Toward Legitimate Governance of Solar Geoengineering Research: A Role for Sub-State Actors

ABSTRACT

Two recently proposed solar radiation management (SRM) experiments in the United States have highlighted the need for governance mechanisms to guide SRM research. This paper draws on the literatures on legitimacy in global governance, responsible innovation, and experimental governance to argue that public engagement is a necessary (but not sufficient) condition for any legitimate SRM governance regime. We then build on the orchestration literature to argue that, in the absence of federal leadership, U.S. states, such as California, New York, and other existing leaders in climate governance more broadly have an important role to play in the near-term development of SRM research governance. Specifically, we propose that one or more U.S. states should establish a new interdisciplinary advisory commission to oversee and review the governance of SRM research in their states. Centrally, we propose that state-level advisory commissions on SRM research could help build legitimacy in SRM research decisions through the inclusion of, at minimum: meaningful public engagement early in the research design process; an iterative and reflexive mechanism for learning and improving both participatory governance mechanisms and broader SRM governance goals over time; as well as mechanisms for adaptation and diffusion of governance mechanisms across jurisdictions and scales.

Introduction

It is increasingly clear that, even if fully implemented, the first round of national pledges to address climate change will not in themselves be enough to meet the temperature targets under the 2015 Paris Agreement (Chen & Xin, 2017; Rogelj et al., 2016). This reality has led to increased scholarly attention, and more recently attention from policy makers and civil society more broadly, on the possibility of developing climate engineering technologies.

Climate engineering is an umbrella term encompassing two different categories of technology – carbon dioxide removal (CDR) technologies that might remove large amounts of carbon from the atmosphere and hold it in storage, and solar radiation management (SRM) or albedo modification technologies that might dampen temperatures by reflecting some amount of incoming solar radiation back into space. We focus here on SRM technologies, and in particular on the exploratory roles that sub-state (that is, sub-national) actors ought to play in the near-term governance of SRM research. Our focus is on the SRM governance roles that might be played by states like California, Washington, and New York in the United States, since this is the country from which the best-known and most advanced proposals for open-air experimentation related to SRM are emanating. The article concludes by advocating for and sketching out design considerations to guide development of one or more sub-state ‘Advisory Commissions on SRM Research,’ as a reasonable and manageable near-term step that a state like California, Washington, or New York might take.

Despite their early stages of investigation, a rich body of scholarship has already emerged surrounding SRM technologies.¹ Two notable strands of this literature explore the ethical and governance dimensions of SRM development and potential deployment. We briefly, here, outline and mine this SRM-focused literature for lessons about legitimate governance (our chief theoretical concern in the following pages is the legitimacy of governance efforts). In the next section we turn to an examination of a number of different strands of literature from political science, international relations, and science and technology studies, for additional insights about what it takes to craft legitimate forms of governance. The paper then translates the insights from this range of complementary literatures into a concrete and actionable proposal for near-term sub-state action.

As a starting point, we note that there already exists a particularly rich literature surrounding the ethical and normative dimensions of climate engineering. Scholars have unpacked, for example, the moral issues raised by climate engineering, the normative implications of technological lock in on future generations, the conditions under which climate engineering could be considered morally permissible, and the ethical frameworks embedded in key climate engineering reports (Gardiner, 2011; Jamieson, 1996; McKinnon, in press; Preston, 2013). Particularly relevant to our work here are ethical arguments that political legitimacy in decision-making are sufficient to guide SRM research (Morrow, Kopp, & Oppenheimer, 2013). Still others have questioned whether humans should be meddling with the climate system via large-scale technological interventions at all (Hamilton, 2013), raising questions about the legitimacy of the entire enterprise, and who should be entrusted with decisions surrounding the shaping of the ‘Synthetic Age’ (Preston, 2018).

Daniel Callies’ contribution to this special issue pushes forward these conversations in proposing a set of normative criteria, which he argues any institution ought to have if it is to legitimately coordinate action around SRM (Callies, this issue). Callies argues that any such institution governing SRM, must at a minimum confer: (1) comparative benefit; (2) accountability; (3) transparency; (4) substantive justice; and (5) procedural justice. His institution-specific, and ‘necessary but not sufficient’ approach to developing legitimate institutions is worth underscoring here, as it aligns with our own view. That is, although we can identify some of the necessary ingredients, institutional legitimacy cannot be derived from a recipe. Rather, on the basis of an ingredient list policy actors can begin to develop institutions, but those undertaking such work must always be reflexive and willing to pivot to the needs of specific contexts and issues as institutions evolve.

Due to the inherent transboundary impacts of these technologies if ever deployed, scholars of global governance and international law have also been active in advancing insights into and proposals for the governance of SRM. Several have put forth specific governance proposals at the international and regional levels (e.g. Armeni & Redgwell, 2015; Hubert, 2017; Lloyd & Oppenheimer, 2014; Nicholson, Jinnah, & Gillespie, 2018; Parson & Ernst, 2013; Reynolds, Jorge, Contreras, & Sarnoff, 2018; Sugiyama et al., 2017), debated various framings and tradeoffs between climate engineering and traditional mitigation and adaptation measures (Horton, 2015; Reynolds, 2015), and argued that expert bodies are already playing an important role in ‘de facto’ governance across scales (Gupta & Moller, in press). In response to calls for more engagement by scholars of international relations (Horton & Reynolds, 2016), recent work has also explored, for example, how international relations theory might contribute to governance design (Jinnah, 2018) and the collective action implications of geoengineering (Sandler, 2018).

With a few notable exceptions (e.g. BPC, 2011; Craik, 2017; Craik, Blackstock, & Hubert, 2013; Winickoff & Brown, 2013; Mahajan, Tingley, & Wagner, 2018; US GAO, 2010; Stilgoe, Owen, & Macnaghten, 2013), the role of domestic politics and institutions for SRM governance has received comparatively little attention. This is surprising given the broader devolutionary trend in the climate governance literature with, for example, the increased attention to city-level governance (e.g. Gordon, 2013), and the fact that some federal government agencies and departments in a range of countries have begun to fund small amounts of SRM research, though the bulk of this research consists of indoor modeling studies and social scientific research.

We aim in this paper to fill this gap by outlining a particular role for sub-state actors in laying the initial governance foundation for SRM research. Although research governance may be necessary at the global scale as well, governance at the sub-state level is more tractable in the near-term given the nascent nature of formal climate engineering governance, its contentious character among many publics, and the proactive role that these actors are already taking on climate change issues, including as related to climate engineering. Moreover, privately-funded small-scale outdoor experiments are currently planned/being planned in U.S. states, often without triggering national or international oversight. Early attempts at governing research at sub-state scales can serve as sites of experimentation and contribute to collective social learning, if designed carefully, and with this goal in mind (Stilgoe, 2016).

Our other concern in this paper is with the legitimacy of governance efforts. Joining other governance scholars who are centrally interested in the intersection between issues of equity and governance in this space (e.g. Burns & Flegal, 2015; Flegal & Gupta, 2018; Svoboda, Buck, & Suarez, 2018), this paper explores how various strands of governance theory can inform the design of legitimate political institutions. To be clear, developing legitimate political institutions cannot be achieved by applying some predetermined formula. Nor are the prescriptions we suggest here alone sufficient to develop legitimate political institutions. However, in looking across several strands of governance theory, there are certain practices, centrally, participatory ones, that can contribute to the development of legitimate political institutions. We are not so naive to think that any participatory process could or would be successful in this regard. To the contrary, participatory processes done poorly can, in fact, undermine efforts to develop political institutions that enjoy high degrees of legitimacy. As such, this paper draws from governance theory to propose some initial steps towards development of one political institution that could, if designed carefully, begin to build norms and mechanisms for broad stakeholder engagement in SRM research governance.

Specifically, building on previous work, this paper makes a policy recommendation to advance a slightly revised version of one of Nicholson et al.’s SRM governance objectives. Namely, we are proposing here policy mechanisms for soliciting active and informed public and expert community engagement with a view towards improving the quality of research conducted, especially its responsiveness to societal goals and norms (2018).² This objective is rooted in the literature on the role of public engagement in climate policy, including – although not exclusively – for SRM (see e.g. Bellamy, Chilvers, & Vaughan, 2016; Corner, Parkhill, Pidgeon, & Vaughan, 2013; Macnaghten & Szerszynski, 2013; Pidgeon, Parkhill, Corner, & Vaughan, 2013; Winickoff, Flegal, & Asrat, 2015).

The rationales for public engagement are not always explicit, but have been described in three categories: instrumental, normative, and substantive (Fiorino, 1990). The instrumental rationale is principally about staving off public resistance, while the normative rationale holds that democratic ideals require that potentially affected parties have a say in scientific decision-making, in accordance with principles of transparency, informed consent, and political legitimacy (NASEM, 2016). The substantive rationale holds that experts do not have a monopoly on expertise and concerns relevant to research, and therefore broader participation can enhance the quality – and legitimacy – of science itself. To the extent that forms of public engagement can open up expert discourse, these efforts can help to facilitate ‘honest brokering’ (Pielke, 2007) of policy options and ensure that parties with potentially quite different worldviews have an opportunity to be heard and responded to by each other (Heyward & Rayner in Heazle & Kane, 2015). Under some circumstances, the interplay of divergent perspectives enabled by robust public engagement can lead to a kind of ‘distributed technology assessment’ and/or social learning (Rayner, 2004) in ways that may contribute to the perceived legitimacy of institutions; as we note above, however, in poor circumstances it can lead to social conflicts, misallocation of resources, and/or the scientization of politics (Rayner, 2004).

To summarize, the goal here, then, is to develop a concrete recommendation for how sub-state actors can promote meaningful public engagement in order to construct legitimate SRM governance. The remainder of the paper proceeds as follows. The next section provides a theoretical foundation for our recommendations on participatory SRM governance. Picking up on the brief introductions to these topics above, we detail how insights from the literatures on legitimacy in global governance, responsible innovation, and experimental governance support central elements of our subsequent policy recommendation that U.S. states should spearhead experimental governance initiatives that are rooted in meaningful public engagement. We then turn to the recommendations themselves. Centrally, we propose that U.S. states establish one or more expert advisory bodies to orchestrate a set of experimental research governance objectives that are deeply rooted in meaningful public engagement surrounding SRM research governance goals. Our conclusions and recommendations reinforce the rationale for such a Commission and point again to the unique leadership role that U.S. states can play in this increasingly salient arena.

Theoretical Foundations for Legitimate Governance of SRM Research

In this section, our focus is on the legitimacy of SRM governance efforts. We look across a number of different strands of governance literature to glean insights about the links between meaningful public engagement and the legitimate governance of something as complex and contentious as SRM research. Although we focus on governance of SRM research, we recognize that other science policy questions – including around mitigation efforts – might also benefit from greater attention to the construction of legitimacy and public engagement. Moreover, we recognize the somewhat artificial separation between research and deployment in the context of SRM. The need for governance of the former is rooted in the risks and uncertainties surrounding potentiality for the latter. We focus on near-term governance of research because activities in the early stages of technological exploration can help to define and enable socially acceptable research, constrain or halt potentially dangerous or undesirable research, begin establishing institutions to manage these technologies if they are ever deployed, and to shape the trajectories of technological development while these technologies are inchoate and therefore susceptible to social steering. Our goal here is to derive a set of design considerations from the broader governance literature that can guide sub-state actors in moving forward with SRM research governance.

To be clear, we do not mean ‘legitimate’ in the narrow sense that public engagement or the development of a governance architecture are a means to rubber-stamp a research effort. Rather, as we explain in some detail below, legitimate research requires, among other things, opportunities for public engagement and feedback that could, ultimately, hold a particular research pathway to be illegitimate or undesirable in some way. That is, in advocating for research governance, we are including the possibility that the mechanisms of governance might themselves result in a closing off of all SRM research or particular avenues for SRM research.

Therefore, we draw insights here from various strands of the governance literature to inform the development of SRM research governance. That is, we review those literatures to inform the design elements of our proposed governance framework that is rooted in principles of transparency, participation, and learning over time. We demonstrate through this review how this multidisciplinary governance literature supports our assertion that there are essential baseline conditions for any legitimate SRM research governance scheme. Specifically, we discuss below the literatures on legitimacy in global governance, responsible innovation, and experimental governance in order to theoretically frame our core policy recommendation in the subsequent section. We further draw from the orchestration literature to explain how sub-state actors, such as U.S. states, can coordinate governance in this space in the near-term while their official political positions on SRM are still undetermined. We argue for an experimental approach to SRM research governance that takes seriously public engagement, social learning, and is orchestrated, in the near-term, by U.S. states. As highlighted by the 2018 Intergovernmental Panel on Climate Change’s (IPCC) special report on global warming of 1.5 degrees, much of the literature on SRM governance appears as ‘commentaries, policy briefs, viewpoints and opinions’ (2018, p. 52). As in our previous work (Nicholson et al., 2018; Chhetri et al., 2018), we seek here to delve into the governance literature to develop a theoretically-derived policy proposal for SRM research governance.

The global governance literature on political legitimacy is instructive in constructing a foundation for SRM research governance. Distilling from this literature a definition that is relevant to the SRM research governance context, we can say the following: political legitimacy refers to the acceptance of a governance institution by its stakeholders who are willing to defer to the decisions of that body in lieu of making their own decisions on specific matters (Bernstein & Cashore, 2007; Bodansky, 1999; Cashore, 2002; Esty, 2006). The conditions under which political actors would be willing to defer in this way are directly related to elements of process in institutional design, and are a central topic of analysis in the literature on political legitimacy.

For example, Bernstein and Cashore (2007) analysis of non-state market driven governance schemes offers some important insights into how elements of institutional design can contribute to (not guarantee) the development of politically legitimate institutions. Importantly, they demonstrate that political legitimacy is often something that institutions work towards rather than something that is inherent in their initial design. They highlight the importance of shared norms, such as transparency and democratic participation, in providing the foundation of political legitimacy of governance systems, but underscore that norms continue to evolve in politically legitimate systems. Specifically, they argue that ‘what is “good” is often precisely the issue to be worked out within politically legitimate arenas’ (p. 355). That is, built-in processes for learning within governance institutions is central to developing legitimacy over time. This can be done through, for example, forums/institutions for expert discussion, which allow for debate, critique, and learning, and building of shared databases of experience (Bernstein and Cashore, 2007, p. 362–3).

In keeping with a point that we underscored in the opening section above, Bernstein has further argued elsewhere that specific determinants of political legitimacy cannot be developed a priori from a checklist (2011). Rather, he argues that stakeholders themselves must identify the specific determinants of legitimacy through interactions over time (2011). Interestingly, Bernstein further notes that understanding the construction of legitimacy frameworks in this way further helps to explain why, in some cases, legitimacy determinants for some institutions can ultimately be dysfunctional for achieving the institution’s core purpose (p. 43). This suggests that in designing an SRM research governance framework that strives for political legitimacy, compromises may have to be made between legitimacy and the speed with which SRM technological development proceeds, at least in the near-term. We come back to this point and discuss implications for our core recommendation in the concluding section.

The literature on science and technology studies (STS) also provides guidance in the quest to move toward legitimate experimentation and governance of SRM. First, STS research suggests that legitimacy is itself a contested concept, and that experts do not have a monopoly on its definition. Rather than orient research toward pre-defined normative goals, STS-inflected frameworks, such as responsible research and innovation (RRI), anticipatory governance, real-time technology assessment, and constructive technology assessment, seek to open up the question of what is good/desirable for broader democratic negotiation and debate, alongside questions such as, ‘what is the purpose [of research]; who will be hurt; who benefits; and how can we know?’ (Jasanoff, 2003, p. 240) As Jack Stilgoe points out, viewing physical science or engineering experiments themselves as socially constructed and determined, at least in part, ‘challenges the attempt to hermetically seal [them] from public scrutiny’ (Stilgoe, 2016). In its focus on anticipation, inclusion, reflexivity, and responsiveness, RRI underscores the importance of ‘experimenting with experimentation,’ considering who should be involved in the definition and conduct of experiments themselves.³

As with the global governance literature on political legitimacy, STS literature also underscores that participation through public engagement is an important, although insufficient, element in legitimate SRM research. The substantive rationale for public engagement described earlier (which holds that experts do not have a monopoly on the kinds of expertise and concerns relevant to research, and therefore broader participation can enhance the quality of science itself), although not explicitly discussed in terms of political legitimacy above, comports with Bernstein’s (2011) argument that stakeholders themselves must define the determinants of political legitimacy in any given context, and Bernstein and Cashore’s call for forums for presentation, discussion, and critique of expert’s proposals/plans/etc. by broader stakeholders. Far from a panacea for the legitimacy of science and technology, public engagement has itself been critiqued in STS literature, both with regard to inputs (Lovbrand et al. point out that engagements are often only seen legitimate for those directly involved in them) (2011, p. 483); and outputs: engagements sometimes reinforce existing power structures, and do not meaningfully impact governance (Scharpf, 1999; Van Oudheusden, 2011).

While public participation can be pursued in light of concerns about the legitimacy of decision-making, it is important to underscore that the act of engagement does not necessarily confer legitimacy on decision-making in any straightforward way. A particularly important challenge, which we take up directly in our recommendation below, relates to making policy decisions responsive to the outputs of public engagement efforts. There is some risk that weak public engagements do not facilitate true deliberation, and instead serve to legitimate existing policies (Rayner, 2003). Furthermore, these mechanisms of engagement are most likely to be impactful when technologies are further ‘upstream,’ or before they are locked-in (Collingridge, 1980). This means that, while especially effective in cases of emerging technologies, such as SRM, public engagements should ideally be initiated in the early stages of research, including in the setting of goals and priorities for research itself.

The literature on experimental governance provides some further and highly complementary insights into elements of a legitimate governance scheme for SRM research. Experimental governance is a ‘recursive process of provisional goal-setting and revision based on learning from comparison of alternative approaches to advancing these goals in different contexts’ (Overdevest & Zeitlin, 2014, p. 25). Although the idea of experimenting with governance in this area may raise concerns for some, experimental governance need not take sweeping, irreversible governance decisions, and need not presume that all research should face additional governance from the top down (Stilgoe, 2016). Rather, experimental governance can proceed in an incremental way, with built-in processes for feedback and adjustment. As such, experimental governance has gained much attention in the global environmental governance literature, particularly surrounding issues related to complex environmental problem-solving that require innovation beyond tried and tested models, such as climate change (e.g. Castán Broto & Bulkeley, 2013a, 2013b; Hoffmann, 2011).⁴ This is in part because, as with RRI described above, its reflexive and adaptive approach allows for suppleness, including in response to unexpected policy options as they arise (Heilmann, Shih, & Hofem, 2017). Experimental governance further provides opportunities for SRM governance to be more publicly-accountable, in that central to any experimental governance architectures are broad participatory processes for goal-setting (Overdevest & Zeitlin, 2014).

Central, therefore, to experimental governance are processes for learning and improvement over time on the basis of multiple experiments and broad participation (McFadgen & Huitema, 2018; Overdevest & Zeitlin, 2014). Importantly, this learning process could include pivoting to the possibility that SRM research should halt. As Castán Broto and Bulkeley have argued in their studies of urban climate politics, governance experiments allow for exploration of ‘unchartered policy territories to either learn or open up new forms of intervention … [Such experimental governance] interventions… try out new ideas and methods in the context of future uncertainties. They serve to understand how interventions work in practice, in new contexts where they are thought of as innovative’ (Castán Broto & Bulkeley, 2013a, p. 1953, 2013b, p. 93). They go on to argue that such experiments can establish new forms of political space that blur public and private authority and, of particular relevance to SRM governance, have the potential to challenge dominant understandings of political response to climate change. This flexible and adaptive approach is particularly well suited to public engagement in SRM research because it allows for policy learning within the nascent and highly uncertain empirical terrain of SRM research and its impacts.

An experimental governance approach could further lead to the development of participatory models of governance design that could be adapted and diffused across jurisdictions. By building in processes for learning and information exchange, experimental governance can aid in diffusion by developing networks across jurisdictions (Hildén, Jordan, & Huitema, 2017). For example, in an experimental governance framework, local peer jurisdictions share information and take corrective measures to better meet centrally defined goals, which themselves are provisional and updated over time based on review of experience on-the-ground (Overdevest & Zeitlin, 2014; Sabel & Zeitlin, 2012). Such review is a critical element of the experimental governance process (Jordan, Huitema, Schoenefeld, Van Asselt, & Forster, 2018). Through this process, experimental governance initiatives can diffuse both horizontally across peer jurisdictions and vertically to other levels of governance (Hildén et al., 2017).

Finally, there is a question of agency here. Who should be responsible for initiating and stewarding public engagement, learning, and review processes surrounding experimental SRM research governance? What role should that actor(s) play? How should that role change over time? The literature on orchestration lends some insights here. Orchestration is a mode of governance in which one actor (the orchestrator) enlists one or more intermediary actors to govern a third actor or set of actors (the targets) in line with the orchestrator’s goals (Abbott et al., 2015). The concept has gained much attention in the global governance literature, particularly surrounding transnational climate governance (Abbott & Hale, 2014; Chan, Brandi, & Bauer, 2016; Chan & Pauw, 2014; Gordon & Johnson, 2017; Hale & Rogers, 2014), and is increasingly recognized as a key mode of global governance (Bäckstrand & Kuyper, 2017). Although initially focused on the orchestrating role played by IGOs and states (Abbott & Snidal, 2010; Hale & Rogers, 2014), more recent works have broadened this to include non-state actors, networks, corporations, foundations, and non-governmental organizations (Chan & Pauw, 2014; Gordon & Johnson, 2017).

We contend here that orchestration can further offer analytical leverage in considering how sub-state actors, such as California or New York in the U.S. context, can engage in SRM research governance by creating intermediaries (e.g. the advisory commission detailed below) to launch and oversee experimental research initiatives. This approach is particularly useful in the case of solar climate engineering because the scientific uncertainties and complex ethical issues surrounding it has led to much political controversy. This controversy makes it incredibly risky for any political actors to take ownership of the issue for fear of electoral or other forms of political backlash. In charging other actors with constructing voluntary near-term governance structures, an orchestrating role can provide political cover for elected officials to catalyze governance at an early stage without taking a big political risk. Centrally therefore, we propose below that one or more U.S. states should create new interdisciplinary advisory commissions on SRM research to serve as a chief intermediary for this work.

U.S. states are well suited to this role because several of them possess the key characteristics necessary to orchestrate SRM research governance. Namely, in order to orchestrate, an actor must possess convening power, moral legitimacy,⁵ financial support, technical expertise, the ability to serve as a focal point (Abbott et al., 2015; Hale & Rogers, 2014), and the availability of or capacity to create appropriate intermediaries and sufficient resources to enlist, support, and steer intermediaries (Abbott & Bernstein, 2015). In addition to having the political, financial, and logistical means to create a new advisory commission, states such as California and New York can also provide the material and/or ideational support to facilitate the intermediary’s work towards specific governance goals. Material support can strengthen the intermediary’s operational capacity, whereas ideational support can include guidance, formal approval, or political endorsement to enhance the intermediary’s effectiveness and legitimacy vis-à-vis private, non-state or state targets (Abbott et al., 2015).

Sub-national political bodies in the U.S. have played this kind of orchestration role in relation to other arenas of scientific and technological development, particularly in instances when research or the implications of research have broader social implications. Illustrative examples of issues and mechanisms include investigation by the interstate Delaware River Basin Commission (among other bodies) into fracking (Davis & Hoffer, 2012); a range of state-level initiatives to examine and regulate research into and the use of products resulting from genetic engineering and biotechnology (Vito, 2018); and the establishment of broad-based commissions and councils empowered to provide independent assessment of issues stemming from scientific and technological developments affecting a given state (see, for instance, the California Council on Science and Technology).

While a substantial literature has explored how and why orchestration takes place, the normative dimensions (e.g. fair and just governance), power relationships, and the effectiveness of orchestration in attaining the intended goals remain underexplored. One exception is Bäckstrand and Kuyper (2017) recent proposal for a ‘democratic values approach’ to evaluating orchestration. Dovetailing with the literatures on political legitimacy in global governance, RRI, and experimental governance discussed above, this approach argues for participation, deliberation, accountability, and transparency as key metrics for evaluating, and in our case, designing and maintaining an experimental governance architecture for SRM research.

In summary, the global governance literature offers several insights into how sub-state actors might begin to build a legitimate governance framework for SRM research. Namely, this literature suggests that:

1. Near-term trade-offs between legitimacy and speed of technological development may be necessary.

2. Evolving and responsive governance processes that allow for institutional learning over time based on public input are critical.

3. The public – and/or its representatives – should have early opportunities to critique and contribute to development and broad assessment of research proposals in order to shape innovation in socially acceptable directions, including in the setting of goals and priorities for research itself.

4. Controlled experimental governance approaches to participatory engagement can be helpful in developing context-specific governance mechanisms for engagement surrounding SRM research over time.

5. Information sharing across jurisdictions about experimental governance approaches will be crucial to learning and adapting governance institutions over time. Such information sharing should allow for the taking of corrective measures to better meet centrally defined governance goals, which themselves are provisional and updated over time based on review of experience on-the-ground.

6. In the absence of clear political positions on SRM research from governments, sub-state actors can orchestrate participatory processes as a way to catalyze governance in this area.

Recommendation: One or More U.S. States Should Establish an Interdisciplinary Advisory Commission on SRM Research Governance

In this section we build from prior work by others to sketch a concrete proposal for an ‘Advisory Commission on SRM Research.’ We build this proposal on theoretically-derived insights from the governance literatures examined in the prior section. We apply those theoretically-derived insights to the particular empirical and political contexts surrounding proposed SRM experiments and emerging leadership in climate governance more broadly. Development of one or more advisory commissions would enable U.S. states such as California or New York to take an orchestrating role in SRM research governance, with an eye toward facilitating early public engagement on questions of research design and governance. Specifically, we suggest here that states should create a new state-level advisory commission to serve as an intermediary (i.e. overseen by but not necessarily representing the state) in catalyzing and shaping the development of experimental SRM research governance in the state.

Our intent here is to sketch a set of design characteristics and features of a legitimate, functioning, and workable sub-state commission. We do so by building on previous proposals for a similar federal-level body (e.g. BPC, 2011; Winickoff & Brown, 2013) and the six core insights from the literature delineated above.

As a starting point, we note that various commentators and experts have put forward substantive standards for the governance of research, including standards for the kinds of research that ought to move forward (or not) (e.g. Parson & Keith, 2013). However, as Winickoff and Brown argue, the articulation of these standards only addresses one piece of the problem, because it does not address procedural questions around the need for standards themselves to ‘emerge from trusted institutions and a transparent process’ (2013, p. 80). Addressing procedural questions is the first critical role that an advisory commission might play.

In terms of outlining design considerations, there are a variety of prior efforts on which to draw. Advisory commissions have been established in a variety of issue areas, such as bioethics and neuroethics, to provide information, analysis, and recommendations to policymakers. In their proposal for a federal-level advisory commission for geoengineering, Winickoff and Brown (2013) look to such prior efforts to outline several necessary characteristics to ensure the effectiveness and legitimacy of any advisory commission developed to look specifically at SRM. Specifically, Winickoff and Brown (2013) argue for a commission that is independent, transparent, deliberative, publicly engaged, and broadly framed. They further suggest that a Commission’s membership be: interdisciplinary, including natural/physical and social scientists; politically diverse, including ‘experience-based experts’ from, for example civil society organizations and the private sector; and, critically in our view, also including representatives from potentially affected communities (p. 83). We endorse those broad and general characteristics here as well, noting that when it comes to ‘affected communities,’ lines will have to be drawn between those potentially impacted in the near-term by small-scale research (e.g. those who are close to experimental sites) and those who would be impacted should SRM technologies ever be deployed at scale (i.e. implying regional or global-scale impacts). The question of who draws those lines and how is also a matter of concern from the perspective of legitimacy; advisory commissions should consider these issues and provide advice.

Advisory commissions also typically establish explicit avenues for public input, either through the hosting of public meetings or the seeking of written commentary on aspects of the commission’s work. Here, the global governance literatures reviewed above suggest a need for special care. This is because exercises designed to elicit public input too often become rote exchanges. Given the broad social and political ramifications of SRM and its status as an imaginary technology at this stage, experts – including ethical experts – do not have a monopoly on the kinds of expertise or concerns relevant to its governance. In order to maintain political legitimacy, then, a Commission on SRM Research must move beyond thinking in terms of public input and the provision of consensus policy recommendations, and instead work to generate meaningful public engagement to clarify value disputes and political disagreements, enabling compromise and honest brokering of policy choices. Indeed, meaningful engagement is a necessary (but not sufficient) design element to ensure any future SRM research governance regime enjoys some degree of political legitimacy. Moreover, advisory commissions should be designed to enable social learning around efforts to govern SRM research (flexibility, adaptability are features that might help enable this).

Research in other emerging technologies has shown that wider publics often raise issues related to innovation missed by experts.⁶ Public engagements should also focus on the kinds of governance architectures that might be appropriate for various kinds of research. This addresses the gap in existing approaches to governing SRM experimentation, which seek to develop substantive standards and/or thresholds for research based on criteria developed in isolation of public involvement (e.g. technically-defined thresholds or ‘allowed zones’ for experimentation). Engagement work might also overlap with the anticipatory functions of the Commission, including via the involvement of lay publics in foresight exercises and scenario-building. Centrally, meaningful public engagement can take many forms, and should be integrated across phases of the research process.

The Commission should ideally operate at the level of research programs, rather than at the level of specific proposed experiments. This means that the public should be involved in early stage discussions about ethics and other social issues that may arise from SRM research. At the same time, public input could prove valuable for individual researchers and experimental efforts, by helping to sharpen research questions and revealing possible risks and concerns that might have been overlooked by principal investigators. Similarly, there are a variety of established methods that can be employed to this end, for example, focus groups, deliberative mapping, citizen panels, and scenario planning (see Stilgoe et al., 2013).

Critically, there must be a mechanism for outputs of such public engagement processes to feed back into governance design, ideally across scales and jurisdictions. Here, again, insights from both experimental governance models and RRI underscore the importance of reflexivity and learning. The Commission should play a central role in developing such processes, through for example, ‘peer review’ of public engagement work across jurisdictions (Overdevest & Zeitlin, 2014, p. 25), and asking localities to report back to the Commission on experience with engagement processes identified above, and by creating pathways for such experiences to feed directly back to policy makers through, for example, formal recommendations from the Commission. Public engagement outputs should also feed into less formal governance mechanisms such as refinement of existing codes of conduct and/or through ties to public funding decisions, which conditionally link responsiveness of project design to outputs of public engagement processes. In addition, the outputs of the engagement work should meaningfully affect the advice the Commission gives to researchers and research funders.

Finally, experimental governance models can provide for diffusion across jurisdictions. As suggested in the RRI literature, governance diffusion can indeed be dangerous if done in a ‘one size fits all’ manner. However, this tension can be managed if procedures for adapting governance models for specific localities are a baked in and expected part of the experimental governance process. States are particularly well suited to orchestrate diffusion due to existing institutional mechanisms, such as the Global Climate Leadership Memorandum of Understanding (‘Under2MOU’) and the C40, which provide policy frameworks to facilitate cooperation between sub-national jurisdictions across the globe, including as related to public outreach and transparency. Specifically, the Under2MOU can serve as an intermediary to organize communication and learning between the Commission and its peer institutions in other jurisdictions.

The Commission should adopt the procedures outlined above across its work program on SRM research governance. In addition, building on the Bipartisan Policy Center’s Report on climate engineering governance, the Commission on SRM Research Governance should undertake, at a minimum, the following three core functions:

1. Identify key research questions and capacities

An important early task for the Commission would be an authoritative review of existing SRM research, from modeling to possible open-air experimentation, with an eye to existing or planned research efforts in the relevant jurisdiction. The goal should be the establishment of a detailed understanding of the existing research landscape and, to the degree possible, the establishment of a set of near- and longer-term research priorities.

SRM technologies, should they ever move from the laboratory into the world, will likely function as only one part of a portfolio of climate change response options. SRM research, this is to say, is best seen as science in the service of important social and environmental aims, rather than as science determined and defined ultimately by the proclivities of individual scientists or investors (Long, 2017). This view suggests the need to ensure that research is multidisciplinary, and/or that it is integrated with stakeholders/users, so that research is responsive to evolving social needs and norms. The Commission will need to build from existing research reviews, most notably those carried out by the National Academies of Sciences (see National Research Council, 2015a, 2015b), direct inputs from relevant scientific communities, and broad and deep public engagement to establish research needs.

In assessing research questions and determining priorities, the Commission will need to pay attention to the trade-offs between legitimacy and efficiency of technological development noted above. That is, meaningful public participation will necessarily take time. However, if done well this added time will not only add legitimacy to the process of policy development, but will also yield stronger policy outcomes in future. As such, the trade-off is a short term one, with public participation yielding efficiency benefits in the longer-term by way of enhanced problem solving. A forward-reaching research agenda that appears intent on bringing SRM technologies quickly into the world would likely generate public backlash and could lock-in undesirable research pathways. On the other hand, too restrictive a research agenda could make the effort appear legitimate in the eyes of certain stakeholders, but could unduly shackle the activities of researchers.

This is all to say that it will be an additional and important early step to have a full accounting of the state’s existing SRM-relevant research capacity and efforts and to establish mechanisms for information sharing across jurisdictions. The Commission will need to identify individuals and research programs that are already undertaking research that is relevant to an understanding of climate engineering and its impacts, along with identification of individuals and research programs that are in a strong position to contribute to the development of such knowledge. This should include identification of key individuals, programs, and perspectives in other jurisdictions, to increase flows of information and learning.

2. Advise state governments on social and ethical issues that may arise from research

Another function of the Commission will be to offer advice on how best to navigate the social and ethical issues that arise from SRM research and potential deployment scenarios, potentially building on substantive ethical frameworks put forward in the literature (Rayner et al., 2013; Bipartisan Policy Center, 2011; Asilomar 2, 2010). This suggests two things. First, the makeup of the Commission must take explicit account of the need for a range of different kinds of expertise and viewpoints. In addition to those with a strong grasp of relevant aspects of physical science, the Commission should have membership that includes ethicists and religious scholars, representatives from non-governmental organizations, along with members versed in social scientific research and policy engagement, and perhaps lay membership. Second, here, too, there is explicit need not just for public input but for public engagement, to spur deep social learning and deliberation on SRM’s ethical dimensions. The learning and diffusion mechanisms described above are easily transferrable to this function as well.

Importantly, in the early stages of SRM governance, this function of the Commission should be seen as advisory and evaluative rather than regulatory. The SRM conversation and research into possible technologies is at too early a stage to make big decisions on formal control over research efforts. The evolving and responsive governance frameworks argued for above will be fostered by information gathering and sharing and the daylighting of activities, coupled with careful consideration of the social and ethical dimensions of present-day and possible future research decisions.

3. Recommend criteria for research oversight and the apportionment of state funding for CE research

Research on SRM is not simply normal scientific investigation, since the technologies to which such research could lead have enormous, world-shaping implications, and are characterized by deep uncertainty and high decision stakes. By this, we mean two things. First, decisions about SRM research taken or supported by sub-state actors could ripple into the global climate change response conversation. If a major US state, for instance, were to signal support for rapid development of SRM deployment capacities, this could force a reevaluation of international commitments to climate change mitigation activities. The political signals sent by research prioritization and funding decisions is, then, an important consideration. Second, if particular SRM technologies were to be shown through a research program to be viable or, by contrast, to be too risky to support, this too would shape understandings of how best to respond to climate change.

Given all of this, any decisions about funding are immensely important. In addition to establishing research priorities, it might be supposed that a Commission could make recommendations about directing public research dollars, or a coordinating and daylighting role for private research dollars by contributing to ongoing efforts to establish clearinghouse and other transparency mechanisms (see e.g. Turkaly, Nicholson, Livingston, & Thompson, 2017). A key additional feature of this work will be the determination and inculcation of norms and rules that would be relevant to publicly funded research. This suggests that any Commission should, working with other relevant bodies, be tasked with determining appropriate forms of research oversight, either through voluntary means (e.g. requesting that researchers adhere to a voluntary established code of conduct) or formal means (e.g. requiring particular forms of reporting and impact assessment of research related to SRM, or requiring that researchers adhere to a code of conduct). This final point relates directly to the insights from the orchestration literature unpacked above. Sub-state actors are in a prime position to develop the guidelines and guardrails for SRM research, and to disseminate rules and norms across jurisdictions.

Conclusions

Governance questions surrounding climate engineering are becoming increasingly salient. This is driven by the sense that traditional forms of mitigation will be insufficient to avoid breaching the 1.5 C threshold of atmospheric warming set out in the Paris Agreement, and by a number of proposed small-scale field experiments in the U.S. In parallel, sub- and non-state actors have demonstrated leadership in climate change response through a bevy of creative and largely experimental climate governance response measures. In light of these developments, we urge states to play a role in the experimental governance of SRM research.

Nevertheless, engagement from U.S. states on research governance is unlikely to be straightforward. Why should states take the lead on this issue when they can demonstrate climate leadership in other less controversial ways? One central reason they may wish to do so is because field experiments are already proposed and may soon be underway in various states. These states can either reactively address SRM experiments and public responses after the fact, or they can do so proactively in an effort to steer the responsible development of SRM research and its governance. There is a non-negligible risk, of course, that efforts to develop research governance at the state level will be unable to contain political backlash or opposition to technology development and associated governance institutions. This could be viewed as a kind of informal technology assessment (Rayner, 2004).

Another key consideration if U.S. states consider developing advisory commissions for SRM research governance is the potential for competition between parallel commissions in various states. To the extent that advisory commissions across states make different recommendations about research governance needs, there is some risk that researchers simply propose experiments in the most permissive states. This issue has been raised in the context of experimentation with driverless vehicles in the U.S., for example (Stilgoe, 2017). Our recommendations here do not preclude governments and political decision-making bodies at a range of scales from governing SRM research, and we do not imply that advisory commissions at sub-national levels ought to substitute for national or even international approaches to governing SRM research. Rather, our more modest ambition is to advise state decision-makers in jurisdictions where experiments have been proposed, in order to enable collective learning and harmonization of research governance approaches over time – precisely the things that might help avoid regulatory competition, at least within a given political culture.

We have suggested above the development of an Advisory Commission on SRM Research as a valuable first step. The design of such a Commission deserves careful consideration. Drawing from several important strands of governance literature, we proposed that sub-state actors, such as California or New York, adopt an experimental governance model to inform and guide the Commission’s work. By this, we mean that the Commission’s governance recommendations and activities should, by design, emerge in an iterative and reflexive fashion. A key piece of this will be the development of a Commission that has not just the seeking of public input but rather the fashioning of ongoing and deep public engagement as one of its chief guiding principles. Such engagement should involve publics early in the research process, in everything from the construction and framing of research questions, to the dissemination of results and reflection on the meaning of those results for the future of SRM research in the state.

Acknowledgments

We are grateful for comments on earlier versions of this manuscript from Steve Gardiner, Catriona McKinnon, Augustine Fragniére, Ted Parson, and an anonymous reviewer. We are also grateful to Zach Dove for this research assistance. This paper was in part supported by a fellowship from the Andrew Carnegie Corporation of New York. All content remains the responsibility of the authors.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Given their centrality to the Paris Agreement’s net zero emissions mandate, the literature surrounding carbon dioxide removal is also quickly growing (e.g. Burns & Nicholson, 2017). Given the divergent political implications and thus governance frameworks demanded by CDR technologies we do not address them directly here. Nonetheless, the core governance proposal detailed below could be adapted to include CDR technologies as well.

2. Nicholson et al. (2018) derive, from a review of governance literatures, four objectives to guide the development of SRM governance: Objective #1: Guard against potential risks and harms; Objective #2: Enable appropriate research and development of scientific knowledge; Objective #3: Legitimize any future research or policy-making through active and informed public and expert community engagement and the development of democratic systems of control; and Objective #4: Ensure that SRM is considered only as a part of a broader portfolio of responses to climate change.

3. For an illustrative table of potential techniques and approaches that correspond to these dimensions, see (Stilgoe et al., 2013).

4. Experimental governance is closely related to what Bryan Norton and others have termed adaptive management, which also focuses on experimentalism, multi-scalar analysis, and place sensitivity (Norton, 2005, p. 92). However, whereas adaptive management is by Norton’s definition used to make decisions impacting the environment specifically, experimental governance is largely used to discuss governance decisions that impact social systems specifically.

5. Moral legitimacy refers to the recognized authority by its polity to govern in a particular space.

6. For example, a 2007 study on public views of nanotechnology found that the public was significantly concerned about the possibility that the technologies might eliminate jobs, while this was of little concern to nanoscientists. (Scheufele et al., 2007).