It is widely accepted in the scientific community that climate change poses a severe threat to current and future generations, as well as to the rest of nature (IPCC, 2014). Nevertheless, the countries of the world are not currently on track to meet their stated goal of avoiding dangerous anthropogenic interference with the climate system (UNFCCC, 1992), understood in terms of the internationally agreed targets of limiting average global temperature rise to 2 degrees Celsius and pursuing efforts to achieve 1.5 degrees (Climate Action Tracker, 2015; UNEP, 2017; UNFCCC, 2015). Moreover, the political prospects for further robust action remain questionable. Consequently, many scientists are concerned that in practice deliberate large-scale technological interventions into the climate system (‘geoengineering’) are already, or may soon become, unavoidable if the 1.5 or 2 degrees targets are to be met (Anderson & Peters, 2016; Bawden, 2016; EASAC, 2018; Kriegler et al., 2018; Shepherd, 2016). At the same time, it is generally recognized that a drive toward geoengineering would have serious social implications, that ‘ethical considerations are central to decision-making in this field’, and that ‘analysis of ethical and social issues associated with research and deployment’ should be a central research priority (Shepherd et al., 2009, pp. 39, 53).
One of the earliest interventions comes in the form of the ground-breaking Oxford Principles (Rayner et al., 2013; Rayner, Redgwell, Savulescu, Pidgeon, & Kruger, 2009), which remain influential.1 In this paper, we build on the Oxford proposal, focusing on its ethical dimensions and in particular the ethical adequacy of its framing of geoengineering.2 First, we offer a detailed constructive critique of the Oxford Principles. Second, we propose an alternative set of principles based both on that critique and also on some standard work in practical ethics. We name these ‘the Tollgate Principles’, in part after the village pub in which the guidelines were originally developed, but also because in our view respecting the principles is ‘the price that must be paid’ by any attempt to frame and introduce an ethically defensible geoengineering policy.3 One upshot of the Tollgate Principles is that geoengineering becomes a much more ethically demanding enterprise than is often suggested. This has implications for how geoengineering policy is likely to evolve and especially for the prospects for ‘ethical geoengineering’.
1. The Oxford Principles
One approach to generating principles is broadly ‘bottom-up’. It proceeds by identifying the ethically salient features of geoengineering based on existing reports, experience from related cases, and so on. Another approach is broadly ‘top-down’. It confronts the issue of geoengineering from the perspective of foundational or mid-level theory (e.g. in moral philosophy, international political theory, global justice, etc.) and seeks to apply such theory directly to geoengineering. In our view, both approaches have a role to play and ideally will ultimately become integrated. As a way to push the debate forward, we employ the bottom-up approach. In doing so, we also identify some of the issues relevant to a top-down strategy.
Our starting point is the influential Oxford Principles, first put forward by a small group of distinguished academics at Oxford University in 2009. These principles were given qualified endorsements by the UK House of Commons report (House of Commons Science and Technology Committee, 2010) and the Asilomar report (ASOC, 2010), and their spirit and content were subsequently fleshed out in an article in Climatic Change (Rayner et al., 2013). The principles have played a pioneering role in the geoengineering debate, and we have a great deal of respect for the authors’ contribution. Our hope is to continue their necessary and important work by enriching the ethical discussion and preparing the ground for a wider, and possibly top-down, debate. Our background concern is that, despite the explicit intention to foster the debate about the ‘overarching societal values’ that should govern geoengineering policy (Rayner et al., 2013, p. 503), the original Oxford Principles are largely instrumental and dominated by procedural considerations. As a result, they do not sufficiently lay the groundwork for the more substantive ethical debate that is needed, especially around values such as justice, respect and legitimacy.4
The Oxford authors summarize their principles as follows:
(OP1) Geoengineering to be regulated as a public good.
(OP2) Public participation in geoengineering decision-making.
(OP3) Disclosure of geoengineering research and open publication of results.
(OP4) Independent assessment of impacts.
(OP5) Governance before deployment.5
Before assessing these principles directly, we offer some quick clarifications about our approach. First, the Oxford group follows the Royal Society in defining geoengineering as: ‘the deliberate large‐scale manipulation of the planetary environment to counteract anthropogenic climate change’ (Shepherd et al., 2009, p. 1). While some would reject this definition (e.g. as too permissive and overly vague) and others are skeptical of the term ‘geoengineering’ itself (e.g. Heyward, 2013; Jamieson, 2013), we aim to sidestep definitional discussions in this paper by assuming that we are discussing the paradigm case of stratospheric sulfate injection (SSI). The extent to which other interventions share the features that make all or some of the Tollgate Principles appropriate, and the question of whether these deserve the label ‘geoengineering’ are topics for another occasion (cf. Gardiner, 2016).6
Second, in their original form, each Oxford principle was accompanied by a brief text (Rayner et al., 2009), and in the later article, each is supplemented by a longer comment (Rayner et al., 2013). While it is not entirely clear whether the supplements are intended to define the principles, draw out implications, or something else, we shall assess each principle in conjunction with its accompanying remarks.
Third, one background question concerns the scope of the Oxford principles, and in particular whether they are intended to guide research and deployment, research alone, or even just early, small-scale research (e.g. excluding large-scale field trials). We believe that these tasks are not entirely separable, since governance of research (even near-term research) cannot help but be influenced by the wider aspiration of potential deployment and the norms that would govern such deployment. Nevertheless, in this paper, we will not take a stand on the interpretive issue. Instead, we assume that the ultimate aim of developing principles is to frame geoengineering – from early research through deployment – in ways that facilitate successful governance. Seen in this light, the question is how far the Oxford Principles assist in this project, however they were initially intended.
Fourth, the Oxford principles are often criticized as being too high-level or abstract to be useful (e.g. Nature Editorial, 2012). This is a common concern about governance principles in all areas. However, we agree with the Oxford authors that abstraction in this context need not be a problem and may be an advantage (Rayner et al., 2013). For instance, offering high-level principles often allows one to avoid prejudging more specific issues prematurely, to identify such issues, and to facilitate appropriately formed, justified and authoritative judgments about them. These are important elements of successful governance. Hence, our Tollgate principles will also be high level.
Finally, in any case, our primary intention is to influence the framing, tone and direction of geoengineering governance, rather than focus on the specific designation or wording of particular principles. Again, our aim is to enrich the ethical dimensions of the conversation and prepare the ground for a wider, and possibly top-down, discussion. Although we believe that the Tollgate principles are useful, we do not see them as the final word, but rather as another step on an ongoing journey.
1.1. Regulating a Public Good
The first Oxford principle states:
Oxford Principle 1 (OP1): Geoengineering to be regulated as a public good
While the involvement of the private sector in the delivery of a geoengineering technique should not be prohibited, and may indeed be encouraged to ensure that deployment of a suitable technique can be effected in a timely and efficient manner, regulation of such techniques should be undertaken in the public interest by the appropriate bodies at the state and/or international levels. (Rayner et al., 2009)
Unfortunately, this principle provides a problematic framing of geoengineering.7 As John Virgoe put it in his testimony to the UK House of Commons, ‘once you peer below the surface of the public good, it becomes quite hard to define it and you get into some difficult ethical territory’ (House of Commons, 2010, Ev 12). On our view, this is due to an awkward ambiguity. The phrase ‘public good’ has informal, colloquial uses, but also a number of closely related technical meanings in economics and international relations, often with specific, but potentially conflicting, policy connotations.8 This threatens to make framing geoengineering as a ‘global public good’ seriously misleading, especially in the public sphere. Consequently, we will advocate for a more transparent, and explicitly ethical, approach which emphasizes central values, such as justice and political legitimacy.
Let us begin with three uses of ‘public good’ and ‘global public good’. First, one minimal colloquial understanding is that of something that is not (or not primarily) a private concern, but a public one that should be governed or regulated as such. This sense of ‘public good’ is suggested by the 2009 text accompanying OP1 (above).
Second, in economics and public policy, the common technical conception of ‘public good’ defines a pure public good as a good that is both nonrival and nonexcludable. A good is nonrival if and only if one person’s consumption of the good does not limit or inhibit another person’s consumption. A good is nonexcludable if and only if, once it is available to some, others cannot be prevented from consuming it. A standard example is the good provided by a lighthouse. It is nonrival: one sailor’s being able to see the rocks does not limit or inhibit others being able to do the same thing, and vice versa. It is also nonexcludable: once the lighthouse is illuminating the rocks for some sailors, others cannot be prevented from seeing them too. These features of pure public goods are also emphasized in the geoengineering context, including by the Oxford authors (Rayner 2011, p. 11).
Third, a further colloquial conception of...