II. Legitimacy in Recent SRM Research Governance Reports

It is important to distinguish between the descriptive sense of legitimacy (also termed sociological or perceived legitimacy), and the normative sense. Descriptive legitimacy is the extent to which, as an empirical matter, an institution is regarded as having authority and is complied with willingly, while normative legitimacy is the extent to which an institution is worthy of support according to a set of standards. The normative sense has received increasing attention in the SRM ethics literature. Most obviously, two special issues have explored questions of political and institutional legitimacy for SRM.Footnote ²⁴ Yet as with the reports noted above, much of the scholarly discussion pertains to SRM deployment, particularly, what would constitute legitimate decision-making leading up to the global-scale deployment of SAI. This leaves unclear the legitimacy standards for the governance of research, and a fortiori, research into non-SAI interventions such as MCB. Different interventions present distinct governance concerns. For instance, to achieve significant cooling MCB would require widely distributed infrastructure, meaning the risk of unilateral action is lower than it is for SAI (though it remains a possibility, especially at regional scale).Footnote ²⁵

There are also distinct conceptions of normative legitimacy. An important point of contention between two influential high-level proposals to guide governance, the Oxford PrinciplesFootnote ²⁶ and the Tollgate Principles,Footnote ²⁷ is their respective interpretations of the principle that public participation is a requirement for decision-making. The Oxford authors ground their call for participation in the Principle of Free, Prior and Informed Consent (FPIC) for research with human subjects,Footnote ²⁸ while the Tollgate authors view the requirement as serving the function of political authorisation, whereby decision-making regarding research programs should be “done by bodies acting on behalf of the global, intergenerational and ecological public.”Footnote ²⁹ These differences in procedural requirements rest on different views of the requirements of legitimacy. While the Oxford authors hold that legitimacy (for decision-making regarding both research and deployment) is to be established relative to existing institutional structures, the Tollgate authors argue that it is an open question whether a given institutional structure attains the standard of legitimacy required for the oversight of SRM-related activities.

The recent high-level reports produced in the United StatesFootnote ³⁰ pay some attention to considerations of legitimacy, but it is questionable whether they engage with institutional legitimacy under the broadly normative conception invoked by the authors of the Oxford and Tollgate principles. The committee that produced the NASEM report notes that it “worked under the basic premise that it is reasonable to proceed with appropriately structured and governed research.”Footnote ³¹ This assumption, while understandable given the committee’s terms of reference, means that the question of the conditions under which institutions may legitimately institute SRM research programs is formally excluded from consideration. NASEM calls for measures to improve the legitimacy of research, though this term is usually grouped together with the concept of “trust” or is referred to as “perceived legitimacy”, implying that a descriptive conception of legitimacy has been foregrounded. Legitimacy is also sometimes referred to as a “substantive” consideration along with research quality, in this context appealing to a conception of legitimacy that draws on the traditions of transdisciplinarity,Footnote ³² post-normal scienceFootnote ³³ and knowledge co-production,Footnote ³⁴ according to which legitimacy refers to the scientific justifiability of research with respect to socially embedded aims. While “political legitimacy” is mentioned when presenting the Oxford and Tollgate principles (meaning, in this context, the normative legitimacy of decision-making institutions), this conception of legitimacy is not used elsewhere in the report.Footnote ³⁵

The EU’s Group of Chief Scientists, meanwhile, propose that a permitting scheme for outdoor research should be managed in the context of a treaty-based international governance regime, with “international legitimacy”Footnote ³⁶ This recognises the role of institutional legitimacy in the authorisation of some forms of research. They do not, however, give a precise specification of the legitimacy conditions for such a regime, or of the legitimacy conditions of institutions authorising research prior to the conclusion of a treaty (which may not be achieved for many years, if at all).

III. Legitimacy in Political Theory

In the ethics and political theory literature, institutional legitimacy–legitimacy as a property of institutions in their capacity as makers of rules–has traditionally been foregrounded. Thus, discussions of legitimacy with respect to SRM research and deployment have tended to be approached via accounts of the legitimacy of the institutions overseeing those activities. These accounts have drawn on one of the central theories of normative legitimacy, developed by Buchanan and KeohaneFootnote ³⁷ and by Buchanan independently.Footnote ³⁸ Buchanan and Keohane developed the Complex Standard of Legitimacy for Global Governance Institutions, which offered substantive criteria for assessing the legitimacy of such institutions. Buchanan went on to explicitly set out his conception view of the function of legitimacy assessment, the Metacoordination view, according to which, to say that an institution is legitimate is to say that it is morally worthy of our support because the benefits of empowering it outweigh the risks of doing so.Footnote ³⁹ This account is grounded in the function that legitimacy assessments serve in practice: solving a “metacoordination” problem, i.e., balancing the burdens and benefits of various actual and potential institutional arrangements, in their own capacity as solutions to first-order coordination problems. Though legitimacy assessment involves consideration of risk, an analogy with risk analysis is apt only in part. This is because legitimacy assessment is fundamentally a social practice. The role of legitimacy assessment is not to provide criteria for the calculation of expected value, but parameters for deliberation about effective collective action. The behaviour of others is an input in such deliberation, while in risk analysis it is a variable whose value must be estimated. Legitimacy assessment is in this sense more tractable than risk analysis, focusing deliberation on clear criteria, current institutional capacities and the near-term feasibility of alternatives.

Buchanan takes his assessment criteria to be supported by the Metacoordination view of the function of legitimacy assessment. Applying the view to specific institutional contexts will yield refinements to the criteria. Buchanan summarises the criteria as follows: (i) comparative benefit (the institution provides a benefit relative to feasible alternatives); (ii) institutional integrity; (iii) minimal moral acceptability; (iv) acceptable origination; (v) non-discrimination.Footnote ⁴⁰ Callies distils this into the following SRM-specific requirements: (1) comparative benefit; (2) accountability; (3) transparency; (4) substantive justice; and (5) procedural justice.Footnote ⁴¹

The two accounts that apply Buchanan’s view to SRM governance broadly agree at the level of assessment criteria, though they diverge in scope. Morrow, Kopp and Oppenheimer present an account of the legitimacy conditions for an institution that would oversee “large-scale” experimentation, defined as experiments large enough to distinguish their effect from normal climate variation.Footnote ⁴² The basic claim is that large-scale experimentation must only be carried out by institutions that meet assessment criteria. Open-air experimentation on a scale large enough to cause a measurable change in regional climate, they argue, effectively constitutes human subject research given such research would have the potential to materially affect human populations in these regions. Ideally, such research should be governed by the FPIC requirement, following standard research ethics norms in biomedical science. The consent of large populations is typically mediated through institutions, which, under the right conditions, allow for collective consent to or dissent from large scale interventions. A necessary condition for the authorisation of large-scale field tests is thus that the decision to carry out such tests is made by an institution that meets criteria for global political legitimacy.

In contrast, Callies’ account is intended to cover decision-making institutions, including research institutions, pertaining to SRM activities generally. Callies discusses SRM without differentiation between interventions. While this means his view elides specific considerations in relation, for example, to targeted geoengineering,Footnote ⁴³ because the precise function of the institution under consideration is something “to be decided together”Footnote ⁴⁴ (presumably, via a political process) his view retains general relevance, though more work is required to operationalise it.

Callies goes beyond Buchanan and Keohane’s requirements in several respects. Notably, he interprets procedural justice as demanding that “all those with legitimate claims to participate in the decision-making process [all affected parties] are included,” and “all those included in the decision-making process are included on justifiable terms.”Footnote ⁴⁵ He argues that while these requirements remain general, they can assist in identifying clear cases of procedural injustice. As a purported example, Callies describes a case in which residents of small island states lacked access to fair terms of participation in “decision-making”, despite facing potentially catastrophic effects both of climate change and of climate engineering. Thus, “any process that failed to include these people”, the claim goes, “and allow their voices to be heard on reasonable terms of participation would be procedurally unjust (regardless of the substantive outcome).”Footnote ⁴⁶

In assessing the adequacy of these accounts, it is noteworthy that Buchanan and Keohane’s criteria were developed primarily for assessing the legitimacy of extant global governance institutions such as the WTO, with a view to offering critical analysis in pursuit of reform. It is less straightforward to use these criteria to advocate for the creation of a novel institution, as both Morrow et al. and Callies do. The application of these criteria is also unclear, when there are already some polycentric governance structures instantiated in a range of institutions (e.g., general research regulations), and various potential functions institutions could serve.

Another complicating factor is scope. Morrow et al. limit the applicability of the assessment criteria to the governance of institutions authorising large-scale open-air experiments. This is because, as noted above, they deploy legitimacy as a proxy for FPIC. Callies quite reasonably takes the view that legitimacy assessment can be applied to any institution, whether or not they coercively enforce the rules they create. The demands of assessment would simply vary with the case. However, Callies does not provide a rationale for why an institution “overseeing geoengineering” should be the object of assessment, rather than a set of institutions with different functions whose legitimacy should be assessed relative to those functions (for example, determining which interventions to pursue, directing research funding, issuing guidelines, authorising experiments, etc.). There is an assumption that the institution would have global scope, which may rest on an assumption that it would have the authority to authorise large-scale SAI experimentation. This means that the question of legitimacy conditions for institutions with more limited authority is not directly addressed. In other words, Morrow et al. are right to tie assessment criteria closely to scope, while Callies is right to try to expand legitimacy assessment to research institutions, he simply pays insufficient attention to distinctions of institutional function and the specific criteria they imply.

That said, it is plausible that institutions overseeing SRM research in general (including for instance indoor research and social scientific research) should be held to the same standard of legitimacy – considered at a high level – given such institutions serve the following basic function: promoting institutional responses to the problem of climate action that align with societal values. This does not mean that legitimacy conditions cannot be more demanding for some institutional functions than for others. Establishing a common standard in broad terms (though with different implications given different specific contexts) would also be expedient for building up institutional competence, so that institutions could fulfil a more demanding governance role in future, if needed. This would also build perceived legitimacy. Assessing the legitimacy of institutions as their competences evolve thus requires dynamic standards.

Buchanan and Keohane do not consider it necessary for a legitimate global governance institution to involve direct representation of the public, so long as it has the consent of democratic states. Callies, meanwhile, argues that specific legitimacy conditions for SRM governance institutions do indeed demand public participation “in some way, shape, or form.”Footnote ⁴⁷ The reason turns on a further claim about representation. Callies claims that institutional legitimacy requires more genuine representation of the interests of those purportedly represented. He points to the common criticism that global institutions, even if well-designed, tend to be less representative than domestic institutions, and that non-democratic participants in such institutions arguably do not represent their constituents’ interests in any real sense. It is, however, unclear in Callies’ account how this additional requirement is justified by the Metacoordination view of the function of legitimacy assessment, which turns on the ability of an institution to perform its practical function of resolving a collective action problem. Callies’ account of the scope of such participatory requirements is also incomplete. He offers “legitimate interest” in SRM as the criterion for inclusion but leaves the meaning of this term unspecified, relying instead on paradigm cases like that of small island states.

Our claim is that we can offer a more detailed story about who has a claim to participate in the oversight of which activities, and why. It is possible to derive an argument for direct rather than mediated representation of stakeholders, and for the inclusion of certain stakeholders, from a functional legitimacy assessment of research policy institutions, and the special function of those bodies within a system of institutions directed towards climate action. Where Callies discusses the general form that a hypothetical institution ought to have, we claim that we can make progress by placing the extant architecture of research oversight in the assessment spotlight, to determine how it could be augmented to serve various potential roles, and thus to appropriately manage research at various stages. In this way, the practice of legitimacy assessment can serve the function of improving the capacity of institutions to balance competing interests in a mutually beneficial way, while also acting as a spur to further institutional reform.

We argue that assessment of existing research governance architecture, taking a Metacoordination view as a model, in the wider institutional context of global climate governance and regional research governance, allows us to provide a richer account of participatory norms and their justification. Our claim is that inclusive participation assessment criteria should include considerations of justice as recognition. Such criteria are context-dependent, grounded in the institutional practice of international climate action. These considerations provide a contextually specific interpretation of Callies’ under-theorised “legitimate interest” criterion. Recognition highlights the danger of domination posed by the development of SRM capabilities, and the prospects of misrecognition of certain groups as a result of SRM-related activities, before any globally significant effects are likely to occur. Based on recognition, we propose to clarify participatory assessment criteria for any research institution overseeing SRM-related research activities. We turn to this argument in the next section.

IV. The Importance of Recognition

Recent ethical debates feature growing attention to recognition as an essential aspect of justice in the context of SRM-related activities,Footnote ^{48 ,} Footnote ^{49 ,} Footnote ^{50 ,} Footnote ⁵¹ and related concerns with domination resulting both from a potential future deployment of SRM, and even from the pursuit of research.Footnote ^{52 ,} Footnote ^{53 ,} Footnote ⁵⁴ Despite this, recognition is not explicitly mentioned in either of the accounts of institutional legitimacy discussed above. In our view, a lens of recognition justice helps to clarify ambiguities surrounding who has a legitimate interest in SRM research and the appropriate terms of participation for research governance.

Before proceeding, we will briefly explain what is distinctive about justice as recognition, and what would constitute morally unacceptable misrecognition of certain groups.Footnote ⁵⁵ On Nancy Fraser’s influential account, marginalised or oppressed collectivities tend to be simultaneously subject to injustices of distribution and of recognition where neither of these are reducible to the other.Footnote ⁵⁶ Distributive injustices arise when unequal access to economic opportunity or economically derived power skews the structure of society, so that some enjoy undeserved benefits while others face undeserved hardship or exploitation. In contrast, recognitional injustices arise when “pervasive cultural patterns of interpretation and evaluation” serve to place a collectivity into an inferior social status.Footnote ⁵⁷ The harms of misrecognition, then, are “status injuries”, as experienced by women as opposed to men or colonised peoples as opposed to colonial peoples.

The core claim here is that such status injuries typically require measures of redress that go beyond material distribution, for instance through the institutional recognition of equality of social status. According to Fraser’s “bivalent” account of injustice, “virtually all real-world oppressed collectivities … suffer both maldistribution and misrecognition in forms where each of those injustices has some independent weight, whatever its ultimate roots” and thus “overcoming injustice in virtually every case requires both redistribution and recognition.”Footnote ⁵⁸ Fraser’s account proposes the norm of “participatory parity” to redress misrecognition. According to this, “justice requires social arrangements that permit all (adult) members of society to interact with one another as peers.”Footnote ⁵⁹ This has two necessary conditions: material distribution must be sufficient for individuals to be able to participate and to express their particular perspectives (what Fraser calls “voice”); and second, “institutionalized cultural patterns of interpretation and evaluation express equal respect for all participants and ensure equal opportunity for achieving social esteem.”Footnote ⁶⁰

V. Deliberating on the Risks of Institutional Empowerment: the Role of Institutional Context

We argue that the procedural and substantive justice criteria of legitimacy assessments for institutions directing SRM research ought to involve considerations of recognition. This is not because of the material impacts of experiments on communities (although these must be considered). Rather, it is because SRM is, by definition (with some caveats),Footnote ⁶¹ a proposal to address a global problem. Under a Metacoordination view of legitimacy assessments, an institution is legitimate if it is worthy of our moral reasons-based support, given the function of the practice of making legitimacy assessments (solving the relevant metacoordination problem). This means the scope of legitimacy assessments is tied to the scope of that metacoordination problem: balancing the benefits of empowering such an institution against the risks of empowering it.

Assessment requires a clear specification of the function of the institution being assessed. Let us focus for now on a more specific function than the one considered by Callies: whether an institution should be empowered to support a programme of SAI and MCB experiments (leaving aside questions of scale for the moment). It is important to be clear about the collective action problem the institution is supposed to address. Here, the distinction between mission-driven and exploratory or investigator-driven research is relevant. For an institution overseeing exploratory research, the relevant collective action problem is ensuring funds are distributed judiciously to promote research of the highest scientific value, while maintaining a resilient scientific research sector. Institutions overseeing mission-driven research, meanwhile, also address the collective action problem embodied by the mission itself. For instance, missions under the Horizon Europe programme round ending in 2030 include supporting 150 European regions to adapt to climate change – coordinating effective adaptation in Europe. The recent high-level science policy reports, in particular NASEM, are moving in the direction of mission-driven research, aiming to assess the potential of SRM as response to global climate change. This means institutions overseeing such research are, and should be assessed as, institutions coordinating the response to global climate change.

The ways in which empowering such an institution could distribute risks and benefits are reasonably well rehearsed, although we should note they are not identical to the risks of prospective experiments themselves. While risk assessment in relation to experimentation focuses on environmental impacts, assessing the risks of inadequate institutional coordination with respect to SRM oversight will foreground considerations arising from power that is inappropriately concentrated or inadequately checked. It will also focus on balancing competing interests between groups with an interest in, or advocating, different responses to climate change. This is why consideration of any potential political distraction from mitigation falls within the competence of SRM research oversight institutions (this paper is neutral on the question of whether mitigation deterrence in fact occurs).

Relevant threats to effective coordination also include the potential for domination in a number of possible domains. Stephens and Surprise identify a risk of “structural power imbalances” in the form of “colonial legacies of paternalistic technocratic humanitarism” that arises from SRM research having been dominated by researchers from the Global North and funded by wealthy elites.Footnote ⁶² Taiwo and Talati argue that overstating this concern risks recapitulating the very colonial legacies it seeks to overcome by excluding Global South researchers from decision-making. They argue that a research program should produce a “collective framework to make collective decisions about SG [solar geoengineering]—importantly, including the decision about whether to abandon such technologies and further research into them.” For instance, they argue that representative bodies based on sortition should be empowered with “direct and consequential decision-making power” and this is “key to their potential to address social inequalities as such.”Footnote ⁶³ There is also a danger of domination created by an imbalance of information. SRM research will create new knowledge which will advance its possessor’s capacity to manipulate the climate to greater or lesser extents, even if that capacity is not exercised or scalable technologies are never developed.Footnote ⁶⁴ Given that such information arises not only from experiments at certain scales, but from SRM research in general, it is the initiation of research programmes that need to be the object of assessment rather than any particular research project or experiment.

Some interventions present a lower risk of creating power imbalances than others; this would need to be considered in assessment. As noted above, because the infrastructure for a large-scale MCB deployment would have to be widely distributed, its potential for militarised or unilateral deployment is lower than for SAI. That said, even research into localised forms of SRM can involve dangers of domination and misrecognition: these dangers form a continuum with the parallel dangers involved in countries’ climate policies overall. Just as wealthy countries have greater adaptation capacity, enabling them to countenance higher-risk emissions pathways, so too might knowledge derived from SRM research improve the freedom of action of certain parties at the expense of others.

This demonstrates that the metacoordination problem for SRM research policy institutions is nested within a wider metacoordination problem of institutional responses to climate action. This is a clear reason that sets such an institution’s function apart from other institutions for the oversight of scientific research. In this sense, it is part of an institutional landscape that includes, for example, the UNFCCC, the UNEA and UNEP, climate focused sub-initiatives of scores of other international organisations, and national-level and transnational institutions. Because these institutions are responses to the same coordination problem, they help to establish parameters for specific interpretations of legitimacy criteria.Footnote ⁶⁵ Of particular relevance for our purposes are the criteria of substantive and procedural justice, which Callies argues must involve prioritising the interest of the most climate-vulnerable, and will likely involve public participation. When we place legitimacy assessment in the context of a constellation of institutional responses to climate action viewed as a coordination problem, the justification for these requirements becomes clearer. This is because the institution authorising experimentation must be judged by normative standards as high or higher than existing institutional practice, as implied by the principle of comparative benefit. This principle states that an institution must do a better job of addressing the relevant collective action problem than non-institutional alternatives. Thus, existing standards that would apply in the absence of the institution under assessment constitute a benchmark against which they should be judged.

Emerging international practice can be used to demonstrate that recognition represents an appropriate way to understand the existing standard. This is apparent from the form and content of the demands of vulnerable groups for participation in international climate action. Preston and Carr, for instance, note that “scientists, ethicists, humanitarian and environmental organisations, and members of the public have all advocated for the inclusion of more geographically and culturally diverse perspectives in future research and decision-making.” Thus, they argue on the basis of interview studies that climate vulnerable groups themselves (indigenous Arctic, sub-Saharan Africa, small island states) conceptualise their interest in climate action as achieving control of their destiny, in contrast to decades of marginalisation in the context of colonialism.Footnote ⁶⁶ They also draw attention to groups with a particular “environmental heritage” that might be marginalised or ignored by researchers conducting SRM-related activities, and thus be made victims of kinds of social status injuries that constitute recognition injustice.Footnote ⁶⁷ Preston and Carr’s interviews also bore out an important concern with a loss of voice. They write, “[i]njustices that respondents worried a great deal about included the ways in which traditions, practices, and ways of knowing might be ignored or made impossible. They included relationships of power, dependence and vulnerability that might be established. They also involved the likelihood of not being equal participants in decisions that would impact them.”Footnote ⁶⁸ As Kyle Powys Whyte has argued, indigenous peoples in particular face significant barriers to equal participation in SRM discussions, including those pertaining to research governance.Footnote ⁶⁹ This is not only due to economic inequalities, but stems from a lack of respect for alternative ways of viewing climate change as a moral and political problem, in particular, through the ongoing significance of colonialism. Such a reframing of the problem would have important implications in light of the FPIC principle.

Applying a recognition lens to research governance would also require a different approach to knowledge production and the science-policy interface. As Hourdequin notes, building recognition considerations into research processes would mean departing fundamentally from “pipeline models of science and ethics, where ‘experts’ develop knowledge and policy makers and ‘the public’ later decide how to apply it.” Instead, a recognition-based approach to SRM research would “involve upstream engagement from diverse publics, give developing countries a leadership role in research on SRM, and prioritise regionally focused research attentive to local effects and impacts.” Such an approach would give greater salience to questions surrounding the “distribution of power, voice, and resources in discourses and decisions surrounding SRM.”Footnote ⁷⁰ This too implies a significant departure from standard research governance, and a reconceived science-policy interface such as one would see in post-normal science contexts.Footnote ⁷¹

The idea that recognition is the proper model for participatory requirements under assessment criteria for institutional legitimacy is already embedded in the norms of international climate institutions. An example is the Escazù Agreement on Access to Information, Public Participation in Decision-making and Access to Justice in Environmental Matters in Latin America and the Caribbean, a region with a high level of cultural diversity and many indigenous communities.Footnote ^{72 ,} Footnote ⁷³ It requires states to actively “identify and support” vulnerable parties to eliminate barriers to participation in environmental decision-making, a turn towards concern not just for recognition but for participatory parity. More recently, the UN Special Rapporteur in the field of cultural rights has recommended “a wide, inclusive and decolonized understanding of science is an important way to fulfil the right to participate in science”, reconfirming that it includes the right to participate in decision-making, and noting “democratic control does not mean state control.”Footnote ⁷⁴

VI. Institutional Implications of Recognition-based Legitimacy Assessments

The small number of experimental research projects on SRM hitherto have attempted to develop their own protocols for accountability, transparency and public engagement. Recent authoritative reports on SRM governance have begun to acknowledge the shortcomings of this approach. The AGU principles, for instance, state that not only research itself but also “the decision to pursue it” should be “based on processes for equitable engagement with all stakeholders involved.”Footnote ⁷⁵ UNESCO COMEST notes “Member States must ensure that vulnerable, neglected, and marginalized individuals and groups, women, youth, and indigenous people are included as key stakeholders in the development and implementation of policies [and] decisions.”Footnote ⁷⁶ Similarly, the EGE notes “it is crucial to involve diverse populations, including vulnerable individuals and groups, underprivileged regions of the globe, and marginalised social groups in the process of setting SRM goals and related research agendas.”Footnote ⁷⁷ It also notes, citing the UN Special Rapporteur on Cultural Rights, that the right to science implies a right to participate in policy decisions relating to science.Footnote ⁷⁸

Locating protocols for accountability, transparency and public engagement at the level of the research project risks placing questions of agenda setting and decisions to pursue research outside the window for public engagement. These recent arguments reflect earlier proposals by social scientists in the US context to build legitimacy for SRM research through substate institutions, in the form of advisory commissions, which should “ideally operate at the level of research programs, rather than at the level of specific proposed experiments,” and involve participation from affected communities.Footnote ⁷⁹ The question of what scope to apply – whether to include only those materially affected by near-term proposed experiments or also by hypothetical large-scale deployments – is a matter to be determined by the commissions themselves.Footnote ⁸⁰ Our view is consistent with this idea: the failure to allow for participatory involvement of globally climate vulnerable groups at the programmatic level would be an unacceptable legitimacy deficit for some functions, but the demandingness of oversight must evolve with the function.

While a full treatment of institutional possibilities is beyond our scope here, there are a number of ways this deficit could be addressed. One would be to expand the definitions of stakeholders in existing research policy oversight committees to include representatives of indigenous Arctic communities, global south communities and representatives of climate-vulnerable populations. In the European context, this could include bodies like the ERA Forum, which already includes representation for industry stakeholders. Another would be via standing committees of the European Research Council. In the first instance it is plausible that the prospects for institutional reform will be pursued via research projects, but in order to build legitimacy further, these need to transition to permanent structures. Here we can learn from ongoing experience internationally. The Queensland-based Reef Restoration and Adaptation Program – a suite of projects on preserving the Great Barrier Reef, with targeted MCB among the methods being investigated – provides useful insights for the European context. The RRAP places research in the context of a wider, mission-driven programme of reef restoration projects, recognising at a high level the importance of co-creating research with local stakeholders, with specific rights being recognised for Traditional Owners of the reef and surrounding sea country, over 70 indigenous groups.Footnote ⁸¹

RRAP is not a “geoengineering” or “SRM” programme, thus its MCB research is also not characterised as such, but the model is nevertheless instructive from a structural perspective. Where SCoPEx was a project to conduct a specific experiment, and plans for stakeholder engagement – including late, unsuccessful attempts to engage the indigenous Sámi – were framed in relation to that experiment,Footnote ⁸² RRAP has been able to engage stakeholders at a programme level.Footnote ⁸³ This simple structural fact enables much richer opportunities for a relationship with stakeholders that embodies recognition. With the aims of projects or particular experiments already wholly determined by researchers, there is little stakeholders can do but register their assent for or dissent from researchers’ proposals. This relationship is too narrow to embody recognition, which as noted above involves, for instance, respect for the ways in which different communities conceptualise the problem under investigation. Engagement at a programme level, meanwhile, creates space for meaningful participation on topics including the determination of research priorities and experimental design. This approach is more broadly “anticipatory” than project-level legitimacy assessment, helping to identify and mitigate risks in advance of their emergence.Footnote ⁸⁴

In the EU context, stakeholder engagement might be integrated at a programme level through “missions” under the Horizon programme. If these missions in relation to climate change continue to focus, as in the current cycle, on adaptation, the extent to which various forms of climate intervention research can be said to contribute to adaptation goals has to be determined. This is an area of policy that carries significant risk of misrecognition. Interventions which are among the most active areas of research in the European context include cryosphere interventions, which are increasing framed as being pursued for the sake of helping local communities adapt.Footnote ⁸⁵ If research oversight institutions under the European research architecture are to promote such research, the conception of institutional legitimacy we advocate implies that local stakeholders are to be involved at a programme level in definitional and agenda-setting questions. This would include helping to determine whether research topics were an appropriate fit with high-level missions and helping to surface key concerns to be included in the terms of reference for funding calls.

A key question that arises here is how much influence groups who are likely to be subjects of misrecognition ought to have over research decisions. While we cannot pursue this question further here, it is a result of our view, echoed by Callies, Morrow et al. and others, that legitimacy is enhanced by the capacity to participate in collective decision-making over the direction of research programmes. One marker of this would be providing a collective capacity to set research agendas. This should not amount to a veto for any interested party, to allow for competing interests to be balanced. Careful work will therefore be needed to ensure institutional integrity is preserved by reforms, especially if the new institutional architecture governing SRM research is to be vested with significant authority.

VII. Objection: Asymmetry

It may be objected that legitimacy assessment places a heavy burden on institutions, one that constrains their accepted or normal operation. This can either be framed as an objection about states’ prerogatives, or about SRM research governance institutions being subject to disproportionate scrutiny, especially given that polluters exert influence over the climate system with relative impunity. It is important to recall that the Metacoordination view implies that legitimacy admits of degrees.Footnote ⁸⁶ Legitimacy criteria on this view present a set of parameters for public deliberation about institutional design. The claim is not that institutions that do not conform to the participatory norms defended in the present paper by that token lack any authority, rather, it is that is it reasonable for members of the public to object to them.

Thus, though states make policy as they see fit within their sovereign competence, this does not mean they are immune from justified criticism under institutional legitimacy criteria. An implication of the Metacoordination view is that, because any suitably situated agent has moral reason to support institutions that meet the criteria, domestic citizens have just as much reason to support the institution as external actors. The sceptic of moralised conceptions of legitimacy will demand to know why they should view institutions as responding to a global collective action problem, rather than simply serving the function of domestic coordination. A non-moralised response is that it is in their interest to do so: the global coordination problem is a more pressing concern than any domestic coordination issue.

It is true that institutions pursuing damaging climate policies have tended to evade legitimacy-based criticism. In this context, the asymmetry challenge consists of the compelling observation that raising questions of legitimacy in relation to SRM research seems an unfair or irrational distribution of critical concern. We note here that questions of legitimacy in relation to climate policy are entirely appropriate, even if somewhat rare or obscure. The actions of climate policy institutions can helpfully be understood as attempting to increase legitimacy according to criteria of the kind adduced by Buchanan. Take, for example, Climate Action Tracker, a project that aims to improve transparency on states’ actual policies, their promises via Nationally Determined Contributions under the Paris Agreement, and the likely impacts of both.Footnote ⁸⁷ As Buchanan and Keohane note, improving transparency can serve as “proxy” for the legitimacy assessment criteria of minimal moral acceptability, comparative benefit and institutional integrity: institutions that systemically fail to be transparent are subject to a justified presumption of illegitimacy.Footnote ⁸⁸ Such transparency mechanisms therefore serve both legitimation and legitimacy assessment functions. Critiques of the UNFCCC process challenge its failure to stabilise global greenhouse gas emissions, its function established by treaty.Footnote ⁸⁹ Such critiques are equally applications of legitimacy assessment, in terms of institutional integrity. Thus, at the level of normative assessment at least, we believe that any problematic asymmetry between SRM research governance and the governance of other activities is overstated.