The Paradox of Climate Engineering

Michael ZürnWissenschaftszentrum Berlin für Sozialforschung

Stefan SchäferBerlin Graduate School for Transnational Studies

AbstractClimate engineering technologies, sometimes also referred to as geoengineering technologies, attempt to ward offthe worst effects of climate change by intervening in the global climate system. We see the potentials offered byclimate engineering technologies in counteracting the threats of climate change but also take into account the risksthat arise from the side effects of these technologies on natural, social and political systems. We find a paradox ofclimate engineering, which consists in the circumstance that exactly those technologies that are capable of acting fastand effectively against rising temperatures at comparatively low costs, are also the technologies that are likely tocreate the greatest amount of social and political conflict. To address this apparent paradox, we argue that aninstitutional setting for researching and potentially deploying climate engineering technologies is needed whichcreates a sufficiently high degree of social and political legitimacy and addresses a set of specified problemsconnected to climate engineering. We present a proposal for such an institutional setting that explicitly addressesthese concerns.

Policy Implications• Research on and potential uses of climate engineering technologies need to be coordinated internationally in a

multilateral institutional setting.• An international climate engineering agency should be created that coordinates and disseminates research on

climate engineering.• Research results should be evaluated by the IPCC.• Decision-making on climate engineering should occur within the UNFCCC, where the states party to that conven-

tion should decide on norms and rules that govern climate engineering (regarding, for example, an upper limit formanipulations of the radiation balance, a uniform metric for making different responses to climate change compa-rable, and a time limited moratorium on field tests and deployments of climate engineering technologies).

1. International cooperation on engineering theclimate?

A set of newly emerging high technologies, aimed ataltering the global climate in an effort to counteract ris-ing global mean temperatures, has recently begun toenter the scientific, public, and political debates. Thesetechnologies, referred to collectively as climate engineer-ing technologies, attempt to offset global climatechange by either reducing the concentration of CO2 inthe atmosphere (carbon dioxide removal (CDR)) or byreflecting sunlight away from earth (solar radiation man-agement (SRM)). The amount of attention directed atclimate engineering technologies will be magnified bythe International Panel on Climate Change’s (IPCC) fifth

assessment report in 2014, which will address thesetechnologies (IPCC, 2012).

Climate engineering promises a cheap solution for glo-bal climate change. At the same time, the unintendedside effects on natural, social and political systems andthe distributional consequences resulting from possibleuses of these technologies are highly uncertain, andlikely to involve ‘unknown unknowns’. Therefore, it doesnot come as a surprise that the development anddeployment of these technologies are contested.

In this contribution, we argue that the advantages ofinvesting in these technologies are dependent upon cer-tain institutional pre-requisites. On the one hand, weassume that it is useful to have a high leverage technol-ogy available if a climate crisis occurs. Climatic changes

Global Policy

Global Policy (2013) doi: 10.1111/1758-5899.12004 ª 2013 University of Durham and John Wiley & Sons, Ltd.

Research

Article

1

can be swift and ugly, creating unpleasant outcomesthat are hard to tolerate. In such a case, adaptation isimpractical (Victor, 2011, p. 185) and additional reduc-tions of CO2 emissions do not help since their effectswould be felt only decades later. In the absence of suffi-cient emission reductions, climate engineering should beresearched as a potential way to avoid such outcomes.On the other hand, we see a paradox of climate engi-neering that can only be overcome if a rightly designedinternational institution is in place. Only very few of theproposed technologies fulfill the promise of acting fastat a low price, making them suitable as responses tointolerable climatic changes. Many other climate engi-neering technologies are as cost-intensive and slow ineffect as reducing CO2 emissions is. Moreover, exactlythose few technologies that promise to act fast at a lowprice would also bear the greatest risk of creating politi-cal and social resistance and conflict. It is for this reasonthat the successful development and deployment of cli-mate engineering depends on an international institu-tional setting which is able to cope with the social andpolitical side effects of climate engineering technologiesthrough appropriate regulation.

We begin by examining the cooperation requirementsfor climate engineering from two perspectives. First, weask what formal cooperation requirements climate engi-neering poses based on its implementation costs and itstechnical requirements. Which climate engineeringoptions are amenable to unilateral or minilateral imple-mentation and at the same time capable of quickly andsignificantly manipulating the global average tempera-ture? It emerges that some – but far from all – climateengineering technologies can actually be implementedon a unilateral or minilateral basis with the goal ofsignificantly influencing global mean temperatures(section 2).

Second, we expand this purely rationalist analysis insection 3 to include sociopolitical factors, asking whatsocial and political side effects might arise from a unilat-eral or minilateral implementation of climate engineeringtechnologies. Which technologies are likely to provokesocial and political conflict in the absence of a legitimatemultilateral approach to their research and implementa-tion? We argue here that a unilateral or minilateralimplementation of climate engineering measures couldtrigger a series of problematic consequences, which ren-der the integration of research on and implementationof climate engineering technologies into a multilateralnegotiation process advisable.

Sections 2 and 3 each result in a typology, whichtogether display analytical power. Most typologies ofclimate engineering technologies are based on natural sci-ence concepts, such as the differentiation between CDRand SRM technologies. We offer two typologies to mapclimate engineering technologies which are based on

social science concepts. The first typology shows whichtechnologies are in fact amenable to unilateral imple-mentation by laying open the requirements for this – theymust be efficient (i.e., cheap and effective) and amenableto centralized implementation (as opposed to technolo-gies that require implementation on the territories ofmany sovereign states to be effective). While this firsttypology explicates well known economic reasoning, thesecond typology uses sociological and political reasoningin order to show which technologies are most likely tocreate considerable social and political contestation andconflict. The fact that these technologies turn out to bethe same as those that are amenable to efficient unilateralimplementation in the first place lets us speak of a ‘para-dox of climate engineering’.

Given the paradox of climate engineering, we ask,thirdly, about the institutional prerequisites for the suc-cessful development and implementation of climateengineering technologies (section 4). We argue that acertain institutional design is necessary and desirable forinternationally governing climate engineering. The insti-tutional solution we propose aims at avoiding severalproblems that arise in the context of high leverage cli-mate engineering: lack of social and political acceptance,moral hazard, slippery slope and prohibitively hightermination costs.

2. The virtues of climate engineering?

Nobel laureate Tom Schelling, in an early piece on thetopic, concludes that climate engineering will fundamen-tally change the way we think about climate change,reducing the issue to the simple question of who isgoing to pay for the costs of engineering the climate(Schelling, 1996, p. 306). We refer to the understandingthat climate engineering provides a unilaterally imple-mentable solution to the problem of rising global meantemperatures as the ‘Schelling thesis’. Schelling’s thesisregarding the politically simple implementation logic ofclimate engineering is ultimately based on the assump-tion of a vast efficiency of such measures from an eco-nomic point of view. This is vividly expressed by the titleof an article by Scott Barrett, ‘The Incredible Economicsof Geoengineering’, in which he states programmatically:‘In contrast to emission reductions, this approach [geo-engineering] is inexpensive and can be undertaken by asingle country, unilaterally’ (Barrett, 2008, p. 45).

The central thesis of the unilateral resolvability of theproblems associated with a rise in global mean tempera-tures through climate engineering is based on twoassumptions. First, the costs of such measures must beso low that they can be burdened by an individual stateunilaterally or by a small group of states (minilateralism).Additionally, their effectiveness must be so great thatthey enable the implementing state or group of states

Michael Zürn and Stefan Schäfer2

ª 2013 University of Durham and John Wiley & Sons, Ltd. Global Policy (2013)

to significantly manipulate global mean temperatures. Ifa technology meets both of these criteria, we consider itto be highly efficient. Second, in order to be implement-able uni- or minilaterally, climate engineering technolo-gies need to be amenable to implementation on theterritory of a single state, on a limited number of stateterritories, or in common spaces outside of nationaljurisdiction.1

The economic efficiency of climate engineering

Are climate engineering technologies indeed cheap andeffective (i.e. efficient), and can they be deployed cen-trally in one location? Only if these conditions are metcan a climate engineering technology be considered uni-laterally or minilaterally implementable and potentiallycapable of ‘totally transform[ing] the greenhouse issue’(Schelling, 1996, p. 305). In order to examine thesepropositions, we need to take into account the results ofeconomic and natural scientific research.2

We operationalize the costs, effectiveness and amena-bility to centralized implementation of climate engineer-ing technologies in the following way. Regarding costs,we are interested only in the direct costs that arise fromthe unilateral implementation of the measure in ques-tion, and not in external costs that may eventually arisein other locations – only the former are decisive in termsof the Schelling thesis. A climate engineering technologyis considered highly efficient if it is not only affordablein this sense but also effective, meaning it must be capa-ble of significantly changing global mean temperaturesas a single measure in a relatively short period of time.Additionally, in order to be unilaterally implementable, aclimate engineering measure must be amenable to cen-tralized implementation. This means that it must be ableto significantly change global mean temperatures whendeployed on the implementing state’s territory, on theterritories of a small group of cooperating states or incommon spaces outside of national jurisdiction. Should aclimate engineering measure require decentralizedimplementation (i.e., in the territories of many states) inorder for it to achieve a significant effect on global meantemperatures, then by its very nature it cannot beemployed unilaterally with the goal of significantly influ-encing global mean temperatures, and would require amultilateral approach.3

Based on this operationalization, it emerges that farfrom all climate engineering technologies conform tothe Schelling thesis. Only stratospheric particle injectionand marine cloud brightening are both efficient and uni-laterally implementable. All the other climate engineer-ing technologies discussed in the literature are either:

• too expensive from the beginning, such as mirrors inspace (see Robock, 2008; Royal Society, 2009, p. 45),

• too expensive when scaled up to an effective scale,such as afforestation of the Sahara and the Australianoutback (see Ornstein, 2009), covering the world’s des-ert areas with reflective material (see Royal Society,2009) and enhancing rooftop reflectivity (see Akbariet al., 2009; Royal Society, 2009; Oleson et al., 2010),

• not effective as a single measure to stabilize globalmean temperatures on a short timescale (enhancedweathering), or require decentralized implementationto be effective in that sense, such as direct air capture.

The only CDR measure that has by some been consid-ered as potentially highly efficient is ocean iron fertiliza-tion. This technology is also amenable to centralizedimplementation. However, the extent of the effectivenessof this measure remains uncertain. Lenton and Vaughan(2009, p. 5593) argue that this measure is not suited forsignificantly influencing global mean temperatures on ashort timescale. There is also uncertainty with regard tothe costs of extensive ocean fertilization. In view of this,the efficiency of this measure can also be consideredinsufficiently high for it to be amenable to unilateral orminilateral implementation.

Figure 1: Typology 1 below sums up these findings. Itemerges that only stratospheric particle injection andmarine cloud brightening are amenable to unilateral orminilateral implementation. Large scale direct air capture,large scale afforestation, the enhancement of rooftopreflectivity, the modification of deserts to reflect moresunlight, ocean fertilization, reflectors in space, andenhanced weathering are either too expensive, not effec-tive enough to significantly influence global mean tem-peratures, or require decentralized implementation, thusrequiring the participation of many states and conse-quently a multilateral approach to their implementation.For example, measures such as the massive afforestationof the Sahara and the Australian Outback or the coveringof the world’s deserts with reflective material cannot becarried out unilaterally, but by their very nature require amultilateral approach. The implementation of such pro-jects appears highly unlikely.

3. The social and political miseries of aunilateral or minilateral climate engineeringpolicy

The thesis of a unilateral or minilateral solution to theproblem of rising global mean temperatures throughclimate engineering is based on the rationalist theory ofcooperation in international relations (compare Schel-ling, 1960; Keohane, 1984, Zürn 1992). The simplest ver-sion of this theory, from which the Schelling thesis isderived, is based on two premises. First, it assumes thatstates are the key actors in international politics andcan act relatively autonomously, meaning without

The Pardox of Climate Engineering3

Global Policy (2013) ª 2013 University of Durham and John Wiley & Sons, Ltd.

consideration for transnational norms and actors orchanging domestic interest constellations, only takinginto account the commitments they made regardinginternational law. Second, it assumes that state interestswith regard to a specific problematic situation are fixedand are not influenced by international negotiation pro-cesses. This approach is powerful for identifying inter-governmental interest constellations, but runs the riskof losing sight of the dynamic component of politicalinteraction processes and the role of transnationalnorms and actors.

On the social and political consequences of aunilateral climate engineering policy

It can be expected that the unilateral implementation ofa climate engineering technology will result in the com-prehensive politicization of such activities. Decision-mak-ing processes, institutions and policy outcomes areconsidered politicized when they become the target ofstrong social mobilization with a high degree of contes-tation (on the politicization of international institutions,see Zürn, Binder and Ecker-Ehrhardt, 2012). Transnationalnorm entrepreneurs such as Greenpeace and other non-governmental organizations (NGOs) can mobilize a highlevel of political resistance by recourse to internationallyrecognized norms (Keck and Sikkink, 1998). It is espe-cially likely for such a development to occur in Europe,since the potential for mobilizing the general publicsagainst the introduction of new technologies has beentraditionally high on the continent. Moreover, one canexpect that the North-South antagonism in internationalclimate policy would likely be intensified in case of

unilateral deployment of efficient climate engineeringtechnologies.

In the context of the Conference of the Parties to theCBD, the international community already submitted inpart to the pressure exercised by several transnationalNGOs with the support of some developing countries.The following was stated in a decision (UNEP ⁄ CBD ⁄ COP10 Decision X ⁄ 33):

[The Conference of the Parties […] invites Par-ties and other Governments […] to […] ensure]that no climate-related geo-engineering activi-ties that may affect biodiversity take place, untilthere is an adequate scientific basis on which tojustify such activities and appropriate consider-ation of the associated risks for the environ-ment and biodiversity and associated social,economic and cultural impacts, with the excep-tion of small scale scientific research studiesthat would be conducted in a controlledsetting.

This very vague formulation and the legally non-bindingnature of the CBD do not make it impossible for anystate to field-test or implement climate engineeringmeasures. Nevertheless, the inclusion of the topic in theconvention is indicative of an increased awareness ofclimate engineering on the part of societal and politicalactors.

Growing social resistance against climate engineeringcan also be witnessed in other places. For example, thetransnational NGO ‘ETC Group’ is active in the area ofclimate engineering. The ETC Group sets forth several

High efficiency Low efficiency

Central implementation possible

(1)� Stratospheric particle

injection� Marine cloud

brightening

(2)� Ocean fertilization� Space reflectors� Enhanced weathering

(land or ocean)

Decentralized implementation required

(3)� Large scale direct air

capture

(4)� Large scale

afforestation� Enhanced rooftop

reflectivity� Desert modification

Figure 1. Typology 1: Unilateral implementability of different climate engineering technologies.4

Source: Author.

Michael Zürn and Stefan Schäfer4

ª 2013 University of Durham and John Wiley & Sons, Ltd. Global Policy (2013)

arguments against climate engineering, includingunequally distributed regional effects (ETC Group, 2009),the possibility of military use, and the potential role pri-vate sector interests could play (ETC Group 2010a, 37ff.).They also recur to international law in their rejection ofclimate engineering (for example to the ENMOD treatyin ETC Group, 2010b).

A further politicization of decisions regarding climateengineering and the institutions involved in makingthese decisions – which can already be witnessed in itsbeginnings – can be expected for several reasons. First,climate engineering fulfills all of the preconditions forthe mobilization of social resistance that generally arisefrom the technocratic imposition of risk technologies ina global risk society (see Beck, 2008). Therefore, a strongconnection can be expected between social movementsin the implementing countries and a correspondingtransnational protest movement. Second, in view of thenon-involvement of technologically less powerful states,the hitherto unclear distribution of climate effects andthe side effects arising from an implementation of cli-mate engineering could evoke the North-South schismon a comprehensive scale and cause permanent damageto the United Nations Framework Convention on ClimateChange (UNFCCC) process as a target of anti-hegemonicresistance (Rajagopal, 2003). Finally, both forms of resis-tance, social and political, could invoke general princi-ples of international law (UNCLOS, Art. 195; Outer SpaceTreaty, right of consultation; Antarctic Treaty, peacefulpurposes; Montreal Protocol, protection of the ozonelayer; etc.), as well as the more specific provisions in theCBD and the London Convention and its Protocol. Thesenorms do not contain clear provisions prohibitingclimate engineering. They provide, however, a suitablenormative anchor to serve as ‘political opportunity struc-ture’ for resistance (della Porta and Tarrow, 2005; Tarrow,2005).5

Moreover, dynamic processes of conflict escalation areconceivable. Third countries negatively affected by uni-lateral climate engineering could resort to radical oppo-sition in the context of the UNFCCC process, and, in avery extreme case, even adopt counter measures –‘counter-climate engineering’, (Lane, 2010). To this end,fluorocarbons could be deployed to counter the coolingeffect from climate engineering, or black coal could bereleased to decrease the earth’s reflectivity (see also Hor-ton, 2011, p. 62).

In general, should environmental damage arise in athird country as a result of a unilateral climate engineer-ing intervention, this is likely to lead to internationalconflict. Conflict might even occur in the absence of aclearly established causal link between environmentaldamage in a third country and a unilateral climate engi-neering intervention. A unilateral climate engineeringintervention could thus be blamed for the occurrence of

weather events even if it is not entirely clear that it wasthe climate engineering intervention that caused theseevents, substantially increasing the potential for interna-tional conflict.

In any case, the current absence of indications for amanifestation of interstate conflicts should not lead topredictions concerning their (un)likelihood in case cli-mate engineering is implemented unilaterally. The regio-nal effects of climate interventions will be unevenlydistributed and thus are likely to lead to conflicts thatcould undermine the basis of cooperation on climatepolicy.

The social and political side effects likely to arise inreaction to a unilateral or minilateral implementation ofclimate engineering technologies bear witness to thelegitimacy deficit inherent to such an approach, and canbe expected to undermine the success of the imple-mented technologies. The most important circumstanceout of which this legitimacy deficit arises is the fact thatnot only those states that proceed to intervene in theglobal climate are affected by this intervention. There isthus a lack of ‘input congruency’ (Zürn, 1998, p. 237) –not all parties affected by the intervention are involvedin the decision-making process leading up to the inter-vention.

The specific social and political consequences ofindividual climate engineering technologies

In general terms we maintain that a unilateral or minilat-eral climate engineering intervention can have social andpolitical consequences that undermine its chances forsuccess. In order to undertake a more detailed assess-ment of this assertion, a second typology (Figure 2) ofclimate engineering technologies is introduced below.Here, climate engineering technologies are differentiatedbased on a political-legal perspective. We examinewhether their implementation involves common spacesoutside national jurisdiction (on this concept see Wolf-rum, 1984) or state territories, and whether the unde-sired side effects resulting from their implementation arelikely to remain local or tend to be global in nature.

First, we expect the social and political side effects ofa unilateral or minilateral climate intervention to beespecially strong in cases which involve the use of com-mon spaces outside national jurisdiction, where no claimto sovereignty can be made. Second, we expect thatmeasures producing undesired side effects that tend tobe global in nature are conducive to strong politiciza-tion. On this basis, we can now identify the measures infield 2 as being at an especially high risk of underminingtheir own effectiveness by provoking social and politicalconflict when implemented unilaterally or minilaterally.This concerns exactly those measures that are sufficientlyefficient to render their unilateral use in a climate

The Pardox of Climate Engineering5

Global Policy (2013) ª 2013 University of Durham and John Wiley & Sons, Ltd.

emergency possible in the first place. In other words,precisely those measures which we identified above aspotentially unilaterally or minilaterally implementable arethose that incorporate the greatest potential in terms ofgenerating politicization and resistance. Here, theexplanatory character of our second typology emerges:even those few technologies that conform to the Schel-ling thesis are unlikely to be successfully implementedby a single state or a small group of states, due to theirinherent conflict potential. This is what we label the par-adox of climate engineering (see also Rayner, 2010).

This paradox is what seems to drive NGOs that opposeclimate engineering. ‘Hands Off Mother Earth’ (HOME) isan association of different organizations which is exclu-sively devoted to resistance against climate engineeringtechnologies.6 Its objective is ‘[…] to build a globalmovement to oppose real world geoengineering experi-ments […]’ (HOME, 2011a). The climate engineeringtechnologies mentioned on the HOME website andagainst which the organization’s protest is directed arethe production and storage of biochar, the introductionof sulfur particles into the stratosphere, marine cloudbrightening and the storage of CO2 in the oceanthrough nutrient fertilization (HOME, 2011b). With theexception of the production and storage of biochar,these are all measures that have extensive trans bound-ary effects and can be implemented in common spacesoutside national jurisdiction.7

This means in turn that climate engineering technolo-gies which are not amenable to unilateral or minilateralimplementation are likely to generate less social and politi-cal conflict. The modification of rooftops, afforestation,the enhancement of natural weathering and air captureoffer a comparatively low conflict potential. As yet, nosignificant social or political mobilization processes have

arisen in opposition to these proposals. Also, proposalsfor scaling up originally local measures to increase theirleverage, such as ‘greening’ the Sahara and the Austra-lian outback or covering the world’s deserts in reflectivematerial, have not led to notable reactions. This is likelyto be to a large degree due to the clear requirement ofmultilateral action on such projects, which makes themappear less threatening, and the associated fact thatsuch projects are highly unlikely to be realized.

The examination of the social and political conse-quences of individual climate engineering technologiesshows that climate engineering faces a seemingly irre-solvable paradox. The promise of a fast and highly effec-tive solution to the problems associated with risingglobal mean temperatures, which is so cheap that it canbe implemented by a few states on behalf of all human-ity, is provided by only two climate engineering technol-ogies: stratospheric particle injection and marine cloudbrightening. However, these are precisely those technol-ogies that are likely to lead to particularly vehementpoliticization and far-reaching social and political resis-tance with potentially devastating consequences for theUNFCCC process.

4. Overcoming the paradox: Proposal of aninstitutional design for the governance ofclimate engineering technologies

In spite of the paradox of climate engineering, researchon high leverage climate engineering technologiesseems in principle advisable for several reasons. Due tothe high costs arising from the necessary transition to acarbon neutral economy, all options that could contrib-ute to reducing the costs of this transition should beexplored. Furthermore, the difficulty of achieving interna-

Undesired side effects tend to remain local

Undesired side effects tend to be global in nature

Implementation occurs in common spaces outside national jurisdiction

(1)� Ocean fertilization� Enhanced weathering

(ocean)

(2)� Mirrors in space� Stratospheric aerosols� Marine cloud

brightening

Implementation occurs on state territory

(3)� Air capture� Enhanced weathering

(land)� Enhanced rooftop

reflectivity

(4)� Large-scale

afforestation� Albedo modification

of deserts

Figure 2. Typology 2: Extent of the social and political consequences of climate engineering unilateralism.

Source: Author.

Michael Zürn and Stefan Schäfer6

ª 2013 University of Durham and John Wiley & Sons, Ltd. Global Policy (2013)

tional agreement on reducing emissions and the longev-ity of greenhouse gases in the atmosphere make it desir-able to have an option available that allows us to shaveoff the worst effects of climate change.8 Most impor-tantly, some authors have suggested that climatechange might occur dramatically and abruptly if so-called ‘tipping points’ are passed (on tipping points inthe climate system see Lenton et al., 2008). If researchedtoday, highly effective climate engineering technologieswould be made available for cases of ‘climate emergen-cies’ (Victor, 2011, chap. 6).

However, research on and implementations of climateengineering technologies can only overcome the para-dox and produce the desired results if four potential sideeffects are addressed. First, research on climate engineer-ing technologies needs to possess a sufficient amount ofsocial and political acceptance. In the absence of thisacceptance, negative social and political reactions arelikely to preempt the development of a possibly impor-tant technological option to counteract climate change,which might be needed in the future. Second, ‘moralhazard’ effects need to be avoided. This term refers tonegative effects of research on or implementations of cli-mate engineering technologies on emission reductionefforts. Third, the slippery slope effect needs to berestricted. This describes situations in which conductingresearch on climate engineering technologies mightincrease the likeliness of deployment. For example, thiscan refer to the fear that an institution conductingresearch on climate engineering measures woulddevelop an interest in conducting further research andeventually pressure for the use of climate engineeringtechnologies, even in the absence of a situation thatwould warrant such a course of action. Finally, since thepremature abandonment of an ongoing climate engi-neering intervention might have catastrophic conse-quences, the fourth criterion consists in avoiding suchpremature abandonment as far as possible and in mini-mizing the risks associated with it. This has also beenreferred to as the termination problem. Research on andpossible implementations of climate engineering tech-nologies thus require an institutional setting which cre-ates a sufficiently high degree of social and politicalacceptance, specifically addresses the moral hazard andthe slippery slope effect, and ameliorates the terminationproblem. The fulfillment of these criteria – each ofthem mostly discussed in separate discourses – is neces-sary to overcome the climate engineering paradox. Thus,an institutional solution is offered to overcome theparadox.

A multilateral approach to climate engineeringwould be capable of reducing these side effects of cli-mate engineering technologies. Below, we propose sixfeatures for a multilateral institutional setting thatdirectly address the criteria developed above in the

context of the paradox of climate engineering, identi-fied in sections 2 and 3.9 This institutional settingdraws on core elements of the existing regime com-plex for international climate policy (Keohane andVictor, 2011), including the UNFCCC, the IPCC, and theKyoto Protocol.10

We develop our proposal for an appropriate institu-tional setting against the background of analyses aboutthe effectiveness of international environmental regimes(Haas et al., 1993; Young, 1999; Miles et al., 2001, Bre-itmeier et al., 2006) and corresponding research on inter-national institutional design (Ostrom, 1990; Koremenoset al., 2001). We also take into account the existing liter-ature on the multilateral regulation of climate engineer-ing.11 Virgoe’s (2009) considerations on the internationalregulation of climate engineering come closest to theproposal presented in the following section. He seesthe danger of strong international tensions emerging inthe event of a unilateral deployment, which would resultfrom the lack of legitimacy inherent to such anapproach. A consortium of states on the other handwould be exposed to a conflict of objectives betweenthe ability to act on the one hand and the requirementof achieving a high degree of legitimacy on the other,and would be unstable in the long run. He thus opts fora multilateral framework. Such a multilateral frameworkcan in our view contain minilateral initiatives by a clubof states capable and willing to do research, as long asthis is embedded within a larger framework (see Victor,2011, chap. 7). Our proposal consists of three principlesand six components. Three institutional principles mustguide research and deployment of climate engineeringin order to overcome the four potential side effects iden-tified above:

• transparent coordination of efforts should dominateresearch and deployment competition in order toachieve social and political acceptance, and to ensurethat a potential future deployment takes place basedon the best available scientific knowledge;

• institutional integration with existing climate policyshould dominate institutional fragmentation in orderto ensure that past achievements in setting emissionreduction requirements do not suffer, and that futureefforts to this end are not preempted, thus avoidingmoral hazard;

• a clear distinction between research and deployment isneeded to allow for political decisions after sufficientknowledge is available, thus avoiding slippery slopeeffects and the termination problem.

More specifically, we see six components of such aninternational institution as necessary.

1. An international climate engineering agency shouldcoordinate research into individual climate engineering

The Pardox of Climate Engineering7

Global Policy (2013) ª 2013 University of Durham and John Wiley & Sons, Ltd.

technologies.12 To this end, a coalition of states needsto be formed that is prepared to finance and carryout the corresponding research in a transparent man-ner. To ensure incentives to participate, memberstates should be able to credit contributions againstCO2 efforts (see also component 4). To ensure broadacceptance, information on funded projects needs tobe made available to the public, perhaps through anonline registry. While members of the club should beasked to finance their climate engineering researchefforts exclusively via the agency, we do not expectthere to be no research on climate engineering out-side of agency-funded projects; however, the availabil-ity of funding per se, paired with the high visibilityand legitimacy gains associated with agency-fundedresearch and the associated assurance of a smoothresearch process, should provide a strong incentivefor scientists to engage with the agency and applyfor funding. Researchers conducting research projectson climate engineering that are not funded throughthe agency should be encouraged to also registertheir projects, leading to a strong norm of transpar-ency. This will create a vibrant research program thatsets standards and norms for research on climateengineering. In sum, the agency is intended to ensurethe availability of funding for climate engineeringresearch, increase the transparency of research, con-tribute to the development of norms and standardsin climate engineering research, and to conduct insti-tutional oversight according to a regulatory code thatis to be developed.13 This would generally increasethe legitimacy of climate engineering research.

2. The assessment of the research outcomes, however,should be undertaken by the Intergovernmental Panelon Climate Change (IPCC). Three reasons speak forthis task sharing arrangement between the IPCC andthe proposed climate engineering agency. First, dueto the enhanced consultation commitments of theIPCC, the range of the actors involved is considerablywider than in the proposed climate engineeringagency, the members of which are to be drawn fromcountries actively involved in research. Charging theIPCC with the assessment of research results wouldthus increase the social and political acceptance ofthis research.14 Second, the assessment of researchinto climate engineering on the one hand and ofresearch into climate change more generally on theother would occur in an integrated manner, thus alsotaking into consideration climate policy alternativesand the overall problem of climate change. Thiswould reduce the risk of climate engineering beingperceived as an alternative to emission reductions,thus reducing the risk of moral hazard. Most impor-tantly, the institutional interest of a climate engineer-ing agency in increasing its own standing and

relevance through positive evaluations of climateengineering research would be preempted, thusreducing the risk of a slippery slope effect. Very muchin line with this suggestion, such an assessment of cli-mate engineering research will be part of the IPCC’sFifth Assessment Report (AR5). The IPCC’s ExpertMeeting on Geoengineering also suggested drafting aspecial report on the topic.

3. On the basis of the IPCC assessment, the memberstates of the UNFCCC should then proceed to makedecisions regarding norms and rules for governing cli-mate engineering.15 These norms and rules shouldgovern which technologies are to be further devel-oped and made ready for deployment, which fieldresearch will be allowed for this purpose, and whichtechnologies are to be deployed under which condi-tions. The checks and balances created by an institu-tional setting in which the climate engineeringagency functions as research coordinator, the IPCC asevaluator and the UNFCCC as political decision makerallows for a clear separation between research anddeployment, thus helping to avoid slippery slopeeffects. While the UNFCCC process, including theIPCC, seems for many discredited after the ongoingfailure to achieve a strong international climateregime to reduce CO2 emissions, there is no alterna-tive for multilateral norms and rules than agreementby the member states. Moreover, by introducing cli-mate engineering to the package that is negotiated,the current deadlock could be overcome since itstrengthens those states that have an interest in act-ing on climate change and broadens the portfolio ofavailable measures. It thus creates new possibilitiesfor issue-linkage, introducing to the negotiationsnovel responses to climate change that are not (yet)strongly politicized.

The following three components provide furthersuggestions for potential rules that should guide suchan institutional effort. Such rules will need to beworked out in more detail, of course.

4. To avoid moral hazard and a one-sided focus on cli-mate engineering, climate engineering technologiesshould be made comparable to conventional emissionreductions by installing a uniform metric for compari-son. Since climate engineering technologies, andespecially SRM technologies, are not of themselvescomparable to mitigation due to the huge differencesin effectiveness and costs, this needs to be achievedvia an external factor. A price conversion mechanismthus should be created which compensates for thehigher effectiveness of climate engineering technolo-gies where this is appropriate. The contributions ofstates to the costs arising from the development anddeployment of climate engineering technologiesshould thus be measured according to how much the

Michael Zürn and Stefan Schäfer8

ª 2013 University of Durham and John Wiley & Sons, Ltd. Global Policy (2013)

same contribution would have achieved wheninvested in conventional mitigation, thus adjustingthe comparison between the costs of mitigation andthe costs of climate engineering according to thedesired goal of avoiding a use of climate engineeringas a substitution for mitigation. This would create costequivalence, as opposed to effectiveness equivalence,between different climate engineering technologiesand conventional emission reductions.

As a result, it would be possible to extend the port-folio of climate-effective measures whose implementa-tion can be accredited to the reduction targets laiddown in the Kyoto process in the future. The imple-mentation of direct air capture, enhanced naturalweathering, and increasing the reflectivity of rooftopscould thus be accredited to international reductiontargets,16 as is already the case with afforestation.17

Within this uniform metric, CO2 could be selected as areference measure, as already applies to the conver-sion of the climate effects of other greenhouse gasessuch as CH4 or N2O. This is intended to preempt aninterpretation of climate engineering technologies asan alternative to emission reductions by ranking theireffectiveness adjusted to the effectiveness of conven-tional mitigation. Similarly as an upper limit on con-scious interventions into the radiation balance (seecomponent 6), this measure would contribute to anunderstanding of SRM as a transition technologycapable of helping societies avoid the worst effectsof global climate change without substituting formitigation.

5. To limit the risk of a slippery slope effect, a time-lim-ited moratorium on the implementation and fieldtesting of climate engineering technologies should beinstalled. The technologies in field 3 of typology 2(those which can be implemented on sovereign terri-tory and whose effects remain local) need to beexempted from such a moratorium. Research on allother technologies would be affected by this mea-sure. The moratorium needs to be time-limited so asnot to prematurely prescribe an institutional block-ade.18 However, so as not to initiate a slippery slopeeffect after the moratorium has expired, field testingand deployment would still need to follow the rulesset down by the UNFCCC.

6. In order to counteract the termination problem, astate should be obliged to significantly increase itsemission reduction efforts should it abandon a multi-lateral climate engineering effort. In order to keep theproblem in a feasible range, a possible base rulecould be to decide not to alter the radiation balanceby more than a certain value, such as 1 W ⁄ m2. Thiswould defuse the termination problem to a certainextent, since the pressure to reduce greenhouse gasconcentrations would remain stronger than under a

scenario in which forcing is offset to a larger degree.This would point in the direction of an understand-ing of SRM as a means to shave off the worsteffects of climate change, rather than as a completesubstitute for emission reductions. In addition, thiswould likely reduce the potential extent of negativeside effects on natural systems in comparison to ascenario in which a stronger intervention is under-taken. Coupling emission reduction requirements tothe exit option of a multilateral climate engineeringintervention would provide a strong legal basis forpressuring individual states to reduce their emis-sions. The costs associated with the avoidance ofaccelerated climate change in the case of prematuretermination could thus be imposed on the countriesthat were involved in driving the climate engineer-ing intervention.

Conclusions

The above analysis shows that an institutional setting forthe governance of climate engineering research andimplementation is desirable which ensures sufficientinternational and transnational social acceptance andintegrates climate engineering with existing climate regu-lations in such a way that negative social and politicaleffects can be avoided. To this end, we presented a rec-ommendation which proposes the creation of a climateengineering agency for coordinating and conductingresearch, an assessment of research results through theIPCC, the definition of norms and rules governing climateengineering through the member states of the UNFCCC,installing a uniform metric for creating comparabilitybetween SRM, CDR and conventional mitigation mea-sures, installing a time-limited moratorium on field test-ing and deployment of certain technologies, and thedefinition of terms for phasing out a climate engineeringintervention. If climate engineering is deployed in theabsence of such a multilateral institutional framework, itis likely that the problems associated with lackingsocial and political acceptance, moral hazard, slipperyslope, and premature termination fully manifest them-selves.

NotesWe would like to thank Peter Haas, Alan Robock, David Victor andOran Young for critical comments on an earlier version of thispiece. We also greatly benefitted from the collaboration with ourcolleagues involved in the drafting of the scoping report on climateengineering (Rickels et al., 2011), commissioned by the GermanFederal Ministry of Education and Research, and from the discussionof an earlier version of this piece in the colloquium of the researchgroup Transnational Conflicts and International Institutions at theSocial Science Research Center Berlin.

The Pardox of Climate Engineering9

Global Policy (2013) ª 2013 University of Durham and John Wiley & Sons, Ltd.

1. In addition, the Schelling thesis assumes that international lawdoes not formally stand in the way of such a solution. Legallybinding prohibitions on the use of climate engineering technol-ogies do indeed not exist in international law. International lawhowever does not explicitly allow the use of climate engineer-ing technologies either. We thus concur with Wiertz andReichwein (2010, p. 17) that whether and to what extent inter-national law applies to research on and uses of climate engi-neering depends on the political interpretation of existingtreaties (see also Virgo, 2009; Zedalis, 2010; Proelss and Güssow,2011).

2. We thus inherit the uncertainty of the respective assessments.While it has been shown by natural scientists that individual cli-mate engineering technologies are capable of manipulatingearth’s energy balance on a large scale, and economists haveshown that the direct costs for these technologies appear to liesignificantly below the costs of conventional emission reduc-tions, these results remain freighted with large uncertainty.

3. Via the criterion of effectiveness we operationalize what isreferred to in the natural scientific literature as ‘scalability’(Caldeira and Keith, 2010, p. 60).

4. In this typology, the efficiency criterion refers exclusively to thecapability of a climate engineering technology to significantlyinfluence global mean temperatures on a short timescale(effectiveness) at relatively low costs (affordability). Measureswhose efficiency is here listed as being ‘low’ can thus neverthe-less, on a medium to long timescale and in combination withemission reductions, be part of a portfolio system of policymeasures aimed at reducing the risks that result from risingglobal mean temperatures, and should not be disregarded.Lenton and Vaughan (2009, p. 5539) state that ‘[s]trong mitiga-tion, combined with global-scale air capture and storage,afforestation, and biochar production, i.e. enhanced CO2 sinks,might be able to bring CO2 back to its pre-industrial level by2100 […]’.

5. As mentioned before, the legal situation is starkly underdeter-mined and thus allows the use of international law for verydifferent positions.

6. The HOME campaign currently lists 107 organisations as ‘alliesand endorsers’ on its website (HOME, 2011c).

7. However, even resistance against the production and storage ofbiochar is justified by referring to the large area that is neces-sary for the use of this technology, thus indirectly conformingto our thesis: ‘The biggest danger of biochar for geoengineer-ing, however, is scale. Hundreds of millions of hectares of landlikely needs to be turned over to new plantations in order toproduce the quantities of biochar many talk about’ (HOME,2011d).

8. Biermann et al. (2010) and Victor (2011) together provide for avery good overview on the efforts and problems in creatingeffective climate governance.

9. While the inclusion of core elements of the existing interna-tional regime complex for climate change and the fact thatcertain aspects of our institutional proposal are already beingimplemented, such as the IPCC assessment of climate engineer-ing technologies in AR5, have the effect that the realization ofour proposal does not appear unfeasible, the issue of feasibilityis not further explored here. For a detailed exploration ofthe feasibility of this institutional proposal, see Schäfer (forth-coming).

10. We are aware that some commentators consider this to becounterproductive, arguing that the introduction of climateengineering into this existing institutional environment will

provide the death blow to international cooperation on climatepolicy. We do not believe this to be the case for the followingtwo reasons. First, the IPCC is capable of providing highly credi-ble assessments of state of the art research, thus substantiallyincreasing the perceived legitimacy of climate science. Subject-ing climate science, and by extension also research on climateengineering to IPCC assessment, can greatly enhance trust inthe results of such research and in policy changes and decisionson research project funding that are made on the basis ofthese results. Second, some critics consider the UNFCCC processtoo feeble to cope with the potentially conflicting results ofintroducing climate engineering to the ongoing negotiations.However, we consider the opposite to be at least as likely.Through the introduction of climate engineering into negotia-tions on climate policy, these discussions could be reinvigo-rated through the possibility of drawing up a portfolioapproach to climate change management that does not relyexclusively on voluntary emission reductions and the transfer offunds for adaptation. Climate engineering could thus depoliti-cize the current negotiations on climate change managementby providing an additional measure that is not (yet) stronglypoliticized itself.

11. For example Barrett (2008); Bodansky (1996); Carlin (2007);House of Commons (2010).

12. This would be another environmental assessment agency. Suchagencies have increased strongly in number during the past10–15 years (see Mitchell et al., 2006). Along the same line,Biermann and Siebenhüner (2009, p. 319) point out that inter-national secretariats can act as negotiation facilitators, capacitybuilders, and knowledge brokers.

13. Points of departure for developing such a regulatory code canbe found in the ‘Oxford Principles’, in SRMGI (2011), and inMorgan and Ricke (2010).

14. While the legitimacy of the IPCC is regularly questioned, thishas in the past always led to reform and consequently toincreases in the perceived legitimacy of the institution. Currentdiscussions need to be seen in this broader context (Beck,2010). Victor (2008) argues that the IPCC is not suited forassessing the results of research on climate engineering since itfocuses on ‘consensus science’, while the ‘improbable, harmfuland unexpected side effects’ are what counts in climate engi-neering. While Victor’s arguments for an alternative approachto climate engineering research assessment are convincing perse, we argue that the legitimacy gains an IPCC assessment ofclimate engineering entails and its expected catalytic effects forUNFCCC negotiations weigh up the tradeoffs. Apart from theseconsiderations, it has already been decided that AR5 will promi-nently address climate engineering research. The IPCC ‘ExpertMeeting on Geoengineering’ also suggested drafting a specialreport on the topic in the near future, very much in line withour suggestions.

15. In fact, most of the regulatory proposals in the literaturesupport a solution within the UN system (see e.g. Barrett, 2008;Virgoe, 2009; House of Commons, 2010). Bodansky (1996, pp.318ff) criticizes the low level of authority of existing institutionsand considers the creation of a new institutional settingdesirable.

16. A similar view is expressed by the The Royal Society (2009): ‘Aquestion for all CDR methods is whether they will be eligiblefor certification under the KP (or its successor instrument) underthe clean development mechanism or joint implementation’.

17. National afforestation measures can be offset against emissionreduction commitments within the framework of the Kyoto

Michael Zürn and Stefan Schäfer10

ª 2013 University of Durham and John Wiley & Sons, Ltd. Global Policy (2013)

Protocol. Should an international accreditation scheme emerge,this should also apply to afforestation measures being financedin other countries.

18. This distinguishes our proposal for a moratorium from Kra-emer’s (2010), which suggests a moratorium as a component ofa political strategy to prevent climate engineering altogether.

ReferencesAkbari, H., Menon, S. and Rosenfeld, A. (2009) ‘Global Cooling:

Increasing World Wide Urban Albedos to Offset CO2’, ClimaticChange, 95, pp. 275–286.

Barrett, S. (2008) ‘The Incredible Economics of Geoengineering‘,Environmental and Resource Economics, 39, pp. 45–54.

Beck, S. (2010) ‘Vertrauen geschmolzen? Zur Glaubwürdigkeit derKlimaforschung’, Aus Politik und Zeitgeschichte, 32–33, pp. 15–21.

Beck, U. (2008) World at Risk. Cambridge: Polity Press.Breitmeier, H., Young, O. R. and Zürn, M. (2006) Analyzing Interna-

tional Environmental Regimes: From Case Study to Database. Cam-bridge, MA: MIT Press.

Biermann, F. and Siebenhüner, B. (2009) (eds) Managers of GlobalChange: The Influence of International Environmental Change.Cambridge, MA: MIT Press.

Biermann, F., Pattberg, P. and Zelli, F. (eds) (2010) Global ClimateGovernance Beyond 2012. Architecture, Agency and Adaptation.Cambridge: Cambridge University Press.

Caldeira, K. and Keith, D. (2010) ‘The Need for GeoengineeringResearch‘, Issues in Science and Technology, Fall 2010, pp. 57–62.

Della Porta, D. and Tarrow, S. (eds) (2005) Transnational Protest andGlobal Activism. Lanham, MD: Rowman & Littlefield Publishers,Inc.

ETC Group (2009) ‘Geoengineering’s Governance Vacuum: Unilater-alism and the Future of the Planet’, prepared by ETC Group forthe US National Academies Workshop Geoengineering Options toRespond to Climate Change: Steps to Establish a Research Agenda,Washington, DC, June 15–16.

ETC Group (2010a) Geopiracy: The Case Against Geoengineering, ETCGroup Communiqué #103. Available from: http: ⁄ ⁄ www.etcgroup.org [Accessed 4 April 2012].

ETC Group (2010b) Geoengineering: Gambling with GAIA, briefingand recommendations for CBD delegates to COP 10. Availablefrom: http: ⁄ ⁄ www.etcgroup.org [Accessed 4 April 2012].

Haas, P., Keohane, R. O. and Levy, M. A. (eds) (1993) Institutions forthe Earth: Sources of Effective International Environmental Protec-tion. Cambridge, MA: MIT Press.

HOME (2011a) What are the aims of Hand Off Mother Earth cam-paign? Available from: http: ⁄ ⁄ www.handsoffmotherearth.org[Accessed 4 April 2012].

HOME (2011b) What is Geoengineering? Available from:http: ⁄ ⁄ www.handsoffmotherearth.org [Accessed 4 April 2012].

HOME (2011c) Organisations. Available from http: ⁄ ⁄ www.handsoff-motherearth.org [Accessed 4 April 2012].

HOME (2011d) Biochar. Available from http: ⁄ ⁄ www.handsoffmotherearth.org [Accessed 4 April 2012].

Horton, J. B. (2011) ‘Geoengineering and the Myth of Unilateralism:Pressures and Prospects for International Cooperation‘, StanfordJournal of Law, Science, and Policy. May Issue, pp. 56–69.

House of Commons (2010) The Regulation of Geoengineering. Lon-don: The Stationery Office Limited.

IPCC (2012) ‘Meeting Report of the Intergovernmental Panel on Cli-mate Change Expert Meeting on Geoengineering’, in O. Edenho-fer, R. Pichs-Madruga, Y. Sokona, C. Field, V. Barros and T.F.

Stockeret al. (eds). Potsdam: IPCC Working Group III TechnicalSupport Unit, Potsdam Institute for Climate Impact Research.

Keck, M. and Sikkink, K. (1998) Activists Beyond Borders: Advocacy Net-works in International Politics. Ithaca, NY: Cornell University Press.

Keith, D. W. (2000) ‘Geoengineering the Climate: History andProspect’, Annual Review of Energy and the Environment, 25, pp.245–284.

Keith, D. W. (2010a) ‘Engineering the Planet’, in S. H. Schneider, A.Rosencranz, M. D. Mastrandrea and K. Kuntz-Duriseti (eds) Cli-mate Change Science and Policy. Washington, DC: Island Press,pp. 494–501.

Keith, D. W. (2010b) ‘Photophoretic levitation of engineered aero-sols for geoengineering’, Proceedings of the National Academy ofSciences, 107 (38), pp. 16428–16431.

Keohane, R. O. (1984) After Hegemony: Cooperation and Discord inthe World Political Economy. Princeton, NJ: Princeton UniversityPress.

Keohane, R. O. and Victor, D. (2011) ‘The Regime Complex forClimate Change’, Perspectives on Politics, 9 (1), pp. 7–23.

Koremenos, B., Lipson, C. and Snidal, D. (2001) ‘The Rational Designof International Institutions’, International Organization, 55 (4),pp. 761–799.

Kraemer, R. A. (2010) ‘Schöner Leben im Labor? Geoengineeringund das Recht, die Welt zu verändern’, Internationale Politik, Jan-uar ⁄ Februar, pp. 70–75.

Lane, L. (2010) ‘Plan B: The last best hope for containing climatechange’, Milken Institute Review, 12 (3). Available from:http: ⁄ ⁄ www.aei.org [Accessed 4 April 2012].

Lenton, T. M., Held, H., Kriegler, E., Hall, J. W., Lucht, W., Rahmstorf,S., et al. (2008) ‘Tipping Elements in the Earth’s Climate System’,Proceedings of the National Academy of Sciences, 105 (6),pp. 1786–1793.

Lenton, T. M. and Vaughan, N. E. (2009) ‘The Radiative ForcingPotential of Different Climate Engineering Options’, AtmosphericChemistry and Physics, 9, pp. 5539–5561.

Miles, E. L., Underdal, A. and Andresen, S. (eds) (2001) EnvironmentalRegime Effectiveness: Confronting Theory with Evidence. Cam-bridge, MA: MIT Press.

Mitchell, R. B., Clark, W. C., Cash, D. W. and Dickson, N. M. (eds)(2006) Global Environmental Assessments: Information andInfluence. Cambridge, MA: MIT Press.

Morgan, M. G. and Ricke, K. (2010) Cooling the Earth through SolarRadiation Management: The Need for Research and an Approachto its Governance. Available from: http: ⁄ ⁄ www.irgc.com[Accessed 10 August 2012].

Oleson, K. W., Bonan, G. B. and Feddema, J. (2010) ‘Effects of whiteroofs on urban temperature in a global climate model’, Geophys-ical Research Letters, 37, pp. 1–7.

Ornstein, L., Aleinov, I. and Rind, D. (2009) ‘Irrigated afforestation ofthe Sahara and Australian Outback to end global warming’, Cli-matic Change, 97, pp. 409–437.

Ostrom, E. (1990) Governing the Commons: The Evolution of Institutionsfor Collective Action. Cambridge: Cambridge University Press.

Proelss, A. and Güssow, K. (2011) Relevante Institutionen und Instru-mente des internationalen Rechts. Trier: Scoping study commis-sioned by the German Federal Ministry for Education andResearch.

Rajagopal, B. (2003) International Law From Below: Development,Social Movements and Third World Resistance. Cambridge: Cam-bridge University Press.

Rayner, S. (2010) ‘The Geoengineering Paradox’, The GeoengineeringQuarterly, online newsletter, 20 March.

The Pardox of Climate Engineering11

Global Policy (2013) ª 2013 University of Durham and John Wiley & Sons, Ltd.

Rickels, W., Klepper, G., Dovern, J., Betz, G., Brachatzek, N.,Cacean, S., et al. (2011) Large-Scale Intentional Interventionsinto the Climate System? Assessing the Climate EngineeringDebate. Scoping report commissioned by the GermanFederal Ministry of Education and Research. Kiel: Kiel EarthInstitute.

Robock, A. (2008) ‘20 Reasons Why Geoengineering May Be a BadIdea’, Bulletin of the Atomic Scientists, 64 (2), pp. 14–18, 59.

Robock, A., Marquardt, A., Kravitz, B. and Stenchikov, G. (2009) ‘Ben-efits, Risks and Costs of Stratospheric Geoengineering’, Geophysi-cal Research Letters, 36, pp. 1–9.

The Royal Society (2009) Geoengineering the Climate: Science, Gover-nance and Uncertainty. RS Policy document 10 ⁄ 09, London: TheRoyal Society.

Schäfer, S. (forthcoming) ‘The International Politics of Engineeringthe Climate’, unpublished dissertation manuscript.

Schelling, T. C. (1960) The Strategy of Conflict. Cambridge, MA:Harvard University Press.

Schelling, T. C. (1996) ‘The Economic Diplomacy of Geoengineering’,Climatic Change, 33, pp. 303–307.

SRMGI (2011) Solar Radiation Management: The Governance ofResearch. Available from: http: ⁄ ⁄ www.srmgi.org [Accessed 10August 2012].

Tarrow, S. (2005) The New Transnational Activism. Cambridge: Cam-bridge University Press.

Victor, D. G. (2008) ‘On the Regulation of Geoengineering’, OxfordReview of Economic Policy, 24, 2, pp. 322–336.

Victor, D. G. (2011) Global Warming Gridlock: Creating More EffectiveStrategies for Protecting the Planet. Cambridge: CambridgeUniversity Press.

Virgoe, J. (2009) ‘International Governance of a Possible Geoengi-neering Intervention to Combat Climate Change’, ClimaticChange, 95, pp. 103–119.

Wiertz, T. and Reichwein, D. (2010) ‘Climate Engineering zwischenKlimapolitik und Völkerrecht‘, Technikfolgenabschätzung - Theorieund Praxis, 19 (2), pp. 17–25.

Young, O. R. (ed.) (1999) The Effectiveness of International Regimes: Cau-sal Connections and Behavioral Mechanisms. Cambridge: MIT Press.

Zedalis, R. (2010) ‘National Academy of Sciences’ Idea of Geoengi-neering’, European Energy and Environmental Law Review, 19 (1),pp. 18–32.

Zürn, M. (1992) Interessen und Institutionen in der internationalen Pol-itik: Grundlegung und Anwendung des situationsstrukturellen Ansa-tzes. Opladen: Leske + Budrich.

Zürn, M. (1998) Regieren jenseits des Nationalstaates. Frankfurt amMain: Suhrkamp.

Zürn, M., Binder, M. and Ecker-Ehrhardt, M. (2012) ‘International Author-ity and its Politicization’, International Theory, 4 (1), pp. 69–10.

Author InformationMichael Zürn is director at the WZB (Social Science Research Cen-ter Berlin) and Professor of International Relations at the Free Uni-versity Berlin.

Stefan Schäfer is a project scientist at the Institute for AdvancedSustainability Studies (IASS). At the time of writing, he was a PhDstudent at the Berlin Graduate School for Transnational Studies(BTS).

Michael Zürn and Stefan Schäfer12

ª 2013 University of Durham and John Wiley & Sons, Ltd. Global Policy (2013)