At last year’s UvA roundtable discussion with Philippe Sands, and a subsequent Rethinking SLIC Panel Discussion on Ecocide with Kate Mackintosh, we presented our views about the scope of liability for acts of ecocide in response to the Stop Ecocide Foundation Independent Expert Panel’s (IEP) proposal for the legal definition of ecocide. We welcomed the IEP’s contribution and agreed with the need for accountability under international law for environmental destruction. However, we also sought to consider an aspect of the IEP’s proposal that we felt had not received much attention: who could be held liable for ecocide? We examined how the Rome Statute’s framework for individual criminal liability could apply to a hypothetical scenario of ecocide. This exercise raised concerns that the proposed Rome Statute crime of ecocide lacks certainty and may lead to overcriminalization. Some reflections from our discussions on ecocide are set out in this blog.

1. The collective nature of ecocide

A holistic understanding of any law requires consideration about how it would apply in practice. We identified two gaps in the critical commentary on the IEP’s proposal, firstly the lack of factual hypotheticals about the types of conduct that could amount to the crime of ecocide. Secondly, the limited discussion about how Article 25 and 28 modes of Rome Statute liability would apply to this crime.

Some of the most egregious examples of environmental destruction, such as climate change, are best characterized as collective acts. The collective nature of environmental destruction shares similarities to occurrences of international crimes. Rome Statute crimes are ‘an expression of collective criminality’. Genocide, crimes against humanity, war crimes, and aggression apply not only to the upper echelons of political and military hierarchies, but also to those who make less significant contributions. Contextual requirements, such as attacks, armed conflicts, plans, and policies, mean that international crimes typically involve groups of individuals contributing to criminality. International modes of liability have developed to assign liability to individuals acting within these collective systems of criminality.

The jurisprudence of the ICC has already started to wrangle with the outer limits of criminal liability for accomplices, most notably in relation to the ‘purpose requirement’ in Article 25(3)(c) and the low actus reus and mens rea requirements of Article 25(3)(d). Under the latter provision, the actus reus for an accomplice who assists a group seems to be one of ‘any other contribution’, which could include even very minor contributions, while the mens rea extends to simple knowledge. The Court recognized in the Mbarushimana case the possibility of so-called ‘neutral’ contributions (Separate Opinion, paras 11-12), where even those whose actions are remote from the crimes could be held responsible.

Based on the IEP’s current definition, an ‘ecocide’ crime could criminalize contributions that collectively lead to a ‘substantial likelihood of severe and either widespread or long-term damage to the environment being caused by those acts’. The distinctiveness of the criminal conduct in putative instances of ecocide (e.g., the excessive burning of fossil fuels) lies in the society-wide extent of participation in these activities. The desirability of criminalizing a broad range of ‘ecocidal conduct’ would quickly lead to normative questions about who is causally responsible for the harm that arises from massive environmental damage. Global warming has arisen from the collective accumulation of uncountably numerous, separate instances of fossil fuel use by individuals, whether directly as they sit in traffic jams or switch on their gas-powered central heating, or more indirectly when they switch on electric lights that are powered by coal plants, or much more remotely when they consume products or services with a high carbon footprint. Still, while innumerable individuals have contributed to climate change throughout history, some contributions are calculably much greater than others.

Due to this abnormally broad range of potential contributors to some of the worst occurrences of ecocide, it is critical which individuals will be implicated by the proposed crime. There is nothing in the IEP’s definition that would limit liability to the actors most responsible for ecocide. Hypothetical versions of the ecocide crime must be read in light of the Rome Statute Article 25 and 28 modes of liability, in order to give shape to the contours of liability surrounding the range of contributions and large numbers of persons involved. The broadness of the IEP’s definition, coupled with the unsettled and potentially expansive ICC interpretations of Articles 25(3)(c) and (d), could extend liability to even quite minor and remote contributions.

The IEP did not highlight any examples of ecocide in its commentary. IEP member Philippe Sands subsequently described how a state’s authorization of the extraction and burning of coal ‘without properly taking into account impacts on the climate system’ could satisfy the definition’s requirements. Kevin Jon Heller countered that this example was unlikely to fall under the IEP’s current definition. In a recent webinar on ecocide’s potential applications in an Asia, Emma Carmody suggested three sets of scenarios: (1) large scale mining operations that have poisoned ecosystems, leading to prosecutions of directors, ministers, or a combination of both (giving the example of Shell and BHP), (2) largescale property developments e.g. hotel constructions that destroy entire protected areas, and (3) single large scale irrigation developments that destroy wetlands. While IEP members must have considered other examples of acts that could amount to ecocide, we are yet to come across a well-developed analysis of how Rome Statute modes of liability, particularly secondary liability, would apply. Our hope, in starting to examine how Rome Statute modes of liability would apply to hypotheticals, is to open the discussion about the scope of liability. From there, we can begin to ask whether the extent of this criminalization would be desirable.

2. Examining the ecocide proposal in light of the ICC’s modes of liability

To examine how the Rome Statute modes of liability could apply, we start with our own expanded version of the hypothetical provided by Philippe Sands concerning the authorisation and operation of a coal mine:

A company seeks authorisation to mine an area of coal. The company consists of directors, senior management, middle management, supervisors, and employees. The operation of this mine requires government officials to approve the mining site. Contracted geologists and engineers help to plan and develop the mines operations. The mining company purchases or rents mining materials, machinery, and other infrastructure to operate the mine. Employees mine coal under management’s supervision. Management reports the mines activities back to the company’s directors. The company sells the coal for use as fuel.

Recent real-world examples of coal mines sharing similarities to this hypothetical include the Aberpergwm mine in the U.K., the Parsa East and Kante Basan mines in India, and the Narrabri mine in Australia. Undoubtedly, there are many more examples of this ilk.

The most readily applicable forms of participation here relate to the ICC’s co-perpetration and complicity types of liability. First, the Rome Statute holds individuals criminally responsible for committing crimes through the doctrine of co-perpetration. Perpetration occurs when a person ‘[c]ommits […] a [Rome Statute] crime, whether as an individual, jointly with another or through another person, regardless of whether that other person is criminally responsible’. According to prevailing ICC jurisprudence, co-perpetration requires that an individual intentionally provides an essential contribution to the crime (Lubanga paras 330, 347).

To establish who perpetrated ecocide in the coal mine hypothetical, we must first identify the act or acts that could cause “a substantial likelihood of severe and either widespread or long-term damage to the environment”. Depending on what is considered damage to the environment, individuals may have committed ecocide in different ways. The mine itself could directly cause significant environmental damage to a particular area of land or ecosystem. Alternatively, the extraction of coal could indirectly contribute to global warming if it is later burned as a fossil fuel. Either way, it is feasible that a large coal mine could inflict ‘severe and either widespread or long-term’ damage either directly on the earth’s surface or indirectly on the atmosphere. We highlight concerns with indirect endangerment in part 3.B below.

If we consider the authorization, development, and operation of the mine as acts that could cause environmental damage, either directly or indirectly, any persons who intentionally provide essential contributions to these activities could be held liable as co-perpetrators, provided they have the adequate mens rea.1 A state official who authorizes the coal mine or a CEO who oversees the coal mine might provide essential contributions to the crime. Employees within the upper management of the coal company or those integral to the mine’s operation might also provide essential contributions.

Second, as noted above, the Rome Statute includes secondary modes of liability in Article 25(3)(c) and 25(3)(d). Aiding and abetting under Article 25(3)(c) of the Rome Statute requires that an individual carries out acts or omissions, for the purpose of facilitating the crime, that encourage or support the principal crime. The act of assistance must affect the crime, and the accomplice must intend their own underlying act of help or influence. Article 25(3)(d) assigns liability for any other contribution to criminality. To trigger liability under 25(3)(d), an accused must provide any other contribution to a group carrying out criminal conduct, with a mens rea as low as knowledge of the crimes to which they are contributing.

Arguably, a broad range of individuals in our hypothetical could have the necessary actus reus for secondary liability under the Rome Statute by providing ‘any other contribution’. The contributions of coal mine employees, individuals who supplied services or machinery used at the mine, or geologists and engineers who helped to develop the mine might be considered sufficient to satisfy the requirements of 25(3)(c) and 25(3)(d) as contributions to acts that have, under the IEP definition, a ‘substantial likelihood’ of causing environmental damage.

Regarding the mens rea requirements for secondary liability under the Rome Statute, at the very least, one can envisage that employees, coal miners, service agents, or engineers who design the plant would all intentionally carry out their contributions. It is less clear whether they would have sufficient simple knowledge under 25(3)(d) that the group of coal company employees, perhaps together with the CEOs and state officials, has a criminal intention to execute the plan to mine and burn coal in a manner that disregards the environmental harm at stake. There is also uncertainty in the ICC jurisprudence over so-called ‘neutral’ contributions’, as raised in Mbarushimana (Separate Opinion, paras 11-12) and the extent of Rome Statute liability over remote contributions. This leaves the possibility of ICC liability even for actions such as sub-contracting minor aspects of the work of the coal mine.

Other Rome Statute modes of liability may raise further, legally interesting questions about who could be held responsible for ecocide. For example, whether a CEO who operates a coal mine could be held liable under civil superior responsibility.

The coal mine hypothetical reiterates the need to understand first the descriptive question of who is implicated by the IEP’s text before we can discuss whether that is merited at the societal level. By exploring how liability models could apply to hypothetical examples of ecocide, we can gain a greater understanding of how this proposed law will function. This allows us to better understand the acts that may trigger criminality, as well as the actors who may be liable for the crime. From there, we can consider whether there is over - or under-criminalization, and whether the crime is indeed enforceable.

3. Additional thoughts on the scope of liability

Here, we add two brief observations related to the scope of ecocide. Each of these issues highlights concerns that the IEP’s current proposed definition could be impermissibly broad and vague, undermining the principle of legality. Moreover, each of these issues, when properly clarified, would alter the range of individuals who are implicated by the IEP’s current definition.

A. The term ‘environment’ lacks specificity

The proposed definition of ‘environment’ lacks specificity. The commentary states that the environment is “the earth, its biosphere, cryosphere, lithosphere, hydrosphere and atmosphere, as well as outer space” (IEP’s commentary, section 2.c). This definition suffers from undetermined scalability, and it is unclear how the lines of criminality would be drawn. Depending on what one considers the ‘environment’, different acts could be criminalized by ecocide.

If we consider the environment to mean goliath concepts such as “the earth, its biosphere, cryosphere, lithosphere, hydrosphere and atmosphere, as well as outer space”, the threshold requirement of ‘severe and either widespread or long-term damage’ becomes extremely high. The world’s largest coal mine may only cover a tiny fraction of the ‘lithosphere’ or be responsible for a fraction of yearly CO2 emissions into the ‘atmosphere’. At this scale, it would be difficult to establish that the mine caused the requisite damage to the environment. On the other hand, if the term ‘environment’ includes concepts like a forest or an animal subspecies, we immediately lower the scope and scale of damage required. Acts such as deforestation or poaching could amount to ecocide. While the threshold requirement of ‘severe and widespread or long-term damage’ attempts to filter out smaller scale acts, this threshold relates only to the damage required. The damage scale is thus dependent on what constitutes an environment.

Without knowing what exactly constitutes an ‘environment’ under the IEP definition, it is very difficult to predict what acts might amount to ecocide, and who might be liable for them. We are thus unable to know the scope of liability for ecocide.

B. The crime of ecocide does not require actual harm to occur

Our second concern is that the actus reus of ecocide can be too remote from any actual harm. This stems from the proposed crime being one of endangerment. As the commentary notes: “[c]ulpability for the crime of ecocide attaches to the creation of a dangerous situation, rather than to a particular outcome. It is the commission of acts with knowledge of the substantial likelihood that they will cause severe and either widespread or long-term damage that is criminalized” (IEP’s commentary, section 4). Based on this explanation, it seems that there is no need for actual harm against the environment. This makes ecocide a blunt tool.

Consider two oil tankers that leave a shipyard, and on both ships the captains know that crucial safety procedures have not been followed. Both captains have knowledge of a substantial likelihood that severe and long-term damage could occur. Ship A ruptures causing a catastrophic oil spill. Ship B sails safely to port. Without the requirement for actual harm, both ship captains are equally guilty of ecocide. It is worth considering whether ecocide should recognize a difference between these two ships. Moreover, it can be extremely difficult to predict what constitutes a ‘risk’, and incorrect determinations of risks could lead to premature prosecutions.

There is also a question of whether the endangerment can be indirect under the IEP’s proposal. The IEP has not clarified whether indirect endangerment would be criminalized under their proposal, but examples provided by their members suggest that it would.2 Direct endangerment would involve an act that is substantially likely to cause direct damage to the environment (e.g., operating a coal mine could directly damage the surrounding land), whereas indirect endangerment would involve an act that is substantially likely to create a situation whereby further actions could cause damage (e.g.,  mining coal could indirectly damage the atmosphere if that coal is later burned by others). Whether the endangerment can be indirect or direct is important for determining who could be liable as a perpetrator for ecocide, and for delineating the difference between perpetrators and accomplices.

With no requirement for actual environmental damage to occur, and with the potential to include acts that might indirectly damage the environment, the scope of ecocide under the IEP’s definition may be impermissibly broad, extending to individuals whose acts are too far removed from actual harm.

4. Rethinking the scope of ecocide

Underlying the concerns highlighted in this blog is a lack of clarity as to how ecocide will apply in practice, and to whom it will apply. The current IEP proposed definition does not sit well with the principle of legality, in particular the principle of specificity. Criminal laws should be adequately precise to guide behavior and put potential defendants on notice as to which acts are indeed criminal.

Of course, the definition could be made more specific by the IEP and is likely to be developed over time by expert input or replaced by alternative approaches to addressing environmental harm at the international level. We appreciate the IEP’s initiative to extend international criminal law to provide greater protection against environmental harms, perhaps reflecting a shift in global priorities since the Rome Statute was negotiated. However, we strongly believe there are significant hurdles in criminal law theory and practice that would need to be addressed before the IEP’s current definition could be workable in actual cases.



1 Kevin Jon Heller noted the difficulties with establishing the mens rea for ecocide, as well as challenges with the wanton requirement, based on the IEP’s proposal. We will assume for the purpose of this hypothetical that the individuals involved are aware that their actions may cause a substantial likelihood of severe and either widespread or long-term damage to the environment.

2 This assumption feels safe. IEP members have referred to fossil fuel extraction companies’ impact on global warming and climate change when discussing ecocide (see here). For the most part, companies that extract fossil fuels indirectly harm the atmosphere, given that others must burn the fossil fuels to release harmful gasses.