Causation in Climate Change: The International Court of Justice (ICJ)’s Climate Change Advisory Opinion vis-à-vis English Tort Law

[Dr. Juan-Pablo Peréz-León-Acevedo is a DPhil in Law candidate and a tutor at the University of Oxford. He also teaches at the Universities of Reading, Southampton, Oslo, and Abo Akademi (Finland)]

While the ICJ’s Obligations of States in respect of Climate Change Advisory Opinion (ICJ-AO) has been extensively analysed, this post examines an issue that remains largely unexplored, namely, how the ICJ-AO and a specific domestic legal system could interact in terms of causation in climate change litigation. Thus, this entry examines whether causation in English tort law (ETL) regarding attribution of liability for climate change diverges or converges from the ICJ-AO findings on causation, which can inform potential climate change litigation. ETL has been selected herein for two main reasons. First, as key legal developments concerning tort-based climate change litigation in other common law jurisdictions, such as New Zealand, arguably show (examined later), English tort case law has usually been influential. Second, since large multinational energy companies have their headquarters or major branches in the UK, tort law litigation before English/British courts for climate change-related damages caused in other countries (especially in the Global South) is expected to take place, as already evidenced by some ongoing proceedings in England (examined later).

Justification and Limits

This comparative analysis is justified because the ICJ-AO examined causation (paras. 433-438), which is central to ETL jurisprudence that has influenced common law jurisprudence, particularly in New Zealand’s tort-based climate change litigation. A common law grammar regarding causation in climate change litigation could potentially emerge. Moreover, the ICJ-AO provides statements about treaty and customary international law obligations binding on the UK, which could potentially involve ETL adaptations to climate change litigation. Finally, the UK participated in the ICJ-AO proceedings, and despite its restrictive approach to applicable state obligations concerning climate change, the UK’s intervention plausibly advanced  links between ETL with the ICJ-AO.

Nonetheless, while ETL considers causation within liability attribution, the ICJ-AO does so regarding remedies: after liability attribution. Yet, causation still has to be determined in climate change litigation. Although ETL targets private actors such as big polluters, and public offices as defendants, the ICJ-AO addresses states. Nevertheless, the ICJ-AO’s substantive law findings can impact future national/transnational litigation against private companies.

Similarities and Differences

Common law courts, such as New Zealand’s Supreme Court (NZSC) in Smith v Fonterra, and the ICJ-AO invoked the UN-Intergovernmental Panel on Climate Change (IPCC) findings to provide scientific background to attribution questions, including causation. Conversely, English courts have not yet explicitly invoked IPCC’s findings. Nevertheless, there is not yet an English court judgment on purely tort-based climate change litigation.

The ICJ-AO rejected two extreme positions: causation cannot be established because of the diffuse nature of climate change, and causation cannot be presumed in climate change: causation is required for reparation (paras. 434-435). ETL, mutatis mutandis, could proceed similarly on tort claim litigation for environmental damage. As the NZSC in Smith v Fonterra remarked:

‘It may indeed be beyond the capacity of the common law to resolve climate change […] but we are not presently convinced […] that the common law is incapable of addressing tortious aspects of climate change’.

(para. 154)

Furthermore, ETL nuisance cases (such as St Helen’s Smelting) were invoked to emphasise that the common law of torts can rise to new environmental challenges:

‘As a consequence of the long, global industrial revolution, the common law had to deal with new, widespread risk and damage caused by air and water pollution and the escape of biohazards’

(Smith v Fonterra, para. 157)

ETL could also balance plaintiffs’ and defendants’ rights similarly to how the ICJ-AO did:

‘How the law of torts should respond to cumulative causation in a public nuisance case involving newer technologies and newer harms (GHGs, rather than sewage and other water pollution) is a matter that should not be answered pre-emptively, without evidence and policy analysis.’

(Smith v Fonterra, para. 166)

Under the ICJ-AO, the ‘sufficiently direct and certain causal nexus’ standard applies to harm from anthropogenic GHG emissions, being flexible enough (para. 436). In ETL, the ‘but for’ test is the basic factual causation standard: ‘but for the defendant’s acts or omissions would the harm have occurred?’ (see, for example, Barnett v. Chelsea). However, the limited binary yes/no answers under this standard in, for instance, industrial activity-related damage caused by two+ factors, prompted English/UK courts to be more flexible in the said standard, which could apply to climate change litigation. Indeed, a manner in which the said ICJ standard and the ETL standard might interact in practice is that, for instance, English/British courts in tort-based climate change litigation could consider the ICJ-AO’s remark regarding the need for sufficient flexibility in causation: ‘to address the challenges arising in respect of the phenomenon of climate change (para. 436)’. In turn, this would enable ETL to better handle traditional tort law issues, such as remoteness, concurrent causation, and contributory negligence, which are also present in tort-based climate change litigation, as adapted or tailored to the specific challenges of the said global phenomenon. ETL cases developed two concepts particularly relevant. First, ‘material contribution’ to harm, namely, scientific/precise quantification of the contribution is unnecessary concerning two+ contributing factors: a factor stemming from the breach of duty above a ‘de minimis’ contribution suffices (see, for example Bonnington Castings). Second, ‘material increase in risk’ of harm (McGhee v. National Coal Board, a dermatitis case in the context of industrial work) applied in asbestos cases. In the seminal Fairchild, since the plaintiffs had worked for several employers, the ‘guilty’ asbestos fibre could not be identified. Thus, the House of Lords considered that each employer enhanced the risk of asbestos-related lung diseases stemming from their breaches of duty. In Sienkiewicz, the UK Supreme Court (Lord Phillips) clarified that the material increase in risk could apply only if it was ‘impossible’ to determine causation through conventional means; and established that epidemiological evidence, or other mathematical modelling/statistical methods, should be used carefully and usually supported by additional proof. ‘Increase in risk’ has been confined to asbestos litigation (‘Fairchild exception’); however, Sienkiewicz could incentivise claims against, e.g., energy companies, while overcoming causation challenges (p. 280).

‘to address the challenges arising in respect of the phenomenon of climate change’ (para. 436). In turn, this would enable ETL to better handle traditional tort law issues, such as remoteness, concurrent causation, and contributory negligence, which are also present in tort-based climate change litigation, as adapted or tailored to the specific challenges of the said global phenomenon. ETL cases developed two concepts particularly relevant. First, ‘material contribution’ to harm, namely, scientific/precise quantification of the contribution is unnecessary concerning two+ contributing factors: a factor stemming from the breach of duty above a ‘de minimis’ contribution suffices (see, for example Bonnington Castings). Second, ‘material increase in risk’ of harm (McGhee v. National Coal Board, a dermatitis case in the context of industrial work) applied in asbestos cases. In the seminal Fairchild, since the plaintiffs had worked for several employers, the ‘guilty’ asbestos fibre could not be identified. Thus, the House of Lords considered that each employer enhanced the risk of asbestos-related lung diseases stemming from their breaches of duty. In Sienkiewicz, the UK Supreme Court (Lord Phillips) clarified that the material increase in risk could apply only if it was ‘impossible’ to determine causation through conventional means; and established that epidemiological evidence, or other mathematical modelling/statistical methods, should be used carefully and usually supported by additional proof. ‘Increase in risk’ has been confined to asbestos litigation (‘Fairchild exception’); however, Sienkiewicz could incentivise claims against, e.g., energy companies, while overcoming causation challenges (p. 280).

Potential Convergences and Divergences

Unlike ETL jurisprudence, the ICJ-AO did not engage with factual causation. While the ICJ acknowledged that ‘it must be shown that the damage for which reparation is claimed has been factually and legally caused by a State’ (para. 422), it established that ‘Factual questions arising in the context of attribution and apportionment of responsibility are to be resolved on a case-by-case basis’ (para. 432). Nevertheless, unlike ETL jurisprudence, the ICJ-AO indicated no factual causation test. ETL on factual causation, also applied in other common law jurisdictions, could inform the ICJ when deciding on climate change litigation.

Concerning legal causation, the ICJ’s ‘sufficiently direct and certain causal nexus’ standard would correspond to the old standard in ETL, namely, ‘direct consequences’: liability for all direct consequences stemming from the defendant’s negligence, even caused by an unexpected event (In re Polemis). Nevertheless, the ‘reasonable foreseeability’ test crafted in Wagon Mound (No 1) became the new standard: whether damage was reasonably foreseeable or predictable. ETL has applied this ever since (see, for example, Hughes v Lord Advocate).

What standard should apply in climate change cases would depend on policy considerations: liability constraint or expansion. Alternatively, a new legal causation standard could be crafted regarding causation-related challenges of climate change litigation.

Operationalizing and Looking Ahead

Under the ICJ-AO, causation involves:

‘first, whether a given climatic event or trend can be attributed to anthropogenic climate change; and second, to what extent damage caused by climate change can be attributed to a particular State or group of States’.

(para. 437)

While science often addresses the former, the latter is established in concreto regarding states’ specific damage claims (para. 437). Scientific evidence invoked in ICJ-AO proceedings, for example, the IPCC’s 2023 Synthesis Report, linking the human contribution to climate change regarding increases in heatwaves, flooding, and drought, suggests that ‘significant harm to the climate system and other parts of the environment has been caused as a result of anthropogenic GHG emissions’ (para. 437). It added that: ‘the damage may be due to several concurrent causes, or the state of science regarding the causal link between the wrongful act and the damage may be uncertain’; the Court has to address these difficulties if/when they arise, considering facts and evidence; and it decides whether there is a sufficient causal nexus (para. 437). Subject to adaptations, these ICJ-AO findings could be useful when English/UK courts decide on tort-based climate change litigation.

In this context, in December 2025, 100+ Filipino victims of a typhoon brought civil claims for personal and property damages against Shell before the England and Wales-Court of Appeals (defendant’s domicile), claiming that Shell propelled this climate disaster with knowledge of the risks. Philippines’ law applies (European Parliament-Regulation 864/2007): the state where the damage occurred. Nevertheless, this tort-based case could arguably lead to jurisprudence integrating the ICJ-AO’s legal findings into English courts’ case law.      

ETL and tort jurisprudence from other common law jurisdictions (such as New Zealand) arguably indicate that ETL could handle causation-related challenges in climate change litigation through reliance on a material increase in risk whereby probabilistic causation is used, namely, to what extent a big polluter’s emissions increase the risk of, e.g., heatwave/forest fires leading to injuries, life loss, and property damages (p. 281). While English courts are acquainted with epidemiological evidence, partially similar to climate change attribution, such evidence has yet to be fully tested as decisive for causation (Corby Group Litigation). ETL would need to be more open to such evidence and expand the application of the increase in risk approach: from asbestos litigation to a general standard, also considering that GHG emissions can be a ‘single agent’ connecting all potential polluters (p. 282).      

Lord Sales of the UK-Supreme Court, in a speech, stated that while tort law is:

‘primarily backwards looking, focusing on apportioning liability for events that have already occurred […] regulatory legislation is more readily able to look to the future, with the Climate Change Act 2008 creating a legally binding obligation for the government to achieve net zero by 2050 and introducing a system of carbon budgeting to give effect to this objective’.

Nonetheless, Lord Sales remarked that in Smith v Fonterra where the defendants ‘were only a handful of the millions of persons responsible for global emissions and climate change would have happened without them’, the plaintiff ‘argued that it was sufficient that the defendants’ emissions had made a material contribution to the damage’, relying on English public nuisance cases concerning air/water pollution:

‘caused by multiple polluters where the courts had held their contribution was sufficient to establish liability’.

He promisingly added that:

‘the merits were not fully examined, and we will have to wait for the trial to see how a common law system similar to our own addresses a climate mitigation claim like this’.

Conclusion

Despite differences, there are potentially converging avenues that can benefit national and international systems in multilayered frameworks under ‘smart’ cross-fertilization to achieve better-tailored and balanced approaches to causation as a key element for climate change-related damages. In practice, ‘smart’ cross-fertilization in ETL litigation for climate-change related damages does not mean replacing the causation standards firmly established by English/British courts with a causation standard relatively recently adopted by the ICJ. Instead, this would mean that, when ETL makers and operators, mainly judges and legislators, engage with causation standards in the challenging climate change scenario, they can consider the jurisprudential principles from the ICJ-AO to better flesh out or apply factual and legal causation in terms of substantive law contents as well as evidentiary and procedural ‘best practices’ or guiding principles.