Institutional Legitimacy and Geoengineering Governance

ABSTRACT

There is general agreement amongst those involved in the normative discussion about geoengineering that if we are to move forward with significant research, development, and certainly any future deployment, legitimate governance is a must. However, while we agree that the abstract concept of legitimacy ought to guide geoengineering governance, agreement surrounding the appropriate conception of legitimacy has yet to emerge. Relying upon Allen Buchanan’s metacoordination view of institutional legitimacy, this paper puts forward a conception of legitimacy appropriate for geoengineering governance, outlining five normative criteria an institution ought to fulfill if it is to justifiably coordinate our action around geoengineering.

KEYWORDS:

• Geoengineering
• climate engineering
• legitimacy
• governance
• metacoordination view

1. Introduction

Geoengineering is a controversial topic.¹ On the one hand, many of the technologies under the geoengineering umbrella boast characteristics that make them appropriate for continued research and possible future deployment. On the other hand, many of these same technologies are dangerous. Take stratospheric aerosol injection, for example. It boasts relatively strong leverage on the climate system. The fact that with relatively little effort we could generate huge impacts on global climate gives us reason to look into the technology. However, that same strong leverage makes the technology troubling. In the hands of an irresponsible actor, such technology could have devastatingly negative effects across the world population and could irreparably damage natural ecosystems and the species that comprise them. In a similar vein, the relatively inexpensive price tag attached to the technology means that we could alleviate some future climatic harms without having to divert scarce resources away from mitigation, adaptation, and other valuable goals such as the eradication of global poverty. However, that same relatively inexpensive price tag makes the technology vulnerable to unjustified unilateral action since a multilateral cost-sharing arrangement is not strictly necessary.²

Given that geoengineering technologies have the potential to bring about both significant benefits and drastic burdens, regulation is important. Nearly everyone involved in the normative discussion about climate engineering agrees that if we are to move forward with significant research, development, and certainly any future implementation of geoengineering technologies, legitimate governance is a must.³ Despite this agreement that further research and certainly any future deployment should be accompanied by legitimate governance, there has been relatively little discussion to date among political philosophers about what would constitute legitimate governance of such a technology.⁴ That is, while we all might recognize that the abstract concept of legitimacy ought to guide geoengineering governance, agreement surrounding the appropriate conception of legitimacy for geoengineering governance has yet to emerge.

The main point of this paper is to introduce a framework out of which an appropriate conception of legitimacy can spring. To do so, I’ll begin in the next section by exploring specific conceptions of legitimacy that have been developed in the philosophical literature to date. I’ll argue that what we need is a general concept of legitimacy that is appropriate for the diversity of institutions that occupy the international realm – the realm where geoengineering governance would take place. Drawing on the recent work of Allen Buchanan, Section 3 then puts forward a general concept of institutional legitimacy. This general concept will pave the way for a more specific normative conception of legitimacy (explored in Section 4) that will serve to coordinate judgements about the legitimacy of an institution set up to oversee climate engineering. Section 5 offers a quick recap and then concludes the paper.

2. Conceptions of (State) Legitimacy

Before looking at specific conceptions of legitimacy, we need to be clear about the general concept of legitimacy.⁵ In the philosophical literature, the concept of legitimacy has generally denoted either the justification of political power – understood as coercive power backed by government sanctions (Rawls, 2005, p. 136) – or the justification of political authority – understood as a right to rule and a correlating obligation to obey (Buchanan, 2003, p. 147). Aiming at these understandings of the concept of legitimacy, various conceptions have been developed that spell out exactly when it is that political power or political authority is justified. For example, according to Rawls, ‘… political power is legitimate only when it is exercised in accordance with a constitution the essentials of which all citizens, as reasonable and rational, can endorse in the light of their common human reason’ (2001, p. 41, as cited in Peter, 2014). Here we see Rawls grounds legitimacy (understood as the justification of coercive power) in the processes whereby political power is exercised.⁶ As long as political power is exercised in accordance with a constitution that citizens could hypothetically endorse, then the exercise of such political power is justified. A. J. Simmons has a more demanding conception of legitimacy. For Simmons, a legitimate government has a right to rule and citizens of a legitimate government are under an obligation to obey. The only way, according to Simmons, for a government to derive such a right and for citizens to be under such an obligation to obey is if the citizens have actually expressed their consent (Simmons, 2001, p. 136). In the absence of such consent, the government has no right to rule and citizens are under no obligation to obey. Given that the legitimacy of the government is conditioned upon voluntary consent, consent theorists like Simmons are called ‘voluntarists.’

Now, the Rawlsian or the voluntarist conception of legitimacy may be an appropriate conception for the legitimacy of state-like institutions.⁷ But these conceptions do not apply to international institutions very well. The Rawlsian conception places a priority on the publicly recognized conception of justice embedded in the constitution. And the voluntarist conception places a priority on the express consent of those within the purview of the state. But international institutions exercise political power without any publicly recognized constitution and without the consent of all of those within their purview. Rather than reach the conclusion that international institutions are illegitimate because they fail to meet either the Rawlsian or the voluntarist standards, I contend that we should rely upon a different conception of legitimacy for two reasons.

First, international institutions are beneficial. For example, the world is a better place with the International Atomic Energy Association (IAEA) and the United Nations Framework Convention on Climate Change (UNFCCC) wielding the power they do, notwithstanding their imperfections. And this is the case despite the fact that they are wielding this power in the absence of any international or global constitution and despite the fact that they do not enjoy the consent of every individual over whom they wield such power. To conclude that these institutions are illegitimate because they fail to meet the standards of state legitimacy outlined by Rawls and Simmons would be counterproductive.⁸ But, second, it makes sense to say that how high we set the bar for an institution to be considered legitimate should be sensitive to the characteristics and the function of the particular institution in question. If an international institution is wielding significant power like a state, then perhaps it should meet the demanding requirements we expect of states for them to be considered legitimate. On the other hand, if the institution is merely providing suggestive guidelines and has no ability to enforce any of its directives, then we may want to relax the criteria it needs to fulfil to be considered legitimate.

To be more explicit, what we need is a general concept of institutional legitimacy; a concept that is malleable enough to give rise to appropriate conceptions of legitimacy for the variety of institutions that occupy our world. Clearly explicating a general concept of institutional legitimacy will prove invaluable in developing a more concrete conception of legitimacy that can be applied to the kind of climate engineering institution that ought to oversee research and possible development.

3. Institutional Legitimacy

To begin the outline of a general concept of institutional legitimacy, we can ask the following question: Why do we have institutions and what is it for an institution, in general, to be legitimate? We have institutions, according to Buchanan, in order solve coordination problems. They supply us with the kind of coordination we need to achieve certain desirable outcomes and do so without the excessive costs or inefficiencies associated with non-institutional alternatives. In order for institutions to effectively coordinate our action and deliver the desirable outcomes, they require us to grant them a certain kind of standing – the kind of standing they need to perform their functions. And in order for us to justifiably⁹ grant them the kind of standing they need in order to perform this function, we first must converge (or coordinate) on particular normative criteria that institutions ought to fulfil to be worthy of that standing. So, legitimacy assessments serve to solve a metacoordination problem: they allow us to justifiably coordinate around normative criteria that institutions must meet if we are to grant them the kind of standing needed to solve the further problem of coordinating our collective action towards certain desirable outcomes. Given that legitimacy assessments allow us to solve this higher order coordination problem, Buchanan aptly refers to this as the Metacoordination View of institutional legitimacy.

3.1. The Metacoordination View of Institutional Legitimacy

The Metacoordination View of institutional legitimacy thus says, abstractly, that an institution is legitimate when it is worthy of the certain kind of standing needed for it to perform its institutional goals. Thus, a geoengineering governance institution would be legitimate if it is worthy of the kind of standing it needs to coordinate our action around research and development (or abandonment!) of technologies capable of modifying the planetary climate. Before moving on to a more specific conception of legitimacy for geoengineering governance, I want to highlight six aspects of the Metacoordination View.

First, this is an elucidation of the concept of institutional legitimacy, not a conception of institutional legitimacy. This general concept of legitimacy can be applied to institutions that rule in the strongest sense – that is, institutions such as the state that claim a monopoly over the use of force for a given jurisdiction. But it can just as coherently be applied to institutions that do not rule even in a weak sense of the word, but instead merely promulgate principles that can be used for coordination across either small or large populations. It may be helpful to think of the concept of institutional legitimacy as a large circle encapsulating smaller circles that articulate conceptions of legitimacy for particular institutions. For example, the particular conceptions of state legitimacy previously surveyed put forth different normative criteria that would count in favour of a state being legitimate. A particular conception for the legitimacy of, say, a national sports league would identify different – presumably, less demanding – normative criteria that would count in favour of the sports league being legitimate. And, similarly, a particular conception for the legitimacy of a geoengineering governance institution would identify normative criteria appropriate for the kind of institution it is. But all of these conceptions of legitimacy – the conception of state legitimacy, the conception of national-sports-league legitimacy, and the conception of geoengineering governance legitimacy – would be encapsulated by the broader concept of institutional legitimacy. The important point to take away is that the general concept of institutional legitimacy says that an institution is legitimate when it merits the kind of standing it needs to function (and it merits that standing when it sufficiently fulfils context-specific normative criteria). A particular conception of legitimacy will then identify and outline the normative criteria appropriate for the kind of institution under consideration. So, while it doesn’t outline specific normative criteria for all kinds of institutions, the Metacoordination View of institutional legitimacy is flexible enough to accommodate various specific conceptions of legitimacy.

Second, the special standing we grant to legitimate institutions can be considered a kind of respect. Thus, if a state is legitimate, it means that the state is worthy of our respect. Similarly, if a sports league is legitimate, it means it is worthy of our respect. And if a geoengineering governance institution is to be considered legitimate, it means it is an institution that is worthy of our respect. Admittedly, ‘respect’ is a complex term with many different variations. Showing respect for a person is certainly distinct from showing respect for, say, the law or another institution. Following Buchanan, we can say that the kind of respect one accords an institution will vary depending upon one’s relationship to the institution. In the case of those who are addressees of the institution’s directives, showing respect need not mean uniformly complying with those directives or regarding oneself as duty-bound to comply with them. Rather, for those who are addressees of the institution, showing respect implies that institutional directives should be taken seriously, independently of the particular content of the directive (Buchanan, 2013, p. 184). In more technical words, addressees of the institution, when showing it proper respect, would take one of its directives as a ‘(defeasible) content-independent, exclusionary reason to comply’ (Buchanan, 2018, p. 74). However, for non-addressees, the appropriate form of respect for the legitimate institution will generally equate to at least some kind of presumption in favour of non-interference.¹⁰

Third, there is a difference between an institution being worthy of respect and an institution actually gaining respect. This is the distinction between normative and descriptive legitimacy. The descriptive or sociological sense of legitimacy refers to the actual attitudes or beliefs of agents (Weber, 1997). For instance, if people believe an institution to be legitimate, their general belief is what confers (descriptive) legitimacy onto the institution – regardless of whether or not it actually ought to be granted such standing. Take North Korea as an example. It could be the case that the North Korean government enjoys descriptive or sociological legitimacy from the standpoint of its people. It is possible that the people of North Korea believe that their government is legitimate. But even if we grant that the people of North Korea believe their government to be legitimate, we can still doubt its normative legitimacy. In other words, we can doubt whether the people actually should consider the government legitimate, that is, whether the government is actually deserving of such a predicate. To bring this back to the Metacoordination View, if an institution is worthy of respect, regardless of whether or not it actually gains the respect it deserves, it has normative legitimacy. Thus, we will be exploring when it is that we should consider a geoengineering governance institution worthy of our respect. However, the Metacoordination View also ‘explains why achieving sociological legitimacy is important from a normative standpoint, namely, because it enables us to empower institutions so that they can provide their distinctive benefits’ (Buchanan, 2018, p. 76). That is, we want an institution to not only be worthy of our respect; we also want it to actually garner our respect, since such actual respect is often necessary for it to deliver its beneficial coordinating functions without relying upon excessive coercion (coercion that could make the institution dangerous). So, the fact that a geoengineering governance institution is worthy of our respect is not enough; we’ll want it to actually garner respect too.

Fourth, the proper standpoint from which to judge the legitimacy of institutions is social rather than individual. When declaring an institution worthy of respect, we are not declaring that it is worthy of my individual respect, in the sense of the first-person singular possessive pronoun. Rather, we are declaring that is worthy of our respect socially, worthy of our respect in the sense of the first-person plural possessive pronoun (Buchanan, 2013, p. 180). This is due to the fact that institutions are attempting to coordinate not my or your action alone, but our action as a group. Whether or not an institution is able to solve the coordination problem it is meant to address will depend upon its success in coordinating the action of the group to a sufficient degree. Therefore, to judge a geoengineering governance institution as legitimate means that it is worthy of the respect of the group, not any particular individual within the group.

Fifth, it is possible for an institution to be legitimate without every single one of its rules or directives being legitimate. That is, a geoengineering governance institution may, in general, be worthy of the respect of those within its purview (and those outside) notwithstanding the fact that certain of its directives are unworthy of respect. This has direct implications for how we ought to respond to a deficient institution. ‘Generally speaking, the proper response to deficiencies in legitimate institutions is reform, not revolution, to attempt to modify it so as to remedy its flaws rather than to scrap it’ (Buchanan, 2013, p. 185). To require that a geoengineering governance institution only be labelled legitimate if every single aspect of its existence is worthy of respect would be too demanding.¹¹ As is commonly said, this would make the Best the enemy of the Good – something I’ll say more about in the penultimate section of the paper.

This leads directly to the final clarification of the concept. Institutional legitimacy is not a bivalent concept, with institutions either being absolutely legitimate or absolutely not. Rather, it is a concept of degree; an institution can be more, or less, legitimate. Thus, the more and/or the better a geoengineering governance institution fulfils the appropriate normative criteria, the stronger the reason we have to consider it worthy of respect – in other words, the more legitimate the institution is. The normative criteria identified as relevant for geoengineering governance will serve as what Rawls called ‘counting principles’ rather than necessary and sufficient conditions (Rawls, 1999, p. 364). ‘[T]he more of them an institution satisfies, and the higher degree to which it satisfies them, the stronger its claim to legitimacy’ (Buchanan & Keohane, 2006, p. 424). There will be institutional arrangements that are closer to the end of the spectrum that characterizes clear legitimacy – we can call this robust legitimacy. And there will be institutional arrangements that are closer to the end of the spectrum that characterizes clear illegitimacy, while nonetheless sufficiently satisfying the relevant criteria to be worthy of respect – we can call this weak legitimacy. Where a geoengineering governance institution falls upon this spectrum will be determined by how well it satisfies the normative criteria that are salient for the kind of institution it is.

3.2. The Normative Criteria

Now, the general concept of institutional legitimacy as adumbrated by the Metacoordination View does not necessarily specify any particular normative criteria. However, there are some things we can say about the normative criteria simply from an analysis of the general concept. First, we know that the normative criteria used in legitimacy assessments will vary depending upon the specific form and function of the institution being assessed. Whereas democracy may be a salient normative criterion for state-like institutions, it may not be entirely relevant when making legitimacy assessments of, say, a national sports league.

Second, we know that the normative criteria will lie somewhere on a continuum between ‘the excessively demanding requirements of full justice or optimal efficacy and the excessively forgiving requirement of bare advantage relative to the non-institutional alternative’ (Buchanan, 2013, p. 193). Moreover, we know that where on this continuum the normative criteria of a given institution fall will depend upon the particulars of the institution and the environment in which the institution is situated. The more important the need for the coordination the institution is providing, the more we should relax our criteria. However, the greater the risk we run by empowering the institution, the more demanding we should make our criteria. There are at least two ways that we can be put at risk by empowering an institution. The first, and perhaps most obvious way, is that the institution could abuse its power. For instance, an institution endowed with significant coercive power could wield that power either through questionable means or to undesirable ends. But, second, we can even be put at risk by institutions that wield mere coordinative power. For instance, an institution setup to coordinate action around geoengineering could suggest that resources be directed towards a particular (perhaps comparatively less optimific) technology or could have a bias in favour of development. In such a situation, even if the institution didn’t back its suggestions with coercive power, even its mere coordinative power could put us at risk of being locked into the development of a suboptimal technology or could even lead us towards objectionable deployment scenarios. The upshot here is that we’ll want to make sure the normative criteria identified as appropriate for judging a given institution as legitimate are sensitive to both the particulars of the institution and the environment in which the institution is situated. This entails that the more we need the coordination the institution can provide and the lower the risk we run by empowering it, the more relaxed we should make the normative criteria; conversely, the weaker the need for coordination and the greater the risk the institution poses, the more demanding the normative criteria ought to be.

Third, we know that the criteria we use to make legitimacy assessments will have to be translated into ‘epistemically accessible standards’ that can serve as proxies for the normative criteria (Buchanan, 2013, p. 193). Imagine that we identify transparency as a normative criterion for geoengineering governance. It may be difficult to agree upon whether or not the abstract normative criterion of transparency is being met. So, in order to coordinate our legitimacy judgements, we could establish a substantive standard relating to the normative criterion of transparency which requires the institution to release its meeting minutes and perhaps its voting record. This would provide a clear point around which addressees of the institution could determine whether or not the criterion of transparency was being fulfilled. Transparency is thus a normative criterion that plausibly can be translated into specific standards for myriad institutions. The point being made here is that if the abstract normative criteria that ought to be used to coordinate legitimacy assessments are incapable of being translated into substantive standards that agents can use to actually inform their attitudes of respect for the institution, the criteria are inadequate (Buchanan, 2013).

One final point to mention again about the normative criteria is that, according to Buchanan, they are not necessary and/or sufficient conditions for an institution being worthy of our respect. Rather, the more of the criteria that are satisfied, and the greater the extent to which they are satisfied, the more legitimate the institution is. Given that institutional legitimacy, as stated previously, is a concept admitting of degree, this makes sense. But I would depart from Buchanan and maintain instead that at least one normative criterion is a necessary condition for any institution to be worthy of our respect – namely, that of comparative benefit.¹² The normative criterion of comparative benefit is, in fact, a necessary condition, and not merely a counting principle. Why is it the case that comparative benefit is a necessary condition and not merely a point in favour of an institution being considered legitimate? The answer is conceptual. Insofar as we have institutions in order to help solve coordination problems and to do so without the costs and inefficiencies associated with a non-institutional alternative, an institution that fails to deliver a comparative benefit is not helping to solve the original coordination problem. Now, it may be the case that institutions that go significantly beyond the minimal satisfaction of the criterion of comparative benefit ought to be considered more legitimate than institutions that barely satisfy such a condition. But when assessing an institution that fails to meet the criterion of mere comparative benefit, we need not even examine the other normative criteria. Any institution that fails the test of comparative benefit ought to be judged illegitimate.

But I should make clear that the comparative benefit criterion is one that applies to the institution, and not to any of its particular policies. Institutions that deliver a comparative benefit on the whole can sometimes issue policies that are comparatively detrimental – this is most clear with patently unjust policies. An institution setup to oversee geoengineering would be making decisions about, most likely, various technologies at various stages of research and development. For instance, a geoengineering governance institution could be tasked with approving or denying field trials across a number of different geoengineering proposals. A particular policy of that institution, on the other hand, could, for example, place a moratorium on field trials of a specific technology (i.e. ocean fertilization). What we want to know is whether the institution as a whole is legitimate – that is, we want know whether the institution as a whole is justified in attempting to coordinate our collective action. And this can be the case even if some of its particular policies are not comparatively beneficial. Again, we don’t want to set our bar of legitimacy so high as to make the Best the enemy of the Good.

4. Normative Criteria for Geoengineering Governance

What, then, are the appropriate normative criteria for an institution overseeing geoengineering? There are two methodological obstacles to proceeding with an adumbration of normative criteria appropriate for a climate engineering regulatory institution. First, without knowing the precise purpose of the institution and the kind of power it has at its disposal to achieve this purpose, nailing down clear and uncontroversial normative criteria for legitimacy assessments is difficult. Not only that, remember that the Metacoordination View sees legitimacy judgements as a social practice. The object of the practice is to reach justified agreement on shared normative criteria and related substantive standards. We’ll need to decide together what the main goal of this institution is and what kind of power it ought to have to achieve such a goal.

What I offer here, then, are not the necessary and sufficient criteria that ought to be used to judge the legitimacy of geoengineering regulation. But given the characteristics of geoengineering technologies, the following proposed criteria seem relevant for assessing whether or not a governance institution is worthy of our respect. Thus, the following criteria are a good place to begin a discourse that will need to gather input from participants from all over the globe. With these words of caution in mind, the following five subsections explore tentative normative criteria for an institution charged with overseeing climate engineering. An institution empowered to oversee climate engineering ought to sufficiently satisfy the normative criteria of: (1) comparative benefit; (2) accountability; (3) transparency; (4) substantive justice; and (5) procedural justice.

4.1. Comparative Benefit

The first normative criterion of comparative benefit is actually a necessary condition for any justified claim to legitimacy. As mentioned previously, a climate engineering regulatory institution that fails to meet the criterion of comparative benefit ought not to receive our respect. This should be fairly straightforward. Insofar as we are not better off with a climate engineering institution than we would be without it, we have no reason to accord it the respect it would require to function.

There are a variety of ways in which a geoengineering governance institution could make the world a better place than the non-institutional alternative. For instance, if the institution made the threat of harmful unilateral deployment less likely, this would count as a benefit, since surely a world in which the threat of harmful unilateral deployment is less likely is better than a world in which it is more likely. Similarly, if the institution aided coordinated research so that resources could be more efficiently used, this would count as a comparative benefit. Or if the institution furthered other things we value – such as substantive and procedural justice – this too would count as a benefit compared to the non-institutional alternative. Conversely, if an institution increased the likelihood that geoengineering would be deployed prematurely, or exacerbated global injustices (again compared to the non-institutional alternative), then this would count against its fulfilment of the criterion of comparative benefit.

Now, the idea of comparative benefit has two readings. The first is the thought that the institution in question should lead to greater social benefit than the non-institutional alternative; that is, the world in which we have an institution to oversee climate engineering must be comparatively better than a world in which no such institution exists, ceteris paribus. I call this the non-institutional alternative reading. The second reading of comparative benefit is also counterfactual, but is not limited to merely a non-institutional alternative. Under this reading of the comparative benefit criterion, the climate engineering regulatory institution must deliver a social benefit compared to possible alternative institutions as well. This second – more demanding – reading, which I call the institutional alternative reading, has four conditions. A climate engineering institution will fail the comparative benefit criterion if there is an institutional alternative that: (a) provides significantly greater benefits, (b) enjoys similar or even greater feasibility, (c) is accessible without unacceptable transition costs, and (d) sufficiently fulfils the other normative criteria (Buchanan & Keohane, 2006, p. 422). Thus, a climate engineering institution will be worthy of our respect if it fulfils the comparative benefit criterion under the non-institutional alternative reading. But insofar as there is an alternative institution that satisfies conditions (a) – (d) mentioned above, it is that institution that should be coordinating our action around geoengineering. That is, a climate engineering regulatory institution ought to fulfil the institutional alternative reading of the criterion of comparative benefit.

4.2. Accountability

A second normative criterion for the legitimacy of geoengineering governance is the idea of accountability. Remember that empowering an institution to help us solve our coordination problem carries with it risks associated with the power the institution wields. In order to make sure that power is being used in the way we intend, we will want institutional agents and the institution itself to be accountable. The norm of accountability has, according the Buchanan and Keohane, three elements:

First, standards that those who are held accountable are expected to meet; second, information available to accountability holders, who can then apply the standards in question to the performance of those who are held to account; and third, the ability of these accountability holders to impose sanctions – to attach costs to the failure to meet the standards (Buchanan & Keohane, 2006, p. 426).

To these three I’d like to add a fourth element that would clarify the appropriate group of accountability holders. It wouldn’t be hard to imagine a scenario in which the three elements above are properly accounted for, and yet the group of accountability holders is insufficiently restricted so as to render the fulfilment of the first three elements hollow.

However, while accountability is a normative criterion that a climate engineering regulatory institution ought to satisfy if it is to warrant our respect, it is also possible for such an institution to be overly accountable in at least two ways. First of all, while we want institutional agents to be accountable to the right group, we also want them to exercise their own judgement and expertise to a certain degree. Presumably, institutional agents are better informed than the general group of accountability holders (Maskin & Tirole, 2004). If institutional agents are to act as responsible representatives for the accountability holders and not merely pander to their wishes, they will require a certain (perhaps small, perhaps larger) degree of insulation.¹³ This leads to the second point, which is that – while they certainly should be part of the group of accountability holders – it is impossible to allow future generations to hold current institutional agents accountable. If the current group of accountability holders has too much sway over institutional agents, they can tilt the institution’s functioning in their favour, disregarding important duties the current generation may have to future generations. This is what Stephen Gardiner has termed ‘the tyranny of the contemporary’ (Gardiner, 2011, ch. 5). For these two reasons, we will want to design our accountability mechanisms with some caution.¹⁴

4.3. Transparency

A third normative criterion relevant to geoengineering governance is that of transparency. Transparency is commonly defined as ‘the principle of enabling the public to gain information about the operations and structures of a given entity’ (Etzioni, 2010). While some have highlighted the costs associated with significant transparency (Bannister & Connolly, 2011), the norm has at least three significant functions. The first significant function correlates to the second element of accountability highlighted above. Transparency aids accountability holders in accessing information useful for determining how they should exercise their power of holding institutional agents accountable. But, secondly, broad transparency gives non-addressees of an institution the ability to analyse the institution’s functioning and enables them to ‘contest the terms of accountability’ (Buchanan & Keohane, 2006, p. 427). Finally, transparency can serve as a kind of public justification and create more trust for institutions and the agents that occupy senior positions (O’Neill, 2006). And it is worth mentioning, given climate engineering’s controversial nature, that transparency will be of the utmost importance in garnering descriptive legitimacy, aiding people in reaching concurrent judgements as to whether or not the institution is worthy of our common respect. However, given the complex and highly technical nature of climate change and especially geoengineering, true transparency will require that the information provided be accessible to both addressees and non-addressees of the institution. This information ought to be accessible not merely in the sense that it is available, but accessible in the sense that it is intelligible to institutional outsiders as well.

4.4. Substantive Justice

A fourth normative criterion that a climate engineering regulatory institution ought to sufficiently satisfy in order to warrant our respect is captured by the idea of substantive justice. Institutions can deliver certain benefits and often carry with them certain burdens as well.¹⁵ This is certainly true of an institution overseeing geoengineering. It is universally recognized that geoengineering has the potential to create novel distributions of climatic benefits and burdens.

In the abstract, the idea of substantive justice refers to just substantive outcomes, or just distributions of the benefits and burdens produced by a geoengineering governance institution. We generally assume that everyone would like to secure as many of the benefits and as few of the burdens associated with the institution’s functioning. We can imagine people having certain claims to some of the benefits and certain claims against the burdens associated with geoengineering governance. Substantive (or distributive) justice obtains when there is a proper balance between the competing claims to those benefits and against the accompanying burdens. While a perfect balance between these competing claims would have to obtain for us to determine the geoengineering governance institution was perfectly just, remember that legitimacy assessments admit of degree. We don’t want to require a climate engineering institution to be perfectly just in order for us to grant it the kind of standing it needs to function. This is for at least the following two reasons. First, we may need such an institution in order to make progress on justice. So, ‘refusing to regard an institution as legitimate unless it is fully just would be self-defeating from the standpoint of justice’ (Buchanan, 2018, p. 55). Second, requiring that the institution be fully just in order for it to be considered legitimate would conflate ‘legitimacy’ and ‘justice’ and impoverish our moral lexicon. Thus, when making legitimacy assessments, we’ll have to determine whether or not a geoengineering governance institution is sufficiently satisfying the demands of substantive distributive justice to be worthy of our collective respect, while nonetheless recognizing that it need not go all the way in satisfying such demands.

Exactly what would count as a just substantive outcome with respect to geoengineering (or climate change in general) is subject to ongoing debate. But there is a modest claim about substantive justice and geoengineering that seems to be on solid ground.¹⁶ The fact that the least well off members of the global community have (a) contributed the least to the genesis of climate change, (b) have benefited the least from previous actions that have brought about climate change, and (c) have the weakest ability to respond to the burdens of climate change all point to the same conclusion regarding the distribution of benefits and burdens related to geoengineering. The three facts listed above lend support to the following conclusion: if a geoengineering governance institution is to minimally meet the normative criterion of substantive justice, then the distribution of benefits and burdens engendered by such an institution should (probably heavily) favour the least well off members of the global community. Exactly how much that distribution should favour them is difficult to say.¹⁷ But what seems clear is that if a geoengineering governance institution were to lead to a world in which the least well off members of the global community were forced to shoulder even greater burdens than they already are in the face of anthropogenic climate change, such an institution would fail to satisfy the normative criterion of substantive justice.

4.5. Procedural Justice

The final normative criterion I put forward as applicable to geoengineering governance is that of procedural justice. As was noted in the previous subsection, we will want to make sure that our governance institution is conforming to norms of substantive justice. But the substantive distributive outcome with respect to geoengineering is not all we care about.¹⁸ We are also concerned with how it is that we go about making decisions regarding research, development, and deployment. That is, along with substantive justice, we also care about procedural justice. The difference between substantive and procedural justice can be understood as the difference between fairness in the result and fairness in the process, respectively (Rawls, 2005, p. 424; Hampshire, 1993). If we think of substantive justice as a proper balance between the competing claims to the benefits and burdens associated with the institution, we can think of procedural justice as a proper balance between the competing claims to participate in the decision-making process that will determine the outcome. Just as in the case of substantive justice, the normative criterion of procedural justice can be satisfied to a greater or lesser degree. The more procedurally just the geoengineering governance institution is, the more confident we can be that we are justified in granting it the kind of respect it needs to coordinate our action. The more procedural injustice that exists within the institution – that is, the more there are people with claims to participation that are being ignored – the more that will count against us collectively appraising the institution as legitimate.

As in the case of substantive justice, it is difficult to say exactly what would count as a just procedure when it comes to making decisions about geoengineering. But in order for such a procedure to minimally fulfil the normative criterion of procedural justice, it would need to go some way towards providing what I call fair terms of inclusion and fair terms of participation.¹⁹ By fair terms of inclusion, I mean that all those with legitimate claims to participate in the decision-making process are included. By fair terms of participation, I mean that all those included in the decision-making process are included on justifiable terms. Again, this doesn’t provide us with specific instructions for securing procedural justice in geoengineering governance. But it does help us pick out clear instances of procedural injustice. For example, given the potentially beneficial and potentially catastrophic effects geoengineering could have on, say, those residing within small island states, any process that failed to include these people and allow their voices to be heard on reasonable terms of participation would be procedurally unjust (regardless of the substantive outcome).

Finally, whatever formal procedure ends up guiding the decision-making process around geoengineering, there is a good reason to think that such a procedure should include public participation in some way, shape, or form.²⁰ From a purely theoretical standpoint, public participation may not be strictly necessary provided that good representatives of all those with legitimate claims are included in the decision-making process. But including public participation in the decision-making process with respect to geoengineering would be valuable on two fronts. First, we know that even the best of representatives do not always fully represent the interests of their constituents – a problem exacerbated at the international level where (a) representatives of democratic states are often too far removed from their constituents to adequately represent their interests, and (b) representatives of non-democratic states often do not represent the interests of their constituents at all. Including public participation could go some minor way towards addressing this issue. Second, embedding public participation of some kind into the decision-making process can help a geoengineering governance institution secure descriptive or sociological legitimacy more easily. That is, having members of the public included in the institution could help a geoengineering governance institution actually garner the respect it needs to perform its function. And given geoengineering’s controversial nature, a governance institution will need all the help it can get in gaining the actual support of the public.

5. Conclusion

A quick recap is probably in order. This paper began by noting that, given geoengineering’s potential to deliver both great benefits and great harms to various populations throughout the world, continued research, development, and any possible deployment ought to be overseen by a legitimate governance institution. And, in order for such an institution to justifiably regulate geoengineering, it will have to be a legitimate institution. While perhaps everyone agrees that geoengineering governance ought to be legitimate, an agreed upon conception of legitimacy has yet to emerge. I posited that standard conceptions of (state) legitimacy may be inadequate to determine the legitimacy of a geoengineering regulatory institution. This is because geoengineering governance is something that will take place on the international stage, and international institutions vary dramatically in their characteristics and functions. Given this diversity, we should adopt the broader concept of institutional legitimacy, and then develop specific normative criteria that each kind of institution ought to fulfil. That is the route this paper has taken, identifying five normative criteria that could be appropriate for making legitimacy assessments of an institution setup to oversee climate engineering.

Now, it’s clear that these normative criteria are somewhat ambiguous. And we need to remember that there may be other normative criteria that are also salient to legitimacy assessments of a geoengineering governance institution.²¹ Furthermore, even upon reaching a justified agreement regarding the appropriate normative criteria for legitimacy assessments, we still need to explore how these normative criteria can be translated into substantive standards. These standards are to serve as proxies for the normative criteria and, thus, they will need to fall between two points. On the one hand, they need to capture the normative criteria as closely as possible. On the other hand, they need to be accessible to both addressees and non-addressees of the institution, and they need to be standards that could be widely accepted. Exactly how these standards are instantiated and the judgement of whether or not they are met must be done socially.

Some might find this conclusion unsatisfying for perhaps two reasons. First, one might think that geoengineering governance is going to require significant coercive power in order to effectively coordinate our action (which may be the case), and so such a loose conception of legitimacy is insufficient. But remember what was said at the end of Section 2. If a geoengineering governance institution is wielding significant coercive power, then it, just like states, ought to meet significantly more demanding normative criteria to be considered legitimate. Second, what we want from a conception of legitimacy, one might argue, are explicit necessary and sufficient conditions that unmistakably spell out exactly when it is that an institution is legitimate and when it lacks legitimacy. The Metacoordination View, clearly, does not do that. But to the extent that we recognize legitimacy assessments as social practices, what the Metacoordination View does do is provide us with the right framework from which to engage in that practice. The hope is that this paper has clearly explicated that general framework, and done some modest work towards generating a conception of legitimacy that will allow us to justifiably coordinate our assessments regarding geoengineering governance.

Acknowledgments

For helpful comments on earlier drafts of this paper, I’d like to thank Darrel Moellendorf, Catriona McKinnon, Clare Heyward, and especially Steven Gardiner and Augustin Fragnière. I’d also like to thank the participants of the following conferences and workshops for fruitful discussions: ‘Geoengineering and Legitimacy Conference’, University of Washington, Seattle (November, 2017); ‘Science, Technology, and Public Policy Workshop’, Harvard University (November, 2016); ‘Climate Scholars’ Conference’, University of Reading, England (November, 2016).

Research for this paper was supported by the DFG-funded Excellence Cluster, The Formation of Normative Orders (Goethe-Universität Frankfurt), and the Bernheim Foundation’s Hoover Chair for Social and Economic Ethics (Université catholique de Louvain)

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. Throughout this paper, I’ll be loose with language and refer to ‘geoengineering’ or ‘climate engineering’ broadly. But it is important to note that there are significant and morally relevant differences between the different proposals (Heyward, 2013; Jamieson, 2014, ch. 7).

2. For an argument against the prospect of unilateral deployment, see (Horton, 2011).

3. Note that one need not think that we should move forward with research and development to endorse this claim. Even if one thinks research should be abandoned today, this conditional claim could still be accepted. And note that by highlighting the importance of legitimate governance for geoengineering, no claim is being made about its relative importance. There are many other emerging technologies (i.e. gene drives, artificial intelligence, etc.) that deserve attention from normative theorists in order to help guide the development of legitimate regulatory institutions. The focus of this paper, however, is on the cluster of emerging technologies known as geoengineering.

4. For notable exceptions, see (Wong, 2016), (Morrow, Kopp, & Oppenheimer, 2013), (Morrow, 2017b), and (Gardiner & Fragniere, 2018). The conception of legitimacy I outline later in the paper is relatively similar to the conception outlined by Morrow, Kopp, and Oppenheimer – which shouldn’t be surprising given that we’re all relying upon work done by Buchanan and Keohane (2006). That being said, there are also significant differences. For example, Morrow, Kopp, and Oppenheimer would claim that if a climate engineering regulatory institution is legitimate that it (a) has a right to rule and (b) that those within its purview are under a duty to obey. The Metacoordination View of Buchanan’s that I am relying upon claims instead that a legitimate institution is (a) justified in coordinating our action and (b) that those within its purview have a defeasible content-independent reason to comply, but not a duty to obey. Morrow, Kopp, and Oppenheimer also claim that the consent of all democratic states would be necessary but not sufficient to confer legitimacy. But what if all democratic states, except France, consented? Contrary to the view put forward by Morrow, Kopp, and Oppenheimer, the Metacoordination View could still consider such an institution legitimate (though this would of course depend upon how well it is satisfying the relevant normative criteria).

5. For explanations of the distinction between concepts and conceptions, see (Dworkin, 2013, p. 167; Hart, 1998, p. 160; Korsgaard, 2003, pp. 115–16; Rawls, 1999, p. 5).

6. Though, I should point out that even procedural conceptions like Rawls’ recognize substantive constraints. See (Rawls, 2005, p. 428). I thank Augustin Fragnière for pointing this out to me.

7. I have serious doubts about the voluntarist conception of legitimacy for even state-like institutions, but I set that concern aside here.

8. I don’t mean to suggest that merely being beneficial confers legitimacy onto an institution. Rather, I mean to suggest that we should recognize what we would be giving up by labelling all international institutions as illegitimate. We need international institutions both to make progress on the legitimacy of domestic institutions and on concerns of justice more globally. Thus, to label international institutions as illegitimate would be counterproductive to the causes of making sure that political power is only exercised legitimately and that it serves the cause of justice. See my fifth point in Section 3.1.

9. It should be noted that we are not looking for arbitrary or unjustified granting of this standing. As Buchanan writes, ‘The goal is not simply agreement that we should support the institution, but rather agreement that is morally appropriate’ (Buchanan, 2013, p. 179).

10. For a slightly different understanding of what it means to show an institution the kind of respect it needs to perform its function, see (Adams, 2018).

11. It could also be noted that sometimes institutions are required in order to further the cause of justice (the natural duty of justice). It would, ironically, undermine the cause of justice to require that all institutions be perfectly just in order for them to be considered legitimate.

12. Buchanan has altered his view and now agrees that comparative benefit is a necessary criterion. See (Buchanan, 2018, p. 60). But the reason for this isn’t fleshed out. As I explain below, I think construing the comparative benefit criterion as necessary is conceptually entailed by the Metacoordination View.

13. Take the case of genetically modified mosquitoes, for instance. Imagine that politicians, speaking with informed scientists, recognize that genetic modification is the all-things-considered best way to stop the spread of infectious disease. If there is too much accountability, they may be reluctant to pursue such a policy option because their constituency may be averse to genetic modification. Adrienne LaFrance writes, ‘A poll conducted by the Annenberg Public Policy Center in February found more than one-third of Americans believed genetically modified mosquitoes were to blame for the spread of Zika. (They’re not.)’ (LaFrance, 2016).

14. We will want to avoid ‘narrow accountability’ in favor of a more broad accountability. Narrow accountability is ‘accountability without provision for contestation of the terms of accountability’. Broad accountability, however, would allow for public discourse and require a fifth element for the criterion, namely, that the terms of accountability themselves be publicly justified and open to revision. See (Buchanan & Keohane, 2006, p. 427).

15. Not only can institutions actually deliver certain benefits and burdens, they can also affect the possibility of certain benefits and burdens coming to fruition. Whether an institution has actually produced particular benefits or burdens can often only be assessed ex post. For instance, if we assume that a geoengineering regulatory institution would deliver a comparative benefit, we should grant it standing ex ante. But if it fails to deliver such benefits and is actually putting us at risk of catastrophe, then we should collectively revoke its standing. We could also recognize that an institution could put us at risk ex ante, in which case we should also withhold a positive legitimacy assessment. See Section 4.1 for more on risk and comparative benefit.

16. While it is not completely free from controversy, the following claim enjoys rather broad support. See (Gardiner, 2004, 2016a; Horton & Keith, 2016; Morrow, 2017a).

17. There are two ways to understand the claim that the least well off should be favoured most heavily: in an absolute sense and in a relative sense. I think that substantive justice would require any use of climate engineering to favour the least well off in both a relative and an absolute sense. This entails that if the use of climate engineering were to make everyone worse off, but were to make the least well off less worse off than others, that it would not further the cause of justice. That is, I am not entertaining ideas of levelling down. I thank the editors for bringing this to my attention.

18. For more about why we should care about more than the mere distributive outcome with respect to climate engineering, see the contributions of Marion Hourdequin, Patrick Taylor Smith, and Christopher Preston to this special issue.

19. While he doesn’t use these terms and focuses more on what I call fair terms of inclusion, these terms are influenced by Charles Beitz’s discussion of political equality. See (Beitz, 1989).

20. For more on what form that public participation could take, see: Callies (2018) and Jinnah (2018).

21. For other papers that address some salient normative criteria for geoengineering governance, see (Jamieson, 1996; Morrow et al., 2013; Rayner et al., 2013) along with the contributions of Christopher Preston, Kyle Powys White, and Marion Hourdequin to this special issue.