The Decision Space: Context and Key Considerations for Solar Geoengineering Research and Research Governance
The fragmented state of the SG research and research governance landscape is a barrier to the efficient and effective advancement of knowledge and the understanding of societal views on this issue. This chapter describes the “SG decision space” that spans a complex, interconnected terrain encompassing scientific research, societal values and perspectives, and governance mechanisms. Section 3.1 considers which decision makers might be primary consumers of information from a research program and discusses their potential information needs. Section 3.2 examines the societal contexts in which SG-related decisions may be made, and Section 3.3 considers the many intertwined scientific, societal, and governance issues that make this such a challenging area of exploration. Section 3.4 explores high-level principles that might guide research and research governance, as context for constructing a research program that is attentive to the many difficult social issues inherent to this topic.
Scientific research is typically conducted to support curiosity-driven knowledge expansion, predetermined objectives, or both. SG research could be viewed as seeking a middle ground, where scientific knowledge is acquired to assess the feasibility and uncertainties of deployable, predictable, and controllable SG technologies, and to provide information that permits society to assess the value of SG as part of climate change response.
At the intersection of science and society are people who must make decisions about research activities, governance mechanisms, and potential deployment actions. Addressing the needs of these diverse decision makers is a primary goal of a research program. To assist with identifying these needs, the committee engaged several representative decision makers over the course of this study (discussed in Box 3.1).
3.1a Who Are the Decision Makers?
A decision maker is anyone with the authority to determine the course of future actions. SG decision makers may exist at all levels of society and include local community leaders, elected national representatives and civil servants, academics, corporate and nongovernmental organization (NGO) executives, and representatives
within multinational bodies. Decision makers are geographically diverse, spanning the globe; economically diverse, spanning developed and developing nations; and politically diverse, spanning a spectrum of political viewpoints. Decisions relating to SG may arise for decision makers at any level, whether it be reviewing a local permit for experimentation or pursuing a globally coordinated initiative to initiate or restrict deployment of a technology.
3.1b Decision-Maker Needs
What information will enable informed decisions about SG, and what research will best provide that information? Answering these questions is an exercise in prediction and anticipation, as there are currently few decision makers who are knowledgeable about and focused specifically on this issue.1 If research activities expand beyond current modeling efforts, or if SG were to be considered for deployment, decision makers will be faced with a wide array of complex questions. This includes general questions such as the following:
- What are the different geoengineering action options available?
- How are these different options likely to be received or accepted by different stakeholders and publics?
- How might SG affect the broader portfolio of climate responses? How can SG research be governed so as to minimize risks of mitigation deterrence?
- What are the known costs and potential liabilities (e.g., fiscal, political, and social) of different SG actions?
- How might SG research or deployment affect international relations?
- What forms of governance are needed to manage research and development, and decisions about deployment?
and includes many questions specifically about the outcomes/impacts of SG deployment:
- What global- and regional-scale impacts might be expected over time?
- What will be the distribution of impacts (i.e., benefits and risks) across different parts of the world, and what are the equity/justice implications of such differences?
- What are the expected risks and benefits in terms of outcomes from different SG approaches, relative to expected outcomes without deployment?
1 One possible exception is the sponsors of Australia’s experimental marine cloud brightening program for the Great Barrier Reef (which is part of a larger program on climate adaptation).
- Can SG actions be effectively managed to deliver the intended outcomes? What would be the (positive or negative) indicators of such outcomes, and on what timescale might these outcomes be known?
- What are the possible unintended consequences associated with different SG actions, and what could be done to mitigate against such consequences?
- If adverse impacts are experienced and are perceived to outweigh the benefits of deployment, can SG actions be reversed?
The purpose of designing an SG research and research governance program is to help address these questions as effectively and efficiently as possible, as a foundation for informing the choices that decision makers may face.
3.1c Scenario Context
Decisions about SG deployment would be made by a wide variety of actors with differing motivations and under differing climate, geopolitical, and societal conditions. It is important to consider a wide range of possible scenarios in order to identify the types of information that could be necessary for robust decision making. Below are a few scenarios that have been used to inform research to date2 (noting that current knowledge is insufficient to ascertain whether SG is a viable option for addressing the challenges encapsulated within scenarios).
The “climate context” for SG decision making will depend in particular on the severity of climate impacts occurring and on the degree to which other climate change responses (e.g., mitigation, carbon dioxide removal [CDR], and adaptation) are being pursued. At least three types of scenarios have been commonly discussed in this regard:
- A common framing scenario is referred to as “peak shaving,” in which mitigation efforts are having a positive effect, but are not sufficient to prevent an “overshoot” of goals for limiting global mean temperature increase. Large-scale CDR could potentially help reduce temperatures but only over time periods of decades to centuries. It has been suggested that, in the interim, SG might be considered for deployment in a temporary way to reduce peak temperatures (MacMartin et al., 2018a; Tilmes et al., 2016; Wigley, 2006). Here “temporary” implies a sustained deployment over multiple decades to a half-century or more, to limit peak temperatures during a period of globally
- sustained decarbonization. In this context, an important factor influencing SG research or deployment considerations will be the trajectory of continuing mitigation measures and what is known about the plausibility of large-scale CDR relative to the required time frame of a commitment to SG once begun.
- An alternative (also time-bound) scenario is one in which there are emission reduction efforts but no reliance on CDR, large-scale adaptation is under way, and SG deployment is considered to slow the rate of temperature rise, with the goal of securing more time for adaptation (Irvine et al., 2019; Keith and MacMartin, 2015; MacMartin et al., 2014a). Current projections of peak temperatures, expected impacts, and the timescale for adaptation would influence considerations of how or whether SG might be utilized.
- A third scenario is one in which efforts for meaningful mitigation, CDR, and adaptation have been inadequate, and SG is considered as an emergency response to blunt destruction caused by rapidly accelerating temperature rise. This scenario is characterized by the need for indefinite and ever-increasing levels of SG, with significant unmitigated direct harms from rising CO2 concentrations (e.g., ocean acidification), and growing risk of unintended side effects from intensifying SG deployment.
Differing political contexts and scenarios will also affect decision making. For instance, a scenario involving deliberative action through a globally representative body or agreement might best address social and governance concerns and provide the most resilient foundation for research activities or sustained deployments. But there may also be scenarios in which regional coalitions or collections of individual state actors act autonomously but with shared views or even scenarios involving a lone actor, perhaps not even representing a sovereign nation, attempting unilateral deployment. In each of these cases, the actor(s)’ record of pursuing climate change mitigation, and the associated inferences regarding underlying motivations and commitments to decarbonization, might be an important part of the decision-making context.
The specific details of a particular SG deployment could also significantly alter the context for decision making. For example, in a lone actor scenario, response to a localized marine cloud brightening (MCB) operation deployed by a small island nation is likely to differ from the response to a large nation experimenting or deploying stratospheric aerosol injection (SAI) at a scale with transboundary or global effects. If that island nation were part of a coalition of island nations mounting a coordinated deployment campaign that started to have measurable global effect, these differences might diminish. Similarly, an SAI deployment focused on Arctic sea ice preservation, while it would still have a global impact, would have different implications and potentially elicit different responses than efforts designed to change global temperature.
The potential of SG activities to cross geo-political boundaries, the long timescales over which SG might be deployed, and the diversity of perceptions about potential SG benefits and risks all raise societal issues that could affect the course of SG research as much as scientific and technical considerations. The following discussion identifies issues that shaped the committee’s exploration of SG research and considers how these issues factor into SG research and research governance program design.
3.2a Diverse views on Solar Geoengineering Research
SG research is controversial within and beyond the climate science community. Some view SG as a potentially critical tool for climate change response and thus argue for an acceleration of, and greater funding for, research. Others argue that without substantial societal demand for SG research, it is inappropriate to redirect funds away from other areas of climate science. There is no consensus, domestically or internationally, on whether and how research should be pursued. Section 2.3b reviews the “conditional” support for SG that is found in social science research studies. Environmental NGO positions are divided on the question of outdoor experiments—some silent, many strongly opposed,3 and others in favor of caution with controls for outdoor experiments and greater engagement of global publics in decision making.4
These widely varying views illustrate why an SG research program needs to be multifaceted, encompassing not only natural science research to better understand the direct and indirect effects of different SG approaches, but also social science research to better understand societal views on risk tolerance and equity, as well as the most effective and appropriate methods to engage stakeholders and to build research and research governance capacity.
3.2b Issues Related to Risk and Uncertainty
Characterizing and Reducing Uncertainty
Risk assessment and uncertainty characterization will be critical for future SG decision making. Risks and uncertainties may be reduced, but not eliminated, by research.
Some degree of uncertainty will be a persistent feature of SG technologies, especially at regional scales, as a direct result of uncertainties in the underlying climate models. In fact, some SG research activities may result in increased characterization of uncertainties. For example, model projections of the climate response to SG currently use a relatively modest range of future climate, socioeconomic, and deployment scenarios. Research designed to improve understanding of the possible future conditions under which SG might be deployed may expand the range of scenarios assessed, and this in turn may expand our understanding of uncertainties about SG efficacy and risks. In other words, further research might (at least initially) help illuminate how little we know.
It is particularly important to understand the extent of uncertainty reduction that could be achieved through research activities versus through actual SG deployment (or at least testing at scales that would be tantamount to deployment). For example, decision makers may reasonably wish to know the likelihood of SG-induced changes in regional precipitation patterns and how such changes could affect agriculture, ecosystems, and public health. If risks are significant, they might further ask whether a consequential precipitation change that is observed following deployment could be attributable among natural climate variability, direct consequences of climate change, or the SG activity.
Many difficulties in reducing uncertainty are linked to challenges associated with performing field experiments on the spatial and temporal scales required to observe climate impacts. Given the backdrop of natural climate variability, a perturbation experiment large enough to produce a detectable change in secondary effects such as regional precipitation patterns would necessarily have to be large enough to affect radiative forcing at hemispheric or global scales for decades. That is, it would need to be carried out at a deployment scale. An SG research program may be able to reduce some current uncertainties, for instance, related to regional precipitation and precipitation-dependent systems; but with current technology, it is unlikely that models will be able to provide high confidence about secondary effects (see Chapter 2). Further characterizing and quantifying how much uncertainty can be reduced through research will be critical for decision makers, and this itself can be viewed as an important research question.
Assessment of Comparative Risk
Some argue that decisions about expanding research on (and considering potential deployment of ) SG strategies should be based on efforts to weigh the risks of climate change against the risks of the particular form of SG in question. Risk-risk assessment
(or risk trade-off analysis) provides a framework wherein the risks of one policy option are comparatively assessed in relation to the risks of others to identify options that maximize benefit. The relevant comparison would characterize the risks of climate change without SG versus the risks of climate change with SG—in both cases, looking across a range of greenhouse gas (GHG) concentration pathway scenarios and including an array of other climate response actions. This is a common methodology for assessing choices between complex and indeterminate options.
Comparative risk assessment, often quantified as the probability of an outcome times the magnitude of that outcome, requires probability and magnitude estimates for key risks associated with different policy choices. The robustness of any comparative risk assessment is dependent upon the accurate identification of the climate context (with its associated risks), the identification of the relevant SG options (with their associated risks), and interactions between these and other factors discussed in the following section. It is likely that, even with substantial further research, significant uncertainties will remain absent full deployment and decades-long observation. This is equally true for comparison to the impacts of climate change without SG, where uncertainties across the broad range of impacts are unlikely to be fully understood or quantifiable. In similar cases in which data-driven risk assessment is limited, other strategies such as assessment by expert judgment can be employed—though such methods increase subjectivity and risk of disagreement, especially if such risk assessments are highly contextualized.
Another particular challenge for comparative risk assessment is the diverse perception of risk across the global communities that may be affected by SG research or deployment. Climate change effects—with or without SG—will not be distributed equally across the globe, and risk perceptions and risk-related values vary significantly across nations and communities. Therefore, any risk-risk analysis should be contextualized to the specific decision maker, and the challenge of reaching consensus on risk assessments will scale to the breadth of impact for the activity under consideration.
For these reasons, operationalizing comparative risk analysis will be challenging and require specific focus in any research and research governance program. A primary implication is the need for research to better identify and understand the potential climate, health, and ecological risks (as well as the broader social, political, and economic risks) that could be associated with specific forms of SG.
The inherent challenges in quantifying risks, together with diversity in risk perception and priorities among different global stakeholders, support the notion that international participation and stakeholder engagement will be important for determining how to address risks and uncertainties in relation to the overall research enterprise. Any consideration of moving from research to deployment would increase the needed
scale and breadth of such engagement. These challenges also support the assertion that an SG research agenda needs to include research on issues of risk and uncertainty, including the scientific, social, political, and ethical dimensions of this topic. New approaches may be necessary to understand, evaluate, manage, and communicate risk and uncertainty in this domain.
In addition to characterizing risks and uncertainties, and attempting to compare a range of GHG concentration pathway scenarios and climate response portfolios, risk assessment can be embedded in a broader approach to risk, often described as risk governance. The risk governance approach evolved out of a framework that includes risk assessment, risk management, and risk communication. However, it involves a more integrated and iterative process of understanding, evaluating, and managing risks and uncertainties by engaging both experts and various publics throughout these ongoing processes (Klinke and Renn, 2019). As such, the risk governance approach is process oriented and addresses both the objective and subjective dimensions of risk and uncertainty.
If pursued for SG research, a risk governance approach might include engaging experts, stakeholders, and diverse publics in exploring risk and uncertainty in relation to various climate scenarios, including scenarios that involve one or more forms of SG. Risks and uncertainties may vary significantly depending on context, and both technical expertise and public and stakeholder engagement could play an important role in characterizing and evaluating risks under different scenarios.
Developing responses to climate change is a dynamic and ongoing process. There will be significant unknowns in any pathway taken, and decisions about SG research, as well as any future decisions about deployment, will be made in an environment of incomplete information. This is the case for all policy decisions but will be particularly acute for decisions (including but not limited to SG) that have global and intergenerational reach. Efforts to understand and characterize the climatic, ecological, and social risks and uncertainties associated with SG, as well as to understand their significance for various groups, will be an important component of any research program. A risk governance approach can incorporate comparative risk assessment into a broader frame that includes collaborative, participatory, and adaptive approaches to managing risk, uncertainty, and ignorance. As Klinke and Renn (2019) put it,“the goal of risk governance is to embrace uncertainty, complexity, and ambiguity as major characteristics of risk governing processes and deal with them upfront.”
3.2c Solar Geoengineering and Society
Societal concerns about climate change specifically, underlying economic and equity conditions more broadly, and perceptions about technological risk versus benefit, factor significantly into considerations of SG research. Such considerations encompass complex justice and equity concerns, discussed in Section 2.3a (“Ethics and Geoengineering”), as well as several other key issues highlighted below.
Failure to Meet Climate Mitigation Goals
SG research emerged largely in response to concerns over inadequate action to mitigate climate change by reducing GHG emissions. Rationales for considering SG as one element of a broader climate response include (i) broadening the array of tools with which to address climate impacts (sometimes justified by appeal to the need to develop “every tool in the toolbox” against climate impacts); (ii) reducing the impacts of a possible “climate emergency”; (iii) buying time for more ambitious mitigation and climate stabilization (the “peak-shaving” scenario); and (iv) reducing near-term impacts of climate change for vulnerable communities.
Philosopher Stephen Gardiner has argued that the failure to significantly reduce emissions represents an ongoing moral failure, which leaves us in an ethically compromised position:“It is mainly because we have failed—and continue to fail—to do what we should have done, ethically speaking…that geoengineering is being considered at all” (Gardiner, 2020). Gardiner questions whether an ethical geoengineering policy is likely in the context of this ongoing moral failure, noting that many of the same problems that plague climate policy (e.g., a tendency to focus on the short term, postpone
action to mitigate, and displace risks onto others) may emerge in the context of SG research, development, and consideration of deployment.
Others suggest that research is important precisely because progress on mitigation has been so slow (Victor et al., 2013). According to this view, even if it would not be needed in an ideal world, research on SG may be important now, given that many parts of the world are already experiencing significant negative impacts of climate change. SG might be able to reduce the severity of some of these impacts (at least temporarily) while GHGs are being stabilized.
Moral Hazard/Mitigation Deterrence
As discussed in Chapter 2, numerous studies and reports have identified the possibility that geoengineering may reduce commitments to climate mitigation, slowing the pace of emissions reductions and the transition away from fossil fuels. As mentioned previously, the idea that geoengineering might undermine mitigation efforts is commonly referred to as “moral hazard” (Keith, 2000; Lin, 2013; McLaren, 2016). The worry is that if geoengineering is viewed as a partial remedy for near-term climate impacts, it may reduce incentives to commit to and invest in mitigation and adaptation efforts.
Societal acceptability of expanded investments in SG research within the United States and internationally may be contingent, in part, on how moral hazard concerns are addressed. An expanded research program can be expected to have greater social acceptability if it is embedded within a portfolio of climate policies and research investments that include a firm policy commitment to decarbonization. In the absence of such a commitment, expanded funding for research risks exacerbating moral hazard concerns and reducing societal acceptability of both research and prospective deployment.
Slippery Slope Concerns
A commonly expressed concern about pursuing an SG research program is that such a program could create a “slippery slope,” or an acceleration toward eventual deployment (should a viable technology emerge). There are various ways in which this acceleration scenario could occur:
- First, a research program could create political momentum toward deployment. Political support will generally be needed to start a state-sponsored, nationally coordinated program. One concern about political acceleration of SG
- is that those same actors who politically supported, defended, or sanctioned a research program may become invested in seeing it used.
- Second, research programs could create sociotechnical communities of interest, motivating further development of the technology. An SG research program could take several decades to produce the desired results. In that time, any number of structural, institutional, and even psychological factors could incline the research community toward a demonstration of the technology platforms under development. In a future in which climate change is increasingly getting worse, some SG researchers or research communities could become focused on advocating for deployment.
- A third line of argument is economic—that those developing the technology will seek an opportunity to monetize it, thus motivating a push for deployment.
An overarching concern is that any one of these pathways to acceleration could result in deployment. If motivated together, they could produce mutually reinforcing dynamics. There are, however, countervailing scenarios in which researched technologies are not deployed or in which governance could effectively mitigate acceleration:
- Research could dissuade deployment if it is found that it is very difficult to produce a technology that is deployable, predictable, and controllable. If it is demonstrated that SG is not an “easy” alternative to GHG mitigation or large-scale adaptation, then it could become less politically attractive.
- Moreover, if it were to take a long time to discern whether an SG intervention is working, it may attract less political support, since any positive outcome might not yield political capital for its proponents in their political lifetime.
- The idea of a “sociotechnical community” accelerating deployment could be counterbalanced by the active cultivation of transparency among the research community. Demonstrations of transparency with regard to risks unearthed by a research program or of an aversion to advocating for deployment in the face of lingering uncertainties about SG’s secondary impacts could reduce an impetus toward deployment.
- The economic acceleration scenario is mitigated by the natural marketplace of capital allocation, as substantial economic forces are already pursuing more traditional climate mitigation measures. The current market for measures such as decarbonization of the energy sector is estimated in the tens of trillions of
- dollars over the next decade.5 For SG deployment to represent a significant financial opportunity for its proponents, it will have to have sufficient supporting evidence to allow it to outcompete other investment streams. This could also counter-balance concerns about mitigation deterrence, as investors in these new energy systems will have little motivation to abandon these revenue streams in the face of relatively small research programs, which cannot be monetized in the near term.
Although the likelihood of a slippery slope from research to deployment is difficult to assess and may change over time,“one of the problems with slippery slopes is that it is not necessarily possible to recognize them until it is too late to implement policies to address them” (Parker, 2014). By designing research efforts that build in mechanisms to detect and prevent slippery slopes, this risk may be reduced. Such mechanisms include (i) stage-gate systems or checkpoints, in which approval is required to proceed in scaling-up research, or moving from laboratory-based to field-based experiments; and (ii) incorporating public engagement into decision processes (Callies, 2019a).
Societal views about SG can be affected by the results of scientific research (i.e., by new insights about risks and benefits) and by the conduct of research itself (i.e., by the inclusivity and transparency of the enterprise); in turn, evolving societal views of climate change or risk management may steer the focus of SG research activities. This “co-evolution” of research (Jasanoff, 2004) and societal views necessarily has implications at the level of geopolitics, which refers to behavior between nation-states as shaped and mediated by geographical characteristics. Technology has long had a role in shaping geopolitics. Just as national security technologies developed during the Cold War shaped geopolitics and governance in that era (Dalby, 2015), geoengineering technologies could likewise influence geopolitics in the coming decades.
SG may have very different implications for different countries, and countries have widely varying interests in and political engagement with this technology (Heyen, 2015). Different national actors may have very different perceptions and opinions on
5 Global Commission on the Economy and Climate Change 2018; International Finance Corporation 2017.
the following questions: Is it appropriate to pursue a geoengineering research program? How does geoengineering relate to other actions seeking to address climate impacts? Under what conditions might it be appropriate to deploy geoengineering? What is the desirable climatic state resulting from geoengineering, and how will it be governed? (Humphreys, 2011). This has led to proposals such as geopolitically relevant ranking criteria to demarcate different geoengineering approaches (Boyd, 2016). Divergent interests among states also raise important questions about meaningful public participation with the geoengineering research enterprise (Jasanoff, 2019).
Research attention to issues such as how preferences about environmental futures are formulated by different people in different places, how conditions of substantial uncertainty influence decision making, and how model-based projections inform these processes might help improve our understanding of various dimensions of the regional and social disparities and related political implications of geoengineering (Heyen, 2015). Such research issues also highlight the importance of attention to the international governance of geoengineering and perhaps to a deeper exploration of “the more difficult, and more interesting question...what kind of planet is it that provides ‘the future that we want’” (Dalby, 2014).
Urgency for Research
There exists no consensus on the particular timeline for various phases of SG research, development, and possible deployment. Time frames for research and development are often unstated or focus on the short term, such as the next 5–10 years. Geoengineering research is not, however, an open-ended, curiosity-driven enterprise, and the technologies being explored are generally envisioned as relevant to addressing climate impacts this century. Many researchers imagine that decisions about whether to proceed in further developing and possibly deploying SG would occur in the next 10–30 years, and some worry that a unilateral actor might try to deploy sooner than that. There is disagreement about the relative urgency of research, with some arguing that the risk of catastrophic unmitigated climate impacts is increasing and that, as this technology represents one of a few known responses that could relieve human suffering, it should be rapidly explored and developed. Others argue that such climate emergency framing is problematic because it has the potential to displace discussion of social and ethical issues critical to SG research and development. As climate change impacts progress, societal feedback on research urgency will change accordingly. The challenge is to design a research agenda and governance framework today that best positions the state of science to be responsive to this feedback, not lagging so far behind that options are foreclosed and not pushing so far ahead that risks are exacerbated.
Science and technology do not exist in a vacuum; there are risks associated with the use of any technology, but risk can be moderated by governance. Take, for example, the concern that deployment would need to be indefinite—in other words, that once initiated it could never be stopped because it is necessary to maintain a target temperature range. While this objection may sound defeating given the risks associated with assuming indefinite maintenance of large-scale delivery and monitoring systems, anticipating this concern could lead to governance frameworks that ensure SG is considered only as part of a broader overall strategy for temperature stabilization. The following sections discuss how research, society, and governance interact in the context of SG.
3.3a Research Governance Considerations
A key near-term goal of SG research governance could be to foster a diverse, socially engaged, responsible, and accountable research program that provides clearer understanding of SG as one possible component of a broader climate response strategy, incorporating the complex and interdisciplinary perspectives inherent to the topic. With that goal in mind, some critical functions of research governance would include the following:
- Ensuring compliance with existing laws and respect for well-established ethical norms (e.g., informed consent) and values (e.g., transparency);
- Enabling responsible and legitimate SG activities to be executed efficiently;
- Promoting development and sharing of socially beneficial knowledge;
- Managing key technical concerns (e.g., unintentional secondary environmental or health impacts);
- Managing key societal concerns (e.g., moral hazard or slippery slope);
- Building trust and legitimacy across stakeholders;
- Aligning interpretation of the above and compliance across potential actors (national, international, and non-state); and
- Helping develop the capacity for future governance to effectively address decisions and responsibilities related to SG deployment.
There currently exists no clear legal framework or institutional locus for global decision making about SG research, development, or deployment. As noted in Chapter 2, some international laws and principles could be applied to certain aspects of research or deployment, but there exists no international governance regime designed specifically for research or development, and there is no one charged with making decisions about whether, how, and when SG should ever be used. Some researchers have argued that SAI would be “easy and cheap”, raising concerns about unilateral geoengineering by an individual country, collection of parties, or independent actor, potentially before the consequences of SG were extensively studied. Even barring such unilateral scenarios, there are significant unanswered questions about international governance approaches and institutions for research, development, and decisions about any large-scale testing or use. As research governance extends to potential deployment governance, the institutional challenges increase dramatically.
Although research and decisions regarding whether to pursue SG further may happen over coming decades, the time frame for deployment to achieve and maintain a desired effect is typically envisioned as significantly longer, on the magnitude of decades to centuries. Thus, global-scale interventions could require multidecadal or multicentury coordination and monitoring. This poses a particular challenge for the development and sustainment of institutions that could manage SG over many generations and with transnational/global cooperation, with no clear historical success models to emulate.
3.3b Research and Research Governance Intersection
“[Science] is intertwined with technology, innovation, and socio-economic change, facilitating the creation of new possibilities. It is this aspect—the role that science plays in creating new futures—that raises the most pressing questions for governance.”
– Global Governance of Science: Report of the Expert Group on Global Governance of Science to the Science, Economy and Society Directorate, Directorate-General for Research, European Commission (Ozolina et al., 2009)
SG research and research governance can be mutually supportive and co-evolve. SG technologies are not well developed or understood, and there is not agreement about whether and under what conditions it would be reasonable or prudent to use these technologies. This is a complex domain that involves scientific, social, political, legal, and ethical questions, and these questions interact. For example, to understand the feasibility and desirability of different approaches requires both scientific and social knowledge, and it involves both descriptive components (e.g., What would the climate effects of a certain SAI deployment be?) and normative ones (e.g.,“How are different objectives with different spatial or temporal [effects] to be balanced?” [Tuana et al., 2012]). Research can generate knowledge important to answering these questions. Research governance can advance and coordinate appropriate research; facilitate inclusive and equitable public and stakeholder engagement; address physical risks and social, ethical, and legal concerns relating to research; and help to guide research toward socially beneficial ends.
Contemporary scientific research is governed to improve both research processes and research outcomes. In the United States, for example, abuses of human subjects as part of scientific research in the 20th century triggered governance to ensure stronger protections for research participants, particularly centered on the norm of informed consent. These governance changes required more thoughtful research design and provoked reflection on research priorities, arguably contributing to better processes and better outcomes for research.
Although governance is well integrated into many scientific research processes, traditional mechanisms focus primarily on protection of human and animal subjects, assessment of direct environmental risks of research, research integrity, transparency, and funding accountability. This research governance paradigm emphasizes
the responsibilities of individual researchers and research groups to ensure that their research conforms to laws, guidelines, and effective practices, and tends to separate governance of research processes from governance of research products. This approach implicitly relies on a linear model of innovation, beginning with basic research, then moving to applied research, (technology) development, and (production and) diffusion (Godin, 2006). The linear model implicitly assumes that the main social and policy decisions regarding the use and regulation of particular technologies occur once a technology is fully formed.
A large body of scholarship in science and technology studies challenges the linear model, however, showing how “ethically significant decisions are often embedded in the scientific analysis itself” (Tuana et al., 2012) and that scientific ideas and beliefs are embedded in and evolve together with representations, social identities, discourses, and institutions (Jasanoff, 2004). Other research suggests that the changing relationship between science and society necessitates a transparent, participative, and context-sensitive scientific process in order for the generated knowledge to be socially robust (Gibbons, 1999; Nowotny et al., 2001). This is particularly true of issue-driven areas with high stakes, differing values, and difficulty in reducing uncertainty, in which a post-normal scientific approach involves broader communities in knowledge generation and validation (Ravetz and Funtowicz, 1999).
Innovation studies literature also indicates that successful innovation is based on interactions between the science and technology community and other actors (e.g., Kline and Rosenberg, 1986) with “user-producer” interactions being particularly important (Lundvall, 1992). Even framing research questions and setting a research agenda involves judgment and values (Jasanoff, 2019), and differences in frames may lead to policy controversies (Schön and Reid, 1994). In the case of SG, for example, one of the main motivations for research is concern about the human and ecological impacts of climate change and a desire to understand whether and how SG could mitigate these impacts and risks. However, other values are also at stake: some believe that large-scale SG would be a hubristic and morally problematic response to climate change (e.g., see Hamilton, 2013), while others worry that geoengineering cannot be democratically governed (Hulme, 2014; Owen, 2014; Szerszynski et al., 2013), and still others argue that geoengineering would be likely to reduce or increase global inequities (Horton and Keith, 2016; Preston, 2012; Svoboda et al., 2011).
Although some of the debate over SG focuses on whether it is “good” or “bad,” a more nuanced approach is useful when considering the governance of research. SG includes a range of technologies that could be developed and utilized in a variety of ways, or not utilized at all, and a blanket assessment can obscure the diverse forms that geoen-
gineering technologies might take (e.g., Flegal and Gupta, 2018; Stilgoe, 2015). Additionally, the research process itself will shape whether and how these technologies are developed, and normative questions can be explicitly discussed and explored as part of this process.
SG research is goal oriented, guided by the aim of exploring approaches that might temporarily alleviate some of the negative effects of global climate change. However, the precise mission for geoengineering research is not always clearly defined. Different rationales and purposes for SG have been offered, and the design of particular geoengineering strategies depends on the central purposes for which it is being developed, as well as on trade-offs between various objectives (offsetting effects of climate change on temperature versus offsetting effects on precipitation).
MacMartin and Kravitz (2019) argue that although SG research thus far has focused on modeling the impacts of particular geoengineering scenarios and identifying the uncertainties associated with those predictions, additional attention is needed to the design question,“How would one deploy to meet specified objectives?” Defining these objectives is a value-laden enterprise, and the particular objectives one identifies shapes the kinds of scenarios that are modeled and explored. International engagement as well as input from diverse publics can support the development of model scenarios that reflect various perspectives and objectives. In addition, stating explicitly the objectives underlying various SG model scenarios can clarify their strengths and limitations and reduce the likelihood that modeling will focus on an overly narrow range of scenarios.
SG research is often framed from the perspective of an unspecified deployer—for example, what kind of geoengineering do we want to consider, and what effects would it have? Given certain objectives, how might we design a geoengineering strategy that would achieve them? Should we focus on optimal strategies or robust ones (cf. Bellamy, 2015)? In each of these questions, the “we” is ambiguous (additionally, the “we” of the designers is not necessarily the same “we” as the decision makers), and there are other perspectives from which the mission of geoengineering research could be understood.
For example, some countries might be less interested in determining how to design an SG strategy than about how to detect and attribute any negative side effects of another country’s intervention; other countries may be legitimately worried about existential threats from unabated climate change and therefore may be particularly interested in engaging with and supporting deployment-oriented research. Other nations might be interested in studying ways to counteract geoengineering efforts that they oppose, and if future generations were to inherit a geoengineered world, they
might be more concerned about how to safely phase out geoengineering than how to start it.
Given the complexities of these interactions, research without continuous co-evolution of research governance striving to incorporate evolving objectives and objections seems certain to result in sub-optimal execution hindered by various forms of backlash.
3.3c Society-Research Governance Intersection
The substantial and growing body of literature on ethics, justice, and equity addresses a range of issues including whether and under what conditions geoengineering would be morally permissible; whether and how SG could be fair and equitable, considering multiple dimensions of justice (e.g., distributive, procedural, recognitional, and intergenerational); what principles might guide ethical governance; and how to evaluate SG in relation to other climate response options and address interactions with other climate responses. Such societal interests link governance of what and how SG research is conducted and its implications for influencing any subsequent consideration of SG deployment.
SAI is typically envisioned as a global-scale intervention, and the effects cannot be isolated to local or regional scales due to dispersion of aerosols throughout the stratosphere (although it could be designed to have relatively larger influence in the Arctic, for example). National-level research, governance, and stakeholder engagement may be starting points, but it should be remembered that stakeholders are global. Values and preferences surrounding research, governance, and whether and under what conditions SG could or should be used vary widely. In addition, SG may interact differently with social and ecological systems in different parts of the world, thus raising distinctive local or regional concerns. It is important not to generalize too widely from a narrow subset of stakeholders, research studies, or governance proposals.
Societally informed implications for research governance go beyond outdoor experimentation. Although some areas of SG research, such as modeling, have no direct climatic effects, the technologies being explored are typically envisioned as regional-to global-scale interventions that would potentially (for small-scale MCB) or necessarily (for global SAI) have transboundary effects. Therefore, the international and transboundary dimensions of SG deserve consideration throughout the full course of research. In addition, although current U.S. and international laws focus primarily on
physical impacts, stakeholders and members of the public have shown much more wide-ranging concerns in relation to SG research. For example, concerns surrounding the Stratospheric Particle Injection for Climate Engineering experiment, which generated significant controversy, centered not on the experiment’s minimal environmental effects but on the potential for the experiment to open the door to further development of SG and the private ownership of the technology.
These concerns also include the fact that the technologies under consideration could be used for decades or centuries; thus, future generations would inherit responsibilities for managing this deployment. Such responsibilities could be viewed as burdensome to future generations (especially if the use of SG were not accompanied by substantial climate mitigation efforts); at the same time, doing research now may benefit future generations by providing knowledge that can inform critical decisions and may provide an additional possible strategy to reduce some negative impacts of climate change.
3.3d The Nexus of Research, Society, and Research Governance
Because SG research and development are controversial and socially consequential, and the technologies themselves could have a range of regional- to global-scale impacts, building trust, legitimacy, accountability, and social responsiveness in both research and research governance are key. Even if some research and research governance initiatives begin within individual nations, the international dimensions of research and development are crucial, and inclusiveness, cooperation, coordination, and trust are critical in developing research and research governance from their earliest stages.
In this context, neither a “governance first” nor “research first” approach will work; rather, research and research governance need sustained interaction over time (see Box 3.2). At this early stage of research, there is an opportunity to co-develop research and research governance and programmatically facilitate their interaction over time. More specifically, one can co-develop governance approaches, governance research, governance capacity, and governance structures alongside research (with mutual learning between the research and research governance efforts), beginning in the early stages, in order to make thoughtful and legitimate decisions about whether or how research should proceed, what directions it should take, and whether and under what conditions SG should ever be used.
In order to integrate the breadth of the complexity of the intertwined research, social, and governance issues associated with SG, the committee explored higher-level principles to inform the design of a research agenda and associated governance mechanisms and to ensure completeness in addressing critical elements. The following section discusses principles of general governance, research governance, and international law that are relevant to SG, as well as some of the specific governance proposals that have been developed in relation to research, development, and any possible future consideration of deployment. Drawing on these examples, the concluding section identifies key guiding principles for research and research governance.
3.4a General Principles of Governance, Research Governance, and International Law
“The governance of science needs to focus on the whole spectrum of scientific activity, from theory construction and basic research to technological development and innovation.” – Ozolina et al., 2009
International governance principles have been proposed for many different domains, and there is often significant overlap among them. For instance, in 1997, the United Nations Development Program identified participation, rule of law, transparency, responsiveness, consensus orientation, equity, effectiveness and efficiency, accountability, and strategic vision as key principles of good governance.6 In 2001, the European Commission identified openness, participation, accountability, effectiveness, and coherence
as central. Woods (1999) focuses on three key principles for good governance in the international domain: public participation, accountability, and fairness—noting that public participation gives “affected parties access to decision making and power so that they have a meaningful stake”; accountability “requires clarity about for whom or on whose behalf the institution is making and implementing decisions”; and fairness applies both to the processes by which decisions are made and the outcomes of those decisions. Woods further notes that accountability depends on transparency, which provides information critical to holding institutions accountable.
Woods’ principles were not developed specifically for global governance of science, but they overlap with science governance recommendations such as those of the 2009 Global Governance of Science report to the European Commission, which endorsed five principles: openness, participation, accountability, effectiveness, and coherence (Ozolina et al., 2009). The recommendations of A Framework for Addressing Ethical Dimensions of Emerging and Innovative Biomedical Technologies reflect many of these same principles, adapted for a biomedical context. These include principles focused on advancing the general public good, protecting the interests of those more specifically affected, ensuring integrity of the research process, engaging relevant communities, and ensuring oversight and accountability (NASEM, 2019b).
More directly applicable governance proposals may be found in the environmental arena. For instance, the Lisbon Principles for sustainable ocean management (Costanza et al., 1998) include responsibility (use resources in ways that are ecologically sustainable, efficient, and fair), scale-matching (consider and integrate across multiple scales), precaution (err on the side of caution, especially with respect to irreversible impacts), adaptive management (iteratively assess and adjust), full cost allocation (consider and include all social and ecological costs and benefits), and participation (engage stakeholders in decision making).
In the realm of natural resource management more broadly, Lockwood et al. (2010) drew on an expert panel, literature review, and work with Australian governance authorities to develop eight recommendations for governance. They argue that governance should be “legitimate, transparent, accountable, inclusive, and fair and… also exhibit functional and structural integration, capability, and adaptability.” These environmental governance recommendations aim to be responsive to social values, as well as to cope with “complexity, uncertainty, interdependency, and deficiencies in resources, expertise, and knowledge” (Lockwood et al., 2010). The emphasis on inclusivity and participation, integration, capability, and iterative assessment and adaptability particularly speak to these features. As indicated below, these latter elements often require approaches to governance that include but extend beyond
traditional mechanisms. In governance of science, this involves strengthening connections and facilitating greater communication between “science” and “society,” through ongoing engagement with stakeholders, publics, and decision makers and increased attention to “usable science,” developed in response to social values and needs.
3.4b Existing Proposals for Geoengineering Research Governance
Developed subsequent to the UK Royal Society report (Shepherd, 2009), the Oxford Principles (Rayner et al.,2009, 2013) were developed and presented to the British House of Commons, then later published in an academic journal. The Oxford Principles, which aimed to address “early research through deployment,” are as follows:
- Geoengineering to be regulated as a public good;
- Public participation in geoengineering decision making;
- Disclosure of geoengineering research and open publication of results;
- Independent assessment of impacts; and
- Governance before deployment.
In 2010, a conference at the Asilomar conference center in California brought together researchers to discuss geoengineering governance. The resulting report identified five principles as the basis for research governance, many of which overlap with the Oxford Principles:
- Promoting collective benefit;
- Establishing responsibility;
- Open and cooperative research;
- Iterative evaluation and assessment; and
- Public involvement and consent, with “consideration of the international; and intergenerational implications of climate engineering” (ASOC, 2010).
As described in Chapter 2, the Code of Conduct for Responsible Geoengineering Research (Hubert and Reichwein, 2015) was developed subsequently to provide a more specific set of rules that could be followed by researchers and others. Provisions relating to research discuss cooperation across jurisdictions, practices for responsible research, assessment of outdoor experiments, public participation, post-project monitoring of outdoor experiments, and open access to information.
More recently, Gardiner and Fragnière, 2018) extended and modified the Oxford Principles in their 10 Tollgate Principles for geoengineering governance:
- Framing. Geoengineering should be administered by or on behalf of the global, intergenerational, and ecological public, in light of their interests and other ethically relevant norms.
- Authorization. Geoengineering decision making (e.g., authorizing research programs, large-scale field trials, and deployment) should be done by bodies acting on behalf of (e.g., representing) the global, intergenerational, and ecological public, with appropriate authority and in accordance with suitably strong ethical norms (e.g., justice and political legitimacy).
- Consultation. Decisions about geoengineering research activities should be made only after proper notification and consultation of those materially affected and their appropriate representatives and after due consideration of their self-declared interests and values.
- Trust. Geoengineering policy should be organized so as to facilitate reliability, trust, and accountability across nations and generations.
- Ethical Accountability. Robust governance systems (including of authority, legitimacy, justification, and management) are increasingly needed and ethically necessary at each stage from advanced research to deployment.
- Technical Availability. For a geoengineering technique to be policy-relevant, ethically defensible forms of it must be technically feasible on the relevant time frame.
- Predictability. For a geoengineering technique to be policy-relevant, ethically defensible forms of it must be reasonably predictable on the relevant time frame and in relation to the threat being addressed.
- Protection. Climate policies that include geoengineering schemes should be socially and ecologically preferable to other available climate policies and focus on protecting basic ethical interests and concerns (e.g., human rights, capabilities, and fundamental ecological values).
- Respecting General Ethical Norms. Geoengineering policy should respect general ethical norms that are well founded and salient to global environmental policy (e.g., autonomy and justice).
- Respecting Ecological Norms. Geoengineering policy should respect well-founded ecological norms, including norms of environmental ethics and governance (e.g., sustainability, precaution, respect for nature, ecological accommodation).
Compared to the Oxford Principles, the Tollgate Principles provide more specificity regarding the interests that geoengineering research or deployment should serve: a “global, intergenerational, and ecological public.” Additionally, compared to the Oxford Principles, the Tollgate Principles make reference to a number of more substantive
ethical norms, including sustainability, precaution, respect for nature, justice, and human rights. The Tollgate Principles also more explicitly assert that forms of geoengineering should be ethically defensible, as well as technically feasible and reasonably predictable. Both the Oxford and Tollgate Principles share a commitment to trust and legitimacy, accountability, and engagement of affected publics or their representatives.
The Oxford Principles, Asilomar recommendations, and Tollgate Principles represent a subset of the proposals for principles of SG governance within a broader landscape of reports and proposals, plus a growing academic literature. Notably, there is significant convergence in certain basic requirements and desired features of research governance, especially in mainstream literature. However, there exists significant divergence in other areas, including regarding whether further research should be pursued. At one end of the spectrum, some oppose further research (Cairns, in Hulme, 2014; Long and Cairns, 2020) or the application of stringent conditions before additional research is undertaken (Whyte, 2012). At the more permissive end, proposals call for de-exceptionalizing SG research and limiting governance below a certain threshold (see, e.g., Parson and Keith, 2013). Despite differences, there seems to be consensus among commentators on governance that clearly defined governance mechanisms are needed for any future expansion of research. Jinnah (2019) argued that “good governance” of SG should promote fair distribution of benefits, protect vulnerable populations, and amplify marginalized voices. This underscores the idea that research governance needs to promote research and development that is fair and equitable, including concern for substantive impacts and inclusive processes.
Across the writing on SG research and research governance, certain key ideas and principles repeatedly emerge:
- International coordination and cooperation;
- International governance of any experiments with transboundary effects (and seeking to avoid transboundary harm);
- Public participation;
- Research in the public interest; and
- Legitimacy and accountability.
Also discussed repeatedly, but not explicit in all proposals, are the following:
- Fairness and inclusion;
- Intergenerational considerations; and
- Maximized benefits and minimized harms and risks.
Our subsequent research governance recommendations, discussed in Chapter 5, are informed by these principles.