2.1. Widespread Adoption

Since the United States (US) adopted the National Environmental Policy Act (NEPA) in 1969,Footnote ¹⁶ most countries, many subnational governments, and some other entities have established procedures for the assessment of the environmental impacts of activities that are likely to affect the environment.Footnote ¹⁷ These procedures aim at ensuring that decisions are well informed, rather than to pre-empt or substitute for political deliberations.Footnote ¹⁸ They typically apply notwithstanding whether the project is planned by a public agency or a private entity. In some jurisdictions, a distinction is drawn between ‘EIA’, which applies to concrete projects, and ‘strategic environmental assessment’ (SEA), which applies to broader policies, plans or programmes.Footnote ¹⁹ However, this terminology is not always used consistently, or at all.Footnote ²⁰ This article uses ‘EIA’ and ‘project’ in a broad and generic sense. EIA frameworks usually involve six components: (i) the screening of planned activities to determine the need for an assessment procedure; (ii) the scoping of the assessment; (iii) the realization of the scientific study; (iv) public consultations; (v) a formal decision on whether to approve, impose conditions on, or reject the project; and (vi) monitoring compliance with the decision.Footnote ²¹

Many EIA laws and regulations did not initially specify whether an assessment of GHG emissions was required. Yet CEA requirements were often identified through an interpretation of these instruments and through a form of contagious lawmaking.Footnote ²² Thus, in as early as 2003, a US court found that the carbon dioxide (CO₂) of a power plant could fall within the scope of the impacts to be assessed under NEPA.Footnote ²³ In the following year, a court in the Australian state of Victoria interpreted the state’s EIA framework as implying that the EIA for a coal-mining project had to include a CEA.Footnote ²⁴ Similar decisions were later reached in non-Western countries, for instance, with regard to coal-fired power plants, including in South Africa (2017),Footnote ²⁵ Kenya (2019),Footnote ²⁶ Chile (2022),Footnote ²⁷ and Indonesia (2022).Footnote ²⁸ In 2019, the National Green Tribunal of India dismissed a petition contending, among other things, that the government had failed to establish a mandatory CEA: in the Tribunal’s view, this requirement already existed, implicitly, in the existing regulatory framework.Footnote ²⁹ Even in the absence of such judicial decisions, administrative agencies have sometimes taken the initiative of requiring project proponents to undertake a CEA – for instance, in Mauritius,Footnote ³⁰ Minnesota (US),Footnote ³¹ and China.Footnote ³²

Over time, a growing number of jurisdictions have sought to enhance legal certainty by including an express CEA requirement in their EIA instrument or in additional documents. In contrast to judicial decisions, these developments often clarify the modalities of a CEA, including the types of climate effect to be assessed and the methodology to be used for their assessment. In the US, the Council on Environmental Quality (CEQ) adopted successive and detailed guidance documents advising national agencies on how to consider GHG emissions and climate change when implementing NEPA.Footnote ³³ In 2014, the European Union (EU) revised its EIA Directive to require Member States to consider ‘the impact of [a] project on climate’ through an assessment of ‘the nature and magnitude of greenhouse gas emissions’.Footnote ³⁴ The European Commission adopted guidance documents,Footnote ³⁵ although not as detailed and specific as the CEQ guidance. An express statutory CEA requirement has also been adopted in several other jurisdictions, including California (US),Footnote ³⁶ Mexico,Footnote ³⁷ New South Wales (Australia),Footnote ³⁸ Norway,Footnote ³⁹ and Quebec (Canada).Footnote ⁴⁰

Non-state actors have also implemented CEA requirements. In particular, the World Bank has long required the assessment of ‘global environmental aspects’ of planned action, including on climate change.Footnote ⁴¹ Other multilateral development banks have progressively adopted comparable rules.Footnote ⁴² In 2020, a CEA requirement was also included in a new edition of the Equator Principles,Footnote ⁴³ a document adopted by over 100 private financial institutions as a pledge to assess the environmental and social risk arising from the projects they finance, in particular, in countries with lax environmental law or poor enforcement capability.Footnote ⁴⁴

2.2. Resistance and Opposition

While EIA is widely applied as a tool for climate change mitigation, it is not a universal practice. A few jurisdictions, like Singapore, do not have a mandatory EIA requirement. Others that do have a mandatory EIA did not extend it as a CEA. This is often related to administrative lethargy and the poor enforcement of EIA instruments, particularly in developing countries. Thus, despite the above-mentioned decision of the National Green Tribunal,Footnote ⁴⁵ Indian authorities generally do not appear to require an assessment of GHG emissions as part of the EIA for relevant projects.Footnote ⁴⁶ In contrast, some other jurisdictions have made a deliberate choice to exclude climate effects from the scope of their EIA framework.

In KazakhstanFootnote ⁴⁷ and, until 2022, in New Zealand,Footnote ⁴⁸ the exclusion of climate effects from EIA was influenced largely by concerns that CEA would be redundant with cap-and-trade mechanisms. In principle, any policy or measure that reduces GHG emissions within the scope of a cap-and-trade mechanism would only cause emission allowances to be sold to and, used by, other actors, with no additional mitigation outcome overall – a phenomenon often referred to as a ‘waterbed effect’.Footnote ⁴⁹ Yet it is noteworthy that some jurisdictions, such as California and the EU, have adopted a CEA requirement in spite of having a cap-and-trade mechanism already in place.Footnote ⁵⁰ Some mechanisms have been devised to adjust the cap of a cap-and-trade mechanism, with the view of ‘puncturing’ the waterbed.Footnote ⁵¹ These include allowing national authorities to limit the distribution of emission allowances or even purchase emission allowances already on the market. Overall, many – perhaps most – of a project’s GHG emissions would take place beyond the sectorial, temporal or geographical scope of a national cap-and-trade mechanism. For instance, New Zealand’s livestock emissions are (still) not included in its cap-and-trade mechanism.Footnote ⁵² As such, while avoiding a waterbed effect should be an important consideration in the design of a cap-and-trade mechanism, this effect is not a convincing reason to exclude the application of EIA as a tool for climate change mitigation.

A more frequent ground for resistance and opposition to CEA relates to the abstract nature of climate effects. Opponents of a CEA requirement have pointed out that, by themselves, one project’s GHG emissions would not directly cause any tangible effect on any specific communities or resources.Footnote ⁵³ Yet scientists suggest that any incremental GHG emissions exacerbate climate change in a near-linear fashion.Footnote ⁵⁴ Each billion tonnes of CO₂ is expected to increase global average temperature by approximately 0.0005°C, thereby amplifying a wide range of associated risks.Footnote ⁵⁵ Thus, the US Supreme Court rightly noted the importance of considering even ‘a small incremental step’ towards addressing climate change.Footnote ⁵⁶

The opponents of a CEA requirement may further argue that an EIA framework does not, or should not, allow consideration of the extraterritorial effects of a project’s GHG emissions. Thus, the Montana Environmental Policy Act (MEPA) formally excludes consideration of impacts unfolding ‘beyond Montana’s borders’, especially those that are ‘global in nature’.Footnote ⁵⁷ Nonetheless, a state court held that an agency still had a duty to consider ‘the greenhouse gas effects of [the] project as it relates to impacts within the Montana borders’.Footnote ⁵⁸ This judgment is based on questionable grounds, as only an insignificant fraction of a project’s climate effects would unfold within Montana.Footnote ⁵⁹ On the other hand, by excluding any consideration of extraterritorial effects, the state legislation may appear incompatible with its international due diligence obligation related to climate change mitigation. In particular, it may conflict with what some refer to as a principle of non-discrimination in environmental protection: the notion that a state should not give less weight to environmental impacts in its decision-making process on the sole ground that these impacts would unfold beyond its borders.Footnote ⁶⁰

A more fundamental concern, however, relates to the effectiveness of EIA in addressing diffuse effects. If one approaches EIA merely as a means to mediate between project proponents and affected communities, its prospects as a tool for climate change mitigation are significantly limited by the absence of any community directly affected by the project’s GHG emissions. Yet richer understandings of EIA are possible, including as a means to prompt stakeholders to ‘move beyond strict self-interest … and thus [to] make decisions based on the common good’.Footnote ⁶¹ Indeed, while CEA does not prevent many GHG-intensive projects from being implemented,Footnote ⁶² this procedure can sometimes make a difference, albeit at the margins.Footnote ⁶³ At times, increased scrutiny of the project’s climate effects leads to an erosion of support for the project.Footnote ⁶⁴ In other instances, a CEA may result in an administrative decision not to approve the project,Footnote ⁶⁵ or to approve it subject to conditions aimed at reducing its GHG emissionsFootnote ⁶⁶ or at eliminating them altogether.Footnote ⁶⁷ While rejection decisions could result in carbon leakage and other undesirable effects,Footnote ⁶⁸ decisions approving activities with conditions are almost certain to achieve genuine mitigation outcomes when these conditions are effective and properly implemented.

3.1. The Problem

One of the central functions of any EIA framework is to determine whether a project would have significant environmental impacts.Footnote ⁶⁹ A preliminary determination of likely significance takes place at the screening and scoping stages, when the authorities decide on the need for and scope of an EIA procedure.Footnote ⁷⁰ This is followed by a substantive determination of significance, informed by scientific research and public deliberations, as part of the decision on the merits of the project.Footnote ⁷¹ The finding that a project would have significant impacts does not typically preclude its approval, but it may result in additional legal and political scrutiny.Footnote ⁷²

In EIA law and practice, the issue of significance has often been raised in litigation against an agency’s determination that a project would have no significant climate effect, based on the argument that it represents only a small fraction of global GHG emissions. These cases sometimes involved very large fossil-fuel projectsFootnote ⁷³ and, at other times, far more anodyne projects, such as road construction.Footnote ⁷⁴ Further, the significance of a project’s climate effect is also relevant in determining the need for a CEA, whether alongside the assessment of other potentially significant environmental impacts or not. Thus, in the US, Hawaii,Footnote ⁷⁵ Minnesota,Footnote ⁷⁶ and New York CityFootnote ⁷⁷ have decided to consider a project’s energy consumption as a source of GHG emissions, in deciding whether a project must undertake an EIA. This measure could potentially extend EIA requirements to various projects, including data centres (for example, crypto-mining and AI facilities)Footnote ⁷⁸ and healthcare facilities.

The EIA literature acknowledges that there is ‘no single agreed method for determining significance’.Footnote ⁷⁹ Whether an impact is deemed significant depends not only on an objective prediction of its magnitude, but also on a value-based threshold that ‘separates the realm of the acceptable from the realm of the unacceptable’.Footnote ⁸⁰ Most environmental impacts considered in EIA are direct and concrete, making it relatively easy for observers to agree on their significance. A project’s impact would be conspicuously significant if it were to pose a real threat to human life, have a far-reaching impact on ecologically sensitive areas, or inhibit the attainment of statutory environmental quality objectives. In contrast, a project’s GHG emissions have a far less direct and concrete effect on the environment and society. Even the largest GHG-emitting projects – some of which exceed a billion tonnes CO₂ in lifetime emissionsFootnote ⁸¹ – contribute only a tiny fraction to cumulative anthropogenic GHG emissions (2 trillion tonnes CO₂).Footnote ⁸² As a result, these projects lead to an extremely small increase in global average temperature,Footnote ⁸³ without any direct and predictable effect on any specific ecosystems, areas or communities.Footnote ⁸⁴

In the debate on whether climate effects should nonetheless be assessed, these effects are often characterized as the archetype of a ‘cumulative impact’.Footnote ⁸⁵ Indeed, climate effects ‘result from the incremental effects of the action when added to the effects of other past, present, and reasonably foreseeable actions’, in a context of ‘individually minor but collectively significant actions taking place over a period of time’.Footnote ⁸⁶ This characterization, however, provides little guidance on how the significance of climate effects is to be determined. Many EIA frameworks recognize the need to account for a project’s cumulative impacts, but it is not clear precisely what that means. EIA scholars recognize cumulative impact assessment as ‘one of the most challenging and least successful components of impact assessments’.Footnote ⁸⁷

3.2. Two Radical Approaches

To assess the significance of a project’s cumulative impacts, including its climate effects, decision makers tend to rely on one of two radical and unsatisfactory approaches.

On the one hand, a radically inclusionary approach suggests that, once cumulative impact is deemed significant, any contribution to it should also be considered significant. In the words of Nelson and Shirley, ‘causing minor harm to an already degraded environment should automatically be considered significant’.Footnote ⁸⁸ As climate change is clearly a significant issue, the argument goes, ‘we cannot afford to ignore even modest contributions to global warming’.Footnote ⁸⁹

However, this approach becomes entirely impractical if it implies that the climate effect of any project, however innocuous it might be, should be deemed significant. Undertaking a thorough assessment of the climate effects before carrying out even the most de minimis activities would surely not be a proper use of public or private resources.

On the other hand, the radically exclusionary approach suggests that it is only when the project’s effects are significant by themselves that they ought to be considered significant overall. Thus, according to the Californian regulator, ‘[t]he mere existence of significant cumulative impacts caused by other projects alone shall not constitute substantial evidence that the proposed project’s incremental effects are cumulatively considerable’.Footnote ⁹⁰ This approach would result in virtually no project being considered to have a significant climate effect, as almost no project would produce any discernible change on a global scale by itself.Footnote ⁹¹

Yet this latter approach ignores the fact that, as stated by a US Court of Appeal, ‘a very small portion of a gargantuan source’ of pollution could still constitute a very significant problem in its own terms.Footnote ⁹² Activities that emit large amounts of GHGs can be said to have a significant ‘climate effect’, if only in an abstract and statistical sense of the word ‘effect’. Taken individually, they cause an extremely minor exacerbation of an extraordinarily broad range of risks. Society recognizes the significance of GHG emissions that do not directly cause any discernible harm, in particular by taking action to avoid such emissions, for instance, by closing coal plants. If society is willing to adopt various measures to avoid millions of tonnes of CO₂ emissions, then it seems reasonable for projects with similar levels of emissions to be subjected to some form of scrutiny.

3.3. Alternative Approaches

Going beyond the radical approaches documented above, EIA law and practice have identified three more promising ways to assess the significance of a project’s climate impact: (i) thresholds, (ii) benchmarks, and (iii) economic valuation. Firstly, thresholds of magnitude may facilitate a preliminary determination of significance at the stages of screening and scoping. Several jurisdictions have adopted clear and specific thresholds. Many jurisdictions, for instance, have found that a project’s climate effect ought to be thoroughly assessed if it exceeds a threshold set between 20 and 200 kilotonnes of CO₂ equivalent (kt CO₂e).Footnote ⁹³ These thresholds have obvious limitations; for instance, they reduce significance to a mono-dimensional approach to magnitude, such as annual emissions, without accounting for the project’s duration. Yet if any CEA is to be implemented at all, a line needs to be drawn between those projects the climate effects of which are or are not to be assessed. The absence of an explicit threshold has only forced agencies and courts to fill the gap in a less predictable way.Footnote ⁹⁴ As such, explicit thresholds of magnitude play a useful role in advancing legal certainty at the scoping and screening stages.Footnote ⁹⁵ On the other hand, this rudimentary approach cannot be usefully relied upon for the final determination of significance at the substantive decision-making stage if this means that a quantity of smaller projects would be preferred over a few larger ones.

Secondly, benchmarks can be used to assess the significance of a project’s climate effect, including at the substantive decision-making stage. Benchmarks do not provide a purely objective test for the assessment of significance of climate effects;Footnote ⁹⁶ nonetheless, they can inform political deliberations by situating a project’s GHG emissions in a relevant context.Footnote ⁹⁷ Several benchmarks can be used in combination, including those that are empirical (e.g., observed levels of GHG emissions),Footnote ⁹⁸ predictive (e.g., the projected level of GHG emissions at a given point in time),Footnote ⁹⁹ and normative (e.g., a carbon budget).Footnote ¹⁰⁰ Scale matters greatly in the selection of relevant benchmarks.Footnote ¹⁰¹ A project’s GHG emissions may appear negligible when compared with a global benchmark,Footnote ¹⁰² but its socio-economic benefits would also appear negligible at that same global scale. For instance, a power plant would make a negligible contribution to meeting global energy needs. An assessment of the merits of the project, which involves a comparison of its costs and benefits, is better informed if these are assessed at the same (local) scale, as the use of vastly different scales would make a comparison extremely difficult.Footnote ¹⁰³ Other potential benchmarks do not require the definition of a scale, but rather the selection of one or several relevant analogues. These sectorial benchmarks point, for instance, to a level of emissions intensity by similar facilities or to the use of the best available technology.Footnote ¹⁰⁴ The climate effect of a cement factory would appear perhaps less significant, or at least more justifiable, if the factory were to emit fewer GHG emissions per tonne of cement compared with other cement factories.

Thirdly, economic valuation tools may also usefully, and perhaps more conveniently, inform the determination of the significance of a project’s climate effects. For instance, a US interagency working group has developed estimates of the social cost of GHG emissions (IWG-SCGHG).Footnote ¹⁰⁵ This tool has been widely used in the US at federalFootnote ¹⁰⁶ and stateFootnote ¹⁰⁷ levels, as well as in several other jurisdictions.Footnote ¹⁰⁸ As the CEQ points out, such economic valuation tools ‘allow … for comparisons to other monetized values, and estimates the damages associated with GHG emissions over time’ in such a way as to ‘assist agencies and the public in assessing the significance of climate impacts’.Footnote ¹⁰⁹ Yet a major difficulty for society is to agree on the value of GHG emissions in the first place. The IWG-SCGHG relies on an economic valuation of the global harm caused by GHG emissions,Footnote ¹¹⁰ an approach that necessarily involves far-reaching judgements about the economic value of non-economic damage and, even more importantly, the discount rate applicable to harm that would unfold in the distant future.Footnote ¹¹¹ Alternative valuation techniques involve an observation of the price that society appears to be willing to pay to avoid GHG emissions. This observation is based on the cost of the measures that are implemented and their predicted mitigation outcomes.Footnote ¹¹² Another approach is to determine the price that needs to be imposed on GHG emissions to ensure the achievement of an emissions-reduction goal, such as the temperature goals of the Paris Agreement.Footnote ¹¹³ Despite major differences in their methodologies and assumptions, institutional studies using these various approaches to economic valuation tend to converge towards a relatively consistent range, typically between US$50 and 380 per tonne CO₂.Footnote ¹¹⁴

There is no single best approach to the determination of the significance of a project’s climate effect. In many cases, a CEA objective of informing the public and decision makers would be most effectively achieved by a combination of these three approaches. Clear and specific thresholds of magnitude can assist in a swift decision as to the need for a thorough assessment of a project’s climate effects. Benchmarks and valuation tools can then inform the public and decision makers about the significance of a predicted level of GHG emissions and, eventually, whether to approve the project. While relying on a single approach runs the risk of presenting the assessment of a project’s climate effect as an ‘objective’ and technical question best left to experts, a combination of these three approaches better reflects the value-based judgements involved in a determination of significance.

4.1. Market Substitution Argument

One common argument against the assessment of indirect GHG emissions is that these emissions, being driven by market forces, would occur regardless of whether the project is approved. For instance, the Land Court of Queensland (Australia) repeatedly found that rejecting coal mine projects would ‘have no impact on climate change because it [would] have no impact on the global demand for coal and therefore no impact on global GHG emissions’.Footnote ¹²⁵ In other words, a new coal mine would simply substitute for coal production from existing coal mines elsewhere, with no net impact on global coal consumption. This market substitution argument can apparently be made with regard to various types of indirect GHG emission, provided only that they take place in a competitive environment. For instance, a new cheese-making factory would simply substitute for other cheese-making factories to meet demand from cheese consumers.Footnote ¹²⁶ This argument is generally made to exclude the assessment of indirect GHG emissions, but it could just as well be deployed in relation to direct climate effects, such as fugitive emissions associated with fossil-fuel production projects.

One objection to the market substitution argument is grounded in moral philosophy. Authors and judges have opposed what they call ‘the drug dealers’ defence’, suggesting that ‘[t]here are few other contexts where a harmful behaviour or action is excused purely because another entity would have otherwise caused the harm’.Footnote ¹²⁷ This objection is supported by non-consequentialist ethical theories, which suggest, for instance, that states ‘have a duty to reduce their carbon footprint irrespective of what others do’,Footnote ¹²⁸ or otherwise must refrain from ‘profit[ing] from an injustice’.Footnote ¹²⁹ By contrast, the market substitution argument is plainly supported by consequentialist ethical theories: if the moral value of one’s action is determined by the consequence, nothing is wrong with an action without consequences.Footnote ¹³⁰ Therefore, whether one is persuaded by the moral objection to the market substitution argument – assuming that the argument is economically valid – depends entirely on the ethical theory on which one relies.

A more compelling objection to the market substitution argument challenges its economic foundation.Footnote ¹³¹ To justify the exclusion of indirect (or direct) climate effects, the market substitution argument needs not only to show that market substitution occurs, but also that it is (at least nearly) ‘perfect’. This means that the argument needs to show that there would be ‘the same amount of GHG emissions’ with or without the project.Footnote ¹³² In reality, while some market substitution is likely to occur in relation to activities that take place in a competitive environment, it is unlikely to be ‘perfect’.Footnote ¹³³ For instance, the coal produced from a new mine would be likely to replace some of the coal from existing mines, but it would also drive prices down – if only slightly – by increasing competition in the relevant coal market, thus incentivizing higher consumption.Footnote ¹³⁴ Perfect market substitution would occur only if either the price elasticity of supply was infinite (that is, if other producers could increase production without any price increase) or the price elasticity of demand was null (if consumers would buy the exact same quantity notwithstanding any price increase). However, either of these hypotheses goes against the basic tenets of economic theoryFootnote ¹³⁵ as well as empirical research on price elasticities, including in the energy sector.Footnote ¹³⁶ As such, the CEQ rightly recommends that agencies ‘should not simply assume that if the federal action does not take place, another action will perfectly substitute for it and generate identical emissions’.Footnote ¹³⁷ In aggregate, the creation of new fossil-fuel projects is instrumental to the increase in global GHG emissions from the combustion of fossil fuels.

A complicating factor is that, to the extent that market substitution does occur, it may not be climate-neutral. For instance, substituting coal from existing mines with coal from a new mine may have important implications for downstream emissions if the substitute coal is of a different quality, with a different energy or carbon intensity.Footnote ¹³⁸ A fortiori, a new natural gas well may have some climate benefits if it leads to substitution of natural gas for coal.Footnote ¹³⁹ It is not entirely inconceivable that such intermodal substitution could balance the climate effect of the additional emissions from increased fossil-fuel consumption.Footnote ¹⁴⁰ Even if the fossil fuel provided is of the same nature and quality, project-specific circumstances, such as the measures taken to reduce fugitive emissions, may affect the emissions associated with their production and distribution.Footnote ¹⁴¹ These observations have led the Supreme Court of Norway to decry that assessing indirect GHG emissions is ‘complicated and controversial’.Footnote ¹⁴² Yet it is precisely because assessing these indirect climate effects is difficult that EIA can play a useful role in informing political deliberations.

4.2. Extraterritoriality Argument

Another frequent concern with indirect climate effects is that they may result from activities conducted in another jurisdiction. For instance, the downstream emissions of a coal mine may be associated with coal consumption overseas.Footnote ¹⁴³ In such circumstances, it has been argued, a CEA should not, or could not, consider indirect climate effects.

An initial question is whether a state has the right to take measures aimed at reducing indirect GHG emissions when these emissions would unfold overseas. In this regard, the Lotus principle establishes that, while a state ‘may not exercise its power … in the territory of another State’, it may nonetheless ‘exercis[e] jurisdiction in its own territory … in respect of any case which relates to acts which have taken place abroad’.Footnote ¹⁴⁴ Accordingly, as far as international law is concerned, a state is allowed to take into account extraterritorial effects when regulating activities occurring within its territory.Footnote ¹⁴⁵

Courts have sometimes postulated that climate treaties establish a ‘division of responsibilities between states’, including a ‘clear principle’ according to which ‘each state is responsible for’ limiting and reducing GHG emissions within its own territory.Footnote ¹⁴⁶ This conventional division of responsibilities, however, exists mainly in relation to states’ obligations to inventory GHG emissions, as a way to avoid omissions and double-counting.Footnote ¹⁴⁷ Regarding substantive obligations, neither the UN Framework Convention on Climate ChangeFootnote ¹⁴⁸ nor the Paris AgreementFootnote ¹⁴⁹ limit a state’s mitigation action to its own territory.Footnote ¹⁵⁰ States have occasionally taken steps,Footnote ¹⁵¹ or formally committed to actions,Footnote ¹⁵² aimed at reducing extraterritorial GHG emissions. Some courts have gone as far as to suggest that states have a duty to reduce emissions overseas whenever possible.Footnote ¹⁵³

A more complex issue relates to the legitimacy for a national government to assess, through a national EIA, the merits of activities that occur overseas. For instance, in determining whether a gas well project should be approved, an EIA may need to consider whether the use of the natural gas in another country is justified. This assessment would need to consider not only national and global climate goals but also the other state’s own development objectives and national circumstances.Footnote ¹⁵⁴ In considering the effect of a project on extraterritorial activities, a national EIA would inevitably involve some interaction with the internal affairs of another state. Yet such interaction can hardly be avoided between states that are trade partners. The decision not to consider a project’s extraterritorial indirect climate effects would equally result in some interaction in the domestic affairs of other states. For instance, the plentiful availability of cheap fossil fuels on global markets would make it politically more challenging for governments to limit their national GHG emissions.

4.3. Uncertainties Argument

A more fundamental concern relates to the very possibility of assessing indirect climate effects. Judges have found that such effects are ‘incapable of measurement or assessment’,Footnote ¹⁵⁵ or are at least not ‘practicable’.Footnote ¹⁵⁶ Almost by definition, a project’s indirect climate effects depend in part on factors extraneous to the project, including economic, technological, regulatory, demographic, and cultural variables.Footnote ¹⁵⁷ The extent of this epistemological issue depends on the type of indirect effects under consideration. The GHG emissions resulting from the consumption of fossil fuels of a known quality are fairly predictable.Footnote ¹⁵⁸ In contrast, the consequences of market substitution for GHG emissions can be far more difficult to predict, in particular, when there is a risk of intermodal substitution, such as between natural gas and coal. A project’s effects can extend ad infinitum, like the waves created by a stone on a pond, to increasingly remote and unpredictable effects, including through technological development, public opinion, and democratic processes.

These epistemological concerns are real and well founded, but they do not justify doing away with an assessment of indirect climate effects that are potentially significant. A general principle in EIA law and practice is that ‘[e]nvironmental effects are not to be disregarded merely because they are difficult to identify and quantify’.Footnote ¹⁵⁹ Ignoring indirect (or other) climate effects merely because they are difficult to predict would result in biased information about the merits of a project. This, in turn, could mislead the public and the decision makers, ultimately undermining the purpose of an EIA. Addressing uncertainties and providing the best information available are among the central functions of EIAs.Footnote ¹⁶⁰ EIAs routinely deal with uncertainties by first seeking to reduce them as far as possible and then by disclosing them in a clear and transparent manner.Footnote ¹⁶¹

Methodological tools already exist to reduce uncertainties relating to indirect climate effects. GHG emissions accounting methodologies have been developed by the Intergovernmental Panel on Climate ChangeFootnote ¹⁶² and by the GHG Protocol.Footnote ¹⁶³ These documents include some suggestions on the way in which one might deal with certain sources of uncertainty, such as the lack of available data.Footnote ¹⁶⁴ They highlight the need to define ‘key’ sources of emissions,Footnote ¹⁶⁵ including ‘significant secondary effects’,Footnote ¹⁶⁶ and to subject them to more intense scrutiny. This is in agreement with EIA law and practice. For instance, the interpretation of NEPA requires agencies to ‘discuss impacts in proportion to their significance’,Footnote ¹⁶⁷ which calls for a thorough discussion of the downstream emissions of fossil-fuel projects,Footnote ¹⁶⁸ while justifying a more succinct treatment of effects that ‘are so remote as to be indiscernible’.Footnote ¹⁶⁹

Further, EIAs must clearly communicate the existence of uncertainties to the public and decision makers. GHG emissions accounting tools note that ‘[u]ncertainty estimates are an essential element of a complete inventory of greenhouse gas emissions and removals’.Footnote ¹⁷⁰ Consistently, the CEQ recommends that CEAs ‘explain any uncertainty’ when quantifying direct and indirect climate effects.Footnote ¹⁷¹ US courts have further noted that a clear acknowledgement of uncertainties would allow readers to take ‘the resulting estimates with the appropriate amount of salt’.Footnote ¹⁷² With any feasible prediction and a clear acknowledgement of remaining uncertainties, the public and decision makers would be better equipped to make what they consider to be the appropriate decision regarding the merits of the project.