A. EA Procedures as a General Practice

EA, as defined by the IAIA, is ‘the process of identifying, predicting, evaluating and mitigating the bio-physical, social and other relevant effects of development proposals prior to major decisions being taken and commitments made’.Footnote ²⁴ The UN Environment Programme (UNEP) describes EA as involving ‘an examination, analysis and assessment of planned activities with a view to ensuring environmentally sound and sustainable development’.Footnote ²⁵ EA typically involves an expert assessment and public consultations before a decision to authorize the proposed activity can be made, possibly with conditions.

The origins of this tool are often traced back to the US National Environmental Policy Act (NEPA) of 1969, which directs federal agencies proposing any ‘major Federal actions’ capable of having a significant impact on the environment to conduct public consultation based on a ‘detailed statement’Footnote ²⁶ of this possible impact. This statement documents the likelihood and the nature of the impact, but also possible alterations or alternatives to the proposed activity which could mitigate the impact. The NEPA review is an ‘ostensibly procedural commitment’,Footnote ²⁷ which, as Judge Stevens noted, ‘merely prohibits uninformed – rather than unwise – agency action’.Footnote ²⁸ Its rationale is that a clear scientific assessment of the proposed activity and public scrutiny could permit balanced consideration for environmental concerns in sensitive activities.

In the half century since the adoption of NEPA, the idea that an EA should be carried out before a project is approved spread to US statesFootnote ²⁹ and, beyond, to most countries, often in the form of a statutory procedure.Footnote ³⁰ Although the trend is clearer in developed countries,Footnote ³¹ most developing countries have also established EA laws,Footnote ³² although implementation is sometimes incomplete.Footnote ³³ The relevance of EA was recognized in the World Charter for Nature of 1982Footnote ³⁴ and in the Rio Declaration on Environment and Development of 1992;Footnote ³⁵ it is also affirmed in the draft of a Global Compact for the Environment currently under negotiations.Footnote ³⁶

A distinction has been made between project-based EA, also called ‘Environmental Impact Assessment’ (EIA), and ‘Strategic Environmental Assessment’ (SEA) conducted in relation to policies, plans and programmes well before particular projects are developed. While NEPA applies the same EA procedure to projects and strategies, distinct procedures have been established in some jurisdictions. In the EU, for instance, SEA procedure is established under a different directive,Footnote ³⁷ while China's EA Law establishes EIA and SE procedures in two different chapters.Footnote ³⁸

B. The Recognition of EAs in a Transboundary Context

With public consultations as a central feature,Footnote ³⁹ EA procedures were first conceived of within the boundaries of a political community and did not originally apply to transboundary environmental impacts.Footnote ⁴⁰ On the other hand, both the Stockholm Declaration on the Human Environment and the Rio Declaration on Environment and Development restated that States have ‘the responsibility to ensure that activities within their jurisdiction or control do not cause damage to the environment of other States or of areas beyond the limits of national jurisdiction’.Footnote ⁴¹ International courts and tribunals recognized the so-called ‘no-harm principle’ as customary international law.Footnote ⁴²

The no-harm principle requires States to implement reasonably appropriate measures or ‘due diligence’.Footnote ⁴³ At the very least, this due diligence obligation requires that a State does no less effort to avoid transboundary environmental harm than it does to avoid environmental harm within its own territory.Footnote ⁴⁴ This means that, if a State has a statutory EA procedure that seeks to avoid or reduce harm, this procedure should apply equally whether the harm would occur within its territory or beyond its borders. More broadly, EA could be approached as a general requirement implied by the due diligence obligation of States to ensure that activities carried out within their jurisdiction do not cause transboundary environmental harm, whether or not a State has such a requirement to avoid or reduce environmental harm within its own territory.Footnote ⁴⁵

UNEP has actively promoted the application of EA procedures to activities likely to cause transboundary environmental harms. In 1978, the agency adopted a guidance on the conservation and harmonious utilization of shared natural resources where it suggested that ‘States should make environmental assessment before engaging in any activity with respect to a shared natural resource which may create a risk of significantly affecting the environment of another State or States sharing that resource.’Footnote ⁴⁶ In 1987, the Goals and Principles of Environmental Impact Assessment called upon States to:

Since the 1980s, provisions were included in issue-specific treaties which reflected the growing recognition of the obligation of State to conduct EAs in relation to activities likely to cause transboundary environmental harm. In particular, the UN Convention on the Law of the Sea (UNCLOS) requires States, when planning activities which could cause significant harm to the marine environment, to ‘assess the potential effects of such activities on the marine environment’ and to ‘communicate reports of the results of such assessments’ to other States via competent international organizations.Footnote ⁴⁸ The Protocol on Environmental Protection to the Antarctic Treaty of 1991 contains detailed provisions on the EA which is to be conducted before any impactful activity can be undertaken in the Antarctic Treaty Area, including the circulation of information to all Parties and to a Committee for Environmental Protection.Footnote ⁴⁹ A year later, a provision was also inserted in the Convention on Biological Diversity, calling upon States, ‘as far as possible and as appropriate’, to introduce EIA procedures for projects likely to have significant adverse effects on biological diversity.Footnote ⁵⁰

Furthermore, in 1991, the Member States of the UN Economic Commission for Europe (UNECE) adopted a dedicated treaty on EIAs in a transboundary context in Espoo, Finland.Footnote ⁵¹ The Espoo Convention provides that the ‘Party of origin’ of certain projects must notify the ‘affected Party’ if the project is likely to cause a significant adverse transboundary impact, leaving to the latter the option to participate in an EIA procedure that the former must conduct before authorizing the project.Footnote ⁵² The procedure must consist in the preparation of the EIA documentation, where the affected Party may be requested to provide information, followed by a phase of consultations, leading to a decision by the Party of origin.Footnote ⁵³ While the Espoo Convention was originally limited to the Member of the UNECE, steps have been taken towards allowing accession by other UN Member States.Footnote ⁵⁴

While the Espoo Convention only imposed the conduct of an EIA, it encouraged the Parties, ‘[t]o the extent appropriate’, to ‘endeavour to apply the principles of environmental impact assessment to policies, plans and programmes’.Footnote ⁵⁵ Further negotiations led to the adoption of the Kiev Protocol on SEA in 2003.Footnote ⁵⁶ The Kiev Protocol requires that States conduct a SEA for certain plans and programmes.Footnote ⁵⁷ It outlines a procedure that States must follow whether or not the plan or programme is likely to have such transboundary effects, including the screening plans and programmes,Footnote ⁵⁸ the scoping and preparation of environmental reports,Footnote ⁵⁹ and public participation.Footnote ⁶⁰ It also defines a procedure for transboundary consultationsFootnote ⁶¹ and for the participation of the public of the affected countryFootnote ⁶² applicable when a plan or programme is likely to produce significant transboundary environmental effects. Participation to the Kiev Protocol is opened to all UN Member States,Footnote ⁶³ although only UNECE Member States have ratified it to date.

The Espoo Convention and its Kiev Protocol contributed, along with UNCLOS, the Protocol on Environmental Protection to the Antarctic Treaty and the Convention on Biological Convention, to the crystallization of a new norm of customary international law:Footnote ⁶⁴ the obligation of States to conduct an EA before approving a proposed activity likely to cause transboundary environmental harm. This new obligation was recognized by the ILC as early as 2001 through the adoption of the Draft Articles on the Prevention of Transboundary Harms from Hazardous Activities, stating that the decision to authorize an activity within the scope of the Draft Articles ‘shall … be based on an assessment of the possible transboundary harm caused by that activity, including any environmental impact assessment’.Footnote ⁶⁵

This view was later confirmed by the ICJ. In its 2010 judgment in Pulp Mills, the ICJ recognized the existence of ‘a requirement under general international law to undertake an environmental impact assessment where there is a risk that the proposed industrial activity may have a significant adverse impact in a transboundary context, in particular, on a shared resource’.Footnote ⁶⁶ The Court added that ‘due diligence, and the duty of vigilance and prevention which it implies, would not be considered to have been exercised’ if a State did not undertake an EIA before approving a project with liable to affect transboundary environmental resources.Footnote ⁶⁷ Yet, the Court stopped short from specifying what this procedure should contain;Footnote ⁶⁸ in particular, it rejected Argentina's claim that Uruguay had an obligation to consult the Argentinian population likely to be affected by the proposed activity.Footnote ⁶⁹

In its 2015 judgment in Certain Activities, the ICJ noted again the existence of ‘an obligation to conduct an environmental impact assessment concerning activities carried out within a State's jurisdiction that risk causing significant harm to other States’.Footnote ⁷⁰ It stated that this obligation was applicable ‘generally to proposed activities which may have a significant adverse impact in a transboundary context’.Footnote ⁷¹ The ICJ noted that, where an EA reveals an actual risk of significant transboundary harm, the State planning to undertake the activity is required ‘to notify and consult in good faith with the potentially affected States, where that is necessary to determine the appropriate measures to prevent or mitigate that risk’.Footnote ⁷² Overall, this judgment confirmed that compliance with the procedural obligation to conduct an EA is to be assessed separately from compliance with the substantive obligation not to cause transboundary environmental harm. In this case, the ICJ concluded that Costa Rica had violated its obligation to conduct an EA in a transboundary context even though no significant transboundary environmental harm had arisen from the realization of the project.Footnote ⁷³

C. The Significance of Existing Authorities to Climate Change

While Pulp Mills and Certain Activities regarded transboundary environmental harm affecting the territory of a third State, the obligation to conduct an EIA also applies in relation to activities which could affect areas beyond national jurisdiction. Treaty-based obligations to conduct an EA apply when harm could affect the AntarcticFootnote ⁷⁴ or the high seas,Footnote ⁷⁵ and the Stockholm Declaration on the Human Environment defined the no-harm principle as applying not only when damage affects the territory of another State, but also when it affects ‘areas beyond the limits of national jurisdiction’.Footnote ⁷⁶ In 2011, the Seabed Chamber of the International Tribunal on the Law of the Seas (ITLOS) in its Advisory Opinion on Activities in the Area recognized that the ICJ's reasoning in a transboundary context in Pulp Mills ‘may also apply to activities with an impact on the environment in an area beyond the limits of national jurisdiction’.Footnote ⁷⁷ EA is among the issues that the UN General Assembly decided to defer to the negotiations that it initiated in 2018 towards the adoption of an international legally binding instrument under UNCLOS on the conservation and sustainable use of marine biological diversity of areas beyond national jurisdiction.Footnote ⁷⁸

Likewise, it is relatively well accepted that the no-harm principle applies to ‘common concerns of humankind’Footnote ⁷⁹ such as climate change. The obligation for a State to ensure that activities conducted within its jurisdiction do not cause transboundary environmental harm suggests a fortiori an obligation to ensure that such activities do not cause harm to the global environment as a whole.Footnote ⁸⁰ Consistently, the Preamble of the UNFCCC,Footnote ⁸¹ the International Law Association's Draft Articles on the Legal Principles relating to Climate Change,Footnote ⁸² the ILC's Draft Guidelines on the Protection of the AtmosphereFootnote ⁸³ and academic analysisFootnote ⁸⁴ suggest that States have an obligation, under general international law, to prevent excessive GHG emissions within their jurisdiction or under their control.

By contrast, there is little authority regarding the application of the obligation to conduct an EA in relation to common concerns. The UNFCCC only defines a very general commitment for Parties to ‘[t]ake climate change considerations into account, to the extent feasible, in their relevant social, economic and environmental policies and actions’.Footnote ⁸⁵ The Convention on Biological Diversity goes arguably further, in relation to another common concern,Footnote ⁸⁶ by requiring that its Parties establish EIA procedures.Footnote ⁸⁷ None of these provisions, however, provides sufficient support, on its own, to the existence of a customary international law obligation to conduct an EA in relation to common concerns.Footnote ⁸⁸

Nevertheless, the ILC's Draft Guidelines on the Protection of the Atmosphere suggest that ‘States have the obligation to ensure that an environmental impact assessment is undertaken of proposed activities under their jurisdiction or control which are likely to cause significant adverse impact on the atmosphere in terms of … atmospheric degradation’Footnote ⁸⁹ such as climate change.Footnote ⁹⁰ The Commentary acknowledges the lack of relevant authority on the obligation to conduct an EA in a global environmental context, but it suggests that precedents developed in a transboundary context could apply ‘a fortiori to those activities potentially causing global atmospheric degradation’.Footnote ⁹¹ While the Draft Guidelines mention ‘intentional large-scale modification of the atmosphere’ (ie geo-engineering activities) as an example,Footnote ⁹² they do not take a clear position on the application of this obligation in relation to activities that involve unintentional and typically smaller-scale degradation of the atmosphere through GHG emissions.

The a fortiori argument which forms the core of the ILC's analysis is valid with regard to a general substantive obligation such as the no-harm principle: as States must take measures to prevent transboundary harm, it is to be expected that they should also a fortiori take measures to prevent atmospheric degradation, as atmospheric degradation is of a similar nature and generally of greater concern than transboundary harm.Footnote ⁹³ By contrast, this argument does not necessarily work when applied in relation to an instrumental obligation such as the obligation to conduct an EA: it all depends, then, upon the relevance of the tool in relation to the case of greater concern, that is, the relevance of EA as a tool to avoid or reduce the contribution of a proposed activity to global environmental harm. In other words, before concluding that the obligation to conduct an EA applies a fortiori in relation to atmospheric degradation, the ILC would need to determine whether and to what extent EA is an effective tool to address atmospheric degradation. As this determination relates to identification of customary international law, it should be done from the perspectives of States. As appears through the next sections, States have progressively accepted EA as a relevant tool for climate change mitigation.

A. A Growing State Practice

A general practice, for the purpose of asserting the existence of a norm of customary international law, does not need to be uniform. Rather, as interpreted by the ILC, it should consist in a practice which is ‘sufficiently widespread and representative, as well as consistent’.Footnote ⁹⁵ In most cases, original EA laws did not include specific provisions on climate change mitigation: whether the GHG emissions implied by a proposed activity had to be included in the EA was decided subsequently, through the interpretation of existing legal framework or sometimes their revision. Either way, developments have been slow, often incremental, yet generally steady and unidirectional. Today, the inclusion of consideration for GHG emissions in national EA procedures has become sufficiently widespread, representative and consistent to constitute a prevailing and, arguably, ‘general’ practice.

First of all, many EA laws require the assessment of all environmental impacts, in terms which are sufficiently broad to include GHG emissions.Footnote ⁹⁶ This enabled courts to play an important role in identifying and interpreting a CA requirement in existing EA frameworks. In particular:

• • In the United States, the Court of Appeal for the Ninth Circuit in Center for Biological Diversity v National Highway Traffic Safety Administration held that a preliminary NEPA assessment of a new fuel economy standard applicable to some vehicles should have assessed the impact of these standards on climate change.Footnote ⁹⁷

• • In the United Kingdom, the Queen's Bench Division confirmed that climate change was a relevant consideration before a local authority could consent to an increase in the capacity or in the number of flights at an airport.Footnote ⁹⁸

• • In New Zealand, the High Court held that a local government could consider the impact of a coal plant on climate change when deciding whether to authorize it.Footnote ⁹⁹

• • In South Africa, the High Court declared that the EIA for a coal-fired power plant should have included a climate change impact assessment.Footnote ¹⁰⁰

• • In Australia, the Civil and Administrative Tribunal of Victoria decided that the EIA of a coal power plant should document the GHG emissions which would result from its operations,Footnote ¹⁰¹ while the Land and Environment Court of New South Wales required the EIA of a coal mining project to include an assessment of the GHG emissions resulting from the end-use of the coal.Footnote ¹⁰²

• • In Austria, the Federal Administrative Court quashed the administrative approval for the construction of a third runway at Vienna's International Airport on the ground that the project's economic benefits did not outweigh its environmental impact, including on climate change.Footnote ¹⁰³ This judgment was overturned by the Constitutional Court which, however, agreed that the impact of the project on climate change had to be taken into account.Footnote ¹⁰⁴

Secondly, national governments have adopted guidelines and lawmakers have revised existing EA procedures in order to establish and clarify a CA requirement. In particular:

• • In the United States, following some initiatives at the state levelFootnote ¹⁰⁵ and previous hints at the Federal level,Footnote ¹⁰⁶ the Council on Environmental Quality (CEQ) issued draft guidance documents on consideration of the effects of climate change and GHG emissions in NEPA reviews in 1997,Footnote ¹⁰⁷ 2010Footnote ¹⁰⁸ and 2014,Footnote ¹⁰⁹ followed by a final guidance in 2016.Footnote ¹¹⁰ The final guidance was withdrawn by executive order in 2017,Footnote ¹¹¹ but courts have continued to rely on it,Footnote ¹¹² or else on the 2014 draft guidance,Footnote ¹¹³ as persuasive authorities.

• • In Canada, a ‘general guidance for practitioners’ was developed by a federal-provincial-territorial committee in 2003,Footnote ¹¹⁴ and related instruments have since been adopted at the provincial level.Footnote ¹¹⁵ Furthermore, at the time of writing this article, the Parliament of Canada was in the process of adopting a Government Bill recognizing ‘that impact assessment contributes to Canada's ability to meet its environmental obligations and its commitments in respect of climate change’.Footnote ¹¹⁶

• • In the EU,Footnote ¹¹⁷ the Commission has long held that GHG emissions should be documented as part of an EA.Footnote ¹¹⁸ At the occasion of a review on the implementation of the EU's EIA directive, however, the Commission noted that the effects of proposed activities on climate change were ‘not adequately identified and assessed within the EIA process’.Footnote ¹¹⁹ In 2013, the Commission adopted two guidance documents on integrating climate change and biodiversity into, respectively, EIA and SEA.Footnote ¹²⁰ The following year, the EIA directive was revised with the addition of a requirement that EIA reports document ‘the impact of the project on climate (for example the nature and magnitude of greenhouse gas emissions)’.Footnote ¹²¹

Thirdly, CA has percolated as a practice in some countries in the absence of any judicial precedent or formal legal development. For instance, China's 2002 EA Law and enforcing regulations contain no reference to climate change,Footnote ¹²² and there is no authoritative judicial decision on the matter. Nevertheless, two documents among the many issued by the Ministry of Environmental Protection (MEP) to guide the implementation of particular aspects of the EA act imply, almost accidentally, that China's EA should include a CA. One guidance document suggests that technical reviews of EIA reports should consider the feasibility and effectiveness of any measure described in the project which seeks to reduce GHG emissions.Footnote ¹²³ Another guidance document includes carbon dioxide (but not other GHGs) among the air pollutants to be documented in SEAs.Footnote ¹²⁴ In practice, it has been reported that about a fifth of SEA reports involve some sort of consideration for GHG emissions, although this rarely amounts to a systematic appraisal.Footnote ¹²⁵ Scholars have advocated for a reform of China's EA framework towards a more effective and systematic approach of CA.Footnote ¹²⁶

Fourthly, this general trend was reflected in the practice of international organizations.Footnote ¹²⁷ The World Bank, for instance, adopted internal rules in August 2016 which require an assessment of the GHG emissions as part of the environmental and social assessment of a project for which support is sought.Footnote ¹²⁸ This measure was included on the ground that ‘[e]stimation of project GHGs is part of good international industry practice’.Footnote ¹²⁹ Other international institutions with various memberships and voting powers have integrated climate change in their safeguard policies, including the Asian Development Bank,Footnote ¹³⁰ the Asian Infrastructure Investment BankFootnote ¹³¹ and the UNDP.Footnote ¹³² Consistently, the Equator Principles, which provide an industry benchmark for EA in financial institutions, require consideration for the GHG emissions of activities for which support is sought.Footnote ¹³³

Yet, while the inclusion of GHG in EA procedures is a clear trend, it is not a universal practice. A few States simply do not have a statutory EA framework at all.Footnote ¹³⁴ Other, developing States have a statutory framework but lack the resources to implement it consistently.Footnote ¹³⁵ In yet other jurisdictions, debates on the inclusion of considerations for GHG emissions in EAs do not seem to have sparked off yet. In India, for instance, the ‘Notification’ defining the regulatory framework for the federal EIA procedure leaves it to ad hoc Expert Appraisal Committees to determine the Terms of Reference of the EIA study, which is to ‘address … all relevant environmental concerns’.Footnote ¹³⁶ In practice, Terms of Reference generally do not include considerations for GHG emissions, even in the most relevant projects. A petition pending before the National Green Tribunal since 2017, but undecided as of February 2019, seeks to direct the Government to mainstream considerations for climate change mitigation in the EIA Notification.Footnote ¹³⁷ In Hong Kong, the scope of EIAs is, in practice, confined to a list of issues included in a ‘Technical Memorandum’, which does not include any proxy for the inclusion of transboundary or global environmental concerns.Footnote ¹³⁸ Here again, the question was never brought to a court, and it has rarely been raised in local debates.Footnote ¹³⁹ The absence of CA in these and other jurisdictions has more to do with political inertia, hesitancy or, at most, reluctance, than with a deliberate exclusion.

By contrast, at least two States have specifically excluded considerations for GHG emissions from domestic EA frameworks. In 2002, the Environmental Court of New Zealand expressed ‘considerable disquiet about the efficacy, appropriateness and reasonableness’ of a condition for GHG emission limitation or reduction that regional authorities had imposed on a gas-fired power plant.Footnote ¹⁴⁰ The Court considered that climate change mitigation should be carried out exclusively through national policies as only those could ‘guarantee an efficiency compatible with achieving best social, environmental and economic outcomes’.Footnote ¹⁴¹ Two years later, this approach was endorsed by a Government Bill amending the Resource Management Act to exclude considerations for ‘the effects on climate change of discharges into air of greenhouse gases’.Footnote ¹⁴² In the government's view, this amendment would avoid duplication of efforts and reduce administrative costs, as a national emissions trading scheme would ensure that New Zealand complies with its quantified emissions limitation and reduction commitment under the Kyoto Protocol.Footnote ¹⁴³ A similar development occurred in Kazakhstan where, in 2011, a statutory reform specifically excluded ‘the impact of greenhouse gas emissions’Footnote ¹⁴⁴ from the scope of the EAs as the country was establishing a carbon market.

Ironically, even where EAs do not document the adverse effects of GHG emissions, they tend to document the benefits of emission reductions where applicable. In New Zealand, the same amendment to the Resource Management Act allows consideration for the effects of GHG emissions in renewable energy projects to the extent that such project enables a reduction of such emissions.Footnote ¹⁴⁵ Likewise, in Hong Kong, reduction in GHG emissions was abundantly documented in the EIA reports relating to a small windfarm projectFootnote ¹⁴⁶ and to the conversion of a coal-fired power plant into gas-fired power plant with higher carbon efficiency.Footnote ¹⁴⁷ This selective inclusion of GHG emissions in EAs is inconsistent with EAs’ aim of providing complete and objective information. If decision makers in Hong Kong and New Zealand are to take into account the benefits of reductions in GHG emission in the assessment of some proposed activities, it is unclear why they should not also recognize the costs of additional GHG emissions in other proposed activities.

B. An Incomplete Acceptance as Law

Acceptance as law (opinio juris) distinguishes a custom from mere usage or habit.Footnote ¹⁴⁸ As interpreted by the ILC, this requirement is that ‘the practice in question must be undertaken with a sense of legal right or obligation’.Footnote ¹⁴⁹ It is notoriously difficult to establish this subjective element, if only because States are actually composed of many individuals with different beliefs and motivations.Footnote ¹⁵⁰ It is generally understood, as international courts and tribunals have held in successive cases,Footnote ¹⁵¹ that States have accepted as law an obligation to carry an EA at least in a transboundary context. The question here is whether States have accepted a similar obligation in the context of a global environmental harm, namely GHG emissions.

Treaties and their interpretation by States can be relevant to identifying acceptance as law, in particular when they are framed not merely as a trade-off between the interests of their parties, but rather as the recognition of a general norm.Footnote ¹⁵² Yet, treaties provide limited support to the acceptance of EA in the context of climate change as law. Climate change treaties, for instance, do not impose or recognize a mandatory EA procedure. To the contrary, the Preamble of the UNFCCC recognizes ‘the principle of sovereignty of States in international cooperation to address climate change’.Footnote ¹⁵³ Accordingly, the UNFCCC regime has generally been left it for States to determine how best to achieve mitigation outcomes.

Nevertheless, some provisions in climate change treaties could be taken as expression of a sense of obligation. Thus, the UNFCCC suggests that all Parties should ‘[t]ake climate change considerations into account, to the extent feasible, in their relevant social, economic and environmental policies and actions’.Footnote ¹⁵⁴ A similar language was echoed in the Sustainable Development Goals when the UN General Assembly recognized the need to ‘[i]ntegrate climate change measures into national policies, strategies and planning’.Footnote ¹⁵⁵ Other treaty provisions calling States to take relevant measures appear to imply an obligation to conduct CA if and inasmuch as it is an effective tool for climate change mitigation. Thus, Annex I Parties committed under the UNFCCC to ‘adopt national policies and take corresponding measures on the mitigation of climate change’,Footnote ¹⁵⁶ whereas all the Parties to the Paris Agreement are to ‘pursue domestic mitigation measures, with the aim of achieving the objectives’Footnote ¹⁵⁷ of their successive NDCs. On the other hand, no State appears to have mentioned EA in their NDC as a tool for climate change mitigation (although a few States highlighted the need to integrate consideration for climate change adaptation in EAs).Footnote ¹⁵⁸

Alternatively, the regime of the Espoo Convention could be relevant in assessing acceptance of CA as law. The Espoo Convention itself appears to focus on typical cases of transboundary harm confined to a geographical area. Although the convention includes ‘climate’ as one of the dimensions of a project's potential impacts,Footnote ¹⁵⁹ ‘transboundary impact’ is defined as any impact ‘not exclusively of a global nature’.Footnote ¹⁶⁰ Nonetheless, the Parties to the Espoo Convention are supportive to a broader vision of EIA which may extend to the context of climate change. The Guidance on the Practical Application of the Espoo Convention, adopted by its Parties, suggests, through somewhat of a stretch, that ‘activities with linkages to climate change’ may be amongst the ‘long-range transboundary impacts’ to be assessed under the Convention,Footnote ¹⁶¹ rather than impacts exclusively of a global nature excluded from its scope. By contrast to the Espoo Convention, the Kiev Protocol on SEA, adopted about a decade later, relates to activities likely to have significant environmental effects (including effects on the climate) notwithstanding whether these effects are local, transboundary or global in nature.Footnote ¹⁶²

Discussions on CA were held under the regime of the Espoo Convention, in particular at the 7th session of the Meeting of the Parties to the Convention and the 3rd session of the Meeting of the Parties to the Protocol held conjointly in Minsk in June 2017. The Convention and Protocol's Secretariat organized a high-level Panel discussion on the role of the two treaties in addressing climate change.Footnote ¹⁶³ A work plan, adopted at the same meetings, introduced measures to raise awareness, exchange experience and train professionals on the opportunity of integrating climate change in EAs, in particular SEAs.Footnote ¹⁶⁴ Overall, a meeting of the UNECE Member States at the same occasion outlined a common vision of SEA as ‘a key tool for the development of national climate change action and planning, and for the incorporation of specific climate change mitigation and adaptation measures into regional development and sectoral plans, programmes and policies’.Footnote ¹⁶⁵ Given the strong prevalence of CA in the practice of the UNECE Member States, further developments in this forum can be anticipated.Footnote ¹⁶⁶

Some other treaty provisions could also be taken to suggest acceptance of CA as a customary obligation. In particular, it is notable that UNCLOS's definition of pollution of the marine environment is wide enough to encompass some of the impacts of climate change on the seas, such as warming, acidification and sea-level rise.Footnote ¹⁶⁷ Accordingly, the obligation of States to assess the potential effects of activities which may cause substantial pollution to the marine environment under Article 206 of UNCLOS may be interpreted as requiring EAs of proposed activities likely to contribute significantly to climate change. Likewise, given the impact of climate change on ecosystems, the provision of the Convention on Biological Diversity on EIA could be interpreted as encompassing GHG emissions.Footnote ¹⁶⁸

However, rather than treaties, it is arguably the very conduct of States which provides the strongest evidence of a growing acceptance of EA as an obligation in the context of climate change.Footnote ¹⁶⁹ A distinction needs to be drawn between a conduct adopted by a State in pursuance of its own interests and that adopted in pursuance to what a State views as its duty, possibly against its immediate interests.Footnote ¹⁷⁰ States pursue no immediate interests by conducting EAs to avoid or reduce the impact of proposed activities on climate change. The adoption of CA procedures in various countries and international organizations cannot be explained by courtesy, comity, political expediency or convenience.Footnote ¹⁷¹ Efforts to comply with national commitments on climate change mitigation can also be excluded when the State has no quantified commitment,Footnote ¹⁷² or inasmuch as the GHG emissions which are the object of the CA take place beyond the State's territory (hence outside the geographical scope of the State's quantified commitment),Footnote ¹⁷³ or beyond the time horizon of any national commitment.Footnote ¹⁷⁴ Rather, States’ conduct generally implies their acceptance of CA as a duty.

A possible difficulty with the identification of an obligation to conduct a CA, as Neil Craik noted, is that of ‘determining to which States the duty is owed, and as a result, which States have a right to be notified and be consulted’.Footnote ¹⁷⁵ In a transboundary context, the obligation to conduct an EA is owed the State likely to be affected.Footnote ¹⁷⁶ In relation to a common concern such as climate change, where the impact would not specially affect any distinct area, population or State, the obligation to conduct an EA—like the more general obligation to prevent excessive GHG emissions—is as an obligation owed to the international community as a whole (erga omnes obligation),Footnote ¹⁷⁷ and any State has the right to seek the performance of this obligation.Footnote ¹⁷⁸ Yet, this does not necessarily mean that every State has a right to be notified and to be consulted, as the obligation to notify and to consult another State can inferred from the obligation to conduct an EA only ‘where that is necessary to determine the appropriate measures to prevent or mitigate’ the risk.Footnote ¹⁷⁹

To conclude, there appears to be a relatively well-established general practice of States and an increasing acceptance as law in support of a customary obligation to conduct a CA, though it is not clear whether this obligation has yet fully emerged. What continues to hinder a complete recognition of this norm is the difficulty of adapting EA to the particular nature of GHG emissions as a global cumulative environmental harm. Consequently, it is essential to determine whether EA is actually a relevant tool for climate change mitigation.

A. Assessing the Significance of the GHG Emissions Caused by a Proposed Activity

Perhaps the most obvious objection to the use of EA as a tool for climate change mitigation relates to the difficulty of asserting the significance of GHG emissions caused by a given activity.

The question of the significance of an activity's GHG emissions is typically raised in relation to the selection of activities that need to undergo a thorough assessment (‘screening’ phase of the EA) and to the identification of the issues to be documented (‘scoping’). At either stage, the finding that an activity could cause a significant environmental impact calls for additional steps to document this possible impact and consider revisions of the proposed activity to avoid or reduce this impact or additional mitigation measures. The magnitude of the GHG emissions likely to result from a proposed activity should determine the degree of scrutiny that the activity should undergo. Crucially, significance is also to be considered at the decision-making stage, where the impact of the proposed activity on the climate system is to be taken into account and weighed along with other costs and benefits.

Taken in isolation, a single activity is unlikely to have a measurable impact on planetary systems, as the increase in the GHG concentrations in the atmosphere results from innumerable sources scattered among and within countries. This ‘“drop in the ocean” problem’Footnote ¹⁸⁰ is most obvious with regard to (project-level) EIAs—and this may be one of the reasons why the Minsk Declaration adopted by the Member States of the UNECE promoted the role of SEA to mitigate climate change.Footnote ¹⁸¹ However, this problem surely concerns most SEAs on plans, programmes and policies as well. Only in a few exceptional cases could a policy be considered to have a more-or-less discernible impact on the entire climate system.Footnote ¹⁸²

The difficulty of assessing the significance of GHG emissions in a proposed activity is the most common ground for excluding consideration for GHG emissions from EA. An instance amongst many, the Federal Court of Australia in Anvil Hill noted that no significant causal relation could be established between the GHG emissions caused by a single coal mine and a measurable increase in global average temperature. On this ground, the Court concluded that the EA did not have to contemplate the impact of the project on the climate system.Footnote ¹⁸³

Likewise, in New Zealand, the statutory exclusion of GHG emission from the scope of EA has been interpreted on the ground that a single project's GHG emissions could never be significant: ‘given … the infinitesimal contribution which any particular project could make, there could be no demonstrable linkage between GHG emissions associated with any particular project and climate change generally’.Footnote ¹⁸⁴ But, as mentioned above, while a subsequent legislative reform endorsed this reasoning, it also permitted consideration for the benefits of emission reductions arising from renewable energy projects,Footnote ¹⁸⁵ even though the contribution of such projects to climate change mitigation is equally likely to be infinitesimal.

On the other side of the world, a debate arose in relation to the significance of the GHG emissions caused by the Kearl Oil Sands Project in the Canadian Province of Alberta. The 2005 EIA report estimated that the project's operations would cause 3.8 MtCO₂eq per year, representing respectively 1.7 per cent and 0.5 per cent of Alberta and Canada's GHG emissions.Footnote ¹⁸⁶ In 2007, a Joint Review Panel in charge of making recommendations to the responsible Federal agency under the relevant EA legislationFootnote ¹⁸⁷ estimated that this impact was not significant and recommended approval of the project.Footnote ¹⁸⁸ In a subsequent application for juridical review, the Federal Court remitted the matter to the same Panel with the direction of providing a rationale for its conclusion that the project's GHG emissions were not significant.Footnote ¹⁸⁹ This led the Joint Review Panel to adopt an addendum to its Report where it highlighted the lack of evidence that the project's GHG emissions would have any significant impact on the entire climate system.Footnote ¹⁹⁰ In June 2008, the relevant authorities approved the project based on the conclusion that it was ‘not likely to cause significant adverse environmental effects’.Footnote ¹⁹¹

The most convincing response to the ‘drop in the ocean’ problem is arguably to situate it in the context of a broader debate on cumulative effects. Many environmental issues do not occur as the result of any single action, but because of the incremental addition of small impacts by multiple actors.Footnote ¹⁹² Most EA frameworks require consideration for the activity's cumulative impact, that is, ‘the incremental impact of the action when added to other past, present, and reasonably foreseeable future actions’.Footnote ¹⁹³ Even though the implementation of such assessments is not always entirely satisfactory,Footnote ¹⁹⁴ there is at least a broad consensus that EA cannot ignore the incremental contribution of an activity to a broader environmental issue. The effectiveness of EA would be considerably reduced if it was to turn a blind eye to other stress factors affecting the environment in which the proposed activity is to take place. As Judge Betty B Fletcher noted in Center for Biological Diversity, ‘[t]he impact of [GHG] emissions on climate change is precisely the kind of cumulative impacts analysis that NEPA requires agencies to conduct’.Footnote ¹⁹⁵ The Environment Court of New Zealand once adopted the same position, when it held that, ‘since [GHG] emissions all contribute cumulatively to the same global atmosphere, every small contribution makes a difference’.Footnote ¹⁹⁶ Even though a given activity may only have a minor effect on climate change, human activities are collectively causing tremendous harm: GHG emissions need to be reduced whenever reasonably practical.

Assessing the cumulative effect of an activity's GHG emissions requires, first of all, documentation of the GHG emissions resulting from this activity. A number of methodologies have been developed in recent years to facilitate the documentation of GHG emissions arising from virtually any given activity.Footnote ¹⁹⁷ The stringency of this assessment should depend on the nature of the activity: strenuous efforts to document GHG emissions which are ‘vanishingly small’ would serve no useful purpose,Footnote ¹⁹⁸ but proposed activities which entail massive GHG emissions should attract the most careful scrutiny.Footnote ¹⁹⁹ In some cases, the Institute of Environmental Management and Assessment concedes that a summary qualitative assessment of the source of GHG emissions would be sufficient, although it should be ‘robust, transparent and justifiable’.Footnote ²⁰⁰ A quantified assessment of the GHG emissions likely to result from a proposed activity should only be required when these emissions are likely to exceed a certain threshold. Unsurprisingly, State practice varies considerably regarding the determination of this threshold, which is generally set somewhere between 1 and 100 kilo-tonnes of carbon dioxide equivalent per year.Footnote ²⁰¹ The duration of the activity ought also to be relevant. Beside absolute levels of emissions, any departure from industry benchmark should also command close scrutiny.

It is also largely understood that the documentation of the GHG emissions entailed by a proposed activity has to have ‘a wide scope and a broad purpose’Footnote ²⁰² in order to shed light on the opportunity to carry out the activity and the possibility of reducing these emissions. In addition to the GHG emissions caused directly by the activity, it is generally accepted that an EA should document the impacts of ‘connected actions’Footnote ²⁰³ which have a ‘reasonably close causal relationship’ with the project.Footnote ²⁰⁴ Besides on-site emissions (‘scope 1’ emissions), indirect GHG emissions resulting from the generation of electricity purchased by the project (‘scope 2 emissions’) or from any other supporting activities or infrastructure (‘scope 3 emissions’) may be included in the documentation of the GHG emissions.Footnote ²⁰⁵ Thus, two US Appeal courts held that the EIAs for the construction of a railwayFootnote ²⁰⁶ and a pipelineFootnote ²⁰⁷ destined to be used for transportation of fossil fuels should document the GHG emissions which would result from the use of these fossil fuels. Likewise, a US district court found that the EIA for the construction and operation of electricity transmission lines to connect two coal-fired power plants in Mexico with South California's power grid should document the GHG emissions from these power plants.Footnote ²⁰⁸ The Land and Environment Court of New South Wales, followed by others, held that the EIA of a coalmine should consider the GHG emissions from the burning of the coal by third parties.Footnote ²⁰⁹

In addition, a meaningful assessment of the GHG emissions which would result from a proposed activity should compare these emissions with benchmarks relevant at the scale and in the context of the activity. GHG emissions arising from a given project appear like a ‘drop in the ocean’ when compared with global GHG emissions, but so do its economic benefits in the context of the global economy and in the long history of human civilization.Footnote ²¹⁰ The same amount of GHG emissions may appear disproportionate when compared with GHG emissions of similar projects, or with the sum of GHG emissions in a region or country. It may also be useful to relate the GHG emissions attributable to an activity with the efforts and resources invested in the same region or country to reduce GHG emissions.Footnote ²¹¹ For instance, a study commissioned by the Hong Kong Airport Authority showed that the GHG emissions resulting from a single project—the construction of a third runway in Hong Kong's International Airport—would cancel out two-thirds of the GHG emissions reduction resulting from the region's mitigation efforts by 2030.Footnote ²¹²

An EA should, at the very least, ensure that an activity's GHG emissions are consistent with relevant commitments and targets, in particular the mitigation targets contained in NDCs.Footnote ²¹³ Thus, the Supreme Court of California held that ‘consistency’ with state-wide emission reduction goals could be a ‘legally permissible criterion’ to determine the significance of GHG emissions arising from the large-scale real estate development project.Footnote ²¹⁴ Likewise, the High Court of South Africa considered that an assessment of the GHG emissions that a proposed coal-fired power plant would generate was necessary to ensure that the project ‘fit … South Africa's peak, plateau and decline trajectory as outlined in the NDC and its commitment to build cleaner and more efficient than existing power stations’.Footnote ²¹⁵

However, these two judgments also recognized the complexity of assessing the consistency of a project or programme with an economy-wide goal. California's Supreme Court noted that, as energy efficiency can more easily be reduced in new buildings than by retrofitting existing ones, ‘a greater degree of reduction may be needed from new land use projects rather than from the economy as a whole’.Footnote ²¹⁶ Similarly, the High Court of South Africa ordered the defendants to produce a ‘professionally researched climate change impact report’Footnote ²¹⁷ which would determine whether the project could fit within South Africa's plan to implement its international commitment on climate change mitigation. Other courts noted that compliance with economy-wide targets helped little in the assessment of the significance of the GHG emissions likely to arise from the extension of an airportFootnote ²¹⁸ or the construction of a stretch of highway.Footnote ²¹⁹

In any case, national commitments and other national or subnational targets provide a relatively undemanding benchmark for evaluating the GHG emissions arising from proposed activities. It is well-recognized that the sum of current national commitments on climate change mitigation is far from enough to achieve the global objectives of avoiding dangerous interference with the climate systemFootnote ²²⁰ or holding the increase in global average temperature well below 2°C, possibly at 1.5°C.Footnote ²²¹ It was argued elsewhere that national commitments specifically agreed upon by States, for instance through their NDCs, do not substitute to their obligations under general international law; rather, negotiated instruments under the UNFCCC regime seek to induce incremental steps towards compliance with States’ obligation not to cause significant transboundary environmental harm.Footnote ²²² National commitments surely do not create a right to emit any given quantity of GHG emissions.Footnote ²²³ In application to their broader obligation to mitigate climate change under general international law, States must do every effort possible to overachieve these targets.Footnote ²²⁴ Likewise, the ratchet and review mechanism established by the Paris Agreement implies that States must prepare themselves to communicate and achieve further, more ambitious mitigation commitment beyond the time-horizon of their present NDC.Footnote ²²⁵ Therefore, the GHG emissions arising from a project should not escape a careful assessment simply because they are consistent with relevant national commitments on climate change mitigation.Footnote ²²⁶ Even when a State is already well on the way to achieve its commitments, any reasonable step to reduce GHG emissions further should be considered until the achievement of global mitigation objectives is secured.Footnote ²²⁷

As an alternative way of evaluating an activity's GHG emissions, some have explored the option of attributing a fixed monetary value to a unit of emissions, as a way to compare the ‘cost’ of a proposed activity on the climate system with its benefits. Thus, in 2010, several US agencies adopted a document seeking to estimate the ‘social cost of carbon’ to support regulatory impact analysis under Executive Order 12866.Footnote ²²⁸ This led a District Court to overturn a project approval on the basis that the NEPA review did not include an economic valuation of the GHG emissions caused by a proposed extension of a coal mine.Footnote ²²⁹ The difficulty with this approach is that an economic valuation of GHG emissions necessarily relies on a range of shaky assumptions, for instance with regard to the discount rate applicable when assessing the present value of future harms, the valuation of non-economic harms affecting human societies or ecosystems, the treatment of fundamental uncertainties about the responses of planetary systems to GHG emissions, and even the ethical appropriateness of, effectively, putting non-economic values such as life or biological diversity on the market.Footnote ²³⁰ Likewise, proposed activities often have social or even environmental benefits which do not have a market value. The risk is that a monetarization of the costs and benefits of a proposed activity would displace essential debates from political institutions to expert bodies and, possibly, put a disproportionate emphasis on tangible economic benefits at the expense of a sustainable development.

In conclusion, there is no unique, universally agreed-upon method for determining when the GHG emissions resulting from a proposed activity are significant enough to justify a CA, or even to prevent the approval of the project. The emerging obligation to conduct a CA leaves it largely to States to determine a relevant threshold of significance. Such determination should arguably be made on the basis of national circumstances and taking into account States’ common but differentiated responsibilities and national capacities,Footnote ²³¹ with developed States taking the lead towards more stringent assessments. States are also to decide how best to scope the assessment, including with regard to off-site emissions, and to decide of any appropriate method for the valuation of these emissions and their comparison with other anticipated costs and benefits. This margin of appreciation, however, is not so loose as to render the rule entirely meaningless, and it appears that a hard core of activities—most evidently any projects resulting in more than 100 kilo-tonnes of carbon-dioxide equivalent of on-site greenhouse gas emissions per year—are consistently viewed in most States and under most circumstances as requiring the conduct of a CA.

B. Deliberative Process in the Absence of Direct Harm

Along with an evidence-based assessment of environmental impacts, EA procedures seek to favour a consensual collective decision. In a domestic context, EA laws often promote public participation,Footnote ²³² which, within the boundaries of a political community, may implement a form of deliberative democracy.Footnote ²³³ Yet, as Craik noted, ‘[t]he degree and effectiveness of participation is influenced by the broader culture of openness with which administrative decisions are made in the country in question’.Footnote ²³⁴ While public participation is a central feature of EA procedures in Western democracies, it often boils down to mere formalities elsewhere.Footnote ²³⁵ Transposing this component of EA procedures in a transboundary context has appeared problematic.

The Espoo Convention and its Kiev Protocol suggest that deliberation should involve a notification of the States likely to be affected by the proposed activity,Footnote ²³⁶ an offer to enter into consultationsFootnote ²³⁷ and the creation of opportunities for the public in areas likely to be affected in foreign countries to take an active part in the deliberations.Footnote ²³⁸ Yet, by contrast with the obligations of notification and consultation, the obligation to create an opportunity for the participation of a foreign public is arguably an aspect of the Espoo Convention and its Kiev Protocol which goes beyond customary international law,Footnote ²³⁹ building upon the particular context of the UNECE,Footnote ²⁴⁰ in particular the shared democratic tradition of UNECE Member States—or even, in the case of the EU and its direct sphere of influence, broadly harmonized legal systems. Even in this context, the implementation of the provisions of the Espoo Convention and its Kiev Protocol on public participation has been hindered by issues of coordination between the Party of origin and the affected Party, whose respective roles are not clearly defined in the treaties.Footnote ²⁴¹

Other instruments relating to EA in a transboundary context generally contain less ambitious provisions on deliberations, especially as regards the participation of foreign populations.Footnote ²⁴² While Principle 19 of the Rio Declaration on Environment and Development relates to notification and consultations, it does not mention public participation in a transboundary context.Footnote ²⁴³ The UNEP Goals and Principles of EIA of 1987 approach consultation as a desirable objective rather than a current practice.Footnote ²⁴⁴ The ILC's Draft Articles on the Prevention of Transboundary Harm elaborate on notification, consultation and public information, but not on public participation.Footnote ²⁴⁵ The draft of a Global Pact for the Environment does not mention any form of notification, consultation or participation of a foreign public in relation to the obligation of States to conduct an EA.Footnote ²⁴⁶

Likewise, in a 2010 judgment, the ICJ could find ‘no legal obligation to consult the affected populations’Footnote ²⁴⁷ applicable in the case of Pulp Mills. Instead, the Court held that ‘it is for each State to determine … the specific content of an [EIA], having regard to the nature and magnitude of the proposed development and its likely adverse impact on the environment as well as to the need to exercise due diligence in conducting such an assessment’.Footnote ²⁴⁸ Thus, the ICJ implied that an EA could be conducted in a transboundary context without consultation with, and possibly even without notification of the concerned States.Footnote ²⁴⁹ In its 2015 judgment in Certain Activities, the ICJ clarified that the State conducting an EA in a transboundary context ‘is required … to notify, and consult with, the potentially affected State in good faith, where that is necessary to determine the appropriate measures to prevent or mitigate that risk’.Footnote ²⁵⁰ This would presumably include circumstances where the harm is unfolding in the affected country in ways that may not be well known in the country of origin and where cooperation can effectively mitigate the harm.

This deliberative process needs to take different forms when applied in the context of climate change. The impact of GHG emissions is not limited to one or several States; all States are affected, though none of them directly. As the consequences of GHG emissions do not depend on the nature of the activity from which they result, notification and consultation in relation to each proposed activity would not be necessary to determine measures to prevent or mitigate the harm. States already have various forums where they can express their views on how GHG emissions from other countries affect the global environment and their national interests;Footnote ²⁵¹ no useful purpose would be served by creating an additional forum to express these same views each time a GHG-intensive activity is under consideration.

As implied by the ICJ judgments in Pulp Mills and in Certain Activities, consultations with a third State and a fortiori the participation of the public affected in another country are not an essential element of an EA procedure in a transboundary context.Footnote ²⁵² If public participation is the ‘soul’Footnote ²⁵³ of the EA in a domestic context, its function is not to ensure the direct representation of all stakeholders haggling for their own interests. EAs are often concerned with purely environmental harm or with harm affecting future generations, neither of which could be directly represented through public participation. Public participation, in this context, seeks to promote a meaningful deliberation among reasonable, well-informed citizens who, looking beyond their own interests, reflect on their vision of the common good.

Thus, in the context of climate change, a deliberative process confined to the State in which jurisdiction the activity is proposed could very well include careful considerations for the impacts of a proposed activity beyond national borders and present generations. Such deliberations could be an opportunity for debates as to whether the benefits of a proposed activity would justify the GHG emissions that it could cause. It could also help ensure that decision-makers let no reasonable step to avoid or decrease GHG emissions unexplored. The circumstances in which these debates take place would differ in every country, reflecting a wide range of political and economic as well as social and cultural circumstances.Footnote ²⁵⁴ These deliberations would inevitably involve deeply political questions, in particular through arbitrages between alternative activities likely to cause GHG emissions such as—at the risk of oversimplifying—between an airport to provide fast transportation to the few and a coal-fired power plant to provide cheap electricity to the many. CA could thus ensure that the need to decrease and cease excessive GHG emissions is properly conveyed in relevant decision-making processes.

As such, international engagement is arguably less central in EAs in the context of climate change than it is in a transboundary context, where associating affected populations and their national government to the decision-making may help avoiding or mitigating harm. In practice, CA do not generally involve the notification and consultation of foreign governments, let alone the participation of foreign populations. In a rather unique case, Micronesia took advantage of a Czech statute to demand that the Czech Government enter consultations in relation to the latter's project to renovate a coal-fired power plant in Prunéřov. This intervention invited sufficient public scrutiny to shame the Czech government into altering its project substantially.Footnote ²⁵⁵ Yet, bilateral consultations do not reflect the global and diffuse nature of the harm caused by GHG emissions. Micronesia was not ‘affected’ by the Prunéřov project in any tangible way.

In comparable situations, States have generally inclined towards giving an opportunity to all States and often their population to access to information about the proposed activity, thus enabling them, at least informally, to express their views. For instance, the EA procedure established under in the Protocol on Environmental Protection to the Antarctic Treaty requires EA reports to be circulated to an intergovernmental committee and to each Party to the Protocol, which, in turn, is to make the report publicly available within its jurisdiction.Footnote ²⁵⁶ UNCLOS requires its Parties to communicate EA reports ‘to the competent international organizations, which should make them available to States,’Footnote ²⁵⁷ although this treaty does not create any specific mechanism. In practice, these reports are often communicated to regional seas commissions established under regional agreements.Footnote ²⁵⁸ The Convention on Biological Diversity encourages the conclusion of bilateral, regional or multilateral arrangements for notification, exchange of information and consultations on proposed activities likely to have a significant impact on biological diversity.Footnote ²⁵⁹

Yet, no agreement established any institutional mechanism for notification and consultations as far as climate change is concerned. By analogy to UNCLOS, a State conducting a CA could take the initiative to communicate a report to a competent international organization such as the UNFCCC Secretariat or UNEP in lieu of notification and to invite the organization to gather views in lieu of consultations. Nevertheless, transnational deliberations would certainly be more effective if an institutional arrangement could be established, for instance in the form of a registry recording CA reports and gathering views by national governments and, possibly, authorized non-governmental organizations. This arrangement would ideally be made through a dedicated treaty, but a non-binding decision under an existing framework—for instance a decision of the Parties to the Espoo Convention—would be sufficient to create an optional institutional framework.Footnote ²⁶⁰

In conclusion, there is no clear rule on how deliberations are to be conducted in relation to CA. Consultations of the national population is an essential trait of the EA process itself, and, by extension, of the CA. Beyond this, there is little practice and no established rule on how foreign governments are to be notified and consulted, and whether foreign populations are to be provided an opportunity to participate. It is certainly desirable, but probably not required, that CAs be opened to extraneous interventions, in particular constructive recommendations and offers of financial or technical assistance.

C. Effects on the Final Decisions

EA is about procedure more than about substance. Under most EA laws, a decision-maker, having taken notice of the impacts of a proposed activity on climate change or any other environmental concerns, may nevertheless approve it.Footnote ²⁶¹ Often, CA makes no difference.Footnote ²⁶² For example, less than a year after the High Court of South Africa overturned the administrative approval for the construction of the Thabametsi coal-fired power plant in Limpopo on the ground that the EIA had failed to assess foreseeable GHG emissions,Footnote ²⁶³ a new administrative approval was issued based on a CA of the project.Footnote ²⁶⁴ While acknowledging that the project would result in significant GHG emissions, the approval authority concluded that such harm ‘were outweighed by the benefit to the country of having the additional energy-generation capacity’.Footnote ²⁶⁵ Courts typically do not interfere with the administrative appraisal of the opportunity to approve the activity, unless the decision is procedurally flawed or evidently unreasonable.Footnote ²⁶⁶

On the other hand, one should not expect every single CA to lead to significant changes in the proposed activity. CA are useful even if only a tiny minority of the proposed activities are substantially altered as a result. In rare cases, an assessment of the GHG emissions may tilt the balance and lead decision-makers to withhold approval. More frequently, a CA may suggest conditions on the approval with the view of reducing GHG emissions. In some cases, activities were approved subject to a very general condition, such as the obligation for a project's proponent to ‘implement all reasonable and feasible measures to minimize the release of [GHG] emissions from the [project] site’.Footnote ²⁶⁷ A review of Canada's practice revealed conditions ranging from the use of best technologies, to the adoption of a detailed GHG emissions management plan.Footnote ²⁶⁸ Less directly, in anticipation to CA, the proponents of large projects are likely to adopt alternatives which reduce GHG emissions, for instance by using the latest technology available when economically feasible. There is also some evidence that innovative conditions imposed as the result of a CA can promote new ideas which are then endorsed by lawmakers and imposed as a matter of general policy to all similar activities.Footnote ²⁶⁹