attribution science establishes the general causal chain from emissions to climate impacts, while wrongful conduct must be tied to specific injuries. “Overall assessment” should be reasoned ecosystem by ecosystem.

Disciplined use of evidentiary flexibility. The Opinion supports flexible proof rules and a causation standard tailored to climate realities, including burden-shifting and reliance on best available science. But flexibility does not excuse weak claims—each head of loss must be documented and causally linked to harm.

Global sums only as a last resort. Where precise valuation is impossible—such as slow-onset ecosystem losses—paragraph 454 permits evidence-based ranges with equitable considerations. Otherwise, Costa Rica’s head-by-head test better serves full reparation and withstands scrutiny.

Responsibility and compensation. Obligations to prevent significant transboundary harm are erga omnes: non-injured States may seek cessation and guarantees, while compensation belongs to the injured and should be pursued with evidence-based claims.

Enforcement matters. Past cases show both resistance (Corfu Channel, Akande) and compliance (Uganda’s instalments). Transparent, well-reasoned calculations enhance respect for judgments, while awards must remain proportionate to avoid overburdening weaker economies.

Conclusion

The Climate Change Advisory Opinion consolidates rather than revolutionises the Court’s approach to compensation: Chorzów’s principle remains the lodestar; environmental damage “in itself” is compensable; the “sufficiently direct and certain” causation standard is reaffirmed as flexible; and Costa Rica’s granular, ecosystem‑level method is implicitly endorsed. The nod to Armed Activities’ “global sum” based on equitable considerations should not be overread. In climate cases it is a narrow, exceptional device for residual uncertainty that remains after serious proof efforts. The Court’s minimal reference to Diallo in the Opinion—viewed primarily through the perspective of Costa Rica—suggests its implicit disapproval of Diallo for placing excessive weight on vague equitable principles rather than on the evidentiary standard.

Thus, the most promising path the Opinion points to is the Costa Rica path: build claims ecosystem‑by‑ecosystem, head‑by‑head; use science to establish general causation; prove, as far as possible, specific links to wrongful conduct; and explain the valuation method used. Deployed that way, compensation in climate disputes will be demanding but doable—and truer to the promise of full reparation than any single, undifferentiated number could be.